Datasets:

ChicagoHAI

/

CaseSumm

like 11

Follow

ChicagoHAI 19

562.US.86

In 1994, deputies called to drug dealer Johnson’s California home found Johnson wounded and Klein fatally wounded. Johnson claimed that he was shot in his bedroom by respondent Richter’s codefendant, Branscombe; that he found Klein on the living room couch; and that his gun safe, a pistol, and cash were missing. His account was corroborated by evidence at the scene, including, relevant here, spent shell casings, blood spatters, and blood pooled in the bedroom doorway. Investigators took a blood sample from a wall near the bedroom door, but not from the blood pool. A search of Richter’s home turned up the safe and ammunition matching evidence at the scene. After his arrest on murder and other charges, Richter initially denied his involvement, but later admitted disposing of Johnson’s and Branscombe’s guns. The prosecution initially built its case on Johnson’s testimony and the circumstantial evidence, but it adjusted its approach after Richter’s counsel, in his opening statement, outlined the theory that Branscombe shot Johnson in self-defense and that Klein was killed in the crossfire in the bedroom doorway, and stressed the lack of forensic support for the prosecution’s case. The prosecution then decided to call an expert in blood pattern evidence, who testified that it was unlikely that Klein had been shot outside the living room and then moved to the couch, and a serologist, who testified that the blood sample taken near the blood pool could be Johnson’s but not Klein’s. Under cross-examination, she conceded that she had not tested the sample for cross-contamination and that a degraded sample would make it difficult to tell if it had blood of Klein’s type. Defense counsel called Richter to tell his conflicting version of events and called other witnesses to corroborate Richter’s version. Richter was convicted and sentenced to life without parole. He later sought habeas relief from the California Supreme Court, asserting, inter alia, that his counsel provided ineffective assistance, see Strickland v. Washington, 466 U. S. 668, when he failed to present expert testimony on blood evidence, because it could have disclosed the blood pool’s source and bolstered Richter’s theory. He also offered affidavits from forensics experts to support his claim. The court denied the petition in a one-sentence summary order. Subsequently, he reasserted his state claims in a federal habeas petition. The District Court denied his petition. A Ninth Circuit panel affirmed, but the en banc court reversed. Initially it questioned whether 28 U. S. C. §2254(d)—which, as amended by the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), limits the availability of federal habeas relief for claims previously “adjudicated on the merits” in state court—applied to Richter’s petition, since the State Supreme Court issued only a summary denial. But it found the state-court decision unreasonable anyway. In its view, trial counsel was deficient in failing to consult blood evidence experts in planning a trial strategy and in preparing to rebut expert evidence the prosecution might—and later did—offer. Held: 1. Section 2254(d) applies to Richter’s petition, even though the state court’s order was unaccompanied by an opinion explaining the court’s reasoning. Pp. 7–10. (a) By its terms, §2254(d) bars relitigation of a claim “adjudicated on the merits” in state court unless, among other exceptions, the earlier state-court “decision” involved “an unreasonable application” of “clearly established Federal law, as determined by” this Court, §2254(d)(1). Nothing in its text—which refers only to a “decision” resulting “from an adjudication”—requires a statement of reasons. Where the state-court decision has no explanation, the habeas petitioner must still show there was no reasonable basis for the state court to deny relief. There is no merit to the assertion that applying §2254(d) when state courts issue summary rulings will encourage those courts to withhold explanations. The issuance of summary dispositions can enable state judiciaries to concentrate resources where most needed. Pp. 7–9. (b) Nor is there merit to Richter’s argument that §2254(d) does not apply because the California Supreme Court did not say it was adjudicating his claim “on the merits.” When a state court has denied relief, adjudication on the merits can be presumed absent any contrary indication or state-law procedural principles. The presumption may be overcome by a more likely explanation for the state court’s decision, but Richter does not make that showing here. Pp. 9–10. 2. Richter was not entitled to the habeas relief ordered by the Ninth Circuit. Pp. 10–24. (a) That court failed to accord the required deference to the decision of a state court adjudicating the same claims later presented in the federal habeas petition. Its opinion shows an improper understanding of §2254(d)’s unreasonableness standard and operation in the context of a Strickland claim. Asking whether the state court’s application of Strickland’s standard was unreasonable is different from asking whether defense counsel’s performance fell below that standard. Under AEDPA, a state court must be granted a deference and latitude that are not in operation in a case involving direct review under Strickland. A state court’s determination that a claim lacks merit precludes federal habeas relief so long as “fair-minded jurists could disagree” on the correctness of that decision. Yarborough v. Alvarado, 541 U. S. 652, 664. And the more general the rule being considered, “the more leeway courts have in reaching outcomes in case-by-case determinations.” Ibid. The Ninth Circuit explicitly conducted a de novo review and found a Strickland violation; it then declared without further explanation that the state court’s contrary decision was unreasonable. But §2254(d) requires a habeas court to determine what arguments or theories supported, or could have supported, the state-court decision; and then to ask whether it is possible fair-minded jurists could disagree that those arguments or theories are inconsistent with a prior decision of this Court. AEDPA’s unreasonableness standard is not a test of the confidence of a federal habeas court in the conclusion it would reach as a de novo matter. Even a strong case for relief does not make the state court’s contrary conclusion unreasonable. Section 2254(d) is designed to confirm that state courts are the principal forum for asserting constitutional challenges to state convictions. Pp. 10–14. (b) The Ninth Circuit erred in concluding that Richter demonstrated an unreasonable application of Strickland by the state court. Pp. 14–23. (1) Richter could have secured relief in state court only by showing both that his counsel provided deficient assistance and that prejudice resulted. To be deficient, counsel’s representation must have fallen “below an objective standard of reasonableness,” Strickland, 466 U. S., at 688; and there is a “strong presumption” that counsel’s representation is within the “wide range” of reasonable professional assistance, id., at 689. The question is whether counsel made errors so fundamental that counsel was not functioning as the counsel guaranteed by the Sixth Amendment. Prejudice requires demonstrating “a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Id., at 694. “Surmounting Strickland’s high bar is never … easy.” Padilla v. Kentucky, 559 U. S.___, ___. Strickland can function as a way to escape rules of waiver and forfeiture. The question is whether an attorney’s representation amounted to incompetence under prevailing professional norms, not whether it deviated from best practices or most common custom. Establishing that a state court’s application of Strickland was unreasonable under §2254(d) is even more difficult, since both standards are “highly deferential,” 466 U. S. at 689, and since Strickland’s general standard has a substantial range of reasonable applications. The question under §2254(d) is not whether counsel’s actions were reasonable, but whether there is any reasonable argument that counsel satisfied Strickland’s deferential standard. Pp. 14–16. (2) The Ninth Circuit erred in holding that because Richter’s attorney had not consulted forensic blood experts or introduced expert evidence, the State Supreme Court could not reasonably have concluded counsel provided adequate representation. A state court could reasonably conclude that a competent attorney could elect a strategy that did not require using blood evidence experts. Rare are the situations in which the latitude counsel enjoys will be limited to any one technique or approach. There were any number of experts whose insight might have been useful to the defense. Counsel is entitled to balance limited resources in accord with effective trial tactics and strategies. In finding otherwise the Ninth Circuit failed to “reconstruct the circumstances of counsel’s challenged conduct” and “evaluate the conduct from counsel’s perspective at the time.” Strickland, supra, at 689. Given the many factual differences between the prosecution and defense versions of events, it was far from evident at the time of trial that the blood source was central to Richter’s case. And relying on “the harsh light of hindsight” to cast doubt on a trial that took place over 15 years ago is precisely what Strickland and AEDPA seek to prevent. See Bell v. Cone, 535 U. S. 685, 702. Even had the value of expert testimony been apparent, it would be reasonable to conclude that a competent attorney might elect not to use it here, where counsel had reason to question the truth of his client’s account. Making blood evidence a central issue could also have led the prosecution to produce its own expert analysis, possibly destroying Richter’s case, or distracted the jury with esoteric questions of forensic science. Defense counsel’s opening statement may have inspired the prosecution to present forensic evidence, but that shows only that the defense strategy did not work out as well as hoped. In light of the record here there was no basis to rule that the state court’s determination was unreasonable. The Court of Appeals erred in dismissing such concern as an inaccurate account of counsel’s actual thinking, since Strickland examined only the objective reasonableness of counsel’s actions. As to whether counsel was constitutionally deficient for not preparing expert testimony as a response to the prosecution’s, an attorney may not be faulted for a reasonable miscalculation or lack of foresight or for failing to prepare for remote possibilities. Here, even if counsel was mistaken, the prosecution itself did not expect to present forensic testimony until the eve of trial. Thus, it is at least debatable whether counsel’s error was so fundamental as to call the trial’s fairness into doubt. Even if counsel should have foreseen the prosecution’s tactic, Richter would still need to show it was indisputable that Strickland required his attorney to rely on a rebuttal witness rather than on cross-examination to discredit the witnesses, but Strickland imposes no such requirement. And while it is possible an isolated error can constitute ineffective assistance if it is sufficiently egregious, it is difficult to establish ineffective assistance where counsel’s overall performance reflects active and capable advocacy. Pp. 16–22. (3) The Ninth Circuit also erred in concluding that Richter had established prejudice under Strickland, which asks whether it is “reasonably likely” the verdict would have been different, 466 U. S., at 696, not whether a court can be certain counsel’s performance had no effect on the outcome or that reasonable doubt might have been established had counsel acted differently. There must be a substantial likelihood of a different result. The State Supreme Court could have reasonably concluded that Richter’s prejudice evidence fell short of this standard. His expert serology evidence established only a theoretical possibility of Klein’s blood being in the blood pool; and at trial, defense counsel extracted a similar concession from the prosecution’s expert. It was also reasonable to find Richter had not established prejudice given that he offered no evidence challenging other conclusions of the prosecution’s experts, e.g., that the blood sample matched Johnson’s blood type. There was, furthermore, sufficient conventional circumstantial evidence pointing to Richter’s guilt, including, e.g., the items found at his home. Pp. 22–23. 578 F. 3d 944, reversed and remanded. Kennedy, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Thomas, Breyer, Alito, and Sotomayor, JJ., joined. Ginsburg, J., filed an opinion concurring in the judgment. Kagan, J., took no part in the consideration or decision of the case.

The writ of habeas corpus stands as a safeguard against imprisonment of those held in violation of the law. Judges must be vigilant and independent in reviewing petitions for the writ, a commitment that entails substantial judicial resources. Those resources are diminished and misspent, however, and confidence in the writ and the law it vindicates undermined, if there is judicial disregard for the sound and established principles that inform its proper issuance. That judicial disregard is inherent in the opinion of the Court of Appeals for the Ninth Circuit here under review. The Court of Appeals, in disagreement with the contrary conclusions of the Supreme Court of the State of California and of a United States District Court, ordered habeas corpus relief granted to set aside the conviction of Joshua Richter, respondent here. This was clear error. Under 28 U. S. C. §2254(d), the availability of federal habeas relief is limited with respect to claims previously “adjudicated on the merits” in state-court proceedings. The first inquiry this case presents is whether that pro-vision applies when state-court relief is denied without an accompanying statement of reasons. If it does, the question is whether the Court of Appeals adhered to the statute’s terms, in this case as it relates to ineffective-assistance claims judged by the standard set forth in Strickland v. Washington, 466 U. S. 668 (1984). A second case decided today, Premo v. Moore, post, p. ___, presents similar issues. Here, as in that case, it is necessary to reverse the Court of Appeals for failing to accord required deference to the decision of a state court. I It is necessary to begin by discussing the details of a crime committed more than a decade and a half ago. A Sometime after midnight on December 20, 1994, sheriff’s deputies in Sacramento County, California, arrived at the home of a drug dealer named Joshua Johnson. Hours before, Johnson had been smoking marijuana in the company of Richter and two other men, Christian Branscombe and Patrick Klein. When the deputies arrived, however, they found only Johnson and Klein. Johnson was hysterical and covered in blood. Klein was lying on a couch in Johnson’s living room, unconscious and bleeding. Klein and Johnson each had been shot twice. Johnson recovered; Klein died of his wounds. Johnson gave investigators this account: After falling asleep, he awoke to find Richter and Branscombe in his bedroom, at which point Branscombe shot him. Johnson heard more gunfire in the living room and the sound of his assailants leaving. He got up, found Klein bleeding on the living room couch, and called 911. A gun safe, a pistol, and $6,000 cash, all of which had been in the bedroom, were missing. Evidence at the scene corroborated Johnson’s account. Investigators found spent shell casings in the bedroom (where Johnson said he had been shot) and in the living room (where Johnson indicated Klein had been shot). In the living room there were two casings, a .32 caliber and a .22 caliber. One of the bullets recovered from Klein’s body was a .32 and the other was a .22. In the bedroom there were two more casings, both .32 caliber. In addition detectives found blood spatter near the living room couch and bloodstains in the bedroom. Pools of blood had collected in the kitchen and the doorway to Johnson’s bedroom. Investigators took only a few blood samples from the crime scene. One was from a blood splash on the wall near the bedroom doorway, but no sample was taken from the doorway blood pool itself. Investigators searched Richter’s residence and found Johnson’s gun safe, two boxes of .22-caliber ammunition, and a gun magazine loaded with cartridges of the same brand and type as the boxes. A ballistics expert later concluded the .22-caliber bullet that struck Klein and the .22-caliber shell found in the living room matched the ammunition found in Richter’s home and bore markings consistent with the model of gun for which the magazine was designed. Richter and Branscombe were arrested. At first Richter denied involvement. He would later admit taking Johnson’s pistol and disposing of it and of the .32-caliber weapon Branscombe used to shoot Johnson and Klein. Richter’s counsel produced Johnson’s missing pistol, but neither of the guns used to shoot Johnson and Klein was found. B Branscombe and Richter were tried together on charges of murder, attempted murder, burglary, and robbery. Only Richter’s case is presented here. The prosecution built its case on Johnson’s testimony and on circumstantial evidence. Its opening statement took note of the shell casings found at the crime scene and the ammunition and gun safe found at Richter’s residence. Defense counsel offered explanations for the circumstantial evidence and derided Johnson as a drug dealer, a paranoid, and a trigger-happy gun fanatic who had drawn a pistol on Branscombe and Richter the last time he had seen them. And there were inconsistencies in Johnson’s story. In his 911 call, for instance, Johnson first said there were four or five men who had broken into his house, not two; and in the call he did not identify Richter and Branscombe among the intruders. Blood evidence does not appear to have been part of the prosecution’s planned case prior to trial, and investigators had not analyzed the few blood samples taken from the crime scene. But the opening statement from the defense led the prosecution to alter its approach. Richter’s attorney outlined the theory that Branscombe had fired on Johnson in self-defense and that Klein had been killed not on the living room couch but in the crossfire in the bedroom doorway. Defense counsel stressed deficiencies in the investigation, including the absence of forensic support for the prosecution’s version of events. The prosecution took steps to adjust to the counterattack now disclosed. Without advance notice and over the objection of Richter’s attorney, one of the detectives who investigated the shootings testified for the prosecution as an expert in blood pattern evidence. He concluded it was unlikely Klein had been shot outside the living room and then moved to the couch, given the patterns of blood on Klein’s face, as well as other evidence including “high velocity” blood spatter near the couch consistent with the location of a shooting. The prosecution also offered testimony from a serologist. She testified the blood sample taken near the pool by the bedroom door could be Johnson’s but not Klein’s. Defense counsel’s cross-examination probed weaknesses in the testimony of these two witnesses. The detective who testified on blood patterns acknowledged that his inferences were imprecise, that it was unlikely Klein had been lying down on the couch when shot, and that he could not say the blood in the living room was from either of Klein’s wounds. Defense counsel elicited from the serologist a concession that she had not tested the bedroom blood sample for cross-contamination. She said that if the year-old sample had degraded, it would be difficult to tell whether blood of Klein’s type was also present in the sample. For the defense, Richter’s attorney called seven witnesses. Prominent among these was Richter himself. Richter testified he and Branscombe returned to Johnson’s house just before the shootings in order to deliver something to one of Johnson’s roommates. By Richter’s account, Branscombe entered the house alone while Richter waited in the driveway; but after hearing screams and gunshots, Richter followed inside. There he saw Klein lying not on the couch but in the bedroom doorway, with Johnson on the bed and Branscombe standing in the middle of the room. According to Richter, Branscombe said he shot at Johnson and Klein after they attacked him. Other defense witnesses provided some corroboration for Richter’s story. His former girlfriend, for instance, said she saw the gun safe at Richter’s house shortly before the shootings. The jury returned a verdict of guilty on all charges. Richter was sentenced to life without parole. On appeal, his conviction was affirmed. People v. Branscombe, 72 Cal. Rptr. 2d 773 (Cal. App. 1998) (officially depublished). The California Supreme Court denied a petition for review, People v. Branscombe, No. S069751, 1998 Cal. LEXIS 4252 (June 24, 1998), and Richter did not file a petition for certiorari with this Court. His conviction became final. C Richter later petitioned the California Supreme Court for a writ of habeas corpus. He asserted a number of grounds for relief, including ineffective assistance of counsel. As relevant here, he claimed his counsel was deficient for failing to present expert testimony on serology, pathology, and blood spatter patterns, testimony that, he argued, would disclose the source of the blood pool in the bedroom doorway. This, he contended, would bolster his theory that Johnson had moved Klein to the couch. He offered affidavits from three types of forensic experts. First, he provided statements from two blood serologists who said there was a possibility Klein’s blood was intermixed with blood of Johnson’s type in the sample taken from near the pool in the bedroom doorway. Second, he provided a statement from a pathologist who said the blood pool was too large to have come from Johnson given the nature of his wounds and his own account of his actions while waiting for the police. Third, he provided a statement from an expert in bloodstain analysis who said the absence of “a large number of satellite droplets” in photographs of the area around the blood in the bedroom doorway was inconsistent with the blood pool coming from Johnson as he stood in the doorway. App. 118. Richter argued this evidence established the possibility that the blood in the bedroom doorway came from Klein, not Johnson. If that were true, he argued, it would confirm his account, not Johnson’s. The California Supreme Court denied Richter’s petition in a one-sentence summary order. See In re Richter, No. S082167 (Mar. 28, 2001), App. to Pet. for Cert. 22a. Richter did not seek certiorari from this Court. After the California Supreme Court issued its summary order denying relief, Richter filed a petition for habeas corpus in United States District Court for the Eastern District of California. He reasserted the claims in his state petition. The District Court denied his petition, and a three-judge panel of the Court of Appeals for the Ninth Circuit affirmed. See Richter v. Hickman, 521 F. 3d 1222 (2008). The Court of Appeals granted rehearing en banc and reversed the District Court’s decision. See Richter v. Hickman, 578 F. 3d 944 (2009). As a preliminary matter, the Court of Appeals questioned whether 28 U. S. C. §2254(d) was applicable to Richter’s petition, since the California Supreme Court issued only a summary denial when it rejected his Strickland claims; but it determined the California decision was unreasonable in any event and that Richter was entitled to relief. The court held Richter’s trial counsel was deficient for failing to consult experts on blood evidence in determining and pursuing a trial strategy and in preparing to rebut expert evidence the prosecution might—and later did—offer. Four judges dissented from the en banc decision. We granted certiorari. 559 U. S. ___ (2010). II The statutory authority of federal courts to issue habeas corpus relief for persons in state custody is provided by 28 U. S. C. §2254, as amended by the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA). The text of §2254(d) states: “An application for a writ of habeas corpus on behalf of a person in custody pursuant to the judgment of a State court shall not be granted with respect to any claim that was adjudicated on the merits in State court proceedings unless the adjudication of the claim— “(1) resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States; or “(2) resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” As an initial matter, it is necessary to decide whether §2254(d) applies when a state court’s order is unaccompanied by an opinion explaining the reasons relief has been denied. By its terms §2254(d) bars relitigation of any claim “adjudicated on the merits” in state court, subject only to the exceptions in §§2254(d)(1) and (d)(2). There is no text in the statute requiring a statement of reasons. The statute refers only to a “decision,” which resulted from an “adjudication.” As every Court of Appeals to consider the issue has recognized, determining whether a state court’s decision resulted from an unreasonable legal or factual conclusion does not require that there be an opinion from the state court explaining the state court’s reasoning. See Chadwick v. Janecka, 312 F. 3d 597, 605–606 (CA3 2002); Wright v. Secretary for Dept. of Corrections, 278 F. 3d 1245, 1253–1254 (CA11 2002); Sellan v. Kuhlman, 261 F. 3d 303, 311–312 (CA2 2001); Bell v. Jarvis, 236 F. 3d 149, 158–162 (CA4 2000) (en banc); Harris v. Stovall, 212 F. 3d 940, 943, n. 1 (CA6 2000); Aycox v. Lytle, 196 F. 3d 1174, 1177–1178 (CA10 1999); James v. Bowersox, 187 F. 3d 866, 869 (CA8 1999). And as this Court has observed, a state court need not cite or even be aware of our cases under §2254(d). Early v. Packer, 537 U. S. 3, 8 (2002) (per curiam). Where a state court’s decision is unaccompanied by an explanation, the habeas petitioner’s burden still must be met by showing there was no reasonable basis for the state court to deny relief. This is so whether or not the state court reveals which of the elements in a multipart claim it found insufficient, for §2254(d) applies when a “claim,” not a component of one, has been adjudicated. There is no merit to the assertion that compliance with §2254(d) should be excused when state courts issue summary rulings because applying §2254(d) in those cases will encourage state courts to withhold explanations for their decisions. Opinion-writing practices in state courts are influenced by considerations other than avoiding scrutiny by collateral attack in federal court. Cf. In re Robbins, 18 Cal. 4th 770, 778, n. 1, 959 P. 2d 311, 316, n. 1 (1998) (state procedures limiting habeas are “a means of protecting the integrity of our own appeal and habeas corpus process,” rather than a device for “insulating our judgments from federal court review” (emphasis deleted)). At the same time, requiring a statement of reasons could undercut state practices designed to preserve the integrity of the case-law tradition. The issuance of summary dispositions in many collateral attack cases can enable a state judiciary to concentrate its resources on the cases where opinions are most needed. See Brief for California Attorneys for Criminal Justice et al. as Amici Curiae 8 (noting that the California Supreme Court disposes of close to 10,000 cases a year, including more than 3,400 original habeas corpus petitions). There is no merit either in Richter’s argument that §2254(d) is inapplicable because the California Supreme Court did not say it was adjudicating his claim “on the merits.” The state court did not say it was denying the claim for any other reason. When a federal claim has been presented to a state court and the state court has denied relief, it may be presumed that the state court adjudicated the claim on the merits in the absence of any indication or state-law procedural principles to the contrary. Cf. Harris v. Reed, 489 U. S. 255, 265 (1989) (presumption of a merits determination when it is unclear whether a decision appearing to rest on federal grounds was decided on another basis). The presumption may be overcome when there is reason to think some other explanation for the state court’s decision is more likely. See, e.g., Ylst v. Nunnemaker, 501 U. S. 797, 803 (1991). Richter, however, does not make that showing. He mentions the theoretical possibility that the members of the California Supreme Court may not have agreed on the reasons for denying his petition. It is pure speculation, however, to suppose that happened in this case. And Richter’s assertion that the mere possibility of a lack of agreement prevents any attribution of reasons to the state court’s decision is foreclosed by precedent. See ibid. As has been noted before, the California courts or Legislature can alter the State’s practices or elaborate more fully on their import. See Evans v. Chavis, 546 U. S. 189, 197, 199 (2006). But that has not occurred here. This Court now holds and reconfirms that §2254(d) does not require a state court to give reasons before its decision can be deemed to have been “adjudicated on the merits.” Richter has failed to show that the California Supreme Court’s decision did not involve a determination of the merits of his claim. Section 2254(d) applies to his petition. III Federal habeas relief may not be granted for claims subject to §2254(d) unless it is shown that the earlier state court’s decision “was contrary to” federal law then clearly established in the holdings of this Court, §2254(d)(1); Williams v. Taylor, 529 U. S. 362, 412 (2000); or that it “involved an unreasonable application of” such law, §2254(d)(1); or that it “was based on an unreasonable determination of the facts” in light of the record before the state court, §2254(d)(2). The Court of Appeals relied on the second of these exceptions to §2254(d)’s relitigation bar, the exception in §2254(d)(1) permitting relitigation where the earlier state decision resulted from an “unreasonable application of” clearly established federal law. In the view of the Court of Appeals, the California Supreme Court’s decision on Richter’s ineffective-assistance claim unreasonably applied the holding in Strickland. The Court of Appeals’ lengthy opinion, however, discloses an improper understanding of §2254(d)’s unreasonableness standard and of its operation in the context of a Strickland claim. The pivotal question is whether the state court’s application of the Strickland standard was unreasonable. This is different from asking whether defense counsel’s performance fell below Strickland’s standard. Were that the inquiry, the analysis would be no different than if, for example, this Court were adjudicating a Strickland claim on direct review of a criminal conviction in a United States district court. Under AEDPA, though, it is a necessary premise that the two questions are different. For purposes of §2254(d)(1), “an unreasonable application of federal law is different from an incorrect application of federal law.” Williams, supra, at 410. A state court must be granted a deference and latitude that are not in operation when the case involves review under the Strickland standard itself. A state court’s determination that a claim lacks merit precludes federal habeas relief so long as “fairminded jurists could disagree” on the correctness of the state court’s decision. Yarborough v. Alvarado, 541 U. S. 652, 664 (2004). And as this Court has explained, “[E]valuating whether a rule application was unreasonable requires considering the rule’s specificity. The more general the rule, the more leeway courts have in reaching outcomes in case-by-case determinations.” Ibid. “[I]t is not an unreasonable application of clearly established Federal law for a state court to decline to apply a specific legal rule that has not been squarely established by this Court.” Knowles v. Mirzayance, 556 U. S. ___, ___ (2009) (slip op., at 9–10) (internal quotation marks omitted). Here it is not apparent how the Court of Appeals’ analysis would have been any different without AEDPA. The court explicitly conducted a de novo review, 578 F. 3d, at 952; and after finding a Strickland violation, it declared, without further explanation, that the “state court’s decision to the contrary constituted an unreasonable application of Strickland.” 578 F. 3d, at 969. AEDPA demands more. Under §2254(d), a habeas court must determine what arguments or theories supported or, as here, could have supported, the state court’s decision; and then it must ask whether it is possible fairminded jurists could disagree that those arguments or theories are inconsistent with the holding in a prior decision of this Court. The opinion of the Court of Appeals all but ignored “the only question that matters under §2254(d)(1).” Lockyer v. Andrade, 538 U. S. 63, 71 (2003). The Court of Appeals appears to have treated the unreasonableness question as a test of its confidence in the result it would reach under de novo review: Because the Court of Appeals had little doubt that Richter’s Strickland claim had merit, the Court of Appeals concluded the state court must have been unreasonable in rejecting it. This analysis overlooks arguments that would otherwise justify the state court’s result and ignores further limitations of §2254(d), including its requirement that the state court’s decision be evaluated according to the precedents of this Court. See Renico v. Lett, 559 U. S. ___, ___ (2010) (slip op., at 11–12). It bears repeating that even a strong case for relief does not mean the state court’s contrary conclusion was unreasonable. See Lockyer, supra, at 75. If this standard is difficult to meet, that is because it was meant to be. As amended by AEDPA, §2254(d) stops short of imposing a complete bar on federal court relitigation of claims already rejected in state proceedings. Cf. Felker v. Turpin, 518 U. S. 651, 664 (1996) (discussing AEDPA’s “modified res judicata rule” under §2244). It preserves authority to issue the writ in cases where there is no possibility fairminded jurists could disagree that the state court’s decision conflicts with this Court’s precedents. It goes no farther. Section 2254(d) reflects the view that habeas corpus is a “guard against extreme malfunctions in the state criminal justice systems,” not a substitute for ordinary error correction through appeal. Jackson v. Virginia, 443 U. S. 307, 332, n. 5 (1979) (Stevens, J., concurring in judgment). As a condition for obtaining habeas corpus from a federal court, a state prisoner must show that the state court’s ruling on the claim being presented in federal court was so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement. The reasons for this approach are familiar. “Federal habeas review of state convictions frustrates both the States’ sovereign power to punish offenders and their good-faith attempts to honor constitutional rights.” Calderon v. Thompson, 523 U. S. 538, 555–556 (1998) (internal quotation marks omitted). It “disturbs the State’s significant interest in repose for concluded litigation, denies society the right to punish some admitted offenders, and intrudes on state sovereignty to a degree matched by few exercises of federal judicial authority.” Reed, 489 U. S., at 282 (Kennedy, J., dissenting). Section 2254(d) is part of the basic structure of federal habeas jurisdiction, designed to confirm that state courts are the principal forum for asserting constitutional challenges to state convictions. Under the exhaustion requirement, a habeas petitioner challenging a state conviction must first attempt to present his claim in state court. 28 U. S. C. §2254(b). If the state court rejects the claim on procedural grounds, the claim is barred in federal court unless one of the exceptions to the doctrine of Wainwright v. Sykes, 433 U. S. 72, 82–84 (1977), applies. And if the state court denies the claim on the merits, the claim is barred in federal court unless one of the exceptions to §2254(d) set out in §§2254(d)(1) and (2) applies. Section 2254(d) thus complements the exhaustion requirement and the doctrine of procedural bar to ensure that state proceedings are the central process, not just a preliminary step for a later federal habeas proceeding, see id., at 90. Here, however, the Court of Appeals gave §2254(d) no operation or function in its reasoning. Its analysis illustrates a lack of deference to the state court’s determination and an improper intervention in state criminal processes, contrary to the purpose and mandate of AEDPA and to the now well-settled meaning and function of habeas corpus in the federal system. IV The conclusion of the Court of Appeals that Richter demonstrated an unreasonable application by the state court of the Strickland standard now must be discussed. To have been entitled to relief from the California Supreme Court, Richter had to show both that his counsel provided deficient assistance and that there was prejudice as a result. To establish deficient performance, a person challenging a conviction must show that “counsel’s representation fell below an objective standard of reasonableness.” 466 U. S., at 688. A court considering a claim of ineffective assistance must apply a “strong presumption” that counsel’s representation was within the “wide range” of reasonable professional assistance. Id., at 689. The challenger’s burden is to show “that counsel made errors so serious that counsel was not functioning as the ‘counsel’ guaranteed the defendant by the Sixth Amendment.” Id., at 687. With respect to prejudice, a challenger must demonstrate “a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.” Id., at 694. It is not enough “to show that the errors had some conceivable effect on the outcome of the proceeding.” Id., at 693. Counsel’s errors must be “so serious as to deprive the defendant of a fair trial, a trial whose result is reliable.” Id., at 687. “Surmounting Strickland’s high bar is never an easy task.” Padilla v. Kentucky, 559 U. S. ___, ___ (2010) (slip op., at 14). An ineffective-assistance claim can function as a way to escape rules of waiver and forfeiture and raise issues not presented at trial, and so the Strickland standard must be applied with scrupulous care, lest “intrusive post-trial inquiry” threaten the integrity of the very adversary process the right to counsel is meant to serve. Strickland, 466 U. S., at 689–690. Even under de novo review, the standard for judging counsel’s representation is a most deferential one. Unlike a later reviewing court, the attorney observed the relevant proceedings, knew of materials outside the record, and interacted with the client, with opposing counsel, and with the judge. It is “all too tempting” to “second-guess counsel’s assistance after conviction or adverse sentence.” Id., at 689; see also Bell v. Cone, 535 U. S. 685, 702 (2002); Lockhart v. Fretwell, 506 U. S. 364, 372 (1993). The question is whether an attorney’s representation amounted to incompetence under “prevailing professional norms,” not whether it deviated from best practices or most common custom. Strickland, 466 U. S., at 690. Establishing that a state court’s application of Strickland was unreasonable under §2254(d) is all the more difficult. The standards created by Strickland and §2254(d) are both “highly deferential,” id., at 689; Lindh v. Murphy, 521 U. S. 320, 333, n. 7 (1997), and when the two apply in tandem, review is “doubly” so, Knowles, 556 U. S., at ___ (slip op., at 11). The Strickland standard is a general one, so the range of reasonable applications is substantial. 556 U. S., at ___ (slip op., at 11). Federal habeas courts must guard against the danger of equating unreasonableness under Strickland with unreasonableness under §2254(d). When §2254(d) applies, the question is not whether counsel’s actions were reasonable. The question is whether there is any reasonable argument that counsel satisfied Strickland’s deferential standard. A With respect to defense counsel’s performance, the Court of Appeals held that because Richter’s attorney had not consulted forensic blood experts or introduced expert evidence, the California Supreme Court could not reasonably have concluded counsel provided adequate representation. This conclusion was erroneous. 1 The Court of Appeals first held that Richter’s attorney rendered constitutionally deficient service because he did not consult blood evidence experts in developing the basic strategy for Richter’s defense or offer their testimony as part of the principal case for the defense. Strickland, however, permits counsel to “make a reasonable decision that makes particular investigations unnecessary.” 466 U. S., at 691. It was at least arguable that a reasonable attorney could decide to forgo inquiry into the blood evidence in the circumstances here. Criminal cases will arise where the only reasonable and available defense strategy requires consultation with experts or introduction of expert evidence, whether pretrial, at trial, or both. There are, however, “countless ways to provide effective assistance in any given case. Even the best criminal defense attorneys would not defend a particular client in the same way.” Id., at 689. Rare are the situations in which the “wide latitude counsel must have in making tactical decisions” will be limited to any one technique or approach. Ibid. It can be assumed that in some cases counsel would be deemed ineffective for failing to consult or rely on experts, but even that formulation is sufficiently general that state courts would have wide latitude in applying it. Here it would be well within the bounds of a reasonable judicial determination for the state court to conclude that defense counsel could follow a strategy that did not require the use of experts regarding the pool in the doorway to Johnson’s bedroom. From the perspective of Richter’s defense counsel when he was preparing Richter’s defense, there were any number of hypothetical experts—specialists in psychiatry, psychology, ballistics, fingerprints, tire treads, physiology, or numerous other disciplines and subdisciplines—whose insight might possibly have been useful. An attorney can avoid activities that appear “distractive from more important duties.” Bobby v. Van Hook, 558 U. S. ___, ___ (2009) (per curiam) (slip op., at 8). Counsel was entitled to formulate a strategy that was reasonable at the time and to balance limited resources in accord with effective trial tactics and strategies. See Knowles, supra, at ___ (slip op., at 14–15); Rompilla v. Beard, 545 U. S. 374, 383 (2005); Wiggins v. Smith, 539 U. S. 510, 525 (2003); Strickland, 466 U. S., at 699. In concluding otherwise the Court of Appeals failed to “reconstruct the circumstances of counsel’s challenged conduct” and “evaluate the conduct from counsel’s perspective at the time.” Id., at 689. In its view Klein’s location was “the single most critical issue in the case” given the differing theories of the prosecution and the defense, and the source of the blood in the doorway was therefore of central concern. 578 F. 3d, at 953–954. But it was far from a necessary conclusion that this was evident at the time of the trial. There were many factual differences between prosecution and defense versions of the events on the night of the shootings. It is only because forensic evidence has emerged concerning the source of the blood pool that the issue could with any plausibility be said to stand apart. Reliance on “the harsh light of hindsight” to cast doubt on a trial that took place now more than 15 years ago is precisely what Strickland and AEDPA seek to prevent. Cone, 535 U. S., at 702; see also Lockhart, 506 U. S., at 372. Even if it had been apparent that expert blood testimony could support Richter’s defense, it would be reasonable to conclude that a competent attorney might elect not to use it. The Court of Appeals opinion for the en banc majority rests in large part on a hypothesis that reasonably could have been rejected. The hypothesis is that without jeopardizing Richter’s defense, an expert could have testified that the blood in Johnson’s doorway could not have come from Johnson and could have come from Klein, thus suggesting that Richter’s version of the shooting was correct and Johnson’s a fabrication. This theory overlooks the fact that concentrating on the blood pool carried its own serious risks. If serological analysis or other forensic evidence demonstrated that the blood came from Johnson alone, Richter’s story would be exposed as an invention. An attorney need not pursue an investigation that would be fruitless, much less one that might be harmful to the defense. Strickland, supra, at 691. Here Richter’s attorney had reason to question the truth of his client’s account, given, for instance, Richter’s initial denial of involvement and the subsequent production of Johnson’s missing pistol. It would have been altogether reasonable to conclude that this concern justified the course Richter’s counsel pursued. Indeed, the Court of Appeals recognized this risk insofar as it pertained to the suggestion that counsel should have had the blood evidence tested. 578 F. 3d, at 956, n. 9. But the court failed to recognize that making a central issue out of blood evidence would have increased the likelihood of the prosecution’s producing its own evidence on the blood pool’s origins and composition; and once matters proceeded on this course, there was a serious risk that expert evidence could destroy Richter’s case. Even apart from this danger, there was the possibility that expert testimony could shift attention to esoteric matters of forensic science, distract the jury from whether Johnson was telling the truth, or transform the case into a battle of the experts. Accord, Bonin v. Calderon, 59 F. 3d 815, 836 (CA9 1995). True, it appears that defense counsel’s opening statement itself inspired the prosecution to introduce expert forensic evidence. But the prosecution’s evidence may well have been weakened by the fact that it was assembled late in the process; and in any event the prosecution’s response shows merely that the defense strategy did not work out as well as counsel had hoped, not that counsel was incompetent. To support a defense argument that the prosecution has not proved its case it sometimes is better to try to cast pervasive suspicion of doubt than to strive to prove a certainty that exonerates. All that happened here is that counsel pursued a course that conformed to the first option. If this case presented a de novo review of Strickland, the foregoing might well suffice to reject the claim of inadequate counsel, but that is an unnecessary step. The Court of Appeals must be reversed if there was a reasonable justification for the state court’s decision. In light of the record here there was no basis to rule that the state court’s determination was unreasonable. The Court of Appeals erred in dismissing strategic considerations like these as an inaccurate account of counsel’s actual thinking. Although courts may not indulge “post hoc rationalization” for counsel’s decisionmaking that contradicts the available evidence of counsel’s actions, Wiggins, 539 U. S., at 526–527, neither may they insist counsel confirm every aspect of the strategic basis for his or her actions. There is a “strong presumption” that counsel’s attention to certain issues to the exclusion of others reflects trial tactics rather than “sheer neglect.” Yarborough v. Gentry, 540 U. S. 1, 8 (2003) (per curiam). After an adverse verdict at trial even the most experienced counsel may find it difficult to resist asking whether a different strategy might have been better, and, in the course of that reflection, to magnify their own responsibility for an unfavorable outcome. Strickland, however, calls for an inquiry into the objective reasonableness of counsel’s performance, not counsel’s subjective state of mind. 466 U. S., at 688. 2 The Court of Appeals also found that Richter’s attorney was constitutionally deficient because he had not expected the prosecution to offer expert testimony and therefore was unable to offer expert testimony of his own in response. The Court of Appeals erred in suggesting counsel had to be prepared for “any contingency,” 578 F. 3d, at 946 (internal quotation marks omitted). Strickland does not guarantee perfect representation, only a “ ‘reasonably competent attorney.’ ” 466 U. S., at 687 (quoting McMann v. Richardson, 397 U. S. 759, 770 (1970)); see also Gentry, supra, at 7. Representation is constitutionally ineffective only if it “so undermined the proper functioning of the adversarial process” that the defendant was denied a fair trial. Strickland, supra, at 686. Just as there is no expectation that competent counsel will be a flawless strategist or tactician, an attorney may not be faulted for a reasonable miscalculation or lack of foresight or for failing to prepare for what appear to be remote possibilities. Here, Richter’s attorney was mistaken in thinking the prosecution would not present forensic testimony. But the prosecution itself did not expect to make that presenta- tion and had made no preparations for doing so on the eve of trial. For this reason alone, it is at least debatable whether counsel’s error was so fundamental as to call the fairness of the trial into doubt. Even if counsel should have foreseen that the prosecution would offer expert evidence, Richter would still need to show it was indisputable that Strickland required his attorney to act upon that knowledge. Attempting to establish this, the Court of Appeals held that defense counsel should have offered expert testimony to rebut the evidence from the prosecution. But Strickland does not enact Newton’s third law for the presentation of evidence, requiring for every prosecution expert an equal and opposite expert from the defense. In many instances cross-examination will be sufficient to expose defects in an expert’s presentation. When defense counsel does not have a solid case, the best strategy can be to say that there is too much doubt about the State’s theory for a jury to convict. And while in some instances “even an isolated error” can support an ineffective-assistance claim if it is “sufficiently egregious and prejudicial,” Murray v. Carrier, 477 U. S. 478, 496 (1986), it is difficult to establish ineffective assistance when counsel’s overall performance indicates active and capable advocacy. Here Richter’s attorney represented him with vigor and conducted a skillful cross-examination. As noted, defense counsel elicited concessions from the State’s experts and was able to draw attention to weaknesses in their conclusions stemming from the fact that their analyses were conducted long after investigators had left the crime scene. For all of these reasons, it would have been reasonable to find that Richter had not shown his attorney was deficient under Strickland. B The Court of Appeals further concluded that Richter had established prejudice under Strickland given the expert evidence his attorney could have introduced. It held that the California Supreme Court would have been unreasonable in concluding otherwise. This too was error. In assessing prejudice under Strickland, the question is not whether a court can be certain counsel’s performance had no effect on the outcome or whether it is possible a reasonable doubt might have been established if counsel acted differently. See Wong v. Belmontes, 558 U. S. ___, ___ (2009) (per curiam) (slip op., at 13); Strickland, 466 U. S., at 693. Instead, Strickland asks whether it is “reasonably likely” the result would have been different. Id., at 696. This does not require a showing that counsel’s actions “more likely than not altered the outcome,” but the difference between Strickland’s prejudice standard and a more-probable-than-not standard is slight and matters “only in the rarest case.” Id., at 693, 697. The likelihood of a different result must be substantial, not just conceivable. Id., at 693. It would not have been unreasonable for the California Supreme Court to conclude Richter’s evidence of prejudice fell short of this standard. His expert serology evidence established nothing more than a theoretical possibility that, in addition to blood of Johnson’s type, Klein’s blood may also have been present in a blood sample taken near the bedroom doorway pool. At trial, defense counsel extracted a concession along these lines from the prosecution’s expert. The pathology expert’s claim about the size of the blood pool could be taken to suggest only that the wounded and hysterical Johnson erred in his assessment of time or that he bled more profusely than estimated. And the analysis of the purported blood pattern expert indicated no more than that Johnson was not standing up when the blood pool formed. It was also reasonable to find Richter had not established prejudice given that he offered no evidence directly challenging other conclusions reached by the prosecution’s experts. For example, there was no dispute that the blood sample taken near the doorway pool matched Johnson’s blood type. The California Supreme Court reasonably could have concluded that testimony about patterns that form when blood drips to the floor or about the rate at which Johnson was bleeding did not undermine the results of chemical tests indicating blood type. Nor did Richter provide any direct refutation of the State’s expert testimony describing how blood spatter near the couch suggested a shooting in the living room and how the blood patterns on Klein’s face were inconsistent with Richter’s theory that Klein had been killed in the bedroom doorway and moved to the couch. There was, furthermore, sufficient conventional circumstantial evidence pointing to Richter’s guilt. It included the gun safe and ammunition found at his home; his flight from the crime scene; his disposal of the .32-caliber gun and of Johnson’s pistol; his shifting story concerning his involvement; the disappearance prior to the arrival of the law enforcement officers of the .22-caliber weapon that killed Klein; the improbability of Branscombe’s not being wounded in the shootout that resulted in a combined four bullet wounds to Johnson and Klein; and the difficulties the intoxicated and twice-shot Johnson would have had in carrying the body of a dying man from bedroom doorway to living room couch, not to mention the lack of any obvious reason for him to do so. There was ample basis for the California Supreme Court to think any real possibility of Richter’s being acquitted was eclipsed by the remaining evidence pointing to guilt. * * * The California Supreme Court’s decision on the merits of Richter’s Strickland claim required more deference than it received. Richter was not entitled to the relief ordered by the Court of Appeals. The judgment is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

562.US.428

The Department of Veterans Affairs (VA) has a two-step process for adjudicating veterans’ benefits claims for service-connected disabilities: A VA regional office makes an initial decision on the claim; and a veteran dissatisfied with the decision may then seek de novo review in the Board of Veterans’ Appeals. Before 1988, a veteran whose claim was denied by the Board generally could not obtain further review, but the Veterans’ Judicial Review Act (VJRA) created the Court of Appeals for Veterans Claims (Veterans Court), an Article I tribunal, to review Board decisions adverse to veterans. A veteran must file a notice of appeal with that court within 120 days of the date when the Board’s final decision is properly mailed. 38 U. S. C. §7266(a). After the VA denied David Henderson’s claim for supplemental disability benefits, he filed a notice of appeal in the Veterans Court, missing the 120-day filing deadline by 15 days. Henderson argued that his failure to timely file should be excused under equitable tolling principles. While his appeal was pending, this Court decided Bowles v. Russell, 551 U. S. 205, which held that the statutory limitation on the length of an extension of time to file a notice of appeal in an ordinary civil case is “jurisdictional,” so that a party’s failure to file within that period could not be excused. The Veterans Court concluded that Bowles compelled jurisdictional treatment of the 120-day deadline and dismissed Henderson’s untimely appeal. The Federal Circuit affirmed. Held: The deadline for filing a notice of appeal with the Veterans Court does not have jurisdictional consequences. Pp. 4–13. (a) Branding a procedural rule as going to a court’s subject-matter jurisdiction alters the normal operation of the adversarial system. Federal courts have an independent obligation to ensure that they do not exceed the scope of their subject-matter jurisdiction and thus must raise and decide jurisdictional questions that the parties either overlook or elect not to press. Jurisdictional rules may also cause a waste of judicial resources and may unfairly prejudice litigants, since objections may be raised at any time, even after trial. Because of these drastic consequences, this Court has urged that a rule should not be referred to as jurisdictional unless it governs a court’s adjudicatory capacity, i.e., its subject-matter or personal jurisdiction. E.g., Reed Elsevier, Inc. v. Muchnick, 559 U. S. ___, ___. Among the rules that should not be described as jurisdictional are “claim-processing rules,” which seek to promote the orderly progress of litigation by requiring parties to take certain procedural steps at specified times. Although filing deadlines are quintessential claim-processing rules, Congress is free to attach jurisdictional consequences to such rules. Arbaugh v. Y & H Corp., 546 U. S. 500, applied a “readily administrable bright line” rule to determine whether Congress has done so: There must be a “clear” indication that Congress wanted the rule to be “jurisdictional.” Id., at 515–516. “[C]ontext, including this Court’s interpretation of similar provisions in many years past, is relevant,” Reed Elsevier, supra, at ___, to whether Congress has spoken clearly on this point. Pp. 4–6. (b) Congress did not clearly prescribe that the 120-day deadline here be jurisdictional. Pp. 7–12. (1) None of the precedents cited by the parties controls here. All of the cases they cite—e.g., Bowles, supra; Stone v. INS, 514 U. S. 386; and Bowen v. City of New York, 476 U. S. 467—involved review by Article III courts. This case, by contrast, involves review by an Article I tribunal as part of a unique administrative scheme. Instead of applying a categorical rule regarding review of administrative decisions, this Court attempts to ascertain Congress’ intent regarding the particular type of review at issue. Pp. 7–8. (2) Several factors indicate that 120-day deadline was not meant to be jurisdictional. The terms of §7266(a), which sets the deadline, provide no clear indication that the provision was meant to carry jurisdictional consequences. It neither speaks in “jurisdictional terms” nor refers “in any way to the jurisdiction of the [Veterans Court],” Zipes v. Trans World Airlines, Inc., 455 U. S. 385, 394. Nor does §7266’s placement within the VJRA provide such an indication. Its placement in a subchapter entitled “Procedure,” and not in the subchapter entitled “Organization and Jurisdiction,” suggests that Congress regarded the 120-day limit as a claim-processing rule. Most telling, however, are the singular characteristics of the review scheme that Congress created for adjudicating veterans’ benefits claims. Congress’ longstanding solicitude for veterans, United States v. Oregon, 366 U. S. 643, 647, is plainly reflected in the VJRA and in subsequent laws that place a thumb on the scale in the veteran’s favor in the course of administrative and judicial review of VA decisions. The contrast between ordinary civil litigation—which provided the context in Bowles—and the system Congress created for veterans is dramatic. In ordinary civil litigation suits must generally be commenced within a specified limitations period; the litigation is adversarial; plaintiffs must gather the evidence supporting their claims and generally bear the burden of production and persuasion; both parties may appeal an adverse decision; and a final judgment may be reopened only in narrow circumstances. By contrast, a veteran need not file an initial benefits claim within any fixed period; the VA proceedings are informal and nonadversarial; and the VA assists veterans in developing their supporting evidence and must give them the benefit of any doubt in evaluating that evidence. A veteran who loses before the Board may obtain review in the Veterans Court, but a Board decision in the veteran’s favor is final. And a veteran may reopen a claim simply by presenting new and material evidence. Rigid jurisdictional treatment of the 120-day period would clash sharply with this scheme. Particularly in light of “the canon that provisions for benefits to members of the Armed Services are to be construed in the beneficiaries’ favor,” King v. St. Vincent’s Hospital, 502 U. S. 215, 220–221, n. 9, this Court sees no clear indication that the 120-day limit was intended to carry the harsh consequences that accompany the jurisdiction tag. Contrary to the Government’s argument, the lack of review opportunities for veterans before 1988 is of little help in interpreting §7266(a). Section 7266(a) was enacted as part of the VJRA, and that legislation was decidedly favorable to veterans. Pp. 8–12. 589 F. 3d 1201, reversed and remanded. Alito, J., delivered the opinion of the Court, in which all other Members joined, except Kagan, J., who took no part in the consideration or decision of the case.

A veteran whose claim for federal benefits is denied by the Board of Veterans’ Appeals may appeal to the United States Court of Appeals for Veterans Claims (Veterans Court). To do so, the veteran must file a notice of appeal with the Veterans Court within 120 days after the date when the Board’s final decision is properly mailed. 38 U. S. C. §7266(a). This case presents the question whether a veteran’s failure to file a notice of appeal within the 120-day period should be regarded as having “jurisdictional” consequences. We hold that it should not. I A The Department of Veterans Affairs (VA) administers the federal program that provides benefits to veterans with service-connected disabilities. The VA has a two-step process for the adjudication of these claims. First, a VA regional office receives and processes veterans’ claims and makes an initial decision on whether to grant or deny benefits. Second, if a veteran is dissatisfied with the regional office’s decision, the veteran may obtain de novo review by the Board of Veterans’ Appeals. The Board is a body within the VA that makes the agency’s final decision in cases appealed to it. §§7101, 7104(a). The VA’s adjudicatory “process is designed to function throughout with a high degree of informality and solicitude for the claimant.” Walters v. National Assn. of Radiation Survivors, 473 U. S. 305, 311 (1985). A veteran faces no time limit for filing a claim, and once a claim is filed, the VA’s process for adjudicating it at the regional office and the Board is ex parte and nonadversarial, 38 CFR §§3.103(a), 20.700(c) (2010). The VA has a statutory duty to assist veterans in developing the evidence necessary to substantiate their claims. 38 U. S. C. §§5103(a) (2006 ed., Supp. III), 5103A (2006 ed.). And when evaluating claims, the VA must give veterans the “benefit of the doubt” whenever positive and negative evidence on a material issue is roughly equal. §5107(b). If a regional office denies a claim, the veteran has a generous one-year time limit to initiate an appeal to the Board. §7105(b)(1); 38 CFR §20.302(a). A veteran may also reopen a previously denied claim at any time by presenting “new and material evidence,” 38 U. S. C. §5108, and decisions by a regional office or the Board are subject to challenge at any time based on “clear and unmistakable error,” §§5109A, 7111. Before 1988, a veteran whose claim was rejected by the VA was generally unable to obtain further review. 38 U. S. C. §211(a) (1988 ed.).[Footnote 1] But the Veterans’ Judicial Review Act (VJRA), 102 Stat. 4105 (codified, as amended, in various sections of 38 U. S. C. (2006 ed. and Supp. III)), created the Veterans Court, an Article I tribunal, and authorized that court to review Board decisions adverse to veterans.[Footnote 2] §§7251, 7252(a) (2006 ed.). While proceedings before the Veterans Court are adversarial, see §7263, veterans have a remarkable record of success before that tribunal. Statistics compiled by the Veterans Court show that in the last decade, the court ordered some form of relief in around 79 percent of its “merits decisions.”[Footnote 3] Review of Veterans Court decisions on certain issues of law is available in the United States Court of Appeals for the Federal Circuit. §7292. Federal Circuit decisions may in turn be reviewed by this Court by writ of certiorari. B David Henderson served in the military during the Korean War. In 1992, the VA gave Henderson a 100-percent disability rating for paranoid schizophrenia, and in 2001, he filed a claim for supplemental benefits based on his need for in-home care. After a VA regional office and the Board denied his claim, he filed a notice of appeal with the Veterans Court, but he missed the 120-day filing deadline by 15 days. See §7266(a). The Veterans Court initially dismissed Henderson’s appeal as untimely. It concluded that Henderson was not entitled to equitable tolling of the deadline because he had not shown that his illness had caused his tardy filing. Later, the court granted Henderson’s motion for reconsideration, revoked the dismissal, and set the case for argument. While Henderson’s appeal was pending, however, we decided Bowles v. Russell, 551 U. S. 205 (2007). In Bowles, we held that the statutory limitation on the length of an extension of the time to file a notice of appeal in an ordinary civil case, 28 U. S. C. §2107(c) (2006 ed., Supp. III), is “jurisdictional,” and we therefore held that a party’s failure to file a notice of appeal within that period could not be excused based on equitable factors, or on the opposing party’s forfeiture or waiver of any objection to the late filing. Bowles, supra, at 213–214. After we announced our decision in Bowles, the Veterans Court directed the parties to brief that decision’s effect on prior Federal Circuit precedent that allowed the equitable tolling of the 120-day deadline for filing a notice of appeal in the Veterans Court. A divided panel of the Veterans Court concluded that Bowles compelled jurisdictional treatment of the 120-day deadline and dismissed Henderson’s untimely appeal for lack of jurisdiction. Henderson v. Peake, 22 Vet. App. 217 (2008). Henderson then appealed to the Federal Circuit, and a divided en banc court affirmed. 589 F. 3d 1201 (2009). We granted certiorari. 561 U. S. ___ (2010). II In this case, as in others that have come before us in recent years, we must decide whether a procedural rule is “jurisdictional.” See Reed Elsevier, Inc. v. Muchnick, 559 U. S. ___ (2010); Union Pacific R. Co. v. Locomotive Engineers and Trainmen Gen. Comm. of Adjustment, Central Region, 558 U. S. ___ (2009); Bowles, supra; Arbaugh v. Y & H Corp., 546 U. S. 500 (2006); Eberhart v. United States, 546 U. S. 12 (2005) (per curiam); Scarborough v. Principi, 541 U. S. 401 (2004); Kontrick v. Ryan, 540 U. S. 443 (2004). This question is not merely semantic but one of considerable practical importance for judges and litigants. Branding a rule as going to a court’s subject-matter jurisdiction alters the normal operation of our adversarial system. Under that system, courts are generally limited to addressing the claims and arguments advanced by the parties. See Sanchez-Llamas v. Oregon, 548 U. S. 331, 356–357 (2006). Courts do not usually raise claims or arguments on their own. But federal courts have an independent obligation to ensure that they do not exceed the scope of their jurisdiction, and therefore they must raise and decide jurisdictional questions that the parties either overlook or elect not to press. See Arbaugh, supra, at 514. Jurisdictional rules may also result in the waste of judicial resources and may unfairly prejudice litigants. For purposes of efficiency and fairness, our legal system is replete with rules requiring that certain matters be raised at particular times. See Sanchez-Llamas, supra, at 356–357. Objections to subject-matter jurisdiction, however, may be raised at any time. Thus, a party, after losing at trial, may move to dismiss the case because the trial court lacked subject-matter jurisdiction. Arbaugh, 546 U. S., at 508. Indeed, a party may raise such an objection even if the party had previously acknowledged the trial court’s jurisdiction. Ibid. And if the trial court lacked jurisdiction, many months of work on the part of the attorneys and the court may be wasted. Because the consequences that attach to the jurisdictional label may be so drastic, we have tried in recent cases to bring some discipline to the use of this term. We have urged that a rule should not be referred to as jurisdictional unless it governs a court’s adjudicatory capacity, that is, its subject-matter or personal jurisdiction. Reed Elsevier, supra, at ___ (slip op., at 6); Kontrick, supra, at 455. Other rules, even if important and mandatory, we have said, should not be given the jurisdictional brand. See Union Pacific, 558 U. S., at ___ (slip op., at 12). Among the types of rules that should not be described as jurisdictional are what we have called “claim-processing rules.” These are rules that seek to promote the orderly progress of litigation by requiring that the parties take certain procedural steps at certain specified times. Id., at ___ (slip op., at 14); Eberhart, supra, at 19; Scarborough, supra, at 413–414; Kontrick, supra, at 455–456. Filing deadlines, such as the 120-day filing deadline at issue here, are quintessential claim-processing rules. Accordingly, if we were simply to apply the strict definition of jurisdiction that we have recommended in our recent cases, we would reverse the decision of the Federal Circuit, and this opinion could end at this point. Unfortunately, the question before us is not quite that simple because Congress is free to attach the conditions that go with the jurisdictional label to a rule that we would prefer to call a claim-processing rule. See Bowles, supra, at 212–213. The question here, therefore, is whether Congress mandated that the 120-day deadline be “jurisdictional.” In Arbaugh, we applied a “readily administrable bright line” rule for deciding such questions. 546 U. S., at 515–516. Under Arbaugh, we look to see if there is any “clear” indication that Congress wanted the rule to be “jurisdictional.” Ibid. This approach is suited to capture Congress’ likely intent and also provides helpful guidance for courts and litigants, who will be “duly instructed” regarding a rule’s nature. See id., at 514–515, and n. 11. Congress, of course, need not use magic words in order to speak clearly on this point. “[C]ontext, including this Court’s interpretation of similar provisions in many years past, is relevant.” Reed Elsevier, supra, at ___ (slip op., at 13). When “a long line of this Court’s decisions left undisturbed by Congress,” Union Pacific, supra, at ___ (slip op., at 13), has treated a similar requirement as “jurisdictional,” we will presume that Congress intended to follow that course. See John R. Sand & Gravel Co. v. United States, 552 U. S. 130, 133–134, 139 (2008). III With these principles in mind, we consider whether Congress clearly prescribed that the deadline for filing a notice of appeal with the Veterans Court should be “jurisdictional.” A Contending that the 120-day filing deadline was meant to be jurisdictional, the Government maintains that Bowles is controlling. The Government reads Bowles to mean that all statutory deadlines for taking appeals in civil cases are jurisdictional. Since §7266(a) establishes a statutory deadline for taking an appeal in a civil case, the Government reasons, that deadline is jurisdictional. We reject the major premise of this syllogism. Bowles did not hold categorically that every deadline for seeking judicial review in civil litigation is jurisdictional. Instead, Bowles concerned an appeal from one court to another court. The “century’s worth of precedent and practice in American courts” on which Bowles relied involved appeals of that type. See 551 U. S., at 209–210, and n. 2. Contending that Bowles’ reasoning extends to the judicial review of administrative decisions, the Government relies on Stone v. INS, 514 U. S. 386 (1995). There, without elaboration, we described as “ ‘mandatory and jurisdictional’ ” the deadline for seeking review in the courts of appeals of final removal orders of the Board of Immigration Appeals. Id., at 405 (quoting Missouri v. Jenkins, 495 U. S. 33, 45 (1990)). The Government also notes that lower court decisions have uniformly held that the Hobbs Act’s 60-day time limit for filing a petition for review of certain final agency decisions, 28 U. S. C. §2344, is jurisdictional. Brief for United States 18. Petitioner correctly observes, however, that Veterans Court review of a VA decision denying benefits differs in many respects from court of appeals review of an agency decision under the Hobbs Act. Cf. Shinseki v. Sanders, 556 U. S. ___, ___ (2009) (slip op., at 15) (“Congress has made clear that the VA is not an ordinary agency”). And there is force to petitioner’s argument that a more ap-propriate analog is judicial review of an administrative decision denying Social Security disability benefits. The Social Security disability benefits program, like the veterans benefits program, is “unusually protective” of claimants, Heckler v. Day, 467 U. S. 104, 106–107 (1984). See also Sims v. Apfel, 530 U. S. 103, 110–112 (2000) (plurality opinion). Indeed, the Government acknowledges that “the Social Security and veterans-benefit review mechanisms share significant common attributes.” Brief for United States 16. And long before Congress enacted the VJRA, we held that the deadline for obtaining review of Social Security benefits decisions in district court, 42 U. S. C. §405(g), is not jurisdictional. Bowen v. City of New York, 476 U. S. 467, 478, and n. 10 (1986); Mathews v. Eldridge, 424 U. S. 319, 328, n. 9 (1976); Weinberger v. Salfi, 422 U. S. 749, 763–764 (1975). In the end, however, none of the precedents cited by the parties controls our decision here. All of those cases involved review by Article III courts. This case, by contrast, involves review by an Article I tribunal as part of a unique administrative scheme. Instead of applying a categorical rule regarding review of administrative decisions, we attempt to ascertain Congress’ intent regarding the particular type of review at issue in this case. B Several factors convince us that the 120-day deadline for seeking Veterans Court review was not meant to have jurisdictional attributes. The terms of the provision setting that deadline, 38 U. S. C. §7266(a), do not suggest, much less provide clear evidence, that the provision was meant to carry jurisdictional consequences. Section 7266(a) provides: “In order to obtain review by the Court of Appeals for Veterans Claims of a final decision of the Board of Veterans’ Appeals, a person adversely affected by such decision shall file a notice of appeal with the Court within 120 days after the date on which notice of the decision is mailed pursuant to section 7104(e) of this title.” This provision “does not speak in jurisdictional terms or refer in any way to the jurisdiction of the [Veterans Court],” Zipes v. Trans World Airlines, Inc., 455 U. S. 385, 394 (1982). If Congress had wanted the 120-day time to be treated as jurisdictional, it could have cast that provision in language like that in the provision of the VJRA that governs Federal Circuit review of decisions of the Veterans Court. This latter provision states that Federal Circuit review must be obtained “within the time and in the manner prescribed for appeal to United States courts of appeals from United States district courts.” §7292(a). Because the time for taking an appeal from a district court to a court of appeals in a civil case has long been understood to be jurisdictional, see Bowles, supra, at 209–210, and n. 2, this language clearly signals an intent to impose the same restrictions on appeals from the Veterans Court to the Federal Circuit. But the 120-day limit at issue in this case is not framed in comparable terms. It is true that §7266 is cast in mandatory language, but we have rejected the notion that “all mandatory prescriptions, however emphatic, are … properly typed jurisdictional.” Union Pacific, 558 U. S., at ___ (slip op., at 12) (quoting Arbaugh, 546 U. S., at 510; internal quotation marks omitted). Thus, the language of §7266 provides no clear indication that Congress wanted that provision to be treated as having jurisdictional attributes. Nor does §7266’s placement within the VJRA provide such an indication. Congress placed §7266, numbered §4066 in the enacting legislation, in a subchapter entitled “Procedure.” See VJRA, §301, 102 Stat. 4113, 4115–4116. That placement suggests Congress regarded the 120-day limit as a claim-processing rule. Cf. INS v. National Center for Immigrants’ Rights, Inc., 502 U. S. 183, 189 (1991) (“[T]he title of a statute or section can aid in resolving an ambiguity in the legislation’s text”). Congress elected not to place the 120-day limit in the VJRA subchapter entitled “Organization and Jurisdiction.” See 102 Stat. 4113–4115. Within that subchapter, a separate provision, captioned “Jurisdiction; finality of decisions,” prescribes the jurisdiction of the Veterans Court. Id., at 4113–4114. Subsection (a) of that provision, numbered §4052 in the enacting legislation, grants the Veterans Court “exclusive jurisdiction to review decisions of the Board of Veterans’ Appeals” and the “power to affirm, modify, or reverse a decision of the Board or to remand the matter, as appropriate.” Id., at 4113. It also prohibits the court from hearing appeals by the VA Secretary. Subsection (b) limits the court’s review to “the record of proceedings before the [VA],” specifies the scope of that review, and precludes review of the VA’s disability ratings schedule. Ibid. Nothing in this provision or in the “Organization and Jurisdiction” subchapter addresses the time for seeking Veterans Court review. While the terms and placement of §7266 provide some indication of Congress’ intent, what is most telling here are the singular characteristics of the review scheme that Congress created for the adjudication of veterans’ benefits claims. “The solicitude of Congress for veterans is of long standing.” United States v. Oregon, 366 U. S. 643, 647 (1961); see also Sanders, 556 U. S., at ___ (slip op., at 15). And that solicitude is plainly reflected in the VJRA, as well as in subsequent laws that “place a thumb on the scale in the veteran’s favor in the course of administrative and judicial review of VA decisions,” id., at ___ (Souter, J., dissenting) (slip op., at 2). See, e. g., Veterans Claims Assistance Act of 2000, 114 Stat. 2096; Act of Nov. 21, 1997, 111 Stat. 2271; VJRA, §103, 102 Stat. 4106–4107. The contrast between ordinary civil litigation—which provided the context of our decision in Bowles—and the system that Congress created for the adjudication of veterans’ benefits claims could hardly be more dramatic. In ordinary civil litigation, plaintiffs must generally commence their suits within the time specified in a statute of limitations, see 28 U. S. C. §1658, and the litigation is adversarial. Plaintiffs must gather the evidence that supports their claims and generally bear the burden of production and persuasion. Both parties may appeal an adverse trial-court decision, see §1291, and a final judgment may be reopened only in narrow circumstances. See Fed. Rule Civ. Proc. 60. By contrast, a veteran seeking benefits need not file an initial claim within any fixed period after the alleged onset of disability or separation from service. When a claim is filed, proceedings before the VA are informal and nonadversarial. The VA is charged with the responsibility of assisting veterans in developing evidence that supports their claims, and in evaluating that evidence, the VA must give the veteran the benefit of any doubt. If a veteran is unsuccessful before a regional office, the veteran may obtain de novo review before the Board, and if the veteran loses before the Board, the veteran can obtain further review in the Veterans Court. A Board decision in the veteran’s favor, on the other hand, is final. And even if a veteran is denied benefits after exhausting all avenues of administrative and judicial review, a veteran may reopen a claim simply by presenting “new and material evidence.” Rigid jurisdictional treatment of the 120-day period for filing a notice of appeal in the Veterans Court would clash sharply with this scheme. We have long applied “the canon that provisions for benefits to members of the Armed Services are to be construed in the beneficiaries’ favor.” King v. St. Vincent’s Hospital, 502 U. S. 215, 220–221, n. 9 (1991); see also Coffy v. Republic Steel Corp., 447 U. S. 191, 196 (1980); Fishgold v. Sullivan Drydock & Repair Corp., 328 U. S. 275, 285 (1946). Particularly in light this canon, we do not find any clear indication that the 120-day limit was intended to carry the harsh consequences that accompany the jurisdiction tag. The Government argues that there is no reason to think that jurisdictionally time-limited review is inconsistent with a pro-veteran administrative scheme because, prior to the enactment of the VJRA in 1988, VA decisions were not subject to any further review at all. Brief for United States 29. The provision at issue here, however, was enacted as part of the VJRA, and that legislation was decidedly favorable to veterans. Accordingly, the review opportunities available to veterans before the VJRA was enacted are of little help in interpreting 38 U. S. C. §7266(a). IV We hold that the deadline for filing a notice of appeal with the Veterans Court does not have jurisdictional attributes. The 120-day limit is nevertheless an important procedural rule. Whether this case falls within any exception to the rule is a question to be considered on remand.[Footnote 4] The judgment of the United States Court of Appeals for the Federal Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 Section 211(a) did not foreclose judicial review of constitutional challenges to veterans’ benefits legislation, Johnson v. Robison, 415 U. S. 361, 366–374 (1974), or of challenges to VA benefits regulations based on later-in-time statutes that the VA did not administer exclusively, Traynor v. Turnage, 485 U. S. 535, 541–545 (1988). Footnote 2 When such an appeal is taken, the Veterans Court’s scope of review, §7261, is similar to that of an Article III court reviewing agency action under the Administrative Procedure Act, 5 U. S. C. §706. Footnote 3 See United States Court of Appeals for Veterans Claims, Annual Reports 2000–2009, http://uscourts.cavc.gov/documents/Annual_Report_ FY_2009_October_1_2008_to_September_30_2009.pdf (as visited Feb. 25, 2011, and available in Clerk of Court’s case file). Footnote 4 The parties have not asked us to address whether the 120-day deadline in 38 U. S. C. §7266(a) is subject to equitable tolling, nor has the Government disputed that the deadline is subject to equitable tolling if it is not jurisdictional. See Brief for Petitioner 18. Accordingly, we express no view on this question.

564.US.2010_09-1343

Respondent Nicastro injured his hand while using a metal-shearing machine that petitioner J. McIntyre Machinery, Ltd. (J. McIntyre), manufactured in England, where the company is incorporated and operates. Nicastro filed this products-liability suit in a state court in New Jersey, where the accident occurred, but J. McIntyre sought to dismiss the suit for want of personal jurisdiction. Nicastro’s jurisdictional claim was based on three primary facts: A U. S. distributor agreed to sell J. McIntyre’s machines in this country; J. McIntyre officials attended trade shows in several States, albeit not in New Jersey; and no more than four J. McIntyre machines (the record suggests only one), including the one at issue, ended up in New Jersey. The State Supreme Court held that New Jersey’s courts can exercise jurisdiction over a foreign manufacturer without contravening the Fourteenth Amendment’s Due Process Clause so long as the manufacturer knew or reasonably should have known that its products are distributed through a nationwide distribution system that might lead to sales in any of the States. Invoking this “stream-of-commerce” doctrine of jurisdiction, the court relied in part on Asahi Metal Industry Co. v. Superior Court of Cal., Solano Cty., 480 U. S. 102. Applying its test, the court concluded that J. McIntyre was subject to jurisdiction in New Jersey, even though at no time had it advertised in, sent goods to, or in any relevant sense targeted the State. Held: The judgment is reversed. 201 N. J. 48, 987 A. 2d 575, reversed. Justice Kennedy, joined by The Chief Justice, Justice Scalia, and Justice Thomas, concluded that because J. McIntyre never engaged in any activities in New Jersey that revealed an intent to invoke or benefit from the protection of the State’s laws, New Jersey is without power to adjudge the company’s rights and liabilities, and its exercise of jurisdiction would violate due process. Pp. 4–12. (a) Due process protects the defendant’s right not to be coerced except by lawful judicial power. A court may subject a defendant to judgment only when the defendant has sufficient contacts with the sovereign “such that the maintenance of the suit does not offend ‘traditional notions of fair play and substantial justice.’ ” International Shoe Co. v. Washington, 326 U. S. 310, 316. Freeform fundamental fairness notions divorced from traditional practice cannot transform a judgment rendered without authority into law. As a general rule, the sovereign’s exercise of power requires some act by which the defendant “purposefully avails itself of the privilege of conducting activities within the forum State, thus invoking the benefits and protections of its laws.” Hanson v. Denckla, 357 U. S. 235, 253. In cases like this one, it is the defendant’s purposeful availment that makes jurisdiction consistent with “fair play and substantial justice” notions. No “stream-of-commerce” doctrine can displace that general rule for products-liability cases. The rules and standards for determining state jurisdiction over an absent party have been unclear because of decades-old questions left open in Asahi. The imprecision arising from Asahi, for the most part, results from its statement of the relation between jurisdiction and the “stream of commerce.” That concept, like other metaphors, has its deficiencies as well as its utilities. It refers to the movement of goods from manufacturers through distributors to consumers, yet beyond that descriptive purpose its meaning is far from exact. A defendant’s placement of goods into commerce “with the expectation that they will be purchased by consumers within the forum State” may indicate purposeful availment. World-Wide Volkswagen Corp. v. Woodson, 444 U. S. 286, 298. But that does not amend the general rule of personal jurisdiction. The principal inquiry in cases of this sort is whether the defendant’s activities manifest an intention to submit to the power of a sovereign. See, e.g., Hanson, supra, at 253. In Asahi, Justice Brennan’s concurrence (joined by three other Justices) discarded the central concept of sovereign authority in favor of fairness and foreseeability considerations on the theory that the defendant’s ability to anticipate suit is the touchstone of jurisdiction. 480 U. S., at 117. However, Justice O’Connor’s lead opinion (also for four Justices) stated that “[t]he ‘substantial connection’ between the defendant and the forum State necessary for a finding of minimum contacts must come about by an action of the defendant purposefully directed toward the forum State.” Id., at 112. Since Asahi, the courts have sought to reconcile the competing opinions. But Justice Brennan’s rule based on general notions of fairness and foreseeability is inconsistent with the premises of lawful judicial power under this Court’s precedents. Today’s conclusion that the authority to subject a defendant to judgment depends on purposeful availment is consistent with Justice O’Connor’s Asahi opinion. Pp. 4–10. (b) Nicastro has not established that J. McIntyre engaged in conduct purposefully directed at New Jersey. The company had no office in New Jersey; it neither paid taxes nor owned property there; and it neither advertised in, nor sent any employees to, the State. Indeed, the trial court found that petitioner did not have a single contact with the State apart from the fact that the machine in question ended up there. Neither these facts, nor the three on which Nicastro centered his jurisdictional claim, show that J. McIntyre purposefully availed itself of the New Jersey market. Pp. 10–12. Justice Breyer, joined by Justice Alito, agreed that the New Jersey Supreme Court’s judgment must be reversed, but concluded that because this case does not present issues arising from recent changes in commerce and communication, it is unwise to announce a rule of broad applicability without fully considering modern-day consequences. Rather, the outcome of the case is determined by the Court’s precedents. Pp. 2–7. (a) Based on the record, respondent Nicastro failed to meet his burden to demonstrate that it was constitutionally proper to exercise jurisdiction over petitioner J. McIntyre Machinery, Ltd. (British Manufacturer). The three primary facts the state high court relied on do not satisfy due process. None of the Court’s precedents finds that a single isolated sale, even if accompanied by the kind of sales effort indicated here, is sufficient. See World-Wide Volkswagen Corp. v. Woodson, 444 U. S. 286; Asahi Metal Industry Co. v. Superior Court of Cal., Solano Cty., 480 U. S. 102. Here, the relevant facts show no “regular … flow” or “regular course” of sales in New Jersey, id., at 117 (Brennan, J., concurring in part and concurring in judgment); id., at 122 (Stevens, J., concurring in part and concurring in judgment); and there is no “something more,” such as special state-related design, advertising, advice, or marketing, id., at 111, 112 (opinion of O’Connor, J.), that would warrant the assertion of jurisdiction. Nicastro has shown no specific effort by the British Manufacturer to sell in New Jersey. And he has not otherwise shown that the British Manufacturer “ ‘purposefully avail[ed] itself of the privilege of conducting activities’ ” within New Jersey, or that it delivered its goods in the stream of commerce “with the expectation that they will be purchased” by New Jersey users. World-Wide Volkswagen, supra, at 297–298. Pp. 2–4. (b) Justice Breyer would not go further. Because the incident at issue does not implicate modern concerns, and because the factual record leaves many open questions, this is an unsuitable vehicle for making broad pronouncements that refashion basic jurisdictional rules. At a minimum, he would not work such a change to the law in the way either the plurality or the New Jersey Supreme Court suggests without a better understanding of the relevant contemporary commercial circumstances. Insofar as such considerations are relevant to any change in present law, they might be presented in a case (unlike the present one) in which the Solicitor General participates. Pp. 4–7. Kennedy, J., announced the judgment of the Court and delivered an opinion, in which Roberts, C. J., and Scalia and Thomas, JJ., joined. Breyer, J., filed an opinion concurring in the judgment, in which Alito, J., joined. Ginsburg, J., filed a dissenting opinion, in which Sotomayor and Kagan, JJ., joined.

in which the Chief Justice, Justice Scalia, and Justice Thomas join. Whether a person or entity is subject to the jurisdiction of a state court despite not having been present in the State either at the time of suit or at the time of the alleged injury, and despite not having consented to the exercise of jurisdiction, is a question that arises with great frequency in the routine course of litigation. The rules and standards for determining when a State does or does not have jurisdiction over an absent party have been unclear because of decades-old questions left open in Asahi Metal Industry Co. v. Superior Court of Cal., Solano Cty., 480 U. S. 102 (1987). Here, the Supreme Court of New Jersey, relying in part on Asahi, held that New Jersey’s courts can exercise jurisdiction over a foreign manufacturer of a product so long as the manufacturer “knows or reasonably should know that its products are distributed through a nationwide distribution system that might lead to those products being sold in any of the fifty states.” Nicastro v. McIntyre Machinery America, Ltd., 201 N. J. 48, 76, 77, 987 A. 2d 575, 591, 592 (2010). Applying that test, the court concluded that a British manufacturer of scrap metal machines was subject to jurisdiction in New Jersey, even though at no time had it advertised in, sent goods to, or in any relevant sense targeted the State. That decision cannot be sustained. Although the New Jersey Supreme Court issued an extensive opinion with care-ful attention to this Court’s cases and to its own pre-cedent, the “stream of commerce” metaphor carried the decision far afield. Due process protects the defendant’s right not to be coerced except by lawful judicial power. As a general rule, the exercise of judicial power is not lawful unless the defendant “purposefully avails itself of the privilege of conducting activities within the forum State, thus invoking the benefits and protections of its laws.” Hanson v. Denckla, 357 U. S. 235, 253 (1958). There may be exceptions, say, for instance, in cases involving an intentional tort. But the general rule is applicable in this products-liability case, and the so-called “stream-of-commerce” doctrine cannot displace it. I This case arises from a products-liability suit filed in New Jersey state court. Robert Nicastro seriously injured his hand while using a metal-shearing machine manufactured by J. McIntyre Machinery, Ltd. (J. McIntyre). The accident occurred in New Jersey, but the machine was manufactured in England, where J. McIntyre is incorporated and operates. The question here is whether the New Jersey courts have jurisdiction over J. McIntyre, notwithstanding the fact that the company at no time either marketed goods in the State or shipped them there. Nicastro was a plaintiff in the New Jersey trial court and is the respondent here; J. McIntyre was a defendant and is now the petitioner. At oral argument in this Court, Nicastro’s counsel stressed three primary facts in defense of New Jersey’s as-sertion of jurisdiction over J. McIntyre. See Tr. of Oral Arg. 29–30. First, an independent company agreed to sell J. McIntyre’s machines in the United States. J. McIntyre itself did not sell its machines to buyers in this country beyond the U. S. distributor, and there is no allegation that the distributor was under J. McIntyre’s control. Second, J. McIntyre officials attended annual conventions for the scrap recycling industry to advertise J. Mc-Intyre’s machines alongside the distributor. The conventions took place in various States, but never in New Jersey. Third, no more than four machines (the record suggests only one, see App. to Pet. for Cert. 130a), including the machine that caused the injuries that are the basis for this suit, ended up in New Jersey. In addition to these facts emphasized by petitioner, the New Jersey Supreme Court noted that J. McIntyre held both United States and European patents on its recycling technology. 201 N. J., at 55, 987 A. 2d, at 579. It also noted that the U. S. distributor “structured [its] adver-tising and sales efforts in accordance with” J. McIntyre’s “direction and guidance whenever possible,” and that “at least some of the machines were sold on consignment to” the distributor. Id., at 55, 56, 987 A. 2d, at 579 (internal quotation marks omitted). In light of these facts, the New Jersey Supreme Court concluded that New Jersey courts could exercise jurisdiction over petitioner without contravention of the Due Process Clause. Jurisdiction was proper, in that court’s view, because the injury occurred in New Jersey; because petitioner knew or reasonably should have known “that its products are distributed through a nationwide distribution system that might lead to those products being sold in any of the fifty states”; and because petitioner failed to “take some reasonable step to prevent the distribution of its prod-ucts in this State.” Id., at 77, 987 A. 2d, at 592. Both the New Jersey Supreme Court’s holding and its account of what it called “[t]he stream-of-commerce doctrine of jurisdiction,” id., at 80, 987 A. 2d, at 594, were incorrect, however. This Court’s Asahi decision may be responsible in part for that court’s error regarding the stream of commerce, and this case presents an opportunity to provide greater clarity. II The Due Process Clause protects an individual’s right to be deprived of life, liberty, or property only by the exercise of lawful power. Cf. Giaccio v. Pennsylvania, 382 U. S. 399, 403 (1966) (The Clause “protect[s] a person against having the Government impose burdens upon him except in accordance with the valid laws of the land”). This is no less true with respect to the power of a sovereign to resolve disputes through judicial process than with respect to the power of a sovereign to prescribe rules of conduct for those within its sphere. See Steel Co. v. Citizens for Bet-ter Environment, 523 U. S. 83, 94 (1998) (“Jurisdiction is power to declare the law”). As a general rule, neither statute nor judicial decree may bind strangers to the State. Cf. Burnham v. Superior Court of Cal., County of Marin, 495 U. S. 604, 608–609 (1990) (opinion of Scalia, J.) (invoking “the phrase coram non judice, ‘before a person not a judge’—meaning, in effect, that the proceeding in question was not a judicial proceeding because lawful judicial authority was not present, and could therefore not yield a judgment”) A court may subject a defendant to judgment only when the defendant has sufficient contacts with the sovereign “such that the maintenance of the suit does not offend ‘traditional notions of fair play and substantial justice.’ ” International Shoe Co. v. Washington, 326 U. S. 310, 316 (1945) (quoting Milliken v. Meyer, 311 U. S. 457, 463 (1940)). Freeform notions of fundamental fairness divorced from traditional practice cannot transform a judgment rendered in the absence of authority into law. As a general rule, the sovereign’s exercise of power requires some act by which the defendant “purposefully avails itself of the privilege of conducting activities within the forum State, thus invoking the benefits and protections of its laws,” Hanson, 357 U. S., at 253, though in some cases, as with an intentional tort, the defendant might well fall within the State’s authority by reason of his attempt to obstruct its laws. In products-liability cases like this one, it is the defendant’s purposeful availment that makes jurisdiction consistent with “traditional notions of fair play and substantial justice.” A person may submit to a State’s authority in a number of ways. There is, of course, explicit consent. E.g., In-surance Corp. of Ireland v. Compagnie des Bauxites de Guinee, 456 U. S. 694, 703 (1982). Presence within a State at the time suit commences through service of process is another example. See Burnham, supra. Citizenship or domicile—or, by analogy, incorporation or principal place of business for corporations—also indicates general submission to a State’s powers. Goodyear Dunlop Tires Operations, S. A. v. Brown, post, p. __. Each of these examples reveals circumstances, or a course of conduct, from which it is proper to infer an intention to benefit from and thus an intention to submit to the laws of the forum State. Cf. Burger King Corp. v. Rudzewicz, 471 U. S. 462, 476 (1985). These examples support exercise of the general jurisdiction of the State’s courts and allow the State to resolve both matters that originate within the State and those based on activities and events elsewhere. Helicopteros Nacionales de Colombia, S. A. v. Hall, 466 U. S. 408, 414, and n. 9 (1984). By contrast, those who live or operate primarily outside a State have a due process right not to be subjected to judgment in its courts as a general matter. There is also a more limited form of submission to a State’s authority for disputes that “arise out of or are con-nected with the activities within the state.” International Shoe Co., supra, at 319. Where a defendant “purposefully avails itself of the privilege of conducting activities within the forum State, thus invoking the benefits and protections of its laws,” Hanson, supra, at 253, it submits to the judicial power of an otherwise foreign sovereign to the extent that power is exercised in connection with the defendant’s activities touching on the State. In other words, submission through contact with and activity directed at a sovereign may justify specific jurisdiction “in a suit arising out of or related to the defendant’s contacts with the forum.” Helicopteros, supra, at 414, n. 8; see also Goodyear, post, at 2. The imprecision arising from Asahi, for the most part, results from its statement of the relation between jurisdiction and the “stream of commerce.” The stream of commerce, like other metaphors, has its deficiencies as well as its utility. It refers to the movement of goods from manufacturers through distributors to consumers, yet beyond that descriptive purpose its meaning is far from exact. This Court has stated that a defendant’s placing goods into the stream of commerce “with the expectation that they will be purchased by consumers within the forum State” may indicate purposeful availment. World-Wide Volkswagen Corp. v. Woodson, 444 U. S. 286, 298 (1980) (finding that expectation lacking). But that statement does not amend the general rule of personal jurisdiction. It merely observes that a defendant may in an appropriate case be subject to jurisdiction without entering the forum—itself an unexceptional proposition—as where man-ufacturers or distributors “seek to serve” a given State’s market. Id., at 295. The principal inquiry in cases of this sort is whether the defendant’s activities manifest an intention to submit to the power of a sovereign. In other words, the defendant must “purposefully avai[l] it-self of the privilege of conducting activities within the forum State, thus invoking the benefits and protections of its laws.” Hanson, supra, at 253; Insurance Corp., supra, at 704–705 (“[A]ctions of the defendant may amount to a legal submission to the jurisdiction of the court”). Sometimes a defendant does so by sending its goods rather than its agents. The defendant’s transmission of goods permits the exercise of jurisdiction only where the defendant can be said to have targeted the forum; as a general rule, it is not enough that the defendant might have predicted that its goods will reach the forum State. In Asahi, an opinion by Justice Brennan for four Justices outlined a different approach. It discarded the central concept of sovereign authority in favor of considerations of fairness and foreseeability. As that concurrence contended, “jurisdiction premised on the placement of a product into the stream of commerce [without more] is consistent with the Due Process Clause,” for “[a]s long as a participant in this process is aware that the final product is being marketed in the forum State, the possibility of a lawsuit there cannot come as a surprise.” 480 U. S., at 117 (opinion concurring in part and concurring in judgment). It was the premise of the concurring opinion that the defendant’s ability to anticipate suit renders the assertion of jurisdiction fair. In this way, the opinion made foreseeability the touchstone of jurisdiction. The standard set forth in Justice Brennan’s concurrence was rejected in an opinion written by Justice O’Connor; but the relevant part of that opinion, too, commanded the assent of only four Justices, not a majority of the Court. That opinion stated: “The ‘substantial connection’ between the defendant and the forum State necessary for a finding of minimum contacts must come about by an action of the defendant purposefully directed toward the forum State. The placement of a product into the stream of commerce, without more, is not an act of the defendant purposefully directed toward the forum State.” Id., at 112 (emphasis deleted; citations omitted). Since Asahi was decided, the courts have sought to rec-oncile the competing opinions. But Justice Brennan’s con-currence, advocating a rule based on general notions of fairness and foreseeability, is inconsistent with the premises of lawful judicial power. This Court’s precedents make clear that it is the defendant’s actions, not his expectations, that empower a State’s courts to subject him to judgment. The conclusion that jurisdiction is in the first instance a question of authority rather than fairness explains, for example, why the principal opinion in Burnham “conducted no independent inquiry into the desirability or fairness” of the rule that service of process within a State suffices to establish jurisdiction over an otherwise foreign defendant. 495 U. S., at 621. As that opinion explained, “[t]he view developed early that each State had the power to hale before its courts any individual who could be found within its borders.” Id., at 610. Furthermore, were general fairness considerations the touchstone of jurisdiction, a lack of purposeful availment might be excused where carefully crafted judicial procedures could otherwise protect the defendant’s interests, or where the plaintiff would suffer substantial hardship if forced to litigate in a foreign forum. That such considerations have not been deemed controlling is instructive. See, e.g., World-Wide Volkswagen, supra, at 294. Two principles are implicit in the foregoing. First, per-sonal jurisdiction requires a forum-by-forum, or sovereign-by-sovereign, analysis. The question is whether a de-fendant has followed a course of conduct directed at the society or economy existing within the jurisdiction of a given sovereign, so that the sovereign has the power to subject the defendant to judgment concerning that conduct. Personal jurisdiction, of course, restricts “judicial power not as a matter of sovereignty, but as a matter of individual liberty,” for due process protects the individual’s right to be subject only to lawful power. Insurance Corp., 456 U. S., at 702. But whether a judicial judgment is lawful depends on whether the sovereign has authority to render it. The second principle is a corollary of the first. Because the United States is a distinct sovereign, a defendant may in principle be subject to the jurisdiction of the courts of the United States but not of any particular State. This is consistent with the premises and unique genius of our Constitution. Ours is “a legal system unprecedented in form and design, establishing two orders of government, each with its own direct relationship, its own privity, its own set of mutual rights and obligations to the people who sustain it and are governed by it.” U. S. Term Limits, Inc. v. Thornton, 514 U. S. 779, 838 (1995) (Kennedy, J., concurring). For jurisdiction, a litigant may have the requisite relationship with the United States Government but not with the government of any individual State. That would be an exceptional case, however. If the defendant is a domestic domiciliary, the courts of its home State are available and can exercise general jurisdiction. And if another State were to assert jurisdiction in an inappropriate case, it would upset the federal balance, which posits that each State has a sovereignty that is not subject to unlawful intrusion by other States. Furthermore, foreign corporations will often target or concentrate on particular States, subjecting them to specific jurisdiction in those forums. It must be remembered, however, that although this case and Asahi both involve foreign manufacturers, the undesirable consequences of Justice Brennan’s approach are no less significant for domestic producers. The owner of a small Florida farm might sell crops to a large nearby distributor, for example, who might then distribute them to grocers across the country. If foreseeability were the controlling criterion, the farmer could be sued in Alaska or any number of other States’ courts without ever leaving town. And the issue of foreseeability may itself be contested so that significant expenses are incurred just on the preliminary issue of jurisdiction. Jurisdictional rules should avoid these costs whenever possible. The conclusion that the authority to subject a defendant to judgment depends on purposeful availment, consistent with Justice O’Connor’s opinion in Asahi, does not by itself resolve many difficult questions of jurisdiction that will arise in particular cases. The defendant’s conduct and the economic realities of the market the defendant seeks to serve will differ across cases, and judicial exposition will, in common-law fashion, clarify the contours of that principle. III In this case, petitioner directed marketing and sales efforts at the United States. It may be that, assuming it were otherwise empowered to legislate on the subject, the Congress could authorize the exercise of jurisdiction in appropriate courts. That circumstance is not presented in this case, however, and it is neither necessary nor appropriate to address here any constitutional concerns that might be attendant to that exercise of power. See Asahi, 480 U. S., at 113, n. Nor is it necessary to determine what substantive law might apply were Congress to authorize jurisdiction in a federal court in New Jersey. See Hanson, 357 U. S., at 254 (“The issue is personal jurisdiction, not choice of law”). A sovereign’s legislative authority to regulate conduct may present considerations different from those presented by its authority to subject a defendant to judgment in its courts. Here the question concerns the authority of a New Jersey state court to exercise ju-risdiction, so it is petitioner’s purposeful contacts with New Jersey, not with the United States, that alone are relevant. Respondent has not established that J. McIntyre engaged in conduct purposefully directed at New Jersey. Recall that respondent’s claim of jurisdiction centers on three facts: The distributor agreed to sell J. McIntyre’s machines in the United States; J. McIntyre officials attended trade shows in several States but not in New Jersey; and up to four machines ended up in New Jersey. The British manufacturer had no office in New Jersey; it neither paid taxes nor owned property there; and it neither advertised in, nor sent any employees to, the State. Indeed, after discovery the trial court found that the “defendant does not have a single contact with New Jersey short of the machine in question ending up in this state.” App. to Pet. for Cert. 130a. These facts may reveal an intent to serve the U. S. market, but they do not show that J. McIntyre purposefully availed itself of the New Jersey market. It is notable that the New Jersey Supreme Court appears to agree, for it could “not find that J. McIntyre had a presence or minimum contacts in this State—in any jurisprudential sense—that would justify a New Jersey court to exercise jurisdiction in this case.” 201 N. J., at 61, 987 A. 2d, at 582. The court nonetheless held that petitioner could be sued in New Jersey based on a “stream-of-commerce theory of jurisdiction.” Ibid. As discussed, however, the stream-of-commerce metaphor cannot supersede either the mandate of the Due Process Clause or the limits on judicial authority that Clause ensures. The New Jersey Supreme Court also cited “significant policy reasons” to justify its holding, including the State’s “strong interest in protecting its citizens from defective products.” Id., at 75, 987 A. 2d, at 590. That interest is doubtless strong, but the Constitution commands restraint before discarding liberty in the name of expediency. * * * Due process protects petitioner’s right to be subject only to lawful authority. At no time did petitioner engage in any activities in New Jersey that reveal an intent to invoke or benefit from the protection of its laws. New Jersey is without power to adjudge the rights and liabilities of J. McIntyre, and its exercise of jurisdiction would violate due process. The contrary judgment of the New Jersey Supreme Court is Reversed.

564.US.2010_09-11121

Police stopped and questioned petitioner J. D. B., a 13-year-old, seventh-grade student, upon seeing him near the site of two home break-ins. Five days later, after a digital camera matching one of the stolen items was found at J. D. B.’s school and seen in his possession, Investigator DiCostanzo went to the school. A uniformed police officer on detail to the school took J. D. B. from his classroom to a closed-door conference room, where police and school administrators questioned him for at least 30 minutes. Before beginning, they did not give him Miranda warnings or the opportunity to call his grandmother, his legal guardian, nor tell him he was free to leave the room. He first denied his involvement, but later confessed after officials urged him to tell the truth and told him about the prospect of juvenile detention. DiCostanzo only then told him that he could refuse to answer questions and was free to leave. Asked whether he understood, J. D. B. nodded and provided further detail, including the location of the stolen items. He also wrote a statement, at DiCostanzo’s request. When the school day ended, he was permitted to leave to catch the bus home. Two juvenile petitions were filed against J. D. B., charging him with breaking and entering and with larceny. His public defender moved to suppress his statements and the evidence derived therefrom, arguing that J. D. B. had been interrogated in a custodial setting without being afforded Miranda warnings and that his statements were involuntary. The trial court denied the motion. J. D. B. entered a transcript of admission to the charges, but renewed his objection to the denial of his motion to suppress. The court adjudicated him delinquent, and the North Carolina Court of Appeals and State Supreme Court affirmed. The latter court declined to find J. D. B.’s age relevant to the determination whether he was in police custody. Held: A child’s age properly informs Miranda’s custody analysis. Pp. 5–18. (a) Custodial police interrogation entails “inherently compelling pressures,” Miranda v. Arizona, 384 U. S. 436, 467, that “can induce a frighteningly high percentage of people to confess to crimes they never committed,” Corley v. United States, 556 U. S. ___, ___. Recent studies suggest that risk is all the more acute when the subject of custodial interrogation is a juvenile. Whether a suspect is “in custody” for Miranda purposes is an objective determination involving two discrete inquires: “first, what were the circumstances surrounding the interrogation; and second, given those circumstances, would a reasonable person have felt he or she was at liberty to terminate the interrogation and leave.” Thompson v. Keohane, 516 U. S. 99, 112 (footnote omitted). The police and courts must “examine all of the circumstances surrounding the interrogation,” Stansbury v. California, 511 U. S. 318, 322, including those that “would have affected how a reasonable person” in the suspect’s position “would perceive his or her freedom to leave,” id., at 325. However, the test involves no consideration of the particular suspect’s “actual mindset.” Yarborough v. Alvarado, 541 U. S. 652, 667. By limiting analysis to objective circumstances, the test avoids burdening police with the task of anticipating each suspect’s idiosyncrasies and divining how those particular traits affect that suspect’s subjective state of mind. Berkemer v. McCarty, 468 U. S. 420, 430–431. Pp. 5–8. (b) In some circumstances, a child’s age “would have affected how a reasonable person” in the suspect’s position “would perceive his or her freedom to leave.” Stansbury, 511 U. S., at 325. Courts can account for that reality without doing any damage to the objective nature of the custody analysis. A child’s age is far “more than a chronological fact.” Eddings v. Oklahoma, 455 U. S. 104, 115. It is a fact that “generates commonsense conclusions about behavior and perception,” Alvarado, 541 U. S., at 674, that apply broadly to children as a class. Children “generally are less mature and responsible than adults,” Eddings, 455 U. S., at 115; they “often lack the experience, perspective, and judgment to recognize and avoid choices that could be detrimental to them,” Bellotti v. Baird, 443 U. S. 622, 635; and they “are more vulnerable or susceptible to … outside pressures” than adults, Roper v. Simmons, 543 U. S. 551, 569. In the specific context of police interrogation, events that “would leave a man cold and unimpressed can overawe and overwhelm a” teen. Haley v. Ohio, 332 U. S. 596, 599. The law has historically reflected the same assumption that children characteristically lack the capacity to exercise mature judgment and possess only an incomplete ability to understand the world around them. Legal disqualifications on children as a class—e.g., limitations on their ability to marry without parental consent—exhibit the settled understanding that the differentiating characteristics of youth are universal. Given a history “replete with laws and judicial recognition” that children cannot be viewed simply as miniature adults, Eddings, 455 U. S., at 115–116, there is no justification for taking a different course here. So long as the child’s age was known to the officer at the time of the interview, or would have been objectively apparent to a reasonable officer, including age as part of the custody analysis requires officers neither to consider circumstances “unknowable” to them, Berkemer, 468 U. S., at 430, nor to“ ‘ “anticipat[e] the frailties or idiosyncrasies” of the particular suspect being questioned.” ’ ” Alvarado, 541 U. S., at 662. Precisely because childhood yields objective conclusions, considering age in the custody analysis does not involve a determination of how youth affects a particular child’s subjective state of mind. In fact, were the court precluded from taking J. D. B.’s youth into account, it would be forced to evaluate the circumstances here through the eyes of a reasonable adult, when some objective circumstances surrounding an interrogation at school are specific to children. These conclusions are not undermined by the Court’s observation in Alvarado that accounting for a juvenile’s age in the Miranda custody analysis “could be viewed as creating a subjective inquiry,” 541 U. S., at 668. The Court said nothing about whether such a view would be correct under the law or whether it simply merited deference under the Antiterrorism and Effective Death Penalty Act of 1996, 110 Stat. 1214. So long as the child’s age was known to the officer, or would have been objectively apparent to a reasonable officer, including age in the custody analysis is consistent with the Miranda test’s objective nature. This does not mean that a child’s age will be a determinative, or even a significant, factor in every case, but it is a reality that courts cannot ignore. Pp. 8–14. (c) Additional arguments that the State and its amici offer for excluding age from the custody inquiry are unpersuasive. Pp. 14–18. (d) On remand, the state courts are to address the question whether J. D. B. was in custody when he was interrogated, taking account of all of the relevant circumstances of the interrogation, including J. D. B.’s age at the time. P. 18. 363 N. C. 664, 686 S. E. 2d 135, reversed and remanded. Sotomayor, J., delivered the opinion of the Court, in which Kennedy, Ginsburg, Breyer, and Kagan, JJ., joined. Alito, J., filed a dissenting opinion, in which Roberts, C. J., and Scalia and Thomas, JJ., joined.

This case presents the question whether the age of a child subjected to police questioning is relevant to the custody analysis of Miranda v. Arizona, 384 U. S. 436 (1966). It is beyond dispute that children will often feel bound to submit to police questioning when an adult in the same circumstances would feel free to leave. Seeing no reason for police officers or courts to blind themselves to that commonsense reality, we hold that a child’s age properly informs the Miranda custody analysis. I A Petitioner J. D. B. was a 13-year-old, seventh-grade student attending class at Smith Middle School in Chapel Hill, North Carolina when he was removed from his classroom by a uniformed police officer, escorted to a closed-door conference room, and questioned by police for at least half an hour. This was the second time that police questioned J. D. B. in the span of a week. Five days earlier, two home break-ins occurred, and various items were stolen. Police stopped and questioned J. D. B. after he was seen behind a residence in the neighborhood where the crimes occurred. That same day, police also spoke to J. D. B.’s grandmother—his legal guardian—as well as his aunt. Police later learned that a digital camera matching the description of one of the stolen items had been found at J. D. B.’s middle school and seen in J. D. B.’s possession. Investigator DiCostanzo, the juvenile investigator with the local police force who had been assigned to the case, went to the school to question J. D. B. Upon arrival, DiCostanzo informed the uniformed police officer on detail to the school (a so-called school resource officer), the assistant principal, and an administrative intern that he was there to question J. D. B. about the break-ins. Although DiCostanzo asked the school administrators to verify J. D. B.’s date of birth, address, and parent contact in-formation from school records, neither the police offi- cers nor the school administrators contacted J. D. B.’s grandmother. The uniformed officer interrupted J. D. B.’s afternoon social studies class, removed J. D. B. from the classroom, and escorted him to a school conference room.[Footnote 1] There, J. D. B. was met by DiCostanzo, the assistant principal, and the administrative intern. The door to the conference room was closed. With the two police officers and the two administrators present, J. D. B. was questioned for the next 30 to 45 minutes. Prior to the commencement of questioning, J. D. B. was given neither Miranda warnings nor the opportunity to speak to his grandmother. Nor was he informed that he was free to leave the room. Questioning began with small talk—discussion of sports and J. D. B.’s family life. DiCostanzo asked, and J. D. B. agreed, to discuss the events of the prior weekend. Denying any wrongdoing, J. D. B. explained that he had been in the neighborhood where the crimes occurred because he was seeking work mowing lawns. DiCostanzo pressed J. D. B. for additional detail about his efforts to obtain work; asked J. D. B. to explain a prior incident, when one of the victims returned home to find J. D. B. behind her house; and confronted J. D. B. with the stolen camera. The assistant principal urged J. D. B. to “do the right thing,” warning J. D. B. that “the truth always comes out in the end.” App. 99a, 112a. Eventually, J. D. B. asked whether he would “still be in trouble” if he returned the “stuff.” Ibid. In response, DiCostanzo explained that return of the stolen items would be helpful, but “this thing is going to court” regardless. Id., at 112a; ibid. (“[W]hat’s done is done[;] now you need to help yourself by making it right”); see also id., at 99a. DiCostanzo then warned that he may need to seek a secure custody order if he believed that J. D. B. would continue to break into other homes. When J. D. B. asked what a secure custody order was, DiCostanzo explained that “it’s where you get sent to juvenile detention before court.” Id., at 112a. After learning of the prospect of juvenile detention, J. D. B. confessed that he and a friend were responsible for the break-ins. DiCostanzo only then informed J. D. B. that he could refuse to answer the investigator’s questions and that he was free to leave.[Footnote 2] Asked whether he understood, J. D. B. nodded and provided further detail, including information about the location of the stolen items. Eventually J. D. B. wrote a statement, at DiCostanzo’s request. When the bell rang indicating the end of the schoolday, J. D. B. was allowed to leave to catch the bus home. B Two juvenile petitions were filed against J. D. B., each alleging one count of breaking and entering and one count of larceny. J. D. B.’s public defender moved to suppress his statements and the evidence derived therefrom, arguing that suppression was necessary because J. D. B. had been “interrogated by police in a custodial setting without being afforded Miranda warning[s],” App. 89a, and because his statements were involuntary under the totality of the circumstances test, id., at 142a; see Schneckloth v. Bustamonte, 412 U. S. 218, 226 (1973) (due process precludes admission of a confession where “a defendant’s will was overborne” by the circumstances of the interrogation). After a suppression hearing at which DiCostanzo and J. D. B. testified, the trial court denied the motion, deciding that J. D. B. was not in custody at the time of the schoolhouse interrogation and that his statements were voluntary. As a result, J. D. B. entered a transcript of admission to all four counts, renewing his objection to the denial of his motion to suppress, and the court adjudicated J. D. B. delinquent. A divided panel of the North Carolina Court of Appeals affirmed. In re J. D. B., 196 N. C. App. 234, 674 S. E. 2d 795 (2009). The North Carolina Supreme Court held, over two dissents, that J. D. B. was not in custody when he confessed, “declin[ing] to extend the test for custody to include consideration of the age … of an individual subjected to questioning by police.” In re J. D. B., 363 N. C. 664, 672, 686 S. E. 2d 135, 140 (2009).[Footnote 3] We granted certiorari to determine whether the Miranda custody analysis includes consideration of a juvenile suspect’s age. 562 U. S. ___ (2010). II A Any police interview of an individual suspected of a crime has “coercive aspects to it.” Oregon v. Mathiason, 429 U. S. 492, 495 (1977) (per curiam). Only those interrogations that occur while a suspect is in police custody, however, “heighte[n] the risk” that statements obtained are not the product of the suspect’s free choice. Dickerson v. United States, 530 U. S. 428, 435 (2000). By its very nature, custodial police interrogation entails “inherently compelling pressures.” Miranda, 384 U. S., at 467. Even for an adult, the physical and psychological isolation of custodial interrogation can “undermine the individual’s will to resist and … compel him to speak where he would not otherwise do so freely.” Ibid. Indeed, the pressure of custodial interrogation is so immense that it “can induce a frighteningly high percentage of people to confess to crimes they never committed.” Corley v. United States, 556 U. S. __, __ (2009) (slip op., at 16) (citing Drizin & Leo, The Problem of False Confessions in the Post-DNA World, 82 N. C. L. Rev. 891, 906–907 (2004)); see also Miranda, 384 U. S., at 455, n. 23. That risk is all the more troubling—and recent studies suggest, all the more acute—when the subject of custodial interrogation is a juvenile. See Brief for Center on Wrongful Convictions of Youth et al. as Amici Curiae 21–22 (collecting empirical studies that “illustrate the heightened risk of false confessions from youth”). Recognizing that the inherently coercive nature of custodial interrogation “blurs the line between voluntary and involuntary statements,” Dickerson, 530 U. S., at 435, this Court in Miranda adopted a set of prophylactic mea-sures designed to safeguard the constitutional guarantee against self-incrimination. Prior to questioning, a suspect “must be warned that he has a right to remain silent, that any statement he does make may be used as evidence against him, and that he has a right to the presence of an attorney, either retained or appointed.” 384 U. S., at 444; see also Florida v. Powell, 559 U. S. ___, ___ (2010) (slip op., at 8) (“The four warnings Miranda requires are invariable, but this Court has not dictated the words in which the essential information must be conveyed”). And, if a suspect makes a statement during custodial interrogation, the burden is on the Government to show, as a “prerequisit[e]” to the statement’s admissibility as evidence in the Government’s case in chief, that the defendant “voluntarily, knowingly and intelligently” waived his rights.[Footnote 4] Miranda, 384 U. S., at 444, 475–476; Dickerson, 530 U. S., at 443–444. Because these measures protect the individual against the coercive nature of custodial interrogation, they are required “ ‘only where there has been such a restriction on a person’s freedom as to render him “in custody.” ’ ” Stansbury v. California, 511 U. S. 318, 322 (1994) (per curiam) (quoting Oregon v. Mathiason, 429 U. S. 492, 495 (1977) (per curiam)). As we have repeatedly emphasized, whether a suspect is “in custody” is an objective inquiry. “Two discrete inquiries are essential to the determination: first, what were the circumstances surrounding the interrogation; and second, given those circumstances, would a reasonable person have felt he or she was at liberty to terminate the interrogation and leave. Once the scene is set and the players’ lines and actions are reconstructed, the court must apply an objective test to resolve the ultimate inquiry: was there a formal arrest or restraint on freedom of movement of the degree associated with formal arrest.” Thompson v. Keohane, 516 U. S. 99, 112 (1995) (internal quotation marks, alteration, and footnote omitted). See also Yarborough v. Alvarado, 541 U. S. 652, 662–663 (2004); Stansbury, 511 U. S., at 323; Berkemer v. McCarty, 468 U. S. 420, 442, and n. 35 (1984). Rather than demarcate a limited set of relevant circumstances, we have required police officers and courts to “examine all of the circumstances surrounding the interrogation,” Stansbury, 511 U. S., at 322, including any circumstance that “would have affected how a reasonable person” in the suspect’s position “would perceive his or her freedom to leave,” id., at 325. On the other hand, the “subjective views harbored by either the interrogating officers or the person being questioned” are irrelevant. Id., at 323. The test, in other words, involves no consideration of the “actual mindset” of the particular suspect subjected to police questioning. Alvarado, 541 U. S., at 667; see also California v. Beheler, 463 U. S. 1121, 1125, n. 3 (1983) (per curiam). The benefit of the objective custody analysis is that it is “designed to give clear guidance to the police.” Alvarado, 541 U. S., at 668. But see Berkemer, 468 U. S., at 441 (recognizing the “occasiona[l] … difficulty” that police and courts nonetheless have in “deciding exactly when a suspect has been taken into custody”). Police must make in-the-moment judgments as to when to administer Miranda warnings. By limiting analysis to the objective circumstances of the interrogation, and asking how a reasonable person in the suspect’s position would understand his freedom to terminate questioning and leave, the objective test avoids burdening police with the task of anticipating the idiosyncrasies of every individual suspect and divining how those particular traits affect each person’s subjective state of mind. See id., at 430–431 (officers are not required to “make guesses” as to circumstances “unknowable” to them at the time); Alvarado, 541 U. S., at 668 (officers are under no duty “to consider … contingent psychological factors when deciding when suspects should be advised of their Miranda rights”). B The State and its amici contend that a child’s age has no place in the custody analysis, no matter how young the child subjected to police questioning. We cannot agree. In some circumstances, a child’s age “would have affected how a reasonable person” in the suspect’s position “would perceive his or her freedom to leave.” Stansbury, 511 U. S., at 325. That is, a reasonable child subjected to police questioning will sometimes feel pressured to submit when a reasonable adult would feel free to go. We think it clear that courts can account for that reality without doing any damage to the objective nature of the custody analysis. A child’s age is far “more than a chronological fact.” Eddings v. Oklahoma, 455 U. S. 104, 115 (1982); accord, Gall v. United States, 552 U. S. 38, 58 (2007); Roper v. Simmons, 543 U. S. 551, 569 (2005); Johnson v. Texas, 509 U. S. 350, 367 (1993). It is a fact that “generates commonsense conclusions about behavior and perception.” Alvarado, 541 U. S., at 674 (Breyer, J., dissenting). Such conclusions apply broadly to children as a class. And, they are self-evident to anyone who was a child once himself, including any police officer or judge. Time and again, this Court has drawn these commonsense conclusions for itself. We have observed that children “generally are less mature and responsible than adults,” Eddings, 455 U. S., at 115–116; that they “often lack the experience, perspective, and judgment to recognize and avoid choices that could be detrimental to them,” Bellotti v. Baird, 443 U. S. 622, 635 (1979) (plurality opinion); that they “are more vulnerable or susceptible to … outside pressures” than adults, Roper, 543 U. S., at 569; and so on. See Graham v. Florida, 560 U. S. ___, ___ (2010) (slip op., at 17) (finding no reason to “reconsider” these observations about the common “nature of juveniles”). Addressing the specific context of police interrogation, we have observed that events that “would leave a man cold and unimpressed can overawe and overwhelm a lad in his early teens.” Haley v. Ohio, 332 U. S. 596, 599 (1948) (plurality opinion); see also Gallegos v. Colorado, 370 U. S. 49, 54 (1962) (“[N]o matter how sophisticated,” a juvenile subject of police interrogation “cannot be compared” to an adult subject). Describing no one child in particular, these observations restate what “any parent knows”—indeed, what any person knows—about children generally. Roper, 543 U. S., at 569.[Footnote 5] Our various statements to this effect are far from unique. The law has historically reflected the same assumption that children characteristically lack the capacity to exercise mature judgment and possess only an incomplete ability to understand the world around them. See, e.g., 1 W. Blackstone, Commentaries on the Laws of England *464–*465 (hereinafter Blackstone) (explaining that limits on children’s legal capacity under the common law “secure them from hurting themselves by their own improvident acts”). Like this Court’s own generalizations, the legal disqualifications placed on children as a class—e.g., limitations on their ability to alienate property, enter a binding contract enforceable against them, and marry without parental consent—exhibit the settled understanding that the differentiating characteristics of youth are universal.[Footnote 6] Indeed, even where a “reasonable person” standard otherwise applies, the common law has reflected the reality that children are not adults. In negligence suits, for instance, where liability turns on what an objectively reasonable person would do in the circumstances, “[a]ll American jurisdictions accept the idea that a person’s childhood is a relevant circumstance” to be considered. Restatement (Third) of Torts §10, Comment b, p. 117 (2005); see also id., Reporters’ Note, pp. 121–122 (collecting cases); Restatement (Second) of Torts §283A, Comment b, p. 15 (1963–1964) (“[T]here is a wide basis of community experience upon which it is possible, as a practical matter, to determine what is to be expected of [children]”). As this discussion establishes, “[o]ur history is replete with laws and judicial recognition” that children cannot be viewed simply as miniature adults. Eddings, 455 U. S., at 115–116. We see no justification for taking a different course here. So long as the child’s age was known to the officer at the time of the interview, or would have been objectively apparent to any reasonable officer, including age as part of the custody analysis requires officers neither to consider circumstances “unknowable” to them, Berkemer, 468 U. S., at 430, nor to “anticipat[e] the frailties or idiosyncrasies” of the particular suspect whom they question, Alvarado, 541 U. S., at 662 (internal quotation marks omitted). The same “wide basis of community experience” that makes it possible, as an objective matter, “to determine what is to be expected” of children in other contexts, Restatement (Second) of Torts §283A, at 15; see supra, at 10, and n. 6, likewise makes it possible to know what to expect of children subjected to police questioning. In other words, a child’s age differs from other personal characteristics that, even when known to police, have no objectively discernible relationship to a reasonable person’s understanding of his freedom of action. Alvarado, holds, for instance, that a suspect’s prior interrogation history with law enforcement has no role to play in the custody analysis because such experience could just as easily lead a reasonable person to feel free to walk away as to feel compelled to stay in place. 541 U. S., at 668. Because the effect in any given case would be “contingent [on the] psycholog[y]” of the individual suspect, the Court explained, such experience cannot be considered without compromising the objective nature of the custody analysis. Ibid. A child’s age, however, is different. Precisely because childhood yields objective conclusions like those we have drawn ourselves—among others, that children are “most susceptible to influence,” Eddings, 455 U. S., at 115, and “outside pressures,” Roper, 543 U. S., at 569—considering age in the custody analysis in no way involves a determination of how youth “subjectively affect[s] the mindset” of any particular child, Brief for Respondent 14.[Footnote 7] In fact, in many cases involving juvenile suspects, the custody analysis would be nonsensical absent some consideration of the suspect’s age. This case is a prime example. Were the court precluded from taking J. D. B.’s youth into account, it would be forced to evaluate the circumstances present here through the eyes of a reasonable person of average years. In other words, how would a reasonable adult understand his situation, after being removed from a seventh-grade social studies class by a uniformed school resource officer; being encouraged by his assistant principal to “do the right thing”; and being warned by a police investigator of the prospect of juvenile detention and separation from his guardian and primary caretaker? To describe such an inquiry is to demonstrate its absurdity. Neither officers nor courts can reasonably evaluate the effect of objective circumstances that, by their nature, are specific to children without accounting for the age of the child subjected to those circumstances. Indeed, although the dissent suggests that concerns “regarding the application of the Miranda custody rule to minors can be accommodated by considering the unique circumstances present when minors are questioned in school,” post, at 17 (opinion of Alito, J.), the effect of the schoolhouse setting cannot be disentangled from the identity of the person questioned. A student—whose presence at school is compulsory and whose disobedience at school is cause for disciplinary action—is in a far different position than, say, a parent volunteer on school grounds to chaperone an event, or an adult from the community on school grounds to attend a basketball game. Without asking whether the person “questioned in school” is a “minor,” ibid., the coercive effect of the schoolhouse setting is unknowable. Our prior decision in Alvarado in no way undermines these conclusions. In that case, we held that a state-court decision that failed to mention a 17-year-old’s age as part of the Miranda custody analysis was not objectively unreasonable under the deferential standard of review set forth by the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), 110 Stat. 1214. Like the North Carolina Supreme Court here, see 363 N. C., at 672, 686 S. E. 2d, at 140, we observed that accounting for a juvenile’s age in the Miranda custody analysis “could be viewed as creating a subjective inquiry,” 541 U. S., at 668. We said nothing, however, of whether such a view would be correct under the law. Cf. Renico v. Lett, 559 U. S. ___, ____, n. 3 (2010) (slip op., at 11, n. 3) (“[W]hether the [state court] was right or wrong is not the pertinent question under AEDPA”). To the contrary, Justice O’Connor’s concurring opinion explained that a suspect’s age may indeed “be relevant to the ‘custody’ inquiry.” Alvarado, 541 U. S., at 669. Reviewing the question de novo today, we hold that so long as the child’s age was known to the officer at the time of police questioning, or would have been objectively apparent to a reasonable officer, its inclusion in the custody analysis is consistent with the objective nature of that test.[Footnote 8] This is not to say that a child’s age will be a determinative, or even a significant, factor in every case. Cf. ibid. (O’Connor, J., concurring) (explaining that a state-court decision omitting any mention of the defendant’s age was not unreasonable under AEDPA’s deferential standard of review where the defendant “was almost 18 years old at the time of his interview”); post, at 17 (suggesting that “teenagers nearing the age of majority” are likely to react to an interrogation as would a “typical 18-year-old in similar circumstances”). It is, however, a reality that courts cannot simply ignore. III The State and its amici offer numerous reasons that courts must blind themselves to a juvenile defendant’s age. None is persuasive. To start, the State contends that a child’s age must be excluded from the custody inquiry because age is a personal characteristic specific to the suspect himself rather than an “external” circumstance of the interrogation. Brief for Respondent 21; see also id., at 18–19 (distinguishing “personal characteristics” from “objective facts related to the interrogation itself” such as the location and duration of the interrogation). Despite the supposed significance of this distinction, however, at oral argument counsel for the State suggested without hesitation that at least some undeniably personal characteristics—for instance, whether the individual being questioned is blind—are circumstances relevant to the custody analysis. See Tr. of Oral Arg. 41. Thus, the State’s quarrel cannot be that age is a personal characteristic, without more.[Footnote 9] The State further argues that age is irrelevant to the custody analysis because it “go[es] to how a suspect may internalize and perceive the circumstances of an interrogation.” Brief for Respondent 12; see also Brief for United States as Amicus Curiae 21 (hereinafter U. S. Brief) (arguing that a child’s age has no place in the custody analysis because it goes to whether a suspect is “particularly susceptible” to the external circumstances of the interrogation (some internal quotation marks omitted)). But the same can be said of every objective circumstance that the State agrees is relevant to the custody analysis: Each circumstance goes to how a reasonable person would “internalize and perceive” every other. See, e.g., Stansbury, 511 U. S., at 325. Indeed, this is the very reason that we ask whether the objective circumstances “add up to custody,” Keohane, 516 U. S., at 113, instead of evaluating the circumstances one by one. In the same vein, the State and its amici protest that the “effect of … age on [the] perception of custody is internal,” Brief for Respondent 20, or “psychological,” U. S. Brief 21. But the whole point of the custody analysis is to determine whether, given the circumstances, “a reasonable person [would] have felt he or she was … at liberty to terminate the interrogation and leave.” Keohane, 516 U. S., at 112. Because the Miranda custody inquiry turns on the mindset of a reasonable person in the suspect’s position, it cannot be the case that a circumstance is subjective simply because it has an “internal” or “psychological” impact on perception. Were that so, there would be no objective circumstances to consider at all. Relying on our statements that the objective custody test is “designed to give clear guidance to the police,” Alvarado, 541 U. S., at 668, the State next argues that a child’s age must be excluded from the analysis in order to preserve clarity. Similarly, the dissent insists that the clarity of the custody analysis will be destroyed unless a “one-size-fits-all reasonable-person test” applies. Post, at 13. In reality, however, ignoring a juvenile defendant’s age will often make the inquiry more artificial, see supra, at 12–13, and thus only add confusion. And in any event, a child’s age, when known or apparent, is hardly an obscure factor to assess. Though the State and the dissent worry about gradations among children of different ages, that concern cannot justify ignoring a child’s age altogether. Just as police officers are competent to account for other objective circumstances that are a matter of degree such as the length of questioning or the number of officers present, so too are they competent to evaluate the effect of relative age. Indeed, they are competent to do so even though an interrogation room lacks the “reflective atmosphere of a [jury] deliberation room,” post, at 15. The same is true of judges, including those whose childhoods have long since passed, see post, at 14. In short, officers and judges need no imaginative powers, knowledge of developmental psychology, training in cognitive science, or expertise in social and cultural anthropology to account for a child’s age. They simply need the common sense to know that a 7-year-old is not a 13-year-old and neither is an adult. There is, however, an even more fundamental flaw with the State’s plea for clarity and the dissent’s singular focus on simplifying the analysis: Not once have we excluded from the custody analysis a circumstance that we determined was relevant and objective, simply to make the fault line between custodial and noncustodial “brighter.” Indeed, were the guiding concern clarity and nothing else, the custody test would presumably ask only whether the suspect had been placed under formal arrest. Berkemer, 468 U. S., at 441; see ibid. (acknowledging the “occasiona[l] … difficulty” police officers confront in determining when a suspect has been taken into custody). But we have rejected that “more easily administered line,” recognizing that it would simply “enable the police to circumvent the constraints on custodial interrogations established by Miranda.” Ibid.; see also ibid., n. 33.[Footnote 10] Finally, the State and the dissent suggest that excluding age from the custody analysis comes at no cost to juveniles’ constitutional rights because the due process voluntariness test independently accounts for a child’s youth. To be sure, that test permits consideration of a child’s age, and it erects its own barrier to admission of a defendant’s inculpatory statements at trial. See Gallegos, 370 U. S., at 53–55; Haley, 332 U. S., at 599–601; see also post, at 17–18 (“[C]ourts should be instructed to take particular care to ensure that [young children’s] incriminating statements were not obtained involuntarily”). But Miranda’s procedural safeguards exist precisely because the voluntariness test is an inadequate barrier when custodial interrogation is at stake. See 384 U. S., at 458 (“Unless adequate protective devices are employed to dispel the compulsion inherent in custodial surroundings, no statement obtained from the defendant can truly be the product of his free choice”); Dickerson, 530 U. S., at 442 (“[R]eliance on the traditional totality-of-the-circumstances test raise[s] a risk of overlooking an involuntary custodial confession”); see also supra, at 5–6. To hold, as the State requests, that a child’s age is never relevant to whether a suspect has been taken into custody—and thus to ignore the very real differences between children and adults—would be to deny children the full scope of the procedural safeguards that Miranda guarantees to adults. * * * The question remains whether J. D. B. was in custody when police interrogated him. We remand for the state courts to address that question, this time taking account of all of the relevant circumstances of the interrogation, including J. D. B.’s age at the time. The judgment of the North Carolina Supreme Court is reversed, and the case is remanded for proceedings not inconsistent with this opinion. It is so ordered. Footnote 1 Although the State suggests that the “record is unclear as to who brought J. D. B. to the conference room, and the trial court made no factual findings on this specific point,” Brief for Respondent 3, n. 1, the State agreed at the certiorari stage that “the SRO [school resource officer] escorted petitioner” to the room, Brief in Opposition 3. Footnote 2 The North Carolina Supreme Court noted that the trial court’s factual findings were “uncontested and therefore . . . binding” on it. In re J. D. B., 363 N. C. 664, 668, 686 S. E. 2d 135, 137 (2009). The court described the sequence of events set forth in the text. See id., at 670–671, 686 S. E. 2d, at 139. (“Immediately following J. D. B.’s initial confession, Investigator DiCostanzo informed J. D. B. that he did not have to speak with him and that he was free to leave” (internal quotation marks and alterations omitted)). Though less than perfectly explicit, the trial court’s order indicates a finding that J. D. B. initially confessed prior to DiCostanzo’s warnings. See App. 99a. Nonetheless, both parties’ submissions to this Court suggest that the warnings came after DiCostanzo raised the possibility of a secure custody order but before J. D. B. confessed for the first time. See Brief for Petitioner 5; Brief for Respondent 5. Because we remand for a determination whether J. D. B. was in custody under the proper analysis, the state courts remain free to revisit whether the trial court made a conclusive finding of fact in this respect. Footnote 3 J. D. B.’s challenge in the North Carolina Supreme Court focused on the lower courts’ conclusion that he was not in custody for purposes of Miranda v. Arizona, 384 U. S. 436 (1966). The North Carolina Supreme Court did not address the trial court’s holding that the statements were voluntary, and that question is not before us. Footnote 4 Amici on behalf of J. D. B. question whether children of all ages can comprehend Miranda warnings and suggest that additional procedural safeguards may be necessary to protect their Miranda rights. Brief for Juvenile Law Center et al. as Amici Curiae 13–14, n. 7. Whatever the merit of that contention, it has no relevance here, where no Miranda warnings were administered at all. Footnote 5 Although citation to social science and cognitive science authorities is unnecessary to establish these commonsense propositions, the literature confirms what experience bears out. See, e.g., Graham v. Florida, 560 U. S. ___, ___ (2010) (slip op., at 17) (“[D]evelopments in psychology and brain science continue to show fundamental differences between juvenile and adult minds”). Footnote 6 See, e.g., 1 E. Farnsworth, Contracts §4.4, p. 379, and n. 1 (1990) (“Common law courts early announced the prevailing view that a minor’s contract is ‘voidable’ at the instance of the minor” (citing 8 W. Holdsworth, History of English Law 51 (1926))); 1 D. Kramer, Legal Rights of Children §8.1, p. 663 (rev. 2d ed. 2005) (“[W]hile minor children have the right to acquire and own property, they are considered incapable of property management” (footnote omitted)); 2 J. Kent, Commentaries on American Law *78–*79, *90 (G. Comstock ed., 11th ed. 1867); see generally id., at *233 (explaining that, under the common law, “[t]he necessity of guardians results from the inability of infants to take care of themselves . . . and this inability continues, in contemplation of law, until the infant has attained the age of [21]”); 1 Blackstone *465 (“It is generally true, that an infant can neither aliene his lands, nor do any legal act, nor make a deed, nor indeed any manner of contract, that will bind him”); Roper v. Simmons, 543 U. S. 551, 569 (2005) (“In recognition of the comparative immaturity and irresponsibility of juveniles, almost every State prohibits those under 18 years of age from voting, serving on juries, or marrying without parental consent”). Footnote 7 Thus, contrary to the dissent’s protestations, today’s holding neither invites consideration of whether a particular suspect is “unusually meek or compliant,” post, at 9 (opinion of Alito, J.), nor “expan[ds]” the Miranda custody analysis, post, at 8, into a test that requires officers to anticipate and account for a suspect’s every personal characteristic, see post, at 11–12. Footnote 8 This approach does not undermine the basic principle that an interrogating officer’s unarticulated, internal thoughts are never—in and of themselves—objective circumstances of an interrogation. See supra, at 7; Stansbury v. California, 511 U. S. 318, 323 (1994) (per curiam). Unlike a child’s youth, an officer’s purely internal thoughts have no conceivable effect on how a reasonable person in the suspect’s position would understand his freedom of action. See id., at 323–325; Berkemer v. McCarty, 468 U. S. 420, 442 (1984). Rather than “overtur[n]” that settled principle, post, at 13, the limitation that a child’s age may inform the custody analysis only when known or knowable simply reflects our unwillingness to require officers to “make guesses” as to circumstances “unknowable” to them in deciding when to give Miranda warnings, Berkemer, 468 U. S., at 430–431. Footnote 9 The State’s purported distinction between blindness and age—that taking account of a suspect’s youth requires a court “to get into the mind” of the child, whereas taking account of a suspect’s blindness does not, Tr. of Oral Arg. 41–42—is mistaken. In either case, the question becomes how a reasonable person would understand the circumstances, either from the perspective of a blind person or, as here, a 13-year-old child. Footnote 10 Contrary to the dissent’s intimation, see post, at 8, Miranda does not answer the question whether a child’s age is an objective circumstance relevant to the custody analysis. Miranda simply holds that warnings must be given once a suspect is in custody, without “paus[ing] to inquire in individual cases whether the defendant was aware of his rights without a warning being given.” 384 U. S., at 468; see also id., at 468–469 (“Assessments of the knowledge the defendant possessed, based on information as to age, education, intelligence, or prior contact with authorities, can never be more than speculation; a warning is a clearcut fact” (footnote omitted)). That conclusion says nothing about whether age properly informs whether a child is in custody in the first place.

564.US.2010_09-525

Respondent First Derivative Traders (First Derivative), representing a class of stockholders in petitioner Janus Capital Group, Inc. (JCG), filed this private action under Securities and Exchange Commission (SEC) Rule 10b–5, which forbids “any person … [t]o make any untrue statement of a material fact” in connection with the purchase or sale of securities. The complaint alleged, inter alia, that JCG and its wholly owned subsidiary, petitioner Janus Capital Management LLC (JCM), made false statements in mutual fund prospectuses filed by Janus Investment Fund—for which JCM was the investment adviser and administrator—and that those statements affected the price of JCG’s stock. Although JCG created Janus Investment Fund, it is a separate legal entity owned entirely by mutual fund investors. The District Court dismissed the complaint for failure to state a claim. The Fourth Circuit reversed, holding that First Derivative had sufficiently alleged that JCG and JCM, by participating in the writing and dissemination of the prospectuses, made the misleading statements contained in the documents. Before this Court, First Derivative continues to argue that JCM made the statements but seeks to hold JCG liable only as a control person of JCM under §20(a). Held: Because the false statements included in the prospectuses were made by Janus Investment Fund, not by JCM, JCM and JCG cannot be held liable in a private action under Rule 10b–5. Pp. 5–12. (a) Although neither Rule 10b–5 nor the statute it interprets, §10(b) of the Act, expressly creates a private right of action, such an “action is implied under §10(b).” Superintendent of Ins. of N. Y. v. Bankers Life & Casualty Co., 404 U. S. 6, 13, n. 9. That holding “remains the law,” Stoneridge Investment Partners, LLC v. Scientific-Atlanta, Inc., 552 U. S. 148, 165, but, in analyzing the question at issue, the Court is mindful that it must give “narrow dimensions … to a right … Congress did not authorize when it first enacted the statute and did not expand when it revisited” it, id., at 167. Pp. 5–10. (1) For Rule 10b–5 purposes, the maker of a statement is the person or entity with ultimate authority over the statement, including its content and whether and how to communicate it. Without control, a person or entity can merely suggest what to say, not “make” a statement in its own right. This rule follows from Central Bank of Denver, N. A. v. First Interstate Bank of Denver, N. A., 511 U. S. 164, 180, which held that Rule 10b–5’s private right of action does not include suits against aiders and abettors who contribute “substantial assistance” to the making of a statement but do not actually make it. Reading “make” more broadly, to include persons or entities lacking ultimate control over a statement, would substantially undermine Central Bank by rendering aiders and abettors almost nonexistent. The Court’s interpretation is also suggested by Stoneridge, 552 U. S., at 161, and accords with the narrow scope that must be given the implied private right of action, id., at 167. Pp. 6–8. (2) The Court rejects the Government’s contention that “make” should be defined as “create,” thereby allowing private plaintiffs to sue a person who provides the false or misleading information that another person puts into a statement. Adopting that definition would be inconsistent with Stoneridge, supra, at 161, which rejected a private Rule 10b–5 suit against companies involved in deceptive transactions, even when information about those transactions was later incorporated into false public statements. First Derivative notes the uniquely close relationship between a mutual fund and its investment adviser, but the corporate formalities were observed, and reapportionment of liability in light of this close relationship is properly the responsibility of Congress, not the courts. Furthermore, First Derivative’s rule would read into Rule 10b–5 a theory of liability similar to—but broader than—control-person liability under §20(a). Pp. 8–10. (b) Although JCM may have been significantly involved in preparing the prospectuses, it did not itself “make” the statements at issue for Rule 10b–5 purposes. Its assistance in crafting what was said was subject to Janus Investment Fund’s ultimate control. Pp. 10–12. 566 F. 3d 111, reversed. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, and Alito, JJ., joined. Breyer, J., filed a dissenting opinion, in which Ginsburg, Sotomayor, and Kagan, JJ., joined.

This case requires us to determine whether Janus Capital Management LLC (JCM), a mutual fund investment adviser, can be held liable in a private action under Securities and Exchange Commission (SEC) Rule 10b–5 for false statements included in its client mutual funds’ prospectuses. Rule 10b–5 prohibits “mak[ing] any untrue statement of a material fact” in connection with the purchase or sale of securities. 17 CFR §240.10b–5 (2010). We conclude that JCM cannot be held liable because it did not make the statements in the prospectuses. I Janus Capital Group, Inc. (JCG), is a publicly traded company that created the Janus family of mutual funds. These mutual funds are organized in a Massachusetts business trust, the Janus Investment Fund. Janus Investment Fund retained JCG’s wholly owned subsidiary, JCM, to be its investment adviser and administrator. JCG and JCM are the petitioners here. Although JCG created Janus Investment Fund, Janus Investment Fund is a separate legal entity owned entirely by mutual fund investors. Janus Investment Fund has no assets apart from those owned by the investors. JCM provides Janus Investment Fund with investment advisory services, which include “the management and administrative services necessary for the operation of [Janus] Fun[d],” App. 225a, but the two entities maintain legal independence. At all times relevant to this case, all of the officers of Janus Investment Fund were also officers of JCM, but only one member of Janus Investment Fund’s board of trustees was associated with JCM. This is more independence than is required: By statute, up to 60 percent of the board of a mutual fund may be composed of “interested persons.” See 54 Stat. 806, as amended, 15 U. S. C. §80a–10(a); see also 15 U. S. C. A. §80a–2(a)(19) (2009 ed. and Feb. 2011 Supp.) (defining “interested person”). As the securities laws require, Janus Investment Fund issued prospectuses describing the investment strategy and operations of its mutual funds to investors. See 15 U. S. C. §§77b(a)(10), 77e(b)(2), 80a–8(b), 80a–2(a)(31), 80a–29(a)–(b). The prospectuses for several funds represented that the funds were not suitable for market timing and can be read to suggest that JCM would implement policies to curb the practice.[Footnote 1] For example, the Janus Mercury Fund prospectus dated February 25, 2002, stated that the fund was “not intended for market timing or excessive trading” and represented that it “may reject any purchase request … if it believes that any combination of trading activity is attributable to market timing or is otherwise excessive or potentially disruptive to the Fund.” App. 141a. Although market timing is legal, it harms other investors in the mutual fund. In September 2003, the Attorney General of the State of New York filed a complaint against JCG and JCM alleging that JCG entered into secret arrangements to permit market timing in several funds run by JCM. After the complaint’s allegations became public, investors withdrew significant amounts of money from the Janus Investment Fund mutual funds.[Footnote 2] Because Janus Investment Fund compensated JCM based on the total value of the funds and JCM’s management fees comprised a significant percentage of JCG’s income, Janus Investment Fund’s loss of value affected JCG’s value as well. JCG’s stock price fell nearly 25 percent, from $17.68 on September 2 to $13.50 on September 26. Respondent First Derivative Traders (First Derivative) represents a class of plaintiffs who owned JCG stock as of September 3, 2003. Its complaint asserts claims against JCG and JCM for violations of Rule 10b–5 and §10(b) of the Securities Exchange Act of 1934, 48 Stat. 891, as amended, 15 U. S. C. §78j(b). First Derivative alleges that JCG and JCM “caused mutual fund prospectuses to be issued for Janus mutual funds and made them available to the investing public, which created the misleading impression that [JCG and JCM] would implement measures to curb market timing in the Janus [mutual funds].” App. to Pet. for Cert. 60a. “Had the truth been known, Janus [mutual funds] would have been less attractive to investors, and consequently, [JCG] would have realized lower revenues, so [JCG’s] stock would have traded at lower prices.” Id., at 72a. First Derivative contends that JCG and JCM “materially misled the investing public” and that class members relied “upon the integrity of the market price of [JCG] securities and market information relating to [JCG and JCM].” Id., at 109a. The complaint also alleges that JCG should be held liable for the acts of JCM as a “controlling person” under 15 U. S. C. A. §78t(a) (Feb. 2011 Supp.) (§20(a) of the Act). The District Court dismissed the complaint for failure to state a claim.[Footnote 3] In re Mutual Funds Inv. Litigation, 487 F. Supp. 2d 618, 620 (D Md. 2007). The Court of Appeals for the Fourth Circuit reversed, holding that First Derivative had sufficiently alleged that “JCG and JCM, by participating in the writing and dissemination of the prospectuses, made the misleading statements contained in the documents.” In re Mutual Funds Inv. Litigation, 566 F. 3d 111, 121 (2009) (emphasis in original). With respect to the element of reliance, the court found that investors would infer that JCM “played a role in preparing or approving the content of the Janus fund prospectuses,” id., at 127, but that investors would not infer the same about JCG, which could be liable only as a “control person” of JCM under §20(a). Id., at 128, 129–130. II We granted certiorari to address whether JCM can be held liable in a private action under Rule 10b–5 for false statements included in Janus Investment Fund’s pro-spectuses. 561 U. S. ___ (2010). Under Rule 10b–5, it is unlawful for “any person, directly or indirectly, … [t]o make any untrue statement of a material fact” in connection with the purchase or sale of securities. 17 CFR §240.10b–5(b).[Footnote 4] To be liable, therefore, JCM must have “made” the material misstatements in the prospectuses. We hold that it did not.[Footnote 5] A The SEC promulgated Rule 10b–5 pursuant to authority granted under §10(b) of the Securities Exchange Act of 1934, 15 U. S. C. §78j(b). Although neither Rule 10b–5 nor §10(b) expressly creates a private right of action, this Court has held that “a private right of action is implied under §10(b).” Superintendent of Ins. of N. Y. v. Bankers Life & Casualty Co., 404 U. S. 6, 13, n. 9 (1971). That holding “remains the law,” Stoneridge Investment Partners, LLC v. Scientific-Atlanta, Inc., 552 U. S. 148, 165 (2008), but “[c]oncerns with the judicial creation of a private cause of action caution against its expansion,” ibid. Thus, in analyzing whether JCM “made” the statements for purposes of Rule 10b–5, we are mindful that we must give “narrow dimensions … to a right of action Congress did not authorize when it first enacted the statute and did not expand when it revisited the law.” Id., at 167. 1 One “makes” a statement by stating it. When “make” is paired with a noun expressing the action of a verb, the resulting phrase is “approximately equivalent in sense” to that verb. 6 Oxford English Dictionary 66 (def. 59) (1933) (hereinafter OED); accord, Webster’s New International Dictionary 1485 (def. 43) (2d ed. 1934) (“Make followed by a noun with the indefinite article is often nearly equivalent to the verb intransitive corresponding to that noun”). For instance, “to make a proclamation” is the approximate equivalent of “to proclaim,” and “to make a promise” approximates “to promise.” See 6 OED 66 (def. 59). The phrase at issue in Rule 10b–5, “[t]o make any … statement,” is thus the approximate equivalent of “to state.” For purposes of Rule 10b–5, the maker of a statement is the person or entity with ultimate authority over the statement, including its content and whether and how to communicate it. Without control, a person or entity can merely suggest what to say, not “make” a statement in its own right. One who prepares or publishes a statement on behalf of another is not its maker. And in the ordinary case, attribution within a statement or implicit from surrounding circumstances is strong evidence that a statement was made by—and only by—the party to whom it is attributed. This rule might best be exemplified by the relationship between a speechwriter and a speaker. Even when a speechwriter drafts a speech, the content is entirely within the control of the person who delivers it. And it is the speaker who takes credit—or blame—for what is ultimately said. This rule follows from Central Bank of Denver, N. A. v. First Interstate Bank of Denver, N. A., 511 U. S. 164 (1994), in which we held that Rule 10b–5’s private right of action does not include suits against aiders and abettors. See id., at 180. Such suits—against entities that contribute “substantial assistance” to the making of a statement but do not actually make it—may be brought by the SEC, see 15 U. S. C. A. §78t(e), but not by private parties. A broader reading of “make,” including persons or entities without ultimate control over the content of a statement, would substantially undermine Central Bank. If persons or entities without control over the content of a statement could be considered primary violators who “made” the statement, then aiders and abettors would be almost nonexistent.[Footnote 6] This interpretation is further supported by our recent decision in Stoneridge. There, investors sued “entities who, acting both as customers and suppliers, agreed to arrangements that allowed the investors’ company to mislead its auditor and issue a misleading financial statement.” 552 U. S., at 152–153. We held that dismissal of the complaint was proper because the public could not have relied on the entities’ undisclosed deceptive acts. Id., at 166–167. Significantly, in reaching that conclusion we emphasized that “nothing [the defendants] did made it necessary or inevitable for [the company] to record the transactions as it did.” Id., at 161.[Footnote 7] This emphasis suggests the rule we adopt today: that the maker of a statement is the entity with authority over the content of the statement and whether and how to communicate it. Without such authority, it is not “necessary or inevitable” that any falsehood will be contained in the statement. Our holding also accords with the narrow scope that we must give the implied private right of action. Id., at 167. Although the existence of the private right is now settled, we will not expand liability beyond the person or entity that ultimately has authority over a false statement. 2 The Government contends that “make” should be defined as “create.” Brief for United States as Amicus Curiae 14–15 (citing Webster’s New International Dictionary 1485 (2d ed. 1958) (defining “make” as “[t]o cause to exist, appear, or occur”)). This definition, although perhaps appropriate when “make” is directed at an object unassociated with a verb (e.g., “to make a chair”), fails to capture its meaning when directed at an object expressing the action of a verb. Adopting the Government’s definition of “make” would also lead to results inconsistent with our precedent. The Government’s definition would permit private plaintiffs to sue a person who “provides the false or misleading information that another person then puts into the statement.” Brief for United States as Amicus Curiae 13.[Footnote 8] But in Stoneridge, we rejected a private Rule 10b–5 suit against companies involved in deceptive transactions, even when information about those transactions was later incorporated into false public statements. 552 U. S., at 161. We see no reason to treat participating in the drafting of a false statement differently from engaging in deceptive transactions, when each is merely an undisclosed act preceding the decision of an independent entity to make a public statement. For its part, First Derivative suggests that the “well-recognized and uniquely close relationship between a mutual fund and its investment adviser” should inform our decision. Brief for Respondent 21. It suggests that an investment adviser should generally be understood to be the “maker” of statements by its client mutual fund, like a playwright whose lines are delivered by an actor. We decline this invitation to disregard the corporate form. Although First Derivative and its amici persuasively argue that investment advisers exercise significant influence over their client funds, see Jones v. Harris Associates L. P., 559 U. S. ___, ___ (2010) (slip op., at 1–2), it is undisputed that the corporate formalities were observed here. JCM and Janus Investment Fund remain legally separate entities, and Janus Investment Fund’s board of trustees was more independent than the statute requires. 15 U. S. C. §80a–10.[Footnote 9] Any reapportionment of liability in the securities industry in light of the close relationship between investment advisers and mutual funds is properly the responsibility of Congress and not the courts. Moreover, just as with the Government’s theory, First Derivative’s rule would create the broad liability that we rejected in Stoneridge. Congress also has established liability in §20(a) for “[e]very person who, directly or indirectly, controls any person liable” for violations of the securities laws. 15 U. S. C. A. §78t(a). First Derivative’s theory of liability based on a relationship of influence resembles the liability imposed by Congress for control. To adopt First Derivative’s theory would read into Rule 10b–5 a theory of liability similar to—but broader in application than, see post, at 9—what Congress has already created expressly elsewhere.[Footnote 10] We decline to do so. B Under this rule, JCM did not “make” any of the statements in the Janus Investment Fund prospectuses; Janus Investment Fund did. Only Janus Investment Fund—not JCM—bears the statutory obligation to file the prospectuses with the SEC. 15 U. S. C. §§77e(b)(2), 80a–8(b), 80a–29(a)–(b); see also 17 CFR §230.497 (imposing requirements on “investment companies”). The SEC has recorded that Janus Investment Fund filed the prospectuses. See JIF Group1 Standalone Prospectuses (Feb. 25, 2002), online at http://www.sec.gov/Archives/edgar/data/ 277751 / 000027775102000049 / 0000277751-02-000049.txt (as visited June 10, 2011, and available in Clerk of Court’s case file) (recording the “Filer” of the Janus Mercury Fund prospectus as “Janus Investment Fund”). There is no allegation that JCM in fact filed the prospectuses and falsely attributed them to Janus Investment Fund. Nor did anything on the face of the prospectuses indicate that any statements therein came from JCM rather than Janus Investment Fund—a legally independent entity with its own board of trustees.[Footnote 11] First Derivative suggests that both JCM and Janus Investment Fund might have “made” the misleading statements within the meaning of Rule 10b–5 because JCM was significantly involved in preparing the prospectuses. But this assistance, subject to the ultimate control of Janus Investment Fund, does not mean that JCM “made” any statements in the prospectuses. Although JCM, like a speechwriter, may have assisted Janus Investment Fund with crafting what Janus Investment Fund said in the prospectuses, JCM itself did not “make” those statements for purposes of Rule 10b–5.[Footnote 12] * * * The statements in the Janus Investment Fund prospectuses were made by Janus Investment Fund, not by JCM. Accordingly, First Derivative has not stated a claim against JCM under Rule 10b–5. The judgment of the United States Court of Appeals for the Fourth Circuit is reversed. It is so ordered. Footnote 1 Market timing is a trading strategy that exploits time delay in mutual funds’ daily valuation system. The price for buying or selling shares of a mutual fund is ordinarily determined by the next net asset value (NAV) calculation after the order is placed. The NAV calculation usually happens once a day, at the close of the major U. S. markets. Because of certain time delays, however, the values used in these calculations do not always accurately reflect the true value of the underlying assets. For example, a fund may value its foreign securities based on the price at the close of the foreign market, which may have occurred several hours before the calculation. But events might have taken place after the close of the foreign market that could be expected to affect their price. If the event were expected to increase the price of the foreign securities, a market-timing investor could buy shares of a mutual fund at the artificially low NAV and sell the next day when the NAV corrects itself upward. See Disclosure Regarding Market Timing and Selective Disclosure of Portfolio Holdings, 68 Fed. Reg. 70402 (proposed Dec. 17, 2003). Footnote 2 In 2004, JCG and JCM settled these allegations and agreed to reduce their fees by $125 million and pay $50 million in civil penalties and $50 million in disgorgement to the mutual fund investors. Footnote 3 The elements of a private action under Rule 10b–5 are “(1) a material misrepresentation or omission by the defendant; (2) scienter; (3) a connection between the misrepresentation or omission and the purchase or sale of a security; (4) reliance upon the misrepresentation or omission; (5) economic loss; and (6) loss causation.” Stoneridge Investment Partners, LLC v. Scientific-Atlanta, Inc., 552 U. S. 148, 157 (2008). Footnote 4 Rule 10b–5 makes it “unlawful for any person, directly or indirectly, by the use of any means or instrumentality of interstate commerce, or of the mails or of any facility of any national securities exchange, … [t]o make any untrue statement of a material fact or to omit to state a material fact necessary in order to make the statements made, in the light of the circumstances under which they were made, not misleading … .” 17 CFR §240.10b–5(b). Footnote 5 Although First Derivative argued below that JCG violated Rule 10b–5 by making the statements in the prospectuses, it now seeks to hold JCG liable solely as a control person of JCM under §20(a). The only question we must answer, therefore, is whether JCM made the misstatements. Whether First Derivative has stated a claim against JCG as a control person depends on whether it has stated a claim against JCM. Footnote 6 The dissent correctly notes that Central Bank involved secondary, not primary, liability. Post, at 4 (opinion of Breyer, J.). But for Central Bank to have any meaning, there must be some distinction between those who are primarily liable (and thus may be pursued in private suits) and those who are secondarily liable (and thus may not be pursued in private suits). We draw a clean line between the two—the maker is the person or entity with ultimate authority over a statement and others are not. In contrast, the dissent’s only limit on primary liability is not much of a limit at all. It would allow for primary liability whenever “[t]he specific relationships alleged … warrant [that] conclusion”—whatever that may mean. Post, at 11. Footnote 7 We agree that “no one in Stoneridge contended that the equipment suppliers were, in fact, the makers of the cable company’s misstatements.” Post, at 8. If Stoneridge had addressed whether the equipment suppliers were “makers,” today’s decision would be unnecessary. The point is that Stoneridge’s analysis suggests that they were not. Footnote 8 Because we do not find the meaning of “make” in Rule 10b–5 to be ambiguous, we need not consider the Government’s assertion that we should defer to the SEC’s interpretation of the word elsewhere. Brief for United States as Amicus Curiae 13 (citing Brief for SEC as Amicus Curiae in Pacific Inv. Mgmt. Co. LLC v. Mayer Brown LLP, No. 09–1619 (CA2), p. 7); see Christensen v. Harris County, 529 U. S. 576, 588 (2000). We note, however, that we have previously expressed skepticism over the degree to which the SEC should receive deference regarding the private right of action. See Piper v. Chris-Craft Industries, Inc., 430 U. S. 1, 41, n. 27 (1977) (noting that the SEC’s presumed expertise “is of limited value” when analyzing “whether a cause of action should be implied by judicial interpretation in favor of a particular class of litigants”). This also is not the first time this Court has disagreed with the SEC’s broad view of §10(b) or Rule 10b–5. See, e.g., Central Bank of Denver, N. A. v. First Interstate Bank of Denver, N. A., 511 U. S. 164, 188–191 (1994); Dirks v. SEC, 463 U. S. 646, 666, n. 27 (1983); Ernst & Ernst v. Hochfelder, 425 U. S. 185, 207 (1976); Blue Chip Stamps v. Manor Drug Stores, 421 U. S. 723, 746, n. 10 (1975). Footnote 9 Nor does First Derivative contend that any statements made by JCM to Janus Investment Fund were “public statements” for the purposes of Basic Inc. v. Levinson, 485 U. S. 224, 227–228 (1988). We do not address whether and in what circumstances statements would qualify as “public.” Cf. post, at 12–13 (citing cases involving liability for statements made to analysts); In re Aetna, Inc. Securities Litigation, 617 F. 3d 272, 275–277 (CA3 2010) (involving allegations that defendants “publicly tout[ed]” falsities on analyst conference calls). Footnote 10 We do not address whether Congress created liability for entities that act through innocent intermediaries in 15 U. S. C. A. §78t(b). See Tr. of Oral Arg. 6, 61. Footnote 11 First Derivative suggests that “indirectly” in Rule 10b–5 may broaden the meaning of “make.” We disagree. The phrase “directly or indirectly” is set off by itself in Rule 10b–5 and modifies not just “to make,” but also “to employ” and “to engage.” We think the phrase merely clarifies that as long as a statement is made, it does not matter whether the statement was communicated directly or indirectly to the recipient. A different understanding of “indirectly” would, like a broad definition of “make,” threaten to erase the line between primary violators and aiders and abettors established by Central Bank. In this case, we need not define precisely what it means to communicate a “made” statement indirectly because none of the statements in the prospectuses were attributed, explicitly or implicitly, to JCM. Without attribution, there is no indication that Janus Investment Fund was quoting or otherwise repeating a statement originally “made” by JCM. Cf. Anixter v. Home-Stake Production Co., 77 F. 3d 1215, 1220, and n. 4 (CA10 1996) (quoting a signed “ ‘Auditor’s Report’ ” included in a pro-spectus); Basic, supra, at 227, n. 4 (quoting a news item reporting a state-ment by Basic’s president). More may be required to find that a person or entity made a statement indirectly, but attribution is necessary. Footnote 12 That JCM provided access to Janus Investment Fund’s prospectuses on its Web site is also not a basis for liability. Merely hosting a document on a Web site does not indicate that the hosting entity adopts the document as its own statement or exercises control over its content. Cf. United States v. Ware, 577 F. 3d 442, 448 (CA2 2009) (involving the issuance of false press releases through innocent companies). In doing so, we do not think JCM made any of the statements in Janus Investment Fund’s prospectuses for purposes of Rule 10b–5 liability, just as we do not think that the SEC “makes” the statements in the many prospectuses available on its Web site.

563.US.2010_09-834

Petitioner Kasten brought an antiretaliation suit against his former employer, respondent (Saint-Gobain), under the Fair Labor Standards Act of 1938 (Act), which provides minimum wage, maximum hour, and overtime pay rules; and which forbids employers “to discharge … any employee because such employee has filed any complaint” alleging a violation of the Act, 29 U. S. C. §215(a)(3). In a related suit, the District Court found that Saint-Gobain violated the Act by placing timeclocks in a location that prevented workers from receiving credit for the time they spent donning and doffing work-related protective gear. In this suit Kasten claims that he was discharged because he orally complained to company officials about the timeclocks. The District Court granted Saint-Gobain summary judgment, concluding that the Act’s antiretaliation provision did not cover oral complaints. The Seventh Circuit affirmed. Held: The scope of statutory term “filed any complaint” includes oral, as well as written, complaints. Pp. 4–15. (a) The interpretation of the statutory phrase “depends upon reading the whole statutory text, considering the [statute’s] purpose and context … , and consulting any precedents or authorities that inform the analysis.” Dolan v. Postal Service, 546 U. S. 481, 486. The text, taken alone, cannot provide a conclusive answer here. Some dictionary definitions of “filed” contemplate a writing while others permit using “file” in conjunction with oral material. In addition to dictionary definitions, state statutes and federal regulations sometimes contemplate oral filings, and contemporaneous judicial usage shows that oral filings were a known phenomenon at the time of the Act’s passage. Even if “filed,” considered alone, might suggest a narrow interpretation limited to writings, “any complaint” suggests a broad interpretation that would include an oral complaint. Thus, the three-word phrase, taken by itself, cannot answer the interpretive question. The Act’s other references to “filed” also do not resolve the linguistic question. Some of those provisions involve filed material that is virtually always in writing; others specifically require a writing, and the remainder, like the provision here, leave the oral/written question unresolved. Since “filed any complaint” lends itself linguistically to the broader, “oral” interpretation, the use of broader language in other statutes’ antiretaliation provisions does not indicate whether Congress did or did not intend to leave oral grievances unprotected here. Because the text, taken alone, might, or might not, encompass oral complaints, the Court must look further. Pp. 4–8. (b) Several functional considerations indicate that Congress intended the antiretaliation provision to cover oral, as well as written, complaints. Pp. 8–14. (1) A narrow interpretation would undermine the Act’s basic objective, which is to prohibit “labor conditions detrimental to the maintenance of the minimum standard of living necessary for health, efficiency, and general well-being of workers,” 29 U. S. C. §202(a). The Act relies for enforcement of its substantive standards on “information and complaints received from employees,” Mitchell v. Robert DeMario Jewelry, Inc., 361 U. S. 288, 292, and its antiretaliation provision makes the enforcement scheme effective by preventing “fear of economic retaliation” from inducing workers “quietly to accept substandard conditions,” ibid. Why would Congress want to limit the enforcement scheme’s effectiveness by inhibiting use of the Act’s complaint procedure by those who would find it difficult to reduce their complaints to writing, particularly the illiterate, less educated, or overworked workers who were most in need of the Act’s help at the time of passage? Limiting the provision’s scope to written complaints could prevent Government agencies from using hotlines, interviews, and other oral methods to receive complaints. And insofar as the provision covers complaints made to employers, a limiting reading would discourage using informal workplace grievance procedures to secure compliance with the Act. The National Labor Relations Act’s antiretaliation provision has been broadly interpreted as protecting workers who simply “participate[d] in a [National Labor Relations] Board investigation.” NLRB v. Scrivener, 405 U. S. 117, 123. The similar enforcement needs of this related statute argue for a broad interpretation of “complaint.” The Act’s requirement that an employer receive fair notice of an employee’s complaint can be met by oral, as well as written, complaints. Pp. 8–12. (2) Given the delegation of enforcement powers to federal administrative agencies, their views about the meaning of the phrase should be given a degree of weight. The Secretary of Labor has consistently held the view that “filed any complaint” covers both oral and written complaints. The Equal Employment Opportunity Commission has set out a similar view in its Compliance Manual and in multiple briefs. These views are reasonable and consistent with the Act. And the length of time they have been held suggests that they reflect careful consideration, not “post hoc rationalizatio[n].” Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U. S. 29, 50. Pp. 12–13. (3) After engaging in traditional statutory interpretation methods, the statute does not remain sufficiently ambiguous to warrant application of the rule of lenity. Pp. 13–14. (c) This Court will not consider Saint-Gobain’s alternative claim that the antiretaliation provision applies only to complaints filed with the Government, since that claim was not raised in the certiorari briefs and since its resolution is not a “ ‘predicate to an intelligent resolution’ ” of the oral/written question at issue, Caterpillar Inc. v. Lewis, 519 U. S. 61, 75, n. 13. Pp. 14–15. 570 F. 3d 834, vacated and remanded. Breyer, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Ginsburg, Alito, and Sotomayor, JJ., joined. Scalia, J., filed a dissenting opinion, in which Thomas, J., joined as to all but n. 6. Kagan, J., took no part in the consideration or decision of the case.

The Fair Labor Standards Act of 1938 (Act) sets forth employment rules concerning minimum wages, maximum hours, and overtime pay. 52 Stat. 1060, 29 U. S. C. §201 et seq. The Act contains an antiretaliation provision that forbids employers “to discharge or in any other manner discriminate against any employee because such employee has filed any complaint or instituted or caused to be instituted any proceeding under or related to [the Act], or has testified or is about to testify in such proceeding, or has served or is about to serve on an industry committee.” §215(a)(3) (emphasis added). We must decide whether the statutory term “filed any complaint” includes oral as well as written complaints within its scope. We conclude that it does. I The petitioner, Kevin Kasten, brought this antiretaliation lawsuit against his former employer, Saint-Gobain Performance Plastics Corporation. Kasten says that Saint-Gobain located its timeclocks between the area where Kasten and other workers put on (and take off) their work-related protective gear and the area where they carry out their assigned tasks. That location prevented workers from receiving credit for the time they spent putting on and taking off their work clothes—contrary to the Act’s requirements. In a related suit the District Court agreed with Kasten, finding that Saint-Gobain’s “practice of not compensating … for time spent donning and doffing certain required protective gear and walking to work areas” violated the Act. Kasten v. Saint-Gobain Performance Plastics Corp., 556 F. Supp. 2d 941, 954 (WD Wis. 2008). In this suit Kasten claims unlawful retaliation. He says that Saint-Gobain discharged him because he orally complained to Saint-Gobain officials about the timeclocks. In particular, Kasten says that he repeatedly called the unlawful timeclock location to Saint-Gobain’s attention—in accordance with Saint-Gobain’s internal grievance-resolution procedure. See Brief for Petitioner 4 (quoting Saint-Gobain’s Code of Ethics and Business Conduct as imposing upon every employee “the responsibility to report … suspected violations of … any applicable law of which he or she becomes aware”); id., at 4–5 (quoting Saint-Gobain’s Employee Policy Handbook as instructing employees with “questions, complaints, and problems” to “[c]ontact” their “supervisor[s] immediately” and if necessary “take the issue to the next level of management,” then to the “local Human Resources Manager,” then to “Human Resources” personnel at the “Regional” or “Headquarters” level). Kasten adds that he “raised a concern” with his shift supervisor that “it was illegal for the time clocks to be where they were” because of Saint-Gobain’s exclusion of “the time you come in and start doing stuff”; he told a human resources employee that “if they were to get challenged on” the location in court, “they would lose”; he told his lead operator that the location was illegal and that he “was thinking about starting a lawsuit about the placement of the time clocks”; and he told the human resources manager and the operations manager that he thought the location was illegal and that the company would “lose” in court. Record in No. 3:07–cv–00686–bbc (WD Wis.), Doc. 87–3, pp. 31–34 (deposition of Kevin Kasten). This activity, Kasten concludes, led the company to discipline him and, in December 2006, to dismiss him. Saint-Gobain presents a different version of events. It denies that Kasten made any significant complaint about the timeclock location. And it says that it dismissed Kasten simply because Kasten, after being repeatedly warned, failed to record his comings and goings on the timeclock. For present purposes we accept Kasten’s version of these contested events as valid. See Scott v. Harris, 550 U. S. 372, 380 (2007). That is because the District Court entered summary judgment in Saint-Gobain’s favor. 619 F. Supp. 2d 608, 610 (WD Wis. 2008). And it did so, not because it doubted Kasten’s ability to prove the facts he alleged, but because it thought the Act did not protect oral complaints. Id., at 611–613. On appeal, the Seventh Circuit agreed with the District Court that the Act’s antiretaliation provision does not cover oral complaints. 570 F. 3d 834, 838–840 (2009). Kasten sought certiorari. And in light of conflict among the Circuits as to whether an oral complaint is protected, we granted Kasten’s petition. Compare Hagan v. Echostar Satellite, L. L. C., 529 F. 3d 617, 625–626 (CA5 2008) (antiretaliation provision covers oral complaints); Lambert v. Ackerley, 180 F. 3d 997, 1007 (CA9 1999) (en banc) (same); with Lambert v. Genesee Hospital, 10 F. 3d 46, 55–56 (CA2 1993) (antiretaliation provision does not cover informal complaints to supervisors). See also Pacheco v. Whiting Farms, Inc., 365 F. 3d 1199, 1206 (CA10 2004) (antiretaliation provision covers unofficial assertion of rights); EEOC v. White & Son Enterprises, 881 F. 2d 1006, 1011–1012 (CA11 1989) (same); Moore v. Freeman, 355 F. 3d 558, 562–563 (CA6 2004) (assuming without discussion that oral complaints are covered); Brennan v. Maxey’s Yamaha, Inc., 513 F. 2d 179, 181 (CA8 1975) (same). II The sole question presented is whether “an oral complaint of a violation of the Fair Labor Standards Act” is “protected conduct under the [Act’s] anti-retaliation provision.” Pet. for Cert. i. The Act protects employees who have “filed any complaint,” 29 U. S. C. §215(a)(3), and interpretation of this phrase “depends upon reading the whole statutory text, considering the purpose and context of the statute, and consulting any precedents or authorities that inform the analysis,” Dolan v. Postal Service, 546 U. S. 481, 486 (2006). This analysis leads us to conclude that the language of the provision, considered in isolation, may be open to competing interpretations. But considering the provision in conjunction with the purpose and context leads us to conclude that only one interpretation is permissible. A We begin with the text of the statute. The word “filed” has different relevant meanings in different contexts. Some dictionary definitions of the word contemplate a writing. See, e.g., Webster’s New International Dictionary 945 (2d ed. 1934) (def. 4(a)) (to file is to “deliver (a paper or instrument) to the proper officer so that it is received by him to be kept on file, or among the records of his office” (emphasis added)); Webster’s Ninth New Collegiate Dictionary 462 (1983) (def. 2(a)) (one definition of “file” is “to place among official records as prescribed by law”). But other dictionaries provide different definitions that permit the use of the word “file” in conjunction with oral material. One can, for example, file an oral statement that enters a matter “into the order of business.” 1 Funk & Wagnalls New Standard Dictionary of the English Language 920 (rev. ed. 1938) (def. 2) (to file is to “present in the regular way, as to a judicial or legislative body, so that it shall go upon the records or into the order of business”). This possibility is significant because it means that dictionary meanings, even if considered alone, do not necessarily limit the scope of the statutory phrase to written complaints. Cf. Crawford v. Metropolitan Government of Nashville and Davidson Cty., 555 U. S. ___, ___ (2009) (slip op., at 5) (looking for the “limits” of a linguistic phrase rather than what “exemplif[ies]” its application). In addition to the dictionary definitions, we have found that legislators, administrators, and judges have all sometimes used the word “file” in conjunction with oral statements. Thus state statutes sometimes contemplate oral filings. See, e.g., Alaska Stat. §47.32.090(a) (2008) (“file a verbal or written complaint”); Cal. Health & Safety Code Ann. §17055(a) (West 2006) (“file an administra- tive complaint orally or in writing”); D. C. Code §7–1231.12(a)(2)(B) (2001) (“filing his or her grievance, orally or in writing”); Ga. Code Ann. §§31–8–124(a), (c), 31–8–134(b) (2009) (“to file a grievance,” a person may “submit an oral or written complaint”); Ind. Code §27–8–28–14(a) (2009) (“file a grievance orally or in writing”); Me. Rev. Stat. Ann., Tit. 34–B, §5604(3)(B) (2009) (“filed through an oral request”); Miss. Code Ann. §69–47–23(4) (2005) (“file a written or oral complaint”); Mo. Rev. Stat. §198.088.3(3) (2009) (to have a complaint “filed,” a person “shall write or cause to be written his grievance or shall state it orally”); Nev. Rev. Stat. §§618.336(2)(a), 618.341(1)(a) (2009) (“oral or written complaint filed”); N. J. Stat. Ann. §30:4C–12 (West 2008) (“written or oral complaint may be filed”); N. Y. Ins. Law Ann. §§3217–a(a)(7), 4324(a)(7) (West 2006) (“file a grievance orally”); N. Y. Pub. Health Law Ann. §§4408(1)(g) (West Supp. 2010) (“file a grievance orally”); Pa. Stat. Ann., Tit. 40, §§991.2141(a)–(b) (Purdon 1999) (“file a … written or oral complaint”); Tex. Ins. Code Ann. §§1305.401(a)–(b) (West 2009) (“oral or written complaint” must be “file[d]”); Wash. Rev. Code §§90.64.030(3), (5) (2008) (“complaints have been filed … as the result of either an oral or a written complaint”). Regulations promulgated by various federal agencies sometimes permit complaints to be filed orally. See, e.g., 32 CFR §842.20 (2010) (“[f]iling a claim” may proceed “orally or in writing”); 42 CFR §422.564(d)(1) (2009) (“file a grievance … either orally or in writing”); §423.564(d)(1) (same); §438.402(b)(3)(i) (“file a grievance either orally or in writing”); §494.180(e) (“file an oral or written grievance”); 49 CFR §1503.629(c) (2009) (“[f]iling of motions … must be in writing or orally on the record” (italics omitted)); 42 CFR §438.402(b)(3)(ii) (2009) (“file an appeal either orally or in writing”). And a review of contemporaneous judicial usage, cf. Utah v. Evans, 536 U. S. 452, 475 (2002), shows that oral filings were a known phenomenon when the Act was passed. See, e.g., Reed Oil Co. v. Cain, 169 Ark. 309, 312, 275 S. W. 333, 334 (1925) (“appellee filed … an oral complaint”); Tingler v. Lahti, 87 W. Va. 499, 503, 105 S. E. 810, 812 (1921) (“complaint subsequently filed, either oral or written”); Ex parte Mosgrove, 47 Okla. Cr. 40, 287 P. 795 (1930) (only “complaint … filed against him” was “oral complaint of the town marshal”); Indian Fred v. State, 36 Ariz. 48, 52–53, 282 P. 930, 932 (1929) (“filed an oral motion to quash”); Dunn v. State, 60 Okla. Cr. 201, 203, 63 P. 2d 772, 773 (1936) (“filed an oral demurrer”); Morrison v. Lewis, 58 Ga. App. 677, 199 S. E. 782 (1938) (“filed an oral motion” demurring); Brock v. Cullum Bros., 263 S. W. 335 (Tex. Civ. App. 1924) (“filed an oral motion to quash”); Fike v. Allen, 269 S. W. 179, 180 (Tex. Civ. App. 1925) (“filed oral pleadings”). Filings may more often be made in writing. See, e.g., Ritter v. United States, 28 F. 2d 265, 267 (CA3 1928) (finding words “file a claim for refund” to require a written request in context of tax code). But we are interested in the filing of “any complaint.” So even if the word “filed,” considered alone, might suggest a narrow interpretation limited to writings, the phrase “any complaint” suggests a broad interpretation that would include an oral complaint. See, e.g., Republic of Iraq v. Beaty, 556 U. S. ___, ___ (2009) (slip op., at 7). The upshot is that the three-word phrase, taken by itself, cannot answer the interpretive question. We can look further to other appearances of the word “filed” in the Act. See MCI Telecommunications Corp. v. American Telephone & Telegraph Co., 512 U. S. 218, 226 (1994) (examining “contextual indications” of the meaning of a term). That word (or a variant) appears in numerous other provisions. But its appearance elsewhere in the Act does not resolve the linguistic question before us. Some of those other provisions (1) involve filed material that, unlike a complaint, is of a kind that is virtually always in writing. See, e.g., 29 U. S. C. §203(l) (employers must “have on file an unexpired certificate” (emphasis added)); §210(a) (Secretary must “file in the court the record of the industry committee” (emphasis added)); ibid. (industry committee must “file” its findings and recommendations). Others (2) specifically require a writing, see, e.g., §214(c)(5)(A) (requiring employee’s “consent in writing” to join collective action to be “filed” (emphasis added)); §216(b) (same). And the remainder (3) leave the oral/ written question unresolved—just as does the provision before us. See, e.g., §210(b) (prohibiting a stay unless movant “file[s] in court an undertaking” (emphasis added)); §214(c)(5)(A) (employee “may file … a petition” for review of a special wage rate (emphasis added)). Looking beyond the Act, we find other statutes that contain antiretaliation provisions. Those statutes, however, use somewhat different language. See, e.g., §158(a)(4) (protecting an employee who has “filed charges or given testimony”); §623(d) (protecting those who “opposed any [unlawful] practice” (emphasis added)); 42 U. S. C. §§2000e–3(a), 12203(a) (same); 29 U. S. C. §2615(a)(2) (similar). See also, e.g., 15 U. S. C. §2087(a)(1) (2006 ed., Supp. III) (“provided … to the employer … information relating to any violation” (emphasis added)); §2651(a) (similar); 30 U. S. C. §815(c)(1) (“filed or made a complaint” (emphasis added)); 42 U. S. C. §5851(a)(1)(A) (“notified his employer” (emphasis added)); 49 U. S. C. §42121(a)(1) (“provided … information” (emphasis added)); §60129(a)(1) (same). Some of this language is broader than the phrase before us, but, given the fact that the phrase before us lends itself linguistically to the broader, “oral” interpretation, the use of broader language elsewhere may mean (1) that Congress wanted to limit the scope of the phrase before us to writings, or (2) that Congress did not believe the different phraseology made a significant difference in this respect. The language alone does not tell us whether Congress, if intending to protect orally expressed grievances elsewhere, did or did not intend to leave those oral grievances unprotected here. The bottom line is that the text, taken alone, cannot provide a conclusive answer to our interpretive question. The phrase “filed any complaint” might, or might not, encompass oral complaints. We must look further. B 1 Several functional considerations indicate that Congress intended the antiretaliation provision to cover oral, as well as written, “complaint[s].” First, an interpretation that limited the provision’s coverage to written complaints would undermine the Act’s basic objectives. The Act seeks to prohibit “labor conditions detrimental to the maintenance of the minimum standard of living necessary for health, efficiency, and general well-being of workers.” 29 U. S. C. §202(a). It does so in part by setting forth substantive wage, hour, and overtime standards. It relies for enforcement of these standards, not upon “continuing detailed federal supervision or inspection of payrolls,” but upon “information and complaints received from employees seeking to vindicate rights claimed to have been denied.” Mitchell v. Robert DeMario Jewelry, Inc., 361 U. S. 288, 292 (1960). And its antiretaliation provision makes this enforcement scheme effective by preventing “fear of economic retaliation” from inducing workers “quietly to accept substandard conditions.” Ibid. Why would Congress want to limit the enforcement scheme’s effectiveness by inhibiting use of the Act’s complaint procedure by those who would find it difficult to reduce their complaints to writing, particularly illiterate, less educated, or overworked workers? President Franklin Roosevelt pointed out at the time that these were the workers most in need of the Act’s help. See Message to Congress, May 24, 1937, H. R. Doc. No. 255, 75th Cong., 1st Sess., 4 (seeking a bill to help the poorest of “those who toil in factory”). In the years prior to the passage of the Act, illiteracy rates were particularly high among the poor. See E. Gordon & E. Gordon, Literacy in America 273 (2003) (one-quarter of World War I conscripts were illiterate); Dept. of Commerce, Bureau of Census, Sixteenth Census of the United States, 1940, Population: The Labor Force (Sample Statistics): Occupational Characteristics 60 (1943) (20.8% of manufacturing laborers in 1940 had less than five years of schooling). Those rates remained high in certain industries for many years after the Act’s passage. In 1948, for example, the National War Labor Board wrote: “In many plants where there is a high degree of illiteracy, the writing of grievances by employees works a substantial hardship. In other plants where there is considerable dirt and special clothes must be worn, it is often not practicable to write up grievances during work hours.” 1 The Termination Report of the National War Labor Board, p. 122. To limit the scope of the antiretaliation provision to the filing of written complaints would also take needed flexibility from those charged with the Act’s enforcement. It could prevent Government agencies from using hotlines, interviews, and other oral methods of receiving complaints. And insofar as the antiretaliation provision covers complaints made to employers (a matter we need not decide, see infra, at 14–15), it would discourage the use of desirable informal workplace grievance procedures to secure compliance with the Act. Cf. Burlington Industries, Inc. v. Ellerth, 524 U. S. 742, 764 (1998) (reading Title VII to encourage the development of effective grievance procedures to deter misconduct); D. McPherson, C. Gates, & K. Rogers, Resolving Grievances: A Practical Approach 38–40 (1983) (describing the significant benefits of unwritten complaints). Given the need for effective enforcement of the National Labor Relations Act (NLRA), this Court has broadly interpreted the language of the NLRA’s antiretaliation provision—“filed charges or given testimony,” 29 U. S. C. §158(a)(4)—as protecting workers who neither filed charges nor were “called formally to testify” but simply “participate[d] in a [National Labor Relations] Board investigation.” NLRB v. Scrivener, 405 U. S. 117, 123 (1972) (emphasis added). The similar enforcement needs of this related statute argue for an interpretation of the word “complaint” that would provide “broad rather than narrow protection to the employee,” id., at 122 (and would do so here without pressing statutory language to its limit). See also Tennessee Coal, Iron & R. Co. v. Muscoda Local No. 123, 321 U. S. 590, 597 (1944) (the Act’s “remedial and humanitarian … purpose” cautions against “narrow, grudging” interpretations of its language). Saint-Gobain replies that worker protection is not the only relevant statutory objective. The Act also seeks to establish an enforcement system that is fair to employers. To do so, the employer must have fair notice that an employee is making a complaint that could subject the employer to a later claim of retaliation. If oral complaints suffice, Saint-Gobain adds, employers too often will be left in a state of uncertainty about whether an employee (particularly an employee who seems unusually angry at the moment) is in fact making a complaint about an Act violation or just letting off steam. We agree with Saint-Gobain that the statute requires fair notice. Although the dictionary definitions, statutes, regulations, and judicial opinions we considered, see supra, at 4–7, do not distinguish between writings and oral statements, they do suggest that a “filing” is a serious occasion, rather than a triviality. As such, the phrase “filed any complaint” contemplates some degree of formality, certainly to the point where the recipient has been given fair notice that a grievance has been lodged and does, or should, reasonably understand the matter as part of its business concerns. Moreover, the statute prohibits employers from discriminating against an employee “because such employee has filed any complaint.” §215(a)(3) (emphasis added). And it is difficult to see how an employer who does not (or should not) know an employee has made a complaint could discriminate because of that complaint. But we also believe that a fair notice requirement does not necessarily mean that notice must be in writing. At oral argument, the Government said that a complaint is “filed” when “a reasonable, objective person would have understood the employee” to have “put the employer on notice that [the] employee is asserting statutory rights under the [Act].” Tr. of Oral Arg. 23, 26. We agree. To fall within the scope of the antiretaliation provision, a complaint must be sufficiently clear and detailed for a reasonable employer to understand it, in light of both content and context, as an assertion of rights protected by the statute and a call for their protection. This standard can be met, however, by oral complaints, as well as by written ones. 2 Second, given Congress’ delegation of enforcement powers to federal administrative agencies, we also give a degree of weight to their views about the meaning of this enforcement language. See 29 U. S. C. §216(c) (vesting enforcement power in Secretary of Labor); Reorganization Plan No. 1 of 1978, 5 U. S. C. App. §1, p. 664 (transferring to Equal Employment Opportunity Commission (EEOC) enforcement of this antiretaliation provision as part of its Equal Pay Act enforcement responsibilities); Skidmore v. Swift & Co., 323 U. S. 134, 140 (1944) (giving weight to a persuasive articulation of views within an agency’s area of expertise). The Secretary of Labor has consistently held the view that the words “filed any complaint” cover oral, as well as written, complaints. The Department of Labor articulated that view in an enforcement action filed many years ago, Goldberg v. Zenger, 43 CCH LC ¶31,155, pp. 40,985, 40,986 (D Utah 1961). It has subsequently reaffirmed that view in briefs. See, e.g., Brief for Secretary of Labor as Amicus Curiae Supporting Petition for Rehearing with Suggestion for Rehearing En Banc in Lambert v. Ackerley, No. 96–36017 etc. (CA9), pp. 6–7. And more recently it has acted in accordance with that view by creating a hotline to receive oral complaints, see Dept. of Labor, Compliance Assistance By Law—The Fair Labor Standards Act (FLSA), http://www.dol.gov/compliance/laws/comp-flsa.htm (as visited Mar. 18, 2011, and available in Clerk of Court’s case file) (directing participants who wish to “file a complaint” to contact a local office “or call the Department’s Toll-Free Wage and Hour Help Line at 1–866–4–US–WAGE”). The EEOC has set forth a similar view in its Compliance Manual, Vol. 2, §8–II(B)(1), p. 8–3, and n. 12 (1998), and in multiple briefs, see, e.g., Brief for EEOC as Amicus Curiae in Support of Petition for Rehearing with Suggestion for Rehearing En Banc in Lambert v. Ackerley, No. 96–36017 etc. (CA9), pp. 8–13; Brief for Appellee in EEOC v. White & Son Enterprises, Inc., No. 88–7658 (CA11), pp. 29–30. These agency views are reasonable. They are consistent with the Act. The length of time the agencies have held them suggests that they reflect careful consideration, not “post hoc rationalizatio[n].” Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U. S. 29, 50 (1983). And they consequently add force to our conclusion. Skidmore, supra, at 140; cf. United States v. Mead Corp., 533 U. S. 218, 229, 234–235 (2001) (Court sometimes finds judicial deference intended even in absence of rulemaking authority); Babbitt v. Sweet Home Chapter, Communities for Great Ore., 515 U. S. 687, 703–704, and n. 18 (1995) (agency views, where the law counsels deference, can clarify otherwise ambiguous statutory provisions). 3 Finally, we note that Saint-Gobain invokes the “rule of lenity” in support of its “written complaint” interpretation. That rule applies primarily to the interpretation of criminal statutes. It leads us to favor a more lenient interpretation of a criminal statute “when, after consulting traditional canons of statutory construction, we are left with an ambiguous statute.” United States v. Shabani, 513 U. S. 10, 17 (1994). We agree with Saint-Gobain that those who violate the antiretaliation provision before us are subject to criminal sanction, 29 U. S. C. §216(a). And we have said that the rule of lenity can apply when a statute with criminal sanctions is applied in a noncriminal context. See Leocal v. Ashcroft, 543 U. S. 1, 11, n. 8 (2004). But after engaging in traditional methods of statutory interpretation, we cannot find that the statute remains sufficiently ambiguous to warrant application of the rule of lenity here. C Alternatively, Saint-Gobain claims that it should prevail because Kasten complained to a private employer, not to the Government; and, in Saint-Gobain’s view, the antiretaliation provision applies only to complaints filed with the Government. Saint-Gobain advanced this claim in the lower courts, which held to the contrary. 570 F. 3d, at 837–838; 619 F. Supp. 2d, at 613. But Saint-Gobain said nothing about it in response to Kasten’s petition for certiorari. Indeed, it did not mention the claim in this Court until it filed its brief on the merits. We do not normally consider a separate legal question not raised in the certiorari briefs. See this Court’s Rule 15.2; Caterpillar Inc. v. Lewis, 519 U. S. 61, 75, n. 13 (1996). We see no reason to make an exception here. Resolution of the Government/private employer question is not a “ ‘ “predicate to an intelligent resolution” ’ ” of the oral/written question that we granted certiorari to decide. See ibid. (quoting Ohio v. Robinette, 519 U. S. 33, 38 (1996)). That is to say, we can decide the oral/written question separately—on its own. And we have done so. Thus, we state no view on the merits of Saint-Gobain’s alternative claim. Cf. post, at 1–5 (Scalia, J., dissenting). * * * We conclude that the Seventh Circuit erred in determining that oral complaints cannot fall within the scope of the phrase “filed any complaint” in the Act’s antiretaliation provision. We leave it to the lower courts to decide whether Kasten will be able to satisfy the Act’s notice requirement. We vacate the Circuit’s judgment and remand the case for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

563.US.452

Police officers in Lexington, Kentucky, followed a suspected drug dealer to an apartment complex. They smelled marijuana outside an apartment door, knocked loudly, and announced their presence. As soon as the officers began knocking, they heard noises coming from the apartment; the officers believed that these noises were consistent with the destruction of evidence. The officers announced their intent to enter the apartment, kicked in the door, and found respondent and others. They saw drugs in plain view during a protective sweep of the apartment and found additional evidence during a subsequent search. The Circuit Court denied respondent’s motion to suppress the evidence, holding that exigent circumstances—the need to prevent destruction of evidence—justified the warrantless entry. Respondent entered a conditional guilty plea, reserving his right to appeal the suppression ruling, and the Kentucky Court of Appeals affirmed. The Supreme Court of Kentucky reversed. The court assumed that exigent circumstances existed, but it nonetheless invalidated the search. The exigent circumstances rule did not apply, the court held, because the police should have foreseen that their conduct would prompt the occupants to attempt to destroy evidence. Held: 1. The exigent circumstances rule applies when the police do not create the exigency by engaging or threatening to engage in conduct that violates the Fourth Amendment. Pp. 5–16. (a) The Fourth Amendment expressly imposes two requirements: All searches and seizures must be reasonable; and a warrant may not be issued unless probable cause is properly established and the scope of the authorized search is set out with particularity. Although “ ‘searches and seizures inside a home without a warrant are presumptively unreasonable,’ ” Brigham City v. Stuart, 547 U. S. 398, 403, this presumption may be overcome when “ ‘the exigencies of the situation’ make the needs of law enforcement so compelling that [a] warrantless search is objectively reasonable under the Fourth Amendment,” Mincey v. Arizona, 437 U. S. 385, 394. One such exigency is the need “to prevent the imminent destruction of evidence.” Brigham City, supra, at 403. Pp. 5–6. (b) Under the “police-created exigency” doctrine, which lower courts have developed as an exception to the exigent circumstances rule, exigent circumstances do not justify a warrantless search when the exigency was “created” or “manufactured” by the conduct of the police. The lower courts have not agreed, however, on the test for determining when police impermissibly create an exigency. Pp. 7–8. (c) The proper test follows from the principle that permits warrantless searches: warrantless searches are allowed when the circumstances make it reasonable, within the meaning of the Fourth Amendment, to dispense with the warrant requirement. Thus, a warrantless entry based on exigent circumstances is reasonable when the police did not create the exigency by engaging or threatening to engage in conduct violating the Fourth Amendment. A similar approach has been taken in other cases involving warrantless searches. For example, officers may seize evidence in plain view if they have not violated the Fourth Amendment in arriving at the spot from which the observation of the evidence is made, see Horton v. California, 496 U. S. 128, 136–140; and they may seek consent-based encounters if they are lawfully present in the place where the consensual encounter occurs, see INS v. Delgado, 466 U. S. 210, 217, n. 5. Pp. 8–10. (d) Some courts, including the Kentucky Supreme Court, have imposed additional requirements—asking whether officers “ ‘deliberately created the exigent circumstances with the bad faith intent to avoid the warrant requirement,’ ” 302 S. W. 3d 649, 656 (case below); reasoning that police may not rely on an exigency if “ ‘it was reasonably foreseeable that [their] investigative tactics … would create the exigent circumstances,’ ”ibid.; faulting officers for knocking on a door when they had sufficient evidence to seek a warrant but did not do so; and finding that officers created or manufactured an exigency when their investigation was contrary to standard or good law enforcement practices. Such requirements are unsound and are thus rejected. Pp. 10–14. (e) Respondent contends that an exigency is impermissibly created when officers engage in conduct that would cause a reasonable person to believe that entry was imminent and inevitable, but that approach is also flawed. The ability of officers to respond to an exigency cannot turn on such subtleties as the officers’ tone of voice in announcing their presence and the forcefulness of their knocks. A forceful knock may be necessary to alert the occupants that someone is at the door, and unless officers identify themselves loudly enough, occupants may not know who is at their doorstep. Respondent’s test would make it extremely difficult for officers to know how loudly they may announce their presence or how forcefully they may knock without running afoul of the police-created exigency rule. And in most cases, it would be nearly impossible for a court to determine whether that threshold had been passed. Pp. 14–15. 2. Assuming that an exigency existed here, there is no evidence that the officers either violated the Fourth Amendment or threatened to do so prior to the point when they entered the apartment. Pp. 16–19. (a) Any question about whether an exigency existed here is better addressed by the Kentucky Supreme Court on remand. P. 17. (b) Assuming an exigency did exist, the officers’ conduct—banging on the door and announcing their presence—was entirely consistent with the Fourth Amendment. Respondent has pointed to no evidence supporting his argument that the officers made any sort of “demand” to enter the apartment, much less a demand that amounts to a threat to violate the Fourth Amendment. If there is contradictory evidence that has not been brought to this Court’s attention, the state court may elect to address that matter on remand. Finally, the record makes clear that the officers’ announcement that they were going to enter the apartment was made after the exigency arose. Pp. 17–19. 302 S. W. 3d 649, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Thomas, Breyer, Sotomayor, and Kagan, JJ., joined. Ginsburg, J., filed a dissenting opinion.

It is well established that “exigent circumstances,” including the need to prevent the destruction of evidence, permit police officers to conduct an otherwise permissible search without first obtaining a warrant. In this case, we consider whether this rule applies when police, by knocking on the door of a residence and announcing their presence, cause the occupants to attempt to destroy evidence. The Kentucky Supreme Court held that the exigent circumstances rule does not apply in the case at hand because the police should have foreseen that their conduct would prompt the occupants to attempt to destroy evidence. We reject this interpretation of the exigent circumstances rule. The conduct of the police prior to their entry into the apartment was entirely lawful. They did not violate the Fourth Amendment or threaten to do so. In such a situation, the exigent circumstances rule applies. I A This case concerns the search of an apartment in Lexington, Kentucky. Police officers set up a controlled buy of crack cocaine outside an apartment complex. Undercover Officer Gibbons watched the deal take place from an unmarked car in a nearby parking lot. After the deal occurred, Gibbons radioed uniformed officers to move in on the suspect. He told the officers that the suspect was moving quickly toward the breezeway of an apartment building, and he urged them to “hurry up and get there” before the suspect entered an apartment. App. 20. In response to the radio alert, the uniformed officers drove into the nearby parking lot, left their vehicles, and ran to the breezeway. Just as they entered the breezeway, they heard a door shut and detected a very strong odor of burnt marijuana. At the end of the breezeway, the officers saw two apartments, one on the left and one on the right, and they did not know which apartment the suspect had entered. Gibbons had radioed that the suspect was running into the apartment on the right, but the officers did not hear this statement because they had already left their vehicles. Because they smelled marijuana smoke emanating from the apartment on the left, they approached the door of that apartment. Officer Steven Cobb, one of the uniformed officers who approached the door, testified that the officers banged on the left apartment door “as loud as [they] could” and announced, “ ‘This is the police’ ” or “ ‘Police, police, police.’ ” Id., at 22–23. Cobb said that “[a]s soon as [the officers] started banging on the door,” they “could hear people inside moving,” and “[i]t sounded as [though] things were being moved inside the apartment.” Id., at 24. These noises, Cobb testified, led the officers to believe that drug-related evidence was about to be destroyed. At that point, the officers announced that they “were going to make entry inside the apartment.” Ibid. Cobb then kicked in the door, the officers entered the apartment, and they found three people in the front room: respondent Hollis King, respondent’s girlfriend, and a guest who was smoking marijuana.[Footnote 1] The officers performed a protective sweep of the apartment during which they saw marijuana and powder cocaine in plain view. In a subsequent search, they also discovered crack cocaine, cash, and drug paraphernalia. Police eventually entered the apartment on the right. Inside, they found the suspected drug dealer who was the initial target of their investigation. B In the Fayette County Circuit Court, a grand jury charged respondent with trafficking in marijuana, first-degree trafficking in a controlled substance, and second-degree persistent felony offender status. Respondent filed a motion to suppress the evidence from the warrantless search, but the Circuit Court denied the motion. The Circuit Court concluded that the officers had probable cause to investigate the marijuana odor and that the officers “properly conducted [the investigation] by initially knocking on the door of the apartment unit and awaiting the response or consensual entry.” App. to Pet. for Cert. 9a. Exigent circumstances justified the warrantless entry, the court held, because “there was no response at all to the knocking,” and because “Officer Cobb heard movement in the apartment which he reasonably concluded were persons in the act of destroying evidence, particularly narcotics because of the smell.” Ibid. Respondent then entered a conditional guilty plea, reserving his right to appeal the denial of his suppression motion. The court sentenced respondent to 11 years’ imprisonment. The Kentucky Court of Appeals affirmed. It held that exigent circumstances justified the warrantless entry because the police reasonably believed that evidence would be destroyed. The police did not impermissibly create the exigency, the court explained, because they did not deliberately evade the warrant requirement. The Supreme Court of Kentucky reversed. 302 S. W. 3d 649 (2010). As a preliminary matter, the court observed that there was “certainly some question as to whether the sound of persons moving [inside the apartment] was sufficient to establish that evidence was being destroyed.” Id., at 655. But the court did not answer that question. Instead, it “assume[d] for the purpose of argument that exigent circumstances existed.” Ibid. To determine whether police impermissibly created the exigency, the Supreme Court of Kentucky announced a two-part test. First, the court held, police cannot “deliberately creat[e] the exigent circumstances with the bad faith intent to avoid the warrant requirement.” Id., at 656 (internal quotation marks omitted). Second, even absent bad faith, the court concluded, police may not rely on exigent circumstances if “it was reasonably foreseeable that the investigative tactics employed by the police would create the exigent circumstances.” Ibid. (internal quotation marks omitted). Although the court found no evidence of bad faith, it held that exigent circumstances could not justify the search because it was reasonably foreseeable that the occupants would destroy evidence when the police knocked on the door and announced their presence. Ibid. We granted certiorari. 561 U. S. ___ (2010).[Footnote 2] II A The Fourth Amendment provides: “The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.” The text of the Amendment thus expressly imposes two requirements. First, all searches and seizures must be reasonable. Second, a warrant may not be issued unless probable cause is properly established and the scope of the authorized search is set out with particularity. See Payton v. New York, 445 U. S. 573, 584 (1980). Although the text of the Fourth Amendment does not specify when a search warrant must be obtained, this Court has inferred that a warrant must generally be secured. “It is a ‘basic principle of Fourth Amendment law,’ ” we have often said, “ ‘that searches and seizures inside a home without a warrant are presumptively unreasonable.’ ” Brigham City v. Stuart, 547 U. S. 398, 403 (2006) (quoting Groh v. Ramirez, 540 U. S. 551, 559 (2004)). But we have also recognized that this presumption may be overcome in some circumstances because “[t]he ultimate touchstone of the Fourth Amendment is ‘reasonableness.’ ” Brigham City, supra, at 403; see also Michigan v. Fisher, 558 U. S. ___, ___ (2009) (per curiam) (slip op., at 2). Accordingly, the warrant requirement is subject to certain reasonable exceptions. Brigham City, supra, at 403. One well-recognized exception applies when “ ‘the exigencies of the situation’ make the needs of law en-forcement so compelling that [a] warrantless search is objectively reasonable under the Fourth Amendment.” Mincey v. Arizona, 437 U. S. 385, 394 (1978); see also Payton, supra, at 590 (“[T]he Fourth Amendment has drawn a firm line at the entrance to the house. Absent exigent circumstances, that threshold may not reasonably be crossed without a warrant”). This Court has identified several exigencies that may justify a warrantless search of a home. See Brigham City, 547 U. S., at 403. Under the “emergency aid” exception, for example, “officers may enter a home without a warrant to render emergency assistance to an injured occupant or to protect an occupant from imminent injury.” Ibid.; see also, e.g., Fisher, supra, at ___ (slip op., at 5) (upholding warrantless home entry based on emergency aid exception). Police officers may enter premises without a warrant when they are in hot pursuit of a fleeing suspect. See United States v. Santana, 427 U. S. 38, 42–43 (1976). And—what is relevant here—the need “to prevent the imminent destruction of evidence” has long been recognized as a sufficient justification for a warrantless search. Brigham City, supra, at 403; see also Georgia v. Randolph, 547 U. S. 103, 116, n. 6 (2006); Minnesota v. Olson, 495 U. S. 91, 100 (1990).[Footnote 3] B Over the years, lower courts have developed an exception to the exigent circumstances rule, the so-called “police-created exigency” doctrine. Under this doctrine, police may not rely on the need to prevent destruction of evidence when that exigency was “created” or “manufactured” by the conduct of the police. See, e.g., United States v. Chambers, 395 F. 3d 563, 566 (CA6 2005) (“[F]or a warrantless search to stand, law enforcement officers must be responding to an unanticipated exigency rather than simply creating the exigency for themselves”); United States v. Gould, 364 F. 3d 578, 590 (CA5 2004) (en banc) (“[A]lthough exigent circumstances may justify a warrantless probable cause entry into the home, they will not do so if the exigent circumstances were manufactured by the agents” (internal quotation marks omitted)). In applying this exception for the “creation” or “manufacturing” of an exigency by the police, courts require something more than mere proof that fear of detection by the police caused the destruction of evidence. An additional showing is obviously needed because, as the Eighth Circuit has recognized, “in some sense the police always create the exigent circumstances.” United States v. Duchi, 906 F. 2d 1278, 1284 (CA8 1990). That is to say, in the vast majority of cases in which evidence is destroyed by persons who are engaged in illegal conduct, the reason for the destruction is fear that the evidence will fall into the hands of law enforcement. Destruction of evidence issues probably occur most frequently in drug cases because drugs may be easily destroyed by flushing them down a toilet or rinsing them down a drain. Persons in possession of valuable drugs are unlikely to destroy them unless they fear discovery by the police. Consequently, a rule that precludes the police from making a warrantless entry to prevent the destruction of evidence whenever their conduct causes the exigency would unreasonably shrink the reach of this well-established exception to the warrant requirement. Presumably for the purpose of avoiding such a result, the lower courts have held that the police-created exigency doctrine requires more than simple causation, but the lower courts have not agreed on the test to be applied. Indeed, the petition in this case maintains that “[t]here are currently five different tests being used by the United States Courts of Appeals,” Pet. for Cert. 11, and that some state courts have crafted additional tests, id., at 19–20. III A Despite the welter of tests devised by the lower courts, the answer to the question presented in this case follows directly and clearly from the principle that permits warrantless searches in the first place. As previously noted, warrantless searches are allowed when the circumstances make it reasonable, within the meaning of the Fourth Amendment, to dispense with the warrant requirement. Therefore, the answer to the question before us is that the exigent circumstances rule justifies a warrantless search when the conduct of the police preceding the exigency is reasonable in the same sense. Where, as here, the police did not create the exigency by engaging or threatening to engage in conduct that violates the Fourth Amendment, warrantless entry to prevent the destruction of evidence is reasonable and thus allowed.[Footnote 4] We have taken a similar approach in other cases involving warrantless searches. For example, we have held that law enforcement officers may seize evidence in plain view, provided that they have not violated the Fourth Amendment in arriving at the spot from which the observation of the evidence is made. See Horton v. California, 496 U. S. 128, 136–140 (1990). As we put it in Horton, “[i]t is … an essential predicate to any valid warrantless seizure of incriminating evidence that the officer did not violate the Fourth Amendment in arriving at the place from which the evidence could be plainly viewed.” Id., at 136. So long as this prerequisite is satisfied, however, it does not matter that the officer who makes the observation may have gone to the spot from which the evidence was seen with the hope of being able to view and seize the evidence. See id., at 138 (“The fact that an officer is interested in an item of evidence and fully expects to find it in the course of a search should not invalidate its seizure”). Instead, the Fourth Amendment requires only that the steps preceding the seizure be lawful. See id., at 136–137. Similarly, officers may seek consent-based encounters if they are lawfully present in the place where the consensual encounter occurs. See INS v. Delgado, 466 U. S. 210, 217, n. 5 (1984) (noting that officers who entered into consent-based encounters with employees in a factory building were “lawfully present [in the factory] pursuant to consent or a warrant”). If consent is freely given, it makes no difference that an officer may have approached the person with the hope or expectation of obtaining consent. See id., at 216 (“While most citizens will respond to a police request, the fact that people do so, and do so without being told they are free not to respond, hardly eliminates the consensual nature of the response”). B Some lower courts have adopted a rule that is similar to the one that we recognize today. See United States v. MacDonald, 916 F. 2d 766, 772 (CA2 1990) (en banc) (law enforcement officers “do not impermissibly create exigent circumstances” when they “act in an entirely lawful manner”); State v. Robinson, 2010 WI 80, ¶32, 327 Wis. 2d 302, 326–328, 786 N. W. 2d 463, 475–476 (2010). But others, including the Kentucky Supreme Court, have imposed additional requirements that are unsound and that we now reject. Bad faith. Some courts, including the Kentucky Supreme Court, ask whether law enforcement officers “ ‘deliberately created the exigent circumstances with the bad faith intent to avoid the warrant requirement.’ ” 302 S. W. 3d, at 656 (quoting Gould, 364 F. 3d, at 590); see also, e.g., Chambers, 395 F. 3d, at 566; United States v. Socey, 846 F. 2d 1439, 1448 (CADC 1988); United States v. Rengifo, 858 F. 2d 800, 804 (CA1 1988). This approach is fundamentally inconsistent with our Fourth Amendment jurisprudence. “Our cases have repeatedly rejected” a subjective approach, asking only whether “the circumstances, viewed objectively, justify the action.” ’Brigham City, 547 U. S., at 404 (alteration and internal quotation marks omitted); see also Fisher, 558 U. S., at ___ (slip op., at 3–5). Indeed, we have never held, outside limited contexts such as an “inventory search or administrative inspection … , that an officer’s motive invalidates objectively justifiable behavior under the Fourth Amendment.” Whren v. United States, 517 U. S. 806, 812 (1996); see also Brigham City, supra, at 405. The reasons for looking to objective factors, rather than subjective intent, are clear. Legal tests based on reasonableness are generally objective, and this Court has long taken the view that “evenhanded law enforcement is best achieved by the application of objective standards of conduct, rather than standards that depend upon the subjective state of mind of the officer.” Horton, supra, at 138. Reasonable foreseeability. Some courts, again including the Kentucky Supreme Court, hold that police may not rely on an exigency if “ ‘it was reasonably foreseeable that the investigative tactics employed by the police would create the exigent circumstances.’ ” 302 S. W. 3d, at 656 (quoting Mann v. State, 357 Ark. 159, 172, 161 S. W. 3d 826, 834 (2004)); see also, e.g., United States v. Mowatt, 513 F. 3d 395, 402 (CA4 2008). Courts applying this test have invalidated warrantless home searches on the ground that it was reasonably foreseeable that police officers, by knocking on the door and announcing their presence, would lead a drug suspect to destroy evidence. See, e.g., id., at 402–403; 302 S. W. 3d, at 656. Contrary to this reasoning, however, we have rejected the notion that police may seize evidence without a warrant only when they come across the evidence by happenstance. In Horton, as noted, we held that the police may seize evidence in plain view even though the officers may be “interested in an item of evidence and fully expec[t] to find it in the course of a search.” 496 U. S., at 138. Adoption of a reasonable foreseeability test would also introduce an unacceptable degree of unpredictability. For example, whenever law enforcement officers knock on the door of premises occupied by a person who may be involved in the drug trade, there is some possibility that the occupants may possess drugs and may seek to destroy them. Under a reasonable foreseeability test, it would be necessary to quantify the degree of predictability that must be reached before the police-created exigency doctrine comes into play. A simple example illustrates the difficulties that such an approach would produce. Suppose that the officers in the present case did not smell marijuana smoke and thus knew only that there was a 50% chance that the fleeing suspect had entered the apartment on the left rather than the apartment on the right. Under those circumstances, would it have been reasonably foreseeable that the occupants of the apartment on the left would seek to destroy evidence upon learning that the police were at the door? Or suppose that the officers knew only that the suspect had disappeared into one of the apartments on a floor with 3, 5, 10, or even 20 units? If the police chose a door at random and knocked for the purpose of asking the occupants if they knew a person who fit the description of the suspect, would it have been reasonably foreseeable that the occupants would seek to destroy evidence? We have noted that “[t]he calculus of reasonableness must embody allowance for the fact that police officers are often forced to make split-second judgments—in circumstances that are tense, uncertain, and rapidly evolving.” Graham v. Connor, 490 U. S. 386, 396–397 (1989). The reasonable foreseeability test would create unacceptable and unwarranted difficulties for law enforcement officers who must make quick decisions in the field, as well as for judges who would be required to determine after the fact whether the destruction of evidence in response to a knock on the door was reasonably foreseeable based on what the officers knew at the time. Probable cause and time to secure a warrant. Some courts, in applying the police-created exigency doctrine, fault law enforcement officers if, after acquiring evidence that is sufficient to establish probable cause to search particular premises, the officers do not seek a warrant but instead knock on the door and seek either to speak with an occupant or to obtain consent to search. See, e.g., Chambers, supra, at 569 (citing “[t]he failure to seek a warrant in the face of plentiful probable cause” as a factor indicating that the police deliberately created the exigency). This approach unjustifiably interferes with legitimate law enforcement strategies. There are many entirely proper reasons why police may not want to seek a search warrant as soon as the bare minimum of evidence needed to establish probable cause is acquired. Without attempting to provide a comprehensive list of these reasons, we note a few. First, the police may wish to speak with the occupants of a dwelling before deciding whether it is worthwhile to seek authorization for a search. They may think that a short and simple conversation may obviate the need to apply for and execute a warrant. See Schneckloth v. Bustamonte, 412 U. S. 218, 228 (1973). Second, the police may want to ask an occupant of the premises for consent to search because doing so is simpler, faster, and less burdensome than applying for a warrant. A consensual search also “may result in considerably less inconvenience” and embarrassment to the occupants than a search conducted pursuant to a warrant. Ibid. Third, law enforcement officers may wish to obtain more evidence before submitting what might otherwise be considered a marginal warrant application. Fourth, prosecutors may wish to wait until they acquire evidence that can justify a search that is broader in scope than the search that a judicial officer is likely to authorize based on the evidence then available. And finally, in many cases, law enforcement may not want to execute a search that will disclose the existence of an investigation because doing so may interfere with the acquisition of additional evidence against those already under suspicion or evidence about additional but as yet unknown participants in a criminal scheme. We have said that “[l]aw enforcement officers are under no constitutional duty to call a halt to criminal investigation the moment they have the minimum evidence to establish probable cause.” Hoffa v. United States, 385 U. S. 293, 310 (1966). Faulting the police for failing to apply for a search warrant at the earliest possible time after obtaining probable cause imposes a duty that is nowhere to be found in the Constitution. Standard or good investigative tactics. Finally, some lower court cases suggest that law enforcement officers may be found to have created or manufactured an exigency if the court concludes that the course of their investigation was “contrary to standard or good law enforcement practices (or to the policies or practices of their jurisdictions).” Gould, 364 F. 3d, at 591. This approach fails to provide clear guidance for law enforcement officers and authorizes courts to make judgments on matters that are the province of those who are responsible for federal and state law enforcement agencies. C Respondent argues for a rule that differs from those discussed above, but his rule is also flawed. Respondent contends that law enforcement officers impermissibly create an exigency when they “engage in conduct that would cause a reasonable person to believe that entry is imminent and inevitable.” Brief for Respondent 24. In respondent’s view, relevant factors include the officers’ tone of voice in announcing their presence and the forcefulness of their knocks. But the ability of law enforcement officers to respond to an exigency cannot turn on such subtleties. Police officers may have a very good reason to announce their presence loudly and to knock on the door with some force. A forceful knock may be necessary to alert the occupants that someone is at the door. Cf. United States v. Banks, 540 U. S. 31, 33 (2003) (Police “rapped hard enough on the door to be heard by officers at the back door” and announced their presence, but defendant “was in the shower and testified that he heard nothing”). Furthermore, unless police officers identify themselves loudly enough, occupants may not know who is at their doorstep. Officers are permitted—indeed, encouraged—to identify themselves to citizens, and “in many circumstances this is cause for assurance, not discomfort.” United States v. Drayton, 536 U. S. 194, 204 (2002). Citizens who are startled by an unexpected knock on the door or by the sight of unknown persons in plain clothes on their doorstep may be relieved to learn that these persons are police officers. Others may appreciate the opportunity to make an informed decision about whether to answer the door to the police. If respondent’s test were adopted, it would be extremely difficult for police officers to know how loudly they may announce their presence or how forcefully they may knock on a door without running afoul of the police-created exigency rule. And in most cases, it would be nearly impossible for a court to determine whether that threshold had been passed. The Fourth Amendment does not require the nebulous and impractical test that respondent proposes.[Footnote 5] D For these reasons, we conclude that the exigent circumstances rule applies when the police do not gain entry to premises by means of an actual or threatened violation of the Fourth Amendment. This holding provides ample protection for the privacy rights that the Amendment protects. When law enforcement officers who are not armed with a warrant knock on a door, they do no more than any private citizen might do. And whether the person who knocks on the door and requests the opportunity to speak is a police officer or a private citizen, the occupant has no obligation to open the door or to speak. Cf. Florida v. Royer, 460 U. S. 491, 497–498 (1983). (“[H]e may decline to listen to the questions at all and may go on his way”). When the police knock on a door but the occupants choose not to respond or to speak, “the investigation will have reached a conspicuously low point,” and the occupants “will have the kind of warning that even the most elaborate security system cannot provide.” Chambers, 395 F. 3d, at 577 (Sutton, J., dissenting). And even if an occupant chooses to open the door and speak with the officers, the occupant need not allow the officers to enter the premises and may refuse to answer any questions at any time. Occupants who choose not to stand on their constitutional rights but instead elect to attempt to destroy evidence have only themselves to blame for the warrantless exigent-circumstances search that may ensue. IV We now apply our interpretation of the police-created exigency doctrine to the facts of this case. A We need not decide whether exigent circumstances existed in this case. Any warrantless entry based on exigent circumstances must, of course, be supported by a genuine exigency. See Brigham City, 547 U. S., at 406. The trial court and the Kentucky Court of Appeals found that there was a real exigency in this case, but the Kentucky Supreme Court expressed doubt on this issue, observing that there was “certainly some question as to whether the sound of persons moving [inside the apartment] was sufficient to establish that evidence was being destroyed.” 302 S. W. 3d, at 655. The Kentucky Supreme Court “assum[ed] for the purpose of argument that exigent circumstances existed,” ibid., and it held that the police had impermissibly manufactured the exigency. We, too, assume for purposes of argument that an exigency existed. We decide only the question on which the Kentucky Supreme Court ruled and on which we granted certiorari: Under what circumstances do police impermissibly create an exigency? Any question about whether an exigency actually existed is better addressed by the Kentucky Supreme Court on remand. See Kirk v. Louisiana, 536 U. S. 635, 638 (2002) (per curiam) (reversing state-court judgment that exigent circumstances were not required for warrantless home entry and remanding for state court to determine whether exigent circumstances were present). B In this case, we see no evidence that the officers either violated the Fourth Amendment or threatened to do so prior to the point when they entered the apartment. Officer Cobb testified without contradiction that the officers “banged on the door as loud as [they] could” and announced either “ ‘Police, police, police’ ” or “ ‘This is the police.’ ” App. 22–23. This conduct was entirely consistent with the Fourth Amendment, and we are aware of no other evidence that might show that the officers either violated the Fourth Amendment or threatened to do so (for example, by announcing that they would break down the door if the occupants did not open the door voluntarily). Respondent argues that the officers “demanded” entry to the apartment, but he has not pointed to any evidence in the record that supports this assertion. He relies on a passing statement made by the trial court in its opinion denying respondent’s motion to suppress. See App. to Pet. for Cert. 3a–4a. In recounting the events that preceded the search, the judge wrote that the officers “banged on the door of the apartment on the back left of the breezeway identifying themselves as police officers and demanding that the door be opened by the persons inside.” Ibid. (emphasis added and deleted). However, at a later point in this opinion, the judge stated that the officers “initially knock[ed] on the door of the apartment unit and await[ed] the response or consensual entry.” Id., at 9a. This later statement is consistent with the testimony at the suppression hearing and with the findings of the state appellate courts. See 302 S. W. 3d, at 651 (The officers “knocked loudly on the back left apartment door and announced ‘police’ ”); App. to Pet. for Cert. 14a (The officers “knock[ed] on the door and announc[ed] themselves as police”); App. 22–24. There is no evidence of a “demand” of any sort, much less a demand that amounts to a threat to violate the Fourth Amendment. If there is contradictory evidence that has not been brought to our attention, the state court may elect to address that matter on remand. Finally, respondent claims that the officers “explained to [the occupants that the officers] were going to make entry inside the apartment,” id., at 24, but the record is clear that the officers did not make this statement until after the exigency arose. As Officer Cobb testified, the officers “knew that there was possibly something that was going to be destroyed inside the apartment,” and “[a]t that point, … [they] explained … [that they] were going to make entry.” Ibid. (emphasis added). Given that this announcement was made after the exigency arose, it could not have created the exigency. * * * Like the court below, we assume for purposes of argument that an exigency existed. Because the officers in this case did not violate or threaten to violate the Fourth Amendment prior to the exigency, we hold that the exigency justified the warrantless search of the apartment. The judgment of the Kentucky Supreme Court is reversed, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered. Footnote 1 Respondent’s girlfriend leased the apartment, but respondent stayed there part of the time, and his child lived there. Based on these facts, Kentucky conceded in state court that respondent has Fourth Amendment standing to challenge the search. See App. to Pet. for Cert. 7a; see also 302 S. W. 3d 649, 652 (Ky. 2010). Footnote 2 After we granted certiorari, respondent filed a motion to dismiss the petition as improvidently granted, which we denied. 562 U. S. ___ (2010). Respondent’s principal argument was that the case was moot because, after the Kentucky Supreme Court reversed his conviction, the Circuit Court dismissed the charges against him. Respondent’s argument is foreclosed by United States v. Villamonte-Marquez, 462 U. S. 579, 581, n. 2 (1983). As we explained in Villamonte-Marquez, our reversal of the Kentucky Supreme Court’s decision “would reinstate the judgment of conviction and the sentence entered” by the Circuit Court. Ibid. The absence of an indictment does not change matters. See ibid. (“Upon respondents’ conviction and sentence, the indictment that was returned against them was merged into their convictions and sentences”). Footnote 3 Preventing the destruction of evidence may also justify dispensing with Fourth Amendment requirements in other contexts. See, e.g., Richards v. Wisconsin, 520 U. S. 385, 395–396 (1997) (failure to comply with the knock-and-announce requirement was justified because “the circumstances … show[ed] that the officers had a reasonable suspicion that [a suspect] might destroy evidence if given further opportunity to do so”); Schmerber v. California, 384 U. S. 757, 770–771 (1966) (warrantless testing for blood-alcohol content was justified based on potential destruction of evidence); cf. United States v. Banks, 540 U. S. 31, 37–40 (2003) (15 to 20 seconds was a reasonable time for officers to wait after knocking and announcing their presence where there was a risk that suspect would dispose of cocaine). Footnote 4 There is a strong argument to be made that, at least in most circumstances, the exigent circumstances rule should not apply where the police, without a warrant or any legally sound basis for a warrantless entry, threaten that they will enter without permission unless admitted. In this case, however, no such actual threat was made, and therefore we have no need to reach that question. Footnote 5 Contrary to respondent’s argument, see Brief for Respondent 13–18, Johnson v. United States, 333 U. S. 10 (1948), does not require affirmance in this case. In Johnson, officers noticed the smell of burning opium emanating from a hotel room. They then knocked on the door and demanded entry. Upon seeing that Johnson was the only occupant of the room, they placed her under arrest, searched the room, and discovered opium and drug paraphernalia. Id., at 11. Defending the legality of the search, the Government attempted to justify the warrantless search of the room as a valid search incident to a lawful arrest. See Brief for United States in Johnson v. United States, O. T. 1947, No. 329, pp. 13, 16, 36. The Government did not contend that the officers entered the room in order to prevent the destruction of evidence. Although the officers said that they heard a “ ‘shuffling’ ” noise inside the room after they knocked on the door, 333 U. S., at 12, the Government did not claim that this particular noise was a noise that would have led a reasonable officer to think that evidence was about to be destroyed. Thus, Johnson is simply not a case about exigent circumstances. See id., at 14–15 (noting that if “exceptional circumstances” existed—for example, if a “suspect was fleeing or likely to take flight” or if “evidence or contraband was threatened with removal or destruction”—then “it may be contended that a magistrate’s warrant for search may be dispensed with”).

562.US.29

The Humphries (hereinafter respondents) were accused of child abuse in California, but were later exonerated. However, under California law, their names were added to a Child Abuse Central Index (Index), where they would remain available to various state agencies for at least 10 years. The statute has no procedures for allowing individuals to challenge their inclusion in the Index, and neither California nor Los Angeles County has created such procedures. Respondents filed suit under §1983, seeking damages, an injunction, and a declaration that public officials and petitioner Los Angeles County had deprived them of their constitutional rights by failing to create a mechanism through which they could contest inclusion in the Index. The District Court granted the defendants summary judgment, but the Ninth Circuit disagreed, holding that the Fourteenth Amendment required the State to provide those on the list with notice and a hearing, and thus respondents were entitled to declaratory relief. The court also held that respondents were prevailing parties entitled to attorney’s fees, including $60,000 from the county. The county objected, claiming that as a municipal entity, it was liable only if its “policy or custom” caused the deprivation of a plaintiff’s federal right, Monell v. New York City Dept. of Social Servs., 436 U. S. 658, 694; but a state policy caused any deprivation here. The Ninth Circuit, inter alia, found that respondents did prevail against the county on their claim for declaratory relief because Monell did not apply to prospective relief claims. Held: Monell’s “policy or custom” requirement applies in §1983 cases irrespective of whether the relief sought is monetary or prospective. Pp. 4–10. (a) In Monroe v. Pape, 365 U. S. 167, this Court based its holding that municipal entities were not “person[s]” under §1983 on the provision’s legislative history, particularly Congress’ rejection of the so-called Sherman amendment, which would have made municipalities liable for damages done by private persons “ ‘riotously and tumultuously assembled,’ ” id., at 188–190, and n. 38. Reexamining this legislative history in Monell, the Court overruled Monroe. It concluded that Congress had rejected the Sherman amendment, not because it would have imposed liability on municipalities, but because it would have imposed such liability solely based on the acts of others. The Court, on the basis of the statutory text and the legislative history, went on to explain what acts are the municipality’s own for purposes of liability. The Court held that “a municipality cannot be held liable” solely for the acts of others, e.g., “solely because it employs a tortfeasor,” 436 U. S., at 691, but it may be held liable “when execution of a government’s policy or custom … inflicts the injury,” id., at 694. Pp. 4–7. (b) Section 1983, read in light of Monell’s understanding of the legislative history, explains why claims for prospective relief, like claims for money damages, fall within the scope of the “policy or custom” requirement. Nothing in §1983 suggests that the causation requirement should change with the form of relief sought. In fact, the text suggests the opposite when it provides that a person who meets §1983’s elements “shall be liable … in an action at law, suit in equity, or other proper proceeding for redress.” Thus, as Monell explicitly stated, “local governing bodies … can be sued directly under §1983 for monetary, declaratory, or injunctive relief where, as here, the action that is alleged to be unconstitutional implements or executes” a policy or custom. 436 U. S., at 690. To find the “policy or custom” requirement inapplicable in prospective relief cases would also undermine Monell’s logic. For whether an action or omission is a municipality’s “own” has to do with the nature of the action or omission, not with the nature of the relief that is later sought in court. Pp. 7–8. (c) Respondents’ arguments to the contrary are unconvincing. Pp. 8–9. Reversed and remanded. Breyer, J., delivered the opinion of the Court, in which all other Members joined, except Kagan, J., who took no part in the consideration or decision of the case.

In Monell v. New York City Dept. of Social Servs., 436 U. S. 658 (1978), this Court held that civil rights plaintiffs suing a municipal entity under 42 U. S. C. §1983 must show that their injury was caused by a municipal policy or custom. The case before the Court in Monell directly involved monetary damages. The question presented is whether the “policy or custom” requirement also applies when plaintiffs seek prospective relief, such as an injunction or a declaratory judgment. We conclude that it does so apply. I The case arises out of the following circumstances: The California Child Abuse and Neglect Reporting Act, Cal. Penal Code Ann. §11164 et seq. (West Rev. Supp. 2010), requires law enforcement and other state agencies to investigate allegations of child abuse. These agencies must report to the California Department of Justice all instances of reported child abuse the agency finds “not unfounded,” even if they are “inconclusive or unsubstantiated.” §§11169(a), 11170(a)(3). The statute requires the department to include all these reports in a Child Abuse Central Index (Index), where they remain available to various state agencies for at least 10 years. §11170(a). The statute also says that if “a report has previously been filed which subsequently proves to be unfounded, the Department of Justice shall be notified in writing of that fact and shall not retain the report.” §11169(a). The statute, however, does not set forth procedures for reviewing whether a previously filed report is unfounded, or for allowing individuals to challenge their inclusion in the Index. Nor, up until the time of this lawsuit, had California or Los Angeles County created any such procedures. But cf. §11170(a)(2) (“The submitting agencies are responsible for the accuracy, completeness, and retention of the reports described in this section”). The two plaintiffs in this case were initially accused of child abuse. But they were later exonerated. They sought to have their names removed from the Index. Unable to convince the Los Angeles Sheriff’s Department to remove them, they filed this §1983 case against the attorney general of California, the Los Angeles County sheriff, two detectives in the sheriff’s department, and the County of Los Angeles. They sought damages, an injunction, and a declaration that the defendants had deprived them of their constitutional rights by failing to create a procedural mechanism through which one could contest inclusion on the Index. See U. S. Const., Amdt. 14; Rev. Stat. §1979, 42 U. S. C. §1983. The District Court for the Central District of California granted summary judgment to all of the defendants on the ground that California had not deprived plaintiffs of a constitutionally protected “liberty” interest. But on appeal the Ninth Circuit disagreed. The Ninth Circuit held that the Fourteenth Amendment required the State to provide those included on the list notice and “ ‘some kind of hearing.’ ” 554 F. 3d 1170, 1201 (2009). Thus the Circuit held that the plaintiffs were entitled to declaratory relief, and it believed that (on remand) they might prove damages as well. Ibid. The Ninth Circuit also held that the plaintiffs were prevailing parties, thereby entitled to approximately $600,000 in attorney’s fees. 42 U. S. C. §1988(b) (providing for payment of attorney’s fees to parties prevailing on §1983 claims). See No. 05–56467 (June 22, 2009), App. to Pet. for Cert. 1–4 (hereinafter First Fee Order); No. 05–56467 (Dec. 2, 2009), App. to Reply to Brief in Opposition 1–2 (hereinafter Second Fee Order). The Ninth Circuit wrote that Los Angeles County must pay approximately $60,000 of this amount. First Fee Order 3; Second Fee Order 2. Los Angeles County denied that it was liable and therefore that it could be held responsible for attorney’s fees. It argued that, in respect to the county, the plaintiffs were not prevailing parties. That is because the county is a municipal entity. Under Monell’s holding a municipal entity is liable under §1983 only if a municipal “policy or custom” caused a plaintiff to be deprived of a federal right. 436 U. S., at 694 (emphasis added). And it was state policy, not county policy, that brought about any deprivation here. The Ninth Circuit responded to this argument as follows: First, it said that county policy might be responsible for the deprivation. It “is possible,” the Ninth Circuit said, that the county, “[b]y failing to” “creat[e] an independent procedure that would allow” the plaintiffs “to challenge their listing[,] . . . adopted a custom and policy that violated” the plaintiffs’ “constitutional rights.” 554 F. 3d, at 1202. Second, it said that “because this issue is not clear based on the record before us on appeal … we remand to the district court to determine the County’s liability under Monell.” Ibid. Third, it saw no reason to remand in respect to the county’s obligation to pay $60,000 in attorney’s fees. That, it wrote, is because “in our circuit … the limitations to liability established in Monell do not apply to claims for prospective relief,” such as the declaratory judgment that the Circuit had ordered entered. First Fee Order 3–4 (citing Chaloux v. Killeen, 886 F. 2d 247, 250 (CA9 1989); Truth v. Kent School Dist., 542 F. 3d 634, 644 (CA9 2008); emphasis added). The county then asked us to review this last-mentioned Ninth Circuit holding, namely, the holding that Monell’s “policy or custom” requirement applies only to claims for damages but not to claims for prospective relief. Because the Courts of Appeals are divided on this question, we granted the county’s petition for certiorari. Compare Reynolds v. Giuliani, 506 F. 3d 183, 191 (CA2 2007) (holding that Monell’s “policy or custom” requirement applies to claims for prospective relief as well as claims for damages); Dirrane v. Brookline Police Dept., 315 F. 3d 65, 71 (CA1 2002) (same); Greensboro Professional Fire Fighters Assn., Local 3157 v. Greensboro, 64 F. 3d 962, 967, n. 6 (CA4 1995) (applying the Monell requirement to a prospective relief claim); Church v. Huntsville, 30 F. 3d 1332, 1347 (CA11 1994) (same), with Chaloux, supra, at 251 (holding that Monell does not apply to prospective relief claims). See also Gernetzke v. Kenosha Unified School Dist. No. 1, 274 F. 3d 464, 468 (CA7 2001) (reserving the question but noting the “predominant” view that “Monell’s holding applies regardless of the nature of the relief sought”). We conclude that Monell’s holding applies to §1983 claims against municipalities for prospective relief as well as to claims for damages. II A We begin with §1983 itself, which provides: “Every person who, under color of any [state] statute, ordinance, regulation, custom, or usage … subjects, or causes to be subjected, any … other person … to the deprivation of any rights … secured by the Constitution and laws [of the United States], shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress.” (Emphasis added.) In 1961, in Monroe v. Pape, 365 U. S. 167, this Court held that municipal entities were not “person[s]” under §1983. The Court based this conclusion on the history of the Civil Rights Act of 1871’s enactment. It noted that Congress rejected an amendment (called the Sherman amendment) that would have made municipalities liable for damage done by private persons “ ‘riotously and tumultuously assembled.’ ” Id., at 188–190, and n. 38 (quoting Cong. Globe, 42d Cong., 1st Sess., 663 (1871)). This rejection, the Court thought, reflected a determination by the 1871 House of Representatives that “ ‘Congress had no constitutional power to impose any obligation upon county and town organizations, the mere instrumentality for the administration of state law.’ ” 365 U. S., at 190 (quoting Cong. Globe, supra, at 804 (statement of Rep. Poland); emphasis added). The Court concluded that Congress must have doubted its “constitutional power . . . to impose civil liability on municipalities.” 365 U. S., at 190. And for that reason, Congress must have intended to exclude municipal corporations as §1983 defendants. The statute’s key term “person” therefore did not cover municipal entities. Id., at 191. Sixteen years later, in Monell, the Court reconsidered the question of municipal liability. After reexamining the 1871 legislative history in detail, the Court concluded that Congress had rejected the Sherman amendment, not because it would have imposed liability upon municipalities, but because it would have imposed liability upon municipalities based purely upon the acts of others. That is to say, the rejected amendment would have imposed liability upon local governments “without regard to whether a local government was in any way at fault for the breach of the peace for which it was to be held for damages.” 436 U. S., at 681, n. 40 (emphasis added). In Monell’s view Congress may have thought that it lacked the power to impose that kind of indirect liability upon municipalities, id., at 679, but “nothing said in debate on the Sherman amendment would have prevented holding a municipality liable … for its own violations of the Fourteenth Amendment,” id., at 683 (emphasis added). The Court, overruling Monroe, held that municipalities were “persons” under §1983. 436 U. S., at 690. The Court also concluded that a municipality could not be held liable under §1983 solely because it employed a tortfeasor. The Court’s conclusion rested on “the language of §1983, read against the background of the same legislative history.” Id., at 691. Section 1983’s causation language imposes liability on a “ ‘person who … shall subject, or cause to be subjected, any person’ ” to a deprivation of federal rights. Ibid. (quoting 17 Stat. 13; emphasis deleted). That language, the Court observed, could not “be easily read to impose liability vicariously … solely on the basis of the existence of an employer-employee relationship with a tortfeasor.” 436 U. S., at 692. The statute’s legislative history, in particular the constitutional objections that had been raised to the Sherman amendment, supported this conclusion. Id., at 692–94, and n. 57. For these reasons, the Court concluded that a municipality could be held liable under §1983 only for its own violations of federal law. Id., at 694. The Court described what made a violation a municipality’s own violation: “Local governing bodies, therefore, can be sued directly under §1983 for monetary, declaratory, or injunctive relief where, as here, the action that is alleged to be unconstitutional implements or executes a policy statement, ordinance, regulation, or decision officially adopted and promulgated by that body’s officers. … [They can also be sued for] deprivations visited pursuant to governmental ‘custom’ even though such a custom has not received formal approval through the body’s official decisionmaking channels.” Id., at 690–691 (footnote omitted). The Court has also included the terms “usage” and “practice” as customs for which liability is appropriate. See ibid. The length of this list of types of municipal action leads us here to use a shorthand term “policy or custom,” but when we do so, we mean to refer to the entire list. See id., at 694 (using the shorthand “policy or custom”); see also, e.g., Fitzgerald v. Barnstable School Comm., 555 U. S. 246, ___ (2009) (slip op., at 10) (using the phrase “custom, policy, or practice,” to describe municipal liability under §1983). In sum, in Monell the Court held that “a municipality cannot be held liable” solely for the acts of others, e.g., “solely because it employs a tortfeasor.” 436 U. S., at 691. But the municipality may be held liable “when execution of a government’s policy or custom … inflicts the injury.” Id., at 694 (emphasis added). B The language of §1983 read in light of Monell’s understanding of the legislative history explains why claims for prospective relief, like claims for money damages, fall within the scope of the “policy or custom” requirement. Nothing in the text of §1983 suggests that the causation requirement contained in the statute should change with the form of relief sought. In fact, the text suggests the opposite when it provides that a person who meets §1983’s elements “shall be liable … in an action at law, suit in equity, or other proper proceeding for redress.” Thus, as Monell explicitly stated, “[l]ocal governing bodies … can be sued directly under §1983 for monetary, declaratory, or injunctive relief where, as here, the action that is alleged to be unconstitutional implements or executes” a policy or custom. 436 U. S., at 690 (emphasis added). Monell went on to quote this Court’s statement in a 1973 case, Kenosha v. Bruno, 412 U. S. 507, 513, to the effect that the Congress that enacted §1983 did not intend the “ ‘generic word “person” … to have a bifurcated application to municipal corporations depending on the nature of the relief sought against them.’ ” 436 U. S., at 701, n. 66 (emphasis added). Monell added that “[n]othing we say today affects” this pre-Monell “conclusion.” Ibid. Monell’s logic also argues against any such relief-based bifurcation. The Monell Court thought that Congress intended potential §1983 liability where a municipality’s own violations were at issue but not where only the violations of others were at issue. The “policy or custom” requirement rests upon that distinction and embodies it in law. To find the requirement inapplicable where prospective relief is at issue would undermine Monell’s logic. For whether an action or omission is a municipality’s “own” has to do with the nature of the action or omission, not with the nature of the relief that is later sought in court. C The Humphries’ (hereinafter respondents) arguments to the contrary are unconvincing. Respondents correctly note that by the time Monell reached the Supreme Court only the plaintiffs’ damages claim remained live. See id., at 661. From this fact they conclude that the Court’s holding applies directly only to claims for monetary damages. A holding, however, can extend through its logic beyond the specific facts of the particular case. It does so here. Respondents add that not only did Monell involve a damages claim, but its holding rests upon the concern that municipalities might have to pay large damages awards. The Court so suggests when it points out that municipalities should not be liable for an employee’s wrongful acts, simply by applying agency-based principles of respondeat superior. But as we have pointed out, the Court’s rejection of respondeat superior liability primarily rested not on the municipality’s economic needs, but on the fact that liability in such a case does not arise out of the municipality’s own wrongful conduct. Respondents further claim that, where prospective relief is at issue, Monell is redundant. They say that a court cannot grant prospective relief against a municipality unless the municipality’s own conduct has caused the violation. Hence, where such relief is otherwise proper, the Monell requirement “shouldn’t screen out any case.” Tr. of Oral Arg. 48. To argue that a requirement is necessarily satisfied, however, is not to argue that its satisfaction is unnecessary. If respondents are right, our holding may have limited practical significance. But that possibility does not provide us with a convincing reason to sow confusion by adopting a bifurcated relief-based approach to municipal liability that the Court has previously rejected. Finally, respondents make the mirror-image argument that applying Monell’s requirement to prospective relief claims will leave some set of ongoing constitutional violations beyond redress. Despite the fact that four Circuits apply Monell’s requirement to prospective relief, however, respondents have not presented us with any actual or hypothetical example that provides serious cause for concern. * * * For these reasons, we hold that Monell’s “policy or custom” requirement applies in §1983 cases irrespective of whether the relief sought is monetary or prospective. The Ninth Circuit’s contrary judgment is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

563.US.27

Respondents filed this securities fraud class action, alleging that petitioners (hereinafter Matrixx) violated §10(b) of the Securities Exchange Act of 1934 and Securities and Exchange Commission Rule 10b–5 by failing to disclose reports of a possible link between Matrixx’s leading product, Zicam Cold Remedy, and loss of smell (anosmia), rendering statements made by Matrixx misleading. Matrixx moved to dismiss the complaint, arguing that respondents had not pleaded the element of a material misstatement or omission and the element of scienter. The District Court granted the motion, but the Ninth Circuit reversed. It held that the District Court erred in requiring an allegation of statistical significance to establish materiality, concluding instead that the complaint adequately alleged information linking Zicam and anosmia that would have been significant to a reasonable investor. It also held that Matrixx’s withholding of information about reports of adverse effects and about pending lawsuits by Zicam users gave rise to a strong inference of scienter. Held: Respondents have stated a claim under §10(b) and Rule 10b–5. Pp. 8–22. (a) To prevail on their claim, respondents must prove, as relevant here, a material misrepresentation or omission by Matrixx and scienter. See Stoneridge Investment Partners, LLC v. Scientific-Atlanta, Inc., 552 U. S. 148, 157. Matrixx contends that they failed to plead these required elements because they did not allege that the reports Matrixx received reflected statistically significant evidence that Zicam caused anosmia. Pp. 8–9. (b) Respondents have adequately pleaded materiality. Pp. 9–19. (1) Under Basic Inc. v. Levinson, 485 U. S. 224, §10(b)’s materiality requirement is satisfied when there is “ ‘a substantial likelihood that the disclosure of the omitted fact would have been viewed by the reasonable investor as having significantly altered the “total mix” of information made available.’ ” Id., at 231–232. The Court declined to adopt a bright-line rule for determining materiality in Basic, observing that “[a]ny approach that designates a single fact or occurrence as always determinative of an inherently fact-specific finding such as materiality, must necessarily be overinclusive or underinclusive.” Id., at 236. Here, Matrixx’s bright-line rule—that adverse event reports regarding a pharmaceutical company’s products are not material absent a sufficient number of such reports to establish a statistically significant risk that the product is causing the events—would “artificially exclud[e]” information that “would otherwise be considered significant to [a reasonable investor’s] trading decision.” Ibid. Matrixx’s premise that statistical significance is the only reliable indication of causation is flawed. Both medical experts and the Food and Drug Administration rely on evidence other than statistically significant data to establish an inference of causation. It thus stands to reason that reasonable investors would act on such evidence. Because adverse reports can take many forms, assessing their materiality is a fact-specific inquiry, requiring consideration of their source, content, and context. The question is whether a reasonable investor would have viewed the nondisclosed information “ ‘as having significantly altered the “total mix” of information made available.’ ” Id., at 232. Something more than the mere existence of adverse event reports is needed to satisfy that standard, but that something more is not limited to statistical significance and can come from the source, content, and context of the reports. Pp. 9–16. (2) Applying Basic’s “total mix” standard here, respondents adequately pleaded materiality. The complaint’s allegations suffice to “raise a reasonable expectation that discovery will reveal evidence” satisfying the materiality requirement, Bell Atlantic Corp. v. Twombly, 550 U. S. 544, 556, and to “allo[w] the court to draw the reasonable inference that the defendant is liable,” Ashcroft v. Iqbal, 556 U. S. ___, ___. Assuming the complaint’s allegations to be true, Matrixx received reports from medical experts and researchers that plausibly indicated a reliable causal link between Zicam and anosmia. Consumers likely would have viewed Zicam’s risk as substantially outweighing its benefit. Viewing the complaint’s allegations as a whole, the complaint alleges facts suggesting a significant risk to the commercial viability of Matrixx’s leading product. It is substantially likely that a reasonable investor would have viewed this information “ ‘as having significantly altered the “total mix” of information made available.’ ” Basic, supra, at 232. Assuming the complaint’s allegations to be true, Matrixx told the market that revenues were going to rise 50 and then 80 percent when it had information indicating a significant risk to its leading revenue-generating product. It also publicly dismissed reports linking Zicam and anosmia and stated that zinc gluconate’s safety was well established, when it had evidence of a biological link between Zicam’s key ingredient and anosmia and had conducted no studies to disprove that link. Pp. 16–19. (c) Respondents have also adequately pleaded scienter, “ ‘a mental state embracing intent to deceive, manipulate, or defraud,’ ” Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U. S. 308, 319. This Court assumes, without deciding, that the scienter requirement may be satisfied by a showing of deliberate recklessness. Under the Private Securities Litigation Reform Act of 1995, a complaint adequately pleads scienter “only if a reasonable person would deem the inference of scienter cogent and at least as compelling as any opposing inference one could draw from the facts alleged.” Id., at 324. Matrixx’s proposed bright-line rule requiring an allegation of statistical significance to establish a strong inference of scienter is once again flawed. The complaint’s allegations, “taken collectively,” give rise to a “cogent and compelling” inference that Matrixx elected not to disclose adverse event reports not because it believed they were meaningless but because it understood their likely effect on the market. Id., at 323, 324. “[A] reasonable person” would deem the inference that Matrixx acted with deliberate recklessness “at least as compelling as any [plausible] opposing inference.” Id., at 324. Pp. 19–22. 585 F. 3d 1167, affirmed. Sotomayor, J., delivered the opinion for a unanimous Court.

This case presents the question whether a plaintiff can state a claim for securities fraud under §10(b) of the Securities Exchange Act of 1934, 48 Stat. 891, as amended, 15 U. S. C. §78j(b), and Securities and Exchange Commission (SEC) Rule 10b–5, 17 CFR §240.10b–5 (2010), based on a pharmaceutical company’s failure to disclose reports of adverse events associated with a product if the reports do not disclose a statistically significant number of adverse events. Respondents, plaintiffs in a securities fraud class action, allege that petitioners, Matrixx Initiatives, Inc., and three of its executives (collectively Matrixx), failed to disclose reports of a possible link between its leading product, a cold remedy, and loss of smell, rendering statements made by Matrixx misleading. Matrixx contends that respondents’ complaint does not adequately allege that Matrixx made a material representation or omission or that it acted with scienter because the complaint does not allege that Matrixx knew of a statistically significant number of adverse events requiring disclosure. We conclude that the materiality of adverse event reports cannot be reduced to a bright-line rule. Although in many cases reasonable investors would not consider reports of adverse events to be material information, respondents have alleged facts plausibly suggesting that reasonable investors would have viewed these particular reports as material. Respondents have also alleged facts “giving rise to a strong inference” that Matrixx “acted with the required state of mind.” 15 U. S. C. A. §78u–4(b)(2)(A) (Feb. 2011 Supp.). We therefore hold, in agreement with the Court of Appeals for the Ninth Circuit, that respondents have stated a claim under §10(b) and Rule 10b–5. I A Through a wholly owned subsidiary, Matrixx develops, manufactures, and markets over-the-counter pharmaceutical products. Its core brand of products is called Zicam. All of the products sold under the name Zicam are used to treat the common cold and associated symptoms. At the time of the events in question, one of Matrixx’s products was Zicam Cold Remedy, which came in several forms including nasal spray and gel. The active ingredient in Zicam Cold Remedy was zinc gluconate. Respondents allege that Zicam Cold Remedy accounted for approximately 70 percent of Matrixx’s sales. Respondents initiated this securities fraud class action against Matrixx on behalf of individuals who purchased Matrixx securities between October 22, 2003, and February 6, 2004.[Footnote 1] The action principally arises out of statements that Matrixx made during the class period relating to revenues and product safety. Respondents claim that Matrixx’s statements were misleading in light of reports that Matrixx had received, but did not disclose, about consumers who had lost their sense of smell (a condition called anosmia) after using Zicam Cold Remedy. Respondents’ consolidated amended complaint alleges the following facts, which the courts below properly assumed to be true. See Ashcroft v. Iqbal, 556 U. S. ___, ___ (2009) (slip op., at 14). In 1999, Dr. Alan Hirsch, neurological director of the Smell & Taste Treatment and Research Foundation, Ltd., called Matrixx’s customer service line after discovering a possible link between Zicam nasal gel and a loss of smell “in a cluster of his patients.” App. 67a–68a. Dr. Hirsch told a Matrixx employee that “previous studies had demonstrated that intranasal application of zinc could be problematic.” Id., at 68a. He also told the employee about at least one of his patients who did not have a cold and who developed anosmia after using Zicam. In September 2002, Timothy Clarot, Matrixx’s vice president for research and development, called Miriam Linschoten, Ph.D., at the University of Colorado Health Sciences Center after receiving a complaint from a per- son Linschoten was treating who had lost her sense of smell after using Zicam. Clarot informed Linschoten that Matrixx had received similar complaints from other customers. Linschoten drew Clarot’s attention to “previous studies linking zinc sulfate to loss of smell.” Ibid. Clarot gave her the impression that he had not heard of the studies. She asked Clarot whether Matrixx had done any studies of its own; he responded that it had not but that it had hired a consultant to review the product. Soon thereafter, Linschoten sent Clarot abstracts of the studies she had mentioned. Research from the 1930’s and 1980’s had confirmed “[z]inc’s toxicity.” Id., at 69a. Clarot called Linschoten to ask whether she would be willing to participate in animal studies that Matrixx was planning, but she declined because her focus was human research. By September 2003, one of Linschoten’s colleagues at the University of Colorado, Dr. Bruce Jafek, had observed 10 patients suffering from anosmia after Zicam use. Linschoten and Jafek planned to present their findings at a meeting of the American Rhinologic Society in a poster presentation entitled “Zicam® Induced Anosmia.” Ibid. (internal quotation marks omitted). The American Rhinologic Society posted their abstract in advance of the meeting. The presentation described in detail a 55-year-old man with previously normal taste and smell who experienced severe burning in his nose, followed immediately by a loss of smell, after using Zicam. It also reported 10 other Zicam users with similar symptoms. Matrixx learned of the doctors’ planned presentation. Clarot sent a letter to Dr. Jafek warning him that he did not have permission to use Matrixx’s name or the names of its products. Dr. Jafek deleted the references to Zicam in the poster before presenting it to the American Rhinologic Society. The following month, two plaintiffs commenced a product liability lawsuit against Matrixx alleging that Zicam had damaged their sense of smell. By the end of the class period on February 6, 2004, nine plaintiffs had filed four lawsuits. Respondents allege that Matrixx made a series of public statements that were misleading in light of the foregoing information. In October 2003, after they had learned of Dr. Jafek’s study and after Dr. Jafek had presented his findings to the American Rhinologic Society, Matrixx stated that Zicam was “ ‘poised for growth in the upcoming cough and cold season’ ” and that the company had “ ‘very strong momentum.’ ”[Footnote 2] Id., at 72a–74a. Matrixx further expressed its expectation that revenues would “ ‘be up in excess of 50% and that earnings, per share for the full year [would] be in the 25 to 30 cent range.’ ” Id., at 74a. In January 2004, Matrixx raised its revenue guidance, predicting an increase in revenues of 80 percent and earnings per share in the 33-to-38-cent range. In its Form 10–Q filed with the SEC in November 2003, Zicam warned of the potential “ ‘material adverse effect’ ” that could result from product liability claims, “ ‘whether or not proven to be valid.’ ” Id., at 75a–76a. It stated that product liability actions could materially affect Matrixx’s “ ‘product branding and goodwill,’ ” leading to reduced customer acceptance.[Footnote 3] Id., at 76a. It did not disclose, however, that two plaintiffs had already sued Matrixx for allegedly causing them to lose their sense of smell. On January 30, 2004, Dow Jones Newswires reported that the Food and Drug Administration (FDA) was “ ‘looking into complaints that an over-the-counter common-cold medicine manufactured by a unit of Matrixx Initiatives, Inc. (MTXX) may be causing some users to lose their sense of smell’ ” in light of at least three product liability lawsuits. Id., at 79a–80a. Matrixx’s stock fell from $13.55 to $11.97 per share after the report. In response, on February 2, Matrixx issued a press release that stated: “All Zicam products are manufactured and marketed according to FDA guidelines for homeopathic medicine. Our primary concern is the health and safety of our customers and the distribution of fac- tual information about our products. Matrixx believes statements alleging that intranasal Zicam products caused anosmia (loss of smell) are completely unfounded and misleading. “In no clinical trial of intranasal zinc gluconate gel products has there been a single report of lost or diminished olfactory function (sense of smell). Rather, the safety and efficacy of zinc gluconate for the treatment of symptoms related to the common cold have been well established in two double-blind, placebo-controlled, randomized clinical trials. In fact, in neither study were there any reports of anosmia related to the use of this compound. The overall incidence of adverse events associated with zinc gluconate was extremely low, with no statistically significant difference between the adverse event rates for the treated and placebo subsets. “A multitude of environmental and biologic influences are known to affect the sense of smell. Chief among them is the common cold. As a result, the population most likely to use cold remedy products is already at increased risk of developing anosmia. Other common causes of olfactory dysfunction include age, nasal and sinus infections, head trauma, anatomical obstructions, and environmental irritants.” Id., at 77a–78a (internal quotation marks omitted). The day after Matrixx issued this press release, its stock price bounced back to $13.40 per share. On February 6, 2004, the end of the class period, Good Morning America, a nationally broadcast morning news program, highlighted Dr. Jafek’s findings. (The complaint does not allege that Matrixx learned of the news story before its broadcast.) The program reported that Dr. Jafek had discovered more than a dozen patients suffering from anosmia after using Zicam. It also noted that four lawsuits had been filed against Matrixx. The price of Matrixx stock plummeted to $9.94 per share that same day. Zicam again issued a press release largely repeating its February 2 statement. On February 19, 2004, Matrixx filed a Form 8–K with the SEC stating that it had “ ‘convened a two-day meeting of physicians and scientists to review current information on smell disorders’ ” in response to Dr. Jafek’s presentation. Id., at 82a. According to the Form 8–K, “ ‘In the opinion of the panel, there is insufficient scientific evidence at this time to determine if zinc gluconate, when used as recommended, affects a person’s ability to smell.’ ” Ibid. A few weeks later, a reporter quoted Matrixx as stating that it would begin conducting “ ‘animal and human studies to further characterize these post-marketing complaints.’ ” Id., at 84a. On the basis of these allegations, respondents claimed that Matrixx violated §10(b) of the Securities Exchange Act and SEC Rule 10b–5 by making untrue statements of fact and failing to disclose material facts necessary to make the statements not misleading in an effort to maintain artificially high prices for Matrixx securities. B Matrixx moved to dismiss respondents’ complaint, arguing that they had failed to plead the elements of a material misstatement or omission and scienter. The District Court granted the motion to dismiss. Relying on In re Carter-Wallace, Inc., Securities Litigation, 220 F. 3d 36 (CA2 2000), it held that respondents had not alleged a “statistically significant correlation between the use of Zicam and anosmia so as to make failure to public[ly] disclose complaints and the University of Colorado study a material omission.” App. to Pet. for Cert. 50a. The District Court similarly agreed that respondents had not stated with particularity facts giving rise to a strong inference of scienter. See 15 U. S. C. A. §78u–4(b)(2)(A) (Feb. 2011 Supp.). It noted that the complaint failed to allege that Matrixx disbelieved its statements about Zicam’s safety or that any of the defendants profited or attempted to profit from Matrixx’s public statements. App. to Pet. for Cert. 52a. The Court of Appeals reversed. 585 F. 3d 1167 (CA9 2009). Noting that “ ‘[t]he determination [of materiality] requires delicate assessments of the inferences a “reasonable shareholder” would draw from a given set of facts and the significance of those inferences to him,’ ” id., at 1178 (quoting Basic Inc. v. Levinson, 485 U. S. 224, 236 (1988); some internal quotation marks omitted; alterations in original), the Court of Appeals held that the District Court had erred in requiring an allegation of statistical significance to establish materiality. It concluded, to the contrary, that the complaint adequately alleged “information regarding the possible link between Zicam and anosmia” that would have been significant to a reasonable investor. 585 F. 3d, at 1179, 1180. Turning to scienter, the Court of Appeals concluded that “[w]ithholding reports of adverse effects of and lawsuits concerning the product responsible for the company’s remarkable sales increase is ‘an extreme departure from the standards of ordinary care,’ ” giving rise to a strong inference of scienter. Id., at 1183. We granted certiorari, 560 U. S. ___ (2010), and we now affirm. II Section 10(b) of the Securities Exchange Act makes it unlawful for any person to “use or employ, in connection with the purchase or sale of any security . . . any manipulative or deceptive device or contrivance in contravention of such rules and regulations as the Commission may prescribe as necessary or appropriate in the public interest or for the protection of investors.” 15 U. S. C. §78j(b). SEC Rule 10b–5 implements this provision by making it unlawful to, among other things, “make any untrue statement of a material fact or to omit to state a material fact necessary in order to make the statements made, in the light of the circumstances under which they were made, not misleading.” 17 CFR §240.10b–5(b). We have implied a private cause of action from the text and purpose of §10(b). See Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U. S. 308, 318 (2007). To prevail on their claim that Matrixx made material misrepresentations or omissions in violation of §10(b) and Rule 10b–5, respondents must prove “(1) a material misrepresentation or omission by the defendant; (2) scienter; (3) a connection between the misrepresentation or omission and the purchase or sale of a security; (4) reliance upon the misrepresentation or omission; (5) economic loss; and (6) loss causation.” Stoneridge Investment Partners, LLC v. Scientific-Atlanta, Inc., 552 U. S. 148, 157 (2008). Matrixx contends that respondents have failed to plead both the element of a material misrepresentation or omission and the element of scienter because they have not alleged that the reports received by Matrixx reflected statistically significant evidence that Zicam caused anosmia. We disagree. A We first consider Matrixx’s argument that “adverse event reports that do not reveal a statistically significant increased risk of adverse events from product use are not material information.” Brief for Petitioners 17 (capitalization omitted). 1 To prevail on a §10(b) claim, a plaintiff must show that the defendant made a statement that was “misleading as to a material fact.”[Footnote 4] Basic, 485 U. S., at 238. In Basic, we held that this materiality requirement is satisfied when there is “ ‘a substantial likelihood that the disclosure of the omitted fact would have been viewed by the reasonable investor as having significantly altered the “total mix” of information made available.’ ” Id., at 231–232 (quoting TSC Industries, Inc. v. Northway, Inc., 426 U. S. 438, 449 (1976)). We were “careful not to set too low a standard of materiality,” for fear that management would “ ‘bury the shareholders in an avalanche of trivial information.’ ” 485 U. S., at 231 (quoting TSC Industries, 426 U. S., at 448–449). Basic involved a claim that the defendant had made misleading statements denying that it was engaged in merger negotiations when it was, in fact, conducting preliminary negotiations. See 485 U. S., at 227–229. The defendant urged a bright-line rule that preliminary merger negotiations are material only once the parties to the negotiations reach an agreement in principle. Id., at 232–233. We observed that “[a]ny approach that designates a single fact or occurrence as always determinative of an inherently fact-specific finding such as materiality, must necessarily be overinclusive or underinclusive.” Id., at 236. We thus rejected the defendant’s proposed rule, explaining that it would “artificially exclud[e] from the definition of materiality information concerning merger discussions, which would otherwise be considered sig- nificant to the trading decision of a reasonable investor.” Ibid. Like the defendant in Basic, Matrixx urges us to adopt a bright-line rule that reports of adverse events[Footnote 5] associated with a pharmaceutical company’s products cannot be material absent a sufficient number of such reports to establish a statistically significant risk that the product is in fact causing the events.[Footnote 6] Absent statistical significance, Matrixx argues, adverse event reports provide only “anecdotal” evidence that “the user of a drug experienced an adverse event at some point during or following the use of that drug.” Brief for Petitioners 17. Accordingly, it contends, reasonable investors would not consider such reports relevant unless they are statistically significant because only then do they “reflect a scientifically reliable basis for inferring a potential causal link between product use and the adverse event.” Id., at 32. As in Basic, Matrixx’s categorical rule would “artificially exclud[e]” information that “would otherwise be considered significant to the trading decision of a reasonable investor.” 485 U. S., at 236. Matrixx’s argument rests on the premise that statistical significance is the only reliable indication of causation. This premise is flawed: As the SEC points out, “medical researchers . . . consider multiple factors in assessing causation.” Brief for United States as Amicus Curiae 12. Statistically significant data are not always available. For example, when an adverse event is subtle or rare, “an inability to obtain a data set of appropriate quality or quantity may preclude a finding of statistical significance.” Id., at 15; see also Brief for Medical Researchers as Amici Curiae 11. Moreover, ethical considerations may prohibit researchers from conducting randomized clinical trials to confirm a suspected causal link for the purpose of obtaining statistically significant data. See id., at 10–11. A lack of statistically significant data does not mean that medical experts have no reliable basis for inferring a causal link between a drug and adverse events. As Matrixx itself concedes, medical experts rely on other evidence to establish an inference of causation. See Brief for Petitioners 44–45, n. 22.[Footnote 7] We note that courts frequently permit expert testimony on causation based on evidence other than statistical significance. See, e.g., Best v. Lowe’s Home Centers, Inc., 563 F. 3d 171, 178 (CA6 2009); Westberry v. Gislaved Gummi AB, 178 F. 3d 257, 263–264 (CA4 1999) (citing cases); Wells v. Ortho Pharmaceutical Corp., 788 F. 2d 741, 744–745 (CA11 1986). We need not consider whether the expert testimony was properly admitted in those cases, and we do not attempt to define here what constitutes reliable evidence of causation. It suffices to note that, as these courts have recognized, “medical professionals and researchers do not limit the data they consider to the results of randomized clinical trials or to statistically significant evidence.” Brief for Medical Researchers as Amici Curiae 31. The FDA similarly does not limit the evidence it considers for purposes of assessing causation and taking regulatory action to statistically significant data. In assessing the safety risk posed by a product, the FDA considers factors such as “strength of the association,” “temporal relationship of product use and the event,” “consistency of findings across available data sources,” “evidence of a dose-response for the effect,” “biologic plausibility,” “seriousness of the event relative to the disease being treated,” “potential to mitigate the risk in the population,” “feasibility of further study using observational or controlled clinical study designs,” and “degree of benefit the product provides, including availability of other therapies.”[Footnote 8] FDA, Guidance for Industry: Good Pharmacovigilance Prac- tices and Pharmacoepidemiologic Assessment 18 (2005) (capitalization omitted), http://www.fda.gov/downloads/ RegulatingInformation/Guidances/UCM126834.pdf (all Internet materials as visited Mar. 17, 2011, and available in Clerk of Court’s case file); see also Brief for United States as Amicus Curiae 19–20 (same); FDA, The Clinical Im- pact of Adverse Event Reporting 6 (1996) (similar), http://www.fda.gov/downloads/safety/MedWatch/UCM168505.pdf. It “does not apply any single metric for determining when additional inquiry or action is necessary, and it certainly does not insist upon ‘statistical significance.’ ” Brief for United States as Amicus Curiae 19. Not only does the FDA rely on a wide range of evidence of causation, it sometimes acts on the basis of evidence that suggests, but does not prove, causation. For example, the FDA requires manufacturers of over-the-counter drugs to revise their labeling “to include a warning as soon as there is reasonable evidence of an association of a serious hazard with a drug; a causal relationship need not have been proved.” 21 CFR §201.80(e). More generally, the FDA may make regulatory decisions against drugs based on postmarketing evidence that gives rise to only a suspicion of causation. See FDA, The Clinical Impact of Adverse Event Reporting, supra, at 7 (“[A]chieving certain proof of causality through postmarketing surveillance is unusual. Attaining a prominent degree of suspicion is much more likely, and may be considered a sufficient basis for regulatory decisions” (footnote omitted)).[Footnote 9] This case proves the point. In 2009, the FDA issued a warning letter to Matrixx stating that “[a] significant and growing body of evidence substantiates that the Zicam Cold Remedy intranasal products may pose a serious risk to consumers who use them.” App. 270a. The letter cited as evidence 130 reports of anosmia the FDA had received, the fact that the FDA had received few reports of anosmia associated with other intranasal cold remedies, and “evidence in the published scientific literature that various salts of zinc can damage olfactory function in animals and humans.” Ibid. It did not cite statistically significant data. Given that medical professionals and regulators act on the basis of evidence of causation that is not statistically significant, it stands to reason that in certain cases reasonable investors would as well. As Matrixx acknowledges, adverse event reports “appear in many forms, including direct complaints by users to manufacturers, reports by doctors about reported or observed patient reactions, more detailed case reports published by doctors in medical journals, or larger scale published clinical studies.” Brief for Petitioners 17. As a result, assessing the materiality of adverse event reports is a “fact-specific” inquiry, Basic, 485 U. S., at 236, that requires consideration of the source, content, and context of the reports. This is not to say that statistical significance (or the lack thereof) is irrelevant—only that it is not dispositive of every case. Application of Basic’s “total mix” standard does not mean that pharmaceutical manufacturers must dis- close all reports of adverse events. Adverse event reports are daily events in the pharmaceutical industry; in 2009, the FDA entered nearly 500,000 such reports into its reporting system, see FDA, Reports Received and Reports Entered in AERS by Year (as of Mar. 31, 2010), http://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation /Surveillance/ AdverseDrugEffects/ ucm070434. htm. The fact that a user of a drug has suffered an adverse event, standing alone, does not mean that the drug caused that event. See FDA, Annual Adverse Drug Experience Report: 1996, p. 2 (1997), http://drugand devicelaw.net/Annual%20Adverse%20Drug%20Experience %20Report%201996.pdf. The question remains whether a reasonable investor would have viewed the nondisclosed information “ ‘as having significantly altered the “total mix” of information made available.’ ” Basic, 485 U. S., at 232 (quoting TSC Industries, 426 U. S., at 449; emphasis added). For the reasons just stated, the mere existence of reports of adverse events—which says nothing in and of itself about whether the drug is causing the adverse events—will not satisfy this standard. Something more is needed, but that something more is not limited to statistical significance and can come from “the source, content, and context of the reports,” supra, at 15. This contextual inquiry may reveal in some cases that reasonable investors would have viewed reports of adverse events as material even though the reports did not provide statistically significant evidence of a causal link.[Footnote 10] Moreover, it bears emphasis that §10(b) and Rule 10b–5(b) do not create an affirmative duty to disclose any and all material information. Disclosure is required under these provisions only when necessary “to make . . . statements made, in the light of the circumstances under which they were made, not misleading. 17 CFR §240.10b–5(b); see also Basic, 485 U. S., at 239, n. 17 (“Silence, absent a duty to disclose, is not misleading under Rule 10b–5”). Even with respect to information that a reasonable investor might consider material, companies can control what they have to disclose under these provisions by controlling what they say to the market. 2 Applying Basic’s “total mix” standard in this case, we conclude that respondents have adequately pleaded materiality. This is not a case about a handful of anecdotal reports, as Matrixx suggests. Assuming the complaint’s allegations to be true, as we must, Matrixx received information that plausibly indicated a reliable causal link between Zicam and anosmia. That information included reports from three medical professionals and researchers about more than 10 patients who had lost their sense of smell after using Zicam. Clarot told Linschoten that Matrixx had received additional reports of anosmia. (In addition, during the class period, nine plaintiffs commenced four product liability lawsuits against Matrixx alleging a causal link between Zicam use and anosmia.)[Footnote 11] Further, Matrixx knew that Linschoten and Dr. Jafek had presented their findings about a causal link between Zicam and anosmia to a national medical conference devoted to treatment of diseases of the nose.[Footnote 12] Their presentation described a patient who experienced severe burning in his nose, followed immediately by a loss of smell, after using Zicam—suggesting a temporal relationship between Zicam use and anosmia. Critically, both Dr. Hirsch and Linschoten had also drawn Matrixx’s attention to previous studies that had demonstrated a biological causal link between intranasal application of zinc and anosmia.[Footnote 13] Before his conversation with Linschoten, Clarot, Matrixx’s vice president of research and development, was seemingly unaware of these studies, and the complaint suggests that, as of the class period, Matrixx had not conducted any research of its own relating to anosmia. See, e.g., App. 84a (referencing a press report, issued after the end of the class period, noting that Matrixx said it would begin conducting “ ‘animal and human studies to further characterize these post-marketing complaints’ ”). Accordingly, it can reasonably be inferred from the complaint that Matrixx had no basis for rejecting Dr. Jafek’s findings out of hand. We believe that these allegations suffice to “raise a reasonable expectation that discovery will reveal evidence” satisfying the materiality requirement, Bell Atlantic Corp. v. Twombly, 550 U. S. 544, 556 (2007), and to “allo[w] the court to draw the reasonable inference that the defendant is liable for the misconduct alleged,” Iqbal, 556 U. S., at ___ (slip op., at 14). The information provided to Matrixx by medical experts revealed a plausible causal relationship between Zicam Cold Remedy and anosmia. Consumers likely would have viewed the risk associated with Zicam (possible loss of smell) as substantially outweighing the benefit of using the product (alleviating cold symptoms), particularly in light of the existence of many alternative products on the market. Importantly, Zicam Cold Remedy allegedly accounted for 70 percent of Matrixx’s sales. Viewing the allegations of the complaint as a whole, the complaint alleges facts suggesting a significant risk to the commercial viability of Matrixx’s leading product. It is substantially likely that a reasonable investor would have viewed this information “ ‘as having significantly altered the “total mix” of information made available.’ ” Basic, 485 U. S., at 232 (quoting TSC Industries, 426 U. S., at 449). Matrixx told the market that revenues were going to rise 50 and then 80 percent. Assuming the complaint’s allegations to be true, however, Matrixx had information indicating a significant risk to its leading revenue-generating product. Matrixx also stated that reports indicating that Zicam caused anosmia were “ ‘completely unfounded and misleading’ ” and that “ ‘the safety and efficacy of zinc gluconate for the treatment of symptoms related to the common cold have been well established.’ ” App. 77a–78a. Importantly, however, Matrixx had evidence of a biological link between Zicam’s key ingredient and anosmia, and it had not conducted any studies of its own to disprove that link. In fact, as Matrixx later revealed, the scientific evidence at that time was “ ‘insufficient . . . to determine if zinc gluconate, when used as recommended, affects a person’s ability to smell.’ ” Id., at 82a. Assuming the facts to be true, these were material facts “necessary in order to make the statements made, in the light of the circumstances under which they were made, not misleading.” 17 CFR §240.10b–5(b). We therefore affirm the Court of Appeals’ holding that respondents adequately pleaded the element of a material misrepresentation or omission. B Matrixx also argues that respondents failed to allege facts plausibly suggesting that it acted with the required level of scienter. “To establish liability under §10(b) and Rule 10b–5, a private plaintiff must prove that the defendant acted with scienter, ‘a mental state embracing intent to deceive, manipulate, or defraud.’ ” Tellabs, 551 U. S., at 319 (quoting Ernst & Ernst v. Hochfelder, 425 U. S. 185, 193–194, and n. 12 (1976)). We have not decided whether recklessness suffices to fulfill the scienter requirement. See Tellabs, 551 U. S., at 319, n. 3. Because Matrixx does not challenge the Court of Appeals’ holding that the scienter requirement may be satisfied by a showing of “deliberate recklessness,” see 585 F. 3d, at 1180 (internal quotation marks omitted), we assume, without deciding, that the standard applied by the Court of Appeals is sufficient to establish scienter.[Footnote 14] Under the PSLRA, a plaintiff must “state with par- ticularity facts giving rise to a strong inference that the defendant acted with the required state of mind.” 15 U. S. C. A. §78u–4(b)(2)(A) (Feb. 2011 Supp.). This standard requires courts to take into account “plausible opposing inferences.” Tellabs, 551 U. S., at 323. A complaint adequately pleads scienter under the PSLRA “only if a reasonable person would deem the inference of scienter cogent and at least as compelling as any opposing inference one could draw from the facts alleged.” Id., at 324. In making this determination, the court must review “all the allegations holistically.” Id., at 326. The absence of a motive allegation, though relevant, is not dispositive. Id., at 325. Matrixx argues, in summary fashion, that because respondents do not allege that it knew of statistically significant evidence of causation, there is no basis to consider the inference that it acted recklessly or knowingly to be at least as compelling as the alternative inferences. “Rather,” it argues, “the most obvious inference is that petitioners did not disclose the [reports] simply because petitioners believed they were far too few . . . to indicate anything meaningful about adverse reactions to use of Zicam.” Brief for Petitioners 49. Matrixx’s proposed bright-line rule requiring an allegation of statistical significance to establish a strong inference of scienter is just as flawed as its approach to materiality. The inference that Matrixx acted recklessly (or intentionally, for that matter) is at least as compelling, if not more compelling, than the inference that it simply thought the reports did not indicate anything meaningful about adverse reactions. According to the complaint, Matrixx was sufficiently concerned about the information it received that it informed Linschoten that it had hired a consultant to review the product, asked Linschoten to participate in animal studies, and convened a panel of physicians and scientists in response to Dr. Jafek’s presentation. It successfully prevented Dr. Jafek from using Zicam’s name in his presentation on the ground that he needed Matrixx’s permission to do so. Most significantly, Matrixx issued a press release that suggested that studies had confirmed that Zicam does not cause anosmia when, in fact, it had not conducted any studies relating to anosmia and the scientific evidence at that time, according to the panel of scientists, was insufficient to determine whether Zicam did or did not cause anosmia.[Footnote 15] These allegations, “taken collectively,” give rise to a “cogent and compelling” inference that Matrixx elected not to disclose the reports of adverse events not because it believed they were meaningless but because it understood their likely effect on the market. Tellabs, 551 U. S., at 323, 324. “[A] reasonable person” would deem the inference that Matrixx acted with deliberate recklessness (or even intent) “at least as compelling as any opposing inference one could draw from the facts alleged.” Id., at 324. We conclude, in agreement with the Court of Appeals, that respondents have adequately pleaded scienter. Whether respondents can ultimately prove their allegations and establish scienter is an altogether different question. * * * For the reasons stated, the judgment of the Court of Appeals for the Ninth Circuit is Affirmed. Footnote 1 According to the complaint, Matrixx securities were traded on the NASDAQ National Market. App. 99a. Footnote 2 At oral argument, counsel for the United States, which submitted an amicus curiae brief in support of respondents, suggested that some of these statements might qualify as nonactionable “puffery.” Tr. of Oral Arg. 51–52. This question is not before us, as Matrixx has not advanced such an argument. Footnote 3 Respondents also allege that Matrixx falsely reported its financial results in the Form 10–Q by failing to reserve for or disclose potential liability, in violation of Generally Accepted Accounting Principles. The Court of Appeals did not rely on these allegations. Footnote 4 Under the Private Securities Litigation Reform Act of 1995 (PSLRA), when a plaintiff’s claim is based on alleged misrepresentations or omissions of a material fact, “the complaint shall specify each statement alleged to have been misleading, [and] the reason or reasons why the statement is misleading.” 15 U. S. C. §78u–4(b)(1). Footnote 5 The FDA defines an “[a]dverse drug experience” as “[a]ny adverse event associated with the use of a drug in humans, whether or not considered drug related.” 21 CFR §314.80(a) (2010). Federal law imposes certain obligations on pharmaceutical manufacturers to report adverse events to the FDA. During the class period, manufacturers of over-the-counter drugs such as Zicam Cold Remedy had no obligation to report adverse events to the FDA. In 2006, Congress enacted legislation to require manufacturers of over-the-counter drugs to report any “serious adverse event” to the FDA within 15 business days. See 21 U. S. C. §§379aa(b), (c). Footnote 6 “A study that is statistically significant has results that are unlikely to be the result of random error . . . .” Federal Judicial Center, Reference Manual on Scientific Evidence 354 (2d ed. 2000). To test for significance, a researcher develops a “null hypothesis”—e.g., the assertion that there is no relationship between Zicam use and anosmia. See id., at 122. The researcher then calculates the probability of obtaining the observed data (or more extreme data) if the null hypothesis is true (called the p-value). Ibid. Small p-values are evidence that the null hypothesis is incorrect. See ibid. Finally, the researcher compares the p-value to a preselected value called the significance level. Id., at 123. If the p-value is below the preselected value, the difference is deemed “significant.” Id., at 124. Footnote 7 Matrixx and its amici list as relevant factors the strength of the association between the drug and the adverse effects; a temporal relationship between exposure and the adverse event; consistency across studies; biological plausibility; consideration of alternative explanations; specificity (i.e., whether the specific chemical is associated with the specific disease); the dose-response relationship; and the clinical and pathological characteristics of the event. Brief for Petitioners 44–45, n. 22; Brief for Consumer Healthcare Products Assn. et al. as Amici Curiae 12–13. These factors are similar to the factors the FDA considers in taking action against pharmaceutical products. See infra, at 13–14. Footnote 8 See also n. 7, supra. Footnote 9 See also GAO, M. Crosse et al., Drug Safety: Improvement Needed in FDA’s Postmarket Decision-making and Oversight Process 7 (GAO–06–402, 2006) (“If FDA has information that a drug on the market may pose a significant health risk to consumers, it weighs the effect of the adverse events against the benefit of the drug to determine what actions, if any, are warranted. This decision-making process is complex and encompasses many factors, such as the medical importance and utility of the drug, the drug’s extent of usage, the severity of the disease being treated, the drug’s efficacy in treating this disease, and the availability of other drugs to treat the same disorder”), http://www.gao.gov/new.items/d06402.pdf; Federal Judicial Center, supra n. 6, at 33 (“[R]isk assessors may pay heed to any evidence that points to a need for caution, rather than assess the likelihood that a causal relationship in a specific case is more likely than not”). Footnote 10 We note that our conclusion accords with views of the SEC, as expressed in an amicus curiae brief filed in this case. See Brief for United States as Amicus Curiae 11–12; see also TSC Industries, Inc. v. Northway, Inc., 426 U. S. 438, 449, n. 10 (1976) (“[T]he SEC’s view of the proper balance between the need to insure adequate disclosure and the need to avoid the adverse consequences of setting too low a threshold for civil liability is entitled to consideration”). Footnote 11 It is unclear whether these plaintiffs were the same individuals whose symptoms were reported by the medical professionals. Footnote 12 Matrixx contends that Dr. Jafek and Linschoten’s study was not reliable because they did not sufficiently rule out the common cold as a cause for their patients’ anosmia. We note that the complaint alleges that, in one instance, a consumer who did not have a cold lost his sense of smell after using Zicam. More importantly, to survive a motion to dismiss, respondents need only allege “enough facts to state a claim to relief that is plausible on its face.” Bell Atlantic Corp. v. Twombly, 550 U. S. 544, 570 (2007). For all the reasons we state in the opinion, respondents’ allegations plausibly suggest that Dr. Jafek and Linschoten’s conclusions were based on reliable evidence of a causal link between Zicam and anosmia. Footnote 13 Matrixx contends that these studies are not reliable evidence of causation because the studies used zinc sulfate, whereas the active ingredient in Matrixx is zinc gluconate. Respondents’ complaint, however, alleges that the studies confirmed the toxicity of “zinc.” App. 68a. Matrixx further contends that studies relating to fish cannot reliably prove causation with respect to humans. The complaint references several studies, however, only one of which involved fish. In any event, the existence of the studies suggests a plausible biological link between zinc and anosmia, which, in combination with the other allegations, is sufficient to survive a motion to dismiss. Footnote 14 Under the PSLRA, if the alleged misstatement or omission is a “forward-looking statement,” the required level of scienter is “actual knowledge.” 15 U. S. C. §78u–5(c)(1)(B). Matrixx has not argued that the statements or omissions here are “forward-looking statement[s].” Footnote 15 One of Matrixx’s amici argues that “the most cogent inference regarding Matrixx’s state of mind is that it delayed releasing information regarding anosmia complaints in order to provide itself an opportunity to carefully review all evidence regarding any link between Zicam and anosmia.” Brief for Washington Legal Foundation as Amicus Curiae 26. We do not doubt that this may be the most cogent inference in some cases. Here, however, the misleading nature of Matrixx’s press release is sufficient to render the inference of scienter at least as compelling as the inference suggested by amicus.

562.US.44

Petitioners (hereinafter Mayo) offer residency programs to doctors who have graduated from medical school and seek additional instruction in a chosen specialty. Those programs train doctors primarily through hands-on experience. Although residents are required to take part in formal educational activities, these doctors generally spend the bulk of their time—typically 50 to 80 hours a week—caring for patients. Mayo pays its residents annual “stipends” of over $40,000 and also provides them with health insurance, malpractice insurance, and paid vacation time. The Federal Insurance Contributions Act (FICA) requires employees and employers to pay taxes on all “wages” employees receive, 26 U. S. C. §§3101(a), 3111(a), and defines “wages” to include “all remuneration for employment,” §3121(a). FICA defines “employment” as “any service … performed … by an employee for the person employing him,” §3121(b), but excludes from taxation any “service performed in the employ of … a school, college, or university … if such service is performed by a student who is enrolled and regularly attending classes at [the school],” §3121(b)(10). Since 1951, the Treasury Department has construed the student exception to exempt from taxation students who work for their schools “as an incident to and for the purpose of pursuing a course of study.” 16 Fed. Reg. 12474. In 2004, the Department issued regulations providing that “[t]he services of a full-time employee”—which includes an employee normally scheduled to work 40 hours or more per week—“are not incident to and for the purpose of pursuing a course of study.” 26 CFR §31.3121(b)(10)–2(d)(3)(iii). The Department explained that this analysis “is not affected by the fact that the services … may have an educational, instructional, or training aspect.” Ibid. The rule offers as an example a medical resident whose normal schedule requires him to perform services 40 or more hours per week, and concludes that the resident is not a student. Mayo filed suit asserting that this rule was invalid, and the District Court agreed. It found the full-time employee rule inconsistent with §3121’s unambiguous text and concluded that the factors governing this Court’s analysis in National Muffler Dealers Assn., Inc. v. United States, 440 U. S. 472, also indicated that the rule was invalid. The Eighth Circuit reversed. Applying Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837, the Court of Appeals concluded that the Department’s regulation was a permissible interpretation of an ambiguous statute. Held: The Treasury Department’s full-time employee rule is a reasonable construction of §3121(b)(10). Pp. 6–15. (a) Under Chevron’s two-part framework, the Court first asks whether Congress has “directly addressed the precise question at issue.” 467 U. S., at 842-843. Congress has not done so here; the statute does not define “student” or otherwise attend to the question whether medical residents are subject to FICA. Pp. 6–7. (b) The parties debate whether the Court should next apply Chevron step two or the multi-factor analysis used to review a tax regulation in National Muffler. Absent a justification to do so, this Court is not inclined to apply a less deferential framework to evaluate Treasury Department regulations than it uses to review rules adopted by any other agency. The Court has “[r]ecogniz[ed] the importance of maintaining a uniform approach to judicial review of administrative action.” Dickinson v. Zurko, 527 U. S. 150, 154. And the principles underlying Chevron apply with full force in the tax context. Chevron recognized that an agency’s power “ ‘to administer a congressionally created … program necessarily requires the formulation of policy and the making of rules to fill any gap left … by Congress.’ ” 467 U. S., at 843. Filling gaps in the Internal Revenue Code plainly requires the Treasury Department to make interpretive choices for statutory implementation at least as complex as the ones made by other agencies in administering their statutes. It is true that the full-time employee rule, like the rule at issue in National Muffler, was promulgated under the Department’s general authority to “prescribe all needful rules and regulations for the enforcement” of the Internal Revenue Code. 26 U. S. C. §7805(a). It is also true that this Court, in opinions predating Chevron, stated that it owed less deference to a rule adopted under that general grant of authority than it would afford rules issued pursuant to more specific grants. See Rowan Cos. v. United States, 452 U. S. 247, 253; United States v. Vogel Fertilizer Co., 455 U. S. 16, 24. Since then, however, the Court has found Chevron deference appropriate “when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law, and that the agency interpretation claiming deference was promulgated in the exercise of that authority.” United States v. Mead Corp., 533 U. S. 218, 226–227. Chevron and Mead provide the appropriate framework for evaluating the full-time employee rule. The Department issued the rule pursuant to an explicit authorization to prescribe needful rules and regulations, and only after notice-and-comment procedures. The Court has recognized these to be good indicators of a rule meriting Chevron deference, Mead, 533 U. S., at 229–231. Pp. 7–12. (c) The rule easily satisfies Chevron’s second step. Mayo accepts the Treasury Department’s determination that an individual may not qualify for the student exception unless the educational aspect of his relationship with his employer predominates over the service aspect of that relationship, but objects to the Department’s conclusion that residents working more than 40 hours per week categorically cannot satisfy that requirement. Mayo argues that the Treasury Department should be required to engage in a case-by-case inquiry into what each employee does and why he does it, and that the Department has arbitrarily distinguished between hands-on training and classroom instruction. But regulation, like legislation, often requires drawing lines. The Department reasonably sought to distinguish between workers who study and students who work. Focusing on the hours spent working and those spent in studies is a sensible way to accomplish that goal. The Department thus has drawn a distinction between education and service, not between classroom instruction and hands-on training. The Treasury Department also reasonably concluded that its full-time employee rule would “improve administrability,” 69 Fed. Reg. 76405, and thereby “has avoided the wasteful litigation and continuing uncertainty that would inevitably accompany [a] case-by-case approach” like the one Mayo advocates, United States v. Correll, 389 U. S. 299, 302. Moreover, the rule reasonably takes into account the Social Security Administration’s concern that exempting residents from FICA would deprive them and their families of vital social security disability and survivorship benefits. Pp. 12–15. 568 F. 3d 675, affirmed. Roberts, C. J., delivered the opinion of the Court, in which all other Members joined, except Kagan, J., who took no part in the consideration or decision of the case.

Nearly all Americans who work for wages pay taxes on those wages under the Federal Insurance Contributions Act (FICA), which Congress enacted to collect funds for Social Security. The question presented in this case is whether doctors who serve as medical residents are properly viewed as “student[s]” whose service Congress has exempted from FICA taxes under 26 U. S. C. §3121(b)(10). I A Most doctors who graduate from medical school in the United States pursue additional education in a specialty to become board certified to practice in that field. Petitioners Mayo Foundation for Medical Education and Research, Mayo Clinic, and the Regents of the University of Minnesota (collectively Mayo) offer medical residency programs that provide such instruction. Mayo’s residency programs, which usually last three to five years, train doctors primarily through hands-on experience. Residents often spend between 50 and 80 hours a week caring for patients, typically examining and diagnosing them, prescribing medication, recommending plans of care, and performing certain procedures. Residents are generally supervised in this work by more senior residents and by faculty members known as attending physicians. In 2005, Mayo paid its residents annual “stipends” ranging between $41,000 and $56,000 and provided them with health insurance, malpractice insurance, and paid vacation time. Mayo residents also take part in “a formal and structured educational program.” Brief for Petitioners 5 (internal quotation marks omitted). Residents are assigned textbooks and journal articles to read and are expected to attend weekly lectures and other conferences. Residents also take written exams and are evaluated by the attending faculty physicians. But the parties do not dispute that the bulk of residents’ time is spent caring for patients. B Through the Social Security Act and related legislation, Congress has created a comprehensive national insurance system that provides benefits for retired workers, disabled workers, unemployed workers, and their families. See United States v. Lee, 455 U. S. 252, 254, 258, and nn. 1, 7 (1982). Congress funds Social Security by taxing both employers and employees under FICA on the wages employees earn. See 26 U. S. C. §3101(a) (tax on employees); §3111(a) (tax on employers). Congress has defined “wages” broadly, to encompass “all remuneration for employment.” §3121(a) (2006 ed. and Supp. III). The term “employment” has a similarly broad reach, extending to “any service, of whatever nature, performed … by an employee for the person employing him.” §3121(b). Congress has, however, exempted certain categories of service and individuals from FICA’s demands. As relevant here, Congress has excluded from taxation “service performed in the employ of … a school, college, or university … if such service is performed by a student who is enrolled and regularly attending classes at such school, college, or university.” §3121(b)(10) (2006 ed.). The Social Security Act, which governs workers’ eligibility for benefits, contains a corresponding student exception materially identical to §3121(b)(10). 42 U. S. C. §410(a)(10). Since 1951, the Treasury Department has applied the student exception to exempt from taxation students who work for their schools “as an incident to and for the purpose of pursuing a course of study” there. 16 Fed. Reg. 12474 (adopting Treas. Regs. 127, §408.219(c)); see Treas. Reg. §31.3121(b)(10)–2(d), 26 CFR §31.3121(b)(10)–2(d) (2010). Until 2005, the Department determined whether an individual’s work was “incident to” his studies by performing a case-by-case analysis. The primary considerations in that analysis were the number of hours worked and the course load taken. See, e.g., Rev. Rul. 78–17, 1978–1 Cum. Bull. 307 (services of individual “employed on a full-time basis” with a part-time course load are “not incident to and for the purpose of pursuing a course of study”). For its part, the Social Security Administration (SSA) also articulated in its regulations a case-by-case approach to the corresponding student exception in the Social Security Act. See 20 CFR §404.1028(c) (1998). The SSA has, however, “always held that resident physicians are not students.” SSR 78–3, 1978 Cum. Bull. 55–56. In 1998, the Court of Appeals for the Eighth Circuit held that the SSA could not categorically exclude residents from student status, given that its regulations provided for a case-by-case approach. See Minnesota v. Apfel, 151 F. 3d 742, 747–748. Following that decision, the Internal Revenue Service received more than 7,000 claims seeking FICA tax refunds on the ground that medical residents qualified as students under §3121(b)(10) of the Internal Revenue Code. 568 F. 3d 675, 677 (CA8 2009). Facing that flood of claims, the Treasury Department “determined that it [wa]s necessary to provide additional clarification of the ter[m]” “student” as used in §3121(b)(10), particularly with respect to individuals who perform “services that are in the nature of on the job training.” 69 Fed. Reg. 8605 (2004). The Department proposed an amended rule for comment and held a public hearing on it. See id., at 76405. On December 21, 2004, the Department adopted an amended rule prescribing that an employee’s service is “incident” to his studies only when “[t]he educational aspect of the relationship between the employer and the employee, as compared to the service aspect of the relationship, [is] predominant.” Id., at 76408; Treas. Reg. §31.3121(b)(10)–2(d)(3)(i), 26 CFR §31.3121(b)(10)–2(d)(3)(i) (2005). The rule categorically provides that “[t]he services of a full-time employee”—as defined by the employer’s policies, but in any event including any employee normally scheduled to work 40 hours or more per week—“are not incident to and for the purpose of pursuing a course of study.” 69 Fed. Reg. 76408; Treas. Reg. §31.3121(b)(10)–2(d)(3)(iii), 26 CFR §31.3121(b)(10)–2(d)(3)(iii) (the full-time employee rule). The amended provision clarifies that the Department’s analysis “is not affected by the fact that the services performed … may have an educational, instructional, or training aspect.” Ibid. The rule also includes as an example the case of “Employee E,” who is employed by “University V” as a medical resident. 69 Fed. Reg. 76409; Treas. Reg. §31.3121(b)(10)–2(e), 26 CFR §31.3121(b)(10)–2(e) (Example 4). Because Employee E’s “normal work schedule calls for [him] to perform services 40 or more hours per week,” the rule provides that his service is “not incident to and for the purpose of pursuing a course of study,” and he accordingly is not an exempt “student” under §3121(b)(10). 69 Fed. Reg. 76409, 76410; Treas. Reg. §31.3121(b)(10)–2(e), 26 CFR §31.3121(b)(10)–2(e) (Example 4). C After the Department promulgated the full-time employee rule, Mayo filed suit seeking a refund of the money it had withheld and paid on its residents’ stipends during the second quarter of 2005. 503 F. Supp. 2d 1164, 1166–1167 (Minn. 2007); Regents of Univ. of Minn. v. United States, No. 06–5084 (D Minn., Apr. 1, 2008), App. to Pet. for Cert. C–47a. Mayo asserted that its residents were exempt under §3121(b)(10) and that the Treasury Department’s full-time employee rule was invalid. The District Court granted Mayo’s motion for summary judgment. The court held that the full-time employee rule is inconsistent with the unambiguous text of §3121, which the court understood to dictate that “an employee is a ‘student’ so long as the educational aspect of his service predominates over the service aspect of the relationship with his employer.” 503 F. Supp. 2d, at 1175. The court also determined that the factors governing this Court’s analysis of regulations set forth in National Muffler Dealers Assn., Inc. v. United States, 440 U. S. 472 (1979), “indicate that the full-time employee exception is invalid.” 503 F. Supp. 2d, at 1176; see App. to Pet. for Cert. C–54a. The Government appealed, and the Court of Appeals reversed. 568 F. 3d 675. Applying our opinion in Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837 (1984), the Court of Appeals concluded that “the statute is silent or ambiguous on the question whether a medical resident working for the school full-time is a ‘student’ ” for purposes of §3121(b)(10), and that the Department’s amended regulation “is a permissible interpretation of the statut[e].” 568 F. 3d, at 679–680, 683. We granted Mayo’s petition for certiorari. 560 U. S. ___ (2010). II A We begin our analysis with the first step of the two-part framework announced in Chevron, supra, at 842–843, and ask whether Congress has “directly addressed the precise question at issue.” We agree with the Court of Appeals that Congress has not done so. The statute does not define the term “student,” and does not otherwise attend to the precise question whether medical residents are subject to FICA. See 26 U. S. C. §3121(b)(10). Mayo nonetheless contends that the Treasury Department’s full-time employee rule must be rejected under Chevron step one. Mayo argues that the dictionary definition of “student”—one “who engages in ‘study’ by applying the mind ‘to the acquisition of learning, whether by means of books, observation, or experiment’ ”—plainly encompasses residents. Brief for Petitioners 22 (quoting Oxford Universal Dictionary 2049–2050 (3d ed. 1955)). And, Mayo adds, residents are not excluded from that category by the only limitation on students Congress has imposed under the statute—that they “be ‘enrolled and regularly attending classes at [a] school.’ ” Brief for Petitioners 22 (quoting 26 U. S. C. §3121(b)(10)). Mayo’s reading does not eliminate the statute’s ambiguity as applied to working professionals. In its reply brief, Mayo acknowledges that a full-time professor taking evening classes—a person who presumably would satisfy the statute’s class-enrollment requirement and apply his mind to learning—could be excluded from the exemption and taxed because he is not “ ‘predominant[ly]’ ” a student. Reply Brief for Petitioners 7. Medical residents might likewise be excluded on the same basis; the statute itself does not resolve the ambiguity. The District Court interpreted §3121(b)(10) as unambiguously foreclosing the Department’s rule by mandating that an employee be deemed “a ‘student’ so long as the educational aspect of his service predominates over the service aspect of the relationship with his employer.” 503 F. Supp. 2d, at 1175. We do not think it possible to glean so much from the little that §3121 provides. In any event, the statutory text still would offer no insight into how Congress intended predominance to be determined or whether Congress thought that medical residents would satisfy the requirement. To the extent Congress has specifically addressed medical residents in §3121, moreover, it has expressly excluded these doctors from exemptions they might otherwise invoke. See 26 U. S. C. §§3121(b)(6)(B), (7)(C)(ii) (excluding medical residents from exemptions available to employees of the District of Columbia and the United States). That choice casts doubt on any claim that Congress specifically intended to insulate medical residents from FICA’s reach in the first place. In sum, neither the plain text of the statute nor the District Court’s interpretation of the exemption “speak[s] with the precision necessary to say definitively whether [the statute] applies to” medical residents. United States v. Eurodif S. A., 555 U. S. ___, ___ (2009) (slip op., at 13). B In the typical case, such an ambiguity would lead us inexorably to Chevron step two, under which we may not disturb an agency rule unless it is “ ‘arbitrary or capricious in substance, or manifestly contrary to the statute.’ ” Household Credit Services, Inc. v. Pfennig, 541 U. S. 232, 242 (2004) (quoting United States v. Mead Corp., 533 U. S. 218, 227 (2001)). In this case, however, the parties disagree over the proper framework for evaluating an ambiguous provision of the Internal Revenue Code. Mayo asks us to apply the multi-factor analysis we used to review a tax regulation in National Muffler, 440 U. S. 472. There we explained: “A regulation may have particular force if it is a substantially contemporaneous construction of the statute by those presumed to have been aware of congressional intent. If the regulation dates from a later period, the manner in which it evolved merits inquiry. Other relevant considerations are the length of time the regulation has been in effect, the reliance placed on it, the consistency of the Commissioner’s interpretation, and the degree of scrutiny Congress has devoted to the regulation during subsequent re-enactments of the statute.” Id., at 477. The Government, on the other hand, contends that the National Muffler standard has been superseded by Chevron. The sole question for the Court at step two under the Chevron analysis is “whether the agency’s answer is based on a permissible construction of the statute.” 467 U. S., at 843. Since deciding Chevron, we have cited both National Muffler and Chevron in our review of Treasury Department regulations. See, e.g., United States v. Cleveland Indians Baseball Co., 532 U. S. 200, 219 (2001) (citing National Muffler); Cottage Savings Assn. v. Commissioner, 499 U. S. 554, 560–561 (1991) (same); United States v. Boyle, 469 U. S. 241, 246, n. 4 (1985) (citing Chevron); see also Atlantic Mut. Ins. Co. v. Commissioner, 523 U. S. 382, 387, 389 (1998) (citing Chevron and Cottage Savings). Although we have not thus far distinguished between National Muffler and Chevron, they call for different analyses of an ambiguous statute. Under National Muffler, for example, a court might view an agency’s interpretation of a statute with heightened skepticism when it has not been consistent over time, when it was promulgated years after the relevant statute was enacted, or because of the way in which the regulation evolved. 440 U. S., at 477. The District Court in this case cited each of these factors in rejecting the Treasury Department’s rule, noting in particular that the regulation had been promulgated after an adverse judicial decision. See 503 F. Supp. 2d, at 1176; see also Brief for Petitioners 41–44 (relying on the same considerations). Under Chevron, in contrast, deference to an agency’s interpretation of an ambiguous statute does not turn on such considerations. We have repeatedly held that “[a]gency inconsistency is not a basis for declining to analyze the agency’s interpretation under the Chevron framework.” National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U. S. 967, 981 (2005); accord, Eurodif S. A., supra, at ___ (slip op., at 10). We have instructed that “neither antiquity nor contemporaneity with [a] statute is a condition of [a regulation’s] validity.” Smiley v. Citibank (South Dakota), N. A., 517 U. S. 735, 740 (1996). And we have found it immaterial to our analysis that a “regulation was prompted by litigation.” Id., at 741. Indeed, in United Dominion Industries, Inc. v. United States, 532 U. S. 822, 838 (2001), we expressly invited the Treasury Department to “amend its regulations” if troubled by the consequences of our resolution of the case. Aside from our past citation of National Muffler, Mayo has not advanced any justification for applying a less deferential standard of review to Treasury Department regulations than we apply to the rules of any other agency. In the absence of such justification, we are not inclined to carve out an approach to administrative review good for tax law only. To the contrary, we have expressly “[r]ecogniz[ed] the importance of maintaining a uniform approach to judicial review of administrative action.” Dickinson v. Zurko, 527 U. S. 150, 154 (1999). See, e.g., Skinner v. Mid-America Pipeline Co., 490 U. S. 212, 222–223 (1989) (declining to apply “a different and stricter nondelegation doctrine in cases where Congress delegates discretionary authority to the Executive under its taxing power”). The principles underlying our decision in Chevron apply with full force in the tax context. Chevron recognized that “[t]he power of an administrative agency to administer a congressionally created … program necessarily requires the formulation of policy and the making of rules to fill any gap left, implicitly or explicitly, by Congress.” 467 U. S., at 843 (internal quotation marks omitted). It acknowledged that the formulation of that policy might require “more than ordinary knowledge respecting the matters subjected to agency regulations.” Id., at 844 (internal quotation marks omitted). Filling gaps in the Internal Revenue Code plainly requires the Treasury Department to make interpretive choices for statutory implementation at least as complex as the ones other agencies must make in administering their statutes. Cf. Bob Jones Univ. v. United States, 461 U. S. 574, 596 (1983) (“[I]n an area as complex as the tax system, the agency Congress vests with administrative responsibility must be able to exercise its authority to meet changing conditions and new problems”). We see no reason why our review of tax regulations should not be guided by agency expertise pursuant to Chevron to the same extent as our review of other regulations. As one of Mayo’s amici points out, however, both the full-time employee rule and the rule at issue in National Muffler were promulgated pursuant to the Treasury Department’s general authority under 26 U. S. C. §7805(a) to “prescribe all needful rules and regulations for the enforcement” of the Internal Revenue Code. See Brief for Carlton M. Smith 4–7. In two decisions predating Chevron, this Court stated that “we owe the [Treasury Department’s] interpretation less deference” when it is contained in a rule adopted under that “general authority” than when it is “issued under a specific grant of authority to define a statutory term or prescribe a method of executing a statutory provision.” Rowan Cos. v. United States, 452 U. S. 247, 253 (1981); United States v. Vogel Fertilizer Co., 455 U. S. 16, 24 (1982) (quoting Rowan). Since Rowan and Vogel were decided, however, the administrative landscape has changed significantly. We have held that Chevron deference is appropriate “when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law, and that the agency interpretation claiming deference was promulgated in the exercise of that authority.” Mead, 533 U. S., at 226–227. Our inquiry in that regard does not turn on whether Congress’s delegation of authority was general or specific. For example, in National Cable & Telecommunications Assn., supra, we held that the Federal Communications Commission was delegated “the authority to promulgate binding legal rules” entitled to Chevron deference under statutes that gave the Commission “the authority to ‘execute and enforce,’ ” and “to ‘prescribe such rules and regulations as may be necessary in the public interest to carry out the provisions’ of,” the Communications Act of 1934. 545 U. S., at 980–981 (quoting 47 U. S. C. §§151, 201(b)). See also Sullivan v. Everhart, 494 U. S. 83, 87, 88–89 (1990) (applying Chevron deference to rule promulgated pursuant to delegation of “general authority to ‘make rules and regulations and to establish procedures, not inconsistent with the provisions of this subchapter, which are necessary or appropriate to carry out such provisions’ ” (quoting 42 U. S. C. §405(a) (1982 ed.))). We believe Chevron and Mead, rather than National Muffler and Rowan, provide the appropriate framework for evaluating the full-time employee rule. The Department issued the full-time employee rule pursuant to the explicit authorization to “prescribe all needful rules and regulations for the enforcement” of the Internal Revenue Code. 26 U. S. C. §7805(a). We have found such “express congressional authorizations to engage in the process of rulemaking” to be “a very good indicator of delegation meriting Chevron treatment.” Mead, supra, at 229. The Department issued the full-time employee rule only after notice-and-comment procedures, 69 Fed. Reg. 76405, again a consideration identified in our precedents as a “significant” sign that a rule merits Chevron deference. Mead, supra, at 230–231; see, e.g., Long Island Care at Home, Ltd. v. Coke, 551 U. S. 158, 173–174 (2007). We have explained that “the ultimate question is whether Congress would have intended, and expected, courts to treat [the regulation] as within, or outside, its delegation to the agency of ‘gap-filling’ authority.” Id., at 173 (emphasis deleted). In the Long Island Care case, we found that Chevron provided the appropriate standard of review “[w]here an agency rule sets forth important individual rights and duties, where the agency focuses fully and directly upon the issue, where the agency uses full notice-and-comment procedures to promulgate a rule, [and] where the resulting rule falls within the statutory grant of authority.” 551 U. S., at 173. These same considerations point to the same result here. This case falls squarely within the bounds of, and is properly analyzed under, Chevron and Mead. C The full-time employee rule easily satisfies the second step of Chevron, which asks whether the Department’s rule is a “reasonable interpretation” of the enacted text. 467 U. S., at 844. To begin, Mayo accepts that “the ‘educational aspect of the relationship between the employer and the employee, as compared to the service aspect of the relationship, [must] be predominant’ ” in order for an individual to qualify for the exemption. Reply Brief for Petitioners 6–7 (quoting Treas. Reg. §31.3121(b)(10)–2(d)(3)(i), 26 CFR §31.3121(b)(10)–2(d)(3)(i)). Mayo objects, however, to the Department’s conclusion that residents who work more than 40 hours per week categorically cannot satisfy that requirement. Because residents’ employment is itself educational, Mayo argues, the hours a resident spends working make him “more of a student, not less of one.” Reply Brief for Petitioners 15, n. 3 (emphasis deleted). Mayo contends that the Treasury Department should be required to engage in a case-by-case inquiry into “what [each] employee does [in his service] and why” he does it. Id., at 7. Mayo also objects that the Department has drawn an arbitrary distinction between “hands-on training” and “classroom instruction.” Brief for Petitioners 35. We disagree. Regulation, like legislation, often requires drawing lines. Mayo does not dispute that the Treasury Department reasonably sought a way to distinguish between workers who study and students who work, see IRS Letter Ruling 9332005 (May 3, 1993). Focusing on the hours an individual works and the hours he spends in studies is a perfectly sensible way of accomplishing that goal. The Department explained that an individual’s service and his “course of study are separate and distinct activities” in “the vast majority of cases,” and reasoned that “[e]mployees who are working enough hours to be considered full-time employees . . . have filled the conventional measure of available time with work, and not study.” 69 Fed. Reg. 8607. The Department thus did not distinguish classroom education from clinical training but rather education from service. The Department reasonably concluded that its full-time employee rule would “improve administrability,” id., at 76405, and it thereby “has avoided the wasteful litigation and continuing uncertainty that would inevitably accompany any purely case-by-case approach” like the one Mayo advocates, United States v. Correll, 389 U. S. 299, 302 (1967). As the Treasury Department has explained, moreover, the full-time employee rule has more to recommend it than administrative convenience. The Department reasonably determined that taxing residents under FICA would further the purpose of the Social Security Act and comport with this Court’s precedent. As the Treasury Department appreciated, this Court has understood the terms of the Social Security Act to “ ‘import a breadth of coverage,’ ” 69 Fed. Reg. 8605 (quoting Social Security Bd. v. Nierotko, 327 U. S. 358, 365 (1946)), and we have instructed that “exemptions from taxation are to be construed narrowly,” Bingler v. Johnson, 394 U. S. 741, 752 (1969). Although Mayo contends that medical residents have not yet begun their “working lives” because they are not “fully trained,” Reply Brief for Petitioners 13 (internal quotation marks omitted), the Department certainly did not act irrationally in concluding that these doctors—“who work long hours, serve as highly skilled professionals, and typically share some or all of the terms of employment of career employees”—are the kind of workers that Congress intended to both contribute to and benefit from the Social Security system. 69 Fed. Reg. 8608. The Department’s rule takes into account the SSA’s concern that exempting residents from FICA would deprive residents and their families of vital disability and survivorship benefits that Social Security provides. Id., at 8605. Mayo wonders whether the full-time employee rule will result in residents being taxed under FICA but denied coverage by the SSA. The Government informs us, however, that the SSA continues to adhere to its longstanding position that medical residents are not students and thus remain eligible for coverage. Brief for United States 29–30; Tr. of Oral Arg. 33–34. * * * We do not doubt that Mayo’s residents are engaged in a valuable educational pursuit or that they are students of their craft. The question whether they are “students” for purposes of §3121, however, is a different matter. Because it is one to which Congress has not directly spoken, and because the Treasury Department’s rule is a reasonable construction of what Congress has said, the judgment of the Court of Appeals must be affirmed. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

563.US.816

Under the Armed Career Criminal Act (ACCA), a felon unlawfully in possession of a firearm, 18 U. S. C. §922(g)(1), is subject to a 15-year minimum prison sentence if he has three prior convictions for a “violent felony” or “serious drug offense.” As relevant here, a “serious drug offense” is defined as “an offense under State law … , for which a maximum term of imprisonment of ten years or more is prescribed by law,” §924(e)(2)(A)(ii). In sentencing petitioner McNeill for violating §922(g), the District Court determined that he qualified for ACCA’s sentencing enhancement based in part on six prior North Carolina drug trafficking convictions. When McNeill committed those crimes, each carried a 10-year maximum sentence, which McNeill in fact received. However, because the State later reduced the maximum sentence for those offenses to fewer than 10 years, McNeill argued that none of his six prior convictions were for “serious drug offenses” within the meaning of §924(e)(2)(A)(ii). The District Court rejected McNeill’s request that it look to current state law and instead relied on the 10-year maximum sentence that applied at the time he committed his state offenses. The Fourth Circuit affirmed. Held: 1. A federal sentencing court must determine whether “an offense under State law” is a “serious drug offense” by consulting the “maximum term of imprisonment” applicable to a defendant’s prior state drug offense at the time of the defendant’s conviction for that offense. §924(e)(2)(A)(ii). Pp. 3–7. (a) ACCA’s plain text requires this result by mandating that the court determine whether a “previous conviction” was for a serious drug offense. The only way to answer this backward-looking question is to consult the law that applied at the time of that conviction. ACCA’s use of the present tense in defining a “serious drug offense” as, inter alia, “an offense … for which a maximum [10-year] term … is prescribed by law” does not suggest otherwise. McNeill’s argument that this language looks to the state law in effect at the time of the federal sentencing ignores ACCA’s focus on convictions that have already occurred. Pp. 3–4. (b) The statute’s broader context, specifically the adjacent definition of “violent felony,” confirms this interpretation. Although Congress used the present tense in defining “violent felony,” see §924(e)(2)(B), this Court has repeatedly turned to the version of state law that the defendant was actually convicted of violating in determining whether he was convicted of such a felony, see, e.g., Taylor v. United States, 495 U. S. 575, 602. The Court sees no reason to interpret “serious drug offenses” any differently. Cf. Nijhawan v. Holder, 557 U. S. ___, ___. Pp. 5–6. (c) This natural reading of ACCA also avoids the absurd results that would follow from consulting current state law to define a previous offense. Pp. 6–7. 2. The District Court properly applied ACCA’s sentencing enhancement to McNeill because all six of his prior drug convictions were for “serious drug offenses.” Pp. 7–8. 598 F. 3d 161, affirmed. Thomas, J., delivered the opinion for a unanimous Court.

Under the Armed Career Criminal Act (ACCA), a prior state drug-trafficking conviction is for a “serious drug of-fense” if “a maximum term of imprisonment of ten years or more is prescribed by law” for the offense. 18 U. S. C. §924(e)(2)(A)(ii). The question in this case concerns how a federal court should determine the maximum sentence for a prior state drug offense for ACCA purposes. We hold that the “maximum term of imprisonment” for a defendant’s prior state drug offense is the maximum sentence applicable to his offense when he was convicted of it. I After an extended chase, police officers in Fayetteville, North Carolina apprehended petitioner Clifton Terelle McNeill. McNeill was caught with 3.1 grams of crack cocaine packaged for distribution and a .38-caliber revolver. In August 2008, he pleaded guilty to unlawful possession of a firearm by a felon, 18 U. S. C. §922(g)(1), and possession with intent to distribute cocaine base, 21 U. S. C. §841(a)(1). At sentencing, the District Court determined that McNeill qualified for ACCA’s sentencing enhancement. Under ACCA, a person who violates 18 U. S. C. §922(g) and “has three previous convictions … for a violent felony or a serious drug offense” is subject to a 15-year minimum prison sentence. §924(e)(1). McNeill conceded that two of his prior convictions—assault with a deadly weapon and robbery—were for “violent felonies.” McNeill argued, however, that none of his six state drug trafficking convictions were for “serious drug offenses” be-cause those crimes no longer carried a “maximum term of imprisonment of ten years or more.” §924(e)(2)(A)(ii). When McNeill committed those crimes between 1991 and 1994, each carried a 10-year maximum sentence, and McNeill in fact received 10-year sentences. See N. C. Gen. Stat. §§14–1.1(a)(8), 90–95(a)(1) and (b)(1) (Michie 1993) (sale of cocaine and possession with intent to sell cocaine). But as of October 1, 1994, North Carolina reduced the maximum sentence for selling cocaine to 38 months and the maximum sentence for possessing cocaine with intent to sell to 30 months. See N. C. Gen. Stat. Ann. §§15A–1340.17(c) and (d), 90–95(a)(1) and (b)(1) (Lexis 2009). The District Court rejected McNeill’s request that it look to current state law and instead relied on the 10-year maximum sentence that applied to McNeill’s drug offenses at the time he committed them. No. 5:08–CR–2–D–1 (EDNC, Jan. 26, 2009), App. 118. Finding that McNeill therefore had three prior convictions for violent felonies or serious drug offenses, the court applied ACCA’s sentencing enhancement. The court then departed upward from the advisory Sentencing Guidelines range and sentenced McNeill to 300 months in prison in light of his “long and unrelenting history of serious criminal conduct” and “near certain likelihood of recidivism.” Id., at 119, 121. The Court of Appeals for the Fourth Circuit affirmed. Although the court consulted the maximum sentence under current state law, it reached the same conclusion as the District Court because North Carolina’s revised sentencing scheme does not apply to crimes committed before October 1, 1994. 598 F. 3d 161, 165 (2010) (agreeing with United States v. Hinojosa, 349 F. 3d 200 (CA5 2003), and disagreeing with United States v. Darden, 539 F. 3d 116 (CA2 2008)). Thus, even if McNeill were convicted today for his 1991, 1992, and September 1994 drug offenses, he would still be subject to the old 10-year statutory maximum. 598 F. 3d, at 165 (citing N. C. Gen. Stat. §15A–1340.10 and State v. Branch, 134 N. C. App. 637, 639–640, 518 S. E. 2d 213, 215 (1999)). We granted certiorari, 562 U. S. ___ (2011), and now affirm, albeit for a different reason. II A As in all statutory construction cases, we begin with “the language itself [and] the specific context in which that language is used.” Robinson v. Shell Oil Co., 519 U. S. 337, 341 (1997). ACCA’s sentencing enhancement applies to individuals who have “three previous convictions … for a violent felony or a serious drug offense.” §924(e)(1). As relevant here, the statute defines a “serious drug offense” as “an offense under State law, involving manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance … , for which a maximum term of imprisonment of ten years or more is prescribed by law.” §924(e)(2)(A)(ii). The plain text of ACCA requires a federal sentencing court to consult the maximum sentence applicable to a defendant’s previous drug offense at the time of his conviction for that offense. The statute requires the court to determine whether a “previous conviction” was for a serious drug offense. The only way to answer this backward-looking question is to consult the law that applied at the time of that conviction. We did precisely that in United States v. Rodriquez, 553 U. S. 377 (2008), where we addressed whether the “maximum term of imprisonment” in-cludes recidivism enhancements. In assessing the “max-imum term of imprisonment” for Rodriguez’s state drug offenses, we consulted the version of state law “that [he] was convicted of violating,” that is, the 1994 statutes and penalties that applied to his offenses at the time of his state convictions. Id., at 380–381. Use of the present tense in the definition of “serious drug offense” does not suggest otherwise. McNeill argues that the present-tense verb in the phrase “is prescribed by law” requires federal courts to determine the maximum sentence for a potential predicate offense by looking to the state law in effect at the time of the federal sentencing, as if the state offense were committed on the day of federal sentencing. That argument overlooks the fact that ACCA is concerned with convictions that have already occurred. Whether the prior conviction was for an offense “involving manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance” can only be answered by reference to the law under which the defendant was convicted. Likewise, the maximum sentence that “is prescribed by law” for that offense must also be determined according to the law applicable at that time. McNeill’s interpretation contorts the plain meaning of the statute. Although North Carolina courts actually sentenced him to 10 years in prison for his drug offenses, McNeill now contends that the “maximum term of imprisonment” for those offenses is 30 or 38 months. We find it “hard to accept the proposition that a defendant may lawfully [have] be[en] sentenced to a term of imprisonment that exceeds the ‘maximum term of imprisonment … prescribed by law.’ ” Id., at 383. B The “broader context of the statute as a whole,” specifically the adjacent definition of “violent felony,” confirms this interpretation. Robinson, supra, at 341. ACCA defines “violent felony” in part as a crime that “has as an element the use, attempted use, or threatened use of physical force against the person of another” or “is burglary, arson, or extortion, involves the use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.” §924(e)(2)(B) (emphasis added). Despite Congress’ use of present tense in that definition, when determining whether a defendant was convicted of a “violent felony,” we have turned to the version of state law that the defendant was actually convicted of violating. In Taylor v. United States, 495 U. S. 575 (1990), the Court held that whether Taylor’s 1963 and 1971 convictions were for a crime that “is burglary” depended on the “former Missouri statutes defining second-degree burglary” that “were the bases for Taylor’s prior convictions.” Id., at 602; see id., at 578, n. 1 (noting a subsequent change in state law, but relying on the burglary statutes in force “[i]n those years” in which Taylor was convicted). Similarly, in James v. United States, 550 U. S. 192 (2007), this Court looked to the versions of Florida’s burglary and criminal attempt statutes that were in effect “at the time of James’ [1993 state] conviction.” Id., at 197; see ibid. (quoting the 1993 versions of the Florida statutes). The present-tense verbs in the definition of “violent felony” did not persuade us to look anywhere other than the law under which the defendants were actually convicted to determine the elements of their offenses. Having repeatedly looked to the historical statute of conviction in the context of violent felonies, we see no reason to interpret “serious drug offenses” in the adjacent section of the same statute any differently. In both definitions, Congress used the present tense to refer to past convictions. Cf. Nijhawan v. Holder, 557 U. S. ___, ___ (2009) (slip op., at 8) (“Where, as here, Congress uses similar statutory language … in two adjoining provisions, it normally intends similar interpretations”). C This natural reading of ACCA also avoids the absurd results that would follow from consulting current state law to define a previous offense. See United States v. Wilson, 503 U. S. 329, 334 (1992) (“[A]bsurd results are to be avoided”). For example, McNeill concedes that under his approach, a prior conviction could “disappear” entirely for ACCA purposes if a State reformulated the offense between the defendant’s state conviction and federal sentencing. Tr. of Oral Arg. 12–13. The Sixth Circuit confronted a similar scenario in Mallett v. United States, 334 F. 3d 491 (2003), where Ohio had substantially changed how drug quantities were measured since Mallett’s state drug conviction. Id., at 502 (addressing this issue in the context of the career offender provision of the Sentencing Guidelines). The Sixth Circuit could not “determine how Mallett would now be sentenced under Ohio’s revised drug laws” because the offense for which he had been convicted “no longer exist[ed] and no conversion between the former and amended statutes [wa]s facially apparent.” Ibid. The court therefore was compelled to look to state law “as of the time of the state-court conviction” to determine the maximum possible sentence for Mallet’s prior offense. Id., at 503. It cannot be correct that subsequent changes in state law can erase an earlier conviction for ACCA purposes. A defendant’s history of criminal activity—and the culpability and dangerousness that such history demonstrates—does not cease to exist when a State reformulates its criminal statutes in a way that prevents precise translation of the old conviction into the new statutes. Congress based ACCA’s sentencing enhancement on prior convictions and could not have expected courts to treat those convictions as if they had simply disappeared. To the contrary, Congress has expressly directed that a prior violent felony conviction remains a “conviction” unless it has been “expunged, or set aside or [the] person has been pardoned or has had civil rights restored.” 18 U. S. C. §921(a)(20); see also Custis v. United States, 511 U. S. 485, 491 (1994) (explaining that §921(a)(20) “creates a clear negative implication that courts may count a conviction that has not been set aside”). In addition, McNeill’s interpretation would make ACCA’s applicability depend on the timing of the federal sentencing proceeding. McNeill cannot explain why two defendants who violated §922(g) on the same day and who had identical criminal histories—down to the dates on which they committed and were sentenced for their prior offenses—should receive dramatically different federal sentences solely because one’s §922(g) sentencing happened to occur after the state legislature amended the punishment for one of the shared prior offenses. In contrast, the interpretation we adopt permits a defendant to know even before he violates §922(g) whether ACCA would apply. III Applying our holding to this case, we conclude that the District Court properly applied ACCA’s sentencing enhancement to McNeill. In light of his two admitted violent felony convictions, McNeill needed only one conviction for a “serious drug offense” to trigger ACCA, but we note that all six of his prior drug convictions qualify. In November 1992, McNeill pleaded guilty and was sentenced in a North Carolina court for five offenses: selling cocaine on four separate occasions in October 1991 and possessing cocaine with intent to sell on one occasion in February 1992. At the time of McNeill’s November 1992 conviction and sentencing, North Carolina law dictated that the maximum sentence for selling cocaine in 1991 and the maximum sentence for possessing cocaine with intent to sell in 1992 was 10 years in prison. See N. C. Gen. Stat. §§14–1.1(a)(8), 90–95(a)(1) and (b)(1) (Michie 1991). McNeill’s 1992 convictions were therefore for “serious drug offenses” within the meaning of ACCA. McNeill’s sixth drug offense was possessing cocaine with intent to sell in September 1994. He pleaded guilty and was sentenced in a North Carolina court in April 1995. By April 1995, North Carolina had changed the sentence applicable to that type of drug offense but still provided that the maximum sentence for possessing cocaine with intent to sell in September 1994 was 10 years in prison. See 1993 N. C. Sess. Laws, ch. 538, §2 (repealing N. C. Gen Stat. §14–1.1); 1993 N. C. Sess. Laws, ch. 538, §56 (as modified by Extra Session 1994 N. C. Sess. Laws, ch. 24, §14(b)) (“This act becomes effective October 1, 1994, and applies only to offenses occurring on or after that date. Prosecutions for, or sentences based on, offenses occurring before the effective date of this act [are controlled by] the statutes that would be applicable to those prosecutions or sentences but for the provisions of this act”). Therefore, McNeill’s 1995 conviction was also for a “serious drug offense.” * * * We conclude that a federal sentencing court must determine whether “an offense under State law” is a “serious drug offense” by consulting the “maximum term of imprisonment” applicable to a defendant’s previous drug offense at the time of the defendant’s state conviction for that offense. §924(e)(2)(A)(ii). The judgment of the United States Court of Appeals for the Fourth Circuit is affirmed. It is so ordered. As the Government notes, this case does not concern a situation in which a State subsequently lowers the maximum penalty applicable to an offense and makes that reduction available to defendants previously convicted and sentenced for that offense. Brief for United States 18, n. 5; cf. 18 U. S. C. §3582(c)(2). We do not address whether or under what circumstances a federal court could consider the effect of that state action.

562.US.344

Michigan police dispatched to a gas station parking lot found Anthony Covington mortally wounded. Covington told them that he had been shot by respondent Bryant outside Bryant’s house and had then driven himself to the lot. At trial, which occurred before Crawford v. Washington, 541 U. S. 36, and Davis v. Washington, 547 U. S. 813, were decided, the officers testified about what Covington said. Bryant was found guilty of, inter alia, second-degree murder. Ultimately, the Michigan Supreme Court reversed his conviction, holding that the Sixth Amendment’s Confrontation Clause, as explained in Crawford and Davis, rendered Covington’s statements inadmissible testimonial hearsay. Held: Covington’s identification and description of the shooter and the location of the shooting were not testimonial statements because they had a “primary purpose . . . to enable police assistance to meet an ongoing emergency.” Davis, 547 U. S., at 822. Therefore, their admission at Bryant’s trial did not violate the Confrontation Clause. Pp. 5–32. (a) In Crawford, this Court held that in order for testimonial evidence to be admissible, the Sixth Amendment “demands … unavailability and a prior opportunity for cross-examination.” 541 U. S., at 68. Crawford did not “spell out a comprehensive definition of ‘testimonial,’ ” but it noted that testimonial evidence includes, among other things, “police interrogations.” Ibid. Thus, Sylvia Crawford’s statements during a station-house interrogation about a stabbing were testimonial, and their admission when her husband, the accused, had “no opportunity” for cross-examination due to spousal privilege made out a Sixth Amendment violation. In Davis and Hammon, both domestic violence cases, the Court explained that “[s]tatements are nontestimonial when made in the course of police interrogation under circumstances objectively indicating that the [interrogation’s] primary purpose … is to enable police assistance to meet an ongoing emergency,” but they “are testimonial when the circumstances objectively indicate that there is no such ongoing emergency, and that the [interrogation’s] primary purpose is to establish or prove past events potentially relevant to later criminal prosecution.” 547 U. S., at 822. Thus, a recording of a 911 call describing an ongoing domestic disturbance was nontestimonial in Davis, where the victim’s “elicited statements were necessary to be able to resolve [the ongoing] emergency,” and the statements were not formal. Id., at 827. But the statements in Hammon were testimonial, where the victim was interviewed after the event in a room separate from her husband and “deliberately recounted, in response to police questioning” the past events. Id., at 830. Here, the context is a nondomestic dispute, with the “ongoing emergency” extending beyond an initial victim to a potential threat to the responding police and the public. This context requires additional clarification of what Davis meant by “the primary purpose of the interrogation is to enable police assistance to meet an ongoing emergency.” Id., at 822. Pp. 5–12. (b) To make the “primary purpose” determination, the Court must objectively evaluate the circumstances in which the encounter between the individual and the police occurs and the parties’ statements and actions. Pp. 12–23. (1) The primary purpose inquiry is objective. The circumstances in which an encounter occurs—e.g., at or near a crime scene versus at a police station, during an ongoing emergency or afterwards—are clearly matters of objective fact. And the relevant inquiry into the parties’ statements and actions is not the subjective or actual purpose of the particular parties, but the purpose that reasonable participants would have had, as ascertained from the parties’ statements and actions and the circumstances in which the encounter occurred. P. 13. (2) The existence of an “ongoing emergency” at the time of the encounter is among the most important circumstances informing the interrogation’s “primary purpose.” See, e.g., Davis, 547 U. S., at 828–830. An emergency focuses the participants not on “prov[ing] past events potentially relevant to later criminal prosecution,” id., at 822, but on “end[ing] a threatening situation,” id., at 832. The Michigan Supreme Court failed to appreciate that whether an emergency exists and is ongoing is a highly context-dependent inquiry. An assessment of whether an emergency threatening the police and public is ongoing cannot narrowly focus on whether the threat to the first victim has been neutralized because the threat to the first responders and public may continue. The State Supreme Court also did not appreciate that an emergency’s duration and scope may depend in part on the type of weapon involved; the court below relied on Davis and Hammon, where the assailants used their fists, as controlling the scope of an emergency involving a gun. A victim’s medical condition is important to the primary purpose inquiry to the extent that it sheds light on the victim’s ability to have any purpose at all in responding to police questions and on the likelihood that any such purpose would be a testimonial one. It also provides important context for first responders to judge the existence and magnitude of a continuing threat to the victim, themselves, and the public. This does not mean that an emergency lasts the entire time that a perpetrator is on the loose, but trial courts can determine in the first instance when an interrogation transitions from nontestimonial to testimonial. Finally, whether an ongoing emergency exists is simply one factor informing the ultimate inquiry regarding an interrogation’s “primary purpose.” Another is the encounter’s informality. Formality suggests the absence of an emergency, but informality does not necessarily indicate the presence of an emergency or the lack of testimonial intent. The facts here—the questioning occurred in an exposed, public area, before emergency medical services arrived, and in a disorganized fashion—distinguish this case from Crawford’s formal station-house interrogation. Pp. 14–20. (3) The statements and actions of both the declarant and interrogators also provide objective evidence of the interrogation’s primary purpose. Looking to the contents of both the questions and the answers ameliorates problems that could arise from looking solely to one participant, since both interrogators and declarants may have mixed motives. Police officers’ dual responsibilities as both first responders and criminal investigators may lead them to act with different motives simultaneously or in quick succession. And during an ongoing emergency, victims may want the threat to end, but may not envision prosecution. Alternatively, a severely injured victim may have no purpose at all in answering questions. Taking into account such injuries does not make the inquiry subjective. The inquiry still focuses on the understanding and purpose of a reasonable victim in the actual victim’s circumstances, which prominently include the victim’s physical state. Objectively ascertaining the primary purpose of the interrogation by examining the statements and actions of all participants is also consistent with this Court’s prior holdings. E.g., Davis, 547 U. S., at 822–823, n. 1. Pp. 20–23. (c) Here, the circumstances of the encounter as well as the statements and actions of Covington and the police objectively indicate that the interrogation’s “primary purpose” was “to enable police assistance to meet an ongoing emergency,” 547 U. S., at 822. The circumstances of the interrogation involved an armed shooter, whose motive for and location after the shooting were unknown and who had mortally wounded Covington within a few blocks and a few minutes of the location where police found Covington. Unlike the emergencies in Davis and Hammon, this dispute’s potential scope and thus the emergency encompassed a potential threat to the police and the public. And since this case involved a gun, the physical separation that was sufficient to end the emergency in Hammon was not necessarily sufficient to end the threat here. Informed by the circumstances of the ongoing emergency, the Court now turns to determining the “primary purpose of the interrogation” as evidenced by the statements and actions of Covington and the police. The circumstances of the encounter provide important context for understanding Covington’s statements to the police. When he responded to their questions, he was lying in a gas station parking lot bleeding from a mortal gunshot wound, and his answers were punctuated with questions about when emergency medical services would arrive. Thus, this Court cannot say that a person in his situation would have had a “primary purpose” “to establish or prove past events potentially relevant to later criminal prosecution.” Ibid. For their part, the police responded to a call that a man had been shot. They did not know why, where, or when the shooting had occurred; the shooter’s location; or anything else about the crime. They asked exactly the type of questions necessary to enable them “to meet an ongoing emergency.” Ibid. Nothing in Covington’s responses indicated to the police that there was no emergency or that the emergency had ended. Finally, this situation is more similar to the informal, harried 911 call in Davis than to the structured, station-house interview in Crawford. The officers all arrived at different times; asked, upon arrival, what had happened; and generally did not conduct a structured interrogation. The informality suggests that their primary purpose was to address what they considered to be an ongoing emergency, and the circumstances lacked a formality that would have alerted Covington to or focused him on the possible future prosecutorial use of his statements. Pp. 23–32. 483 Mich. 132, 768 N. W. 2d 65, vacated and remanded. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Breyer, and Alito, JJ., joined. Thomas, J., filed an opinion concurring in the judgment. Scalia, J., and Ginsburg, J., filed dissenting opinions. Kagan, J., took no part in the consideration or decision of the case.

At respondent Richard Bryant’s trial, the court admitted statements that the victim, Anthony Covington, made to police officers who discovered him mortally wounded in a gas station parking lot. A jury convicted Bryant of, inter alia, second-degree murder. 483 Mich. 132, 137, 768 N. W. 2d 65, 67–68 (2009). On appeal, the Supreme Court of Michigan held that the Sixth Amendment’s Confrontation Clause, as explained in our decisions in Crawford v. Washington, 541 U. S. 36 (2004), and Davis v. Washington, 547 U. S. 813 (2006), rendered Covington’s statements inadmissible testimonial hearsay, and the court reversed Bryant’s conviction. 483 Mich., at 157, 768 N. W. 2d, at 79. We granted the State’s petition for a writ of certiorari to consider whether the Confrontation Clause barred the admission at trial of Covington’s statements to the police. We hold that the circumstances of the interaction between Covington and the police objectively indicate that the “primary purpose of the interrogation” was “to enable police assistance to meet an ongoing emergency.” Davis, 547 U. S., at 822. Therefore, Covington’s identification and description of the shooter and the location of the shooting were not testimonial statements, and their admission at Bryant’s trial did not violate the Confrontation Clause. We vacate the judgment of the Supreme Court of Michigan and remand. I Around 3:25 a.m. on April 29, 2001, Detroit, Michigan police officers responded to a radio dispatch indicating that a man had been shot. At the scene, they found the victim, Anthony Covington, lying on the ground next to his car in a gas station parking lot. Covington had a gunshot wound to his abdomen, appeared to be in great pain, and spoke with difficulty. The police asked him “what had happened, who had shot him, and where the shooting had occurred.” 483 Mich., at 143, 768 N. W. 2d, at 71. Covington stated that “Rick” shot him at around 3 a.m. Id., at 136, and n. 1, 768 N. W. 2d, at 67, and n. 1. He also indicated that he had a conversation with Bryant, whom he recognized based on his voice, through the back door of Bryant’s house. Covington explained that when he turned to leave, he was shot through the door and then drove to the gas station, where police found him. Covington’s conversation with the police ended within 5 to 10 minutes when emergency medical services arrived. Covington was transported to a hospital and died within hours. The police left the gas station after speaking with Covington, called for backup, and traveled to Bryant’s house. They did not find Bryant there but did find blood and a bullet on the back porch and an apparent bullet hole in the back door. Police also found Covington’s wallet and identification outside the house. At trial, which occurred prior to our decisions in Crawford, 541 U. S. 36, and Davis, 547 U. S. 813, the police officers who spoke with Covington at the gas station testified about what Covington had told them. The jury returned a guilty verdict on charges of second-degree murder, being a felon in possession of a firearm, and possession of a firearm during the commission of a felony. Bryant appealed, and the Michigan Court of Appeals affirmed his conviction. No. 247039, 2004 WL 1882661 (Aug. 24, 2004) (per curiam). Bryant then appealed to the Supreme Court of Michigan, arguing that the trial court erred in admitting Covington’s statements to the police. The Supreme Court of Michigan eventually remanded the case to the Court of Appeals for reconsideration in light of our 2006 decision in Davis. 477 Mich. 902, 722 N. W. 2d 797 (2006). On remand, the Court of Appeals again affirmed, holding that Covington’s statements were properly admitted because they were not testimonial. No. 247039, 2007 WL 675471 (Mar. 6, 2007) (per curiam). Bryant again appealed to the Supreme Court of Michigan, which reversed his conviction. 483 Mich. 132, 768 N. W. 2d 65. Before the Supreme Court of Michigan, Bryant argued that Covington’s statements to the police were testimonial under Crawford and Davis and were therefore inadmissible. The State, on the other hand, argued that the statements were admissible as “excited utterances” under the Michigan Rules of Evidence. 483 Mich., at 142, and n. 6, 768 N. W. 2d, at 70, and n. 6. There was no dispute that Covington was unavailable at trial and Bryant had no prior opportunity to cross-examine him. The court therefore assessed whether Covington’s statements to the police identifying and describing the shooter and the time and location of the shooting were testimonial hearsay for purposes of the Confrontation Clause. The court concluded that the circumstances “clearly indicate that the ‘primary purpose’ of the questioning was to establish the facts of an event that had already occurred; the ‘primary purpose’ was not to enable police assistance to meet an ongoing emergency.” Id., at 143, 768 N. W. 2d, at 71. The court explained that, in its view, Covington was describing past events and as such, his “primary purpose in making these statements to the police . . . was . . . to tell the police who had committed the crime against him, where the crime had been committed, and where the police could find the criminal.” Id., at 144, 768 N. W. 2d, at 71. Noting that the officers’ actions did not suggest that they perceived an ongoing emergency at the gas station, the court held that there was in fact no ongoing emergency. Id., at 145–147, 768 N. W. 2d, at 71–73. The court distinguished the facts of this case from those in Davis, where we held a declarant’s statements in a 911 call to be nontestimonial. It instead analogized this case to Hammon v. Indiana, which we decided jointly with Davis and in which we found testimonial a declarant’s statements to police just after an assault. See 547 U. S., at 829–832. Based on this analysis, the Supreme Court of Michigan held that the admission of Covington’s statements constituted prejudicial plain error warranting reversal and ordered a new trial. 483 Mich., at 151–153, 768 N. W. 2d, at 75–76. The court did not address whether, absent a Confrontation Clause bar, the statements’ admission would have been otherwise consistent with Michigan’s hearsay rules or due process.[Footnote 1] The majority’s opinion provoked two dissents, both of which would have held Covington’s statements admissible because they were made in circumstances indicating that their “primary purpose” was to assist police in addressing an ongoing emergency. Id., at 157, 768 N. W. 2d, at 79 (opinion of Weaver, J.); id., at 157–158, 768 N. W. 2d, at 79 (opinion of Corrigan, J.). Justice Corrigan’s dissent explained that the time and space between “the onset of an emergency and statements about that emergency clearly must be considered in context.” Id., at 161, 768 N. W. 2d, at 80. Justice Corrigan concluded that the objective circumstances of Covington’s interaction with police rendered this case more similar to the nontestimonial statements in Davis than to the testimonial statements in Crawford. 483 Mich., at 164, 768 N. W. 2d, at 82. We granted certiorari to determine whether the Confrontation Clause barred admission of Covington’s statements. 559 U. S. ___ (2010). II The Confrontation Clause of the Sixth Amendment states: “In all criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him.” The Fourteenth Amendment renders the Clause binding on the States. Pointer v. Texas, 380 U. S. 400, 403 (1965). In Ohio v. Roberts, 448 U. S. 56, 66 (1980), we explained that the confrontation right does not bar admission of statements of an unavailable witness if the statements “bea[r] adequate ‘indicia of reliability.’ ” We held that reliability can be established if “the evidence falls within a firmly rooted hearsay exception,” or if it does not fall within such an exception, then if it bears “particularized guarantees of trustworthiness.” Ibid. Nearly a quarter century later, we decided Crawford v. Washington, 541 U. S. 36. Petitioner Michael Crawford was prosecuted for stabbing a man who had allegedly attempted to rape his wife, Sylvia. Sylvia witnessed the stabbing, and later that night, after she and her husband were both arrested, police interrogated her about the incident. At trial, Sylvia Crawford claimed spousal privilege and did not testify, but the State introduced a tape recording of Sylvia’s statement to the police in an effort to prove that the stabbing was not in self-defense, as Michael Crawford claimed. The Washington Supreme Court affirmed Crawford’s conviction because it found Sylvia’s statement to be reliable, as required under Ohio v. Roberts. We reversed, overruling Ohio v. Roberts. 541 U. S., at 60–68; see also Davis, 547 U. S., at 825, n. 4. Crawford examined the common-law history of the confrontation right and explained that “the principal evil at which the Confrontation Clause was directed was the civil-law mode of criminal procedure, and particularly its use of ex parte examinations as evidence against the accused.” 541 U. S., at 50. We noted that in England, pretrial examinations of suspects and witnesses by government officials “were sometimes read in court in lieu of live testimony.” Id., at 43. In light of this history, we emphasized the word “witnesses” in the Sixth Amendment, defining it as “those who ‘bear testimony.’ ” Id., at 51 (quoting 2 N. Webster, An American Dictionary of the English Language (1828)). We defined “testimony” as “ ‘ [a] solemn declaration or affirmation made for the purpose of establishing or proving some fact.’ ” 541 U. S., at 51 (quoting Webster). We noted that “[a]n accuser who makes a formal statement to government officers bears testimony in a sense that a person who makes a casual remark to an acquaintance does not.” Ibid. We therefore limited the Confrontation Clause’s reach to testimonial statements and held that in order for testimonial evidence to be admissible, the Sixth Amendment “demands what the common law required: unavailability and a prior opportunity for cross-examination.” Id., at 68. Although “leav[ing] for another day any effort to spell out a comprehensive definition of ‘testimonial,’ ” Crawford noted that “at a minimum” it includes “prior testimony at a preliminary hearing, before a grand jury, or at a former trial; and . . . police interrogations.” Ibid. Under this reasoning, we held that Sylvia Crawford’s statements in the course of police questioning were testimonial and that their admission when Michael Crawford “had no opportunity to cross-examine her” due to spousal privilege was “sufficient to make out a violation of the Sixth Amendment.” Ibid. In 2006, the Court in Davis v. Washington and Hammon v. Indiana, 547 U. S. 813, took a further step to “determine more precisely which police interrogations produce testimony” and therefore implicate a Confrontation Clause bar. Id., at 822. We explained that when Crawford said that “ ‘interrogations by law enforcement officers fall squarely within [the] class’ of testimonial hearsay, we had immediately in mind (for that was the case before us) interrogations solely directed at establishing the facts of a past crime, in order to identify (or provide evidence to convict) the perpetrator. The product of such interrogation, whether reduced to a writing signed by the declarant or embedded in the memory (and perhaps notes) of the interrogating officer, is testimonial.” Davis, 547 U. S., at 826. We thus made clear in Davis that not all those questioned by the police are witnesses and not all “interrogations by law enforcement officers,” Crawford, 541 U. S., at 53, are subject to the Confrontation Clause.[Footnote 2] Davis and Hammon were both domestic violence cases. In Davis, Michelle McCottry made the statements at issue to a 911 operator during a domestic disturbance with Adrian Davis, her former boyfriend. McCottry told the operator, “ ‘He’s here jumpin’ on me again,’ ” and, “ ‘He’s usin’ his fists.’ ” 547 U. S., at 817. The operator then asked McCottry for Davis’ first and last names and middle initial, and at that point in the conversation McCottry reported that Davis had fled in a car. Id., at 818. McCottry did not appear at Davis’ trial, and the State introduced the recording of her conversation with the 911 operator. Id., at 819. In Hammon, decided along with Davis, police responded to a domestic disturbance call at the home of Amy and Hershel Hammon, where they found Amy alone on the front porch. Ibid. She appeared “ ‘somewhat frightened,’ ” but told them “ ‘nothing was the matter.’ ” Ibid. (quoting Hammon v. State, 829 N. E. 2d 444, 446–447 (Ind. 2005)). She gave the police permission to enter the house, where they saw a gas heating unit with the glass front shattered on the floor. One officer remained in the kitchen with Hershel, while another officer talked to Amy in the living room about what had happened. Hershel tried several times to participate in Amy’s conversation with the police and became angry when the police required him to stay separated from Amy. 547 U. S., at 819–820. The police asked Amy to fill out and sign a battery affidavit. She wrote: “ ‘Broke our Furnace & shoved me down on the floor into the broken glass. Hit me in the chest and threw me down. Broke our lamps & phone. Tore up my van where I couldn’t leave the house. Attacked my daughter.’ ” Id., at 820. Amy did not appear at Hershel’s trial, so the police officers who spoke with her testified as to her statements and authenticated the affidavit. Ibid. The trial court admitted the affidavit as a present sense impression and admitted the oral statements as excited utterances under state hearsay rules. Ibid. The Indiana Supreme Court affirmed Hammon’s conviction, holding that Amy’s oral statements were not testimonial and that the admission of the affidavit, although erroneous because the affidavit was testimonial, was harmless. Hammon v. State, 829 N. E. 2d, at 458–459. To address the facts of both cases, we expanded upon the meaning of “testimonial” that we first employed in Crawford and discussed the concept of an ongoing emergency. We explained: “Statements are nontestimonial when made in the course of police interrogation under circumstances objectively indicating that the primary purpose of the interrogation is to enable police assistance to meet an ongoing emergency. They are testimonial when the circumstances objectively indicate that there is no such ongoing emergency, and that the primary purpose of the interrogation is to establish or prove past events potentially relevant to later criminal prosecution.” Davis, 547 U. S., at 822. Examining the Davis and Hammon statements in light of those definitions, we held that the statements at issue in Davis were nontestimonial and the statements in Hammon were testimonial. We distinguished the statements in Davis from the testimonial statements in Crawford on several grounds, including that the victim in Davis was “speaking about events as they were actually happening, rather than ‘describ[ing] past events,’ ” that there was an ongoing emergency, that the “elicited statements were necessary to be able to resolve the present emergency,” and that the statements were not formal. 547 U. S., at 827. In Hammon, on the other hand, we held that, “[i]t is entirely clear from the circumstances that the interrogation was part of an investigation into possibly criminal past conduct.” Id., at 829. There was “no emergency in progress.” Ibid. The officer questioning Amy “was not seeking to determine . . . ‘what is happening,’ but rather ‘what happened.’ ” Id., at 830. It was “formal enough” that the police interrogated Amy in a room separate from her husband where, “some time after the events described were over,” she “deliberately recounted, in response to police questioning, how potentially criminal past events began and progressed.” Ibid. Because her statements “were neither a cry for help nor the provision of information enabling officers immediately to end a threatening situation,” id., at 832, we held that they were testimonial. Davis did not “attemp[t] to produce an exhaustive classification of all conceivable statements—or even all conceivable statements in response to police interrogation—as either testimonial or nontestimonial.” Id., at 822.[Footnote 3] The basic purpose of the Confrontation Clause was to “targe[t]” the sort of “abuses” exemplified at the notorious treason trial of Sir Walter Raleigh. Crawford, 541 U. S., at 51. Thus, the most important instances in which the Clause restricts the introduction of out-of-court statements are those in which state actors are involved in a formal, out-of-court interrogation of a witness to obtain evidence for trial.[Footnote 4] See id., at 43–44. Even where such an interrogation is conducted with all good faith, introduction of the resulting statements at trial can be unfair to the accused if they are untested by cross-examination. Whether formal or informal, out-of-court statements can evade the basic objective of the Confrontation Clause, which is to prevent the accused from being deprived of the opportunity to cross-examine the declarant about statements taken for use at trial. When, as in Davis, the primary purpose of an interrogation is to respond to an “ongoing emergency,” its purpose is not to create a record for trial and thus is not within the scope of the Clause. But there may be other circumstances, aside from ongoing emergencies, when a statement is not procured with a primary purpose of creating an out-of-court substitute for trial testimony. In making the primary purpose determination, standard rules of hearsay, designed to identify some statements as reliable, will be relevant. Where no such primary purpose exists, the admissibility of a statement is the concern of state and federal rules of evidence, not the Confrontation Clause.[Footnote 5] Deciding this case also requires further explanation of the “ongoing emergency” circumstance addressed in Davis. Because Davis and Hammon arose in the domestic violence context, that was the situation “we had immediately in mind (for that was the case before us).” 547 U. S., at 826. We now face a new context: a nondomestic dispute, involving a victim found in a public location, suffering from a fatal gunshot wound, and a perpetrator whose location was unknown at the time the police located the victim. Thus, we confront for the first time circumstances in which the “ongoing emergency” discussed in Davis extends beyond an initial victim to a potential threat to the responding police and the public at large. This new context requires us to provide additional clarification with regard to what Davis meant by “the primary purpose of the interrogation is to enable police assistance to meet an ongoing emergency.” Id., at 822. III To determine whether the “primary purpose” of an interrogation is “to enable police assistance to meet an ongoing emergency,” Davis, 547 U. S., at 822, which would render the resulting statements nontestimonial, we objectively evaluate the circumstances in which the encounter occurs and the statements and actions of the parties. A The Michigan Supreme Court correctly understood that this inquiry is objective.[Footnote 6] 483 Mich., at 142, 768 N. W. 2d, at 70. Davis uses the word “objective” or “objectively” no fewer than eight times in describing the relevant inquiry. See 547 U. S., at 822, 826–828, 830–831, and n. 5; see, e.g., id., at 826 (“The question before us in Davis, then, is whether, objectively considered, the interrogation that took place in the course of the 911 call produced testimonial statements”). “Objectively” also appears in the definitions of both testimonial and nontestimonial statements that Davis established. Id., at 822. An objective analysis of the circumstances of an encounter and the statements and actions of the parties to it provides the most accurate assessment of the “primary purpose of the interrogation.” The circumstances in which an encounter occurs—e.g., at or near the scene of the crime versus at a police station, during an ongoing emergency or afterwards—are clearly matters of objective fact. The statements and actions of the parties must also be objectively evaluated. That is, the relevant inquiry is not the subjective or actual purpose of the individuals involved in a particular encounter, but rather the purpose that reasonable participants would have had, as ascertained from the individuals’ statements and actions and the circumstances in which the encounter occurred.[Footnote 7] B As our recent Confrontation Clause cases have explained, the existence of an “ongoing emergency” at the time of an encounter between an individual and the police is among the most important circumstances informing the “primary purpose” of an interrogation. See Davis, 547 U. S., at 828–830; Crawford, 541 U. S., at 65. The existence of an ongoing emergency is relevant to determining the primary purpose of the interrogation because an emergency focuses the participants on something other than “prov[ing] past events potentially relevant to later criminal prosecution.”[Footnote 8] Davis, 547 U. S., at 822. Rather, it focuses them on “end[ing] a threatening situation.” Id., at 832. Implicit in Davis is the idea that because the prospect of fabrication in statements given for the primary purpose of resolving that emergency is presumably significantly diminished, the Confrontation Clause does not require such statements to be subject to the crucible of cross-examination. This logic is not unlike that justifying the excited utterance exception in hearsay law. Statements “relating to a startling event or condition made while the declarant was under the stress of excitement caused by the event or condition,” Fed. Rule Evid. 803(2); see also Mich. Rule Evid. 803(2) (2010), are considered reliable because the declarant, in the excitement, presumably cannot form a falsehood. See Idaho v. Wright, 497 U. S. 805, 820 (1990) (“The basis for the ‘excited utterance’ exception . . . is that such statements are given under circumstances that eliminate the possibility of fabrication, coaching, or confabulation . . . ”); 5 J. Weinstein & M. Berger, Weinstein’s Federal Evidence §803.04[1] (J. McLaughlin ed., 2d ed. 2010) (same); Advisory Committee’s Notes on Fed. Rule Evid. 803(2), 28 U. S. C. App., p. 371 (same). An ongoing emergency has a similar effect of focusing an individual’s attention on responding to the emergency.[Footnote 9] Following our precedents, the court below correctly began its analysis with the circumstances in which Covington interacted with the police. 483 Mich., at 143, 768 N. W. 2d, at 71. But in doing so, the court construed Davis to have decided more than it did and thus employed an unduly narrow understanding of “ongoing emergency” that Davis does not require. First, the Michigan Supreme Court repeatedly and incorrectly asserted that Davis “defined” “ ‘ongoing emergency.’ ” 483 Mich., at 147, 768 N. W. 2d, at 73; see also id., at 144, 768 N. W. 2d, at 71–72. In fact, Davis did not even define the extent of the emergency in that case. The Michigan Supreme Court erroneously read Davis as deciding that “the statements made after the defendant stopped assaulting the victim and left the premises did not occur during an ‘ongoing emergency.’ ” 483 Mich., at 150, n. 15, 768 N. W. 2d, at 75, n. 15. We explicitly explained in Davis, however, that we were asked to review only the testimonial nature of Michelle McCottry’s initial statements during the 911 call; we therefore merely assumed the correctness of the Washington Supreme Court’s holding that admission of her other statements was harmless, without deciding whether those subsequent statements were also made for the primary purpose of resolving an ongoing emergency. 547 U. S., at 829. Second, by assuming that Davis defined the outer bounds of “ongoing emergency,” the Michigan Supreme Court failed to appreciate that whether an emergency exists and is ongoing is a highly context-dependent inquiry. See Brief for United States as Amicus Curiae 20. Davis and Hammon involved domestic violence, a known and identified perpetrator, and, in Hammon, a neutralized threat. Because Davis and Hammon were domestic violence cases, we focused only on the threat to the victims and assessed the ongoing emergency from the perspective of whether there was a continuing threat to them. 547 U. S., at 827, 829–830. Domestic violence cases like Davis and Hammon often have a narrower zone of potential victims than cases involving threats to public safety. An assessment of whether an emergency that threatens the police and public is ongoing cannot narrowly focus on whether the threat solely to the first victim has been neutralized because the threat to the first responders and public may continue. See 483 Mich., at 164, 768 N. W. 2d, at 82 (Corrigan, J., dissenting) (examining the threat to the victim, police, and the public); Brief for United States as Amicus Curiae 19–20 (“An emergency posed by an unknown shooter who remains at large does not automatically abate just because the police can provide security to his first victim”). The Michigan Supreme Court also did not appreciate that the duration and scope of an emergency may depend in part on the type of weapon employed. The court relied on Davis and Hammon, in which the assailants used their fists, as controlling the scope of the emergency here, which involved the use of a gun. The problem with that reasoning is clear when considered in light of the assault on Amy Hammon. Hershel Hammon was armed only with his fists when he attacked his wife, so removing Amy to a separate room was sufficient to end the emergency. 547 U. S., at 830–832. If Hershel had been reported to be armed with a gun, however, separation by a single household wall might not have been sufficient to end the emergency. Id., at 819. The Michigan Supreme Court’s failure to focus on the context-dependent nature of our Davis decision also led it to conclude that the medical condition of a declarant is irrelevant. 483 Mich., at 149, 768 N. W. 2d, at 74 (“The Court said nothing at all that would remotely suggest that whether the victim was in need of medical attention was in any way relevant to whether there was an ‘ongoing emergency’ ”). But Davis and Hammon did not present medical emergencies, despite some injuries to the victims. 547 U. S., at 818, 820. Thus, we have not previously considered, much less ruled out, the relevance of a victim’s severe injuries to the primary purpose inquiry. Taking into account the victim’s medical state does not, as the Michigan Supreme Court below thought, “rende[r] non-testimonial” “all statements made while the police are questioning a seriously injured complainant.” 483 Mich., at 149, 768 N. W. 2d, at 74. The medical condition of the victim is important to the primary purpose inquiry to the extent that it sheds light on the ability of the victim to have any purpose at all in responding to police questions and on the likelihood that any purpose formed would necessarily be a testimonial one. The victim’s medical state also provides important context for first responders to judge the existence and magnitude of a continuing threat to the victim, themselves, and the public. As the Solicitor General’s brief observes, Brief for United States as Amicus Curiae 20, and contrary to the Michigan Supreme Court’s claims, 483 Mich., at 147, 768 N. W. 2d, at 73, none of this suggests that an emergency is ongoing in every place or even just surrounding the victim for the entire time that the perpetrator of a violent crime is on the loose. As we recognized in Davis, “a conversation which begins as an interrogation to determine the need for emergency assistance” can “evolve into testimonial statements.” 547 U. S., at 828 (internal quotation marks omitted). This evolution may occur if, for example, a declarant provides police with information that makes clear that what appeared to be an emergency is not or is no longer an emergency or that what appeared to be a public threat is actually a private dispute. It could also occur if a perpetrator is disarmed, surrenders, is apprehended, or, as in Davis, flees with little prospect of posing a threat to the public. Trial courts can determine in the first instance when any transition from nontestimonial to testimonial occurs,[Footnote 10] and exclude “the portions of any statement that have become testimonial, as they do, for example, with unduly prejudicial portions of otherwise admissible evidence.” Id., at 829. Finally, our discussion of the Michigan Supreme Court’s misunderstanding of what Davis meant by “ongoing emergency” should not be taken to imply that the existence vel non of an ongoing emergency is dispositive of the testimonial inquiry. As Davis made clear, whether an ongoing emergency exists is simply one factor—albeit an important factor—that informs the ultimate inquiry regarding the “primary purpose” of an interrogation. Another factor the Michigan Supreme Court did not sufficiently account for is the importance of informality in an encounter between a victim and police. Formality is not the sole touchstone of our primary purpose inquiry because, although formality suggests the absence of an emergency and therefore an increased likelihood that the purpose of the interrogation is to “establish or prove past events potentially relevant to later criminal prosecution,” id., at 822, informality does not necessarily indicate the presence of an emergency or the lack of testimonial intent. Cf. id., at 826 (explaining that Confrontation Clause requirements cannot “readily be evaded” by the parties deliberately keeping the written product of an interrogation informal “instead of having the declarant sign a deposition”). The court below, however, too readily dismissed the informality of the circumstances in this case in a single brief footnote and in fact seems to have suggested that the encounter in this case was formal. 483 Mich., at 150, n. 16, 768 N. W. 2d, at 75, n. 16. As we explain further below, the questioning in this case occurred in an exposed, public area, prior to the arrival of emergency medical services, and in a disorganized fashion. All of those facts make this case distinguishable from the formal station-house interrogation in Crawford. See Davis, 547 U. S., at 827. C In addition to the circumstances in which an encounter occurs, the statements and actions of both the declarant and interrogators provide objective evidence of the primary purpose of the interrogation. See, e.g., Davis, 547 U. S., at 827 (“[T]he nature of what was asked and answered in Davis, again viewed objectively, was such that the elicited statements were necessary to be able to resolve the present emergency, rather than simply to learn (as in Crawford) what had happened in the past” (first emphasis added)). The Michigan Supreme Court did, at least briefly, conduct this inquiry. 483 Mich., at 144–147, 768 N. W. 2d, at 71–73. As the Michigan Supreme Court correctly recognized, id., at 140, n. 5, 768 N. W. 2d, at 69, n. 5, Davis requires a combined inquiry that accounts for both the declarant and the interrogator.[Footnote 11] In many instances, the primary purpose of the interrogation will be most accurately ascertained by looking to the contents of both the questions and the answers. To give an extreme example, if the police say to a victim, “Tell us who did this to you so that we can arrest and prosecute them,” the victim’s response that “Rick did it,” appears purely accusatory because by virtue of the phrasing of the question, the victim necessarily has prosecution in mind when she answers. The combined approach also ameliorates problems that could arise from looking solely to one participant. Predominant among these is the problem of mixed motives on the part of both interrogators and declarants. Police officers in our society function as both first responders and criminal investigators. Their dual responsibilities may mean that they act with different motives simultaneously or in quick succession. See New York v. Quarles, 467 U. S. 649, 656 (1984) (“Undoubtedly most police officers [deciding whether to give Miranda warnings in a possible emergency situation] would act out of a host of different, instinctive, and largely unverifiable motives—their own safety, the safety of others, and perhaps as well the desire to obtain incriminating evidence from the suspect”); see also Davis, 547 U. S., at 839 (Thomas, J., concurring in judgment in part and dissenting in part) (“In many, if not most, cases where police respond to a report of a crime, whether pursuant to a 911 call from the victim or otherwise, the purposes of an interrogation, viewed from the perspective of the police, are both to respond to the emergency situation and to gather evidence”). Victims are also likely to have mixed motives when they make statements to the police. During an ongoing emergency, a victim is most likely to want the threat to her and to other potential victims to end, but that does not necessarily mean that the victim wants or envisions prosecution of the assailant. A victim may want the attacker to be incapacitated temporarily or rehabilitated. Alternatively, a severely injured victim may have no purpose at all in answering questions posed; the answers may be simply reflexive. The victim’s injuries could be so debilitating as to prevent her from thinking sufficiently clearly to understand whether her statements are for the purpose of addressing an ongoing emergency or for the purpose of future prosecution.[Footnote 12] Taking into account a victim’s injuries does not transform this objective inquiry into a subjective one. The inquiry is still objective because it focuses on the understanding and purpose of a reasonable victim in the circumstances of the actual victim—circumstances that prominently include the victim’s physical state. The dissent suggests, post, at 3–4 (opinion of Scalia, J.), that we intend to give controlling weight to the “intentions of the police,” post, at 4. That is a misreading of our opinion. At trial, the declarant’s statements, not the interrogator’s questions, will be introduced to “establis[h] the truth of the matter asserted,” Crawford, 541 U. S., at 60, n. 9, and must therefore pass the Sixth Amendment test. See n. 11, supra. In determining whether a declarant’s statements are testimonial, courts should look to all of the relevant circumstances. Even Justice Scalia concedes that the interrogator is relevant to this evaluation, post, at 3, and we agree that “[t]he identity of an interrogator, and the content and tenor of his questions,” ibid., can illuminate the “primary purpose of the interrogation.” The dissent, see post, at 3–5 (opinion of Scalia, J.), criticizes the complexity of our approach, but we, at least, are unwilling to sacrifice accuracy for simplicity. Simpler is not always better, and courts making a “primary purpose” assessment should not be unjustifiably restrained from consulting all relevant information, including the statements and actions of interrogators. Objectively ascertaining the primary purpose of the interrogation by examining the statements and actions of all participants is also the approach most consistent with our past holdings. E.g., Davis, 547 U. S., at 822–823, n. 1 (noting that “volunteered testimony” is still testimony and remains subject to the requirements of the Confrontation Clause). IV As we suggested in Davis, when a court must determine whether the Confrontation Clause bars the admission of a statement at trial, it should determine the “primary purpose of the interrogation” by objectively evaluating the statements and actions of the parties to the encounter, in light of the circumstances in which the interrogation occurs. The existence of an emergency or the parties’ perception that an emergency is ongoing is among the most important circumstances that courts must take into account in determining whether an interrogation is testimonial because statements made to assist police in addressing an ongoing emergency presumably lack the testimonial purpose that would subject them to the requirement of confrontation.[Footnote 13] As the context of this case brings into sharp relief, the existence and duration of an emergency depend on the type and scope of danger posed to the victim, the police, and the public. Applying this analysis to the facts of this case is more difficult than in Davis because we do not have the luxury of reviewing a transcript of the conversation between the victim and the police officers. Further complicating our task is the fact that the trial in this case occurred before our decisions in Crawford and Davis. We therefore review a record that was not developed to ascertain the “primary purpose of the interrogation.” We first examine the circumstances in which the interrogation occurred. The parties disagree over whether there was an emergency when the police arrived at the gas station. Bryant argues, and the Michigan Supreme Court accepted, 483 Mich., at 147, 768 N. W. 2d, at 73, that there was no ongoing emergency because “there . . . was no criminal conduct occurring. No shots were being fired, no one was seen in possession of a firearm, nor were any witnesses seen cowering in fear or running from the scene.” Brief for Respondent 27. Bryant, while conceding that “a serious or life-threatening injury creates a medical emergency for a victim,” id., at 30, further argues that a declarant’s medical emergency is not relevant to the ongoing emergency determination. In contrast, Michigan and the Solicitor General explain that when the police responded to the call that a man had been shot and found Covington bleeding on the gas station parking lot, “they did not know who Covington was, whether the shooting had occurred at the gas station or at a different location, who the assailant was, or whether the assailant posed a continuing threat to Covington or others.” Brief for United States as Amicus Curiae 15; Brief for Petitioner 16; see also id., at 15 (“[W]hen an officer arrives on the scene and does not know where the perpetrator is, whether he is armed, whether he might have other targets, and whether the violence might continue at the scene or elsewhere, interrogation that has the primary purpose of establishing those facts to assess the situation is designed to meet the ongoing emergency and is nontestimonial”). The Michigan Supreme Court stated that the police asked Covington, “what had happened, who had shot him, and where the shooting had occurred.” 483 Mich., at 143, 768 N. W. 2d, at 71. The joint appendix contains the transcripts of the preliminary examination, suppression hearing, and trial testimony of five officers who responded to the scene and found Covington. The officers’ testimony is essentially consistent but, at the same time, not specific. The officers basically agree on what information they learned from Covington, but not on the order in which they learned it or on whether Covington’s statements were in response to general or detailed questions. They all agree that the first question was “what happened?” The answer was either “I was shot” or “Rick shot me.”[Footnote 14] As explained above, the scope of an emergency in terms of its threat to individuals other than the initial assailant and victim will often depend on the type of dispute involved. Nothing Covington said to the police indicated that the cause of the shooting was a purely private dispute or that the threat from the shooter had ended. The record reveals little about the motive for the shooting. The police officers who spoke with Covington at the gas station testified that Covington did not tell them what words Covington and Rick had exchanged prior to the shooting.[Footnote 15] What Covington did tell the officers was that he fled Bryant’s back porch, indicating that he perceived an ongoing threat.[Footnote 16] The police did not know, and Covington did not tell them, whether the threat was limited to him. The potential scope of the dispute and therefore the emergency in this case thus stretches more broadly than those at issue in Davis and Hammon and encompasses a threat potentially to the police and the public. This is also the first of our post-Crawford Confrontation Clause cases to involve a gun. The physical separation that was sufficient to end the emergency in Hammon was not necessarily sufficient to end the threat in this case; Covington was shot through the back door of Bryant’s house. Bryant’s argument that there was no ongoing emergency because “[n]o shots were being fired,” Brief for Respondent 27, surely construes ongoing emergency too narrowly. An emergency does not last only for the time between when the assailant pulls the trigger and the bullet hits the victim. If an out-of-sight sniper pauses between shots, no one would say that the emergency ceases during the pause. That is an extreme example and not the situation here, but it serves to highlight the implausibility, at least as to certain weapons, of construing the emergency to last only precisely as long as the violent act itself, as some have construed our opinion in Davis. See Brief for Respondent 23–25. At no point during the questioning did either Covington or the police know the location of the shooter. In fact, Bryant was not at home by the time the police searched his house at approximately 5:30 a.m. 483 Mich., at 136, 768 N. W. 2d, at 67. At some point between 3 a.m. and 5:30 a.m., Bryant left his house. At bottom, there was an ongoing emergency here where an armed shooter, whose motive for and location after the shooting were unknown, had mortally wounded Covington within a few blocks and a few minutes of the location where the police found Covington.[Footnote 17] This is not to suggest that the emergency continued until Bryant was arrested in California a year after the shooting. Id., at 137, 768 N. W. 2d, at 67. We need not decide precisely when the emergency ended because Covington’s encounter with the police and all of the statements he made during that interaction occurred within the first few minutes of the police officers’ arrival and well before they secured the scene of the shooting—the shooter’s last known location. We reiterate, moreover, that the existence vel non of an ongoing emergency is not the touchstone of the testimonial inquiry; rather, the ultimate inquiry is whether the “primary purpose of the interrogation [was] to enable police assistance to meet [the] ongoing emergency.” Davis, 547 U. S., at 822. We turn now to that inquiry, as informed by the circumstances of the ongoing emergency just described. The circumstances of the encounter provide important context for understanding Covington’s statements to the police. When the police arrived at Covington’s side, their first question to him was “What happened?”[Footnote 18] Covington’s response was either “Rick shot me” or “I was shot,” followed very quickly by an identification of “Rick” as the shooter. App. 76. In response to further questions, Covington explained that the shooting occurred through the back door of Bryant’s house and provided a physical description of the shooter. When he made the statements, Covington was lying in a gas station parking lot bleeding from a mortal gunshot wound to his abdomen. His answers to the police officers’ questions were punctuated with questions about when emergency medical services would arrive. Id., at 56–57 (suppression hearing testimony of Officer Brown). He was obviously in considerable pain and had difficulty breathing and talking. Id., at 75, 83–84 (testimony of Officer McCallister); id., at 101, 110–111 (testimony of Sgt. Wenturine); id., at 126, 137 (testimony of Officer Stuglin). From this description of his condition and report of his statements, we cannot say that a person in Covington’s situation would have had a “primary purpose” “to establish or prove past events potentially relevant to later criminal prosecution.” Davis, 547 U. S., at 822. For their part, the police responded to a call that a man had been shot. As discussed above, they did not know why, where, or when the shooting had occurred. Nor did they know the location of the shooter or anything else about the circumstances in which the crime occurred.[Footnote 19] The questions they asked—“what had happened, who had shot him, and where the shooting occurred,” 483 Mich., at 143, 768 N. W. 2d, at 71—were the exact type of questions necessary to allow the police to “ ‘assess the situation, the threat to their own safety, and possible danger to the potential victim’ ” and to the public, Davis, 547 U. S., at 832 (quoting Hiibel v. Sixth Judicial Dist. Court of Nev., Humboldt Cty., 542 U. S. 177, 186 (2004)), including to allow them to ascertain “whether they would be encountering a violent felon,”[Footnote 20] Davis, 547 U. S., at 827. In other words, they solicited the information necessary to enable them “to meet an ongoing emergency.” Id., at 822. Nothing in Covington’s responses indicated to the police that, contrary to their expectation upon responding to a call reporting a shooting, there was no emergency or that a prior emergency had ended. Covington did indicate that he had been shot at another location about 25 minutes earlier, but he did not know the location of the shooter at the time the police arrived and, as far as we can tell from the record, he gave no indication that the shooter, having shot at him twice, would be satisfied that Covington was only wounded. In fact, Covington did not indicate any possible motive for the shooting, and thereby gave no reason to think that the shooter would not shoot again if he arrived on the scene. As we noted in Davis, “initial inquiries” may “often . . . produce nontestimonial statements.” Id., at 832. The initial inquiries in this case resulted in the type of nontestimonial statements we contemplated in Davis. Finally, we consider the informality of the situation and the interrogation. This situation is more similar, though not identical, to the informal, harried 911 call in Davis than to the structured, station-house interview in Crawford. As the officers’ trial testimony reflects, the situation was fluid and somewhat confused: the officers arrived at different times; apparently each, upon arrival, asked Covington “what happened?”; and, contrary to the dissent’s portrayal, post, at 7–9 (opinion of Scalia, J.), they did not conduct a structured interrogation. App. 84 (testimony of Officer McCallister) (explaining duplicate questioning, especially as to “what happened?”); id., at 101–102 (testimony of Sgt. Wenturine) (same); id., at 126–127 (testimony of Officer Stuglin) (same). The informality suggests that the interrogators’ primary purpose was simply to address what they perceived to be an ongoing emergency, and the circumstances lacked any formality that would have alerted Covington to or focused him on the possible future prosecutorial use of his statements. Because the circumstances of the encounter as well as the statements and actions of Covington and the police objectively indicate that the “primary purpose of the interrogation” was “to enable police assistance to meet an ongoing emergency,” Davis, 547 U. S., at 822, Covington’s identification and description of the shooter and the location of the shooting were not testimonial hearsay. The Confrontation Clause did not bar their admission at Bryant’s trial. * * * For the foregoing reasons, we hold that Covington’s statements were not testimonial and that their admission at Bryant’s trial did not violate the Confrontation Clause. We leave for the Michigan courts to decide on remand whether the statements’ admission was otherwise permitted by state hearsay rules. The judgment of the Supreme Court of Michigan is vacated, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 The Supreme Court of Michigan held that the question whether the victim’s statements would have been admissible as “dying declarations” was not properly before it because at the preliminary examination, the prosecution, after first invoking both the dying declaration and excited utterance hearsay exceptions, established the factual foundation only for admission of the statements as excited utterances. The trial court ruled that the statements were admissible as excited utterances and did not address their admissibility as dying declarations. 483 Mich., at 153–154, 768 N. W. 2d, at 76–77. This occurred prior to our 2004 decision in Crawford v. Washington, 541 U. S. 36, where we first suggested that dying declarations, even if testimonial, might be admissible as a historical exception to the Confrontation Clause. Id., at 56, n. 6; see also Giles v. California, 554 U. S. 353, 358–359 (2008). We noted in Crawford that we “need not decide in this case whether the Sixth Amendment incorporates an exception for testimonial dying declarations.” 541 U. S., at 56, n. 6. Because of the State’s failure to preserve its argument with regard to dying declarations, we similarly need not decide that question here. See also post, p. __ (Ginsburg, J., dissenting). Footnote 2 We noted in Crawford that “[w]e use the term ‘interrogation’ in its colloquial, rather than any technical legal, sense,” and that “[j]ust as various definitions of ‘testimonial’ exist, one can imagine various definitions of ‘interrogation,’ and we need not select among them in this case.” 541 U. S., at 53, n. 4. Davis did not abandon those qualifications; nor do we do so here. Footnote 3 Davis explained that 911 operators “may at least be agents of law enforcement when they conduct interrogations of 911 callers,” and therefore “consider[ed] their acts to be acts of the police” for purposes of the opinion. 547 U. S., at 823, n. 2. Davis explicitly reserved the question of “whether and when statements made to someone other than law enforcement personnel are ‘testimonial.’ ” Ibid. We have no need to decide that question in this case either because Covington’s statements were made to police officers. The dissent also claims to reserve this question, see post, at 3, n. 1 (opinion of Scalia, J.), but supports one of its arguments by relying on King v. Brasier, 1 Leach 199, 200, 168 Eng. Rep. 202, 202–203 (K. B. 1779), which involved statements made by a child to her mother—a private citizen—just after the child had been sexually assaulted. See also Crawford v. Washington, 541 U. S. 36, 69–70 (2004) (Rehnquist, C. J., concurring in judgment) (citing King v. Brasier for the different proposition that “out-of-court statements made by someone other than the accused and not taken under oath, unlike ex parte depositions or affidavits, were generally not considered substantive evidence upon which a conviction could be based”). Footnote 4 Contrary to the dissent’s excited suggestion, nothing in this opinion casts “favorable light,” post, at 11 (opinion of Scalia, J.), on the conduct of Sir Walter Raleigh’s trial or other 16th- and 17th-century English treason trials. The dissent is correct that such trials are “unquestionably infamous,” ibid., and our decision here confirms, rather than undermines, that assessment. See also n. 17, infra. For all of the reasons discussed in Justice Thomas’ opinion concurring in the judgment, the situation presented in this case is nothing like the circumstances presented by Sir Walter Raleigh’s trial. See post, p. __. Footnote 5 See Davis v. Washington, 547 U. S. 813, 823–824 (2006) (explaining the question before the Court as “whether the Confrontation Clause applies only to testimonial hearsay” and answering in the affirmative because “[a] limitation so clearly reflected in the text of the constitutional provision must fairly be said to mark out not merely its ‘core,’ but its perimeter”). See also post, at 2 (Scalia, J., dissenting). Footnote 6 Bryant suggests that Michigan is arguing for “a subjective analysis of the intent of the interrogator’s questioning.” Brief for Respondent 12. We do not read Michigan’s brief to be arguing for a subjective inquiry, and any such argument would be in error. We do not understand the dissent to disagree that the inquiry is objective. Footnote 7 This approach is consistent with our rejection of subjective inquiries in other areas of criminal law. See, e.g., Whren v. United States, 517 U. S. 806, 813 (1996) (refusing to evaluate Fourth Amendment reasonableness subjectively in light of the officers’ actual motivations); New York v. Quarles, 467 U. S. 649, 655–656, and n. 6 (1984) (holding that an officer’s subjective motivation is irrelevant to determining the applicability of the public safety exception to Miranda v. Arizona, 384 U. S. 436 (1966)); Rhode Island v. Innis, 446 U. S. 291, 301–302 (1980) (holding that a police officer’s subjective intent to obtain incriminatory statements is not relevant to determining whether an interrogation has occurred). Footnote 8 The existence of an ongoing emergency must be objectively assessed from the perspective of the parties to the interrogation at the time, not with the benefit of hindsight. If the information the parties knew at the time of the encounter would lead a reasonable person to believe that there was an emergency, even if that belief was later proved incorrect, that is sufficient for purposes of the Confrontation Clause. The emergency is relevant to the “primary purpose of the interrogation” because of the effect it has on the parties’ purpose, not because of its actual existence. Footnote 9 Many other exceptions to the hearsay rules similarly rest on the belief that certain statements are, by their nature, made for a purpose other than use in a prosecution and therefore should not be barred by hearsay prohibitions. See, e.g., Fed. Rule Evid. 801(d)(2)(E) (statement by a co-conspirator during and in furtherance of the conspiracy); 803(4) (Statements for Purposes of Medical Diagnosis or Treatment); 803(6) (Records of Regularly Conducted Activity); 803(8) (Public Records and Reports); 803(9) (Records of Vital Statistics); 803(11) (Records of Religious Organizations); 803(12) (Marriage, Baptismal, and Similar Certificates); 803(13) (Family Records); 804(b)(3) (Statement Against Interest); see also Melendez-Diaz v. Massachusetts, 557 U. S. __, __ (2009) (slip op., at 18) (“Business and public records are generally admissible absent confrontation not because they qualify under an exception to the hearsay rules, but because—having been created for the administration of an entity’s affairs and not for the purpose of establishing or proving some fact at trial—they are not testimonial”); Giles v. California, 554 U. S., at 376 (noting in the context of domestic violence that “[s]tatements to friends and neighbors about abuse and intimidation and statements to physicians in the course of receiving treatment would be excluded, if at all, only by hearsay rules”); Crawford, 541 U. S., at 56 (“Most of the hearsay exceptions covered statements that by their nature were not testimonial—for example, business records or statements in furtherance of a conspiracy”). Footnote 10 Recognizing the evolutionary potential of a situation in criminal law is not unique to the Confrontation Clause context. We noted in Davis that “[j]ust as, for Fifth Amendment purposes, ‘police officers can and will distinguish almost instinctively between questions necessary to secure their own safety or the safety of the public and questions designed solely to elicit testimonial evidence from a suspect, … trial courts will recognize the point at which, for Sixth Amendment purposes, statements in response to interrogations become testimonial.” 547 U. S., at 829 (quoting New York v. Quarles, 467 U. S., at 658–659). Footnote 11 Some portions of Davis, however, have caused confusion about whether the inquiry prescribes examination of one participant to the exclusion of the other. Davis’ language indicating that a statement’s testimonial or nontestimonial nature derives from “the primary purpose of the interrogation,” 547 U. S., at 822 (emphasis added), could be read to suggest that the relevant purpose is that of the interrogator. In contrast, footnote 1 in Davis explains, “it is in the final analysis the declarant’s statements, not the interrogator’s questions, that the Confrontation Clause requires us to evaluate.” Id., at 822–823, n. 1. Bryant draws on the footnote to argue that the primary purpose inquiry must be conducted solely from the perspective of the declarant, and argues against adoption of a purpose-of-the-interrogator perspective. Brief for Respondent 10–13; see also Brief for Richard D. Friedman as Amicus Curiae 5–15. But this statement in footnote 1 of Davis merely acknowledges that the Confrontation Clause is not implicated when statements are offered “for purposes other than establishing the truth of the matter asserted.” Crawford, 541 U. S., at 60, n. 9. An interrogator’s questions, unlike a declarant’s answers, do not assert the truth of any matter. The language in the footnote was not meant to determine how the courts are to assess the nature of the declarant’s purpose, but merely to remind readers that it is the statements, and not the questions, that must be evaluated under the Sixth Amendment. Footnote 12 In such a situation, the severe injuries of the victim would undoubtedly also weigh on the credibility and reliability that the trier of fact would afford to the statements. Cf. Advisory Committee’s Notes on Fed. Rule Evid. 803(2), 28 U. S. C. App., p. 371 (noting that although the “theory” of the excited utterance exception “has been criticized on the ground that excitement impairs [the] accuracy of observation as well as eliminating conscious fabrication,” it “finds support in cases without number” (citing 6 J. Wigmore, Evidence §1750 (J. Chadbourn rev. 1976))). Footnote 13 Of course the Confrontation Clause is not the only bar to admissibility of hearsay statements at trial. State and federal rules of evidence prohibit the introduction of hearsay, subject to exceptions. Consistent with those rules, the Due Process Clauses of the Fifth and Fourteenth Amendments may constitute a further bar to admission of, for example, unreliable evidence. See Montana v. Egelhoff, 518 U. S. 37, 53 (1996) (plurality opinion) (“[E]rroneous evidentiary rulings can, in combination, rise to the level of a due process violation”); Dutton v. Evans, 400 U. S. 74, 96–97 (1970) (Harlan, J., concurring in result) (“[T]he Fifth and Fourteenth Amendments’ commands that federal and state trials, respectively, must be conducted in accordance with due process of law” is the “standard” by which to “test federal and state rules of evidence”). Footnote 14 See App. 76 (testimony of Officer McCallister); id., at 101, 113–114 (testimony of Sgt. Wenturine); id., at 127, 131–133 (testimony of Officer Stuglin). Covington told them that Rick had shot him through the back door of Rick’s house, id., at 127–128 (testimony of Officer Stuglin), located at the corner of Pennsylvania and Laura, id., at 102 (testimony of Sgt. Wenturine), and that Covington recognized Rick by his voice, id., at 128 (testimony of Officer Stuglin). Covington also gave them a physical description of Rick. Id., at 84–85, 93–94 (testimony of Officer McAllister); id., at 103, 115 (testimony of Sgt. Wenturine); id., at 134 (testimony of Officer Stuglin). Footnote 15 See id., at 114 (“Q Did he tell you what Rick said? A He said they were having a conversation. Q Did he tell you what Rick said? A He did not” (testimony of Sgt. Wenturine) (paragraph breaks omitted)); see also id., at 79 (testimony of Officer McAllister); id., at 128 (testimony of Officer Stuglin). Footnote 16 See id., at 127–128 (“A He said he’d went up, he went up to the back door of a house; that a person he said he knew, and he was knocking and he was knocking on the door he said he’d talked to somebody through the door. He said he recognized the voice. Q Did he say who it was that he recognized the voice of? A That’s when he told me it was, he said it was Rick a/k/a Buster. Q And did he say what the conversation was about at the door? A I don’t, I don’t believe so. Q All right. And did he say what happened there, whether or not they had a conversation or not, did he say what ended up happening? A He said what happened was that he heard a shot and then he started to turn to get off the porch and then another one and then that’s when he was hit by a gunshot” (testimony of Officer Stuglin) (paragraph breaks omitted)). Unlike the dissent’s apparent ability to read Covington’s mind, post, at 6 (opinion of Scalia, J.), we rely on the available evidence, which suggests that Covington perceived an ongoing threat. Footnote 17 It hardly bears mention that the emergency situation in this case is readily distinguishable from the “treasonous conspiracies of unknown scope, aimed at killing or overthrowing the king,” post, at 11, about which Justice Scalia’s dissent is quite concerned. Footnote 18 Although the dissent claims otherwise, post, at 7 (opinion of Scalia, J.), at least one officer asked Covington something akin to “how was he doing.” App. 131 (testimony of Officer Stuglin) (“A I approached the subject, the victim, Mr. Covington, on the ground and had asked something like what happened or are you okay, something to that line. . . . Q So you asked this man how are you, how are you doing? A Well, basically it’s, you know, what’s wrong, you know” (paragraph breaks omitted)). The officers also testified about their assessment of Covington’s wounds. See id., at 35 (suppression hearing testimony of Officer Brown) (“[H]e had blood . . . on the front of his body”); id., at 75 (testimony of Officer McCallister) (“It appeared he had a stomach wound of a gunshot”); id., at 132 (testimony of Officer Stuglin) (“Q Did you see the wound? A Yes, I did. Q You had to move some clothing to do that? A Yes” (paragraph breaks omitted)). Footnote 19 Contrary to the dissent’s suggestion, post, at 8 (opinion of Scalia, J.), and despite the fact that the record was developed prior to Davis’ focus on the existence of an “ongoing emergency,” the record contains some testimony to support the idea that the police officers were concerned about the location of the shooter when they arrived on the scene and thus to suggest that the purpose of the questioning of Covington was to determine the shooter’s location. See App. 136 (testimony of Officer Stuglin) (stating that upon arrival officers questioned the gas station clerk about whether the shooting occurred in the gas station parking lot and about concern for safety); see also ibid. (testimony of Officer Stuglin) (“Q . . . So you have some concern, there may be a person with a gun or somebody, a shooter right there in the immediate area? A Sure, yes. Q And you want to see that that area gets secured? A Correct. Q For your safety as well as everyone else? A Correct” (paragraph breaks omitted)); id., at 82 (testimony of Officer McCallister). But see id., at 83 (cross-examination of Officer McAllister) (“Q You didn’t, you didn’t look around and say, gee, there might be a shooter around here, I better keep an eye open? A I did not, no. That could have been my partner I don’t know” (paragraph breaks omitted)). Footnote 20 Hiibel, like our post-Crawford Confrontation Clause cases, involved domestic violence, which explains the Court’s focus on the security of the victim and the police: they were the only parties potentially threatened by the assailant. 542 U. S., at 186 (noting that the case involved a “domestic assault”).

564.US.2010_10-290

In asserting patent invalidity as a defense to an infringement action, an alleged infringer must contend with §282 of the Patent Act of 1952 (Act), under which “[a] patent shall be presumed valid” and “[t]he burden of establishing invalidity … shall rest on the party asserting” it. Since 1984, the Federal Circuit has read §282 to require a defendant seeking to overcome the presumption to persuade the factfinder of its invalidity defense by clear and convincing evidence. Respondents (collectively, i4i) hold the patent at issue, which claims an improved method for editing computer documents. After i4i sued petitioner Microsoft Corp. for willful infringement of that patent, Microsoft counterclaimed and sought a declaration that the patent was invalid under §102(b)’s on-sale bar, which precludes patent protection for any “invention” that was “on sale in this country” more than one year prior to the filing of a patent application. The parties agreed that, more than a year before filing its patent application, i4i had sold a software program known as S4 in the United States, but they disagreed over whether that software embodied the invention claimed in i4i’s patent. Relying on the undisputed fact that the S4 software was never presented to the Patent and Trademark Office (PTO) during its examination of the patent application, Microsoft objected to i4i’s proposed jury instruction that the invalidity defense must be proved by clear and convincing evidence. The District Court nevertheless gave that instruction, rejecting Microsoft’s alternative instruction proposing a preponderance of the evidence standard. The jury found that Microsoft willfully infringed the i4i patent and had failed to prove the patent’s invalidity. The Federal Circuit affirmed, relying on its settled interpretation of §282. Held: Section 282 requires an invalidity defense to be proved by clear and convincing evidence. Pp. 5–20. (a) The Court rejects Microsoft’s contention that a defendant need only persuade the jury of a patent invalidity defense by a preponderance of the evidence. Where Congress has prescribed the governing standard of proof, its choice generally controls. Steadman v. SEC, 450 U. S. 91, 95. Congress has made such a choice here. While §282 includes no express articulation of the standard of proof, where Congress uses a common-law term in a statute, the Court assumes the “term … comes with a common law meaning.” Safeco Ins. Co. of America v. Burr, 551 U. S. 47, 58. Here, by stating that a patent is “presumed valid,” §282, Congress used a term with a settled common-law meaning. Radio Corp. of America v. Radio Engineering Laboratories, Inc., 293 U. S. 1 (RCA), is authoritative. There, tracing nearly a century of case law, the Court stated, inter alia, that “there is a presumption of [patent] validity [that is] not to be overthrown except by clear and cogent evidence,” id., at 2. Microsoft’s contention that the Court’s pre-Act precedents applied a clear-and-convincing standard only in two limited circumstances is unavailing, given the absence of those qualifications from the Court’s cases. Also unpersuasive is Microsoft’s argument that the Federal Circuit’s interpretation must fail because it renders superfluous §282’s additional statement that “[t]he burden of establishing invalidity … shall rest on the party asserting” it. The canon against superfluity assists only where a competing interpretation gives effect “ ‘to every clause and word of a statute.’ ” Duncan v. Walker, 533 U. S. 167, 174. Here, no interpretation of §282 avoids excess language because, under either of Microsoft’s alternative theories—that the presumption only allocates the burden of production or that it shifts both the burdens of production and persuasion—the presumption itself would be unnecessary in light of §282’s additional statement as to the challenger’s burden. Pp. 5–13. (b) Also rejected is Microsoft’s argument that a preponderance standard must at least apply where the evidence before the factfinder was not before the PTO during the examination process. It is true enough that, in these circumstances, “the rationale underlying the presumption—that the PTO, in its expertise, has approved the claim—seems much diminished,” KSR Int’l Co. v. Teleflex Inc., 550 U. S. 398, 426, though other rationales may still animate the presumption. But the question remains whether Congress has specified the applicable standard of proof. As established here today, Congress did just that by codifying the common-law presumption of patent validity and, implicitly, the heightened standard of proof attached to it. The Court’s pre-Act cases never adopted or endorsed Microsoft’s fluctuating standard of proof. And they do not indicate, even in dicta, that anything less than a clear-and-convincing standard would ever apply to an invalidity defense. In fact, the Court indicated to the contrary. See RCA, 293 U. S., at 8. Finally, the Court often applied the heightened standard of proof without mentioning whether the relevant prior-art evidence had been before the PTO examiner, in circumstances strongly suggesting it had not. See, e.g., Smith v. Hall, 301 U. S. 216, 227, 233. Nothing in §282’s text suggests that Congress meant to depart from that understanding to enact a standard of proof that would rise and fall with the facts of each case. Indeed, had Congress intended to drop the heightened standard of proof where the evidence before the jury varied from that before the PTO, it presumably would have said so expressly. Those pre-Act cases where various Courts of Appeals observed that the presumption is weakened or dissipated where the evidence was never considered by the PTO should be read to reflect the commonsense principle that if the PTO did not have all material facts before it, its considered judgment may lose significant force. Cf. KSR, 550 U. S., at 427. Consistent with that principle, a jury may be instructed to evaluate whether the evidence before it is materially new, and if so, to consider that fact when determining whether an invalidity defense has been proved by clear and convincing evidence. Pp. 14–18. (c) This Court is in no position to judge the comparative force of the parties’ policy arguments as to the wisdom of the clear-and-convincing-evidence standard that Congress adopted. Congress specified the applicable standard of proof in 1952 when it codified the common-law presumption of patent validity. During the nearly 30 years that the Federal Circuit has interpreted §282 as the Court does today, Congress has often amended §282 and other patent laws, but apparently has never considered any proposal to lower the standard of proof. Indeed, Congress has left the Federal Circuit’s interpretation in place despite ongoing criticism, both from within the Federal Government and without. Accordingly, any recalibration of the standard of proof remains in Congress’ hands. Pp. 18–20. 598 F. 3d 831, affirmed. Sotomayor, J., delivered the opinion of the Court, in which Scalia, Kennedy, Ginsburg, Breyer, Alito, and Kagan, JJ., joined. Breyer, J., filed a concurring opinion, in which Scalia and Alito, JJ., joined. Thomas, J., filed an opinion concurring in the judgment. Roberts, C. J., took no part in the consideration or decision of the case.

Under §282 of the Patent Act of 1952, “[a] patent shall be presumed valid” and “[t]he burden of establishing in-validity of a patent or any claim thereof shall rest on the party asserting such invalidity.” 35 U. S. C. §282. We consider whether §282 requires an invalidity defense to be proved by clear and convincing evidence. We hold that it does. I A Pursuant to its authority under the Patent Clause, U. S. Const., Art. I, §8, cl. 8, Congress has charged the United States Patent and Trademark Office (PTO) with the task of examining patent applications, 35 U. S. C. §2(a)(1), and issuing patents if “it appears that the applicant is entitled to a patent under the law,” §131. Congress has set forth the prerequisites for issuance of a patent, which the PTO must evaluate in the examination process. To receive patent protection a claimed invention must, among other things, fall within one of the express categories of patentable subject matter, §101, and be novel, §102, and nonobvious, §103. Most relevant here, the on-sale bar of §102(b) precludes patent protection for any “invention” that was “on sale in this country” more than one year prior to the filing of a patent application. See generally Pfaff v. Wells Electronics, Inc., 525 U. S. 55, 67–68 (1998). In evaluating whether these and other statutory conditions have been met, PTO examiners must make various factual determinations—for instance, the state of the prior art in the field and the nature of the advancement embodied in the invention. See Dickinson v. Zurko, 527 U. S. 150, 153 (1999). Once issued, a patent grants certain exclusive rights to its holder, including the exclusive right to use the invention during the patent’s duration. To enforce that right, a patentee can bring a civil action for infringement if another person “without authority makes, uses, offers to sell, or sells any patented invention, within the United States.” §271(a); see also §281. Among other defenses under §282 of the Patent Act of 1952 (1952 Act), an alleged infringer may assert the invalidity of the patent—that is, he may attempt to prove that the patent never should have issued in the first place. See §§282(2), (3). A defendant may argue, for instance, that the claimed invention was obvious at the time and thus that one of the conditions of patentability was lacking. See §282(2); see also §103. “While the ultimate question of patent validity is one of law,” Graham v. John Deere Co. of Kansas City, 383 U. S. 1, 17 (1966) (citing Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., 340 U. S. 147, 155 (1950) (Douglas, J., concurring)); see post, at 1 (Breyer, J., concurring), the same factual questions underlying the PTO’s original examination of a patent ap-plication will also bear on an invalidity defense in an in-fringement action. See, e.g., 383 U. S., at 17 (describing the “basic factual inquiries” that form the “background” for evaluating obviousness); Pfaff, 525 U. S., at 67–69 (same, as to the on-sale bar). In asserting an invalidity defense, an alleged infringer must contend with the first paragraph of §282, which provides that “[a] patent shall be presumed valid” and “[t]he burden of establishing invalidity … rest[s] on the party asserting such invalidity.”[Footnote 1] Under the Federal Circuit’s reading of §282, a defendant seeking to overcome this presumption must persuade the factfinder of its in-validity defense by clear and convincing evidence. Judge Rich, a principal drafter of the 1952 Act, articulated this view for the court in American Hoist & Derrick Co. v. Sowa & Sons, Inc., 725 F. 2d 1350 (CA Fed. 1984). There, the Federal Circuit held that §282 codified “the existing presumption of validity of patents,” id., at 1359 (internal quotation marks omitted)—what, until that point, had been a common-law presumption based on “the basic proposition that a government agency such as the [PTO] was presumed to do its job,” ibid. Relying on this Court’s pre-1952 precedent as to the “force of the presumption,” ibid. (citing Radio Corp. of America v. Radio Engineering Laboratories, Inc., 293 U. S. 1 (1934) (RCA)), Judge Rich concluded: “[Section] 282 creates a presumption that a patent is valid and imposes the burden of proving invalidity on the attacker. That burden is constant and never changes and is to convince the court of invalidity by clear evidence.” 725 F. 2d, at 1360. In the nearly 30 years since American Hoist, the Federal Circuit has never wavered in this interpretation of §282. See, e.g., Greenwood v. Hattori Seiko Co., 900 F. 2d 238, 240–241 (CA Fed. 1990); Ultra-Tex Surfaces, Inc. v. Hill Bros. Chemical Co., 204 F. 3d 1360, 1367 (CA Fed. 2000); ALZA Corp. v. Andrx Pharmaceuticals, LLC, 603 F. 3d 935, 940 (CA Fed. 2010). B Respondents i4i Limited Partnership and Infrastructures for Information Inc. (collectively, i4i) hold the patent at issue in this suit. The i4i patent claims an improved method for editing computer documents, which stores a document’s content separately from the metacodes asso-ciated with the document’s structure. In 2007, i4i sued petitioner Microsoft Corporation for willful infringement, claiming that Microsoft’s manufacture and sale of certain Microsoft Word products infringed i4i’s patent. In addition to denying infringement, Microsoft counterclaimed and sought a declaration that i4i’s patent was invalid and unenforceable. Specifically and as relevant here, Microsoft claimed that the on-sale bar of §102(b) rendered the patent invalid, pointing to i4i’s prior sale of a software program known as S4. The parties agreed that, more than one year prior to the filing of the i4i patent application, i4i had sold S4 in the United States. They presented opposing arguments to the jury, however, as to whether that software embodied the invention claimed in i4i’s patent. Because the software’s source code had been destroyed years before the commencement of this litigation, the factual dispute turned largely on trial testimony by S4’s two inventors—also the named inventors on the i4i patent—both of whom testified that S4 did not practice the key invention disclosed in the patent. Relying on the undisputed fact that the S4 software was never presented to the PTO examiner, Microsoft objected to i4i’s proposed instruction that it was required to prove its invalidity defense by clear and convincing evidence. Instead, “if an instruction on the ‘clear and convincing’ burden were [to be] given,” App. 124a, n. 8, Microsoft re-quested the following: “ ‘Microsoft’s burden of proving invalidity and unenforceability is by clear and convincing evidence. However, Microsoft’s burden of proof with regard to its defense of invalidity based on prior art that the ex-aminer did not review during the prosecution of the patent-in-suit is by preponderance of the evidence.’ ” Ibid. Rejecting the hybrid standard of proof that Microsoft advocated, the District Court instructed the jury that “Microsoft has the burden of proving invalidity by clear and convincing evidence.” App. to Pet. for Cert. 195a. The jury found that Microsoft willfully infringed the i4i patent and that Microsoft failed to prove invalidity due to the on-sale bar or otherwise. Denying Microsoft’s post-trial motions, the District Court rejected Microsoft’s contention that the court improperly instructed the jury on the standard of proof. The Court of Appeals for the Federal Circuit affirmed.[Footnote 2] 598 F. 3d 831, 848 (2010). Relying on its settled interpretation of §282, the court explained that it could “discern [no] error” in the jury instruction requiring Microsoft to prove its invalidity defense by clear and convincing evidence. Ibid. We granted certiorari. 562 U. S. ___ (2010). II According to Microsoft, a defendant in an infringement action need only persuade the jury of an invalidity defense by a preponderance of the evidence. In the alternative, Microsoft insists that a preponderance standard must apply at least when an invalidity defense rests on evidence that was never considered by the PTO in the examination process. We reject both contentions.[Footnote 3] A Where Congress has prescribed the governing standard of proof, its choice controls absent “countervailing constitutional constraints.” Steadman v. SEC, 450 U. S. 91, 95 (1981). The question, then, is whether Congress has made such a choice here. As stated, the first paragraph of §282 provides that “[a] patent shall be presumed valid” and “[t]he burden of establishing invalidity of a patent or any claim thereof shall rest on the party asserting such invalidity.” Thus, by its express terms, §282 establishes a presumption of patent validity, and it provides that a challenger must overcome that presumption to prevail on an invalidity defense. But, while the statute explicitly specifies the burden of proof, it includes no express articulation of the standard of proof.[Footnote 4] Our statutory inquiry, however, cannot simply end there. We begin, of course, with “the assumption that the ordinary meaning of the language” chosen by Congress “accurately expresses the legislative purpose.” Engine Mfrs. Assn. v. South Coast Air Quality Management Dist., 541 U. S. 246, 252 (2004) (internal quotation marks omitted). But where Congress uses a common-law term in a statute, we assume the “term … comes with a common law meaning, absent anything pointing another way.” Safeco Ins. Co. of America v. Burr, 551 U. S. 47, 58 (2007) (citing Beck v. Prupis, 529 U. S. 494, 500–501 (2000)). Here, by stating that a patent is “presumed valid,” §282, Congress used a term with a settled meaning in the common law. Our decision in RCA, 293 U. S. 1, is authoritative. There, tracing nearly a century of case law from this Court and others, Justice Cardozo wrote for a unanimous Court that “there is a presumption of validity, a presumption not to be overthrown except by clear and cogent evidence.” Id., at 2. Although the “force” of the presumption found “varying expression” in this Court and elsewhere, id., at 7, Justice Cardozo explained, one “common core of thought and truth” unified the decisions: “[O]ne otherwise an infringer who assails the validity of a patent fair upon its face bears a heavy burden of persuasion, and fails unless his evidence has more than a dubious preponderance. If that is true where the assailant connects himself in some way with the title of the true inventor, it is so a fortiori where he is a stranger to the invention, without claim of title of his own. If it is true where the assailant launches his attack with evidence different, at least in form, from any theretofore produced in opposition to the patent, it is so a bit more clearly where the evidence is even verbally the same.” Id., at 8 (internal citation omitted).[Footnote 5] The common-law presumption, in other words, reflected the universal understanding that a preponderance standard of proof was too “dubious” a basis to deem a patent invalid. Ibid.; see also id., at 7 (“[A] patent … is presumed to be valid until the presumption has been overcome by convincing evidence of error”). Thus, by the time Congress enacted §282 and declared that a patent is “presumed valid,” the presumption of patent validity had long been a fixture of the common law. According to its settled meaning, a defendant raising an invalidity defense bore “a heavy burden of persuasion,” requiring proof of the defense by clear and convincing evidence. Id., at 8. That is, the presumption encompassed not only an allocation of the burden of proof but also an imposition of a heightened standard of proof. Under the general rule that a common-law term comes with its common-law meaning, we cannot conclude that Congress intended to “drop” the heightened standard proof from the presumption simply because §282 fails to reiterate it expressly. Neder v. United States, 527 U. S. 1, 23 (1999); see also id., at 21 (“ ‘Where Congress uses terms that have accumulated settled meaning under . . . the common law, [we] must infer, unless the statute otherwise dictates, that Congress means to incorporate the established meaning of those terms.’ ” (quoting Nationwide Mut. Ins. Co. v. Darden, 503 U. S. 318, 322 (1992))); Standard Oil Co. of N. J. v. United Sates, 221 U. S. 1, 59 (1911) (“[W]here words are employed in a statute which had at the time a well-known meaning at common law or in the law of this country they are presumed to have been used in that sense …”). “On the contrary, we must presume that Congress intended to incorporate” the heightened standard of proof, “unless the statute otherwise dictates.” Neder, 527 U. S., at 23 (internal quotation marks omitted). We recognize that it may be unusual to treat a presumption as alone establishing the governing standard of proof. See, e.g., J. Thayer, Preliminary Treatise on Evidence at the Common Law 336–337 (1898) (hereinafter Thayer) (“When … we read that the contrary of any particular presumption must be proved beyond a reasonable doubt, … it is to be recognized that we have something superadded to the rule of presumption, namely, another rule as to the amount of evidence which is needed to overcome the presumption”). But given how judges, including Justice Cardozo, repeatedly understood and explained the presumption of patent validity, we cannot accept Microsoft’s argument that Congress used the words “presumed valid” to adopt only a procedural device for “shifting the burden of production,” or for “shifting both the burden of production and the burden of persuasion.” Brief for Petitioner 21–22 (emphasis deleted). Whatever the significance of a presumption in the abstract, basic principles of statutory construction require us to assume that Congress meant to incorporate “the cluster of ideas” attached to the common-law term it adopted. Beck, 529 U. S., at 501 (internal quotation marks omitted). And RCA leaves no doubt that attached to the common-law presumption of patent validity was an expression as to its “force,” 293 U. S., at 7—that is, the standard of proof required to overcome it.[Footnote 6] Resisting the conclusion that Congress adopted the heightened standard of proof reflected in our pre-1952 cases, Microsoft contends that those cases applied a clear-and-convincing standard of proof only in two limited circumstances, not in every case involving an invalidity defense. First, according to Microsoft, the heightened standard of proof applied in cases “involving oral testimony of prior invention,” simply to account for the unreliability of such testimony. Brief for Petitioner 25. Second, Microsoft tells us, the heightened standard of proof applied to “invalidity challenges based on priority of invention,” where that issue had previously been litigated between the parties in PTO proceedings. Id., at 28. Squint as we may, we fail to see the qualifications that Microsoft purports to identify in our cases. They certainly make no appearance in RCA’s explanation of the presumption of patent validity. RCA simply said, without qualification, “that one otherwise an infringer who assails the validity of a patent fair upon its face bears a heavy burden of persuasion, and fails unless his evidence has more than a dubious preponderance.” 293 U. S., at 8; see also id., at 7 (“A patent regularly issued, and even more obviously a patent issued after a hearing of all the rival claimants, is presumed to be valid until the presumption has been overcome by convincing evidence of error” (emphasis added)). Nor do they appear in any of our cases as express limitations on the application of the heightened standard of proof. Cf., e.g., Smith v. Hall, 301 U. S. 216, 233 (1937) (citing RCA for the proposition that a “heavy burden of persuasion … rests upon one who seeks to negative novelty in a patent by showing prior use”); Mumm v. Jacob E. Decker & Sons, 301 U. S. 168, 171 (1937) (“Not only is the burden to make good this defense upon the party setting it up, but his burden is a heavy one, as it has been held that every reasonable doubt should be resolved against him” (internal quotation marks omitted)). In fact, Microsoft itself admits that our cases “could be read as announcing a heightened standard applicable to all invalidity assertions.” Brief for Petitioner 30 (emphasis deleted). Furthermore, we cannot agree that Microsoft’s proposed limitations are inherent—even if unexpressed—in our pre-1952 cases. As early as 1874 we explained that the burden of proving prior inventorship “rests upon [the de-fendant], and every reasonable doubt should be resolved against him,” without tying that rule to the vagaries and manipulability of oral testimony. Coffin v. Ogden, 18 Wall. 120, 124 (1874). And, more than 60 years later, we applied that rule where the evidence in support of a prior-use defense included documentary proof—not just oral testimony—in a case presenting no priority issues at all. See Smith, 301 U. S., at 221, 233. Thus, even if Congress searched for some unstated limitations on the heightened standard of proof in our cases, it would have found none.[Footnote 7] Microsoft also argues that the Federal Circuit’s interpretation of §282’s statement that “[a] patent shall be pre-sumed valid” must fail because it renders superfluous the statute’s additional statement that “[t]he burden of establishing invalidity of a patent … shall rest on the party asserting such invalidity.” We agree that if the presumption imposes a heightened standard of proof on the patent challenger, then it alone suffices to establish that the defendant bears the burden of persuasion. Cf. Director, Office of Workers’ Compensation Programs v. Greenwich Collieries, 512 U. S. 267, 278 (1994) (“A standard of proof … can apply only to a burden of persuasion”). Indeed, the Federal Circuit essentially recognized as much in American Hoist. See 725 F. 3d, at 1359. But the canon against superfluity assists only where a competing interpretation gives effect “ ‘to every clause and word of a statute.’ ” Duncan v. Walker, 533 U. S. 167, 174 (2001) (quoting United States v. Menasche, 348 U. S. 528, 538–539 (1955)); see Bruesewitz v. Wyeth LLC, 562 U. S. ___, ___ (2011) (slip op., at 12). Here, no interpretation of §282—including the two alternatives advanced by Microsoft—avoids excess language. That is, if the presumption only “allocates the burden of production,” Brief for Petitioner 21, or if it instead “shift[s] both the burden of production and the burden of persuasion,” id., at 22 (emphasis deleted), then it would be unnecessary in light of §282’s statement that the challenger bears the “burden of establishing invalidity.” See 21B Fed. Practice §5122, at 401 (“[T]he same party who has the burden of persuasion also starts out with the burden of producing evidence”). “There are times when Congress enacts provisions that are superfluous,” Corley v. United States, 556 U. S. ___, ___ (2009) (Alito, J., dissenting) (slip op., at 3), and the kind of excess language that Microsoft identifies in §282 is hardly unusual in comparison to other statutes that set forth a presumption, a burden of persuasion, and a standard of proof. Cf., e.g., 28 U. S. C. §2254(e)(1).[Footnote 8] B Reprising the more limited argument that it pressed below, Microsoft argues in the alternative that a preponderance standard must at least apply where the evidence before the factfinder was not before the PTO during the examination process. In particular, it relies on KSR Int’l Co. v. Teleflex Inc., 550 U. S. 398 (2007), where we ob-served that, in these circumstances, “the rationale underlying the presumption—that the PTO, in its expertise, has approved the claim—seems much diminished.” Id., at 426. That statement is true enough, although other rationales may animate the presumption in such circumstances. See The Barbed Wire Patent, 143 U. S. 275, 292 (1892) (explaining that because the patentee “first published this device; put it upon record; made use of it for a practical purpose; and gave it to the public … . doubts … concerning the actual inventor … should be resolved in favor of the patentee”); cf. Brief for United States as Amicus Curiae 33 (arguing that even when the administrative cor-rectness rationale has no relevance, the heightened standard of proof “serves to protect the patent holder’s reliance interests” in disclosing an invention to the public in exchange for patent protection). The question remains, however, whether Congress has specified the applicable standard of proof. As established, Congress did just that by codifying the common-law presumption of patent validity and, implicitly, the heightened standard of proof attached to it. Our pre-1952 cases never adopted or endorsed the kind of fluctuating standard of proof that Microsoft envisions. And they do not indicate, even in dicta, that anything less than a clear-and-convincing standard would ever apply to an invalidity defense raised in an infringement action. To the contrary, the Court spoke on this issue directly in RCA, stating that because the heightened standard of proof applied where the evidence before the court was “different” from that considered by the PTO, it applied even more clearly where the evidence was identical. 293 U. S., at 8. Likewise, the Court’s statement that a “dubious preponderance” will never suffice to sustain an invalidity defense, ibid., admitted of no apparent exceptions. Finally, this Court often applied the heightened standard of proof without any mention of whether the relevant prior-art evidence had been before the PTO examiner, in circumstances strongly suggesting it had not. See, e.g., Smith, 301 U. S., at 227, 233.[Footnote 9] Nothing in §282’s text suggests that Congress meant to depart from that understanding to enact a standard of proof that would rise and fall with the facts of each case. Indeed, had Congress intended to drop the heightened standard of proof where the evidence before the jury varied from that before the PTO—and thus to take the unusual and impractical step of enacting a variable standard of proof that must itself be adjudicated in each case, cf. Santosky v. Kramer, 455 U. S. 745, 757 (1982)[Footnote 10]—we assume it would have said so expressly. To be sure, numerous courts of appeals in the years preceding the 1952 Act observed that the presumption of validity is “weakened” or “dissipated” in the circumstance that the evidence in an infringement action was never considered by the PTO. See Jacuzzi Bros., Inc. v. Berkeley Pump Co., 191 F. 2d 632, 634 (CA9 1951) (“largely dissipated”); H. Schindler & Co. v. C. Saladino & Sons, 81 F. 2d 649, 651 (CA1 1936) (“weakened”); Gillette Safety Razor Co. v. Cliff Weil Cigar Co., 107 F. 2d 105, 107 (CA4 1939) (“greatly weakened”); Butler Mfg. Co. v. Enterprise Cleaning Co., 81 F. 2d 711, 716 (CA8 1936) (“weakened”). But we cannot read these cases to hold or even to suggest that a preponderance standard would apply in such circumstances, and we decline to impute such a reading to Congress. Instead, we understand these cases to reflect the same commonsense principle that the Federal Circuit has recognized throughout its existence—namely, that new evidence supporting an invalidity defense may “carry more weight” in an infringement action than evidence previously considered by the PTO, American Hoist, 725 F. 2d, at 1360. As Judge Rich explained: “When new evidence touching validity of the patent not considered by the PTO is relied on, the tribunal considering it is not faced with having to disagree with the PTO or with deferring to its judgment or with taking its expertise into account. The evidence may, therefore, carry more weight and go further toward sustaining the attacker’s unchanging burden.” Ibid. (emphasis deleted) See also SIBIA Neurosciences, Inc. v. Cadus Pharmaceutical Corp., 225 F. 3d 1349, 1355–1356 (CA Fed. 2000) (“[T]he alleged infringer’s burden may be more easily car-ried because of th[e] additional [evidence]”); Group One, Ltd. v. Hallmark Cards, Inc., 407 F. 3d 1297, 1306 (CA Fed. 2005) (similar). Simply put, if the PTO did not have all material facts before it, its considered judgment may lose significant force. Cf. KSR, 550 U. S., at 427. And, concomitantly, the challenger’s burden to persuade the jury of its invalidity defense by clear and convincing evidence may be easier to sustain. In this respect, although we have no occasion to endorse any particular formulation, we note that a jury instruction on the effect of new evidence can, and when requested, most often should be given. When warranted, the jury may be instructed to consider that it has heard evidence that the PTO had no opportunity to evaluate before granting the patent. When it is disputed whether the evidence presented to the jury differs from that evaluated by the PTO, the jury may be instructed to consider that question. In either case, the jury may be instructed to evaluate whether the evidence before it is materially new, and if so, to consider that fact when determining whether an invalidity defense has been proved by clear and convincing evidence. Cf., e.g., Mendenhall v. Cedarapids, Inc., 5 F. 3d 1557, 1563–1564 (CA Fed. 1993); see also Brief for International Business Machines Corp. as Amicus Curiae 31–37. Although Microsoft emphasized in its argument to the jury that S4 was never considered by the PTO, it failed to request an instruction along these lines from the District Court. Now, in its reply brief in this Court, Microsoft insists that an instruction of this kind was warranted. Reply Brief for Petitioner 22–23. That argument, however, comes far too late, and we therefore refuse to consider it. See Rent-A-Center, West, Inc. v. Jackson, 561 U. S. ___ , ___ (2010) (slip op., at 12); cf. Fed. Rule Civ. Proc. 51(d)(1)(B). III The parties and their amici have presented opposing views as to the wisdom of the clear-and-convincing-evidence standard that Congress adopted. Microsoft and its amici contend that the heightened standard of proof dampens innovation by unduly insulating “bad” patents from invalidity challenges. They point to the high invalidation rate as evidence that the PTO grants patent protection to too many undeserving “inventions.” They claim that inter partes reexamination proceedings before the PTO cannot fix the problem, as some grounds for invalidation (like the on-sale bar at issue here) cannot be raised in such proceedings. They question the deference that the PTO’s expert determinations warrant, in light of the agency’s resources and procedures, which they deem inadequate. And, they insist that the heightened standard of proof essentially causes juries to abdicate their role in reviewing invalidity claims raised in infringement actions. For their part, i4i and its amici, including the United States, contend that the heightened standard of proof properly limits the circumstances in which a lay jury overturns the considered judgment of an expert agency. They claim that the heightened standard of proof is an essential component of the patent “bargain,” see Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U. S. 141, 150–151 (1989), and the incentives for inventors to disclose their innovations to the public in exchange for patent protection. They disagree with the notion that the patent issuance rate is above the optimal level. They explain that limits on the reexamination process reflect a judgment by Congress as to the appropriate degree of interference with patentees’ reliance interests. Finally, they maintain that juries that are properly instructed as to the application of the clear-and-convincing-evidence standard can, and often do, find an invalidity defense established. We find ourselves in no position to judge the comparative force of these policy arguments. For nearly 30 years, the Federal Circuit has interpreted §282 as we do today. During this period, Congress has often amended §282, see, e.g., Pub. L. 104–141, §2, 109 Stat. 352; Pub. L. 98–417, §203, 98 Stat. 1603; not once, so far as we (and Microsoft) are aware, has it even considered a proposal to lower the standard of proof, see Tr. Oral Arg. 10. Moreover, Congress has amended the patent laws to account for concerns about “bad” patents, including by expanding the reexamination process to provide for inter partes proceedings. See Optional Inter Partes Reexamination Procedure Act of 1999, 113 Stat. 1501A–567, codified at 35 U. S. C. §311 et seq. Through it all, the evidentiary standard adopted in §282 has gone untouched. Indeed, Congress has left the Federal Circuit’s interpretation of §282 in place despite ongoing criticism, both from within the Federal Government and without.[Footnote 11] Congress specified the applicable standard of proof in 1952 when it codified the common-law presumption of patent validity. Since then, it has allowed the Federal Circuit’s correct interpretation of §282 to stand. Any re-calibration of the standard of proof remains in its hands. * * * For the reasons stated, the judgment of the Court of Appeals for the Federal Circuit is Affirmed. The Chief Justice took no part in the consideration or decision of this case. Footnote 1 As originally enacted in 1952, the first paragraph of §282 read: “A patent shall be presumed valid. The burden of establishing invalidity of a patent shall rest on a party asserting it.” 66 Stat. 812. Congress has since amended §282, inserting two sentences not relevant here and modifying the language of the second sentence to that in the text. Footnote 2 Although not relevant here, the Court of Appeals modified the effective date of the permanent injunction that the District Court entered in favor of i4i. 598 F. 3d 831, 863–864 (CA Fed. 2010). Footnote 3 i4i contends that Microsoft forfeited the first argument by failing to raise it until its merits brief in this Court. The argument, however, is within the scope of the question presented, and because we reject it on its merits, we need not decide whether it has been preserved. Footnote 4 A preliminary word on terminology is in order. As we have said, “[t]he term ‘burden of proof’ is one of the ‘the slipperiest members of the family of legal terms.’ ” Schaffer v. Weast, 546 U. S. 49, 56 (2005) (quoting 2 J. Strong, McCormick on Evidence §342, p. 433 (5th ed. 1999) (alteration omitted)). Historically, the term has encompassed two separate burdens: the “burden of persuasion” (specifying which party loses if the evidence is balanced), as well as the “burden of production” (specifying which party must come forward with evidence at various stages in the litigation). Ibid. Adding more confusion, the term “burden of proof” has occasionally been used as a synonym for “standard of proof.” E.g., Grogan v. Garner, 498 U. S. 279, 286 (1991). Here we use “burden of proof” interchangeably with “burden of persuasion” to identify the party who must persuade the jury in its favor to prevail. We use the term “standard of proof” to refer to the degree of certainty by which the factfinder must be persuaded of a factual conclusion to find in favor of the party bearing the burden of persuasion. See Addington v. Texas, 441 U. S. 418, 423 (1979). In other words, the term “standard of proof” specifies how difficult it will be for the party bearing the burden of persuasion to convince the jury of the facts in its favor. Various standards of proof are familiar—beyond a reasonable doubt, by clear and convincing evidence, and by a preponderance of the evidence. See generally 21B C. Wright & K. Graham, Federal Practice & Procedure §5122, pp. 405–411 (2d ed. 2005) (hereinafter Fed. Practice) (describing these and other standards of proof). Footnote 5 Among other cases, Justice Cardozo cited Cantrell v. Wallick, 117 U. S. 689, 695–696 (1886) (“Not only is the burden of proof to make good this defence upon the party setting it up, but … every reasonable doubt should be resolved against him” (internal quotation marks omitted)); Coffin v. Ogden, 18 Wall. 120, 124 (1874) (“The burden of proof rests upon [the defendant], and every reasonable doubt should be resolved against him”); The Barbed Wire Patent, 143 U. S. 275, 285 (1892) (“[This] principle has been repeatedly acted upon in the different circuits”); and Washburn v. Gould, 29 F. Cas. 312, 320 (No. 17,214) (CC Mass. 1844) (charging jury that “[i]f it should so happen, that your minds are led to a reasonable doubt on the question, inasmuch as it is incumbent on the defendant to satisfy you beyond that doubt, you will find for the plaintiff”). Footnote 6 Microsoft objects that this reading of §282 “conflicts with the usual understanding of presumptions.” Reply Brief for Petitioner 4. In sup-port, it relies on the “understanding” reflected in Federal Rule of Evidence 301, which explains the ordinary effect of a presumption in federal civil actions. That Rule, however, postdates the 1952 Act by nearly 30 years, and it is not dispositive of how Congress in 1952 understood presumptions generally, much less the presumption of pat-ent validity. In any event, the word “presumption” has often been used when another term might be more accurate. See Thayer 335 (“Often … maxims and ground principles get expressed in this form of a presumption perversely and inaccurately”). And, to the extent Congress used the words “presumed valid” in an imprecise way, we cannot fault it for following our lead. Footnote 7 In a similar vein, Microsoft insists that there simply was no settled presumption of validity for Congress to codify in 1952. Microsoft points to a handful of district court decisions, which “question[ed] whether any presumption of validity was warranted,” or which “required the patentee to prove the validity of his patent by a preponderance of the evidence.” Brief for Petitioner 24 (emphasis deleted; brackets and internal quotation marks omitted); see, e.g., Ginsberg v. Railway Express Agency, Inc., 72 F. Supp. 43, 44 (SDNY 1947) (stating, in dicta, that “[i]t may now well be said that no presumption whatever arises from the grant of patent”); see also post, at 1 (Thomas, J., concurring in judgment). RCA makes clear, however, that the presumption of patent validity had an established meaning traceable to the mid-19th century, 293 U. S. 1, 7–8 (1934); that some lower courts doubted its wisdom or even pretended it did not exist is of no moment. Microsoft may be correct that Congress enacted §282 to correct lower courts that required the patentee to prove the validity of a patent. See American Hoist & Derrick Co. v. Sowa & Sons, Inc., 725 F. 2d 1350, 1359 (CA Fed. 1984). But the language Congress selected reveals its intent not only to specify that the defendant bears the burden of proving invalidity but also that the evidence in support of the defense must be clear and convincing. Footnote 8 For those of us for whom it is relevant, the legislative history of §282 provides additional evidence that Congress meant to codify the judge-made presumption of validity, not to set forth a new presumption of its own making. The accompanying House and Senate Reports both explain that §282 “introduces a declaration of the presumption of va-lidity of a patent, which is now a statement made by courts in deci-sions, but has had no expression in the statute.” H. R. Rep. No. 1923, 82d Cong., 2d Sess., 10 (1952) (hereinafter H. R. Rep.); S. Rep. No. 1979, 82d Cong., 2d Sess., 9 (1952) (hereinafter S. Rep.). To the same effect, the Reviser’s Note indicates that §282’s “first paragraph declares the existing presumption of validity of patents.” Note following 35 U. S. C. §282 (1952 ed.). Prior to 1952, the existing patent laws already incorporated the sum and substance of the presumption as Microsoft would define it—that is, they “assign[ed] the burden of proving invalidity to the accused infringer,” Brief for Petitioner 14 (emphasis deleted). See 35 U. S. C. §69 (1946 ed.) (providing that a defendant in an infringement action “may plead” and “prove on trial” the invalidity of the patent as a defense); see also Patent Act of 1870, ch. 230, §61, 16 Stat. 208 (same); Patent Act of 1836, ch. 357, §15, 5 Stat. 123 (similar); Patent Act of 1793, ch. II, §6, 1 Stat. 322 (similar); Coffin, 18 Wall., at 124 (explaining that the Patent Act of 1836 “allowed a party sued for infringement to prove, among other defences, that the patentee was not the original and first inventor of the thing patented, or of a substantial and material part thereof claimed to be new” (internal quotation marks omitted)). The House and Senate Reports state, however, that §282 established a principle that previously “had no expression in the statute.” H. R. Rep., at 10; S. Rep., at 9. Thus, because the only thing missing from §282’s predecessor was the heightened standard of proof itself, Congress must have understood the presumption of patent validity to include the heightened standard of proof attached to it. Footnote 9 Microsoft cites numerous court of appeals decisions as support for its claim that a preponderance standard must apply in the event that the evidence in the infringement action varies from that considered by the PTO. We see no hint of the hybrid standard of proof that Microsoft advocates in these cases. Indeed, in some of these cases it appears that the court even evaluated the evidence according to a heightened standard of proof. See Jacuzzi Bros., Inc. v. Berkeley Pump Co., 191 F. 2d 632, 634 (CA9 1951) (“Although it is not expressly stated that th[e] conclusion [of invalidity] is based upon evidence establishing the thesis beyond a reasonable doubt, the Trial Court expressed no doubt. And the record shows that such conclusion was supported by substantial evidence”); Western Auto Supply Co. v. American-National Co., 114 F. 2d 711, 713 (CA6 1940) (concluding that the patent was invalid where the court “entertain[ed] no doubt” on the question). Footnote 10 Not the least of the impracticalities of such an approach arises from the fact that whether a PTO examiner considered a particular reference will often be a question without a clear answer. In granting a patent, an examiner is under no duty to cite every reference he considers. 1 Dept. of Commerce, PTO, Manual of Patent Examining Procedure §904.03, p. 900–51 (8th rev. ed. 2010) (“The examiner is not called upon to cite all references that may be available, but only the ‘best.’ Multiplying references, any one of which is as good as, but no better than, the others, adds to the burden and cost of prosecution and should therefore be avoided” (emphasis deleted)); Manual of Patent Examining Procedure §904.02, p. 129 (1st rev. ed. 1952) (same), http://www.uspto.gov/ web/offices/pac/mpep/old/E1R3_900.pdf (all Internet materials as visited June 6, 2011, and available in Clerk of Court’s case file); see also Brief for Respondents 45–46 (describing additional impracticalities). We see no indication in §282 that Congress meant to require collateral litigation on such an inherently uncertain question. Footnote 11 See, e.g., FTC, To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy 28 (Oct. 2003), http://www.ftc.gov/ os/2003/10/innovationrpt.pdf (recommending that “legislation be enacted specifying that challenges to the validity of a patent be determined based on a preponderance of the evidence”); Alsup, Memo to Congress: A District Judge’s Proposal for Patent Reform, 24 Berkeley Tech. L. J. 1647, 1655 (2009) (same); Lichtman & Lemley, Rethinking Patent Law’s Presumption of Validity, 60 Stan. L. Rev. 45, 60 (2007) (proposing “statutory amendment or … judicial reinterpretation of the existing statute and its associated case law” to lower the standard of proof to a preponderance of the evidence (footnote omitted)).

562.US.562

The Freedom of Information Act (FOIA) requires federal agencies to make Government records available to the public, subject to nine exemptions. This case concerns Exemption 2, which protects from disclosure material “related solely to the internal personnel rules and practices of an agency.” 5 U. S. C.§552(b)(2). This provision replaced an Administrative Procedure Act (APA) exemption for “any matter relating solely to the internal management of an agency,” 5 U. S. C. §1002 (1964 ed.). Congress believed that the “sweep” of the phrase “internal management” had led to excessive withholding, and drafted Exemption 2 “to have a narrower reach.” Department of Air Force v. Rose, 425 U. S. 352, 362–363. In Rose, the Court found that Exemption 2 could not be invoked to withhold Air Force Academy honor and ethics hearing summaries. The exemption, the Court suggested, primarily targets material concerning employee relations or human resources. But the Court stated a possible caveat: That understanding of the provision’s coverage governed “at least where the situation is not one where disclosure may risk circumvention of agency regulation.” Id., at 369. The D. C. Circuit subsequently converted this caveat into a new definition of Exemption 2’s scope, finding that the exemption also covered any “predominantly internal” materials whose disclosure would “significantly ris[k] circumvention of agency regulation or statutes.” Crooker v. Bureau of Alcohol, Tobacco & Firearms, 670 F. 2d 1051, 1056–1057, 1074. Courts now use the term “Low 2” for human resources and employee relations records and “High 2” for records whose disclosure would risk circumvention of the law. Petitioner Milner submitted FOIA requests for explosives data and maps used by respondent Department of the Navy (Navy or Government) in storing munitions at a naval base in Washington State. Stating that disclosure would threaten the security of the base and surrounding community, the Navy invoked Exemption 2 and refused to release the data. The District Court granted the Navy summary judgment, and the Court of Appeals affirmed, relying on the High 2 interpretation. Held: Because Exemption 2 encompasses only records relating to employee relations and human resources issues, the explosives maps and data requested here do not qualify for withholding under that exemption. Pp. 6–19. (a) Exemption 2 shields only those records relating to “personnel rules and practices.” When used as an adjective in this manner, the key statutory word “personnel” refers to human resources matters. For example, a “personnel department” deals with employee problems and interviews applicants for jobs. FOIA Exemption 6 provides another example, protecting certain “personnel … files” from disclosure. §552(b)(6). “[T]he common and congressional meaning of … ‘personnel file’ ” is a file maintained by a human resources office collecting personal information about employees, such as examination results and work performance evaluations. Rose, supra, at 377. Exemption 2 uses “personnel” in the exact same way. An agency’s “personnel rules and practices” all share a critical feature: They concern conditions of employment in federal agencies—such matters as hiring and firing, work rules and discipline, compensation and benefits. These items currently fall within the so-called Low 2 exemption. And under this Court’s construction of the statutory language, Low 2 is all of 2. FOIA’s purpose reinforces this reading. The statute’s goal is “broad disclosure,” and the exemptions must be “given a narrow compass.” Department of Justice v. Tax Analysts, 492 U. S. 136, 151. A narrow construction stands on especially firm footing with respect to Exemption 2, which was intended to hem in the expansive withholding that occurred under the prior APA exemption for “internal management” records. Exemption 2, as interpreted here, does not reach the requested explosives information. The data and maps, which calculate and visually portray the magnitude of hypothetical detonations, in no way relate to “personnel rules and practices,” as that term is most naturally understood. Pp. 6–10. (b) The Government’s two alternative readings of Exemption 2 cannot be squared with the statute. Pp. 10–17. (c) While the Navy has a strong security interest in shielding the explosives data and maps from public disclosure, the Government has other tools at hand to protect such information: FOIA Exemption 1 prevents access to classified documents; Exemption 3 applies to records that any other statute exempts from disclosure; and Exemption 7 protects “information compiled for law enforcement purposes” if its release, inter alia, “could reasonably be expected to endanger the life or physical safety of any individual,” §552(b)(7)(F). The Navy’s argument that the explosives information is exempt under Exemption 7 remains open for the Ninth Circuit to address on remand. And if these or other exemptions do not cover records whose release would threaten the Nation’s vital interests, the Government may of course seek relief from Congress. Pp. 17–18. 575 F. 3d 959, reversed and remanded. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Thomas, Ginsburg, Alito, and Sotomayor, JJ., joined. Alito, J., filed a concurring opinion. Breyer, J., filed a dissenting opinion.

The Freedom of Information Act (FOIA), 5 U. S. C. §552, requires federal agencies to make Government records available to the public, subject to nine exemptions for specific categories of material. This case concerns the scope of Exemption 2, which protects from disclosure material that is “related solely to the internal personnel rules and practices of an agency.” §552(b)(2). Respondent Department of the Navy (Navy or Government) invoked Exemption 2 to deny a FOIA request for data and maps used to help store explosives at a naval base in Washington State. We hold that Exemption 2 does not stretch so far. I Congress enacted FOIA to overhaul the public-disclosure section of the Administrative Procedure Act (APA), 5 U. S. C. §1002 (1964 ed.). That section of the APA “was plagued with vague phrases” and gradually became more “a withholding statute than a disclosure statute.” EPA v. Mink, 410 U. S. 73, 79 (1973). Congress intended FOIA to “permit access to official information long shielded unnecessarily from public view.” Id., at 80. FOIA thus mandates that an agency disclose records on request, unless they fall within one of nine exemptions. These exemptions are “explicitly made exclusive,” id., at 79, and must be “narrowly construed,” FBI v. Abramson, 456 U. S. 615, 630 (1982). At issue here is Exemption 2, which shields from compelled disclosure documents “related solely to the internal personnel rules and practices of an agency.” §552(b)(2). Congress enacted Exemption 2 to replace the APA’s exemption for “any matter relating solely to the internal management of an agency,” 5 U. S. C. §1002 (1964 ed.). Believing that the “sweep” of the phrase “internal management” had led to excessive withholding, Congress drafted Exemption 2 “to have a narrower reach.” Department of Air Force v. Rose, 425 U. S. 352, 362–363 (1976). We considered the extent of that reach in Department of Air Force v. Rose. There, we rejected the Government’s invocation of Exemption 2 to withhold case summaries of honor and ethics hearings at the United States Air Force Academy. The exemption, we suggested, primarily targets material concerning employee relations or human resources: “use of parking facilities or regulations of lunch hours, statements of policy as to sick leave, and the like.” Id., at 363 (quoting S. Rep. No. 813, 89th Cong., 1st Sess., 8 (1965) (hereinafter S. Rep.)); see Rose, 425 U. S., at 367. “[T]he general thrust” of Exemption 2, we explained, “is simply to relieve agencies of the burden of assembling and maintaining [such information] for public inspection.” Id., at 369. We concluded that the case summaries did not fall within the exemption because they “d[id] not concern only routine matters” of “merely internal significance.” Id., at 370. But we stated a possible caveat to our interpretation of Exemption 2: That understanding of the provision’s coverage governed, we wrote, “at least where the situation is not one where disclosure may risk circumvention of agency regulation.” Id., at 369. In Crooker v. Bureau of Alcohol, Tobacco & Firearms, 670 F. 2d 1051 (1981), the D. C. Circuit converted this caveat into a new definition of Exemption 2’s scope. Crooker approved the use of Exemption 2 to shield a manual designed to train Government agents in law enforcement surveillance techniques. The D. C. Circuit noted that it previously had understood Exemption 2 to “refe[r] only to ‘pay, pensions, vacations, hours of work, lunch hours, parking, etc.’ ” Id., at 1056 (quoting Jordan v. Department of Justice, 591 F. 2d 753, 763 (1978)). But the court now thought Exemption 2 should also cover any “predominantly internal” materials,[Footnote 1] Crooker, 670 F. 2d, at 1056–1057, whose disclosure would “significantly ris[k] circumvention of agency regulations or statutes,” id., at 1074. This construction of Exemption 2, the court reasoned, flowed from FOIA’s “overall design,” its legislative history, “and even common sense,” because Congress could not have meant to “enac[t] a statute whose provisions undermined … the effectiveness of law enforcement agencies.” Ibid. In the ensuing years, three Courts of Appeals adopted the D. C. Circuit’s interpretation of Exemption 2. See 575 F. 3d 959, 965 (CA9 2009) (case below); Massey v. FBI, 3 F. 3d 620, 622 (CA2 1993); Kaganove v. EPA, 856 F. 2d 884, 889 (CA7 1988).[Footnote 2] And that interpretation spawned a new terminology: Courts applying the Crooker approach now refer to the “Low 2” exemption when discussing materials concerning human resources and employee relations, and to the “High 2” exemption when assessing records whose disclosure would risk circumvention of the law. See, e.g., 575 F. 3d, at 963; Schiller v. NLRB, 964 F. 2d 1205, 1208 (CADC 1992). Congress, as well, took notice of the D. C. Circuit’s decision, borrowing language from Crooker to amend Exemption 7(E) when next enacting revisions to FOIA. The amended version of Exemption 7(E) shields certain “records or information compiled for law enforcement purposes” if their disclosure “could reasonably be expected to risk circumvention of the law.” §552(b)(7)(E); see Freedom of Information Reform Act of 1986, §1802(a), 100 Stat. 3207–49. II The FOIA request at issue here arises from the Navy’s operations at Naval Magazine Indian Island, a base in Puget Sound, Washington. The Navy keeps weapons, ammunition, and explosives on the island. To aid in the storage and transport of these munitions, the Navy uses data known as Explosive Safety Quantity Distance (ESQD) information. 575 F. 3d, at 962. ESQD information prescribes “minimum separation distances” for explosives and helps the Navy design and construct storage facilities to prevent chain reactions in case of detonation. Ibid. The ESQD calculations are often incorporated into specialized maps depicting the effects of hypothetical explosions. See, e.g., App. 52. In 2003 and 2004, petitioner Glen Milner, a Puget Sound resident, submitted FOIA requests for all ESQD information relating to Indian Island. 575 F. 3d, at 962. The Navy refused to release the data, stating that disclosure would threaten the security of the base and surrounding community. In support of its decision to withhold the records, the Navy invoked Exemption 2. Ibid.[Footnote 3] The District Court granted summary judgment to the Navy, and the Court of Appeals affirmed, relying on the High 2 interpretation developed in Crooker. 575 F. 3d, at 963. The Court of Appeals explained that the ESQD information “is predominantly used for the internal purpose of instructing agency personnel on how to do their jobs.” Id., at 968. And disclosure of the material, the court determined, “would risk circumvention of the law” by “point[ing] out the best targets for those bent on wreaking havoc”—for example, “[a] terrorist who wished to hit the most damaging target.” Id., at 971. The ESQD information, the court concluded, therefore qualified for a High 2 exemption. 575 F. 3d, at 971. We granted certiorari in light of the Circuit split respecting Exemption 2’s meaning, 561 U. S. ___ (2010), and we now reverse. III Our consideration of Exemption 2’s scope starts with its text. See, e.g., Park ’N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U. S. 189, 194 (1985) (“Statutory construction must begin with the language employed by Congress and the assumption that the ordinary meaning of that language accurately expresses the legislative purpose”). Judicial decisions since FOIA’s enactment have analyzed and reanalyzed the meaning of the exemption. But comparatively little attention has focused on the provision’s 12 simple words: “related solely to the internal personnel rules and practices of an agency.” The key word in that dozen—the one that most clearly marks the provision’s boundaries—is “personnel.” When used as an adjective, as it is here to modify “rules and practices,” that term refers to human resources matters. “Personnel,” in this common parlance, means “the selection, placement, and training of employees and … the formulation of policies, procedures, and relations with [or involving] employees or their representatives.” Webster’s Third New International Dictionary 1687 (1966) (hereinafter Webster’s). So, for example, a “personnel department” is “the department of a business firm that deals with problems affecting the employees of the firm and that usually interviews applicants for jobs.” Random House Dictionary 1075 (1966) (hereinafter Random House). “Personnel management” is similarly “the phase of management concerned with the engagement and effective utilization of manpower to obtain optimum efficiency of human resources.” Webster’s 1687. And a “personnel agency” is “an agency for placing employable persons in jobs; employment agency.” Random House 1075. FOIA itself provides an additional example in Exemption 6. See Ratzlaf v. United States, 510 U. S. 135, 143 (1994) (“A term appearing in several places in a statutory text is generally read the same way each time it appears”). That exemption, just a few short paragraphs down from Exemption 2, protects from disclosure “personnel and medical files and similar files the disclosure of which would constitute a clearly unwarranted invasion of personal privacy.” §552(b)(6). Here too, the statute uses the term “personnel” as a modifier meaning “human resources.” See Tr. of Oral Arg. 32 (“[The Court:] It’s [an] H. R. file, right? [The Government:] That’s generally true”). As we recognized in Rose, “the common and congressional meaning of … ‘personnel file’ ” is the file “showing, for example, where [an employee] was born, the names of his parents, where he has lived from time to time, his … school records, results of examinations, [and] evaluations of his work performance.” 425 U. S., at 377. It is the file typically maintained in the human resources office—otherwise known (to recall an example offered above) as the “personnel department.” Ibid. Exemption 2 uses “personnel” in the exact same way. An agency’s “personnel rules and practices” are its rules and practices dealing with employee relations or human resources. The D. C. Circuit, in a pre-Crooker decision, gave as examples “matters relating to pay, pensions, vacations, hours of work, lunch hours, parking, etc.” Jordan, 591 F. 2d, at 763; see supra, at 3. That “etc.” is important; we doubt any court could know enough about the Federal Government’s operations to formulate a comprehensive list. But all the rules and practices referenced in Exemption 2 share a critical feature: They concern the conditions of employment in federal agencies—such matters as hiring and firing, work rules and discipline, compensation and benefits.[Footnote 4] Courts in practice have had little difficulty identifying the records that qualify for withholding under this reading: They are what now commonly fall within the Low 2 exemption. Our construction of the statutory language simply makes clear that Low 2 is all of 2 (and that High 2 is not 2 at all, see infra, at 10–14). The statute’s purpose reinforces this understanding of the exemption. We have often noted “the Act’s goal of broad disclosure” and insisted that the exemptions be “given a narrow compass.” Department of Justice v. Tax Analysts, 492 U. S. 136, 151 (1989); see Department of Interior v. Klamath Water Users Protective Assn., 532 U. S. 1, 7–8 (2001).[Footnote 5] This practice of “constru[ing] FOIA exemptions narrowly,” Department of Justice v. Landano, 508 U. S. 165, 181 (1993), stands on especially firm footing with respect to Exemption 2. As described earlier, Congress worded that provision to hem in the prior APA exemption for “any matter relating solely to the internal management of an agency,” which agencies had used to prevent access to masses of documents. See Rose, 425 U. S., at 362. We would ill-serve Congress’s purpose by construing Exemption 2 to reauthorize the expansive withholding that Congress wanted to halt. Our reading instead gives the exemption the “narrower reach” Congress intended, id., at 363, through the simple device of confining the provision’s meaning to its words. The Government resists giving “personnel” its plain meaning on the ground that Congress, when drafting Exemption 2, considered but chose not to enact language exempting “internal employment rules and practices.” Brief for Respondent 30–34, and n. 11. This drafting history, the Navy maintains, proves that Congress did not wish “to limit the Exemption to employment-related matters,” id., at 31, even if the adjective “personnel” conveys that meaning in other contexts, id., at 41. But we think the Navy’s evidence insufficient: The scant history concerning this word change as easily supports the inference that Congress merely swapped one synonym for another. Cf. Mead Corp. v. Tilley, 490 U. S. 714, 723 (1989) (noting with respect to the “unexplained disappearance of one word from an unenacted bill” that “mute intermediate legislative maneuvers are not reliable” aids to statutory interpretation (internal quotation marks omitted)). Those of us who make use of legislative history believe that clear evidence of congressional intent may illuminate ambiguous text. We will not take the opposite tack of allowing ambiguous legislative history to muddy clear statutory language. Exemption 2, as we have construed it, does not reach the ESQD information at issue here. These data and maps calculate and visually portray the magnitude of hypothetical detonations. By no stretch of imagination do they relate to “personnel rules and practices,” as that term is most naturally understood. They concern the physical rules governing explosives, not the workplace rules governing sailors; they address the handling of dangerous materials, not the treatment of employees. The Navy therefore may not use Exemption 2, interpreted in accord with its plain meaning to cover human resources matters, to prevent disclosure of the requested maps and data. IV The Government offers two alternative readings of Exemption 2 to support withholding the ESQD information. We cannot square either with the statute. A The Navy first encourages us to adopt the construction of Exemption 2 pioneered by Crooker, which shields material not only if it meets the criteria set out above (Low 2), but also if it is “predominant[ly] interna[l]” and its “disclosure would significantly risk[] circumvention of federal agency functions” (High 2). Brief for Respondent 41 (internal quotation marks omitted). The dissent, too, favors this reading of the statute. Post, at 1. But the Crooker interpretation, as already suggested, suffers from a patent flaw: It is disconnected from Exemption 2’s text. The High 2 test (in addition to substituting the word “predominantly” for “solely,” see n. 1, supra) ignores the plain meaning of the adjective “personnel,” see supra, at 6–9, and adopts a circumvention requirement with no basis or referent in Exemption 2’s language. Indeed, the only way to arrive at High 2 is by taking a red pen to the statute—“cutting out some” words and “pasting in others” until little of the actual provision remains. Elliott v. Department of Agriculture, 596 F. 3d 842, 845 (CADC 2010). Because this is so, High 2 is better labeled “Non 2” (and Low 2 … just 2). In support of its text-light approach to the statute, the Government relies primarily on legislative history, placing particular emphasis on the House Report concerning FOIA. See Brief for Respondent 33–38. A statement in that Report buttresses the High 2 understanding of the exemption and, indeed, specifically rejects the Low 2 construction. According to the Report: “Operating rules, guidelines, and manuals of procedure for Government investigators or examiners would be exempt from disclosure [under Exemption 2], but this exemption would not cover … employee relations and working conditions and routine administrative procedures.” H. R. Rep. No. 1497, 89th Cong., 2d Sess., 10 (1966). But the Senate Report says exactly the opposite, explaining in support of a Low 2 interpretation that the phrase “internal personnel rules and practices of an agency” means “rules as to personnel’s use of parking facilities or regulation of lunch hours, statements of policy as to sick leave, and the like.” S. Rep., at 8.[Footnote 6] In Rose, we gave reasons for thinking the Senate Report the more reliable of the two. See 425 U. S., at 366. But the more fundamental point is what we said before: Legislative history, for those who take it into account, is meant to clear up ambiguity, not create it. See supra, at 9; Wong Yang Sung v. McGrath, 339 U. S. 33, 49 (1950) (declining to consult legislative history when that “history is more conflicting than the text is ambiguous”). When presented, on the one hand, with clear statutory language and, on the other, with dueling committee reports, we must choose the language. The Government also advances, in support of Crooker’s High 2 approach, an argument based on subsequent legislative action. Congress, the Government notes, amended Exemption 7(E) in 1986 to cover law enforcement records whose production “would disclose techniques and procedures for law enforcement investigations or prosecutions, or would disclose guidelines for law enforcement investigations or prosecutions if such disclosure could reason-ably be expected to risk circumvention of the law.” §552(b)(7)(E). That amendment, the Government contends, codified Crooker’s “circumvention of the law” standard and, in so doing, ratified Crooker’s holding. Brief for Respondent 42–43. The dissent likewise counts as significant that Congress “t[ook] note” of Crooker in revising FOIA. Post, at 9; see post, at 2. But the Government and the dissent neglect the key feature of the 1986 amendment: Congress modified not Exemption 2 (the subject of Crooker), but instead Exemption 7(E). And the Crooker construction of Exemption 2 renders Exemption 7(E) superfluous and so deprives that amendment of any effect. See, e.g., TRW Inc. v. Andrews, 534 U. S. 19, 31 (2001) (noting canon that statutes should be read to avoid making any provision “superfluous, void, or insignificant” (internal quotation marks omitted)). We cannot think of any document eligible for withholding under Exemption 7(E) that the High 2 reading does not capture: The circumvention standard is the same, and the law enforcement records listed in Exemption 7(E) are “predominantly internal.” So if Congress had agreed with Crooker’s reading of Exemption 2, it would have had no reason to alter Exemption 7(E). In that event, Congress would either have left the statute alone (on the theory that Crooker would do the necessary work) or would have amended Exemption 2 specifically to ratify Crooker. The decision instead to amend Exemption 7(E) suggests that Congress approved the circumvention standard only as to law enforcement materials, and not as to the wider set of records High 2 covers. Perhaps this legislative action does not show that Congress affirmatively disagreed with Crooker; maybe Congress was agnostic about whether the circumvention standard should apply to other records. But one thing is clear: The 1986 amendment does not ratify, approve, or otherwise signal agreement with Crooker’s interpretation of Exemption 2. This argument therefore cannot save the High 2 construction. The dissent offers one last reason to embrace High 2, and indeed stakes most of its wager on this argument. Crooker, the dissent asserts, “has been consistently relied upon and followed for 30 years” by other lower courts. Post, at 9; see post, at 1–2. But this claim, too, trips at the starting gate. It would be immaterial even if true, because we have no warrant to ignore clear statutory language on the ground that other courts have done so. And in any event, it is not true. Prior to Crooker, three Circuits adopted the reading of Exemption 2 we think right, and they have not changed their minds. See n. 2, supra.[Footnote 7] Since Crooker, three other Circuits have accepted the High 2 reading. See supra, at 3. One Circuit has reserved judgment on the High 2-Low 2 debate. See Audubon Society v. Forest Serv., 104 F. 3d 1201, 1203–1204 (CA10 1997). And the rest have not considered the matter. (No one should think Crooker has been extensively discussed or debated in the Courts of Appeals. In the past three decades, Crooker’s analysis of Exemption 2 has been cited a sum total of five times in federal appellate decisions outside the D. C. Circuit—on average, once every six years.) The result is a 4 to 3 split among the Circuits.[Footnote 8] We will not flout all usual rules of statutory interpretation to take the side of the bare majority. B Presumably because Crooker so departs from Exemption 2’s language, the Government also offers another construction, which it says we might adopt “on a clean slate,” “based on the plain text … alone.” Brief for Respondent 15. On this reading, the exemption “encompasses records concerning an agency’s internal rules and practices for its personnel to follow in the discharge of their governmental functions.” Id., at 20; see also id., at 13–14 (Exemption 2 “applies generally to matters concerning internal rules and practices to guide agency personnel in performing their duties”). According to the Government, this interpretation makes sense because “the phrase ‘personnel rules and practices of an agency’ is logically understood to mean an agency’s rules and practices for its personnel.” Id., at 20 (emphasis added). But the purported logic in the Government’s definition eludes us. We would not say, in ordinary parlance, that a “personnel file” is any file an employee uses, or that a “personnel department” is any department in which an employee serves. No more would we say that a “personnel rule or practice” is any rule or practice that assists an employee in doing her job. The use of the term “personnel” in each of these phrases connotes not that the file or department or practice/rule is for personnel, but rather that the file or department or practice/rule is about personnel—i.e., that it relates to employee relations or human resources. This case well illustrates the point. The records requested, as earlier noted, are explosives data and maps showing the distances that potential blasts travel. This information no doubt assists Navy personnel in storing munitions. But that is not to say that the data and maps relate to “personnel rules and practices.” No one staring at these charts of explosions and using ordinary language would describe them in this manner. Indeed, the Government’s “clean slate” construction reaches such documents only by stripping the word “personnel” of any real meaning. Under this interpretation, an agency’s “internal personnel rules and practices” appears to mean all its internal rules and practices. That is because agencies necessarily operate through personnel, and so all their internal rules and practices are for personnel. The modifier “personnel,” then, does no modifying work; it does not limit the class of internal rules and practices that Exemption 2 covers. What is most naturally viewed as the provision’s key word—the term that ought to define its scope—does nothing more than state the truism that in an agency it is “personnel” who follow internal rules and practices. And this odd reading would produce a sweeping exemption, posing the risk that FOIA would become less a disclosure than “a withholding statute.” Mink, 410 U. S., at 79. Many documents an agency generates in some way aid employees in carrying out their responsibilities. If Exemption 2 were to reach all these records, it would tend to engulf other FOIA exemptions, rendering ineffective the limitations Congress placed on their application. Exemption 7, for example, shields records compiled for law enforcement purposes, but only if one of six specified criteria is met. §552(b)(7). Yet on the Government’s view, an agency could bypass these restrictions by invoking Exemption 2 whenever law enforcement records guide personnel in performing their duties. Indeed, an agency could use Exemption 2 as an all-purpose back-up provision to withhold sensitive records that do not fall within any of FOIA’s more targeted exemptions.[Footnote 9] Interpreted in this way, Exemption 2—call it “Super 2” now—would extend, rather than narrow, the APA’s former exemption for records relating to the “internal management of an agency.” 5 U. S. C. §1002 (1964 ed.). We doubt that even the “internal management” provision, which Congress thought allowed too much withholding, see supra, at 2, would have protected all information that guides employees in the discharge of their duties, including the explosives data and maps in this case. And perhaps needless to say, this reading of Exemption 2 violates the rule favoring narrow construction of FOIA exemptions. See, e.g., Abramson, 456 U. S., at 630; Rose, 425 U. S., at 361. Super 2 in fact has no basis in the text, context, or purpose of FOIA, and we accordingly reject it. V Although we cannot interpret Exemption 2 as the Government proposes, we recognize the strength of the Navy’s interest in protecting the ESQD data and maps and other similar information. The Government has informed us that “[p]ublicly disclosing the [ESQD] information would significantly risk undermining the Navy’s ability to safely and securely store military ordnance,” Brief for Respondent 47, and we have no reason to doubt that representation. The Ninth Circuit similarly cautioned that disclosure of this information could be used to “wrea[k] havoc” and “make catastrophe more likely.” 575 F. 3d, at 971. Concerns of this kind—a sense that certain sensitive information should be exempt from disclosure—in part led the Crooker court to formulate the High 2 standard. See 670 F. 2d, at 1074 (contending that “common sense” supported the High 2 interpretation because Congress would not have wanted FOIA to “undermin[e] … the effectiveness of law enforcement agencies”). And we acknowledge that our decision today upsets three decades of agency practice relying on Crooker, and therefore may force considerable adjustments. We also note, however, that the Government has other tools at hand to shield national security information and other sensitive materials. Most notably, Exemption 1 of FOIA prevents access to classified documents. §552(b)(1); see 575 F. 3d, at 980 (W. Fletcher, J., dissenting) (Exemption 1 is “specifically designed to allow government agencies to withhold information that might jeopardize our national security”). The Government generally may classify material even after receiving a FOIA request, see Exec. Order No. 13526, §1.7(d), 75 Fed. Reg. 711 (2009); an agency therefore may wait until that time to decide whether the dangers of disclosure outweigh the costs of classification. See Tr. of Oral Arg. 29–30. Exemption 3 also may mitigate the Government’s security concerns. That provision applies to records that any other statute exempts from disclosure, §552(b)(3), thus offering Congress an established, streamlined method to authorize the withholding of specific records that FOIA would not otherwise protect. And Exemption 7, as already noted, protects “information compiled for law enforcement purposes” that meets one of six criteria, including if its release “could reasonably be expected to endanger the life or physical safety of any individual.” §552(b)(7)(F). The Navy argued below that the ESQD data and maps fall within Exemption 7(F), see n. 3, supra, and that claim remains open for the Ninth Circuit to address on remand. If these or other exemptions do not cover records whose release would threaten the Nation’s vital interests, the Government may of course seek relief from Congress. See Tr. of Oral Arg. 48. All we hold today is that Congress has not enacted the FOIA exemption the Government desires. We leave to Congress, as is appropriate, the question whether it should do so. VI Exemption 2, consistent with the plain meaning of the term “personnel rules and practices,” encompasses only records relating to issues of employee relations and human resources. The explosives maps and data requested here do not qualify for withholding under that exemption. We therefore reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Footnote 1 The court adopted the “predominantly internal” standard as a way of implementing the exemption’s requirement that materials “relat[e] solely to” an agency’s internal personnel rules and practices. The word “solely,” the court reasoned, “has to be given the construction, consonant with reasonableness, of ‘predominantly’ ” because otherwise “solely” would conflict with the expansive term “related.” 670 F. 2d, at 1056 (some internal quotation marks omitted). Footnote 2 Three other Courts of Appeals had previously taken a narrower view of Exemption 2’s scope, consistent with the interpretation adopted in Rose. See Cox v. Department of Justice, 576 F. 2d 1302, 1309–1310 (CA8 1978) (concluding that Exemption 2 covers only an agency’s internal “housekeeping matters” (internal quotation marks omitted)); Stokes v. Brennan, 476 F. 2d 699, 703 (CA5 1973) (holding that Exemption 2 “must not be read so broadly as to exempt” an Occupational Safety and Health Administration manual for training compliance officers); Hawkes v. IRS, 467 F. 2d 787, 797 (CA6 1972) (“[T]he internal practices and policies referred to in [Exemption 2] relate only to … employee-employer type concerns”). These Circuits have never revised their understandings of the exemption. See infra, at 13, n. 7. Footnote 3 The Navy also invoked Exemption 7(F), which applies to “records or information compiled for law enforcement purposes, but only to the extent that the production of such … records … could reasonably be expected to endanger the life or physical safety of any individual.” 5 U. S. C. §552(b)(7)(F). The courts below did not decide whether the Navy could withhold the ESQD data under that exemption. 575 F. 3d 959, 971, n. 8 (CA9 2009); No. CV–06–01301 (WD Wash., Oct. 30, 2007), App. to Pet. for Cert. 4, 25, 2007 WL 3228049, *8. Footnote 4 Government records also must satisfy the other requirements of Exemption 2 to be exempt from disclosure. Information must “relat[e] solely”—meaning, as usual, “exclusively or only,” Random House 1354 —to the agency’s “personnel rules and practices.” And the information must be “internal”; that is, the agency must typically keep the records to itself for its own use. See Webster’s 1180 (“internal” means “existing or situated within the limits … of something”). An agency’s human resources documents will often meet these conditions. Footnote 5 The dissent would reject this longstanding rule of construction in favor of an approach asking courts “to turn Congress’ public information objectives into workable agency practice.” Post, at 8–9 (opinion of Breyer, J.). But nothing in FOIA either explicitly or implicitly grants courts discretion to expand (or contract) an exemption on this basis. In enacting FOIA, Congress struck the balance it thought right—generally favoring disclosure, subject only to a handful of specified exemptions—and did so across the length and breadth of the Federal Government. See, e.g., John Doe Agency v. John Doe Corp., 493 U. S. 146, 152–153 (1989). The judicial role is to enforce that congressionally determined balance rather than, as the dissent suggests, post, at 4–6, to assess case by case, department by department, and task by task whether disclosure interferes with good government. Footnote 6 We are perplexed that the dissent takes seriously Crooker’s notion that the reports are “reconcilable.” Post, at 4. To strip the matter to its essentials, the House Report says: “Exemption 2 means A, but not B.” The Senate Report says: “Exemption 2 means B.” That is the very definition of “irreconcilable.” Footnote 7 The dissent’s view that “two of th[ese] Circuits [have] not adher[ed] to their early positions” is incorrect. Post, at 2. In Abraham & Rose, P.L.C. v. United States, cited by the dissent, the Sixth Circuit rejected the Government’s claim that Exemption 2 shielded records of federal tax lien filings. 138 F. 3d 1075, 1082 (1998). The court nowhere discussed the High 2 versus Low 2 question at issue here. Its only reference to Crooker concerned the part of that decision interpreting “solely” to mean “predominantly.” See 138 F.3d, at 1080; see also n. 1, supra. Subsequently, the Sixth Circuit once again held, in Rugiero v. Department of Justice, that Exemption 2 applies to “routine matters of merely internal significance.” 257 F. 3d 534, 549 (2001). In Sladek v. Bensinger, which the dissent also cites, the Fifth Circuit insisted that the Government disclose a Drug Enforcement Administration agent’s manual because it “is not the type of trivial rule, such as allocation of parking facilities, that is covered by Exemption 2.” 605 F. 2d 899, 902 (1979). In confirming this Low 2 interpretation of the statute, the court acknowledged that another Circuit had embraced the High 2 standard. The court, however, declined to consider this alternative interpretation because it would not have changed the case’s outcome. See ibid. Finally, the Eighth Circuit’s last word on Exemption 2 is clear, and the dissent does not say otherwise. The exemption, according to that most recent Eighth Circuit decision, applies “only [to an agency’s] housekeeping matters.” Cox, 576 F. 2d, at 1309–1310 (internal quotation marks omitted). The dissent is surely right to say, post at 2, that Crooker “has guided nearly every FOIA case decided over the last 30 years” in Circuits applying Crooker; but that statement does not hold in the Circuits using the Low 2 approach. Footnote 8 Notably, even those courts approving Crooker have disagreed about how to apply High 2. Fault lines include whether the risk of circumvention must be significant, see, e.g., Hidalgo v. FBI, 541 F. Supp. 2d 250, 253 (DC 2008); Pet. for Cert. 15–16; whether courts should consider the public interest in disclosure when calculating that risk, see, e.g., Department of Justice, Guide to the Freedom of Information Act, p. 185 (2009); and whether an agency must regulate the person or entity threatening circumvention; compare, e.g., 575 F. 3d, at 971, with, e.g., id., at 978 (W. Fletcher, J., dissenting). The disagreement is not surprising. Because High 2 is nowhere evident in the statute, courts lack the normal guideposts for ascertaining its coverage. Footnote 9 The dissent asserts that “30 years of experience” with a more expansive interpretation of the exemption suggests no “seriou[s] interfere[nce] with … FOIA’s informational objectives.” Post, at 6. But those objectives suffer any time an agency denies a FOIA request based on an improper interpretation of the statute. To give just one example, the U. S. Forest Service has wrongly invoked Exemption 2 on multiple occasions to withhold information about (of all things) bird nesting sites. See Audubon Society v. Forest Serv., 104 F. 3d 1201, 1203 (CA10 1997); Maricopa Audubon Soc. v. Forest Serv., 108 F. 3d 1082, 1084 (CA9 1997). And recent statistics raise a concern that federal agencies may too readily use Exemption 2 to refuse disclosure. According to amicus Public Citizen, “while reliance on exemptions overall rose 83% from 1998 to 2006, reliance on Exemption 2 rose 344% during that same time period.” Brief for Public Citizen et al. as Amici Curiae 24. In 2009 alone, federal departments cited Exemption 2 more than 72,000 times to prevent access to records. See Brief for Allied Daily Newspapers of Washington et al. as Amici Curiae 3. We do not doubt that many of these FOIA denials were appropriate. But we are unable to accept the dissent’s unsupported declaration that a sweeping construction of Exemption 2 has not interfered with Congress’s goal of broad disclosure.

562.US.134

2. Assuming, without deciding, that the Government’s challenged inquiries implicate a privacy interest of constitutional significance, that interest, whatever its scope, does not prevent the Government from asking reasonable questions of the sort included on SF–85 and Form 42 in an employment background investigation that is subject to the Privacy Act’s safeguards against public disclosure. Pp. 10–24. (a) The forms are reasonable in light of the Government interests at stake. Pp. 11–19. (1) Judicial review of the forms must take into account the context in which the Government’s challenged inquiries arise. When the Government acts in its capacity “as proprietor” and manager of its “internal operation,” Cafeteria & Restaurant Workers v. McElroy, 367 U. S. 886, 896, it has a much freer hand than when it regulates as to citizens generally. The questions respondents challenge are part of a standard background check of the sort used by millions of private employers. The Government has been conducting employment investigations since the Republic’s earliest days, and the President has had statutory authority to assess an applicant’s fitness for the civil service since 1871. Standard background investigations similar to those at issue became mandatory for federal civil-service candidates in 1953, and the investigations challenged here arose from a decision to extend that requirement to federal contract employees. This history shows that the Government has an interest in conducting basic background checks in order to ensure the security of its facilities and to employ a competent, reliable workforce to carry out the people’s business. The interest is not diminished by the fact that respondents are contract employees. There are no meaningful distinctions in the duties of NASA’s civil-service and contractor employees, especially at JPL, where contract employees do work that is critical to NASA’s mission and that is funded with a multibillion dollar taxpayer investment. Pp. 12–15. (2) The challenged questions on SF–85 and Form 42 are reasonable, employment-related inquiries that further the Government’s interests in managing its internal operations. SF–85’s “treatment or counseling” question is a followup question to a reasonable inquiry about illegal-drug use. In context, the drug-treatment inquiry is also a reasonable, employment-related inquiry. The Government, recognizing that illegal-drug use is both a criminal and medical issue, seeks to separate out those drug users who are taking steps to address and overcome their problems. Thus, it uses responses to the drug-treatment question as a mitigating factor in its contractor credentialing decisions. The Court rejects the argument that the Government has a constitutional burden to demonstrate that its employment background questions are “necessary” or the least restrictive means of furthering its interests. So exacting a standard runs directly contrary to Whalen. See 429 U. S., at 596–597. Pp. 16–18. (3) Like SF–85’s drug-treatment question, Form 42’s open-ended questions are reasonably aimed at identifying capable employees who will faithfully conduct the Government’s business. Asking an applicant’s designated references broad questions about job suitability is an appropriate tool for separating strong candidates from weak ones. The reasonableness of such questions is illustrated by their pervasiveness in the public and private sectors. Pp. 18–19. (b) In addition to being reasonable in light of the Government interests at stake, SF–85 and Form 42 are also subject to substantial protections against disclosure to the public. Whalen and Nixon recognized that a “statutory or regulatory duty to avoid unwarranted disclosures” generally allays privacy concerns created by government “accumulation” of “personal information” for “public purposes.” Whalen, supra, at 605. Respondents attack only the Government’s collection of information, and here, as in Whalen and Nixon, the information collected is shielded by statute from unwarranted disclosure. The Privacy Act—which allows the Government to maintain only those records “relevant and necessary to accomplish” a purpose authorized by law, 5 U. S. C. §552a(e)(1); requires written consent before the Government may disclose an individual’s records, §552a(b); and imposes criminal liability for willful violations of its nondisclosure obligations, §552a(i)(1)—“evidence[s] a proper concern” for individual privacy. Whalen, supra, at 605; Nixon, supra, at 458–459. Respondents’ claim that the statutory exceptions to the Privacy Act’s disclosure bar, see §§552a(b)(1)–(12), leave its protections too porous to supply a meaningful check against unwarranted disclosures. But that argument rests on an incorrect reading of Whalen, Nixon, and the Privacy Act. Pp. 19–23. 530 F. 3d 865, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Ginsburg, Breyer, and Sotomayor, JJ., joined. Scalia, J., filed an opinion concurring in the judgment, in which Thomas, J., joined. Thomas, J., filed an opinion concurring in the judgment. Kagan, J., took no part in the consideration or decision of the case.

In two cases decided more than 30 years ago, this Court referred broadly to a constitutional privacy “interest in avoiding disclosure of personal matters.” Whalen v. Roe, 429 U. S. 589, 599–600 (1977); Nixon v. Administrator of General Services, 433 U. S. 425, 457 (1977). Respondents in this case, federal contract employees at a Government laboratory, claim that two parts of a standard employment background investigation violate their rights under Whalen and Nixon. Respondents challenge a section of a form questionnaire that asks employees about treatment or counseling for recent illegal-drug use. They also object to certain open-ended questions on a form sent to employees’ designated references. We assume, without deciding, that the Constitution protects a privacy right of the sort mentioned in Whalen and Nixon. We hold, however, that the challenged portions of the Government’s background check do not violate this right in the present case. The Government’s interests as employer and proprietor in managing its internal operations, combined with the protections against public dissemination provided by the Privacy Act of 1974, 5 U. S. C. §552a, satisfy any “interest in avoiding disclosure” that may “arguably ha[ve] its roots in the Constitution.” Whalen, supra, at 599, 605. I A The National Aeronautics and Space Administration (NASA) is an independent federal agency charged with planning and conducting the Government’s “space activities.” Pub. L. 111–314, §3, 124 Stat. 3333, 51 U. S. C. §20112(a)(1). NASA’s workforce numbers in the tens of thousands of employees. While many of these workers are federal civil servants, a substantial majority are employed directly by Government contractors. Contract employees play an important role in NASA’s mission, and their duties are functionally equivalent to those performed by civil servants. One NASA facility, the Jet Propulsion Laboratory (JPL) in Pasadena, California, is staffed exclusively by contract employees. NASA owns JPL, but the California Institute of Technology (Cal Tech) operates the facility under a Government contract. JPL is the lead NASA center for deep-space robotics and communications. Most of this country’s unmanned space missions—from the Explorer 1 satellite in 1958 to the Mars Rovers of today—have been developed and run by JPL. JPL scientists contribute to NASA earth-observation and technology-development projects. Many JPL employees also engage in pure scientific research on topics like “the star formation history of the universe” and “the fundamental properties of quantum fluids.” App. 64–65, 68. Twenty-eight JPL employees are respondents here. Many of them have worked at the lab for decades, and none has ever been the subject of a Government background investigation. At the time when respondents were hired, background checks were standard only for federal civil servants. See Exec. Order No. 10450, 3 CFR 936 (1949–1953 Comp.). In some instances, individual contracts required background checks for the employees of federal contractors, but no blanket policy was in place. The Government has recently taken steps to eliminate this two-track approach to background investigations. In 2004, a recommendation by the 9/11 Commission prompted the President to order new, uniform identification standards for “[f]ederal employees,” including “contractor employees.” Homeland Security Presidential Directive/HSPD–12—Policy for a Common Identification Standard for Federal Employees and Contractors, Public Papers of the President, George W. Bush, Vol. 2, Aug. 27, p. 1765 (2007) (hereinafter HSPD–12), App. 127. The Department of Commerce implemented this directive by mandating that contract employees with long-term access to federal facilities complete a standard background check, typically the National Agency Check with Inquiries (NACI). National Inst. of Standards and Technology, Personal Identity Verification of Federal Employees & Contractors, pp. iii–vi, 1–8, 6 (FIPS PUB 201–1, Mar. 2006) (hereinafter FIPS PUB 201–1), App. 131–150, 144–145.[Footnote 1] An October 2007 deadline was set for completion of these investigations. Memorandum from Joshua B. Bolten, Director, OMB, to the Heads of all Departments and Agencies (Aug. 5, 2005), App. 112. In January 2007, NASA modified its contract with Cal Tech to reflect the new background-check requirement. JPL management informed employees that anyone failing to complete the NACI process by October 2007 would be denied access to JPL and would face termination by Cal Tech. B The NACI process has long been the standard background investigation for prospective civil servants. The process begins when the applicant or employee fills out a form questionnaire. Employees who work in “non-sensitive” positions (as all respondents here do) complete Standard Form 85 (SF–85). Office of Personnel Management (OPM), Standard Form 85, Questionnaire for Non-Sensitive Positions, App. 88–95.[Footnote 2] Most of the questions on SF–85 seek basic biographical information: name, address, prior residences, education, employment history, and personal and professional references. The form also asks about citizenship, selective-service registration, and military service. The last question asks whether the employee has “used, possessed, supplied, or manufactured illegal drugs” in the last year. Id., at 94. If the answer is yes, the employee must provide details, including information about “any treatment or counseling received.” Ibid. A “truthful response,” the form notes, cannot be used as evidence against the employee in a criminal proceeding. Ibid. The employee must certify that all responses on the form are true and must sign a release authorizing the Government to obtain personal information from schools, employers, and others during its investigation. Once a completed SF–85 is on file, the “agency check” and “inquiries” begin. 75 Fed. Reg. 5359 (2010). The Government runs the information provided by the employee through FBI and other federal-agency databases. It also sends out form questionnaires to the former employers, schools, landlords, and references listed on SF–85. The particular form at issue in this case—the Investigative Request for Personal Information, Form 42—goes to the employee’s former landlords and references. Ibid.[Footnote 3] Form 42 is a two-page document that takes about five minutes to complete. See ibid. It explains to the reference that “[y]our name has been provided by” a particular employee or applicant to help the Government determine that person’s “suitability for employment or a security clearance.” App. 96–97. After several preliminary questions about the extent of the reference’s associations with the employee, the form asks if the reference has “any reason to question” the employee’s “honesty or trustworthiness.” Id., at 97. It also asks if the reference knows of any “adverse information” concerning the employee’s “violations of the law,” “financial integrity,” “abuse of alcohol and/or drugs,” “mental or emotional stability,” “general behavior or conduct,” or “other matters.” Ibid. If “yes” is checked for any of these categories, the form calls for an explanation in the space below. That space is also available for providing “additional information” (“derogatory” or “favorable”) that may bear on “suitability for government employment or a security clearance.” Ibid. All responses to SF–85 and Form 42 are subject to the protections of the Privacy Act. The Act authorizes the Government to keep records pertaining to an individual only when they are “relevant and necessary” to an end “required to be accomplished” by law. 5 U. S. C. §552a(e)(1). Individuals are permitted to access their records and request amendments to them. §§552a(d)(1),(2). Subject to certain exceptions, the Government may not disclose records pertaining to an individual without that individual’s written consent. §552a(b). C About two months before the October 2007 deadline for completing the NACI, respondents brought this suit, claiming, as relevant here, that the background-check process violates a constitutional right to informational privacy. App. 82 (Complaint for Injunctive and Declaratory Relief).[Footnote 4] The District Court denied respondents’ motion for a preliminary injunction, but the Ninth Circuit granted an injunction pending appeal, 506 F. 3d 713 (2007), and later reversed the District Court’s order. The court held that portions of both SF–85 and Form 42 are likely unconstitutional and should be preliminarily enjoined. 512 F. 3d 1134, vacated and superseded, 530 F. 3d 865 (2008). Turning first to SF–85, the Court of Appeals noted respondents’ concession “that most of the questions” on the form are “unproblematic” and do not “implicate the constitutional right to informational privacy.” 530 F. 3d, at 878. But the court determined that the “group of questions concerning illegal drugs” required closer scrutiny. Ibid. Applying Circuit precedent, the court upheld SF–85’s inquiries into recent involvement with drugs as “necessary to further the government’s legitimate interest” in combating illegal-drug use. Id., at 879. The court went on to hold, however, that the portion of the form requiring disclosure of drug “treatment or counseling” furthered no legitimate interest and was thus likely to be held unconstitutional. Ibid. Form 42, in the Court of Appeals’ estimation, was even “more problematic.” Ibid. The form’s “open-ended and highly private” questions, the court concluded, were not “narrowly tailored” to meet the Government’s interests in verifying contractors’ identities and “ensuring the security of the JPL.” Id., at 881, 880. As a result, the court held, these “open-ended” questions, like the drug-treatment question on SF–85, likely violate respondents’ informational-privacy rights.[Footnote 5] Over the dissents of five judges, the Ninth Circuit denied rehearing en banc. 568 F. 3d 1028 (2009). We granted certiorari. 559 U. S. ___ (2010). II As noted, respondents contend that portions of SF–85 and Form 42 violate their “right to informational privacy.” Brief for Respondents 15. This Court considered a similar claim in Whalen, 429 U. S. 589, which concerned New York’s practice of collecting “the names and addresses of all persons” prescribed dangerous drugs with both “legitimate and illegitimate uses.” Id., at 591. In discussing that claim, the Court said that “[t]he cases sometimes characterized as protecting ‘privacy’ ” actually involved “at least two different kinds of interests”: one, an “interest in avoiding disclosure of personal matters”;[Footnote 6] the other, an interest in “making certain kinds of important decisions” free from government interference.[Footnote 7] The patients who brought suit in Whalen argued that New York’s statute “threaten[ed] to impair” both their “nondisclosure” interests and their interests in making healthcare decisions independently. Id., at 600. The Court, however, upheld the statute as a “reasonable exercise of New York’s broad police powers.” Id., at 598. Whalen acknowledged that the disclosure of “private information” to the State was an “unpleasant invasion of privacy,” id., at 602, but the Court pointed out that the New York statute contained “security provisions” that protected against “public disclosure” of patients’ information, id., at 600–601. This sort of “statutory or regulatory duty to avoid unwarranted disclosures” of “accumulated private data” was sufficient, in the Court’s view, to protect a privacy interest that “arguably ha[d] its roots in the Constitution.” Id., at 605–606. The Court thus concluded that the statute did not violate “any right or liberty protected by the Fourteenth Amendment.” Id., at 606. Four months later, the Court referred again to a constitutional “interest in avoiding disclosure.” Nixon, 433 U. S., at 457 (internal quotation marks omitted). Former President Nixon brought a challenge to the Presidential Recordings and Materials Preservation Act, 88 Stat. 1695, note following 44 U. S. C. §2111, a statute that required him to turn over his presidential papers and tape recordings for archival review and screening. 433 U. S., at 455–465. In a section of the opinion entitled “Privacy,” the Court addressed a combination of claims that the review required by this Act violated the former President’s “Fourth and Fifth Amendmen[t]” rights. Id., at 455, and n. 18, 458–459. The Court rejected those challenges after concluding that the Act at issue, like the statute in Whalen, contained protections against “undue dissemination of private materials.” 433 U. S., at 458. Indeed, the Court observed that the former President’s claim was “weaker” than the one “found wanting . . . in Whalen,” as the Government was required to return immediately all “purely private papers and recordings” identified by the archivists. Id., at 458–459. Citing Fourth Amendment precedent, the Court also stated that the public interest in preserving presidential papers outweighed any “legitimate expectation of privacy” that the former President may have enjoyed. Id., at 458 (citing Katz v. United States, 389 U. S. 347 (1967); Camara v. Municipal Court of City and County of San Francisco, 387 U. S. 523 (1967); and Terry v. Ohio, 392 U. S. 1 (1968)).[Footnote 8] The Court announced the decision in Nixon in the waning days of October Term 1976. Since then, the Court has said little else on the subject of an “individual interest in avoiding disclosure of personal matters.” Whalen, supra, at 599; Nixon, supra, at 457. A few opinions have mentioned the concept in passing and in other contexts. See Department of Justice v. Reporters Comm. for Freedom of Press, 489 U. S. 749, 762–763 (1989); New York v. Ferber, 458 U. S. 747, 759, n. 10 (1982). But no other decision has squarely addressed a constitutional right to informational privacy.[Footnote 9] III As was our approach in Whalen, we will assume for present purposes that the Government’s challenged inquiries implicate a privacy interest of constitutional significance. 429 U. S., at 599, 605.[Footnote 10] We hold, however, that, whatever the scope of this interest, it does not prevent the Government from asking reasonable questions of the sort included on SF–85 and Form 42 in an employment background investigation that is subject to the Privacy Act’s safeguards against public disclosure. A 1 As an initial matter, judicial review of the Government’s challenged inquiries must take into account the context in which they arise. When the Government asks respondents and their references to fill out SF–85 and Form 42, it does not exercise its sovereign power “to regulate or license.” Cafeteria & Restaurant Workers v. McElroy, 367 U. S. 886, 896 (1961). Rather, the Government conducts the challenged background checks in its capacity “as proprietor” and manager of its “internal operation.” Ibid. Time and again our cases have recognized that the Government has a much freer hand in dealing “with citizen employees than it does when it brings its sovereign power to bear on citizens at large.” Engquist v. Oregon Dept. of Agriculture, 553 U. S. 591, 598 (2008); Waters v. Churchill, 511 U. S. 661, 674 (1994) (plurality opinion). This distinction is grounded on the “common-sense realization” that if every “employment decision became a constitutional matter,” the Government could not function. See Connick v. Myers, 461 U. S. 138, 143 (1983); see also Bishop v. Wood, 426 U. S. 341, 350 (1976) (“The Due Process Clause . . . is not a guarantee against incorrect or ill-advised personnel decisions”). An assessment of the constitutionality of the challenged portions of SF–85 and Form 42 must account for this distinction. The questions challenged by respondents are part of a standard employment background check of the sort used by millions of private employers. See Brief for Consumer Data Indus. Assn. et al. as Amici Curiae 2 (hereinafter CDIA Brief) (“[M]ore than 88% of U. S. companies … perform background checks on their employees”). The Government itself has been conducting employment investigations since the earliest days of the Republic. L. White, The Federalists: A Study in Administrative History 262–263 (1948); see OPM, Biography of An Ideal: History of the Federal Civil Service 8 (2002) (noting that President Washington “set a high standard” for federal office and finalized appointments only after “investigating [candidates’] capabilities and reputations”). Since 1871, the President has enjoyed statutory authority to “ascertain the fitness of applicants” for the civil service “as to age, health, character, knowledge and ability for the employment sought,” Act of Mar. 3, 1871, Rev. Stat. §1753, as amended, 5 U. S. C. §3301(2), and that Act appears to have been regarded as a codification of established practice.[Footnote 11] Standard background investigations similar to those at issue here became mandatory for all candidates for the federal civil service in 1953. Exec. Order No. 10450, 3 CFR 936. And the particular investigations challenged in this case arose from a decision to extend that requirement to federal contract employees requiring long-term access to federal facilities. See HSPD–12, at 1765, App. 127; FIPS PUB 201–1, at iii–vi, 1–8, App. 131–150. As this long history suggests, the Government has an interest in conducting basic employment background checks. Reasonable investigations of applicants and employees aid the Government in ensuring the security of its facilities and in employing a competent, reliable workforce. See Engquist, supra, at 598–599. Courts must keep those interests in mind when asked to go line-by-line through the Government’s employment forms and to scrutinize the choice and wording of the questions they contain. Respondents argue that, because they are contract employees and not civil servants, the Government’s broad authority in managing its affairs should apply with diminished force. But the Government’s interest as “proprietor” in managing its operations, Cafeteria & Restaurant Workers, supra, at 896, does not turn on such formalities. See Board of Comm’rs, Wabaunsee Cty. v. Umbehr, 518 U. S. 668, 678, 679 (1996) (formal distinctions such as whether a “service provider” has a “contract of employment or a contract for services” with the government is a “very poor proxy” for constitutional interests at stake). The fact that respondents’ direct employment relationship is with Cal Tech—which operates JPL under a Government contract—says very little about the interests at stake in this case. The record shows that, as a “practical matter,” there are no “[r]elevant distinctions” between the duties performed by NASA’s civil-service workforce and its contractor workforce. App. 221. The two classes of employees perform “functionally equivalent duties,” and the extent of employees’ “access to NASA … facilities” turns not on formal status but on the nature of “the jobs they perform.” Ibid. At JPL, in particular, the work that contract employees perform is critical to NASA’s mission. Respondents in this case include “the lead trouble-shooter for … th[e] $568 [million]” Kepler space observatory, 7 Record 396; the leader of the program that “tests … all new technology that NASA will use in space,” App. 60; and one of the lead “trajectory designers for … the Galileo Project and the Apollo Moon landings,” id., at 62. This is important work, and all of it is funded with a multibillion dollar investment from the American taxpayer. See NASA, Jet Propulsion Laboratory Annual Report 09, p. 35 (2010), online at http://www.jpl.nasa.gov/annualreport/2009-report.pdf. The Government has a strong interest in conducting basic background checks into the contract employees minding the store at JPL.[Footnote 12] 2 With these interests in view, we conclude that the challenged portions of both SF–85 and Form 42 consist of reasonable, employment-related inquiries that further the Government’s interests in managing its internal operations. See Engquist, 553 U. S., at 598–599; Whalen, 429 U. S., at 597–598. As to SF–85, the only part of the form challenged here is its request for information about “any treatment or counseling received” for illegal-drug use within the previous year. The “treatment or counseling” question, however, must be considered in context. It is a followup to SF–85’s inquiry into whether the employee has “used, possessed, supplied, or manufactured illegal drugs” during the past year. The Government has good reason to ask employees about their recent illegal-drug use. Like any employer, the Government is entitled to have its projects staffed by reliable, law-abiding persons who will “ ‘efficiently and effectively’ ” discharge their duties. See Engquist, supra, at 598–599. Questions about illegal-drug use are a useful way of figuring out which persons have these characteristics. See, e.g., Breen & Matusitz, An Updated Examination of the Effects of Illegal Drug Use in the Workplace, 19 J. Human Behavior in the Social Environment, 434 (2009) (illicit drug use negatively correlated with workplace productivity). In context, the follow-up question on “treatment or counseling” for recent illegal-drug use is also a reasonable, employment-related inquiry. The Government, recognizing that illegal-drug use is both a criminal and a medical issue, seeks to separate out those illegal-drug users who are taking steps to address and overcome their problems. The Government thus uses responses to the “treatment or counseling” question as a mitigating factor in determining whether to grant contract employees long-term access to federal facilities.[Footnote 13] This is a reasonable, and indeed a humane, approach, and respondents do not dispute the legitimacy of the Government’s decision to use drug treatment as a mitigating factor in its contractor credentialing decisions. Respondents’ argument is that, if drug treatment is only used to mitigate, then the Government should change the mandatory phrasing of SF–85—“Include [in your answer] any treatment or counseling received”—so as to make a response optional. App. 94. As it stands, the mandatory “treatment or counseling” question is unconstitutional, in respondents’ view, because it is “more intrusive than necessary to satisfy the government’s objective.” Brief for Respondents 26; 530 F. 3d, at 879 (holding that “treatment or counseling” question should be enjoined because the form “appears to compel disclosure”). We reject the argument that the Government, when it requests job-related personal information in an employment background check, has a constitutional burden to demonstrate that its questions are “necessary” or the least restrictive means of furthering its interests. So exacting a standard runs directly contrary to Whalen. The patients in Whalen, much like respondents here, argued that New York’s statute was unconstitutional because the State could not “demonstrate the necessity” of its program. 429 U. S., at 596. The Court quickly rejected that argument, concluding that New York’s collection of patients’ prescription information could “not be held unconstitutional simply because” a court viewed it as “unnecessary, in whole or in part.” Id., at 596–597. That analysis applies with even greater force where the Government acts, not as a regulator, but as the manager of its internal affairs. See Engquist, supra, at 598–599. SF–85’s “treatment or counseling” question reasonably seeks to identify a subset of acknowledged drug users who are attempting to overcome their problems. The Government’s considered position is that phrasing the question in more permissive terms would result in a lower response rate, and the question’s effectiveness in identifying illegal-drug users who are suitable for employment would be “materially reduced.” Reply Brief for Petitioners 19. That is a reasonable position, falling within the “ ‘wide latitude’ ” granted the Government in its dealings with employees. See Engquist, supra, at 600. 3 The Court of Appeals also held that the broad, “open-ended questions” on Form 42 likely violate respondents’ informational-privacy rights. Form 42 asks applicants’ designated references and landlords for “information” bearing on “suitability for government employment or a security clearance.” App. 97. In a series of questions, the Government asks if the reference has any “adverse information” about the applicant’s “honesty or trustworthiness,” “violations of the law,” “financial integrity,” “abuse of alcohol and/or drugs,” “mental or emotional stability,” “general behavior or conduct,” or “other matters.” Ibid. These open-ended inquiries, like the drug-treatment question on SF–85, are reasonably aimed at identifying capable employees who will faithfully conduct the Government’s business. See Engquist, supra, at 598–599. Asking an applicant’s designated references broad, open-ended questions about job suitability is an appropriate tool for separating strong candidates from weak ones. It would be a truly daunting task to catalog all the reasons why a person might not be suitable for a particular job, and references do not have all day to answer a laundry list of specific questions. See CDIA Brief 6–7 (references “typically have limited time to answer questions from potential employers,” and “open-ended questions” yield more relevant information than narrow inquiries). Form 42, by contrast, takes just five minutes to complete. 75 Fed. Reg. 5359. The reasonableness of such open-ended questions is illustrated by their pervasiveness in the public and private sectors. Form 42 alone is sent out by the Government over 1.8 million times annually. Ibid. In addition, the use of open-ended questions in employment background checks appears to be equally commonplace in the private sector. See, e.g., S. Bock et al., Mandated Benefits 2008 Compliance Guide, Exh. 20.1, A Sample Policy on Reference Checks on Job Applicants (“Following are the guidelines for conducting a telephone reference check: … Ask open-ended questions, then wait for the respondent to answer”); M. Zweig, Human Resources Management 87 (1991) (“Also ask, ‘Is there anything else I need to know about [candidate’s name]?’ This kind of open-ended question may turn up all kinds of information you wouldn’t have gotten any other way”). The use of similar open-ended questions by the Government is reasonable and furthers its interests in managing its operations. B 1 Not only are SF–85 and Form 42 reasonable in light of the Government interests at stake, they are also subject to substantial protections against disclosure to the public. Both Whalen and Nixon recognized that government “accumulation” of “personal information” for “public purposes” may pose a threat to privacy. Whalen, 429 U. S., at 605; see Nixon 433 U. S., at 457–458, 462. But both decisions also stated that a “statutory or regulatory duty to avoid unwarranted disclosures” generally allays these privacy concerns. Whalen, supra, at 605; Nixon, supra, at 458–459. The Court in Whalen, relying on New York’s “security provisions” prohibiting public disclosure, turned aside a challenge to the collection of patients’ prescription information. 429 U. S., at 594, and n. 12, 600–601, 605. In Nixon, the Court rejected what it regarded as an even “weaker” claim by the former President because the Presidential Recordings and Materials Preservation Act “[n]ot only . . . mandate[d] regulations” against “undue dissemination,” but also required immediate return of any “purely private” materials flagged by the Government’s archivists. 433 U. S., at 458–459. Respondents in this case, like the patients in Whalen and former President Nixon, attack only the Government’s collection of information on SF–85 and Form 42. And here, no less than in Whalen and Nixon, the information collected is shielded by statute from “unwarranted disclosur[e].” See Whalen, supra, at 605. The Privacy Act, which covers all information collected during the background-check process, allows the Government to maintain records “about an individual” only to the extent the records are “relevant and necessary to accomplish” a purpose authorized by law. 5 U. S. C. §552a(e)(1). The Act requires written consent before the Government may disclose records pertaining to any individual. §552a(b). And the Act imposes criminal liability for willful violations of its nondisclosure obligations. §552a(i)(1). These requirements, as we have noted, give “forceful recognition” to a Government employee’s interest in maintaining the “confidentiality of sensitive information . . . in his personnel files.” Detroit Edison Co. v. NLRB, 440 U. S. 301, 318, n. 16 (1979). Like the protections against disclosure in Whalen and Nixon, they “evidence a proper concern” for individual privacy. Whalen, supra, at 605; Nixon, supra, at 458–459. 2 Notwithstanding these safeguards, respondents argue that statutory exceptions to the Privacy Act’s disclosure bar, see §§552a(b)(1)–(12), leave its protections too porous to supply a meaningful check against “unwarranted disclosures,” Whalen, supra, at 605. Respondents point in particular to what they describe as a “broad” exception for “routine use[s],” defined as uses that are “compatible with the purpose for which the record was collected.” §§552a(b)(3), (a)(7). Respondents’ reliance on these exceptions rests on an incorrect reading of both our precedents and the terms of the Privacy Act. As to our cases, the Court in Whalen and Nixon referred approvingly to statutory or regulatory protections against “unwarranted disclosures” and “undue dissemination” of personal information collected by the Government. Whalen, supra, at 605; Nixon, supra, at 458. Neither case suggested that an ironclad disclosure bar is needed to satisfy privacy interests that may be “root[ed] in the Constitution.” Whalen, supra, at 605. In Whalen, the New York statute prohibiting “[p]ublic disclosure of the identity of patients” was itself subject to several exceptions. 429 U. S., at 594–595, and n. 12. In Nixon, the protections against “undue dissemination” mentioned in the opinion were not even before the Court, but were to be included in forthcoming regulations “mandate[d]” by the challenged Act. 433 U. S., at 458; see id., at 437–439 (explaining that the Court was limiting its review to the Act’s “facial validity” and was not considering the Administrator’s forthcoming regulations). Thus, the mere fact that the Privacy Act’s nondisclosure requirement is subject to exceptions does not show that the statute provides insufficient protection against public disclosure. Nor does the substance of the “routine use” exception relied on by respondents create any undue risk of public dissemination. None of the authorized “routine use[s]” of respondents’ background-check information allows for release to the public. 71 Fed. Reg. 45859–45860, 45862 (2006); 60 Fed. Reg. 63084 (1995), as amended, 75 Fed. Reg. 28307 (2010). Rather, the established “routine use[s]” consist of limited, reasonable steps designed to complete the background-check process in an efficient and orderly manner. See Whalen, supra, at 602 (approving disclosures to authorized New York Department of Health employees that were not “meaningfully distinguishable” from routine disclosures “associated with many facets of health care”). One routine use, for example, involves a limited disclosure to persons filling out Form 42 so that designated references can “identify the individual” at issue and can understand the “nature and purpose of the investigation.” App. 89. Authorized JPL employees also review each completed SF–85 to verify that all requested information has been provided. Id., at 211. These designated JPL employees may not “disclose any information contained in the form to anyone else,” ibid., and Cal Tech is not given access to adverse information uncovered during the Government’s background check, id., at 207–208. The “remote possibility” of public disclosure created by these narrow “routine use[s]” does not undermine the Privacy Act’s substantial protections. See Whalen, 429 U. S., at 601–602 (“remote possibility” that statutory security provisions will “provide inadequate protection against unwarranted disclosures” not a sufficient basis for striking down statute). Citing past violations of the Privacy Act,[Footnote 14] respondents note that it is possible that their personal information could be disclosed as a result of a similar breach. But data breaches are a possibility any time the Government stores information. As the Court recognized in Whalen, the mere possibility that security measures will fail provides no “proper ground” for a broad-based attack on government information-collection practices. Ibid. Respondents also cite a portion of SF–85 that warns of possible disclosure “[t]o the news media or the general public.” App. 89. By its terms, this exception allows public disclosure only where release is “in the public interest” and would not result in “an unwarranted invasion of personal privacy.” Ibid. Respondents have not cited any example of such a disclosure, nor have they identified any plausible scenario in which their information might be unduly disclosed under this exception.[Footnote 15] In light of the protection provided by the Privacy Act’s nondisclosure requirement, and because the challenged portions of the forms consist of reasonable inquiries in an employment background check, we conclude that the Government’s inquiries do not violate a constitutional right to informational privacy. Whalen, supra, at 605. * * * For these reasons, the judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 As alternatives to the NACI process, the Department of Commerce also authorized federal agencies to use another “Office of Personnel Management … or National Security community investigation required for Federal employment.” App. 145. None of these alternative background checks are at issue here. Footnote 2 For public-trust and national-security positions, more detailed forms are required. See OPM, Standard Form 85P, Questionnaire for Public Trust Positions, online at http://www.opm.gov/Forms/pdf_fill/sf85p.pdf; (all Internet materials as visited Jan. 13, 2011, and available in Clerk of Court’s case file); OPM, Standard Form 86, Questionnaire for National Security Positions, online at http://www.opm.gov/Forms/ pdf_fill/sf86.pdf. Footnote 3 The Government sends separate forms to employers (Form 41), educational institutions (Form 43), record repositories (Form 40), and law enforcement agencies (Form 44). 75 Fed. Reg. 5359. None of these forms are at issue here. Footnote 4 Respondents sought to represent a class of “JPL employees in non-sensitive positions.” App. 79. No class has been certified. Footnote 5 In the Ninth Circuit, respondents also challenged the criteria that they believe the Government will use to determine their “suitability” for employment at JPL. Respondents relied on a document, which had been temporarily posted on the JPL intranet, that listed factors purportedly bearing on suitability for federal employment. App. 98–104. Among the listed factors were a failure to “mee[t] financial obligations,” “health issues,” and “mental, emotional, psychological, or psychiatric issues.” Id., at 98, 102. Other factors, which were listed under the heading “Criminal or Immoral Conduct,” included “indecent exposure,” “voyeurism,” “indecent proposal[s],” and “carnal knowledge.” Id., at 98. The document also stated that while “homosexuality,” “adultery,” and “illegitimate children” were not “suitability” issues in and of themselves, they might pose “security issue[s]” if circumstances indicated a “susceptibility to coercion or blackmail.” Id., at 102. The Court of Appeals rejected respondents’ “challenges to . . . suitability determination[s]” as unripe. 530 F. 3d, at 873. Although respondents did not file a cross-petition from that portion of the Ninth Circuit’s judgment, they nonetheless discuss these suitability criteria at some length in their brief before this Court. Respondents’ challenge to these criteria is not before us. We note, however, the Acting Solicitor General’s statement at oral argument that “NASA will not and does not use” the document to which respondents object “to make contractor credentialing decisions.” Tr. of Oral Arg. 22. Footnote 6 429 U. S., at 598–599, and n. 25 (citing Olmstead v. United States, 277 U. S. 438, 478 (1928) (Brandeis, J., dissenting) (describing “the right to be let alone” as “the right most valued by civilized men”); Griswold v. Connecticut, 381 U. S. 479, 483 (1965) (“[T]he First Amendment has a penumbra where privacy is protected from governmental intrusion”); Stanley v. Georgia, 394 U. S. 557, 559, 568 (1969); California Bankers Assn. v. Shultz, 416 U. S. 21, 79 (1974) (Douglas, J., dissenting); and id., at 78 (Powell, J., concurring)). Footnote 7 429 U. S., at 599–600, and n. 26 (citing Roe v. Wade, 410 U. S. 113 (1973); Doe v. Bolton, 410 U. S. 179 (1973); Loving v. Virginia, 388 U. S. 1 (1967); Griswold v. Connecticut, supra; Pierce v. Society of Sisters, 268 U. S. 510 (1925); Meyer v. Nebraska, 262 U. S. 390 (1923); and Allgeyer v. Louisiana, 165 U. S. 587 (1897)). Footnote 8 The Court continued its discussion of Fourth Amendment principles throughout the “Privacy” section of the opinion. See 433 U. S., at 459 (citing United States v. Miller, 425 U. S. 435 (1976), United States v. Dionisio, 410 U. S. 1 (1973), and Katz, 389 U. S. 347)); 433 U. S., at 460–462 (addressing the former President’s claim that the Act was “tantamount to a general warrant” under Stanford v. Texas, 379 U. S. 476 (1965)); 433 U. S., at 463–465, and n. 26 (concluding that the challenged law was analogous to the wiretapping provisions of Title III of the Omnibus Crime Control and Safe Streets Act of 1968, notwithstanding the lack of a “warrant requirement”). Footnote 9 State and lower federal courts have offered a number of different interpretations of Whalen and Nixon over the years. Many courts hold that disclosure of at least some kinds of personal information should be subject to a test that balances the government’s interests against the individual’s interest in avoiding disclosure. E.g., Barry v. New York, 712 F. 2d 1554, 1559 (CA2 1983); Fraternal Order of Police v. Philadelphia, 812 F. 2d 105, 110 (CA3 1987); Woodland v. Houston, 940 F. 2d 134, 138 (CA5 1991) (per curiam); In re Crawford, 194 F. 3d 954, 959 (CA9 1999); State v. Russo, 259 Conn. 436, 459–464, 790 A. 2d 1132, 1147–1150 (2002). The Sixth Circuit has held that the right to informational privacy protects only intrusions upon interests “that can be deemed fundamental or implicit in the concept of ordered liberty.” J. P. v. DeSanti, 653 F. 2d 1080, 1090 (1981) (internal quotation marks omitted). The D. C. Circuit has expressed “grave doubts” about the existence of a constitutional right to informational privacy. American Federation of Govt. Employees v. HUD, 118 F. 3d 786, 791 (1997). Footnote 10 The opinions concurring in the judgment disagree with this approach and would instead provide a definitive answer to the question whether there is a constitutional right to informational privacy. Post, at 6–7 (opinion of Scalia, J.); post, at 1 (opinion of Thomas, J.). One of these opinions expresses concern that our failure to do so will “har[m] our image, if not our self-respect,” post, at 7 (Scalia, J.), and will cause practical problems, post, at 8–9. There are sound reasons for eschewing the concurring opinions’ recommended course. “The premise of our adversarial system is that appellate courts do not sit as self-directed boards of legal inquiry and research, but essentially as arbiters of legal questions presented and argued by the parties before them.” Carducci v. Regan, 714 F. 2d 171, 177 (CADC 1983) (opinion for the court by Scalia, J.). In this case, petitioners did not ask us to hold that there is no constitutional right to informational privacy, and respondents and their amici thus understandably refrained from addressing that issue in detail. It is undesirable for us to decide a matter of this importance in a case in which we do not have the benefit of briefing by the parties and in which potential amici had little notice that the matter might be decided. See Pet. for Cert. 15 (“no need in this case” for broad decision on “the scope of a constitutionally-based right to privacy for certain information”). Particularly in cases like this one, where we have only the “scarce and open-ended” guideposts of substantive due process to show us the way, see Collins v. Harker Heights, 503 U. S. 115, 125 (1992), the Court has repeatedly recognized the benefits of proceeding with caution. E.g., Herrera v. Collins, 506 U. S. 390, 417 (1993) (joined by Scalia, J.) (assuming “for the sake of argument … that in a capital case a truly persuasive demonstration of ‘actual innocence’ ” made after conviction would render execution unconstitutional); Cruzan v. Director, Mo. Dept. of Health, 497 U. S. 261, 279 (1990) (joined by Scalia, J.) (“[W]e assume that the United States Constitution would grant a competent person a constitutionally protected right to refuse lifesaving hydration and nutrition”); Regents of Univ. of Mich. v. Ewing, 474 U. S. 214, 222–223 (1985) (“assum[ing], without deciding, that federal courts can review an academic decision of a public educational institution under a substantive due process standard”); Board of Curators of Univ. of Mo. v. Horowitz, 435 U. S. 78, 91–92 (1978) (same); see also New York State Club Assn., Inc. v. City of New York, 487 U. S. 1, 20 (1988) (Scalia, J., concurring in part and concurring in judgment) (joining the Court’s opinion on the understanding that it “assumes for purposes of its analysis, but does not hold, the existence of a constitutional right of private association for other than expressive or religious purposes”). Justice Scalia provides no support for his claim that our approach in this case will “dramatically increase the number of lawsuits claiming violations of the right to informational privacy,” post, at 9, and will leave the lower courts at sea. We take the same approach here that the Court took more than three decades ago in Whalen and Nixon, and there is no evidence that those decisions have caused the sky to fall. We therefore decide the case before us and leave broader issues for another day. Footnote 11 The debate on the 1871 Act in the House of Representatives contained this exchange on presidential authority to conduct background checks: “Mr. Peters: Has he not that power [to conduct the proposed investigations of candidates for the civil service] now? “Mr. Dawes: He has all that power. If you will go up to the War Department or the Department of the Interior you will see pretty much all of this nailed up on the doors, in the form of rules and regulations.” Cong. Globe, 41st Cong., 3d Sess., 1935 (1871). Footnote 12 In their brief, respondents also rely on the fact that many of them have been working at JPL for years and that Cal Tech previously vetted them through standard “employment reference checks.” Brief for Respondents 52–53. The record indicates that this may be wrong as a factual matter. E.g., 7 Record 391 (“I have not been required to undergo any type of background investigation to maintain my position with JPL”); id., at 397 (“I have never been required to undergo any type of background investigation to maintain my position with JPL other than … [one] which required that I provide my name, social security number, and current address” to facilitate a “check for outstanding warrants, arrests, or convictions”); id., at 356, 367, 386–387 (similar). Even if it were correct, the fact that Cal Tech once conducted a background check on respondents does not diminish the Government’s interests in conducting its own standard background check to satisfy itself that contract employees should be granted continued access to the Government’s facility. In any event, counsel abandoned this position at oral argument. Tr. of Oral Arg. 38. Footnote 13 Asking about treatment or counseling could also help the Government identify chronic drug abusers for whom, “despite counseling and rehabilitation programs, there is little chance for effective rehabilitation.” 38 Fed. Reg. 33315 (1973). At oral argument, however, the Acting Solicitor General explained that NASA views treatment or counseling solely as a “mitigat[ing]” factor that ameliorates concerns about recent illegal drug use. Tr. of Oral Arg. 19. Footnote 14 E.g., GAO, Personal Information: Data Breaches are Frequent, but Evidence of Resulting Identity Theft is Limited; However, the Full Extent Is Unknown 5, 20 (GAO 07–737, 2007) (over 3-year period, 788 data breaches occurred at 17 federal agencies). Footnote 15 Respondents further contend that the Privacy Act’s ability to deter unauthorized release of private information is significantly hampered by the fact that the statute provides only “an ex post money-damages action,” not injunctive relief. Brief for Respondents 44 (citing Doe v. Chao, 540 U. S. 614, 635 (2004) (Ginsburg, J., dissenting)). Nothing in Whalen or Nixon suggests that any private right of action—for money damages or injunctive relief—is needed in order to provide sufficient protection against public disclosure.

564.US.2010_10-568

Nevada’s Ethics in Government Law requires public officials to recuse themselves from voting on, or advocating the passage or failure of, “a matter with respect to which the independence of judgment of a reasonable person in his situation would be materially affected by,” inter alia, “[h]is commitment in a private capacity to the interests of others,” Nev. Rev. Stat. §281A.420(2) (2007), which includes a “commitment to a [specified] person,” e.g., a member of the officer’s household or the officer’s relative, §281A.420(8)(a)–(d), and “[a]ny other commitment or relationship that is substantially similar” to one enumerated in paragraphs (a)–(d), §281A.420(8)(e). Petitioner (Commission) administers and enforces Nevada’s law. The Commission investigated respondent Carrigan, an elected local official who voted to approve a hotel/casino project proposed by a company that used Carrigan’s long-time friend and campaign manager as a paid consultant. The Commission concluded that Carrigan had a disqualifying conflict of interest under §281A.420(8)(e)’s catchall provision, and censured him for failing to abstain from voting on the project. Carrigan sought judicial review, arguing that the Nevada law violated the First Amendment. The State District Court denied the petition, but the Nevada Supreme Court reversed, holding that voting is protected speech and that §281A.420(8)(e)’s catchall definition is unconstitutionally overbroad. Held: The Nevada Ethics in Government Law is not unconstitutionally overbroad. Pp. 3–11. (a) That law prohibits a legislator who has a conflict both from voting on a proposal and from advocating its passage or failure. If it was constitutional to exclude Carrigan from voting, then his exclusion from advocating during a legislative session was not unconstitutional, for it was a reasonable time, place, and manner limitation. See Clark v. Community for Creative Non-Violence, 468 U. S. 288, 293. Pp. 3–4. (b) “[A] ‘universal and long-established’ tradition of prohibiting certain conduct creates ‘a strong presumption’ that the prohibition is constitutional.’ ” Republican Party of Minn. v. White, 536 U. S. 765, 785. Here, dispositive evidence is provided by “early congressional enactments,” which offer “ ‘contemporaneous and weighty evidence of the Constitution’s meaning,’ ” Printz v. United States, 521 U. S. 898, 905. Within 15 years of the founding, both the House and the Senate adopted recusal rules. Federal conflict-of-interest rules applicable to judges also date back to the founding. The notion that Nevada’s recusal rules violate legislators’ First Amendment rights is also inconsistent with long-standing traditions in the States, most of which have some type of recusal law. Pp. 4–8. (c) Restrictions on legislators’ voting are not restrictions on legislators’ protected speech. A legislator’s vote is the commitment of his apportioned share of the legislature’s power to the passage or defeat of a particular proposal. He casts his vote “as trustee for his constituents, not as a prerogative of personal power.” Raines v. Byrd, 521 U. S. 811, 821. Moreover, voting is not a symbolic action, and the fact that it is the product of a deeply held or highly unpopular personal belief does not transform it into First Amendment speech. Even if the mere vote itself could express depth of belief (which it cannot), this Court has rejected the notion that the First Amendment confers a right to use governmental mechanics to convey a message. See, e.g., Timmons v. Twin Cities Area New Party, 520 U. S. 351. Doe v. Reed, 561 U. S. ___, distinguished. Pp. 8–10. (d) The additional arguments raised in Carrigan’s brief were not decided below or raised in his brief in opposition and are thus considered waived. P. 11. 126 Nev. 28, 236 P. 3d 616, reversed and remanded. Scalia, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined. Kennedy, J., filed a concurring opinion. Alito, J., filed an opinion concurring in part and concurring in the judgment.

The Nevada Supreme Court invalidated a recusal provision of the State’s Ethics in Government Law as unconstitutionally overbroad in violation of the First Amendment. We consider whether legislators have a personal, First Amendment right to vote on any given matter. I Nevada’s Ethics in Government Law provides that “a public officer shall not vote upon or advocate the passage or failure of, but may otherwise participate in the con-sideration of, a matter with respect to which the independence of judgment of a reasonable person in his situation would be materially affected by,” inter alia, “[h]is commitment in a private capacity to the interests of others.” Nev. Rev. Stat. §281A.420(2) (2007).[Footnote 1] Section 281A.420(8)(a)–(d) of the law defines the term “commitment in a private capacity to the interests of others” to mean a “commitment to a person” who is a member of the officer’s household; is related by blood, adoption, or marriage to the officer; employs the officer or a member of his household; or has a substantial and continuing business relationship with the officer. Paragraph (e) of the same subsection adds a catchall to that definition: “[a]ny other commitment or relationship that is substantially similar” to one of those listed in paragraphs (a)–(d). The Ethics in Government Law is administered and enforced by the petitioner in this litigation, the Nevada Commission on Ethics. In 2005, the Commission initiated an investigation of Michael Carrigan, an elected member of the City Council of Sparks, Nevada, in response to complaints that Carrigan had violated §281A.420(2) by voting to approve an application for a hotel/casino project known as the “Lazy 8.” Carrigan, the complaints asserted, had a disabling conflict in the matter because his long-time friend and campaign manager, Carlos Vasquez, worked as a paid consultant for the Red Hawk Land Company, which had proposed the Lazy 8 project and would benefit from its approval. Upon completion of its investigation, the Commission concluded that Carrigan had a disqualifying conflict of interest under §281A.420(8)(e)’s catchall provision because his relationship with Vasquez was “substantially similar” to the prohibited relationships listed in §281A.420(8)(a)–(d). Its written decision censured Carrigan for failing to abstain from voting on the Lazy 8 matter, but did not impose a civil penalty because his violation was not willful, see §281A.480. (Before the hearing, Carrigan had consulted the Sparks city attorney, who advised him that disclosing his relationship with Vasquez before voting on the Lazy 8 project, which he did, would satisfy his obligations under the Ethics in Government Law.) Carrigan filed a petition for judicial review in the First Judicial District Court of the State of Nevada, arguing that the provisions of the Ethics in Government Law that he was found to have violated were unconstitutional under the First Amendment. The District Court denied the petition, but a divided Nevada Supreme Court reversed. The majority held that voting was protected by the First Amendment, and, applying strict scrutiny, found that §281A.420(8)(e)’s catchall definition was unconstitutionally overbroad. 126 Nev. 28, ___–___, 236 P. 3d 616, 621–624 (2010). We granted certiorari, 562 U. S. ___ (2011). II The First Amendment prohibits laws “abridging the freedom of speech,” which, “ ‘as a general matter … means that government has no power to restrict expression because of its message, its ideas, its subject matter, or its content.’ ” Ashcroft v. American Civil Liberties Union, 535 U. S. 564, 573 (2002) (quoting Bolger v. Youngs Drug Products Corp., 463 U. S. 60, 65 (1983)). But the Amendment has no application when what is restricted is not protected speech. See, e.g., Roth v. United States, 354 U. S. 476, 483 (1957) (obscenity not protected speech). The Nevada Supreme Court thought a legislator’s vote to be protected speech because voting “is a core legislative function.” 126 Nev., at ___, 236 P. 3d, at 621 (internal quotation marks omitted). We disagree, for the same reason. But before discussing that issue, we must address a preliminary detail: The challenged law not only prohibits the legislator who has a conflict from voting on the proposal in question, but also forbids him to “advocate the passage or failure” of the proposal—evidently meaning advocating its passage or failure during the legislative debate. Neither Carrigan nor any of his amici contend that the prohibition on advocating can be unconstitutional if the prohibition on voting is not. And with good reason. Legislative sessions would become massive town-hall meetings if those who had a right to speak were not limited to those who had a right to vote. If Carrigan was constitutionally excluded from voting, his exclusion from “advocat[ing]” at the legislative session was a reasonable time, place and manner limitation. See Clark v. Community for Creative Non-Violence, 468 U. S. 288, 293 (1984). III “[A] universal and long-established tradition of prohibiting certain conduct creates a strong presumption that the prohibition is constitutional: Principles of liberty fundamental enough to have been embodied within constitutional guarantees are not readily erased from the Nation’s consciousness.” Republican Party of Minn. v. White, 536 U. S. 765, 785 (2002) (internal quotation marks omitted). Laws punishing libel and obscenity are not thought to violate “the freedom of speech” to which the First Amendment refers because such laws existed in 1791 and have been in place ever since. The same is true of legislative recusal rules. The Nevada Supreme Court and Carrigan have not cited a single decision invalidating a generally applicable conflict-of-interest recusal rule—and such rules have been commonplace for over 200 years. “[E]arly congressional enactments ‘provid[e] contemporaneous and weighty evidence of the Constitution’s meaning,’ ” Printz v. United States, 521 U. S. 898, 905 (1997) (quoting Bowsher v. Synar, 478 U. S. 714, 723–724 (1986)). That evidence is dispositive here. Within 15 years of the founding, both the House of Representatives and the Senate adopted recusal rules. The House rule—to which no one is recorded as having objected, on constitutional or other grounds, see D. Currie, The Constitution in Congress: The Federalist Period 1789–1801, p. 10 (1997)—was adopted within a week of that chamber’s first achieving a quorum.[Footnote 2] The rule read: “No member shall vote on any question, in the event of which he is immediately and particularly interested.” 1 Annals of Cong. 99 (1789). Members of the House would have been subject to this recusal rule when they voted to submit the First Amendment for ratification; their failure to note any inconsistency between the two suggests that there was none. The first Senate rules did not include a recusal requirement, but Thomas Jefferson adopted one when he was President of the Senate. His rule provided as follows: “Where the private interests of a member are concerned in a bill or question, he is to withdraw. And where such an interest has appeared, his voice [is] disallowed, even after a division. In a case so contrary not only to the laws of decency, but to the fundamental principles of the social compact, which denies to any man to be a judge in his own case, it is for the honor of the house that this rule, of immemorial observance, should be strictly adhered to.” A Manual of Parliamentary Practice for the Use of the Senate of the United States 31 (1801). Contemporaneous treatises on parliamentary procedure track parts of Jefferson’s formulation. See, e.g., A. Clark, Manual, Compiled and Prepared for the Use of the [New York] Assembly 99 (1816); L. Cushing, Manual of Parliamentary Practice, Rules of Proceeding and Debate in Deliberative Assemblies 30 (7th ed. 1854). Federal conflict-of-interest rules applicable to judges also date back to the founding. In 1792, Congress passed a law requiring district court judges to recuse themselves if they had a personal interest in a suit or had been counsel to a party appearing before them. Act of May 8, 1792, ch. 36, §11, 1 Stat. 278–279. In 1821, Congress expanded these bases for recusal to include situations in which “the judge … is so related to, or connected with, either party, as to render it improper for him, in his opinion, to sit on the trial of such suit.” Act of Mar. 3, 1821, ch. 51, 3 Stat. 643. The statute was again expanded in 1911, to make any “personal bias or prejudice” a basis for recusal. Act of Mar. 3, 1911, §21, 36 Stat. 1090. The current version, which retains much of the 1911 version’s language, is codified at 28 U. S. C. §144. See generally Liteky v. United States, 510 U. S. 540, 544 (1994); Frank, Disqualification of Judges, 56 Yale L. J. 605, 626–630 (1947) (hereinafter Frank). There are of course differences between a legislator’s vote and a judge’s, and thus between legislative and judicial recusal rules; nevertheless, there do not appear to have been any serious challenges to judicial recusal statutes as having unconstitutionally restricted judges’ First Amendment rights.[Footnote 3] The Nevada Supreme Court’s belief that recusal rules violate legislators’ First Amendment rights is also inconsistent with long-standing traditions in the States. A number of States, by common-law rule, have long required recusal of public officials with a conflict. See, e.g., In re Nashua, 12 N. H. 425, 430 (1841) (“If one of the commissioners be interested, he shall not serve”); Commissioners’ Court v. Tarver, 25 Ala. 480, 481 (1854) (“If any member … has a peculiar, personal interest, such member would be disqualified”); Stubbs v. Florida State Finance Co., 118 Fla. 450, 451, 159 So. 527, 528 (1935) (“[A] public official cannot legally participate in his official capacity in the decision of a question in which he is personally and adversely interested”).[Footnote 4] Today, virtually every State has enacted some type of recusal law, many of which, not unlike Nevada’s, require public officials to abstain from voting on all matters presenting a conflict of interest. See National Conference of State Legisla-tures, Voting Recusal Provisions (2009), online at http:// www.ncsl.org/?TabID=15357 (as visited June 9, 2011, and available in Clerk of Court’s case file). In an attempt to combat this overwhelming evidence of constitutional acceptability, Carrigan relies on a handful of lower-court cases from the 1980’s and afterwards. See Brief for Respondent 25 (citing Clark v. United States, 886 F. 2d 404 (CADC 1989); Miller v. Hull, 878 F. 2d 523 (CA1 1989); and Camacho v. Brandon, 317 F. 3d 153 (CA2 2003)). Even if they were relevant, those cases would be too little and too late to contradict the long-recognized need for legislative recusal. But they are not relevant. The first was vacated as moot, see Clark v. United States, 915 F. 2d 699, 700, 706 (CADC 1990) (en banc), and the other two involve retaliation amounting to viewpoint discrimination. See Miller, supra, at 533; Camacho, supra, at 160. In the past we have applied heightened scrutiny to laws that are viewpoint discriminatory even as to speech not protected by the First Amendment, see R. A. V. v. St. Paul, 505 U. S. 377, 383–386 (1992). Carrigan does not assert that the recusal laws here are viewpoint discriminatory, nor could he: The statute is content-neutral and applies equally to all legislators regardless of party or position. IV But how can it be that restrictions upon legislators’ voting are not restrictions upon legislators’ protected speech? The answer is that a legislator’s vote is the commitment of his apportioned share of the legislature’s power to the passage or defeat of a particular proposal. The legislative power thus committed is not personal to the legislator but belongs to the people; the legislator has no personal right to it. As we said in Raines v. Byrd, 521 U. S. 811, 821 (1997), when denying Article III standing to legislators who claimed that their voting power had been diluted by a statute providing for a line-item veto, the legislator casts his vote “as trustee for his constituents, not as a prerogative of personal power.” In this respect, voting by a legislator is different from voting by a citizen. While “a voter’s franchise is a personal right,” “[t]he procedures for voting in legislative assemblies … pertain to legislators not as individuals but as political representatives executing the legislative process.” Coleman v. Miller, 307 U. S. 433, 469–470 (1939) (opinion of Frankfurter, J.). Carrigan and Justice Alito say that legislators often “ ‘us[e] their votes to express deeply held and highly unpopular views, often at great personal or political peril.’ ” Post, at 1 (opinion concurring in part and concurring in judgment) (quoting Brief for Respondent 23). How do they express those deeply held views, one wonders? Do ballots contain a check-one-of-the-boxes attachment that will be displayed to the public, reading something like “( ) I have a deeply held view about this; ( ) this is probably desirable; ( ) this is the least of the available evils; ( ) my personal view is the other way, but my constituents want this; ( ) my personal view is the other way, but my big contributors want this; ( ) I don’t have the slightest idea what this legislation does, but on my way in to vote the party Whip said vote ‘aye’ ”? There are, to be sure, instances where action conveys a symbolic meaning—such as the burning of a flag to convey disagreement with a country’s policies, see Texas v. Johnson, 491 U. S. 397, 406 (1989). But the act of voting symbolizes nothing. It discloses, to be sure, that the legislator wishes (for whatever reason) that the proposition on the floor be adopted, just as a physical assault discloses that the attacker dislikes the victim. But neither the one nor the other is an act of communication. Cf. Rumsfeld v. Forum for Academic and Institutional Rights, Inc., 547 U. S. 47, 66 (2006) (expressive value was “not created by the conduct itself but by the speech that accompanies it”). Moreover, the fact that a nonsymbolic act is the product of deeply held personal belief—even if the actor would like it to convey his deeply held personal belief—does not transform action into First Amendment speech. Nor does the fact that action may have social consequences—such as the unpopularity that cost John Quincy Adams his Senate seat resulting from his vote in favor of the Embargo Act of 1807, see post, at 1. However unpopular Adams’ vote may have made him, and however deeply Adams felt that his vote was the right thing to do, the act of voting was still nonsymbolic conduct engaged in for an independent governmental purpose. Even if it were true that the vote itself could “express deeply held and highly unpopular views,” the argument would still miss the mark. This Court has rejected the notion that the First Amendment confers a right to use governmental mechanics to convey a message. For example, in Timmons v. Twin Cities Area New Party, 520 U. S. 351 (1997), we upheld a State’s prohibition on multiple-party or “fusion” candidates for elected office against a First Amendment challenge. We admitted that a State’s ban on a person’s appearing on the ballot as the candidate of more than one party might prevent a party from “using the ballot to communicate to the public it supports a particular candidate who is already another party’s candidate,” id., at 362; but we nonetheless were “unpersuaded … by the party’s contention that it has a right to use the ballot itself to send a particularized message.” Id., at 362–363; see also Burdick v. Takushi, 504 U. S. 428, 438 (1992). In like manner, a legislator has no right to use official powers for expressive purposes. Carrigan and Justice Alito also cite Doe v. Reed, 561 U. S. ___ (2010), as establishing “the expressive character of voting.” Post, at 2; see also Brief for Respondent 26. But Reed did no such thing. That case held only that a citizen’s signing of a petition—“ ‘core political speech,’ ” Meyer v. Grant, 486 U. S. 414, 421–422 (1988)—was not deprived of its protected status simply because, under state law, a petition that garnered a sufficient number of signatures would suspend the state law to which it pertained, pending a referendum. See Reed, 561 U. S., at ___ (slip op., at 6); id., at ___ (slip op., at 3) (opinion of Scalia, J.). It is one thing to say that an inherently expressive act remains so despite its having governmental effect, but it is altogether another thing to say that a governmental act becomes expressive simply because the governmental actor wishes it to be so. We have never said the latter is true.[Footnote 5] V Carrigan raises two additional arguments in his brief: that Nevada’s catchall provision unconstitutionally burdens the right of association of officials and supporters, and that the provision is unconstitutionally vague. Whatever the merits of these arguments, we have no occasion to consider them. Neither was decided below: The Nevada Supreme Court made no mention of the former argument and said that it need not address the latter given its resolution of the overbreadth challenge, 126 Nev. ___, n. 4, 236 P. 3d, at 619, n. 4. Nor was either argument raised in Carrigan’s brief in opposition to the petition for writ of certiorari. Arguments thus omitted are normally considered waived, see this Court’s Rule 15.2; Baldwin v. Reese, 541 U. S. 27, 34 (2004), and we find no reason to sidestep that Rule here. * * * The judgment of the Nevada Supreme Court is reversed, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered. Footnote 1 At the time of the relevant events in this case, the disclosure and recusal provisions of the Ethics in Government Law were codified at Nev. Rev. Stat. §281.501 (2003). They were recodified without relevant change in 2007 at §281A.420, and all citations are to that version. The Nevada Legislature further amended the statute in 2009, see Nev. Stats., ch. 257, §9.5, p. 1057, but those changes are not relevant here. Footnote 2 The House first achieved a quorum on April 1, 1789, 1 Annals of Cong. 96, and it adopted rules governing its procedures on April 7, 1789, see id., at 98–99. Footnote 3 We have held that restrictions on judges’ speech during elections are a different matter. See Republican Party of Minn. v. White, 536 U. S. 765, 788 (2002) (holding that it violated the First Amendment to prohibit announcement of views on disputed legal and political issues by candidates for judicial election). Footnote 4 A number of States enacted early judicial recusal laws as well. See, e.g., 1797 Vt. Laws, §23, p. 178 (“[N]o justice of the peace shall take cognizance of any cause, where he shall be within either the first, second, third, or fourth degree of affinity, or consanguinity, to either of the parties, or shall be directly or indirectly interested, in the cause or matter to be determined”); 1818 Mass. Laws, §5, p. 632 (“[W]henever any Judge of Probate shall be interested in the estate of any person deceased, within the county of such Judge, such estate shall be settled in the Probate Court of the most ancient next adjoining county …”); Macon v. Huff, 60 Ga. 221, 223–226 (1878). See generally Frank 609–626. Footnote 5 Justice Alito reasons as follows: (1) If an ordinary citizen were to vote in a straw poll on an issue pending before a legislative body, that vote would be speech; (2) if a member of the legislative body were to do the same, it would be no less expressive; therefore (3) the legislator’s actual vote must also be expressive. This conclusion does not follow. A legislator voting on a bill is not fairly analogized to one simply discussing that bill or expressing an opinion for or against it. The former is performing a governmental act as a representative of his constituents, see supra, at 8; only the latter is exercising personal First Amendment rights.

562.US.180

Petitioner Ortiz, a former inmate in an Ohio reformatory, brought a civil rights action under 42 U. S. C. §1983 seeking a judgment for damages against superintending prison officers. On two consecutive nights during her incarceration, Ortiz stated, she was sexually assaulted by a corrections officer. Although she promptly reported the first assault, she further alleged, respondent Jordan, a case manager in her living unit, did nothing to ward off the second sexual assault, despite Jordan’s awareness of the substantial risk of that occurrence. Ortiz further charged that respondent Bright, a prison investigator, retaliated against Ortiz for her accusations by placing her, shackled and handcuffed, in solitary confinement in a cell without adequate heat, clothing, bedding, or blankets. The responses of both officers, she said, violated her right, safeguarded by the Eighth and Fourteenth Amendments, to reasonable protection from violence while in custody. Jordan and Bright moved for summary judgment on pleas of “qualified immunity.” The District Court, noting factual disputes material to Ortiz’s claims and the officers’ qualified immunity defenses, denied the summary judgment motion. The officers did not appeal that ruling. The case proceeded to trial, and the jury returned a verdict against Jordan and Bright. They sought judgment as a matter of law, pursuant to Federal Rule of Civil Procedure 50(a), both at the close of Ortiz’s evidence and at the close of their own presentation. But they did not contest the jury’s liability finding by renewing, under Rule 50(b), their request for judgment as a matter of law. Nor did they request a new trial under Rule 59(a). The District Court entered judgment for Ortiz. On appeal, Jordan and Bright urged, inter alia, that the District Court should have granted their motion for summary judgment based on their qualified immunity defense. The Sixth Circuit agreed and reversed the judgment entered on the jury’s verdict, holding that both defendants were sheltered from Ortiz’s suit by qualified immunity. Held: A party may not appeal a denial of summary judgment after a district court has conducted a full trial on the merits. A qualified immunity plea, not upheld at the summary judgment stage, may be pursued at trial, but at that stage, the plea must be evaluated in light of the character and quality of the evidence received in court. Ordinarily, orders denying summary judgment are interlocutory and do not qualify as “final decisions” subject to appeal under 28 U. S. C. §1291. Because a qualified immunity plea can spare an official not only from liability but from trial, this Court has recognized a limited exception to the categorization of summary judgment denials as nonappealable orders. Mitchell v. Forsyth, 472 U. S. 511, 525–526. The exception permits an immediate appeal when summary judgment is denied to a defendant who urges that qualified immunity shelters her from suit. Id., at 527. Such an immediate appeal is not available, however, when the district court determines that factual issues genuinely in dispute preclude summary adjudication. Johnson v. Jones, 515 U. S. 304, 313. Here, Jordan and Bright sought no immediate appeal from the denial of their summary judgment motion. Nor did they avail themselves of Rule 50(b), which permits the entry of judgment, postverdict, for the verdict loser if the court finds the evidence legally insufficient to sustain the verdict. Absent such a motion, an appellate court is “powerless” to review the sufficiency of the evidence after trial. Unitherm Food Systems, Inc. v. Swift-Eckrich, Inc., 546 U. S. 394, 405. This Court need not address the officers’ argument that a qualified immunity plea raising a “purely legal” issue is preserved for appeal by an unsuccessful summary judgment motion even if the plea is not reiterated in a Rule 50(b) motion. Cases fitting that bill typically involve disputes about the substance and clarity of pre-existing law. In this case, however, what was controverted was not the pre-existing law, but the facts that could render Jordan and Bright answerable under §1983, e.g., whether Jordan was adequately informed, after the first assault, of the assailant’s identity and of Ortiz’s fear of a further assault. Because the dispositive facts were disputed, the officers’ qualified immunity defenses did not present “ ‘neat abstract issues of law.’ ” Johnson, 515 U. S., at 317. To the extent that Jordan and Bright urge Ortiz has not proved her case, they were, by their own account, obliged to raise that sufficiency-of-the-evidence issue by postverdict motion for judgment as a matter of law under Rule 50(b). They did not do so. The Sixth Circuit, therefore, had no warrant to upset the jury’s decision on their liability. Pp. 6–11. 316 Fed. Appx. 449, reversed and remanded. Ginsburg, J., delivered the opinion of the Court, in which Roberts, C. J., and Breyer, Alito, Sotomayor, and Kagan, JJ., joined. Thomas, J., filed an opinion concurring in the judgment, in which Scalia and Kennedy, JJ., joined.

We address in this case a procedural issue arising in a civil rights action brought under 42 U. S. C. §1983 by Michelle Ortiz, a former inmate at the Ohio Reformatory for Women. Plaintiff below, petitioner here, Ortiz filed a complaint in federal court stating key facts on which she based claims for damages against superintending prison officers. On two consecutive nights during her one-year incarceration, Ortiz stated, she was sexually assaulted by a corrections officer. Although she promptly reported the first incident, she further alleged, prison authorities took no measures to protect her against the second assault. After that assault, Ortiz charged, and in retaliation for accounts she gave of the two episodes, prison officials placed her, shackled and handcuffed, in solitary confinement in a cell without adequate heat, clothing, bedding, or blankets. The treatment to which she was exposed, Ortiz claimed, violated her right, safeguarded by the Eighth and Fourteenth Amendments, to reasonable protection from violence while in custody. Principal defendants in the suit, Paula Jordan, a case manager at Ortiz’s living unit, and Rebecca Bright, a prison investigator, moved for summary judgment on their pleas of “qualified immunity,” a defense that shields officials from suit if their conduct “d[id] not violate clearly established statutory or constitutional rights of which a reasonable person would have known.” Harlow v. Fitzgerald, 457 U. S. 800, 818 (1982). Finding that the qualified immunity defense turned on material facts genuinely in dispute, see Fed. Rule Civ. Proc. 56(a), the District Judge denied summary judgment. Ortiz v. Voinovich, 211 F. Supp. 2d 917, 923–930 (SD Ohio 2002). The case then proceeded to trial, and the jury returned verdicts for Ortiz against both Jordan and Bright. The two officers appealed to the United States Court of Appeals for the Sixth Circuit, targeting, inter alia, the denial of their pretrial motion for summary judgment. “[C]ourts normally do not review the denial of a summary judgment motion after a trial on the merits,” the Court of Appeals recognized. 316 Fed. Appx. 449, 453 (2009). Nevertheless, the Court continued, “denial of summary judgment based on qualified immunity is an exception to this rule.” Ibid. Reversing the judgment entered on the jury’s verdict, the appeals court held that both defendants were sheltered from Ortiz’s suit by qualified immunity. We granted review, 559 U. S. ___ (2010), to decide a threshold question on which the Circuits are split: May a party, as the Sixth Circuit believed, appeal an order denying summary judgment after a full trial on the merits?[Footnote 1] Our answer is no. The order retains its interlocutory character as simply a step along the route to final judgment. See Cohen v. Beneficial Industrial Loan Corp., 337 U. S. 541, 546 (1949). Once the case proceeds to trial, the full record developed in court supersedes the record existing at the time of the summary judgment motion. A qualified immunity defense, of course, does not vanish when a district court declines to rule on the plea summarily. The plea remains available to the defending officials at trial; but at that stage, the defense must be evaluated in light of the character and quality of the evidence received in court. When summary judgment is sought on a qualified immunity defense, the court inquires whether the party opposing the motion has raised any triable issue barring summary adjudication. “[O]nce trial has been had,” however, “the availability of official immunity should be determined by the trial record, not the pleadings nor the summary judgment record.” 15A C. Wright, A. Miller, & E. Cooper, Federal Practice & Procedure §3914.10, p. 684 (2d ed. 1992 and Supp. 2010). After trial, if defendants continue to urge qualified immunity, the decisive question, ordinarily, is whether the evidence favoring the party seeking relief is legally sufficient to overcome the defense. See Fed. Rule Civ. Proc. 50(a), (b) (stating conditions on which judgment may be granted as a matter of law). In the case before us, the Court of Appeals, although purporting to review the District Court’s denial of the prison officials’ pretrial summary-judgment motion, 316 Fed. Appx., at 453, several times pointed to evidence presented only at the trial stage of the proceedings, see id., at 453–454. The appeals court erred, but not fatally, by incorrectly placing its ruling under a summary-judgment headline. Its judgment was infirm, however, because Jordan’s and Bright’s failure to renew their motion for judgment as a matter of law under Federal Rule of Civil Procedure 50(b) left the appellate forum with no warrant to reject the appraisal of the evidence by “the judge who saw and heard the witnesses and ha[d] the feel of the case which no appellate printed transcript can impart.” Cone v. West Virginia Pulp & Paper Co., 330 U. S. 212, 216 (1947). I Michelle Ortiz, serving a sentence for aggravated assault against her husband,[Footnote 2] maintained that she was sexually assaulted on consecutive days by Corrections Officer Douglas Schultz. On Friday, November 8, 1996, Ortiz recounted, Schultz walked up behind her in the washroom of her living quarters and grabbed one of her breasts.[Footnote 3] Ortiz fended off the assault, but Schultz returned later that day and threatened to “see [her] tomorrow,” Tr. 36. The next day, Ortiz described the incident to Jordan. After assuring Ortiz that “no one has the right to touch you,” 316 Fed. Appx., at 451 (internal quotation marks omitted), Jordan told Ortiz that Schultz had been reassigned to another correctional facility and was serving his last day at the reformatory. Ortiz could file a written complaint, Jordan noted. She suggested, however, that Ortiz not do so, in view of Schultz’s imminent departure. Jordan advised Ortiz that she “always ha[d] the right to defend [her]self,” Tr. 43, and counseled her to “ ‘hang out with [her] friends’ for the rest of the day so that Schultz would not have the chance to be alone with her,” 316 Fed. Appx., at 451 (alteration in original). The day of her conversation with Ortiz, Jordan wrote an incident report describing her version of the encounter. In that account, Jordan stated that Ortiz had refused to name her assailant or provide any other information about the assault. Jordan did not immediately notify her superiors of the assault Ortiz reported and Ortiz’s consequent fears about her safety. Taking the incident report home in her workbag, Jordan submitted it upon her return to work two days later. Ortiz endeavored to follow Jordan’s advice about staying in the company of friends. But later in the day, feeling ill, she returned to her room and fell asleep. Three other inmates were in the room when Ortiz went to sleep. They were gone when she awoke to find Schultz standing over her, one hand fondling her left breast, the fingers of the other hand inside her underwear penetrating her vagina. Bright’s investigation began two days after the second assault. During its course, Bright placed Ortiz in solitary confinement. Ortiz maintained that Bright isolated her in retaliation for her accusations against Schultz. Bright, however, testified that she segregated Ortiz because Ortiz continued to discuss the investigation with other inmates, disobeying Bright’s repeated instructions to refrain from speaking about it. In her §1983 action, Ortiz claimed that Jordan did nothing to ward off Schultz’s second sexual assault, despite Jordan’s awareness of the substantial risk of that occurrence. Bright, Ortiz charged, retaliated against her because she resisted Bright’s efforts to induce her to retract her accounts of Schultz’s assaults. (Schultz, having resigned from state employment, could not be found and served with process.) The District Court, noting multiple factual disputes material to Ortiz’s claims and the officers’ defense of qualified immunity, denied summary judgment to Jordan and Bright, 211 F. Supp. 2d, at 923–930;[Footnote 4] neither defendant appealed the District Court’s denial of summary judgment. The case proceeded to trial, and a jury returned a verdict of $350,000 in compensatory and punitive damages against Jordan and $275,000 against Bright. Jordan and Bright sought judgment as a matter of law, pursuant to Rule 50(a), both at the close of Ortiz’s evidence and at the close of their own presentation. But they did not contest the jury’s liability finding by renewing, under Rule 50(b), their request for judgment as a matter of law. Nor did they request a new trial under Rule 59(a). The District Court entered judgment for Ortiz in accordance with the jury’s verdict. On appeal, Jordan and Bright urged both that the District Court should have granted them summary judgment on their defense of qualified immunity and that the ver-dict was “against the weight of the evidence.” Brief for Defendants-Appellants in No. 06–3627 (CA6), pp. 21, 26. Appraising the parties’ evidence under a de novo standard of review, the Court of Appeals “reverse[d] the denial of qualified immunity to both Bright and Jordan.” 316 Fed. Appx., at 455.[Footnote 5] We granted certiorari to resolve the conflict among the Circuits as to whether a party may appeal a denial of summary judgment after a district court has conducted a full trial on the merits. See n. 1, supra. II The jurisdiction of a Court of Appeals under 28 U. S. C. §1291 extends only to “appeals from … final decisions of the district courts.” Ordinarily, orders denying summary judgment do not qualify as “final decisions” subject to appeal. Summary judgment must be denied when the court of first instance determines that a “genuine dispute as to [a] material fact” precludes immediate entry of judgment as a matter of law. Fed. Rule Civ. Proc. 56(a). Such rulings, we have observed, are “by their terms interlocutory.” Liberty Mut. Ins. Co. v. Wetzel, 424 U. S. 737, 744 (1976). Because a plea of qualified immunity can spare an official not only from liability but from trial, we have recognized a limited exception to the categorization of summary judgment denials as nonappealable orders. Mitchell v. Forsyth, 472 U. S. 511, 525–526 (1985). When summary judgment is denied to a defendant who urges that qualified immunity shelters her from suit, the court’s order “finally and conclusively [disposes of] the defendant’s claim of right not to stand trial.” Id., at 527 (emphasis deleted). Therefore, Mitchell held, an immediate appeal may be pursued. Ibid. We clarified in Johnson v. Jones, 515 U. S. 304 (1995), that immediate appeal from the denial of summary judgment on a qualified immunity plea is available when the appeal presents a “purely legal issue,” illustratively, the determination of “what law was ‘clearly established’ ” at the time the defendant acted. Id., at 313. However, instant appeal is not available, Johnson held, when the district court determines that factual issues genuinely in dispute preclude summary adjudication. Ibid. Jordan and Bright sought no immediate appeal from the denial of their motion for summary judgment. In light of Johnson, that abstinence is unsurprising. Moreover, even had instant appellate review been open to them, the time to seek that review expired well in advance of trial. See Fed. Rule App. Proc. 4(a)(1)(A) (notice of appeal must generally be filed “within 30 days after the judgment or order appealed from”). Nor did they avail themselves of Rule 50(b), which permits the entry, postverdict, of judgment for the verdict loser if the court finds that the evidence was legally insufficient to sustain the verdict. See Fed. Rule Civ. Proc. 50(a), (b). Absent such a motion, we have repeatedly held, an appellate court is “powerless” to review the sufficiency of the evidence after trial. Unitherm Food Systems, Inc. v. Swift-Eckrich, Inc., 546 U. S. 394, 405 (2006); see Cone, 330 U. S., at 218.[Footnote 6] “[Q]uestions going to the sufficiency of the evidence are not preserved for appellate review by a summary judgment motion alone,” Jordan and Bright acknowledge; rather, challenges of that order “must be renewed post-trial under Rule 50.” Brief for Respondents 11. Jordan and Bright insist, however, in defense of the Sixth Circuit’s judgment, that sufficiency of the evidence is not what is at stake in this case. A qualified immunity plea raising an issue of a “purely legal nature,” they urge, ibid., is preserved for appeal by an unsuccessful motion for summary judgment, and need not be brought up again under Rule 50(b). Id., at 11–12 (citing as pathmarking Rekhi v. Wildwood Indus., Inc., 61 F. 3d 1313, 1318 (CA7 1995)). Unlike an “evidence sufficiency” claim that necessarily “hinge[s] on the facts adduced at trial,” they maintain, a purely legal issue can be resolved “with reference only to undisputed facts.” Brief for Respondents 16 (quoting Mitchell, 472 U. S., at 530, n. 10). We need not address this argument, for the officials’ claims of qualified immunity hardly present “purely legal” issues capable of resolution “with reference only to undisputed facts.” Cases fitting that bill typically involve contests not about what occurred, or why an action was taken or omitted, but disputes about the substance and clarity of pre-existing law. See Behrens v. Pelletier, 516 U. S. 299, 313 (1996); Johnson, 515 U. S. at 317. Here, however, the pre-existing law was not in controversy. See Farmer v. Brennan, 511 U. S. 825, 834, 847 (1994) (prison official may be held liable for “deliberate indifference” to a prisoner’s Eighth Amendment right to protection against violence while in custody if the official “knows that [the] inmat[e] face[s] a substantial risk of serious harm and disregards that risk by failing to take reasonable measures to abate it” (internal quotation marks omitted)); Crawford-El v. Britton, 523 U. S. 574, 592 (1998) (First Amendment shields prisoners from “retaliation for protected speech”).[Footnote 7] What was controverted, instead, were the facts that could render Jordan and Bright answerable for crossing a constitutional line. Disputed facts relevant to resolving the officials’ immunity pleas included: Was Jordan adequately informed, after the first assault, of the identity of the assailant, see App. 4, and of Ortiz’s fear of a further assault? What, if anything, could Jordan have done to distance Ortiz from the assailant, thereby insulating her against a second assault? See id., at 4–5.[Footnote 8] Did Bright place and retain Ortiz in solitary confinement as a retaliatory measure or as a control needed to safeguard the integrity of the investigation? [Footnote 9] In sum, the qualified immunity defenses asserted by Jordan and Bright do not present “neat abstract issues of law.” See Johnson, 515 U. S., at 317 (quoting 15A C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §3914.10, p. 644 (1992)). To the extent that the officials urge Ortiz has not proved her case, they were, by their own account, obliged to raise that sufficiency-of-the-evidence issue by postverdict motion for judgment as a matter of law under Rule 50(b). See Brief for Respondents 11. They did not do so. The Court of Appeals, therefore, had no warrant to upset the jury’s decision on the officials’ liability. * * * For the reasons stated, the judgment of the Court of Appeals for the Sixth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Footnote 1 Compare, e.g., Black v. J. I. Case Co., 22 F. 3d 568, 570–571 (CA5 1994) (declining to review denial of summary judgment after trial); Price v. Kramer, 200 F. 3d 1237, 1243–1244 (CA9 2000) (no exception where summary judgment rejected assertion of qualified immunity), with Goff v. Bise, 173 F. 3d 1068, 1072 (CA8 1999) (denial of summary judgment based on qualified immunity reviewable after trial on the merits); 316 Fed. Appx. 449, 453 (CA6 2009) (case below) (same). Footnote 2 Ortiz maintained in her criminal prosecution that she acted in retaliation for multiple incidents of domestic violence to which she had been subjected over a span of several years. 316 Fed. Appx., at 450. Footnote 3 The accounts of the episodes-in-suit described here recite facts the jury reasonably could have found from the testimony presented at trial. Footnote 4 As to Jordan, the District Court determined, inter alia, that “a jury could reasonably find … that Ortiz did inform Jordan of Schultz’s identity,” and that “Ortiz made Jordan aware that [Ortiz] reasonably feared a further sexual attack.” 211 F. Supp. 2d, at 925. Concerning Bright, the Court noted that “[a] jury could reasonably find that Bright[’s] … purported reason for placing Ortiz in security control was pretextual.” Id., at 928. Footnote 5 Judge Daughtrey dissented; in her view, the strength of the evidence against Jordan and Bright amply supported the jury’s verdict. 316 Fed. Appx., at 456–457. Quoting Kiphart v. Saturn Corp., 251 F. 3d 573, 581 (CA6 2001), she observed that appellate courts “do not weigh the evidence, evaluate the credibility of witnesses or substitute our own judgment for that of the jury.” 316 Fed. Appx., at 457. Footnote 6 Jordan and Bright contend that their failure to file a Rule 50(b) motion cannot be the basis for overturning the judgment of the Court of Appeals. Because Ortiz presented no argument about the absence of a Rule 50(b) motion in her brief to the Sixth Circuit or in her petition for certiorari, Jordan and Bright argue, she has forfeited the objection. Jordan and Bright are not well positioned to make this argument. They did not suggest that Ortiz had forfeited her Rule 50(b) objection, or argue that such an objection is forfeitable, until their merits brief to this Court. Ordinarily we do not consider “a nonjurisdictional argument not raised in a respondent’s brief in opposition to a petition.” Baldwin v. Reese, 541 U. S. 27, 34 (2004) (internal quotation marks omitted). In any case, we do not see how Ortiz can be held to have forfeited her Rule 50(b) objection. The arguments Jordan and Bright made in the Court of Appeals invited no such objection. Jordan and Bright urged on appeal that the jury’s verdict was “against the weight of the evidence.” Brief for Defendants-Appellants in No. 06-3627 (CA6), pp. 21–43. A plea that a verdict is “against the weight of the evidence,” of course, is not equivalent to a plea that the evidence submitted at trial was insufficient to warrant submission of the case to the jury. 11 C. Wright, A. Miller, & E. Cooper, Federal Practice & Procedure §2806, pp. 65–67 (2d ed. 1995 and Supp. 2010). A determination that a verdict is against the weight of the evidence may gain a new trial for the verdict loser, but never a final judgment in that party’s favor. Montgomery Ward & Co. v. Duncan, 311 U. S. 243, 250–251 (1940). Ortiz objected, accordingly, that Jordan and Bright had not asked the District Court for a new trial and were therefore “bar[red] [from] appeal on this ground.” Brief for Plaintiff-Appellee in No. 06–3627 (CA6), p. 23. Ortiz thus responded altogether appropriately to the arguments Jordan and Bright made. Footnote 7 The Court of Appeals held that Ortiz’s complaint had not properly tied her claim against Bright to the First Amendment. 316 Fed. Appx., at 455. “When an issue not raised by the pleadings is tried by the parties’ express or implied consent,” however, “it must be treated in all respects as if raised in the pleadings.” Fed. Rule Civ. Proc. 15(b)(2). Bright, like the District Court, recognized the First Amendment interests at stake in Ortiz’s claim against her. See App. 11 (District Court, in ruling on Rule 50(a) motion, inquired into Bright’s authority to “regulat[e] speech of inmates”); id., at 20 (Bright’s counsel argued that Ortiz’s segregation “would not have had a chilling effect” on her speech). Footnote 8 Relevant to that matter, Bright testified at trial that, had Jordan “reported the first incident immediately, ‘the proper people would have taken a role in protecting Mrs. Ortiz.’ ” 316 Fed. Appx., at 457 (Daughtrey, J., dissenting). Footnote 9 Apart from Bright’s testimony, the defense produced no evidence that Ortiz “continually” talked about the assaults or in any other way interfered with the investigation. See Tr. 271, 397 (Bright’s testimony acknowledging absence of any documentation regarding Ortiz’s conduct during investigation).

562.US.476

After pleading guilty to drug charges, petitioner Pepper was sentenced under the Federal Sentencing Guidelines to 24 months’ imprisonment, a nearly 75 percent downward departure from the low end of the Guidelines range based in part on his substantial assistance, followed by five years of supervised release. In Pepper I, the Eighth Circuit reversed and remanded for resentencing in light of, inter alia, United States v. Booker, 543 U. S. 220. Pepper, who had begun serving his supervised release, testified at his resentencing hearing that he was no longer a drug addict, having completed a 500-hour drug treatment program while in prison; that he was enrolled in community college and had achieved very good grades; and that he was working part time. Pepper’s father testified that he and his son were no longer estranged, and Pepper’s probation officer testified that a 24-month sentence would be reasonable in light of Pepper’s substantial assistance, postsentencing rehabilitation, and demonstrated low recidivism risk. The District Court again sentenced Pepper to 24 months, granting a 40 percent downward departure based on Pepper’s substantial assistance and a further downward variance based on, inter alia, Pepper’s rehabilitation since his initial sentencing. In Pepper II, the Eighth Circuit again reversed and remanded for resentencing, concluding that Pepper’s postsentencing rehabilitation could not be considered as a factor supporting a downward variance, and directing that the case be assigned to a different district judge. After this Court vacated and remanded the Pepper II judgment in light of Gall v. United States, 552 U. S. 38, the Eighth Circuit, in Pepper III, reversed and remanded once more. At the second resentencing hearing, Pepper informed the new district judge that he was still in school, was about to be promoted at his job, and had married and was supporting his new family. Noting the nearly identical remand language of Pepper II and Pepper III, the court observed that it was not bound to reduce Pepper’s range by 40 percent for substantial assistance. Instead, it found him entitled to a 20 percent reduction and refused to grant a further downward variance for, inter alia, postsentencing rehabilitation. It imposed a 65-month prison term and 12 months of supervised release. In Pepper IV, the Eighth Circuit once again rejected Pepper’s postsentencing rehabilitation argument. It also rejected his claim that the law of the case from Pepper II and Pepper III required the District Court to reduce the applicable Guidelines range by at least 40 percent. Held: 1. When a defendant’s sentence has been set aside on appeal, a district court at resentencing may consider evidence of the defendant’s postsentencing rehabilitation, and such evidence may, in appropriate cases, support a downward variance from the now-advisory Guidelines range. Pp. 9–27. (a) Consistent with the principle that “the punishment should fit the offender and not merely the crime,” Williams v. New York, 337 U. S. 241, 247, this Court has observed a consistent and uniform policy “under which a sentencing judge could exercise a wide discretion in the sources and types of evidence used to assist him in determining the kind and extent of punishment to be imposed within limits fixed by law,” id., at 246, particularly “the fullest information possible concerning the defendant’s life and characteristics,” id., at 247. That principle is codified at 18 U. S. C. §3661, which provides that “[n]o limitation shall be placed on the information” a sentencing court may consider “concerning the [defendant’s] background, character, and conduct,” and at §3553(a), which specifies that sentencing courts must consider, among other things, a defendant’s “history and characteristics,” §3553(a)(1). The Guidelines, which Booker made “effectively advisory,” 543 U. S., at 245, “should be the starting point and the initial benchmark,” but district courts may impose sentences within statutory limits based on appropriate consideration of all of the §3553(a) factors, subject to appellate review for “reasonableness,” Gall v. United States, 552 U. S. 38, 49–51. This sentencing framework applies both at initial sentencing and at any subsequent resentencing after a sentence has been set aside on appeal. Pp. 9–12. (b) Postsentencing rehabilitation evidence may support a downward variance from the advisory Guidelines range. The plain language of §3661 makes clear that there is “[n]o limitation … on … background, character, and conduct” information, and it makes no distinction between an initial sentencing and a subsequent resentencing. In addition, postsentencing rehabilitation evidence may be highly relevant to several §3553(a) factors that district courts are required to consider at sentencing. The extensive evidence of Pepper’s rehabilitation since his initial sentencing is clearly relevant to the selection of an appropriate sentence here. Most fundamentally, that evidence provides the most up-to-date picture of his “history and characteristics.” §3553(a)(1). At the time of his initial sentencing, he was an unemployed drug addict who was estranged from his family and sold drugs. By his second resentencing, he had been drug-free for nearly five years, was attending college, was a top employee slated for promotion, had re-established a relationship with his father, and was married and supporting a family. His postsentencing conduct also sheds light on the likelihood that he will engage in future criminal conduct, a central factor that sentencing courts must consider. See §§3553(a)(2)(B)–(C). Pp. 12–15. (c) The contrary arguments advanced by amicus appointed to defend the judgment are unpersuasive. Pp. 15–26. (1) While §3742(g)(2)—which prohibits a district court at resentencing from imposing a sentence outside the Guidelines range except upon a ground it relied upon at the prior sentencing—effectively precludes a court from considering postsentencing rehabilitation, that provision is invalid after Booker. Like the provisions invalidated in Booker—§§3553(b)(1) and 3742(e)—§3742(g)(2) requires district courts effectively to treat the Guidelines as mandatory in an entire set of cases. Thus, the proper remedy is to invalidate the provision. While applying §3742(g)(2) at resentencing would not always result in a Sixth Amendment violation, this Court rejects a partial invalidation that would leave the Guidelines effectively mandatory in some cases and advisory in others. The fact that §3742(g)(2) permits a resentencing court on remand to impose a non-Guidelines sentence where the prior sentence expressly relied on a departure upheld by the court of appeals also does not cure the constitutional infirmity. And the argument that any constitutional infirmity in §3742(g)(2) can be remedied by invalidating §3742(j)(1)(B) is rejected. Pp. 15–20. (2) This Court finds unpersuasive amicus’ arguments focusing on Congress’ sentencing objectives under §3553(a). Contrary to amicus’ contention, §3742(g)(2) does not reflect a congressional purpose to preclude consideration of postsentencing rehabilitation evidence. Thus, that provision has no bearing on this Court’s analysis of whether §3553(a) permits consideration of such evidence. Nor is the consideration of postsentencing rehabilitation inconsistent with the sentencing factor in §3553(a)(5)—which directs sentencing courts to consider “any pertinent policy statement” of the Sentencing Commission—particularly as the pertinent policy statement in this case is based on unconvincing policy rationales not reflected in the relevant sentencing statutes. Consideration of postsentencing rehabilitation is also not inconsistent with §3553(a)(6)—which requires courts to consider “the need to avoid unwarranted sentenc[ing] disparities among defendants with similar records who have been found guilty of similar conduct”—as any disparity arises only from the normal trial and sentencing process. The differences in procedural opportunity that may result because some defendants are inevitably sentenced in error and must be resentenced are not the kinds of “unwarranted” sentencing disparities that Congress sought to eliminate under §3553(a)(6). Pp. 21–26. (d) On remand, the District Court should consider and give appropriate weight to the postsentencing rehabilitation evidence, as well as any additional evidence concerning Pepper’s conduct since his last sentencing. Pp. 26–27. 2. Because the Eighth Circuit in Pepper III set aside Pepper’s entire sentence and remanded for de novo resentencing, the District Court was not bound by the law of the case doctrine to apply the same 40 percent departure applied by the original sentencing judge. To avoid undermining a district court’s original sentencing intent, an appellate court when reversing one part of a sentence “may vacate the entire sentence … so that, on remand, the trial court can reconfigure the sentencing plan … to satisfy [§3553(a)’s] sentencing factors.” Greenlaw v. United States, 554 U. S. 237, 253. That is what the Eighth Circuit did here. Pp. 27–30. 570 F. 3d 958, vacated in part, affirmed in part, and remanded. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, and Ginsburg, JJ., joined, and in which Breyer and Alito, JJ., joined as to Part III. Breyer, J., filed an opinion concurring in part and concurring in the judgment. Alito, J., filed an opinion concurring in part, concurring in the judgment in part, and dissenting in part. Thomas, J., filed a dissenting opinion. Kagan, J., took no part in the consideration or decision of the case.

This Court has long recognized that sentencing judges “exercise a wide discretion” in the types of evidence they may consider when imposing sentence and that “[h]ighly relevant—if not essential—to [the] selection of an appropriate sentence is the possession of the fullest information possible concerning the defendant’s life and characteristics.” Williams v. New York, 337 U. S. 241, 246–247 (1949). Congress codified this principle at 18 U. S. C. §3661, which provides that “[n]o limitation shall be placed on the information” a sentencing court may consider “concerning the [defendant’s] background, character, and con-duct,” and at §3553(a), which sets forth certain factors that sentencing courts must consider, including “the history and characteristics of the defendant,” §3553(a)(1). The United States Court of Appeals for the Eighth Circuit concluded in this case that the District Court, when resentencing petitioner after his initial sentence had been set aside on appeal, could not consider evidence of petitioner’s rehabilitation since his initial sentencing. That conclusion conflicts with longstanding principles of federal sentenc-ing law and Congress’ express directives in §§3661 and 3553(a). Although a separate statutory provision, §3742(g)(2), prohibits a district court at resentencing from imposing a sentence outside the Federal Sentencing Guidelines range except upon a ground it relied upon at the prior sentencing—thus effectively precluding the court from considering postsentencing rehabilitation for purposes of imposing a non-Guidelines sentence—that provision did not survive our holding in United States v. Booker, 543 U. S. 220 (2005), and we expressly invalidate it today. We hold that when a defendant’s sentence has been set aside on appeal, a district court at resentencing may consider evidence of the defendant’s postsentencing rehabilitation and that such evidence may, in appropriate cases, support a downward variance from the now-advisory Federal Sentencing Guidelines range. Separately, we affirm the Court of Appeals’ ruling that the law of the case doctrine did not require the District Court in this case to apply the same percentage departure from the Guidelines range for substantial assistance that had been applied at petitioner’s prior sentencing. I In October 2003, petitioner Jason Pepper was arrested and charged with conspiracy to distribute 500 grams or more of methamphetamine in violation of 21 U. S. C. §846. After pleading guilty, Pepper appeared for sentencing before then-Chief Judge Mark W. Bennett of the U. S. District Court for the Northern District of Iowa. Pepper’s sentencing range under the Guidelines was 97 to 121 months.[Footnote 1] The Government moved for a downward departure pursuant to USSG §5K1.1 based on Pepper’s substantial assistance and recommended a 15 percent downward departure.[Footnote 2] The District Court, however, sentenced Pepper to a 24-month prison term, resulting in an approximately 75 percent downward departure from the low end of the Guidelines range, to be followed by five years of supervised release. The Government appealed Pepper’s sentence, and in June 2005, the Court of Appeals for the Eighth Circuit reversed and remanded for resentencing in light of our intervening decision in Booker (and for another reason not relevant here). See United States v. Pepper, 412 F. 3d 995, 999 (2005) (Pepper I). Pepper completed his 24-month sentence three days after Pepper I was issued and began serving his term of supervised release. In May 2006, the District Court conducted a resentencing hearing and heard from three witnesses. In his testimony, Pepper first recounted that while he had previously been a drug addict, he successfully completed a 500-hour drug treatment program while in prison and he no longer used any drugs. App. 104–105. Pepper then explained that since his release from prison, he had enrolled at a local community college as a full-time student and had earned A’s in all of his classes in the prior semester. Id., at 106–107. Pepper also testified that he had obtained employment within a few weeks after being released from custody and was continuing to work part-time while attending school. Id., at 106–110. Pepper confirmed that he was in compliance with all the conditions of his supervised release and described his changed attitude since his arrest. See id., at 111 (“[M]y life was basically headed to either where—I guess where I ended up, in prison, or death. Now I have some optimism about my life, about what I can do with my life. I’m glad that I got this chance to try again I guess you could say at a decent life… . My life was going nowhere before, and I think it’s going somewhere now”). Pepper’s father testified that he had virtually no contact with Pepper during the 5-year period leading up to his arrest. Id., at 117. Pepper’s drug treatment program, according to his father, “truly sobered him up” and “made his way of thinking change.” Id., at 121. He explained that Pepper was now “much more mature” and “serious in terms of planning for the future,” id., at 119, and that as a consequence, he had re-established a relationship with his son, id., at 118–119. Finally, Pepper’s probation officer testified that, in his view, a 24-month sentence would be reasonable in light of Pepper’s substantial assistance, postsentencing rehabilitation, and demonstrated low risk of recidivism. Id., at 126–131. The probation officer also prepared a sentencing memorandum that further set forth the reasons supporting his recommendation for a 24-month sentence. The District Court adopted as its findings of fact the testimony of the three witnesses and the probation officer’s sentencing memorandum. The court granted a 40 percent downward departure based on Pepper’s substantial assistance, reducing the bottom of the Guidelines range from 97 to 58 months. The court then granted a further 59 percent downward variance based on, inter alia, Pepper’s rehabilitation since his initial sentencing. Id., at 143–148.[Footnote 3] The court sentenced Pepper to 24 months of imprisonment, concluding that “it would [not] advance any purpose of federal sentencing policy or any other policy behind the federal sentencing guidelines to send this defendant back to prison.” Id., at 149–150. The Government again appealed Pepper’s sentence, and the Court of Appeals again reversed and remanded for resentencing. See United States v. Pepper, 486 F. 3d 408, 410, 413 (CA8 2007) (Pepper II). The court concluded that, while it was “a close call, [it could not] say the district court abused its discretion” by granting the 40 percent downward departure for substantial assistance. Id., at 411. The court found the further 59 percent downward variance, however, to be an abuse of discretion. Id., at 412–413. In doing so, the court held that Pepper’s “post-sentencing rehabilitation was an impermissible factor to consider in granting a downward variance.” Id., at 413. The court stated that evidence of postsentencing reha-bilitation “ ‘is not relevant and will not be permitted at resentencing because the district court could not have considered that evidence at the time of the original sentencing,’ ” and permitting courts to consider post-sentencing rehabilitation at resentencing “would create unwarranted sentencing disparities and inject blatant inequities into the sentencing process.” Ibid.[Footnote 4] The Court of Appeals directed that the case be assigned to a different district judge for resentencing. Ibid. After the Court of Appeals’ mandate issued, Pepper’s case was reassigned on remand to Chief Judge Linda R. Reade. In July 2007, Chief Judge Reade issued an order on the scope of the remand from Pepper II, stating that “[t]he court will not consider itself bound to reduce [Pepper’s] advisory Sentencing Guidelines range by 40% pursuant to USSG §5K1.1.” United States v. Pepper, No. 03–CR–4113–LRR, 2007 WL 2076041, *4 (ND Iowa 2007). In the meantime, Pepper petitioned this Court for a writ of certiorari, and in January 2008, we granted the petition, vacated the judgment in Pepper II, and remanded the case to the Court of Appeals for further consideration in light of Gall v. United States, 552 U. S. 38 (2007). See Pepper v. United States, 552 U. S. 1089 (2008). On remand, the Court of Appeals held that Gall did not alter its prior conclusion that “post-sentence rehabilitation is an impermissible factor to consider in granting a downward variance.” 518 F. 3d 949, 953 (CA8 2008) (Pepper III). The court again reversed the sentence and remanded for resentencing. In October 2008, Chief Judge Reade convened Pepper’s second resentencing hearing. Pepper informed the court that he was still attending school and was now working as a supervisor for the night crew at a warehouse retailer, where he was recently selected by management as “associate of the year” and was likely to be promoted the following January. App. 320, 323. Pepper also stated that he had recently married and was now supporting his wife and her daughter. Id., at 321. Pepper’s father reiterated that Pepper was moving forward in both his career and his family life and that he remained in close touch with his son. See id., at 300–304. In December 2008, Chief Judge Reade issued a sentencing memorandum. Noting that the remand language of Pepper III was nearly identical to the language in Pepper II, the court again observed that it was “not bound to reduce [Pepper’s] advisory Sentencing Guidelines range by 40%” for substantial assistance and concluded that Pepper was entitled only to a 20 percent downward departure because the assistance was “timely, helpful and important” but “in no way extraordinary.” Sealed Sentencing Memorandum in No. 03–CR–4113–LRR (ND Iowa), Doc. 198, pp. 7, 10. The court also rejected Pepper’s request for a downward variance based on, inter alia, his postsentencing rehabilitation. Id., at 16. The District Court reconvened Pepper’s resentencing hearing in January 2009. The court’s decision to grant a 20 percent downward departure for substantial assistance resulted in an advisory Guidelines range of 77 to 97 months. The court also granted the Government’s motion under Rule 35(b) of the Federal Rules of Criminal Procedure to account for investigative assistance Pepper provided after he was initially sentenced. The court imposed a 65-month term of imprisonment, to be followed by 12 months of supervised release.[Footnote 5] The Court of Appeals affirmed Pepper’s 65-month sentence. 570 F. 3d 958 (CA8 2009) (Pepper IV). As relevant here, the Court of Appeals rejected Pepper’s argument that the District Court erred in refusing to consider his postsentencing rehabilitation. The court acknowledged that “Pepper made significant progress during and following his initial period of imprisonment” and “commend[ed] Pepper on the positive changes he has made in his life,” but concluded that Pepper’s argument was foreclosed by Circuit precedent holding that “post-sentencing rehabilitation is not a permissible factor to consider in granting a downward variance.” Id., at 964–965 (citing United States v. Jenners, 473 F. 3d 894, 899 (CA8 2007); United States v. McMannus, 496 F. 3d 846, 852, n. 4 (CA8 2007)). The Court of Appeals also rejected Pepper’s claim that the scope of the remand and the law of the case from Pepper II and Pepper III required the District Court to reduce the applicable Guidelines range by at least 40 percent pursuant to USSG §5K1.1. The court noted that its remand orders in Pepper II and Pepper III were “general remand[s] for resentencing,” which “did not place any limitations on the discretion of the newly assigned district court judge in resentencing.” 570 F. 3d, at 963. The court further noted that, although issues decided by an appellate court become law of the case on remand to the sentencing court, its earlier decisions merely held that a 40 percent downward departure for substantial assistance was not an abuse of discretion, not that the district court would be bound by the 40 percent departure previously granted. Id., at 963–964. We granted Pepper’s petition for a writ of certiorari, 561 U. S. ___ (2010), to decide two questions: (1) whether a district court, after a defendant’s sentence has been set aside on appeal, may consider evidence of a defendant’s postsentencing rehabilitation to support a downward variance when resentencing the defendant, a question that has divided the Courts of Appeals;[Footnote 6] and (2) whether the resentencing court was required, under the law of the case doctrine, to apply the same percentage departure from the Guidelines range for substantial assistance that had been applied at Pepper’s prior sentencing. Because the United States has confessed error in the Court of Appeals’ ruling on the first question, we appointed an amicus curiae to defend the Court of Appeals’ judgment.[Footnote 7] We now vacate the Eighth Circuit’s ruling on the first question and affirm its ruling on the second. II A “It has been uniform and constant in the federal judicial tradition for the sentencing judge to consider every convicted person as an individual and every case as a unique study in the human failings that sometimes mitigate, sometimes magnify, the crime and the punishment to ensue.” Koon v. United States, 518 U. S. 81, 113 (1996). Underlying this tradition is the principle that “the punishment should fit the offender and not merely the crime.” Williams, 337 U. S., at 247; see also Pennsylvania ex rel. Sullivan v. Ashe, 302 U. S. 51, 55 (1937) (“For the determination of sentences, justice generally requires consideration of more than the particular acts by which the crime was committed and that there be taken into account the circumstances of the offense together with the character and propensities of the offender”). Consistent with this principle, we have observed that “both before and since the American colonies became a nation, courts in this country and in England practiced a policy under which a sentencing judge could exercise a wide discretion in the sources and types of evidence used to assist him in determining the kind and extent of punishment to be imposed within limits fixed by law.” Williams, 337 U. S., at 246. In particular, we have emphasized that “[h]ighly relevant—if not essential—to [the] selection of an appropriate sentence is the possession of the fullest information possible concerning the defendant’s life and characteristics.” Id., at 247. Permitting sentencing courts to consider the widest possible breadth of information about a defendant “ensures that the punishment will suit not merely the offense but the individual defendant.” Wasman v. United States, 468 U. S. 559, 564 (1984). In 1970, Congress codified the “longstanding principle that sentencing courts have broad discretion to consider various kinds of information” at 18 U. S. C. §3577 (1970 ed.). United States v. Watts, 519 U. S. 148, 151 (1997) (per curiam). Section 3577 (1970 ed.) provided: “No limitation shall be placed on the information concerning the background, character, and conduct of a person convicted of an offense which a court of the United States may receive and consider for the purpose of imposing an appropriate sentence.” (Emphasis added.) In the Sentencing Reform Act of 1984 (SRA), 18 U. S. C. §3551 et seq., Congress effected fundamental changes to federal sentencing by creating the Federal Sentencing Commission and introducing the Guidelines scheme. In doing so, however, Congress recodified §3577 without change at §3661. The Sentencing Commission, moreover, expressly incorporated §3661 in the Guidelines: “In determining the sentence to impose within the guideline range, or whether a departure from the guidelines is warranted, the court may consider, without limitation, any information concerning the background, character and conduct of the defendant, unless otherwise prohibited by law. See 18 U. S. C. §3661.” USSG §1B1.4 (2010) (emphasis added). Both Congress and the Sentencing Commission thus expressly preserved the traditional discretion of sentencing courts to “conduct an inquiry broad in scope, largely unlimited either as to the kind of information [they] may consider, or the source from which it may come.” United States v. Tucker, 404 U. S. 443, 446 (1972).[Footnote 8] The SRA did constrain sentencing courts’ discretion in important respects, most notably by making the Guidelines mandatory, see 18 U. S. C. §3553(b)(1) (2000 ed., Supp. IV), and by specifying various factors that courts must consider in exercising their discretion, see §3553(a). In our seminal decision in Booker, we held that where facts found by a judge by a preponderance of the evidence increased the applicable Guidelines range, treating the Guidelines as mandatory in those circumstances violated the Sixth Amendment right of criminal defendants to be tried by a jury and to have every element of an offense proved by the Government beyond a reasonable doubt. 543 U. S., at 243–244. Our remedial opinion in Booker invalidated two offending provisions in the SRA, see id., at 245 (invalidating 18 U. S. C. §§3553(b)(1), 3742(e)), and instructed the district courts to treat the Guidelines as “effectively advisory,” 543 U. S., at 245. Our post-Booker opinions make clear that, although a sentencing court must “give respectful consideration to the Guidelines, Booker permits the court to tailor the sentence in light of other statutory concerns as well.” Kimbrough v. United States, 552 U. S. 85, 101 (2007) (internal quotation marks and citation omitted). Accordingly, although the “Guidelines should be the starting point and the initial benchmark,” district courts may impose sentences within statutory limits based on appropriate consideration of all of the factors listed in §3553(a), subject to appellate review for “reasonableness.” Gall, 552 U. S., at 49–51. This sentencing framework applies both at a defendant’s initial sentencing and at any subsequent resentencing after a sentence has been set aside on appeal. See 18 U. S. C. §3742(g) (“A district court to which a case is remanded … shall resentence a defendant in accordance with section 3553”); see also Dillon v. United States, 560 U. S. ___, ___ (2010) (slip op., at 10) (distinguishing between “sentence-modification proceedings” under 18 U. S. C. §3582(c)(2), which “do not implicate the interests identified in Booker,” and “plenary resentencing proceedings,” which do). B In light of the federal sentencing framework described above, we think it clear that when a defendant’s sentence has been set aside on appeal and his case remanded for resentencing, a district court may consider evidence of a defendant’s rehabilitation since his prior sentencing and that such evidence may, in appropriate cases, support a downward variance from the advisory Guidelines range. Preliminarily, Congress could not have been clearer in directing that “[n]o limitation … be placed on the information concerning the background, character, and conduct” of a defendant that a district court may “receive and consider for the purpose of imposing an appropriate sentence.” 18 U. S. C. §3661. The plain language of §3661 makes no distinction between a defendant’s initial sentencing and a subsequent resentencing after a prior sentence has been set aside on appeal. We have recognized that “the broad language of §3661” does not provide “any basis for the courts to invent a blanket prohibition against considering certain types of evidence at sentencing.” Watts, 519 U. S., at 152. A categorical bar on the consideration of postsentencing rehabilitation evidence would directly contravene Congress’ expressed intent in §3661. In addition, evidence of postsentencing rehabilitation may be highly relevant to several of the §3553(a) factors that Congress has expressly instructed district courts to consider at sentencing. For example, evidence of postsentencing rehabilitation may plainly be relevant to “the history and characteristics of the defendant.” §3553(a)(1). Such evidence may also be pertinent to “the need for the sentence imposed” to serve the general purposes of sentencing set forth in §3553(a)(2)—in particular, to “afford adequate deterrence to criminal conduct,” “protect the public from further crimes of the defendant,” and “provide the defendant with needed educational or vocational training … or other correctional treatment in the most ef-fective manner.” §§3553(a)(2)(B)–(D); see McMannus, 496 F. 3d, at 853 (Melloy, J., concurring) (“In assessing … deterrence, protection of the public, and rehabilitation, 18 U. S. C. §3553(a)(2)(B)(C) & (D), there would seem to be no better evidence than a defendant’s post-incarceration conduct”). Postsentencing rehabilitation may also critically inform a sentencing judge’s overarching duty under §3553(a) to “impose a sentence sufficient, but not greater than necessary” to comply with the sentencing purposes set forth in §3553(a)(2). As the original sentencing judge recognized, the extensive evidence of Pepper’s rehabilitation since his initial sentencing is clearly relevant to the selection of an appropriate sentence in this case. Most fundamentally, evidence of Pepper’s conduct since his release from custody in June 2005 provides the most up-to-date picture of Pepper’s “history and characteristics.” §3553(a)(1); see United States v. Bryson, 229 F. 3d 425, 426 (CA2 2000) (“[A] court’s duty is always to sentence the defendant as he stands before the court on the day of sentencing”). At the time of his initial sentencing in 2004, Pepper was a 25-year-old drug addict who was unemployed, estranged from his family, and had recently sold drugs as part of a methamphetamine conspiracy. By the time of his second resentencing in 2009, Pepper had been drug-free for nearly five years, had attended college and achieved high grades, was a top employee at his job slated for a promotion, had re-established a relationship with his father, and was married and supporting his wife’s daughter. There is no question that this evidence of Pepper’s conduct since his initial sentencing constitutes a critical part of the “history and characteristics” of a defendant that Congress intended sentencing courts to consider. §3553(a). Pepper’s postsentencing conduct also sheds light on the likelihood that he will engage in future criminal conduct, a central factor that district courts must assess when imposing sentence. See §§3553(a)(2)(B)–(C); Gall, 552 U. S., at 59 (“Gall’s self-motivated rehabilitation … lends strong support to the conclusion that imprisonment was not necessary to deter Gall from engaging in future criminal conduct or to protect the public from his future criminal acts” (citing §§3553(a)(2)(B)–(C))). As recognized by Pepper’s probation officer, Pepper’s steady employment, as well as his successful completion of a 500-hour drug treatment program and his drug-free condition, also suggest a diminished need for “educational or vocation- al training … or other correctional treatment.” §3553(a)(2)(D). Finally, Pepper’s exemplary postsentencing conduct may be taken as the most accurate indicator of “his present purposes and tendencies and significantly to suggest the period of restraint and the kind of discipline that ought to be imposed upon him.” Ashe, 302 U. S., at 55. Accordingly, evidence of Pepper’s postsentencing rehabilitation bears directly on the District Court’s overarching duty to “impose a sentence sufficient, but not greater than necessary” to serve the purposes of sentencing. §3553(a). In sum, the Court of Appeals’ ruling prohibiting the District Court from considering any evidence of Pepper’s postsentencing rehabilitation at resentencing conflicts with longstanding principles of federal sentencing law and contravenes Congress’ directives in §§3661 and 3553(a). C Amicus nevertheless advances two principal arguments in defense of the Court of Appeals’ ruling: (1) 18 U. S. C. §3742(g)(2), which restricts the discretion of a resentencing court on remand to impose a non-Guidelines sentence, effectively forecloses consideration of a defendant’s postsentencing rehabilitation; and (2) permitting district courts to consider postsentencing rehabilitation would defeat Congress’ objectives under §3553(a). We are not persuaded. 1 Amicus’ main argument relies on 18 U. S. C. §3742(g)(2), a provision that the Court of Appeals did not cite below. That provision states that when a sentence is set aside on appeal, the district court to which the case is remanded: “shall not impose a sentence outside the applicable guidelines range except upon a ground that— “(A) was specifically and affirmatively included in the written statement of reasons required by section 3553(c) in connection with the previous sentencing of the defendant prior to the appeal; and “(B) was held by the court of appeals, in remanding the case, to be a permissible ground of departure.” In operation, §3742(g)(2) restricts the discretion of a district court on remand by precluding the court from imposing a sentence outside the Guidelines range except upon a “ground of departure” that was expressly relied upon in the prior sentencing and upheld on appeal. Amicus thus correctly contends that, on its face, §3742(g)(2) effectively forecloses a resentencing court from considering evidence of a defendant’s postsentencing rehabilitation for purposes of imposing a non-Guidelines sentence because, as a practical matter, such evidence did not exist at the time of the prior sentencing. As the Government concedes, however, §3742(g)(2) is invalid after Booker. As we have explained, Booker held that where judicial factfinding increases a defendant’s applicable Sentencing Guidelines range, treating the Guidelines as mandatory in those circumstances would violate the defendant’s Sixth Amendment right to be tried by a jury and to have every element of an offense proved by the Government beyond a reasonable doubt. See supra, at 11. We recognized in Booker that, although the SRA permitted departures from the applicable Guidelines range in limited circumstances,[Footnote 9] “departures are not available in every case, and in fact are unavailable in most.” 543 U. S., at 234. Because in those instances, “the judge is bound to impose a sentence within the Guidelines range,” we concluded that the availability of departures in certain circumstances “does not avoid the constitutional issue.” Ibid. To remedy the constitutional problem, we rendered the Guidelines effectively advisory by invalidating two provisions of the SRA: 18 U. S. C. §3553(b)(1) (2000 ed., Supp. IV), which generally required sentencing courts to impose a sentence within the applicable Guidelines range, and §3742(e) (2000 ed. and Supp. IV), which prescribed the standard of appellate review, including de novo review of Guidelines departures. 543 U. S., at 259. We invalidated these provisions even though we recognized that mandatory application of the Guidelines would not always result in a Sixth Amendment violation.[Footnote 10] Indeed, although the Government suggested in Booker that we render the Guidelines advisory only in cases in which the Constitution prohibits judicial factfinding, we rejected that two-track proposal, reasoning that “Congress would not have authorized a mandatory system in some cases and a non-mandatory system in others, given the administrative complexities that such a system would create.” Id., at 266; see Dillon, 560 U. S., at ___ (slip op., at 12) (“The incomplete remedy we rejected in Booker would have required courts to treat the Guidelines differently in similar proceedings, leading potentially to unfair results and considerable administrative challenges”). We did not expressly mention §3742(g)(2) in Booker,[Footnote 11] but the rationale we set forth in that opinion for invalidating §§3553(b)(1) and 3742(e) applies equally to §3742(g)(2). As with those provisions, §3742(g)(2) requires district courts effectively to treat the Guidelines as mandatory in an entire set of cases. Specifically, §3742(g)(2) precludes a district court on remand from imposing a sentence “outside the applicable guidelines range” except upon a “ground of departure” that was expressly relied upon by the court at the prior sentencing and upheld by the court of appeals. In circumstances in which the district court did not rely upon such a departure ground at the prior sentencing, §3742(g)(2) would require the court on remand to impose a sentence within the applicable Guidelines range, thus rendering the Guidelines effectively mandatory. Because in a large set of cases, judicial factfinding will increase the applicable Guidelines range beyond that supported solely by the facts established by the jury verdict (or guilty plea), requiring a sentencing judge on remand to apply the Guidelines range, as §3742(g)(2) does, will often result in a Sixth Amendment violation for the reasons we explained in Booker. Accordingly, as with the provisions in Booker, the proper remedy here is to invalidate §3742(g)(2). The sentencing proceeding at issue in Booker itself illustrates why §3742(g)(2) cannot withstand Sixth Amendment scrutiny. The district court in Booker increased the defendant’s then-mandatory Guidelines range based on a drug-quantity finding that it, rather than the jury, made. 543 U. S., at 227. After we held that the Guidelines must be treated as advisory, we remanded the case for resentencing. Id., at 267. Had §3742(g)(2) remained valid after Booker, the district court on remand would have been required to sentence within the Guidelines range because it did not depart from the Guidelines at the original sentencing. Accordingly, the resentencing judge in Booker would have been required under §3742(g)(2) to impose a Guidelines sentence based on judge-found facts concerning drug quantity, the precise result that Booker forbids. The same result would occur in any sentencing in which a district court erroneously refuses to impose a sentence outside the Guidelines range “based on a misunderstanding of its authority to depart under or vary from the Guidelines.” Reply Brief for United States 16. For example, if §3742(g)(2) remained valid, there would be no remedy at resentencing if a district court erroneously believed the Guidelines were presumptively reasonable, see Nelson v. United States, 555 U. S. ___, ___ (2009) (per curiam) (slip op., at 2), or if it mistakenly thought that a non-Guidelines sentence required extraordinary circumstances, see Gall, 552 U. S., at 47, or if it incorrectly concluded that it could not vary from the Guidelines based on a policy disagreement with their disparate treatment of crack and powder cocaine, see Kimbrough, 552 U. S., at 101. In such cases, the district court at the initial sentencing proceeding will necessarily have imposed a sentence within the Guidelines range, and thus §3742(g)(2) would require the imposition of a Guidelines sentence on remand. See Reply Brief for Petitioner 3–5 (describing further categories of cases where “the Booker remedy would be entirely unavailable if §3742(g)(2) were valid”). To be sure, applying §3742(g)(2) at resentencing would not always result in a Sixth Amendment violation. For example, where the applicable Guidelines range rests solely on facts found by a jury beyond a reasonable doubt, application of §3742(g)(2) at resentencing would not render the sentence constitutionally infirm. But, as explained above, that possibility was equally true with respect to the sentencing provisions we invalidated in Booker. See supra, at 16. As with those provisions, “we cannot assume that Congress, if faced with the statute’s invalidity in key applications, would have preferred to apply the statute in as many other instances as possible.” 543 U. S., at 248. Just as we rejected a two-track system in Booker that would have made the Guidelines mandatory in some cases and advisory in others, we reject a partial invalidation of §3742(g)(2) that would leave us with the same result. The fact that §3742(g)(2) permits a resentencing court on remand to impose a non-Guidelines sentence in cases where the prior sentence expressly relied upon a departure upheld by the court of appeals also does not cure the constitutional infirmity. As explained above, we observed in Booker that the availability of departures from the applicable Guidelines ranges in specified circumstances “does not avoid the constitutional issue.” Id., at 234. Because “departures are not available in every case, and in fact are unavailable in most,” ibid., we held that remedying the Sixth Amendment problem required invalidation of §3553(b)(1). That same remedial approach requires us to invalidate §3742(g)(2).[Footnote 12] Amicus contends that any constitutional infirmity in §3742(g)(2) can be remedied by invalidating §3742(j)(1)(B) rather than §3742(g)(2). Brief for Amicus Curiae in Support of Judgment Below 21–22. Section 3742(j)(1)(B) provides that a “ground of departure” is “permissible” for purposes of §3742(g)(2)(B) only if it is, inter alia, “authorized under section 3553(b).” In Booker, we noted that “statutory cross-references” to the SRA provisions we invalidated were also constitutionally infirm. 543 U. S., at 259. Because §3742(j)(1)(B) incorporates a cross-reference to §3553(b)(1), one of the provisions we invalidated in Booker, amicus suggests that invalidating §3742(j)(1)(B) would cure any constitutional defect in §3742(g)(2)(B). As the Government explains, however, even if §3742(j)(1)(B) were invalidated and a district court could depart on any ground at an initial sentencing, the district court would not be able to depart on any new ground at resentencing so long as §3742(g)(2) remains in force. Because amicus’ proposed solution would still result in the Guidelines being effectively mandatory at resentencing in an entire set of cases, it fails to remedy the fundamental constitutional defect of §3742(g)(2). 2 Amicus’ next cluster of arguments focuses on Con- gress’ sentencing objectives under §3553(a). Preliminarily, amicus contends that even if §3742(g)(2) is constitutionally invalid, that provision reflects a congressional policy determination that only information available at the time of original sentencing should be considered, and that this policy determination should inform our analysis of whether §3553(a) permits consideration of postsentencing rehabilitation evidence. This argument, however, is based on a faulty premise. Contrary to amicus’ contention, §3742(g)(2) does not reflect a congressional purpose to preclude consideration of evidence of postsentencing rehabilitation at resentencing. To be sure, §3742(g)(2) has the incidental effect of limiting the weight a sentencing court may place on postsentencing rehabilitation by precluding the court from resentencing outside the Guidelines range on a “ground of departure” on which it did not previously rely. But on its face, nothing in §3742(g)(2) prohibits a district court from considering postsentencing developments—including postsentencing rehabilitation—in selecting a sentence within the applicable Guidelines range. Section 3742(g)(2) also does not apply to resentencings that occur for reasons other than when a sentence is overturned on appeal and the case is remanded (e.g., when a sentence is set aside on collateral review under 28 U. S. C. §2255). In such circumstances, §3742(g)(2) does not restrict a district court at all, much less with respect to consideration of postsentencing developments. Accordingly, because we see no general congressional policy reflected in §3742(g)(2) to preclude resentencing courts from considering postsentencing information,[Footnote 13] that provision has no bearing on our analysis of whether §3553(a) permits consideration of evidence of postsentencing rehabilitation. As we explained above, evidence of postsentencing rehabilitation may be highly relevant to several of the sentencing factors that Congress has specifically instructed district courts to consider. See supra, at 13–15 (discussing §§3553(a), (a)(1), (a)(2)(B)–(D)). Amicus, however, argues that consideration of postsentencing reha-bilitation is inconsistent with two sentencing factors: §3553(a)(5), which directs sentencing courts to consider “any pertinent policy statement” of the Sentencing Commission, and §3553(a)(6), which requires courts to consider “the need to avoid unwarranted sentenc[ing] disparities among defendants with similar records who have been found guilty of similar conduct.” With regard to §3553(a)(5), amicus points to the Sentencing Commission’s policy statement in USSG §5K2.19, which provides that “[p]ost-sentencing rehabilitative efforts, even if exceptional, undertaken by a defendant after imposition of a term of imprisonment for the instant offense[,] are not an appropriate basis for a downward departure when resentencing the defendant for that offense.” According to amicus, that policy statement is “clear and unequivocal,” and as an exercise of the Sentencing Commission’s “core function,” should be given effect. Brief for Amicus Curiae in Support of Judgment Below 31–32. To be sure, we have recognized that the Commission post-Booker continues to “fil[l] an important institutional role” because “[i]t has the capacity courts lack to base its determinations on empirical data and national experience, guided by a professional staff with appropriate expertise.” Kimbrough, 552 U. S., at 109 (internal quotation marks omitted). Accordingly, we have instructed that district courts must still give “respectful consideration” to the now-advisory Guidelines (and their accompanying policy statements). Id., at 101. As amicus acknowledges, however, our post-Booker decisions make clear that a district court may in appropriate cases impose a non-Guidelines sentence based on a disagreement with the Commission’s views. See id., at 109–110. That is particularly true where, as here, the Commission’s views rest on wholly unconvincing policy rationales not reflected in the sentencing statutes Congress enacted. The commentary to USSG §5K2.19 expresses the Commission’s view that departures based on postsentencing rehabilitation would “(1) be inconsistent with the policies established by Congress under 18 U. S. C. §3624(b) [governing good time credit] and other statutory provisions for reducing the time to be served by an imprisoned person; and (2) inequitably benefit only those who gain the opportunity to be resentenced de novo.” With regard to the first proffered rationale, a sentencing reduction based on postsentencing rehabilitation can hardly be said to be “inconsistent with the policies” underlying an award of good time credit under §3624(b) because the two serve distinctly different penological interests.[Footnote 14] Indeed, the difference between the two is reflected most obviously in the fact that the BOP has no authority to award good time credit where, as in this case, the defendant’s good behavior occurs after a sentence has already been served.[Footnote 15] The Commission’s second proffered rationale fares no better. To be sure, allowing district courts to consider evidence of postsentencing rehabilitation may result in disparate treatment between those defendants who are sentenced properly and those who must be resentenced. But that disparity arises not because of arbitrary or random sentencing practices, but because of the ordinary operation of appellate sentencing review. In a closely related vein, amicus argues that consideration of postsentencing rehabilitation is inconsistent with §3553(a)(6), which requires sentencing courts to consider the need to avoid unwarranted sentencing disparities. The Court of Appeals also rested its holding on this ground, reasoning that “ ‘allowing [postsentencing rehabilitation] evidence to influence [defendant’s] sentence would be grossly unfair to the vast majority of defendants who receive no sentencing-court review of any positive post-sentencing rehabilitative efforts.’ ” 570 F. 3d, at 965 (quoting McMannus, 496 F. 3d, at 852, n. 4). But amicus points to no evidence, nor are we aware of any, suggesting that Congress enacted §3553(a)(6) out of a concern with disparities resulting from the normal trial and sentencing process.[Footnote 16] The differences in procedural opportunity that may result because some defendants are inevitably sentenced in error and must be resentenced are not the kinds of “unwarranted” sentencing disparities that Congress sought to eliminate under §3553(a)(6). Cf. United States v. LaBonte, 520 U. S. 751, 761–762 (1997) (disparity arising from exercise of prosecutorial discretion not unwarranted); United States v. Rhodes, 145 F. 3d 1375, 1381 (CADC 1998) (“Distinguishing between prisoners whose convictions are reversed on appeal and all other prisoners hardly seems ‘unwarranted’ ”). As the Government explains, moreover, the logic of the Court of Appeals’ approach below—i.e., that “post-sentence rehabilitation is not relevant . . . because the district court could not have considered that evidence at the time of the original sentencing,” 570 F. 3d, at 965 (internal quotation marks omitted)—would require sentencing courts categorically to ignore not only postsentencing rehabilitation, but any postsentencing information, including, for example, evidence that a defendant had committed postsentencing offenses. Our precedents, however, provide no basis to support such a categorical bar. See, e.g., Wasman, 468 U. S., at 572 (“[A] sentencing authority may justify an increased sentence by affirmatively identifying relevant conduct or events that occurred subsequent to the original sentencing proceedings”); cf. North Carolina v. Pearce, 395 U. S. 711, 723 (1969). Indeed, even the Court of Appeals below does not accept the logical consequence of its approach as it permits district courts to consider postsentencing conduct that would support a higher sentence. See United States v. Stapleton, 316 F. 3d 754, 757 (CA8 2003). Nothing in §§3553(a) and 3661, however, remotely suggests that Congress intended district courts to consider only postsentencing evidence detrimental to a defendant while turning a blind eye to favorable evidence of a defendant’s postsentencing rehabilitation. Cf. United States v. Jones, 460 F. 3d 191, 196 (CA2 2006) (“Obviously, the discretion that Booker accords sentencing judges to impose non-Guidelines sentences cannot be an escalator that only goes up”). Finally, we note that §§3553(a)(5) and (a)(6) describe only two of the seven sentencing factors that courts must consider in imposing sentence. At root, amicus effectively invites us to elevate two §3553(a) factors above all others. We reject that invitation. See Gall, 552 U. S., at 49–50 (instructing sentencing courts to “consider all of the §3553(a) factors” (emphasis added)). D For the reasons stated above, we hold that the Court of Appeals erred in categorically precluding the District Court from considering evidence of Pepper’s postsentencing rehabilitation after his initial sentence was set aside on appeal. District courts post-Booker may consider evidence of a defendant’s postsentencing rehabilitation at resentencing and such evidence may, in appropriate cases, support a downward variance from the advisory Guidelines range.[Footnote 17] The Government informs us that, in granting Pepper’s motion for release pending disposition of this appeal, see n. 5, supra, the District Court stated that it would not have exercised its discretion to grant Pepper a downward variance based on postsentencing rehabilitation. That statement, however, was made in light of the Court of Appeals’ erroneous views regarding postsentencing rehabilitation evidence. Because we expressly reject those views today, it is unclear from the record whether the District Court would have imposed the same sentence had it properly considered the extensive evidence of Pepper’s postsentencing rehabilitation. On remand, the District Court should consider and give appropriate weight to that evidence, as well as any additional evidence concerning Pepper’s conduct since his last sentencing in January 2009. Accordingly, we vacate the Eighth Circuit’s judgment in respect to Pepper’s sentence and remand the case for resentencing consistent with this opinion. III The second question presented in this case merits only a brief discussion. As noted above, the original sentencing judge in this case granted Pepper a 40 percent downward departure pursuant to USSG §5K1.1 based on Pepper’s substantial assistance and sentenced him to 24 months’ imprisonment. When the Court of Appeals vacated that sentence in Pepper II, and again in Pepper III, the case was reassigned on remand to Chief Judge Reade. In resentencing Pepper, Chief Judge Reade ruled that she was not bound by the prior sentencing judge’s decision to grant a 40 percent downward departure and instead granted only a 20 percent downward departure, which the Court of Appeals upheld in Pepper IV. Pepper argues that the law of the case doctrine required Chief Judge Reade to apply the same 40 percent departure granted by the original sentencing judge. We disagree. Preliminarily, we note that the mandates in Pepper II and Pepper III were “general remand[s] for resentencing,” which “did not place any limitations on the discretion of the newly assigned district court judge in resentencing Pepper.” 570 F. 3d, at 963. In his merits briefs to this Court, Pepper does not challenge the scope or validity of the Court of Appeals’ mandate ordering de novo resentencing, and thus has abandoned any argument that the mandate itself restricted the District Court from imposing a different substantial assistance departure.[Footnote 18] The only question before us is whether the law of the case doctrine required Chief Judge Reade to adhere to the original sentencing judge’s decision granting a 40 percent downward departure. Although we have described the “law of the case [a]s an amorphous concept,” “[a]s most commonly defined, the doctrine posits that when a court decides upon a rule of law, that decision should continue to govern the same issues in subsequent stages in the same case.” Arizona v. California, 460 U. S. 605, 618 (1983). This doctrine “directs a court’s discretion, it does not limit the tribunal’s power.” Ibid. Accordingly, the doctrine “does not apply if the court is ‘convinced that [its prior decision] is clearly erroneous and would work a manifest injustice.’ ” Agostini v. Felton, 521 U. S. 203, 236 (1997) (quoting Arizona, 460 U. S., at 618, n. 8; alteration in original). Pepper argues that, because the original sentencing judge’s decision to grant the 40 percent departure was never set aside by the Court of Appeals or this Court, it constituted the law of the case. As such, Pepper contends that Chief Judge Reade should not have disturbed that ruling absent “compelling justification” for overturning it. Brief for Petitioner 56. According to Pepper, because Chief Judge Reade identified no such justification, the law of the case doctrine required her to adhere to the 40 percent departure granted by the original sentencing judge. As the Government explains, however, the Court of Appeals in Pepper III set aside Pepper’s entire sentence and remanded for a de novo resentencing. See 518 F. 3d, at 949, 953. Thus, even assuming, arguendo, that the original sentencing court’s decision to impose a 40 percent departure was at one point law of the case, Pepper III effectively wiped the slate clean. To be sure, Pepper III vacated Pepper’s 24-month sentence on grounds unrelated to the substantial assistance departure, but that fact does not affect our conclusion. “A criminal sentence is a package of sanctions that the district court utilizes to effectuate its sentencing intent.” United States v. Stinson, 97 F. 3d 466, 469 (CA11 1996) (per curiam). Because a district court’s “original sentencing intent may be undermined by altering one portion of the calculus,” United States v. White, 406 F. 3d 827, 832 (CA7 2005), an appellate court when reversing one part of a defendant’s sentence “may vacate the entire sentence … so that, on remand, the trial court can reconfigure the sentencing plan … to satisfy the sentencing factors in 18 U. S. C. §3553(a),” Greenlaw v. United States, 554 U. S. 237, 253 (2008). That is precisely what the Eighth Circuit did here. Accordingly, because the Court of Appeals in Pepper III remanded for de novo resentencing, we conclude that Chief Judge Reade was not bound by the law of the case doctrine to apply the same 40 percent departure that had been applied at Pepper’s prior sentencing. * * * For the reasons stated above, the judgment of the United States Court of Appeals for the Eighth Circuit is vacated in part and affirmed in part, and the case is remanded for resentencing consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 Although the charge to which Pepper pleaded guilty carried a mandatory minimum of 120 months’ imprisonment, the mandatory minimum did not apply because he was eligible for safety-valve relief pursuant to 18 U. S. C. §3553(f) (2000 ed.) and §5C1.2 of the United States Sentencing Guidelines Manual (Nov. 2003) (USSG). Footnote 2 USSG §5K1.1 provides that a court may depart from the Guidelines “[u]pon motion of the government stating that the defendant has provided substantial assistance in the investigation or prosecution of another person who has committed an offense.” Pepper provided information to Government investigators and a grand jury concerning two other individuals involved with illegal drugs and guns. Footnote 3 The court also cited Pepper’s lack of a violent history and, to a lesser extent, the need to avoid unwarranted sentencing disparity with Pepper’s co-conspirators. App. 144–145. Footnote 4 The Court of Appeals also held that the District Court “further erred by considering Pepper’s lack of violent history, which history had already been accounted for in the sentencing Guidelines calculation, and by considering sentencing disparity among Pepper’s co-defendants without adequate foundation and explanation.” Pepper II, 486 F. 3d, at 413. Footnote 5 After the District Court resentenced Pepper to 65 months’ imprisonment, Pepper was returned to federal custody. On July 22, 2010, after we granted Pepper’s petition for a writ of certiorari, the District Court granted his motion for release pending disposition of the case here. Footnote 6 Compare, e.g., United States v. Lorenzo, 471 F. 3d 1219, 1221 (CA11 2006) (per curiam) (precluding consideration of postsentencing rehabilitative conduct); United States v. Sims, 174 F. 3d 911, 913 (CA8 1999) (same), with United States v. Lloyd, 469 F. 3d 319, 325 (CA3 2006) (permitting consideration of postsentencing rehabilitation in exceptional cases); United States v. Hughes, 401 F. 3d 540, 560, n. 19 (CA4 2005) (instructing district court to adjust Guidelines calculation on remand “if new circumstances have arisen or events occurred since [defendant] was sentenced that impact the range prescribed by the guidelines”). Footnote 7 We appointed Adam G. Ciongoli to brief and argue the case, as amicus curiae, in support of the Court of Appeals’ judgment. 561 U. S. ___ (2010). Mr. Ciongoli has ably discharged his assigned responsibilities. Footnote 8 Of course, sentencing courts’ discretion under §3661 is subject to constitutional constraints. See, e.g., United States v. Leung, 40 F. 3d 577, 586 (CA2 1994) (“A defendant’s race or nationality may play no adverse role in the administration of justice, including at sentencing”). Footnote 9 See 18 U. S. C. §3553(b)(1) (2000 ed., Supp. IV) (permitting departures where the judge “finds that there exists an aggravating or mitigating circumstance of a kind, or to a degree, not adequately taken into consideration by the Sentencing Commission”). Footnote 10 For example, in the pre-Booker regime, if the applicable Guidelines range depended solely on facts found by a jury beyond a reasonable doubt, requiring a judge to sentence within that range would not run afoul of the Sixth Amendment. Footnote 11 See Dillon, 560 U. S., at ___, n. 5 (Stevens, J., dissenting) (slip op., at 9, n. 5) (citing §3742(g)(2) as “one additional provision of the [SRA that] should have been excised, but was not, in order to accomplish the Court’s remedy”). Footnote 12 Amicus National Association of Criminal Defense Lawyers (NACDL) argues that, because §3742(g)(2)(B) permits a non-Guidelines sentence only with respect to certain “departures,” that provision “appears to preclude sentencing courts on remand from granting any and all variances under Section 3553(a).” Brief for NACDL as Amicus Curiae 11 (emphasis added). In Irizarry v. United States, 553 U. S. 708 (2008), we held that a “ ‘[d]eparture’ is a term of art under the Guidelines and refers only to non-Guidelines sentences imposed under the framework set out in the Guidelines”; in contrast, a “variance” refers to a non-Guidelines sentence outside the Guidelines framework. Id., at 714. Irizarry’s holding construed the term “departure” in Rule 32(h) of the Federal Rules of Criminal Procedure. Because we conclude that §3742(g)(2) is constitutionally infirm and must be invalidated, we need not decide whether its reference to “departure[s]” includes variances. Footnote 13 For those of us for whom it is relevant, the legislative history of §3742(g)(2) confirms that the provision, enacted as part of the PROTECT Act of 2003, §401(e), 117 Stat. 671, was not aimed at prohibiting district courts from considering postsentencing developments. Rather, it was meant to ensure that under the then-mandatory Guidelines system, when a particular departure was reversed on appeal, the district court could not impose the same sentence on remand on the basis of a different departure. See H. R. Conf. Rep. No. 108–66, pp. 58–59 (2003) (noting that §401 of the PROTECT Act, inter alia, “prevent[s] sentencing courts, upon remand, from imposing the same illegal departure on a different theory”). Like the provisions invalidated in Booker, then, the purpose of §3742(g)(2) was “to make Guidelines sentencing even more mandatory than it had been.” 543 U. S. 220, 261 (2005). As we recognized in Booker, that purpose has “ceased to be relevant.” Ibid. Footnote 14 An award of good time credit by the Bureau of Prisons (BOP) does not affect the length of a court-imposed sentence; rather, it is an administrative reward “to provide an incentive for prisoners to ‘compl[y] with institutional disciplinary regulations.’ ” Barber v. Thomas, 560 U. S. ___, ___ (2010) (slip op., at 7) (quoting 18 U. S. C. §3624(b); alteration in original). Such credits may be revoked at any time before the date of a prisoner’s release. See §3624(b)(2). In contrast, a court’s imposition of a reduced sentence based on postsentencing rehabilitation changes the very terms of imprisonment and “recognizes that the [defendant’s] conduct since his initial sentencing warrants a less severe criminal punishment.” Brief for United States 50. Once imposed, a sentence may be modified only in very limited circumstances. See §3582(c). Footnote 15 Amicus points to two other procedural mechanisms that may shorten a defendant’s sentence—early termination of a term of supervised release, see §3583(e)(1), and the potential for sentencing reductions based on postsentencing substantial assistance, see Fed. Rule Crim. Proc. 35(b)—but neither presents an adequate substitute for a district court’s consideration of postsentencing rehabilitation. Supervised release follows a term of imprisonment and serves an entirely different purpose than the sentence imposed under §3553(a). See United States v. Johnson, 529 U. S. 53, 59 (2000) (“Supervised release fulfills rehabilitative ends, distinct from those served by incarceration”). Rule 35(b) departures address only postsentencing cooperation with the Government, not postsentencing rehabilitation generally, and thus a defendant with nothing to offer the Government can gain no benefit from Rule 35(b). Footnote 16 Indeed, some defendants will have a longer period of time between initial custody and trial, or between trial and sentencing, and those defendants—particularly if they are released on bail—will have a greater opportunity to demonstrate postoffense, presentencing rehabilitation. Even before Booker, the lower courts uniformly held that evidence of such rehabilitation could provide a basis for departing from the applicable Guidelines. See USSG App. C, Amdt. 602, comment., p. 74 (Nov. 2003) (“[D]epartures based on extraordinary post-offense rehabilitative efforts prior to sentencing … have been allowed by every circuit that has ruled on the matter”). Footnote 17 Of course, we do not mean to imply that a district court must reduce a defendant’s sentence upon any showing of postsentencing rehabilitation. Nor do we mean to preclude courts of appeals from issuing limited remand orders, in appropriate cases, that may render evidence of postsentencing rehabilitation irrelevant in light of the narrow purposes of the remand proceeding. See, e.g., United States v. Bernardo Sanchez, 569 F. 3d 995, 1000 (CA9 2009). Footnote 18 In any event, as the Court of Appeals recognized, neither Pepper II nor Pepper III held that a 40 percent downward departure was the only reasonable departure that a sentencing court could grant for Pepper’s substantial assistance; rather, the only issue those opinions actually decided was that a “40% downward departure was not an abuse of discretion.” 570 F. 3d, at 963–964.

564.US.2010_09-993

Five years after the Food and Drug Administration (FDA) first approved metoclopramide, a drug commonly used to treat digestive tract problems, under the brand name Reglan, generic manufacturers such as petitioners also began producing the drug. Because of accumulating evidence that long-term metoclopramide use can cause tardive dyskinesia, a severe neurological disorder, warning labels for the drug have been strengthened and clarified several times, most recently in 2009. Respondents were prescribed Reglan in 2001 and 2002, but both received the generic drug from their pharmacists. After taking the drug as prescribed for several years, both developed tardive dyskinesia. In separate state-court tort actions, they sued petitioners, the generic drug manufacturers that produced the metoclopramide they took (Manufacturers). Each respondent alleged, inter alia, that long-term metoclopramide use caused her disorder and that the Manufacturers were liable under state tort law for failing to provide adequate warning labels. In both suits, the Manufacturers urged that federal statutes and FDA regulations pre-empted the state tort claims by requiring the same safety and efficacy labeling for generic metoclopramide as was mandated at the time for Reglan. The Fifth and Eighth Circuits rejected these arguments, holding that respondents’ claims were not pre-empted. Held: The judgment is reversed, and the cases are remanded. 588 F. 3d 603 and 593 F. 3d 428, reversed and remanded. Justice Thomas delivered the opinion of the Court with respect to all but Part III–B–2, concluding that federal drug regulations applicable to generic drug manufacturers directly conflict with, and thus pre-empt, these state claims. Pp. 4–14, 17–20. (a) Because pre-emption analysis requires a comparison between federal and state law, the Court begins by identifying the state tort duties and federal labeling requirements applicable to the Manufacturers. Pp. 4–10. (1) State tort law requires a manufacturer that is, or should be, aware of its drug’s danger to label it in a way that renders it reasonably safe. Respondents pleaded that the Manufacturers knew, or should have known, both that the long-term use of their products carried a high risk of tardive dyskinesia and that their labels did not adequately warn of that risk. Taking these allegations as true, the state-law duty required the Manufacturers to use a different, stronger label than the one they actually used. Pp. 4–5. (2) On the other hand, federal drug regulations, as interpreted by the FDA, prevented the Manufacturers from independently changing their generic drugs’ safety labels. A manufacturer seeking federal approval to market a new drug must prove that it is safe and effective and that the proposed label is accurate and adequate. Although the same rules originally applied to all drugs, the 1984 law commonly called the Hatch-Waxman Amendments allows a generic drug manufacturer to gain FDA approval simply by showing that its drug is equivalent to an already-approved brand-name drug, and that the safety and efficacy labeling proposed for its drug is the same as that approved for the brand-name drug. Respondents contend that federal law nevertheless provides avenues through which the Manufacturers could have altered their metoclopramide labels in time to prevent the injuries here. These include: (1) the FDA’s “changes-being-effected” (CBE) process, which permits drug manufacturers, without preapproval, to add or strengthen a warning label; and (2) sending “Dear Doctor” letters providing additional warnings to prescribing physicians and other healthcare professionals. However, the FDA denies that the Manufacturers could have used either of these processes to unilaterally strengthen their warning labels. The Court defers to the FDA’s views because they are not plainly erroneous or inconsistent with the regulations, and there is no other reason to doubt that they reflect the FDA’s fair and considered judgment. Auer v. Robbins, 519 U. S. 452, 461, 462. Assuming, without deciding, that the FDA is correct that federal law nevertheless required the Manufacturers to ask for the agency’s assistance in convincing the brand-name manufacturer to adopt a stronger label, the Court turns to the pre-emption question. Pp. 5–10. (b) Where state and federal law directly conflict, state law must give way. See, e.g., Wyeth v. Levine, 555 U. S. 555, 583. Such a conflict exists where it is “impossible for a private party to comply with both state and federal requirements.” Freightliner Corp. v. Myrick, 514 U. S. 280, 287. Pp. 11–14, 17–20. (1) The Court finds impossibility here. If the Manufacturers had independently changed their labels to satisfy their state-law duty to attach a safer label to their generic metoclopramide, they would have violated the federal requirement that generic drug labels be the same as the corresponding brand-name drug labels. Thus, it was impossible for them to comply with both state and federal law. And even if they had fulfilled their federal duty to ask for FDA help in strengthening the corresponding brand-name label, assuming such a duty exists, they would not have satisfied their state tort-law duty. State law demanded a safer label; it did not require communication with the FDA about the possibility of a safer label. Pp. 11–12. (2) The Court rejects the argument that the Manufacturers’ pre-emption defense fails because they failed to ask the FDA for help in changing the corresponding brand-name label. The proper question for “impossibility” analysis is whether the private party could independently do under federal law what state law requires of it. See Wyeth, supra, at 573. Accepting respondents’ argument would render conflict pre-emption largely meaningless by making most conflicts between state and federal law illusory. In these cases, it is possible that, had the Manufacturers asked the FDA for help, they might have eventually been able to strengthen their warning label. But it is also possible that they could have convinced the FDA to reinterpret its regulations in a manner that would have opened the CBE process to them, persuaded the FDA to rewrite its generic drug regulations entirely, or talked Congress into amending the Hatch-Waxman Amendments. If these conjectures sufficed to prevent federal and state law from conflicting, it is unclear when, outside of express pre-emption, the Supremacy Clause would have any force. That Clause—which makes federal law “the supreme Law of the Land … any Thing in the Constitution or Laws of any State to the Contrary notwithstanding,” U. S. Const., Art. VI, cl. 2—cannot be read to permit an approach to pre-emption that renders conflict pre-emption all but meaningless. Here, it is enough to hold that when a party cannot satisfy its state duties without the Federal Government’s special permission and assistance, which is dependent on the exercise of judgment by a federal agency, that party cannot independently satisfy those state duties for pre-emption purposes. Pp. 12–14, 17. (3) Wyeth is not to the contrary. The Court there held that a state tort action against a brand-name drug manufacturer for failure to provide an adequate warning label was not pre-empted because it was possible for the manufacturer to comply with both state and federal law under the FDA’s CBE regulation. 555 U. S., at 572–573. The federal statutes and regulations that apply to brand-name drug manufacturers differ, by Congress’ design, from those applicable to generic drug manufacturers. And different federal statutes and regulations may, as here, lead to different pre-emption results. This Court will not distort the Supremacy Clause in order to create similar pre-emption across a dissimilar statutory scheme. Congress and the FDA retain authority to change the law and regulations if they so desire. Pp. 17–20. Thomas, J., delivered the opinion of the Court, except as to Part III–B–2. Roberts, C. J., and Scalia and Alito, JJ., joined that opinion in full, and Kennedy, J., joined as to all but Part III–B–2. Sotomayor, J., filed a dissenting opinion, in which Ginsburg, Breyer, and Kagan, JJ., joined. Together with No. 09–1039, Actavis Elizabeth, LLC v. Mensing, also on certiorari to the same court, and No. 09–1501, Actavis, Inc. v. Demahy, on certiorari to the United States Court of Appeals for the Fifth Circuit.

ex-cept as to Part III–B–2.* These consolidated lawsuits involve state tort-law claims based on certain drug manufacturers’ alleged failure to provide adequate warning labels for generic metoclopramide. The question presented is whether federal drug regulations applicable to generic drug manufacturers directly conflict with, and thus pre-empt, these state-law claims. We hold that they do. I Metoclopramide is a drug designed to speed the movement of food through the digestive system. The Food and Drug Administration (FDA) first approved metoclopramide tablets, under the brand name Reglan, in 1980. Five years later, generic manufacturers also began producing meto-clopramide. The drug is commonly used to treat diges- tive tract problems such as diabetic gastroparesis and gastroesophageal reflux disorder. Evidence has accumulated that long-term metoclopramide use can cause tardive dyskinesia, a severe neurological disorder. Studies have shown that up to 29% of patients who take metoclopramide for several years develop this condition. McNeil v. Wyeth, 462 F. 3d 364, 370, n. 5 (CA5 2006); see also Shaffer, Butterfield, Pamer, & Mackey, Tardive Dyskinesia Risks and Metoclopramide Use Before and After U. S. Market Withdrawal of Cisapride, 44 J. Am. Pharmacists Assn. 661, 663 (2004) (noting 87 cases of metoclopramide-related tardive dyskinesia reported to the FDA’s adverse event reporting system by mid-2003). Accordingly, warning labels for the drug have been strengthened and clarified several times. In 1985, the label was modified to warn that “tardive dyskinesia … may develop in patients treated with metoclopramide,” and the drug’s package insert added that “[t]herapy longer than 12 weeks has not been evaluated and cannot be recommended.” Physician’s Desk Reference 1635–1636 (41st ed. 1987); see also Brief for Petitioner PLIVA et al. 21–22 (hereinafter PLIVA Brief). In 2004, the brand-name Reglan manufacturer requested, and the FDA approved, a label change to add that “[t]herapy should not exceed 12 weeks in duration.” Brief for United States as Amicus Curiae 8 (hereinafter U. S. Brief). And in 2009, the FDA ordered a black box warning—its strongest—which states: “Treatment with metoclopramide can cause tardive dyskinesia, a serious movement disorder that is often irreversible… . Treatment with metoclopramide for longer than 12 weeks should be avoided in all but rare cases.” See Physician’s Desk Reference 2902 (65th ed. 2011). Gladys Mensing and Julie Demahy, the plaintiffs in these consolidated cases, were prescribed Reglan in 2001 and 2002, respectively. Both received generic metoclopramide from their pharmacists. After taking the drug as prescribed for several years, both women developed tardive dyskinesia. In separate suits, Mensing and Demahy sued the generic drug manufacturers that produced the metoclopramide they took (Manufacturers). Each alleged, as relevant here, that long-term metoclopramide use caused her tar-dive dyskinesia and that the Manufacturers were liable under state tort law (specifically, that of Minnesota and Louisiana) for failing to provide adequate warning labels. They claimed that “despite mounting evidence that long term metoclopramide use carries a risk of tardive dyskinesia far greater than that indicated on the label,” none of the Manufacturers had changed their labels to adequately warn of that danger. Mensing v. Wyeth, Inc., 588 F. 3d 603, 605 (CA8 2009); see also Demahy v. Actavis, Inc., 593 F. 3d 428, 430 (CA5 2010). In both suits, the Manufacturers urged that federal law pre-empted the state tort claims. According to the Manufacturers, federal statutes and FDA regulations required them to use the same safety and efficacy labeling as their brand-name counterparts. This means, they argued, that it was impossible to simultaneously comply with both federal law and any state tort-law duty that required them to use a different label. The Courts of Appeals for the Fifth and Eighth Circuits rejected the Manufacturers’ arguments and held that Men-sing and Demahy’s claims were not pre-empted. See 588 F. 3d, at 614; 593 F. 3d, at 449. We granted certiorari, 562 U. S. ___ (2010), consolidated the cases, and now reverse each. II Pre-emption analysis requires us to compare federal and state law. We therefore begin by identifying the state tort duties and federal labeling requirements applicable to the Manufacturers. A It is undisputed that Minnesota and Louisiana tort law require a drug manufacturer that is or should be aware of its product’s danger to label that product in a way that renders it reasonably safe. Under Minnesota law, which applies to Mensing’s lawsuit, “where the manufacturer … of a product has actual or constructive knowledge of danger to users, the … manufacturer has a duty to give warning of such dangers.” Frey v. Montgomery Ward & Co., 258 N. W. 2d 782, 788 (Minn. 1977). Similarly, under Louisiana law applicable to Demahy’s lawsuit, “a manufacturer’s duty to warn includes a duty to provide adequate instructions for safe use of a product.” Stahl v. Novartis Pharmaceuticals Corp., 283 F. 3d 254, 269–270 (CA5 2002); see also La. Rev. Stat. Ann. §9:2800.57 (West 2009). In both States, a duty to warn falls specifically on the manufacturer. See Marks v. OHMEDA, Inc., 2003–1446, pp. 8–9 (La. App. 3/31/04), 871 So. 2d 1148, 1155; Gray v. Badger Min. Corp., 676 N. W. 2d 268, 274 (Minn. 2004). Mensing and Demahy have pleaded that the Manufacturers knew or should have known of the high risk of tardive dyskinesia inherent in the long-term use of their product. They have also pleaded that the Manufacturers knew or should have known that their labels did not adequately warn of that risk. App. 437–438, 67–69, 94–96. The parties do not dispute that, if these allegations are true, state law required the Manufacturers to use a different, safer label. B Federal law imposes far more complex drug labeling requirements. We begin with what is not in dispute. Under the 1962 Drug Amendments to the Federal Food, Drug, and Cosmetic Act, 76 Stat. 780, 21 U. S. C. §301 et seq., a manufacturer seeking federal approval to market a new drug must prove that it is safe and effective and that the proposed label is accurate and adequate.[Footnote 1] See, e.g., 21 U. S. C. §§355(b)(1), (d); Wyeth v. Levine, 555 U. S. 555, 567 (2009). Meeting those requirements involves costly and lengthy clinical testing. §§355(b)(1)(A), (d); see also D. Beers, Generic and Innovator Drugs: A Guide to FDA Approval Requirements §2.02[A] (7th ed. 2008). Originally, the same rules applied to all drugs. In 1984, however, Congress passed the Drug Price Competition and Patent Term Restoration Act, 98 Stat. 1585, commonly called the Hatch-Waxman Amendments. Under this law, “generic drugs” can gain FDA approval simply by showing equivalence to a reference listed drug that has already been approved by the FDA.[Footnote 2] 21 U. S. C. §355(j)(2)(A). This allows manufacturers to develop generic drugs in-expensively, without duplicating the clinical trials already performed on the equivalent brand-name drug. A generic drug application must also “show that the [safety and efficacy] labeling proposed … is the same as the labeling approved for the [brand-name] drug.” §355(j)(2)(A)(v); see also §355(j)(4)(G); Beers §§3.01, 3.03[A]. As a result, brand-name and generic drug manufacturers have different federal drug labeling duties. A brand-name manufacturer seeking new drug approval is responsible for the accuracy and adequacy of its label. See, e.g., 21 U. S. C. §§355(b)(1), (d); Wyeth, supra, at 570–571. A manufacturer seeking generic drug approval, on the other hand, is responsible for ensuring that its warning label is the same as the brand name’s. See, e.g., §355(j)(2)(A)(v); §355(j)(4)(G); 21 CFR §§314.94(a)(8), 314.127(a)(7). The parties do not disagree. What is in dispute is whether, and to what extent, generic manufacturers may change their labels after initial FDA approval. Mensing and Demahy contend that federal law provided several avenues through which the Manufacturers could have altered their metoclopramide labels in time to prevent the injuries here. The FDA, however, tells us that it interprets its regulations to require that the warning labels of a brand-name drug and its generic copy must always be the same—thus, generic drug manufacturers have an ongoing federal duty of “sameness.” U. S. Brief 16; see also 57 Fed. Reg. 17961 (1992) (“[T]he [generic drug’s] labeling must be the same as the listed drug product’s labeling because the listed drug product is the basis for [generic drug] ap-proval”). The FDA’s views are “controlling unless plainly erroneous or inconsistent with the regulation[s]” or there is any other reason to doubt that they reflect the FDA’s fair and considered judgment. Auer v. Robbins, 519 U. S. 452, 461, 462 (1997) (internal quotation marks omitted).[Footnote 3] 1 First, Mensing and Demahy urge that the FDA’s “changes-being-effected” (CBE) process allowed the Manufacturers to change their labels when necessary. See Brief for Respondents 33–35; see also 593 F. 3d, at 439–444; Gaeta v. Perrigo Pharmaceuticals Co., 630 F. 3d 1225, 1231 (CA9 2011); Foster v. American Home Prods. Corp., 29 F. 3d 165, 170 (CA4 1994). The CBE process permits drug manufacturers to “add or strengthen a contraindication, warning, [or] precaution,” 21 CFR §314.70(c)(6)(iii)(A) (2006), or to “add or strengthen an instruction about dosage and administration that is intended to increase the safe use of the drug product,” §314.70(c)(6)(iii)(C). When making labeling changes using the CBE process, drug man-ufacturers need not wait for preapproval by the FDA, which ordinarily is necessary to change a label. Wyeth, supra, at 568. They need only simultaneously file a supplemental application with the FDA. 21 CFR §314.70(c)(6). The FDA denies that the Manufacturers could have used the CBE process to unilaterally strengthen their warning labels. The agency interprets the CBE regulation to allow changes to generic drug labels only when a generic drug manufacturer changes its label to match an updated brand-name label or to follow the FDA’s instructions. U. S. Brief 15, 16, n. 7 (interpreting 21 CFR §314.94(a)(8)(iv)); U. S. Brief 16, n. 8. The FDA argues that CBE changes unilaterally made to strengthen a generic drug’s warning label would violate the statutes and regulations requiring a generic drug’s label to match its brand-name counterpart’s. Id., at 15–16; see also 21 U. S. C. §355(j)(4)(G); 21 CFR §§314.94(a)(8)(iii), 314.150(b)(10) (approval may be withdrawn if the generic drug’s label “is no longer consistent with that for [the brand-name]”). We defer to the FDA’s interpretation of its CBE and generic labeling regulations. Although Mensing and Demahy offer other ways to interpret the regulations, see Brief for Respondents 33–35, we do not find the agency’s interpretation “plainly erroneous or inconsistent with the regulation.” Auer, supra, at 461 (internal quotation marks omitted). Nor do Mensing and Demahy suggest there is any other reason to doubt the agency’s reading. We therefore conclude that the CBE process was not open to the Manufacturers for the sort of change required by state law. 2 Next, Mensing and Demahy contend that the Manufacturers could have used “Dear Doctor” letters to send ad-ditional warnings to prescribing physicians and other healthcare professionals. See Brief for Respondents 36; 21 CFR §200.5. Again, the FDA disagrees, and we defer to the agency’s views. The FDA argues that Dear Doctor letters qualify as “labeling.” U. S. Brief 18; see also 21 U. S. C. §321(m); 21 CFR §202.1(l)(2). Thus, any such letters must be “consistent with and not contrary to [the drug’s] approved … labeling.” 21 CFR §201.100(d)(1). A Dear Doctor letter that contained substantial new warning information would not be consistent with the drug’s approved labeling. Moreover, if generic drug manufacturers, but not the brand-name manufacturer, sent such letters, that would inaccurately imply a therapeutic difference between the brand and generic drugs and thus could be impermissibly “misleading.” U. S. Brief 19; see 21 CFR §314.150(b)(3) (FDA may withdraw approval of a generic drug if “the labeling of the drug … is false or misleading in any particular”). As with the CBE regulation, we defer to the FDA. Mensing and Demahy offer no argument that the FDA’s interpretation is plainly erroneous. See Auer, 519 U. S., at 461. Accordingly, we conclude that federal law did not permit the Manufacturers to issue additional warnings through Dear Doctor letters. 3 Though the FDA denies that the Manufacturers could have used the CBE process or Dear Doctor letters to strengthen their warning labels, the agency asserts that a different avenue existed for changing generic drug labels. According to the FDA, the Manufacturers could have proposed—indeed, were required to propose—stronger warning labels to the agency if they believed such warnings were needed. U. S. Brief 20; 57 Fed. Reg. 17961. If the FDA had agreed that a label change was necessary, it would have worked with the brand-name manufacturer to create a new label for both the brand-name and generic drug. Ibid. The agency traces this duty to 21 U. S. C. §352(f)(2), which provides that a drug is “misbranded … [u]nless its labeling bears … adequate warnings against … unsafe dosage or methods or duration of administration or application, in such manner and form, as are necessary for the protection of users.” See U. S. Brief 12. By regulation, the FDA has interpreted that statute to require that “labeling shall be revised to include a warning as soon as there is reasonable evidence of an association of a serious hazard with a drug.” 21 CFR §201.57(e). According to the FDA, these requirements apply to ge-neric drugs. As it explains, a “ ‘central premise of fed- eral drug regulation is that the manufacturer bears responsibility for the content of its label at all times.’ ” U. S. Brief 12–13 (quoting Wyeth, 555 U. S., at 570–571). The FDA reconciles this duty to have adequate and accurate labeling with the duty of sameness in the following way: Generic drug manufacturers that become aware of safety problems must ask the agency to work toward strengthening the label that applies to both the generic and brand-name equivalent drug. U. S. Brief 20. The Manufacturers and the FDA disagree over whether this alleged duty to request a strengthened label actually existed. The FDA argues that it explained this duty in the preamble to its 1992 regulations implementing the Hatch-Waxman Amendments. Ibid.; see 57 Fed. Reg. 17961 (“If a [generic drug manufacturer] believes new safety information should be added to a product’s labeling, it should contact FDA, and FDA will determine whether the labeling for the generic and listed drugs should be revised”). The Manufacturers claim that the FDA’s 19-year-old statement did not create a duty, and that there is no evi-dence of any generic drug manufacturer ever acting pursuant to any such duty. See Tr. of Oral Arg. 19–24; Reply Brief for Petitioner PLIVA et al. 18–22. Because we ultimately find pre-emption even assuming such a duty existed, we do not resolve the matter. C To summarize, the relevant state and federal requirements are these: State tort law places a duty directly on all drug manufacturers to adequately and safely label their products. Taking Mensing and Demahy’s allegations as true, this duty required the Manufacturers to use a different, stronger label than the label they actually used. Federal drug regulations, as interpreted by the FDA, prevented the Manufacturers from independently changing their generic drugs’ safety labels. But, we assume, federal law also required the Manufacturers to ask for FDA assistance in convincing the brand-name manufacturer to adopt a stronger label, so that all corresponding generic drug manufacturers could do so as well. We turn now to the question of pre-emption. III The Supremacy Clause establishes that federal law “shall be the supreme Law of the Land … any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.” U. S. Const., Art. VI, cl. 2. Where state and federal law “directly conflict,” state law must give way. Wyeth, supra, at 583 (Thomas, J., concurring in judgment); see also Crosby v. National Foreign Trade Council, 530 U. S. 363, 372 (2000) (“[S]tate law is naturally preempted to the extent of any conflict with a federal statute”). We have held that state and federal law conflict where it is “impossible for a private party to comply with both state and federal requirements.”[Footnote 4] Freightliner Corp. v. Myrick, 514 U. S. 280, 287 (1995) (internal quotation marks omitted).[Footnote 5] A We find impossibility here. It was not lawful under federal law for the Manufacturers to do what state law required of them. And even if they had fulfilled their federal duty to ask for FDA assistance, they would not have satisfied the requirements of state law. If the Manufacturers had independently changed their labels to satisfy their state-law duty, they would have violated federal law. Taking Mensing and Demahy’s allegations as true, state law imposed on the Manufac-turers a duty to attach a safer label to their generic metoclopramide. Federal law, however, demanded that ge-neric drug labels be the same at all times as the corres-ponding brand-name drug labels. See, e.g., 21 CFR §314.150(b)(10). Thus, it was impossible for the Manufacturers to comply with both their state-law duty to change the label and their federal law duty to keep the label the same. The federal duty to ask the FDA for help in strengthening the corresponding brand-name label, assuming such a duty exists, does not change this analysis. Although requesting FDA assistance would have satisfied the Man-ufacturers’ federal duty, it would not have satisfied their state tort-law duty to provide adequate labeling. State law demanded a safer label; it did not instruct the Manufacturers to communicate with the FDA about the possibility of a safer label. Indeed, Mensing and Demahy deny that their state tort claims are based on the Manufacturers’ alleged failure to ask the FDA for assistance in changing the labels. Brief for Respondents 53–54; cf. Buckman Co. v. Plaintiffs’ Legal Comm., 531 U. S. 341 (2001) (holding that federal drug and medical device laws pre-empted a state tort-law claim based on failure to properly communicate with the FDA). B 1 Mensing and Demahy contend that, while their state-law claims do not turn on whether the Manufacturers asked the FDA for assistance in changing their labels, the Manufacturers’ federal affirmative defense of pre-emption does. Mensing and Demahy argue that if the Manufacturers had asked the FDA for help in changing the corresponding brand-name label, they might eventually have been able to accomplish under federal law what state law requires. That is true enough. The Manufacturers “freely concede” that they could have asked the FDA for help. PLIVA Brief 48. If they had done so, and if the FDA decided there was sufficient supporting information, and if the FDA undertook negotiations with the brand-name manufacturer, and if adequate label changes were decided on and implemented, then the Manufacturers would have started a Mouse Trap game that eventually led to a better label on generic metoclopramide. This raises the novel question whether conflict pre-emption should take into account these possible actions by the FDA and the brand-name manufacturer. Here, what federal law permitted the Manufacturers to do could have changed, even absent a change in the law itself, depending on the actions of the FDA and the brand-name manufacturer. Federal law does not dictate the text of each generic drug’s label, but rather ties those labels to their brand-name counterparts. Thus, federal law would permit the Manufacturers to comply with the state labeling requirements if, and only if, the FDA and the brand-name manufacturer changed the brand-name label to do so. Mensing and Demahy assert that when a private party’s ability to comply with state law depends on approval and assistance from the FDA, proving pre-emption requires that party to demonstrate that the FDA would not have allowed compliance with state law. Here, they argue, the Manufacturers cannot bear their burden of proving impossibility because they did not even try to start the process that might ultimately have allowed them to use a safer label. Brief for Respondents 47. This is a fair argument, but we reject it. The question for “impossibility” is whether the private party could independently do under federal law what state law requires of it. See Wyeth, 555 U. S., at 573 (finding no pre-emption where the defendant could “unilaterally” do what state law required). Accepting Mensing and Demahy’s argument would render conflict pre-emption largely meaningless because it would make most conflicts between state and federal law illusory. We can often imagine that a third party or the Federal Government might do something that makes it lawful for a private party to accomplish under federal law what state law requires of it. In these cases, it is certainly possible that, had the Manufacturers asked the FDA for help, they might have eventually been able to strengthen their warning label. Of course, it is also possible that the Manufacturers could have convinced the FDA to reinterpret its regulations in a manner that would have opened the CBE process to them. Following Mensing and Demahy’s argument to its logical conclusion, it is also possible that, by asking, the Manufacturers could have persuaded the FDA to rewrite its generic drug regulations entirely or talked Congress into amending the Hatch-Waxman Amendments. If these conjectures suffice to prevent federal and state law from conflicting for Supremacy Clause purposes, it is unclear when, outside of express pre-emption, the Supremacy Clause would have any force.[Footnote 6] We do not read the Supremacy Clause to permit an approach to pre-emption that renders conflict pre-emption all but meaningless. The Supremacy Clause, on its face, makes federal law “the supreme Law of the Land” even absent an express statement by Congress. U. S. Const., Art. VI, cl. 2. 2 Moreover, the text of the Clause—that federal law shall be supreme, “any Thing in the Constitution or Laws of any State to the Contrary notwithstanding”—plainly contemplates conflict pre-emption by describing federal law as effectively repealing contrary state law. Ibid.; see Nelson, Preemption, 86 Va. L. Rev. 225, 234 (2000); id., at 252–253 (describing discussion of the Supremacy Clause in state ratification debates as concerning whether federal law could repeal state law, or vice versa). The phrase “any [state law] to the Contrary notwithstanding” is a non ob-stante provision. Id., at 238–240, nn. 43–45. Eighteenth-century legislatures used non obstante provisions to specify the degree to which a new statute was meant to repeal older, potentially conflicting statutes in the same field. Id., at 238–240 (citing dozens of statutes from the 1770’s and 1780’s with similar provisions). A non obstante provision “in a new statute acknowledged that the statute might contradict prior law and instructed courts not to apply the general presumption against implied repeals.” Id., at 241–242; 4 M. Bacon, A New Abridgment of the Law 639 (4th ed. 1778) (“Although two Acts of Parliament are seemingly repugnant, yet if there be no Clause of non Obstante in the latter, they shall if possible have such Construction, that the latter may not be a Repeal of the former by Implication”). The non obstante provision in the Supremacy Clause therefore suggests that federal law should be understood to impliedly repeal conflicting state law. Further, the provision suggests that courts should not strain to find ways to reconcile federal law with seemingly conflicting state law. Traditionally, courts went to great lengths attempting to harmonize conflicting statutes, in order to avoid implied repeals. Warder v. Arell, 2 Va. 282, 296 (1796) (opinion of Roane, J.) (“[W]e ought to seek for such a construction as will reconcile [the statutes] together”); Ludlow’s Heirs v. Johnston, 3 Ohio 553, 564 (1828) (“[I]f by any fair course of reasoning the two [statutes] can be reconciled, both shall stand”); Doolittle v. Bryan, 14 How. 563, 566 (1853) (requiring “the repugnance be quite plain” before finding implied repeal). A non obstante provision thus was a useful way for legislatures to specify that they did not want courts distorting the new law to accommodate the old. Nelson, supra, at 240–242; see also J. Sutherland, Statutes and Statutory Construction §147, p. 199 (1891) (“[W]hen there is inserted in a statute a provision [of non obstante] … . It is to be supposed that courts will be less inclined against recognizing repugnancy in applying such statutes”); Weston’s Case, 73 Eng. Rep. 780, 781 (K. B. 1576) (“[W]hen there are two statutes, one in appearance crossing the other, and no clause of non obstante is contained in the second statute … the exposition ought to be that both should stand in force”); G. Jacob, A New Law Dictionary (J. Morgan ed., 10th ed. 1782) (definition of “statute,” ¶6: “[W]hen there is a seeming variance between two statutes, and no clause of non obstante in the latter, such construction shall be made that both may stand”). The non obstante provision of he Supremacy Clause indicates that a court need look no further than “the ordinary meanin[g]” of federal law, and should not distort federal law to accommodate conflicting state law. Wyeth, 555 U. S., at 588 (Thomas, J., concurring in judgment) (internal quotation marks omitted). To consider in our pre-emption analysis the contingencies inherent in these cases—in which the Manufacturers’ ability to comply with state law depended on uncertain federal agency and third-party decisions—would be inconsistent with the non obstante provision of the Supremacy Clause. The Manufacturers would be required continually to prove the counterfactual conduct of the FDA and brand-name manufacturer in order to establish the supremacy of federal law. We do not think the Supremacy Clause contemplates that sort of contingent supremacy. The non obstante provision suggests that pre-emption analysis should not involve speculation about ways in which federal agency and third-party actions could potentially reconcile federal duties with conflicting state duties. When the “ordinary meaning” of federal law blocks a private party from independently accomplishing what state law requires, that party has established pre-emption. 3 To be sure, whether a private party can act sufficiently independently under federal law to do what state law requires may sometimes be difficult to determine. But this is not such a case. Before the Manufacturers could satisfy state law, the FDA—a federal agency—had to undertake special effort permitting them to do so. To decide these cases, it is enough to hold that when a party cannot satisfy its state duties without the Federal Government’s special permission and assistance, which is dependent on the exercise of judgment by a federal agency, that party cannot independently satisfy those state duties for pre-emption purposes. Here, state law imposed a duty on the Manufacturers to take a certain action, and federal law barred them from taking that action. The only action the Manufacturers could independently take—asking for the FDA’s help—is not a matter of state-law concern. Mensing and Demahy’s tort claims are pre-empted. C Wyeth is not to the contrary. In that case, as here, the plaintiff contended that a drug manufacturer had breached a state tort-law duty to provide an adequate warning label. 555 U. S., at 559–560. The Court held that the lawsuit was not pre-empted because it was possible for Wyeth, a brand-name drug manufacturer, to comply with both state and federal law. Id., at 572–573.[Footnote 7] Specifically, the CBE regulation, 21 CFR §314.70(c)(6)(iii), permitted a brand-name drug manufacturer like Wyeth “to unilaterally strengthen its warning” without prior FDA approval. 555 U. S., at 573; cf. supra, at 7–8. Thus, the federal regulations applicable to Wyeth allowed the company, of its own volition, to strengthen its label in compliance with its state tort duty.[Footnote 8] We recognize that from the perspective of Mensing and Demahy, finding pre-emption here but not in Wyeth makes little sense. Had Mensing and Demahy taken Reglan, the brand-name drug prescribed by their doctors, Wyeth would control and their lawsuits would not be pre-empted. But because pharmacists, acting in full accord with state law, substituted generic metoclopramide instead, federal law pre-empts these lawsuits. See, e.g., Minn. Stat. §151.21 (2010) (describing when pharmacists may substitute ge-neric drugs); La. Rev. Stat. Ann. §37:1241(A)(17) (West 2007) (same). We acknowledge the unfortunate hand that federal drug regulation has dealt Mensing, Demahy, and others similarly situated.[Footnote 9] But “it is not this Court’s task to decide whether the statutory scheme established by Congress is unusual or even bizarre.” Cuomo v. Clearing House Assn., L. L. C., 557 U. S. ___, ___ (2009) (Thomas, J., concurring in part and dissenting in part) (slip op., at 21) (internal quotation marks and brackets omitted). It is beyond dispute that the federal statutes and regulations that apply to brand-name drug manufacturers are meaningfully different than those that apply to generic drug manufacturers. Indeed, it is the special, and different, regulation of generic drugs that allowed the generic drug market to expand, bringing more drugs more quickly and cheaply to the public. But different federal statutes and regulations may, as here, lead to different pre-emption results. We will not distort the Supremacy Clause in order to create similar pre-emption across a dissimilar statutory scheme. As always, Congress and the FDA retain the authority to change the law and regulations if they so desire. * * * The judgments of the Fifth and Eighth Circuits are reversed, and the cases are remanded for further proceedings consistent with this opinion. It is so ordered. * Justice Kennedy joins all but Part III–B–2 of this opinion. Footnote 1 All relevant events in these cases predate the Food and Drug Administration Amendments Act of 2007, 121 Stat. 823. We therefore refer exclusively to the pre-2007 statutes and regulations and express no view on the impact of the 2007 Act. Footnote 2 As we use it here, “generic drug” refers to a drug designed to be a copy of a reference listed drug (typically a brand-name drug), and thus identical in active ingredients, safety, and efficacy. See, e.g., United States v. Generix Drug Corp., 460 U. S. 453, 454–455 (1983); 21 CFR §314.3(b) (2006) (defining “reference listed drug”). Footnote 3 The brief filed by the United States represents the views of the FDA. Cf. Talk America, Inc. v. Michigan Bell Telephone Co., 564 U. S. ___, ___, n. 1 (2011) (slip op., at 1, n. 1); Chase Bank USA, N. A. v. McCoy, 562 U. S. ___, ___ (2011) (slip op., at 8). Although we defer to the agency’s interpretation of its regulations, we do not defer to an agency’s ultimate conclusion about whether state law should be pre-empted. Wyeth v. Levine, 555 U. S. 555, 576 (2009). Footnote 4 We do not address whether state and federal law “directly conflict” in circumstances beyond “impossibility.” See Wyeth, 555 U. S., at 582, 590–591 (Thomas, J., concurring in judgment) (suggesting that they might). Footnote 5 The Hatch-Waxman Amendments contain no provision expressly pre-empting state tort claims. See post, at 9, 19 (Sotomayor, J., dissenting). Nor do they contain any saving clause to expressly preserve state tort claims. Cf. Williamson v. Mazda Motor of America, Inc., 562 U. S. ___, ___ (2011) (Thomas, J., concurring in judgment) (discussing the saving clause in the National Traffic and Motor Vehicle Safety Act of 1966, 49 U. S. C. §30103(e)). Although an express statement on pre-emption is always preferable, the lack of such a statement does not end our inquiry. Contrary to the dissent’s suggestion, the absence of express pre-emption is not a reason to find no conflict pre-emption. See post, at 19. Footnote 6 The dissent asserts that we are forgetting “purposes-and-objectives” pre-emption. Post, at 15–16. But as the dissent acknowledges, purposes-and-objectives pre-emption is a form of conflict pre-emption. Post, at 9, 16. If conflict pre-emption analysis must take into account hypothetical federal action, including possible changes in Acts of Congress, then there is little reason to think that pre-emption based on the purposes and objectives of Congress would survive either. Footnote 7 Wyeth also urged that state tort law “creat[ed] an unacceptable ‘obstacle to the accomplishment and execution of the full purposes and objectives of Congress.’ ” 555 U. S., at 563–564 (quoting Hines v. Davidowitz, 312 U. S. 52, 67 (1941)). The Court rejected that argument, and that type of pre-emption is not argued here. Cf. post, at 16, n. 13 (opinion of Sotomayor, J.). Footnote 8 The FDA, however, retained the authority to eventually rescind Wyeth’s unilateral CBE changes. Accordingly, the Court noted that Wyeth could have attempted to show, by “clear evidence,” that the FDA would have rescinded any change in the label and thereby demonstrate that it would in fact have been impossible to do under federal law what state law required. Wyeth, supra, at 571. Wyeth offered no such evidence. That analysis is consistent with our holding today. The Court in Wyeth asked what the drug manufacturer could independently do under federal law, and in the absence of clear evidence that Wyeth could not have accomplished what state law required of it, found no pre-emption. The Wyeth Court held that, because federal law accommodated state law duties, “the possibility of impossibility” was “not enough.” Post, at 10; see also Rice v. Norman Williams Co., 458 U. S. 654, 659 (1982) (rejecting “hypothetical” impossibility). But here, “existing” federal law directly conflicts with state law. Post, at 15 (“Conflict analysis necessarily turns on existing law”). The question in these cases is not whether the possibility of impossibility establishes pre-emption, but rather whether the possibility of possibility defeats pre-emption. Post, at 10. Footnote 9 That said, the dissent overstates what it characterizes as the “many absurd consequences” of our holding. Post, at 18. First, the FDA in-forms us that “[a]s a practical matter, genuinely new information about drugs in long use (as generic drugs typically are) appears infre-quently.” U. S. Brief 34–35. That is because patent protections ordinarily prevent generic drugs from arriving on the market for a number of years after the brand-name drug appears. Indeed, situations like the one alleged here are apparently so rare that the FDA has no “formal regulation” establishing generic drug manufacturers’ duty to initiate a label change, nor does it have any regulation setting out that label-change process. Id., at 20–21. Second, the dissent admits that, even under its approach, generic drug manufacturers could establish pre-emption in a number of scenarios. Post, at 12–13.

562.US.115

Respondent Moore and two accomplices attacked and bloodied Kenneth Rogers, tied him up, and threw him in the trunk of a car before driving into the Oregon countryside, where Moore fatally shot him. Afterwards, Moore and one accomplice told Moore’s brother and the accomplice’s girlfriend that they had intended to scare Rogers, but that Moore had accidentally shot him. Moore and the accomplice repeated this account to the police. On the advice of counsel, Moore agreed to plead no contest to felony murder in exchange for the minimum sentence for that offense. He later sought postconviction relief in state court, claiming that he had been denied effective assistance of counsel. He complained that his lawyer had not moved to suppress his confession to police in advance of the lawyer’s advice that Moore considered before accepting the plea offer. The court concluded the suppression motion would have been fruitless in light of Moore’s other admissible confession to two witnesses. Counsel gave that as his reason for not making the motion. He added that he had advised Moore that, because of the abuse Rogers suffered before the shooting, Moore could be charged with aggravated murder. That crime was punishable by death or life in prison without parole. These facts led the state court to conclude Moore had not established ineffective assistance of counsel under Strickland v. Washington, 466 U. S. 668. Moore sought federal habeas relief, renewing his ineffective-assistance claim. The District Court denied the petition, but the Ninth Circuit reversed, holding that the state court’s conclusion was an unreasonable application of clearly established law in light of Strickland and was contrary to Arizona v. Fulminante, 499 U. S. 279. Held: Moore was not entitled to the habeas relief ordered by the Ninth Circuit. Pp. 4–17. (a) Under 28 U. S. C. §2254(d), federal habeas relief may not be granted with respect to any claim a state court has adjudicated on the merits unless, among other exceptions, the state-court decision denying relief involves “an unreasonable application” of “clearly established Federal law, as determined by” this Court. The relevant federal law is the standard for ineffective assistance of counsel under Strickland, which requires a showing of “both deficient performance by counsel and prejudice.” Knowles v. Mirzayance, 556 U. S. ___, ___. Pp. 4–6. (b) The state-court decision was not an unreasonable application of either part of the Strickland rule. Pp. 6–16. (1) The state court would not have been unreasonable to accept as a justification for counsel’s action that suppression would have been futile in light of Moore’s other admissible confession to two witnesses. This explanation confirms that counsel’s representation was adequate under Strickland, so it is unnecessary to consider the reasonableness of his other justification—that a suppression motion would have failed. Plea bargains involve complex negotiations suffused with uncertainty, and defense counsel must make strategic choices in balancing opportunities—pleading to a lesser charge and obtaining a lesser sentence—and risks—that the plea bargain might come before the prosecution finds its case is getting weaker, not stronger. Failure to respect the latitude Strickland requires can create at least two problems. First, the potential for distortions and imbalance that can inhere in a hindsight perspective may become all too real; and habeas courts must be mindful of their limited role, to assess deficiency in light of information then available to counsel. Second, ineffective-assistance claims that lack necessary foundation may bring instability to the very process the inquiry seeks to protect because prosecutors must have assurances that a plea will not be undone in court years later. In applying and defining the Strickland standard—reasonable competence in representing the accused—substantial deference must be accorded to counsel’s judgment. The absence of a developed and extensive record and well-defined prosecution or defense case creates a particular risk at the early plea stage. Here, Moore’s prospects at trial were anything but certain. Counsel knew that the two witnesses presented a serious strategic concern and that delaying the plea for further proceedings might allow the State to uncover additional incriminating evidence in support of a capital prosecution. Under these circumstances, counsel made a reasonable choice. At the very least, the state court would not have been unreasonable to so conclude. The Court of Appeals relied further on Fulminante, but a state-court adjudication of counsel’s performance under the Sixth Amendment cannot be “contrary to” Fulminante, for Fulminante—which involved the admission of an involuntary confession in violation of the Fifth Amendment—says nothing about Strickland’s effectiveness standard. Pp. 6–12. (2) The state court also reasonably could have concluded that Moore was not prejudiced by counsel’s actions. To prevail in state court, he had to demonstrate “a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” Hill v. Lockhart, 474 U. S. 52, 59. Deference to the state court’s prejudice determination is significant, given the uncertainty inherent in plea negotiations. That court reasonably could have determined that Moore would have accepted the plea agreement even if his second confession had been ruled inadmissible. The State’s case was already formidable with two witnesses to an admissible confession, and it could have become stronger had the investigation continued. Moore also faced the possibility of grave punishments. Counsel’s bargain for the minimum sentence for the crime of conviction was thus favorable, and forgoing a challenge to the confession may have been essential to securing that agreement. Again, the state court’s finding could not be contrary to Fulminante, which does not speak to Strickland’s prejudice standard or contemplate prejudice in the plea bargain context. To the extent Fulminante’s harmless-error analysis sheds any light on this case, it suggests that the state court’s prejudice determination was reasonable. Pp. 12–16. 574 F. 3d 1092, reversed and remanded. Kennedy, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Thomas, Breyer, Alito, and Sotomayor, JJ., joined. Ginsburg, J., filed an opinion concurring in the judgment. Kagan, J., took no part in the consideration or decision of the case.

This case calls for determinations parallel in some respects to those discussed in today’s opinion in Harrington v. Richter, ante, p. ___. Here, as in Richter, the Court reviews a decision of the Court of Appeals for the Ninth Circuit granting federal habeas corpus relief in a challenge to a state criminal conviction. Here, too, the case turns on the proper implementation of one of the stated premises for issuance of federal habeas corpus contained in 28 U. S. C. §2254(d), the instruction that federal habeas corpus relief may not be granted with respect to any claim a state court has adjudicated on the merits unless, among other exceptions, the state court’s decision denying relief involves “an unreasonable application” of “clearly established Federal law, as determined by the Supreme Court of the United States.” And, as in Richter, the relevant clearly established law derives from Strickland v. Washington, 466 U. S. 668 (1984), which provides the standard for inadequate assistance of counsel under the Sixth Amendment. Richter involves a California conviction and addresses the adequacy of representation when counsel did not consult or use certain experts in pretrial preparation and at trial. The instant case involves an unrelated Oregon conviction and concerns the adequacy of representation in providing an assessment of a plea bargain without first seeking suppression of a confession assumed to have been improperly obtained. I On December 7, 1995, respondent Randy Moore and two confederates attacked Kenneth Rogers at his home and bloodied him before tying him with duct tape and throwing him in the trunk of a car. They drove into the Oregon countryside, where Moore shot Rogers in the temple, killing him. Afterwards, Moore and one of his accomplices told two people—Moore’s brother and the accomplice’s girlfriend—about the crimes. According to Moore’s brother, Moore and his accomplice admitted: “[T]o make an example and put some scare into Mr. Rogers …, they had blind-folded him [and] duct taped him and put him in the trunk of the car and took him out to a place that’s a little remote … . [T]heir intent was to leave him there and make him walk home … [Moore] had taken the revolver from Lonnie and at the time he had taken it, Mr. Rogers had slipped backwards on the mud and the gun discharged.” App. 157–158. Moore and his accomplice repeated this account to the police. On the advice of counsel Moore agreed to plead no contest to felony murder in exchange for a sentence of 300 months, the minimum sentence allowed by law for the offense. Moore later filed for postconviction relief in an Oregon state court, alleging that he had been denied his right to effective assistance of counsel. He complained that his lawyer had not filed a motion to suppress his confession to police in advance of the lawyer’s advice that Moore considered before accepting the plea offer. After an evidentiary hearing, the Oregon court concluded a “motion to suppress would have been fruitless” in light of the other admissible confession by Moore, to which two witnesses could testify. Id., at 140. As the court noted, Moore’s trial counsel explained why he did not move to exclude Moore’s confession to police: “Mr. Moore and I discussed the possibility of filing a Motion to Suppress and concluded that it would be unavailing, because . . . he had previously made a full confession to his brother and to [his accomplice’s girlfriend], either one of whom could have been called as a witness at any time to repeat his confession in full detail.” Jordan Affidavit (Feb. 26, 1999), App. to Pet. for Cert. 70, ¶ 4. Counsel added that he had made Moore aware of the possibility of being charged with aggravated murder, which carried a potential death sentence, as well as the possibility of a sentence of life imprisonment without parole. See Ore. Rev. Stat. §163.105(1)(a) (1995). The intense and serious abuse to the victim before the shooting might well have led the State to insist on a strong response. In light of these facts the Oregon court concluded Moore had not established ineffective assistance of counsel under Strickland. Moore filed a petition for habeas corpus in the United States District Court for the District of Oregon, renewing his ineffective-assistance claim. The District Court denied the petition, finding sufficient evidence to support the Oregon court’s conclusion that suppression would not have made a difference. A divided panel of the United States Court of Appeals for the Ninth Circuit reversed. Moore v. Czerniak, 574 F. 3d 1092 (2009). In its view the state court’s conclusion that counsel’s action did not constitute ineffective assistance was an unreasonable application of clearly established law in light of Strickland and was contrary to Arizona v. Fulminante, 499 U. S. 279 (1991). Six judges dissented from denial of rehearing en banc. 574 F. 3d, at 1162. We granted certiorari. 559 U. S. ___ (2010). II The statutory authority of federal courts to issue habeas corpus relief for persons in state custody is defined by 28 U. S. C. §2254, as amended by the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA). The text of §2254(d) states: “An application for a writ of habeas corpus on behalf of a person in custody pursuant to the judgment of a State court shall not be granted with respect to any claim that was adjudicated on the merits in State court proceedings unless the adjudication of the claim— “(1) resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States; or “(2) resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” AEDPA prohibits federal habeas relief for any claim adjudicated on the merits in state court, unless one of the exceptions listed in §2254(d) obtains. Relevant here is §2254(d)(1)’s exception “permitting relitigation where the earlier state decision resulted from an ‘unreasonable application of’ clearly established federal law.” Richter, ante, at 11. The applicable federal law consists of the rules for determining when a criminal defendant has received inadequate representation as defined in Strickland. To establish ineffective assistance of counsel “a defendant must show both deficient performance by counsel and prejudice.” Knowles v. Mirzayance, 556 U. S. ___, ___ (2009) (slip op., at 10). In addressing this standard and its relationship to AEDPA, the Court today in Richter, ante, at 14–16, gives the following explanation: “To establish deficient performance, a person challenging a conviction must show that ‘counsel’s representation fell below an objective standard of rea-sonableness.’ [Strickland,] 466 U. S., at 688. A court considering a claim of ineffective assistance must apply a ‘strong presumption’ that counsel’s representation was within the ‘wide range’ of reasonable professional assistance. Id., at 689. The challenger’s burden is to show ‘that counsel made errors so serious that counsel was not functioning as the “counsel” guaranteed the defendant by the Sixth Amendment.’ Id., at 687. “With respect to prejudice, a challenger must demonstrate ‘a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.’ … “ ‘Surmounting Strickland’s high bar is never an easy task.’ Padilla v. Kentucky, 559 U. S. ___, ___ (2010) (slip op., at 14). An ineffective-assistance claim can function as a way to escape rules of waiver and forfeiture and raise issues not presented at trial [or in pretrial proceedings], and so the Strickland standard must be applied with scrupulous care, lest ‘intrusive post-trial inquiry’ threaten the integrity of the very adversary process the right to counsel is meant to serve. Strickland, 466 U. S., at 689–690. Even under de novo review, the standard for judging counsel’s representation is a most deferential one. Unlike a later reviewing court, the attorney observed the relevant proceedings, knew of materials outside the record, and interacted with the client, with opposing counsel, and with the judge. It is ‘all too tempting’ to ‘second-guess counsel’s assistance after conviction or adverse sentence.’ Id., at 689; see also Bell v. Cone, 535 U. S. 685, 702 (2002); Lockhart v. Fretwell, 506 U. S. 364, 372 (1993). The question is whether an attorney’s representation amounted to incompetence under ‘prevailing professional norms,’ not whether it deviated from best practices or most common custom. Strickland, 466 U. S., at 690. “Establishing that a state court’s application of Strickland was unreasonable under §2254(d) is all the more difficult. The standards created by Strickland and §2254(d) are both ‘highly deferential,’ id., at 689; Lindh v. Murphy, 521 U. S. 320, 333, n. 7 (1997), and when the two apply in tandem, review is ‘doubly’ so, Knowles, 556 U. S., at ___ (slip op., at 11). The Strickland standard is a general one, so the range of reasonable applications is substantial. 556 U. S., at ___ (slip op., at 11). Federal habeas courts must guard against the danger of equating unreasonableness under Strickland with unreasonableness under §2254(d). When §2254(d) applies, the question is not whether counsel’s actions were reasonable. The question is whether there is any reasonable argument that counsel satisfied Strickland’s deferential standard.” III The question becomes whether Moore’s counsel provided ineffective assistance by failing to seek suppression of Moore’s confession to police before advising Moore regarding the plea. Finding that any “motion to suppress would have been fruitless,” the state postconviction court concluded that Moore had not received ineffective assistance of counsel. App. 140. The state court did not specify whether this was because there was no deficient performance under Strickland or because Moore suffered no Strickland prejudice, or both. To overcome the limitation imposed by § 2254(d), the Court of Appeals had to conclude that both findings would have involved an unreasonable application of clearly established federal law. See Richter, ante, at 19–20. In finding that this standard was met, the Court of Appeals erred, for the state-court decision was not an unreasonable application of either part of the Strickland rule. A The Court of Appeals was wrong to accord scant deference to counsel’s judgment, and doubly wrong to conclude it would have been unreasonable to find that the defense attorney qualified as counsel for Sixth Amendment purposes. Knowles, supra, at ––– (slip op., at 11); Strickland, 466 U. S., at 687. Counsel gave this explanation for his decision to discuss the plea bargain without first challenging Moore’s confession to the police: that suppression would serve little purpose in light of Moore’s other full and admissible confession, to which both his brother and his accomplice’s girlfriend could testify. The state court would not have been unreasonable to accept this explanation. Counsel also justified his decision by asserting that any motion to suppress was likely to fail. Reviewing the reasonableness of that justification is complicated by the possibility that petitioner forfeited one argument that would have supported its position: The Court of Appeals assumed that a motion would have succeeded because the warden did not argue otherwise. Of course that is not the same as a concession that no competent attorney would think a motion to suppress would have failed, which is the relevant question under Strickland. See Kimmelman v. Morrison, 477 U. S. 365, 382 (1986); Richter, ante, at 19–20. It is unnecessary to consider whether counsel’s second justification was reasonable, however, since the first and independent explanation—that suppression would have been futile—confirms that his representation was adequate under Strickland, or at least that it would have been reasonable for the state court to reach that conclusion. Acknowledging guilt and accepting responsibility by an early plea respond to certain basic premises in the law and its function. Those principles are eroded if a guilty plea is too easily set aside based on facts and circumstances not apparent to a competent attorney when actions and advice leading to the plea took place. Plea bargains are the result of complex negotiations suffused with uncertainty, and defense attorneys must make careful strategic choices in balancing opportunities and risks. The opportunities, of course, include pleading to a lesser charge and obtaining a lesser sentence, as compared with what might be the outcome not only at trial but also from a later plea offer if the case grows stronger and prosecutors find stiffened resolve. A risk, in addition to the obvious one of losing the chance for a defense verdict, is that an early plea bargain might come before the prosecution finds its case is getting weaker, not stronger. The State’s case can begin to fall apart as stories change, witnesses become unavailable, and new suspects are identified. These considerations make strict adherence to the Strickland standard all the more essential when reviewing the choices an attorney made at the plea bargain stage. Failure to respect the latitude Strickland requires can create at least two problems in the plea context. First, the potential for the distortions and imbalance that can inhere in a hindsight perspective may become all too real. The art of negotiation is at least as nuanced as the art of trial advocacy and it presents questions farther removed from immediate judicial supervision. There are, moreover, special difficulties in evaluating the basis for counsel’s judgment: An attorney often has insights borne of past dealings with the same prosecutor or court, and the record at the pretrial stage is never as full as it is after a trial. In determining how searching and exacting their review must be, habeas courts must respect their limited role in determining whether there was manifest deficiency in light of information then available to counsel. Lockhart v. Fretwell, 506 U. S. 364, 372 (1993). AEDPA compounds the imperative of judicial caution. Second, ineffective-assistance claims that lack necessary foundation may bring instability to the very process the inquiry seeks to protect. Strickland allows a defendant “to escape rules of waiver and forfeiture,” Richter, ante, at 15. Prosecutors must have assurance that a plea will not be undone years later because of infidelity to the requirements of AEDPA and the teachings of Strickland. The prospect that a plea deal will afterwards be unraveled when a court second-guesses counsel’s decisions while failing to accord the latitude Strickland mandates or disregarding the structure dictated by AEDPA could lead prosecutors to forgo plea bargains that would benefit defendants, a result favorable to no one. Whether before, during, or after trial, when the Sixth Amendment applies, the formulation of the standard is the same: reasonable competence in representing the accused. Strickland, 466 U. S., at 688. In applying and defining this standard substantial deference must be accorded to counsel’s judgment. Id., at 689. But at different stages of the case that deference may be measured in different ways. In the case of an early plea, neither the prosecution nor the defense may know with much certainty what course the case may take. It follows that each side, of necessity, risks consequences that may arise from contingencies or circumstances yet unperceived. The absence of a developed or an extensive record and the circumstance that neither the prosecution nor the defense case has been well defined create a particular risk that an after-the-fact assessment will run counter to the deference that must be accorded counsel’s judgment and perspective when the plea was negotiated, offered, and entered. Prosecutors in the present case faced the cost of litigation and the risk of trying their case without Moore’s confession to the police. Moore’s counsel could reasonably believe that a swift plea bargain would allow Moore to take advantage of the State’s aversion to these hazards. And whenever cases involve multiple defendants, there is a chance that prosecutors might convince one defendant to testify against another in exchange for a better deal. Moore’s plea eliminated that possibility and ended an ongoing investigation. Delaying the plea for further proceedings would have given the State time to uncover additional incriminating evidence that could have formed the basis of a capital prosecution. It must be remem-bered, after all, that Moore’s claim that it was an accident when he shot the victim through the temple might be disbelieved. It is not clear how the successful exclusion of the confession would have affected counsel’s strategic calculus. The prosecution had at its disposal two witnesses able to relate another confession. True, Moore’s brother and the girlfriend of his accomplice might have changed their accounts in a manner favorable to Moore. But the record before the state court reveals no reason to believe that either witness would violate the legal obligation to convey the content of Moore’s confession. And to the extent that his accomplice’s girlfriend had an ongoing interest in the matter, she might have been tempted to put more blame, not less, on Moore. Then, too, the accomplices themselves might have decided to implicate Moore to a greater extent than his own confession did, say by indicating that Moore shot the victim deliberately, not accidentally. All these possibilities are speculative. What counsel knew at the time was that the existence of the two witnesses to an additional confession posed a serious strategic concern. Moore’s prospects at trial were thus anything but certain. Even now, he does not deny any involvement in the kidnaping and killing. In these circumstances, and with a potential capital charge lurking, Moore’s counsel made a reasonable choice to opt for a quick plea bargain. At the very least, the state court would not have been unreasonable to so conclude. Cf. Yarborough v. Alvarado, 541 U. S. 652, 664 (2004) (explaining that state courts enjoy “more leeway” under AEDPA in applying general standards). The Court of Appeals’ contrary holding rests on a case that did not involve ineffective assistance of counsel: Arizona v. Fulminante, 499 U. S. 279 (1991). To reach that result, it transposed that case into a novel context; and novelty alone—at least insofar as it renders the relevant rule less than “clearly established”—provides a reason to reject it under AEDPA. See Yarborough, supra, at 666 (“Section 2254(d)(1) would be undermined if habeas courts introduced rules not clearly established under the guise of extensions to existing law …[, although c]ertain principles are fundamental enough that when new factual permutations arise, the necessity to apply the earlier rule will be beyond doubt”). And the transposition is improper even on its own terms. According to the Court of Appeals, “Fulminante stands for the proposition that the admission of an additional confession ordinarily reinforces and corroborates the others and is therefore prejudicial.” 574 F. 3d, at 1111. Based on that reading, the Court of Appeals held that the state court’s decision “was contrary to Fulminante.” Id., at 1102. But Fulminante may not be so incorporated into the Strickland performance inquiry. A state-court adjudication of the performance of coun- sel under the Sixth Amendment cannot be “contrary to” Fulminante, for Fulminante—which involved the admission of an involuntary confession in violation of the Fifth Amendment—says nothing about the Strickland standard of effectiveness. See Bell v. Cone, 535 U. S. 685, 694 (2002) (“A federal habeas court may issue the writ under the ‘contrary to’ clause if the state court applies a rule different from the governing law set forth in our cases, or if it decides a case differently than we have done on a set of materially indistinguishable facts”). The Fulminante prejudice inquiry presumes a constitutional violation, whereas Strickland seeks to define one. The state court accepted counsel’s view that seeking to suppress Moore’s second confession would have been “fruitless.” It would not have been unreasonable to conclude that counsel could incorporate that view into his assessment of a plea offer, a subject with which Fulminante is in no way concerned. A finding of constitutionally adequate performance under Strickland cannot be contrary to Fulminante. The state court likely reached the correct result under Strickland. And under §2254(d), that it reached a reasonable one is sufficient. See Richter, ante, at 19. B The Court of Appeals further concluded that it would have been unreasonable for the state postconviction court to have found no prejudice in counsel’s failure to suppress Moore’s confession to police. To prevail on prejudice before the state court Moore had to demonstrate “a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” Hill v. Lockhart, 474 U. S. 52, 59 (1985). Deference to the state court’s prejudice determination is all the more significant in light of the uncertainty inherent in plea negotiations described above: The stakes for defendants are high, and many elect to limit risk by forgoing the right to assert their innocence. A defendant who accepts a plea bargain on counsel’s advice does not necessarily suffer prejudice when his counsel fails to seek suppression of evidence, even if it would be reversible error for the court to admit that evidence. The state court here reasonably could have determined that Moore would have accepted the plea agreement even if his second confession had been ruled inadmissible. By the time the plea agreement cut short investigation of Moore’s crimes, the State’s case was already formidable and included two witnesses to an admissible confession. Had the prosecution continued to investigate, its case might well have become stronger. At the same time, Moore faced grave punishments. His decision to plead no contest allowed him to avoid a possible sentence of life without parole or death. The bargain counsel struck was thus a favorable one—the statutory minimum for the charged offense—and the decision to forgo a challenge to the confession may have been essential to securing that agreement. Once again the Court of Appeals reached a contrary conclusion by pointing to Fulminante: “The state court’s finding that a motion to suppress a recorded confession to the police would have been ‘fruitless’ … was without question contrary to clearly established federal law as set forth in Fulminante.” 574 F. 3d, at 1112. And again there is no sense in which the state court’s finding could be contrary to Fulminante, for Fulminante says nothing about prejudice for Strickland purposes, nor does it contemplate prejudice in the plea bargain context. The Court of Appeals appears to have treated Fulminante as a per se rule of prejudice, or something close to it, in all cases involving suppressible confessions. It is not. In Fulminante five Justices made the uncontroversial observation that many confessions are powerful evidence. See, e.g., 499 U. S., at 296. Fulminante’s prejudice analysis arose on direct review following an acknowledged constitutional error at trial. The State therefore had the burden of showing that it was “clear beyond a reasonable doubt that a rational jury would have found the defendant guilty absent the error.” Neder v. United States, 527 U. S. 1, 18 (1999) (paraphrasing Fulminante, supra). That standard cannot apply to determinations of whether inadequate assistance of counsel prejudiced a defendant who entered into a plea agreement. Many defendants reasonably enter plea agreements even though there is a significant probability—much more than a reasonable doubt—that they would be acquitted if they proceeded to trial. Thus, the question in the present case is not whether Moore was sure beyond a reasonable doubt that he would still be convicted if the extra confession were suppressed. It is whether Moore established the reasonable probability that he would not have entered his plea but for his counsel’s deficiency, Hill, supra, at 59, and more to the point, whether a state court’s decision to the contrary would be unreasonable. To the extent Fulminante’s application of the harmless-error standard sheds any light on the present case, it suggests that the state court’s prejudice determination was reasonable. Fulminante found that an improperly admitted confession was not harmless under Chapman v. California, 386 U. S. 18 (1967) because the remaining evidence against the defendant was weak. The additional evidence consisted primarily of a second confession that Fulminante had made to the informant’s fiancée. But many of its details were not corroborated, the fiancée had not reported the confession for a long period of time, the State had indicated that both confessions were essential to its case, and the fiancée potentially “had a motive to lie.” 499 U. S., at 300. Moore’s plea agreement, by contrast, ended the government’s investigation well before trial, yet the evidence against Moore was strong. The accounts of Moore’s second confession to his brother and his accomplice’s girlfriend corroborated each other, were given to people without apparent reason to lie, and were reported without delay. The State gave no indication that its felony-murder prosecution depended on the admission of the police confession, and Moore does not now deny that he kidnaped and killed Rogers. Given all this, an unconstitutional admission of Moore’s confession to police might well have been found harmless even on direct review if Moore had gone to trial after the denial of a suppression motion. Other than for its discussion of the basic proposition that a confession is often powerful evidence, Fulminante is not relevant to the present case. The state postconviction court reasonably could have concluded that Moore was not prejudiced by counsel’s actions. Under AEDPA, that finding ends federal review. See Richter, ante, at 19. Judge Berzon’s concurring opinion in the Court of Appeals does not provide a basis for issuance of the writ. The concurring opinion would have found the state court’s prejudice determination unreasonable in light of Kimmelman. It relied on Kimmelman to find that Moore suffered prejudice for Strickland purposes because there was a reasonable possibility that he would have obtained a better plea agreement but for his counsel’s errors. But Kimmelman concerned a conviction following a bench trial, so it did not establish, much less clearly establish, the appropriate standard for prejudice in cases involving plea bargains. See 477 U. S., at 389. That standard was established in Hill, which held that a defendant who enters a plea agreement must show “a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” 474 U. S., at 59. Moore’s failure to make that showing forecloses relief under AEDPA. IV There are certain differences between inadequate assistance of counsel claims in cases where there was a full trial on the merits and those, like this one, where a plea was entered even before the prosecution decided upon all of the charges. A trial provides the full written record and factual background that serve to limit and clarify some of the choices counsel made. Still, hindsight cannot suffice for relief when counsel’s choices were reasonable and legitimate based on predictions of how the trial would proceed. See Richter, ante, at 18. Hindsight and second guesses are also inappropriate, and often more so, where a plea has been entered without a full trial or, as in this case, even before the prosecution decided on the charges. The added uncertainty that results when there is no extended, formal record and no actual history to show how the charges have played out at trial works against the party alleging inadequate assistance. Counsel, too, faced that uncertainty. There is a most substantial burden on the claimant to show ineffective assistance. The plea process brings to the criminal justice system a stability and a certainty that must not be undermined by the prospect of collateral challenges in cases not only where witnesses and evidence have disappeared, but also in cases where witnesses and evidence were not presented in the first place. The substantial burden to show ineffective assistance of counsel, the burden the claimant must meet to avoid the plea, has not been met in this case. The state postconviction court’s decision involved no unreasonable application of Supreme Court precedent. Because the Court of Appeals erred in finding otherwise, its judgment is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

562.US.61

Chapter 13 of the Bankruptcy Code uses a statutory formula known as the “means test” to help ensure that debtors who can pay creditors do pay them. The means test instructs a debtor to determine his “disposable income”—the amount he has available to reimburse creditors—by deducting from his current monthly income “amounts reasonably necessary to be expended” for, inter alia, “maintenance or support.” 11 U. S. C. §1325(b)(2)(A)(i). For a debtor whose income is above the median for his State, the means test indentifies which expenses qualify as “amounts reasonably necessary to be expended.” As relevant here, the statute provides that “[t]he debtor’s monthly expenses shall be the debtor’s applicable monthly expense amounts specified under the National Standards and Local Standards, and the debtor’s actual monthly expenses for the categories specified as Other Necessary Expenses issued by the Internal Revenue Service [IRS] for the area in which the debtor resides.” §707(b)(2)(A)(ii)(I). The Standards are tables listing standardized expense amounts for basic necessities, which the IRS prepares to help calculate taxpayers’ ability to pay overdue taxes. The IRS also creates supplemental guidelines known as the “Collection Financial Standards,” which describe how to use the tables and what the amounts listed in them mean. The Local Standards include an allowance for transportation expenses, divided into vehicle “Ownership Costs” and vehicle “Operating Costs.” The Collection Financial Standards explain that “Ownership Costs” cover monthly loan or lease payments on an automobile; the expense amounts listed are based on nationwide car financing data. The Collection Financial Standards further state that a taxpayer who has no car payment may not claim an allowance for ownership costs. When petitioner Ransom filed for Chapter 13 bankruptcy relief, he listed respondent (FIA) as an unsecured creditor. Among his assets, Ransom reported a car that he owns free of any debt. In determining his monthly expenses, he nonetheless claimed a car-ownership deduction of $471, the full amount specified in the “Ownership Costs” table, as well as a separate $388 deduction for car-operating costs. Based on his means-test calculations, Ransom proposed a bankruptcy plan that would result in repayment of approximately 25% of his unsecured debt. FIA objected on the ground that the plan did not direct all of Ransom’s disposable income to unsecured creditors. FIA contended that Ransom should not have claimed the car-ownership allowance because he does not make loan or lease payments on his car. Agreeing, the Bankruptcy Court denied confirmation of the plan. The Ninth Circuit Bankruptcy Appellate Panel and the Ninth Circuit affirmed. Held: A debtor who does not make loan or lease payments may not take the car-ownership deduction. Pp. 6–18. (a) This Court’s interpretation begins with the language of the Bankruptcy Code, which provides that a debtor may claim only “applicable” expense amounts listed in the Standards. Because the Code does not define the key word “applicable,” the term carries its ordinary meaning of appropriate, relevant, suitable, or fit. What makes an expense amount “applicable” in this sense is most naturally understood to be its correspondence to an individual debtor’s financial circumstances. Congress established a filter, permitting a debtor to claim a deduction from a National or Local Standard table only if that deduction is appropriate for him. And a deduction is so appropriate only if the debtor will incur the kind of expense covered by the table during the life of the plan. Had Congress not wanted to separate debtors who qualify for an allowance from those who do not, it could have omitted the term “applicable” altogether. Without that word, all debtors would be eligible to claim a deduction for each category listed in the Standards. Interpreting the statute to require a threshold eligibility determination thus ensures that “applicable” carries meaning, as each word in a statute should. This reading draws support from the statute’s context and purpose. The Code initially defines a debtor’s disposable income as his “current monthly income … less amounts reasonably necessary to be expended.” §1325(b)(2). It then instructs that such reasonably necessary amounts “shall be determined in accordance with” the means test. §1325(b)(3). Because Congress intended the means test to approximate the debtor’s reasonable expenditures on essential items, a debtor should be required to qualify for a deduction by actually incurring an expense in the relevant category. Further, the statute’s purpose—to ensure that debtors pay creditors the maximum they can afford—is best achieved by interpreting the means test, consistent with the statutory text, to reflect a debtor’s ability to afford repayment. Pp. 6–9. (b) The vehicle-ownership category covers only the costs of a car loan or lease. The expense amount listed ($471) is the average monthly payment for loans and leases nationwide; it is not intended to estimate other conceivable expenses associated with maintaining a car. Maintenance expenses are the province of the separate “Operating Costs” deduction. A person who owns a car free and clear is entitled to the “Operating Costs” deduction for all driving-related expenses. But such a person may not claim the “Ownership Costs” deduction, because that allowance is for the separate costs of a car loan or lease. The IRS’ Collection Financial Standards reinforce this conclusion by making clear that individuals who have a car but make no loan or lease payments may take only the operating-costs deduction. Because Ransom owns his vehicle outright, he incurs no expense in the “Ownership Costs” category, and that expense amount is therefore not “applicable” to him. Pp. 9–11. (c) Ransom’s arguments to the contrary—an alternative interpretation of the key word “applicable,” an objection to the Court’s view of the scope of the “Ownership Costs” category, and a criticism of the policy implications of the Court’s approach—are unpersuasive. Pp. 11–18. 577 F. 3d 1026, affirmed. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, Ginsburg, Breyer, Alito, and Sotomayor, JJ., joined. Scalia, J., filed a dissenting opinion.

Chapter 13 of the Bankruptcy Code enables an individual to obtain a discharge of his debts if he pays his creditors a portion of his monthly income in accordance with a court-approved plan. 11 U. S. C. §1301 et seq. To determine how much income the debtor is capable of paying, Chapter 13 uses a statutory formula known as the “means test.” §§707(b)(2) (2006 ed. and Supp. III), 1325(b)(3)(A) (2006 ed.). The means test instructs a debtor to deduct specified expenses from his current monthly income. The result is his “disposable income”—the amount he has available to reimburse creditors. §1325(b)(2). This case concerns the specified expense for vehicle-ownership costs. We must determine whether a debtor like petitioner Jason Ransom who owns his car outright, and so does not make loan or lease payments, may claim an allowance for car-ownership costs (thereby reducing the amount he will repay creditors). We hold that the text, context, and purpose of the statutory provision at issue preclude this result. A debtor who does not make loan or lease payments may not take the car-ownership deduction. I A “Congress enacted the Bankruptcy Abuse Prevention and Consumer Protection Act of 2005 (BAPCPA or Act) to correct perceived abuses of the bankruptcy system.” Milavetz, Gallop & Milavetz, P. A. v. United States, 559 U. S. ___, ___ (2010) (slip op., at 1). In particular, Congress adopted the means test—“[t]he heart of [BAPCPA’s] consumer bankruptcy reforms,” H. R. Rep. No. 109–31, pt. 1, p. 2 (2005) (hereinafter H. R. Rep.), and the home of the statutory language at issue here—to help ensure that debtors who can pay creditors do pay them. See, e.g., ibid. (under BAPCPA, “debtors [will] repay creditors the maximum they can afford”). In Chapter 13 proceedings, the means test provides a formula to calculate a debtor’s disposable income, which the debtor must devote to reimbursing creditors under a court-approved plan generally lasting from three to five years. §§1325(b)(1)(B) and (b)(4).[Footnote 1] The statute defines “disposable income” as “current monthly income” less “amounts reasonably necessary to be expended” for “maintenance or support,” business expenditures, and certain charitable contributions. §§1325(b)(2)(A)(i) and (ii). For a debtor whose income is above the median for his State, the means test identifies which expenses qualify as “amounts reasonably necessary to be expended.” The test supplants the pre-BAPCPA practice of calculating debtors’ reasonable expenses on a case-by-case basis, which led to varying and often inconsistent determinations. See, e.g., In re Slusher, 359 B. R. 290, 294 (Bkrtcy. Ct. Nev. 2007). Under the means test, a debtor calculating his “reasonably necessary” expenses is directed to claim allowances for defined living expenses, as well as for secured and priority debt. §§707(b)(2)(A)(ii)–(iv). As relevant here, the statute provides: “The debtor’s monthly expenses shall be the debtor’s applicable monthly expense amounts specified under the National Standards and Local Standards, and the debtor’s actual monthly expenses for the categories specified as Other Necessary Expenses issued by the Internal Revenue Service [IRS] for the area in which the debtor resides.” §707(b)(2)(A)(ii)(I). These are the principal amounts that the debtor can claim as his reasonable living expenses and thereby shield from creditors. The National and Local Standards referenced in this provision are tables that the IRS prepares listing standardized expense amounts for basic necessities.[Footnote 2] The IRS uses the Standards to help calculate taxpayers’ ability to pay overdue taxes. See 26 U. S. C. §7122(d)(2). The IRS also prepares supplemental guidelines known as the Collection Financial Standards, which describe how to use the tables and what the amounts listed in them mean. The Local Standards include an allowance for transportation expenses, divided into vehicle “Ownership Costs” and vehicle “Operating Costs.”[Footnote 3] At the time Ransom filed for bankruptcy, the “Ownership Costs” table appeared as follows: Ownership Costs National First Car Second Car $471 $332 App. to Brief for Respondent 5a. The Collection Financial Standards explain that these ownership costs represent “nationwide figures for monthly loan or lease payments,” id., at 2a; the numerical amounts listed are “base[d] … on the five-year average of new and used car financing data compiled by the Federal Reserve Board,” id., at 3a. The Collection Financial Standards further instruct that, in the tax-collection context, “[i]f a taxpayer has no car payment, … only the operating costs portion of the transportation standard is used to come up with the allowable transportation expense.” Ibid. B Ransom filed for Chapter 13 bankruptcy relief in July 2006. App. 1, 54. Among his liabilities, Ransom itemized over $82,500 in unsecured debt, including a claim held by respondent FIA Card Services, N. A. (FIA). Id., at 41. Among his assets, Ransom listed a 2004 Toyota Camry, valued at $14,000, which he owns free of any debt. Id., at 38, 49, 52. For purposes of the means test, Ransom reported income of $4,248.56 per month. Id., at 46. He also listed monthly expenses totaling $4,038.01. Id., at 53. In determining those expenses, Ransom claimed a car-ownership deduction of $471 for the Camry, the full amount specified in the IRS’s “Ownership Costs” table. Id., at 49. Ransom listed a separate deduction of $338 for car-operating costs. Ibid. Based on these figures, Ransom had disposable income of $210.55 per month. Id., at 53. Ransom proposed a 5-year plan that would result in repayment of approximately 25% of his unsecured debt. Id., at 55. FIA objected to confirmation of the plan on the ground that it did not direct all of Ransom’s disposable income to unsecured creditors. Id., at 64. In particular, FIA argued that Ransom should not have claimed the car-ownership allowance because he does not make loan or lease payments on his car. Id., at 67. FIA noted that without this allowance, Ransom’s disposable income would be $681.55—the $210.55 he reported plus the $471 he deducted for vehicle ownership. Id., at 71. The difference over the 60 months of the plan amounts to about $28,000. C The Bankruptcy Court denied confirmation of Ransom’s plan. App. to Pet. for Cert. 48. The court held that Ransom could deduct a vehicle-ownership expense only “if he is currently making loan or lease payments on that vehicle.” Id., at 41. Ransom appealed to the Ninth Circuit Bankruptcy Appellate Panel, which affirmed. In re Ransom, 380 B. R. 799, 808–809 (2007). The panel reasoned that an “expense [amount] becomes relevant to the debtor (i.e., appropriate or applicable to the debtor) when he or she in fact has such an expense.” Id., at 807. “[W]hat is important,” the panel noted, “is the payments that debtors actually make, not how many cars they own, because [those] payments … are what actually affect their ability to” reimburse unsecured creditors. Ibid. The United States Court of Appeals for the Ninth Circuit affirmed. In re Ransom, 577 F. 3d 1026, 1027 (2009). The plain language of the statute, the court held, “does not allow a debtor to deduct an ‘ownership cost’ … that the debtor does not have.” Id., at 1030. The court observed that “[a]n ‘ownership cost’ is not an ‘expense’—either actual or applicable—if it does not exist, period.” Ibid. We granted a writ of certiorari to resolve a split of authority over whether a debtor who does not make loan or lease payments on his car may claim the deduction for vehicle-ownership costs. 559 U. S. ___ (2010).[Footnote 4] We now affirm the Ninth Circuit’s judgment. II Our interpretation of the Bankruptcy Code starts “where all such inquiries must begin: with the language of the statute itself.” United States v. Ron Pair Enterprises, Inc., 489 U. S. 235, 241 (1989). As noted, the provision of the Code central to the decision of this case states: “The debtor’s monthly expenses shall be the debtor’s applicable monthly expense amounts specified under the National Standards and Local Standards, and the debtor’s actual monthly expenses for the categories specified as Other Necessary Expenses issued by the [IRS] for the area in which the debtor resides.” §707(b)(2)(A)(ii)(I). The key word in this provision is “applicable”: A debtor may claim not all, but only “applicable” expense amounts listed in the Standards. Whether Ransom may claim the $471 car-ownership deduction accordingly turns on whether that expense amount is “applicable” to him. Because the Code does not define “applicable,” we look to the ordinary meaning of the term. See, e.g., Hamilton v. Lanning, 560 U. S. ___, ___ (2010) (slip op., at 6). “Applicable” means “capable of being applied: having relevance” or “fit, suitable, or right to be applied: appropriate.” Webster’s Third New International Dictionary 105 (2002). See also New Oxford American Dictionary 74 (2d ed. 2005) (“relevant or appropriate”); 1 Oxford English Dictionary 575 (2d ed. 1989) (“[c]apable of being applied” or “[f]it or suitable for its purpose, appropriate”). So an expense amount is “applicable” within the plain meaning of the statute when it is appropriate, relevant, suitable, or fit. What makes an expense amount “applicable” in this sense (appropriate, relevant, suitable, or fit) is most naturally understood to be its correspondence to an individual debtor’s financial circumstances. Rather than authorizing all debtors to take deductions in all listed categories, Congress established a filter: A debtor may claim a deduction from a National or Local Standard table (like “[Car] Ownership Costs”) if but only if that deduction is appropriate for him. And a deduction is so appropriate only if the debtor has costs corresponding to the category covered by the table—that is, only if the debtor will incur that kind of expense during the life of the plan. The statute underscores the necessity of making such an individualized determination by referring to “the debtor’s applicable monthly expense amounts,” §707(b)(2)(A)(ii)(I) (emphasis added)—in other words, the expense amounts applicable (appropriate, etc.) to each particular debtor. Identifying these amounts requires looking at the financial situation of the debtor and asking whether a National or Local Standard table is relevant to him. If Congress had not wanted to separate in this way debtors who qualify for an allowance from those who do not, it could have omitted the term “applicable” altogether. Without that word, all debtors would be eligible to claim a deduction for each category listed in the Standards. Congress presumably included “applicable” to achieve a different result. See Leocal v. Ashcroft, 543 U. S. 1, 12 (2004) (“[W]e must give effect to every word of a statute wherever possible”). Interpreting the statute to require a threshold determination of eligibility ensures that the term “applicable” carries meaning, as each word in a statute should. This reading of “applicable” also draws support from the statutory context. The Code initially defines a debtor’s disposable income as his “current monthly income … less amounts reasonably necessary to be expended.” §1325(b)(2) (emphasis added). The statute then instructs that “[a]mounts reasonably necessary to be expended … shall be determined in accordance with” the means test. §1325(b)(3). Because Congress intended the means test to approximate the debtor’s reasonable expenditures on essential items, a debtor should be required to qualify for a deduction by actually incurring an expense in the relevant category. If a debtor will not have a particular kind of expense during his plan, an allowance to cover that cost is not “reasonably necessary” within the meaning of the statute.[Footnote 5] Finally, consideration of BAPCPA’s purpose strengthens our reading of the term “applicable.” Congress designed the means test to measure debtors’ disposable income and, in that way, “to ensure that [they] repay creditors the maximum they can afford.” H. R. Rep., at 2. This purpose is best achieved by interpreting the means test, consistent with the statutory text, to reflect a debtor’s ability to afford repayment. Cf. Hamilton, 560 U. S., at ___ (slip op., at 14) (rejecting an interpretation of the Bankruptcy Code that “would produce [the] senseless resul[t]” of “deny[ing] creditors payments that the debtor could easily make”). Requiring a debtor to incur the kind of expenses for which he claims a means-test deduction thus advances BAPCPA’s objectives. Because we conclude that a person cannot claim an allowance for vehicle-ownership costs unless he has some expense falling within that category, the question in this case becomes: What expenses does the vehicle-ownership category cover? If it covers loan and lease payments alone, Ransom does not qualify, because he has no such expense. Only if that category also covers other costs associ- ated with having a car would Ransom be entitled to this deduction. The less inclusive understanding is the right one: The ownership category encompasses the costs of a car loan or lease and nothing more. As noted earlier, the numerical amounts listed in the “Ownership Costs” table are “base[d] … on the five-year average of new and used car financing data compiled by the Federal Reserve Board.” App. to Brief for Respondent 3a. In other words, the sum $471 is the average monthly payment for loans and leases nationwide; it is not intended to estimate other conceivable expenses associated with maintaining a car. The Standards do account for those additional expenses, but in a different way: They are mainly the province of the separate deduction for vehicle “Operating Costs,” which include payments for “[v]ehicle insurance, … maintenance, fuel, state and local registration, required inspection, parking fees, tolls, [and] driver’s license.” Internal Rev- enue Manual §§5.15.1.7 and 5.15.1.8 (May 1, 2004), reprinted in App. to Brief for Respondent 16a, 20a; see also IRS, Collection Financial Standards (Feb. 19, 2010), http://www.irs.gov/individuals/article/0,,id=96543,00.html.[Footnote 6] A person who owns a car free and clear is entitled to claim the “Operating Costs” deduction for all these expenses of driving—and Ransom in fact did so, to the tune of $338. But such a person is not entitled to claim the “Ownership Costs” deduction, because that allowance is for the separate costs of a car loan or lease. The Collection Financial Standards—the IRS’s explanatory guidelines to the National and Local Standards—explicitly recognize this distinction between ownership and operating costs, making clear that individuals who have a car but make no loan or lease payments may claim only the operating allowance. App. to Brief for Respondent 3a; see supra, at 4. Although the statute does not incorporate the IRS’s guidelines, courts may consult this material in interpreting the National and Local Standards; after all, the IRS uses those tables for a similar purpose—to determine how much money a delinquent taxpayer can afford to pay the Government. The guidelines of course cannot control if they are at odds with the statutory language. But here, the Collection Financial Standards’ treatment of the car-ownership deduction reinforces our conclusion that, under the statute, a debtor seeking to claim this deduction must make some loan or lease payments.[Footnote 7] Because Ransom owns his vehicle free and clear of any encumbrance, he incurs no expense in the “Ownership Costs” category of the Local Standards. Accordingly, the car-ownership expense amount is not “applicable” to him, and the Ninth Circuit correctly denied that deduction. III Ransom’s argument to the contrary relies on a different interpretation of the key word “applicable,” an objection to our view of the scope of the “Ownership Costs” category, and a criticism of the policy implications of our approach. We do not think these claims persuasive. A Ransom first offers another understanding of the term “applicable.” A debtor, he says, determines his “applicable” deductions by locating the box in each National or Local Standard table that corresponds to his geographic location, income, family size, or number of cars. Under this approach, a debtor “consult[s] the table[s] alone” to determine his appropriate expense amounts. Reply Brief for Petitioner 16. Because he has one car, Ransom argues that his “applicable” allowance is the sum listed in the first column of the “Ownership Costs” table ($471); if he had a second vehicle, the amount in the second column ($332) would also be “applicable.” On this approach, the word “applicable” serves a function wholly internal to the tables; rather than filtering out debtors for whom a deduction is not at all suitable, the term merely directs each debtor to the correct box (and associated dollar amount of deduction) within every table. This alternative reading of “applicable” fails to comport with the statute’s text, context, or purpose. As intimated earlier, supra, at 7–8, Ransom’s interpretation would render the term “applicable” superfluous. Assume Congress had omitted that word and simply authorized a deduction of “the debtor’s monthly expense amounts” specified in the Standards. That language, most naturally read, would direct each debtor to locate the box in every table corresponding to his location, income, family size, or number of cars and to deduct the amount stated. In other words, the language would instruct the debtor to use the exact approach Ransom urges. The word “applicable” is not necessary to accomplish that result; it is necessary only for the different purpose of dividing debtors eligible to make use of the tables from those who are not. Further, Ransom’s reading of “applicable” would sever the connection between the means test and the statutory provision it is meant to implement—the authorization of an allowance for (but only for) “reasonably necessary” expenses. Expenses that are wholly fictional are not easily thought of as reasonably necessary. And finally, Ransom’s interpretation would run counter to the statute’s overall purpose of ensuring that debtors repay creditors to the extent they can—here, by shielding some $28,000 that he does not in fact need for loan or lease payments. As against all this, Ransom argues that his reading is necessary to account for the means test’s distinction between “applicable” and “actual” expenses—more fully stated, between the phrase “applicable monthly expense amounts” specified in the Standards and the phrase “actual monthly expenses for … Other Necessary Expenses.” §707(b)(2)(A)(ii)(I) (emphasis added). The latter phrase enables a debtor to deduct his actual expenses in particular categories that the IRS designates relating mainly to taxpayers’ health and welfare. Internal Revenue Manual §5.15.1.10(1), http://www.irs.gov/irm/part5/ irm_05-015-001.html#d0e1381. According to Ransom, “applicable” cannot mean the same thing as “actual.” Brief for Petitioner 40. He thus concludes that “an ‘applicable’ expense can be claimed [under the means test] even if no ‘actual’ expense was incurred.” Ibid. Our interpretation of the statute, however, equally avoids conflating “applicable” with “actual” costs. Although the expense amounts in the Standards apply only if the debtor incurs the relevant expense, the debtor’s out-of-pocket cost may well not control the amount of the deduction. If a debtor’s actual expenses exceed the amounts listed in the tables, for example, the debtor may claim an allowance only for the specified sum, rather than for his real expenditures.[Footnote 8] For the Other Necessary Expense categories, by contrast, the debtor may deduct his actual expenses, no matter how high they are.[Footnote 9] Our reading of the means test thus gives full effect to “the distinction between ‘applicable’ and ‘actual’ without taking a further step to conclude that ‘applicable’ means ‘nonexistent.’ ” In re Ross-Tousey, 368 B. R. 762, 765 (Bkrtcy. Ct. ED Wis. 2007), rev’d, 549 F. 3d 1148 (CA7 2008). Finally, Ransom’s reading of “applicable” may not even answer the essential question: whether a debtor may claim a deduction. “[C]onsult[ing] the table[s] alone” to determine a debtor’s deduction, as Ransom urges us to do, Reply Brief for Petitioner 16, often will not be sufficient because the tables are not self-defining. This case provides a prime example. The “Ownership Costs” table features two columns labeled “First Car” and “Second Car.” See supra, at 4. Standing alone, the table does not specify whether it refers to the first and second cars owned (as Ransom avers), or the first and second cars for which the debtor incurs ownership costs (as FIA maintains)—and so the table does not resolve the issue in dispute.[Footnote 10] See In re Kimbro, 389 B. R. 518, 533 (Bkrtcy. App. Panel CA6 2008) (Fulton, J., dissenting) (“[O]ne cannot really ‘just look up’ dollar amounts in the tables without either referring to IRS guidelines for using the tables or imposing pre-existing assumptions about how [they] are to be navigated” (footnote omitted)). Some amount of interpretation is necessary to decide what the deduction is for and whether it is applicable to Ransom; and so we are brought back full circle to our prior analysis. B Ransom next argues that viewing the car-ownership deduction as covering no more than loan and lease payments is inconsistent with a separate sentence of the means test that provides: “Notwithstanding any other provision of this clause, the monthly expenses of the debtor shall not include any payments for debts.” §707(b)(2)(A)(ii)(I). The car-ownership deduction cannot comprise only loan and lease payments, Ransom contends, because those payments are always debts. See Brief for Petitioner 28, 44–45. Ransom ignores that the “notwithstanding” sentence governs the full panoply of deductions under the National and Local Standards and the Other Necessary Expense categories. We hesitate to rely on that general provision to interpret the content of the car-ownership deduction because Congress did not draft the former with the latter specially in mind; any friction between the two likely reflects only a lack of attention to how an across-the-board exclusion of debt payments would correspond to a particular IRS allowance.[Footnote 11] Further, the “notwithstanding” sentence by its terms functions only to exclude, and not to authorize, deductions. It cannot establish an allowance for non-loan or -lease ownership costs that no National or Local Standard covers. Accordingly, the “notwithstanding” sentence does nothing to alter our conclusion that the “Ownership Costs” table does not apply to a debtor whose car is not encumbered. C Ransom finally contends that his view of the means test is necessary to avoid senseless results not intended by Congress. At the outset, we note that the policy concerns Ransom emphasizes pale beside one his reading creates: His interpretation, as we have explained, would frustrate BAPCPA’s core purpose of ensuring that debtors devote their full disposable income to repaying creditors. See supra, at 8–9. We nonetheless address each of Ransom’s policy arguments in turn. Ransom first points out a troubling anomaly: Under our interpretation, “[d]ebtors can time their bankruptcy filing to take place while they still have a few car payments left, thus retaining an ownership deduction which they would lose if they filed just after making their last payment.” Brief for Petitioner 54. Indeed, a debtor with only a single car payment remaining, Ransom notes, is eligible to claim a monthly ownership deduction. Id., at 15, 52. But this kind of oddity is the inevitable result of a standardized formula like the means test, even more under Ransom’s reading than under ours. Such formulas are by their nature over- and under-inclusive. In eliminating the pre-BAPCPA case-by-case adjudication of above-median-income debtors’ expenses, on the ground that it leant itself to abuse, Congress chose to tolerate the occasional peculiarity that a brighter-line test produces. And Ransom’s alternative reading of the statute would spawn its own anomalies—even placing to one side the fundamental strangeness of giving a debtor an allowance for loan or lease payments when he has not a penny of loan or lease costs. On Ransom’s view, for example, a debtor entering bankruptcy might purchase for a song a junkyard car—“an old, rusted pile of scrap metal [that would] si[t] on cinder blocks in his backyard,” In re Brown, 376 B. R. 601, 607 (Bkrtcy. Ct. SD Tex. 2007)—in order to deduct the $471 car-ownership expense and reduce his payment to creditors by that amount. We do not see why Congress would have preferred that result to the one that worries Ransom. That is especially so because creditors may well be able to remedy Ransom’s “one payment left” problem. If car payments cease during the life of the plan, just as if other financial circumstances change, an unsecured creditor may move to modify the plan to increase the amount the debtor must repay. See 11 U. S. C. §1329(a)(1). Ransom next contends that denying the ownership allowance to debtors in his position “sends entirely the wrong message, namely, that it is advantageous to be deeply in debt on motor vehicle loans, rather than to pay them off.” Brief for Petitioner 55. But the choice here is not between thrifty savers and profligate borrowers, as Ransom would have it. Money is fungible: The $14,000 that Ransom spent to purchase his Camry outright was money he did not devote to paying down his credit card debt, and Congress did not express a preference for one use of these funds over the other. Further, Ransom’s argument mistakes what the deductions in the means test are meant to accomplish. Rather than effecting any broad federal policy as to saving or borrowing, the deductions serve merely to ensure that debtors in bankruptcy can afford essential items. The car-ownership allowance thus safeguards a debtor’s ability to retain a car throughout the plan period. If the debtor already owns a car outright, he has no need for this protection. Ransom finally argues that a debtor who owns his car free and clear may need to replace it during the life of the plan; “[g]ranting the ownership cost deduction to a vehicle that is owned outright,” he states, “accords best with economic reality.” Id., at 52. In essence, Ransom seeks an emergency cushion for car owners. But nothing in the statute authorizes such a cushion, which all debtors presumably would like in the event some unexpected need arises. And a person who enters bankruptcy without any car at all may also have to buy one during the plan period; yet Ransom concedes that a person in this position cannot claim the ownership deduction. Tr. of Oral Arg. 20. The appropriate way to account for unanticipated expenses like a new vehicle purchase is not to distort the scope of a deduction, but to use the method that the Code provides for all Chapter 13 debtors (and their creditors): modification of the plan in light of changed circumstances. See §1329(a)(1); see also supra, at 17. IV Based on BAPCPA’s text, context, and purpose, we hold that the Local Standard expense amount for transportation “Ownership Costs” is not “applicable” to a debtor who will not incur any such costs during his bankruptcy plan. Because the “Ownership Costs” category covers only loan and lease payments and because Ransom owns his car free from any debt or obligation, he may not claim the allowance. In short, Ransom may not deduct loan or lease expenses when he does not have any. We therefore affirm the judgment of the Ninth Circuit. It is so ordered. Footnote 1 Chapter 13 borrows the means test from Chapter 7, where it is used as a screening mechanism to determine whether a Chapter 7 proceeding is appropriate. Individuals who file for bankruptcy relief under Chapter 7 liquidate their nonexempt assets, rather than dedicate their future income, to repay creditors. See 11 U. S. C. §§704(a)(1), 726. If the debtor’s Chapter 7 petition discloses that his disposable income as calculated by the means test exceeds a certain threshold, the petition is presumptively abusive. §707(b)(2)(A)(i). If the debtor cannot rebut the presumption, the court may dismiss the case or, with the debtor’s consent, convert it into a Chapter 13 proceeding. §707(b)(1). Footnote 2 The National Standards designate allowances for six categories of expenses: (1) food; (2) housekeeping supplies; (3) apparel and services; (4) personal care products and services; (5) out-of-pocket health care costs; and (6) miscellaneous expenses. Internal Revenue Manual §5.15.1.8 (Oct. 2, 2009), http://www.irs.gov/irm/part5/irm_05-015-001.html#d0e1012 (all Internet materials as visited Jan. 7, 2011, and available in Clerk of Court’s case file). The Local Standards authorize deductions for two kinds of expenses: (1) housing and utilities; and (2) transportation. Id., §5.15.1.9. Footnote 3 Although both components of the transportation allowance are listed in the Local Standards, only the operating-cost expense amounts vary by geography; in contrast, the IRS provides a nationwide figure for ownership costs. Footnote 4 Compare In re Ransom, 577 F. 3d 1026, 1027 (CA9 2009) (case below), with In re Washburn, 579 F. 3d 934, 935 (CA8 2009) (permitting the allowance), In re Tate, 571 F. 3d 423, 424 (CA5 2009) (same), and In re Ross-Tousey, 549 F. 3d 1148, 1162 (CA7 2008) (same). The question has also divided bankruptcy courts. See, e.g., In re Canales, 377 B. R. 658, 662 (Bkrtcy. Ct. CD Cal. 2007) (citing dozens of cases reaching opposing results). Footnote 5 This interpretation also avoids the anomalous result of granting preferential treatment to individuals with above-median income. Because the means test does not apply to Chapter 13 debtors whose incomes are below the median, those debtors must prove on a case-by-case basis that each claimed expense is reasonably necessary. See §§1325(b)(2) and (3). If a below-median-income debtor cannot take a deduction for a nonexistent expense, we doubt Congress meant to provide such an allowance to an above-median-income debtor—the very kind of debtor whose perceived abuse of the bankruptcy system inspired Congress to enact the means test. Footnote 6 In addition, the IRS has categorized taxes, including those associated with car ownership, as an “Other Necessary Expens[e],” for which a debtor may take a deduction. See App. to Brief for Respondent 26a; Brief for United States as Amicus Curiae 16, n. 4. Footnote 7 Because the dissent appears to misunderstand our use of the Collection Financial Standards, and because it may be important for future cases to be clear on this point, we emphasize again that the statute does not “incorporat[e]” or otherwise “impor[t]” the IRS’s guidance. Post, at 1, 4 (opinion of Scalia, J.). The dissent questions what possible basis except incorporation could justify our consulting the IRS’s view, post, at 4, n., but we think that basis obvious: The IRS creates the National and Local Standards referenced in the statute, revises them as it deems necessary, and uses them every day. The agency might, therefore, have something insightful and persuasive (albeit not controlling) to say about them. Footnote 8 The parties and the Solicitor General as amicus curiae dispute the proper deduction for a debtor who has expenses that are lower than the amounts listed in the Local Standards. Ransom argues that a debtor may claim the specified expense amount in full regardless of his out-of-pocket costs. Brief for Petitioner 24–27. The Government concurs with this view, provided (as we require) that a debtor has some expense relating to the deduction. See Brief for United States as Amicus Curiae 19–21. FIA, relying on the IRS’s practice, contends to the contrary that a debtor may claim only his actual expenditures in this circumstance. Brief for Respondent 12, 45–46 (arguing that the Local Standards function as caps). We decline to resolve this issue. Because Ransom incurs no ownership expense at all, the car-ownership allowance is not applicable to him in the first instance. Ransom is therefore not entitled to a deduction under either approach. Footnote 9 For the same reason, the allowance for “applicable monthly expense amounts” at issue here differs from the additional allowances that the dissent cites for the deduction of actual expenditures. See post, at 3–4 (noting allowances for “actual expenses” for care of an elderly or chronically ill household member, §707(b)(2)(A)(ii)(II), and for home energy costs, §707(b)(2)(A)(ii)(V)). Footnote 10 The interpretive problem is not, as the dissent suggests, “whether to claim a deduction for one car or for two,” post, at 3, but rather whether to claim a deduction for any car that is owned if the debtor has no ownership costs. Indeed, if we had to decide this question on the basis of the table alone, we might well decide that a debtor who does not make loan or lease payments cannot claim an allowance. The table, after all, is titled “Ownership Costs”—suggesting that it applies to those debtors who incur such costs. And as noted earlier, the dollar amounts in the table represent average automobile loan and lease payments nationwide (with all other car-related expenses approximated in the separate “Operating Costs” table). See supra, at 9–10. Ransom himself concedes that not every debtor falls within the terms of this table; he would exclude, and thus prohibit from taking a deduction, a person who does not own a car. Brief for Petitioner 33. In like manner, the four corners of the table appear to exclude an additional group—debtors like Ransom who own their cars free and clear and so do not make the loan or lease payments that constitute “Ownership Costs.” Footnote 11 Because Ransom does not make payments on his car, we need not and do not resolve how the “notwithstanding” sentence affects the vehicle-ownership deduction when a debtor has a loan or lease expense. See Brief for United States as Amicus Curiae 23, n. 5 (offering alternative views on this question); Tr. of Oral Arg. 51–52.

563.US.401

The public disclosure bar of the False Claims Act (FCA) generally forecloses private parties from bringing qui tam suits to recover falsely or fraudulently obtained federal payments where those suits are “based upon the public disclosure of allegations or transactions in a criminal, civil, or administrative hearing, in a congressional, administrative, or Government Accounting Office report, hearing, audit, or investigation, or from the news media.” 31 U. S. C. §3730(e)(4)(A). Respondent Kirk brought such a suit, alleging that his former employer, petitioner Schindler Elevator Corp., had submitted hundreds of false claims for payment under its federal contracts. To support his allegations, Kirk pointed to information his wife received from the Labor Department (DOL) in response to three requests for records she filed under the Freedom of Information Act (FOIA), 5 U. S. C. §552. Granting Schindler’s motion to dismiss, the District Court concluded, inter alia, that the FCA’s public disclosure bar deprived it of jurisdiction over Kirk’s allegations that were based on information disclosed in a Government “report” or “investigation.” The Second Circuit vacated and remanded, holding, in effect, that an agency’s response to a FOIA request is neither a “report” nor an “investigation.” Held: A federal agency’s written response to a FOIA request for records constitutes a “report” within the meaning of the FCA’s public disclosure bar. Pp. 4–14. (a) “[R]eport” in this context carries its ordinary meaning. Pp. 4–8. (1) Because the FCA does not define “report,” the Court looks first to the word’s ordinary meaning. See, e.g., Gross v. FBL Financial Services, Inc., 557 U. S. ___, ___. Dictionaries define “report” as, for example, something that gives information. This ordinary meaning is consistent with the public disclosure bar’s generally broad scope, see, e.g., Graham County Soil and Water Conservation Dist. v. United States ex rel. Wilson, 559 U. S. ___, ___, as is evidenced by the other sources of public disclosure in §3730(e)(4)(A), especially “news media.” Pp. 4–6. (2) Nor is there any textual basis for adopting a narrower definition of “report.” The Second Circuit committed the very error this Court reversed in Graham County. In applying the noscitur a sociis canon to conclude that a narrower meaning for “report” was mandated, the court failed to consider all of the sources of public disclosure listed in the statute—in particular, the reference to “news media.” See 559 U. S., at ___. Applying the ordinary meaning of “report” also does not render superfluous the other sources of public disclosure in §3730(e)(4)(A). Pp. 6–8. (b) The DOL’s three written FOIA responses in this case, along with the accompanying records produced to Mrs. Kirk, are “reports” within the public disclosure bar’s ordinary meaning. FOIA requires each agency receiving a request to “notify the person making such request of [its] determination and the reasons therefor.” 5 U. S. C. §552(a)(6)(A)(i). Like other federal agencies, the DOL has adopted FOIA regulations mandating a written response. Such agency responses plainly fall within the broad, ordinary meaning of “report” as, e.g., something that gives information. Moreover, any records produced along with such responses are part of the responses, just as if they had been produced as an appendix to a printed report. Pp. 8–9. (c) This Court is not persuaded by assertions that it would be anomalous to read the public disclosure bar to encompass written FOIA responses. Pp. 9–14. (1) The Court’s holding is not inconsistent with the public disclosure bar’s drafting history. If anything, the drafting history supports this Court’s holding. Kirk’s case seems a classic example of the “opportunistic” litigation that the public disclosure bar is designed to discourage. Id., at ___. Anyone could identify a few regulatory filing and certification requirements, submit FOIA requests until he discovers a federal contractor who is out of compliance, and potentially reap a windfall in a qui tam action under the FCA. Pp. 9–11. (2) Nor will extending the public disclosure bar to written FOIA responses necessarily lead to unusual consequences. Kirk argues that the Court’s ruling would allow a suit by a qui tam relator possessing records whose release was required by FOIA even absent a request, but bar an action by a relator who got the same documents by way of a FOIA request. Even assuming, as Kirk does, that unrequested records are not covered by the public disclosure bar, the Court is not troubled by the different treatment. By its plain terms, the bar applies to some methods of public disclosure and not to others. See Graham County, 559 U. S., at ___. It would not be anomalous if some methods of FOIA disclosure fell within the bar’s scope and some did not. Moreover, Kirk’s assertion that potential defendants will now insulate themselves from liability by making a FOIA request for incriminating documents is pure speculation. Cf. id., at ___. There is no suggestion that this has occurred in those Circuits that have long held that FOIA responses are “reports” within the public disclosure bar’s meaning. Pp. 11–13. (3) Even if the foregoing extratextual arguments were accepted, Kirk and his amici have provided no principled way to define “report” to exclude FOIA responses without excluding other documents—e.g., the Justice Department’s annual report of FOIA statistics—that are indisputably reports. Pp. 13–14. (d) Whether Kirk’s suit is “based upon … allegations or transactions” disclosed in the reports at issue is a question to be resolved on remand. P. 14. 601 F. 3d 94, reversed and remanded. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, and Alito, JJ., joined. Ginsburg, J., filed a dissenting opinion, in which Breyer and Sotomayor, JJ., joined. Kagan, J., took no part in the consideration or decision of the case.

The False Claims Act (FCA), 31 U. S. C. §§3729–3733, prohibits submitting false or fraudulent claims for payment to the United States, §3729(a), and authorizes qui tam suits, in which private parties bring civil actions in the Government’s name, §3730(b)(1). This case concerns the FCA’s public disclosure bar, which generally forecloses qui tam suits that are “based upon the public disclosure of allegations or transactions … in a congressional, administrative, or Government Accounting Office report, hearing, audit, or investigation.” §3730(e)(4)(A) (footnote omitted).[Footnote 1] We must decide whether a federal agency’s written response to a request for records under the Freedom of Information Act (FOIA), 5 U. S. C. §552, constitutes a “report” within the meaning of the public disclosure bar. We hold that it does. I Petitioner Schindler Elevator Corporation manufactures, installs, and services elevators and escalators.[Footnote 2] In 1989, Schindler acquired Millar Elevator Industries, Inc., and the two companies merged in 2002. Since 1999, Schindler and the United States have entered into hundreds of contracts that are subject to the Vietnam Era Veterans’ Readjustment Assistance Act of 1972 (VEVRAA). That Act requires contractors like Schindler to report certain information to the Secretary of Labor, including how many of its employees are “qualified covered veterans” under the statute. 38 U. S. C. §4212(d)(1). VEVRAA regulations required Schindler to agree in each of its contracts that it would “submit VETS–100 Reports no later than September 30 of each year.” 48 CFR §52.222–37(c) (2008); see also §22.1310(b). Respondent Daniel Kirk, a United States Army veteran who served in Vietnam, was employed by Millar and Schindler from 1978 until 2003. In August 2003, Kirk resigned from Schindler in response to what he saw as Schindler’s efforts to force him out.[Footnote 3] In March 2005, Kirk filed this action against Schindler under the False Claims Act, which imposes civil penalties and treble damages on persons who submit false or fraudulent claims for payment to the United States. 31 U. S. C. §3729(a). The FCA authorizes both civil actions by the Attorney General and private qui tam actions to enforce its provisions. §3730. When, as here, the Government chooses not to intervene in a qui tam action, the private relator stands to receive between 25% and 30% of the proceeds of the action. §3730(d)(2). In an amended complaint filed in June 2007, Kirk alleged that Schindler had submitted hundreds of false claims for payment under its Government contracts. According to Kirk, Schindler had violated VEVRAA’s reporting requirements by failing to file certain required VETS–100 reports and including false information in those it did file. The company’s claims for payment were false, Kirk alleged, because Schindler had falsely certified its compliance with VEVRAA. Kirk did not specify the amount of damages he sought on behalf of the United States, but he asserted that the value of Schindler’s VEVRAA-covered contracts exceeded $100 million. To support his allegations, Kirk pointed to information his wife, Linda Kirk, received from the Department of Labor (DOL) in response to three FOIA requests. Mrs. Kirk had sought all VETS–100 reports filed by Schindler for the years 1998 through 2006. The DOL responded by letter or e-mail to each request with information about the records found for each year, including years for which no responsive records were located. The DOL informed Mrs. Kirk that it found no VETS–100 reports filed by Schindler in 1998, 1999, 2000, 2002, or 2003. For the other years, the DOL provided Mrs. Kirk with copies of the reports filed by Schindler, 99 in all. Schindler moved to dismiss on a number of grounds, including that the FCA’s public disclosure bar deprived the District Court of jurisdiction. See §3730(e)(4)(A). The District Court granted the motion, concluding that most of Kirk’s allegations failed to state a claim and that the remainder were based upon the public disclosure of alle-gations or transactions in an administrative “report” or “investigation.” 606 F. Supp. 2d 448 (SDNY 2009). The Court of Appeals for the Second Circuit vacated and remanded. 601 F. 3d 94 (2010). The court effectively held that an agency’s response to a FOIA request is neither a “report” nor an “investigation” within the meaning of the FCA’s public disclosure bar. See id., at 103–111 (agreeing with United States ex rel. Haight v. Catholic Healthcare West, 445 F. 3d 1147 (CA9 2006), and disagreeing with United States ex rel. Mistick PBT v. Housing Auth. of Pittsburgh, 186 F. 3d 376 (CA3 1999)). We granted certiorari, 561 U. S. ___ (2010), and now reverse and remand. II Schindler argues that “report” in the FCA’s public disclosure bar carries its ordinary meaning and that the DOL’s written responses to Mrs. Kirk’s FOIA requests are therefore “reports.” We agree.[Footnote 4] A 1 Adopted in 1986, the FCA’s public disclosure bar provides: “No court shall have jurisdiction over an action under this section based upon the public disclosure of allegations or transactions in a criminal, civil, or administrative hearing, in a congressional, administrative, or Government Accounting Office report, hearing, audit, or investigation, or from the news media, unless the action is brought by the Attorney General or the person bringing the action is an original source of the information.” 31 U. S. C. §3730(e)(4)(A) (footnote omitted). Because the statute does not define “report,” we look first to the word’s ordinary meaning. See Gross v. FBL Financial Services, Inc., 557 U. S. ___, ___ (2009) (slip op., at 7) (“Statutory construction must begin with the language employed by Congress and the assumption that the ordinary meaning of that language accurately expresses the legislative purpose” (internal quotation marks omitted)); Asgrow Seed Co. v. Winterboer, 513 U. S. 179, 187 (1995) (“When terms used in a statute are undefined, we give them their ordinary meaning”). A “report” is “something that gives information” or a “notification,” Webster’s Third New International Dictionary 1925 (1986), or “[a]n official or formal statement of facts or proceedings,” Black’s Law Dictionary 1300 (6th ed. 1990). See also 13 Oxford English Dictionary 650 (2d ed. 1989) (“[a]n account brought by one person to another”); American Heritage Dictionary 1103 (1981) (“[a]n account or announcement that is prepared, presented, or delivered, usually in formal or organized form”); Random House Dictionary 1634 (2d ed. 1987) (“an account or statement describing in detail an event, situation, or the like”). This broad ordinary meaning of “report” is consistent with the generally broad scope of the FCA’s public disclosure bar. As we explained last Term, to determine the meaning of one word in the public disclosure bar, we must consider the provision’s “entire text,” read as an “integrated whole.” Graham County Soil and Water Conservation Dist. v. United States ex rel. Wilson, 559 U. S. ___, ___, ___, n. 12 (2010) (slip op., at 8, 12, n. 12); see also Tyler v. Cain, 533 U. S. 656, 662 (2001) (“We do not … construe the meaning of statutory terms in a vacuum”). The other sources of public disclosure in §3730(e)(4)(A), especially “news media,” suggest that the public disclosure bar provides “a broa[d] sweep.” Graham County, supra, at ___ (slip op., at 8). The statute also mentions “administrative hearings” twice, reflecting intent to avoid underinclusiveness even at the risk of redundancy. The phrase “allegations or transactions” in §3730(e) (4)(A) additionally suggests a wide-reaching public disclosure bar. Congress covered not only the disclosure of “allegations” but also “transactions,” a term that courts have recognized as having a broad meaning. See, e.g., Moore v. New York Cotton Exchange, 270 U. S. 593, 610 (1926) (“ ‘Transaction’ is a word of flexible meaning”); Hamilton v. United Healthcare of La., Inc., 310 F. 3d 385, 391 (CA5 2002) (“[T]he ordinary meaning of the term ‘transaction’ is a broad reference to many different types of business dealings between parties”). 2 Nor is there any textual basis for adopting a narrower definition of “report.” The Court of Appeals, in holding that FOIA responses were not “reports,” looked to the words “hearing, audit, or investigation,” and the phrase “criminal, civil, [and] administrative hearings.” It concluded that all of these sources “connote the synthesis of information in an investigatory context” to “serve some end of the government.” 601 F. 3d, at 107; cf. Brief for Respondent 30, n. 15 (“Each is part of the government’s ongoing effort to fight fraud”). Applying the noscitur a sociis canon, the Court of Appeals then determined that these “ ‘neighboring words’ ” mandated a narrower meaning for “report” than its ordinary meaning. 601 F. 3d, at 107. The Court of Appeals committed the very error we reversed in Graham County. Like the Fourth Circuit in that case, the Second Circuit here applied the noscitur a sociis canon only to the immediately surrounding words, to the exclusion of the rest of the statute. See 601 F. 3d, at 107, n. 6. We emphasized in Graham County that “all of the sources [of public disclosure] listed in §3730(e)(4)(A) provide interpretive guidance.” 559 U. S., at ___ (slip op., at 8). When all of the sources are considered, the reference to “news media”—which the Court of Appeals did not consider—suggests a much broader scope. Ibid. The Government similarly errs by focusing only on the adjectives “congressional, administrative, or [GAO],”[Footnote 5] which precede “report.” Brief for United States as Amicus Curiae 18. It contends that these adjectives suggest that the public disclosure bar applies only to agency reports “analogous to those that Congress and the GAO would issue or conduct.” Ibid. As we explained in Graham County, however, those three adjectives tell us nothing more than that a “report” must be governmental. See 559 U. S., at ___, n. 7 (slip op., at 7, n. 7). The governmental nature of the FOIA responses at issue is not disputed. Finally, applying the ordinary meaning of “report” does not render superfluous the other sources of public disclosure in §3730(e)(4)(A). Kirk argues that reading “report” to mean “something that gives information” would subsume the other words in the phrase “report, hearing, audit, or investigation.” Brief for Respondent 23. But Kirk admits that hearings, audits, and investigations are processes “to obtain information.” Ibid. (emphasis added). Those processes are thus clearly different from “something that gives information.” Moreover, the statute contemplates some redundancy: An “audit,” for example, will often be a type of “investigation.” We are not persuaded that we should adopt a “different, somewhat special meaning” of “report” over the word’s “primary meaning.” Muscarello v. United States, 524 U. S. 125, 130, 128 (1998). Indeed, we have cautioned recently against interpreting the public disclosure bar in a way inconsistent with a plain reading of its text. In Graham County, we rejected several arguments for construing the statute narrowly, twice emphasizing that the sole “touchstone” in the statutory text is “public disclosure.” 559 U. S., at ___, ___ (slip op., at 11, 19). We chose in that case simply to give the text its “most natura[l] read[ing],” id., at ___ (slip op., at 5), and we do so again here. B A written agency response to a FOIA request falls within the ordinary meaning of “report.” FOIA requires each agency receiving a request to “notify the person making such request of [its] determination and the reasons therefor.” 5 U. S. C. §552(a)(6)(A)(i). When an agency denies a request in whole or in part, it must additionally “set forth the names and titles or positions of each person responsible for the denial,” “make a reasonable effort to estimate the volume of any [denied] matter,” and “provide any such estimate to the person making the request.” §§552(a)(6)(C)(i), (F). The DOL has adopted more detailed regulations implementing FOIA and mandating a response in writing. See 29 CFR §70.21(a) (2009) (requiring written notice of the grant of a FOIA request and a description of the manner in which records will be disclosed); §§70.21(b)–(c) (requiring a “brief statement of the reason or reasons for [a] denial,” as well as written notification if a record “cannot be located or has been destroyed” (italics deleted)). So, too, have other federal agencies. See, e.g., 28 CFR §16.6 (2010) (Dept. of Justice); 43 CFR §2.21 (2009) (Dept. of Interior); 7 CFR §1.7 (2010) (Dept. of Agriculture). Such an agency response plainly is “something that gives information,” a “notification,” and an “official or formal statement of facts.” Any records the agency produces along with its written FOIA response are part of that response, “just as if they had been reproduced as an appendix to a printed report.” Mistick, 186 F. 3d, at 384, n. 5. Nothing in the public disclosure bar suggests that a document and its attachments must be disaggregated and evaluated individually. If an allegation or transaction is disclosed in a record attached to a FOIA response, it is disclosed “in” that FOIA response and, therefore, disclosed “in” a report for the purposes of the public disclosure bar.[Footnote 6] The DOL’s three written FOIA responses to Mrs. Kirk, along with their attached records, are thus reports within the meaning of the public disclosure bar. Each response was an “official or formal statement” that “[gave] information” and “notif[ied]” Mrs. Kirk of the agency’s resolution of her FOIA request. III A In interpreting a statute, “[o]ur inquiry must cease if the statutory language is unambiguous,” as we have found, and “ ‘the statutory scheme is coherent and consistent.’ ” Robinson v. Shell Oil Co., 519 U. S. 337, 340 (1997) (quoting United States v. Ron Pair Enterprises, Inc., 489 U. S. 235, 240 (1989)). We are not persuaded by assertions that it would be anomalous to read the public disclosure bar to encompass written FOIA responses. 1 The drafting history of the public disclosure bar does not contradict our holding. As originally enacted in 1863, the FCA placed no restriction on the sources from which a qui tam relator could acquire information on which to base a lawsuit. See Graham County, 559 U. S., at ___ (slip op., at 12). Accordingly, this Court upheld the recovery of a relator, even though the Government claimed that he had discovered the basis for his lawsuit by reading a federal criminal indictment. See United States ex rel. Marcus v. Hess, 317 U. S. 537 (1943). In response, Congress amended the statute to preclude such “parasitic” qui tam actions based on “evidence or information in the possession of the United States … at the time such suit was brought.” 559 U. S., at ___ (slip op., at 12–13) (internal quotation marks omitted). Then, in 1986, Congress replaced the so-called Government knowledge bar with the narrower public disclosure bar. Id., at ___ (slip op., at 13). The Court of Appeals concluded that it would be inconsistent with this drafting history to hold that written FOIA responses are reports. The court reasoned that doing so would “essentially resurrect, in a significant subset of cases, the government possession standard … repudiated in 1986.” 601 F. 3d, at 109. We disagree with the Court of Appeals’ conclusion. As a threshold matter, “the drafting history of the public disclosure bar raises more questions than it answers.” Graham County, supra, at ___ (slip op., at 14). In any event, it is hardly inconsistent with the drafting history to read the public disclosure bar as operating similarly to the Government knowledge bar in a “subset of cases.” 601 F. 3d, at 109. As we have observed, “[r]ather than simply repeal the Government knowledge bar,” the public disclosure bar was “an effort to strike a balance between encouraging private persons to root out fraud and stifling parasitic lawsuits.” 559 U. S., at ___ (slip op., at 13) (emphasis added). If anything, the drafting history supports our holding. The sort of case that Kirk has brought seems to us a classic example of the “opportunistic” litigation that the public disclosure bar is designed to discourage. Ibid. (internal quotation marks omitted). Although Kirk alleges that he became suspicious from his own experiences as a veteran working at Schindler, anyone could have filed the same FOIA requests and then filed the same suit. Similarly, anyone could identify a few regulatory filing and certification requirements, submit FOIA requests until he discovers a federal contractor who is out of compliance, and potentially reap a windfall in a qui tam action under the FCA. See Brief for Chamber of Commerce of the United States of America et al. as Amici Curiae 20 (“Government contractors … are required to submit certifications related to everything from how they dispose of hazardous materials to their affirmative action plans” (citing 40 U. S. C. §3142 and 29 U. S. C. §793)).[Footnote 7] 2 Nor will extending the public disclosure bar to written FOIA responses necessarily lead to unusual consequences. FOIA requires agencies to release some records even absent a request. See 5 U. S. C. §§552(a)(1), (2). Kirk argues that it would be strange that two relators could obtain copies of the same document but that only the relator who got the document in response to a FOIA request would find his case barred. This argument assumes that records released under FOIA, but not attached to a written FOIA response, do not fall within the public disclosure bar. We do not decide that question. But even assuming, as Kirk does, that such records are not covered by the public disclosure bar, we are not troubled by the different treatment. By its plain terms, the public disclosure bar applies to some meth- ods of public disclosure and not to others. See Graham County, supra, at ___ (slip op., at 4) (“[T]he FCA’s public disclosure bar … deprives courts of jurisdiction over qui tam suits when the relevant information has already entered the public domain through certain channels” (emphasis added)). It would not be anomalous if some methods of FOIA disclosure fell within the scope of the public disclosure bar and some did not. We also are not concerned that potential defendants will now insulate themselves from liability by making a FOIA request for incriminating documents. This argument assumes that the public disclosure of information in a written FOIA response forever taints that information for purposes of the public disclosure bar. But it may be that a relator who comes by that information from a different source has a legitimate argument that his lawsuit is not “based upon” the initial public disclosure. 31 U. S. C. §3730(e)(4)(A). That question has divided the Courts of Appeals, and we do not resolve it here. See Glaser v. Wound Care Consultants, Inc., 570 F. 3d 907, 915 (CA7 2009) (describing the split in authority). It may also be that such a relator qualifies for the “original source” exception.[Footnote 8] In any event, the notion that potential defendants will make FOIA requests to insulate themselves from liability is pure speculation. Cf. Graham County, 559 U. S., at ___ (slip op., at 19) (rejecting as “strained speculation” an argument that local governments will manipulate the public disclosure bar to escape liability). There is no suggestion that this has occurred in those Circuits that have long held that FOIA responses are “reports” within the meaning of the public disclosure bar. B Even if we accepted these extratextual arguments, Kirk and his amici have provided no principled way to define “report” to exclude FOIA responses without excluding other documents that are indisputably reports. The Government, for example, struggled to settle on a single definition. Compare Brief for United States as Amicus Curiae 19 (“report” must be read to “reflect a focus on situations in which the government is conducting, or has completed, some focused inquiry or analysis concerning the relevant facts”) with id., at 21 (“A FOIA response is not a ‘report’ … because the federal agency is not charged with uncovering the truth of any matter”), and Tr. of Oral Arg. 33 (“[T]he way to think about it is whether or not the agency … is engaging in a substantive inquiry into and a substantive analysis of information”). It is difficult to see how the Department of Justice’s “Annual Report” of FOIA statistics—something that is indisputably a Government report—would qualify under the latter two definitions. See Dept. of Justice, Freedom of Information Act An- nual Report, Fiscal Year 2010, http://www.justice.gov/oip/ annual_report/2010/cover.htm (as visited May 12, 2011, and available in Clerk of Court’s case file); see also Tr. of Oral Arg. 19 (Kirk conceding that the DOJ annual report is a report). And even if the first definition arguably encompasses that report, it would seem also to include FOIA responses, which convey the results of a Government agency’s “focused inquiry.” Kirk also was unable to articulate a workable definition. His various proposed definitions suffer the same deficiencies as the Government’s. Compare Brief for Respondent 27 and Tr. of Oral Arg. 17–18 with Brief for Respondent 34–39 and Tr. of Oral Arg. 23. Kirk’s first suggestion would exclude “a lot of things that are labeled … report,” id., at 22, and the second—the definition advanced by the Court of Appeals—would seem to include written FOIA responses, id., at 28–29. In the end, it appears that the “only argument is that FOIA is a different kind of mission”—“a special case.” Id., at 31. We see no basis for that distinction and adhere to the principle that undefined statutory terms carry their ordinary meaning. * * * The DOL’s three written FOIA responses in this case, along with the accompanying records produced to Mrs. Kirk, are reports within the meaning of the public dis-closure bar. Whether Kirk’s suit is “based upon … allegations or transactions” disclosed in those reports is a question for the Court of Appeals to resolve on remand. The judgment of the United States Court of Appeals for the Second Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 During the pendency of this case, the Patient Protection and Affordable Care Act, 124 Stat. 119, amended the public disclosure bar. Because the amendments are not applicable to pending cases, Graham County Soil and Water Conservation Dist. v. United States ex rel. Wilson, 559 U. S. ___, ___, n. 1 (2010) (slip op., at 1, n. 1), this opinion refers to the statute as it existed when the suit was filed. Footnote 2 The facts in this Part, which we must accept as true, are taken from the amended complaint and the filings submitted in opposition to Schindler’s motion to dismiss. Footnote 3 Kirk filed a complaint with the Department of Labor’s Office of Federal Contract Compliance Programs (OFCCP), claiming that he had been “improperly demoted and constructively terminated by Schindler despite his status as a Vietnam era veteran.” App. 23a. The OFCCP investigated Schindler’s compliance with VEVRAA and found insufficient evidence to support Kirk’s claim. In November 2009, the Department of Labor affirmed the OFCCP’s finding. 601 F. 3d 94, 99 (CA2 2010). Footnote 4 Because we conclude that a written response to a FOIA request qualifies as a “report” within the meaning of the public disclosure bar, we need not address whether an agency’s search in response to a FOIA request also qualifies as an “investigation.” Footnote 5 Although the statute refers to the “Government Accounting Office,” it is undisputed that Congress meant the General Accounting Office, also known as GAO and now renamed the Government Accountability Office. See Graham County, 559 U. S., at ___, n. 6 (slip op., at 6, n. 6). Footnote 6 It is irrelevant whether a particular record is itself a report. The attached records do not “becom[e]” reports, 601 F. 3d, at 109, but simply are part of a report. Footnote 7 There is no merit to the suggestion that the public disclosure bar is intended only to exclude qui tam suits that “ride the investigatory coattails of the government’s own processes.” Brief for Taxpayers Against Fraud Education Fund as Amicus Curiae 25, 26; see Graham County, 559 U. S., at ___ (slip op., at 19) (rejecting the argument that the public disclosure bar applies only to allegations or transactions that “have landed on the desk of a DOJ lawyer”). Footnote 8 An “original source” is “an individual who has direct and independent knowledge of the information on which the allegations are based and has voluntarily provided the information to the Government before filing an action under this section which is based on the information.” §3730(e)(4)(B). Some Courts of Appeals have narrowly construed the exception to limit “original sources” to those who were the cause of the public disclosure, while others have been more generous. See United States ex rel. Duxbury v. Ortho Biotech Prods., L. P., 579 F. 3d 13, 22 (CA1 2009) (describing a three-way split among the Courts of Appeals). That question is not before us, and we do not decide it.

562.US.521

District Attorney’s Office for Third Judicial Dist. v. Osborne, 557 U. S. ___, ___, left unresolved the question whether a convicted state prisoner seeking DNA testing of crime-scene evidence may assert that claim in a civil rights action under 42 U. S. C. §1983 or may assert the claim in federal court only in a petition for a writ of habeas corpus under 28 U. S. C. §2254. A Texas jury convicted petitioner Skinner and sentenced him to death for murdering his girlfriend and her sons. He claimed that a potent alcohol and drug mix rendered him physically unable to commit the brutal murders, and he identified his girlfriend’s uncle as the likely perpetrator. In preparation for trial, the State tested some of the physical evidence, but left untested several items, including knives found on the premises, an axe handle, vaginal swabs, fingernail clippings, and certain hair samples. More than six years later, Texas enacted Article 64, which allows prisoners to gain postconviction DNA testing in limited circumstances. Invoking Article 64, Skinner twice moved in state court for DNA testing of the untested biological evidence. Both motions were denied. The Texas Court of Criminal Appeals (CCA) affirmed the first denial of relief on the ground that Skinner had not shown, as required by Article 64.03(a)(2), that he “would not have been convicted if exculpatory results had been obtained through DNA testing.” The CCA affirmed the second denial of relief on the ground that Skinner had not shown, as required by Article 64.01(b)(1)(B), that the evidence was not previously tested “through no fault” on his part. Skinner next filed the instant federal action for injunctive relief under §1983, naming as defendant respondent Switzer, the District Attorney who has custody of the evidence that Skinner would like to have tested. Skinner alleged that Texas violated his Fourteenth Amendment right to due process by refusing to provide for the DNA testing he requested. The Magistrate Judge recommended dismissal of the complaint for failure to state a claim, reasoning that postconviction requests for DNA evidence are cognizable only in habeas corpus, not under §1983. Adopting that recommendation, the District Court dismissed Skinner’s suit. The Fifth Circuit affirmed. Held: There is federal-court subject-matter jurisdiction over Skinner’s complaint, and the claim he presses is cognizable under §1983. Pp. 7–15. (a) Federal Rule of Civil Procedure 8(a)(2) generally requires only a plausible “short and plain” statement of the plaintiff’s claim, not an exposition of his legal argument. Skinner stated his due process claim in a paragraph alleging that the State’s refusal “to release the biological evidence for testing … deprived [him] of his liberty interests in utilizing state procedures to obtain reversal of his conviction and/or to obtain a pardon or reduction of his sentence … .” His counsel has clarified that Skinner does not challenge the prosecutor’s conduct or the CCA’s decisions; instead, he challenges Texas’ postconviction DNA statute “as construed” by the Texas courts. Pp. 7–8. (b) The Rooker-Feldman doctrine does not bar Skinner’s suit. This Court has applied the doctrine only in the two cases from which it takes its name, Rooker v. Fidelity Trust Co., 263 U. S. 413, District of Columbia Court of Appeals v. Feldman, 460 U. S. 462. See Exxon Mobil Corp. v. Saudi Basic Industries Corp., 544 U. S. 280. Given “the narrow ground” the doctrine occupies, id., at 284, the Court has confined Rooker-Feldman “to cases … brought by state-court losers … inviting district court review and rejection of [a state court’s] judgments.” Ibid. Skinner’s complaint encounters no Rooker-Feldman shoal. “If a federal plaintiff ‘present[s] [an] independent claim,’ ” it is not an impediment to the exercise of federal jurisdiction that the “same or a related question” was earlier aired between the parties in state court. Id., at 292–293. A state-court decision is not reviewable by lower federal courts, but a statute or rule governing the decision may be challenged in a federal action. See, e.g., Feldman, 460 U. S., at 487. Because Skinner’s federal case—which challenges not the adverse state-court decisions but the Texas statute they authoritatively construed—falls within the latter category, there was no lack of subject-matter jurisdiction over his federal suit. Pp. 8–10. (c) Measured against this Court’s prior holdings, Skinner has properly invoked §1983. This Court has several times considered when a state prisoner, complaining of unconstitutional state action, may pursue a civil rights claim under §1983, and when habeas corpus is the prisoner’s sole remedy. The pathmarking decision, Heck v. Humphrey, 512 U. S. 477, concerned a state prisoner who brought a §1983 action for damages, alleging that he had been unlawfully investigated, arrested, tried, and convicted. This Court held that §1983 was not an available remedy because any award in the plaintiff’s favor would “necessarily imply” the invalidity of his conviction. See id., at 487. In contrast, in Wilkinson v. Dotson, 544 U. S. 74, the Court held that prisoners who challenged the constitutionality of administrative decisions denying them parole eligibility, could proceed under §1983, for they sought no “injunction ordering … immediate or speedier release into the community,” id., at 82, and “a favorable judgment [would] not ‘necessarily imply’ the invalidity of [their] conviction[s] or sentence[s],” ibid. Here, success in Skinner’s suit for DNA testing would not “necessarily imply” the invalidity of his conviction. Test results might prove exculpatory, but that outcome is hardly inevitable, for those results could also prove inconclusive or incriminating. Switzer argues that, although Skinner’s immediate aim is DNA testing, his ultimate aim is to use the test results as a platform for attacking his conviction. But she has found no case in which the Court has recognized habeas as the sole remedy where the relief sought would not terminate custody, accelerate the date of release, or reduce the custody level. Contrary to the fears of Switzer and her amici, in the Circuits that currently allow §1983 claims for DNA testing, there has been no flood of litigation seeking postconviction discovery of evidence associated with the questions of guilt or punishment. The projected toll on federal courts is all the more implausible regarding DNA testing claims, for Osborne has rejected substantive due process as a basis for such claims. More generally, in the Prison Litigation Reform Act of 1995, Congress has placed constraints on prisoner suits in order to prevent sportive federal-court filings. Nor is there cause for concern that the instant ruling will spill over to claims relying on Brady v. Maryland, 373 U. S. 83. Brady, which announced a constitutional requirement addressed to the prosecution’s conduct pretrial, proscribes withholding evidence “favorable to an accused” and “material to [his] guilt or to punishment.” Cone v. Bell, 556 U. S. ___, ___. Unlike DNA testing, which may yield exculpatory, incriminating, or inconclusive results, a successful Brady claim necessarily yields evidence undermining a conviction: Brady claims therefore rank within the traditional core of habeas corpus and outside the province of §1983. Pp. 10–14. (d) Switzer’s several arguments why Skinner’s complaint should fail for lack of merit, unaddressed by the courts below, are ripe for consideration on remand. P. 14. 363 Fed. Appx. 302, reversed and remanded. Ginsburg, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Breyer, Sotomayor, and Kagan, JJ., joined. Thomas, J., filed a dissenting opinion, in which Kennedy and Alito, JJ., joined.

We granted review in this case to decide a question presented, but left unresolved, in District Attorney’s Office for Third Judicial Dist. v. Osborne, 557 U. S. ___, ___ (2009) (slip op., at 12–13): May a convicted state prisoner seeking DNA testing of crime-scene evidence assert that claim in a civil rights action under 42 U. S. C. §1983, or is such a claim cognizable in federal court only when asserted in a petition for a writ of habeas corpus under 28 U. S. C. §2254? The Courts of Appeals have returned diverse responses. Compare McKithen v. Brown, 481 F. 3d 89, 99 (CA2 2007) (claim seeking DNA testing is cognizable under §1983); Savory v. Lyons, 469 F. 3d 667, 669 (CA7 2006) (same); Bradley v. Pryor, 305 F. 3d 1287, 1290–1291 (CA11 2002) (same), with Harvey v. Horan, 278 F. 3d 370, 375 (CA4 2002) (claim is not cognizable under §1983) and Kutzner v. Montgomery County, 303 F. 3d 339, 341 (CA5 2002) (per curiam) (same). In Wilkinson v. Dotson, 544 U. S. 74 (2005), we comprehensively surveyed this Court’s decisions on the respective provinces of §1983 civil rights actions and §2254 federal habeas petitions. Habeas is the exclusive remedy, we reaffirmed, for the prisoner who seeks “immediate or speedier release” from confinement. Id., at 82. Where the prisoner’s claim would not “necessarily spell speedier release,” however, suit may be brought under §1983. Ibid. Adhering to our opinion in Dotson, we hold that a postconviction claim for DNA testing is properly pursued in a §1983 action. Success in the suit gains for the prisoner only access to the DNA evidence, which may prove exculpatory, inculpatory, or inconclusive. In no event will a judgment that simply orders DNA tests “necessarily impl[y] the unlawfulness of the State’s custody.” Id., at 81. We note, however, that the Court’s decision in Osborne severely limits the federal action a state prisoner may bring for DNA testing. Osborne rejected the extension of substantive due process to this area, 557 U. S., at ___ (slip op., at 19), and left slim room for the prisoner to show that the governing state law denies him procedural due process, see id., at ___ (slip op., at 18). I In 1995, a Texas jury convicted petitioner Henry Skinner and sentenced him to death for murdering his live-in girlfriend, Twila Busby, and her two sons. Busby was bludgeoned and choked with an axe handle and her sons were stabbed to death; the murders were committed in the house Busby shared with Skinner. Skinner never denied his presence in the house when the killings occurred. He claimed, however, that he was incapacitated by large quantities of alcohol and codeine. The potent alcohol and drug mix, Skinner maintained at trial, rendered him physically unable to commit the brutal murders charged against him. Skinner identified, as a likely perpetrator, Busby’s uncle, Robert Donnell (now deceased), an ex-convict with a history of physical and sexual abuse.[Footnote 1] On direct appeal, the Texas Court of Criminal Appeals (CCA) affirmed Skinner’s conviction and sentence. Skinner v. State, 956 S. W. 2d 532, 546 (1997). The CCA’s opinion described the crime-scene evidence in detail: “As they approached the house … , the police noticed a trail of blood spots on the ground running from the front porch to the fence line. There was a blood smear on the glass storm door and a knife on the front porch. Upon entering the residence, the police found Twila’s dead body on the living room floor… . An ax handle stained with blood and hair was leaning against the couch near her body and a black plastic trash bag containing a knife and a towel with wet brownish stains on it was laying between the couch and the coffee table. “[One officer] proceeded to the bedroom where [Busby’s two sons] usually slept in bunk beds. [The officer] found [one] dead body laying face down on the upper bunk, covered by a blood spotted blanket… . A door leading out of the bedroom and into a utility room yielded further evidence. [He] noticed a bloody handprint located about 24 inches off the floor on the frame of this door. He also noted a bloody handprint on the door knob of the door leading from the kitchen to the utility room and a handprint on the knob of the door exiting from the utility room into the backyard. “[When] police arrested [Skinner] … [t]hey found him standing in a closet wearing blood-stained socks and blood-stained blue jeans.” Id., at 536. Investigators also retained vaginal swabs taken from Busby. In preparation for trial, “the State tested the blood on [Skinner’s] clothing, blood and hair from a blanket that partially covered one of the victims, and hairs on one of the victim’s back and cheeks.” Skinner v. State, 122 S. W. 3d 808, 810 (Tex. Crim. App. 2003). The State also tested fingerprint evidence. Some of this evidence—including bloody palm prints in the room where one victim was killed—implicated Skinner, but “fingerprints on a bag containing one of the knives” did not. Ibid. Items left untested included the knives found on the premises, the axe handle, vaginal swabs, fingernail clippings, and additional hair samples. See ibid.[Footnote 2] In the decade following his conviction, Skinner unsuccessfully sought state and federal postconviction relief. See Skinner v. Quarterman, 576 F. 3d 214 (CA5 2009), cert. denied, 559 U. S. ___ (2010). He also pursued informal efforts to gain access to untested biological evidence the police had collected at the scene of the crime.[Footnote 3] In 2001, more than six years after Skinner’s conviction, Texas enacted Article 64, a statute allowing prisoners to gain postconviction DNA testing in limited circumstances. Tex. Code Crim. Proc. Ann., Art. 64.01(a) (Vernon Supp. 2010). To obtain DNA testing under Article 64, a prisoner must meet one of two threshold criteria. He may show that, at trial, testing either was “not available” or was “available, but not technologically capable of providing probative results.” Art. 64.01(b)(1)(A). Alternatively, he may show that the evidence was not previously tested “through no fault” on his part, and that “the interests of justice” require a postconviction order for testing. Art. 64.01(b)(1)(B). To grant a motion for postconviction testing, a court must make further findings, prime among them, the movant “would not have been convicted if exculpatory results had been obtained through DNA testing,” and “the [Article 64] request … [was] not made to unreasonably delay the execution of sentence or administration of justice.” Art. 64.03(a)(2). Invoking Article 64, Skinner twice moved in state court, first in 2001 and again in 2007, for DNA testing of yet untested biological evidence. See supra, at 4, n. 3. Both motions were denied. Affirming the denial of Skinner’s first motion, the CCA held that he had failed to demonstrate a “reasonable probability … that he would not have been … convicted if the DNA test results were exculpatory.” Skinner v. State, 122 S. W. 3d, at 813. Skinner’s second motion was bolstered by discovery he had obtained in the interim.[Footnote 4] The CCA again affirmed the denial of relief under Article 64, this time on the ground that Skinner failed to meet the “no fault” requirement. See Skinner v. State, 293 S. W. 3d 196, 200 (2009).[Footnote 5] During postconviction proceedings, the CCA noted, trial counsel testified that he had not “ask[ed] for testing because he was afraid the DNA would turn out to be [Skinner’s].” Id., at 202. That decision, the CCA concluded, constituted “a reasonable trial strategy” that the court had no cause to second-guess. Id., at 209. Skinner next filed the instant federal action for injunctive relief under §1983, naming as defendant respondent Lynn Switzer, the District Attorney whose office prosecuted Skinner and has custody of the evidence Skinner would like to have DNA tested. Skinner’s federal-court complaint alleged that Texas violated his Fourteenth Amendment right to due process by refusing to provide for the DNA testing he requested. Complaint ¶33, App. 20–21. The Magistrate Judge recommended dismissal of the complaint for failure to state a claim upon which relief can be granted. App. 24–41. Under the governing Circuit precedent, Kutzner v. Montgomery County, 303 F. 3d 339, the Magistrate Judge observed, postconviction requests for DNA evidence are cognizable only in habeas corpus, not under §1983. App. 39. Adopting the Magistrate Judge’s recommendation, the District Court dismissed Skinner’s suit. Id., at 44–45. On appeal, the United States Court of Appeals for the Fifth Circuit affirmed, 363 Fed. Appx. 302 (2010) (per curiam), reiterating that “an action by a prisoner for post-conviction DNA testing is not cognizable under §1983 and must instead be brought as a petition for writ of habeas corpus,” id., at 303. On Skinner’s petition,[Footnote 6] we granted certiorari, 560 U. S. ___ (2010), and now reverse the Fifth Circuit’s judgment. II A Because this case was resolved on a motion to dismiss for failure to state a claim, the question below was “not whether [Skinner] will ultimately prevail” on his procedural due process claim, see Scheuer v. Rhodes, 416 U. S. 232, 236 (1974), but whether his complaint was sufficient to cross the federal court’s threshold, see Swierkiewicz v. Sorema N. A., 534 U. S. 506, 514 (2002). Skinner’s complaint is not a model of the careful drafter’s art, but under the Federal Rules of Civil Procedure, a complaint need not pin plaintiff’s claim for relief to a precise legal theory. Rule 8(a)(2) of the Federal Rules of Civil Procedure generally requires only a plausible “short and plain” statement of the plaintiff’s claim, not an exposition of his legal argument. See 5 C. Wright & A. Miller, Federal Practice & Procedure §1219, pp. 277–278 (3d ed. 2004 and Supp. 2010). Skinner stated his due process claim in a paragraph alleging that the State’s refusal “to release the biological evidence for testing … has deprived [him] of his liberty interests in utilizing state procedures to obtain reversal of his conviction and/or to obtain a pardon or reduction of his sentence … .” Complaint ¶33, App. 20–21. As earlier recounted, see supra, at 5–6, Skinner had twice requested and failed to obtain DNA testing under the only state-law procedure then available to him. See Complaint ¶¶22–31, App. 14–20.[Footnote 7] At oral argument in this Court, Skinner’s counsel clarified the gist of Skinner’s due process claim: He does not challenge the prosecutor’s conduct or the decisions reached by the CCA in applying Article 64 to his motions; instead, he challenges, as denying him procedural due process, Texas’ postconviction DNA statute “as construed” by the Texas courts. Tr. of Oral Arg. 56. See also id., at 52 (Texas courts, Skinner’s counsel argued, have “construed the statute to completely foreclose any prisoner who could have sought DNA testing prior to trial[,] but did not[,] from seeking testing” postconviction).[Footnote 8] The merits of Skinner’s federal-court complaint assailing the Texas statute as authoritatively construed, and particularly the vitality of his claim in light of Osborne, see supra, at 2—unaddressed by the District Court or the Fifth Circuit—are not ripe for review. We take up here only the questions whether there is federal-court subject-matter jurisdiction over Skinner’s complaint, and whether the claim he presses is cognizable under §1983. B Respondent Switzer asserts that Skinner’s challenge is “[j]urisdictionally [b]arred” by what has come to be known as the Rooker-Feldman doctrine. Brief for Respondent 48–49 (boldface deleted). In line with the courts below, we conclude that Rooker-Feldman does not bar Skinner’s suit. As we explained in Exxon Mobil Corp. v. Saudi Basic Industries Corp., 544 U. S. 280 (2005), the Rooker-Feldman doctrine has been applied by this Court only twice, i.e., only in the two cases from which the doctrine takes its name: first, Rooker v. Fidelity Trust Co., 263 U. S. 413 (1923), then 60 years later, District of Columbia Court of Appeals v. Feldman, 460 U. S. 462 (1983). Both cases fit this pattern: The losing party in state court[Footnote 9] filed suit in a U. S. District Court after the state proceedings ended, complaining of an injury caused by the state-court judgment and seeking federal-court review and rejection of that judgment. Alleging federal-question jurisdiction, the plaintiffs in Rooker and Feldman asked the District Court to overturn the injurious state-court judgment. We held, in both cases, that the District Courts lacked subject-matter jurisdiction over such claims, for 28 U. S. C. §1257 “vests authority to review a state court’s judgment solely in this Court.” See Exxon, 544 U. S., at 292. We observed in Exxon that the Rooker-Feldman doctrine had been construed by some federal courts “to extend far beyond the contours of the Rooker and Feldman cases.” Id., at 283. Emphasizing “the narrow ground” occupied by the doctrine, id., at 284, we clarified in Exxon that Rooker-Feldman “is confined to cases of the kind from which the doctrine acquired its name: cases brought by state-court losers . . . inviting district court review and rejection of [the state court’s] judgments.” Ibid. Skinner’s litigation, in light of Exxon, encounters no Rooker-Feldman shoal. “If a federal plaintiff ‘present[s] [an] independent claim,’ ” it is not an impediment to the exercise of federal jurisdiction that the “same or a related question” was earlier aired between the parties in state court. id., at 292–293 (quoting GASH Assocs. v. Rosemont, 995 F. 2d 726, 728 (CA7 1993); first alteration in original); see In re Smith, 349 Fed. Appx. 12, 18 (CA6 2009) (Sutton, J., concurring in part and dissenting in part) (a defendant’s federal challenge to the adequacy of state-law procedures for postconviction DNA testing is not within the “limited grasp” of Rooker-Feldman). As earlier noted, see supra, at 7–8, Skinner does not challenge the adverse CCA decisions themselves; instead, he targets as unconstitutional the Texas statute they authoritatively construed. As the Court explained in Feldman, 460 U. S., at 487, and reiterated in Exxon, 544 U. S., at 286, a state-court decision is not reviewable by lower federal courts, but a statute or rule governing the decision may be challenged in a federal action.[Footnote 10] Skinner’s federal case falls within the latter category. There was, therefore, no lack of subject-matter jurisdiction over Skinner’s federal suit.[Footnote 11] C When may a state prisoner, complaining of unconstitutional state action, pursue a civil rights claim under §1983, and when is habeas corpus the prisoner’s sole remedy? This Court has several times considered that question. Pathmarking here is Heck v. Humphrey, 512 U. S. 477 (1994). Plaintiff in that litigation was a state prisoner serving time for manslaughter. He brought a §1983 action for damages, alleging that he had been unlawfully investigated, arrested, tried, and convicted. Although the complaint in Heck sought monetary damages only, not release from confinement, we ruled that the plaintiff could not proceed under §1983. Any award in his favor, we observed, would “necessarily imply” the invalidity of his conviction. See id., at 487. When “a judgment in favor of the plaintiff would necessarily imply the invalidity of his conviction or sentence,” the Court held, §1983 is not an available remedy. Ibid. “But if . . . the plaintiff’s action, even if successful, will not demonstrate the invalidity of [his conviction or sentence], the [§1983] action should be allowed to proceed … .” Ibid. We summarized the relevant case law most recently in Wilkinson v. Dotson, 544 U. S. 74 (2005). That case involved prisoners who challenged the constitutionality of administrative decisions denying them parole eligibility. They could proceed under §1983, the Court held, for they sought no “injunction ordering … immediate or speedier release into the community,” id., at 82, and “a favorable judgment [would] not ‘necessarily imply the invalidity of [their] conviction[s] or sentence[s],’ ” ibid. (quoting Heck, 512 U. S., at 487; first alteration added). Measured against our prior holdings, Skinner has properly invoked §1983. Success in his suit for DNA testing would not “necessarily imply” the invalidity of his conviction. While test results might prove exculpatory, that outcome is hardly inevitable; as earlier observed, see supra, at 2, results might prove inconclusive or they might further incriminate Skinner. See Nelson v. Campbell, 541 U. S. 637, 647 (2004) (“[W]e were careful in Heck to stress the importance of the term ‘necessarily.’ ”).[Footnote 12] Respondent Switzer nevertheless argues, in line with Fifth Circuit precedent, see Kutzner, 303 F. 3d, at 341, that Skinner’s request for DNA testing must be pursued, if at all, in an application for habeas corpus, not in a §1983 action. The dissent echoes Switzer’s argument. See post, at 3. Although Skinner’s immediate plea is simply for an order requiring DNA testing, his ultimate aim, Switzer urges, is to use the test results as a platform for attacking his conviction. It suffices to point out that Switzer has found no case, nor has the dissent, in which the Court has recognized habeas as the sole remedy, or even an available one, where the relief sought would “neither terminat[e] custody, accelerat[e] the future date of release from custody, nor reduc[e] the level of custody.” Dotson, 544 U. S., at 86 (Scalia, J., concurring). Respondent Switzer and her amici forecast that a “vast expansion of federal jurisdiction … would ensue” were we to hold that Skinner’s complaint can be initiated under §1983. See Brief for National District Attorneys Association as Amicus Curiae 8. In particular, they predict a proliferation of federal civil actions “seeking postconviction discovery of evidence [and] other relief inescapably associated with the central questions of guilt or punishment.” Id., at 6. These fears, shared by the dissent, post, at 6, are unwarranted.[Footnote 13] In the Circuits that currently allow §1983 claims for DNA testing, see supra, at 1, no evidence tendered by Switzer shows any litigation flood or even rainfall. The projected toll on federal courts is all the more implausible regarding DNA testing claims, for Osborne has rejected substantive due process as a basis for such claims. See supra, at 2. More generally, in the Prison Litigation Reform Act of 1995 (PLRA), 110 Stat. 1321–66, Congress has placed a series of controls on prisoner suits, constraints designed to prevent sportive filings in federal court. See, e.g., PLRA §803(d) (adding 42 U. S. C. §1997e to create new procedures and penalties for prisoner lawsuits under §1983); PLRA §804(a)(3) (adding 28 U. S. C. §1915(b)(1) to require any prisoner proceeding in forma pauperis to pay the full filing fee out of a percentage of his prison trust account); PLRA §804(c)(3) (adding 28 U. S. C. §1915(f) to require prisoners to pay the full amount of any cost assessed against them out of their prison trust account); PLRA §804(d) (adding 28 U. S. C. §1915(g) to revoke, with limited exception, in forma pauperis privileges for any prisoner who has filed three or more lawsuits that fail to state a claim, or are malicious or frivolous). See also Crawford-El v. Britton, 523 U. S. 574, 596–597 (1998) (PLRA aims to “discourage prisoners from filing claims that are unlikely to succeed,” and statistics suggest that the Act is “having its intended effect”). Nor do we see any cause for concern that today’s ruling will spill over to claims relying on Brady v. Maryland, 373 U. S. 83 (1963); indeed, Switzer makes no such assertion. Brady announced a constitutional requirement addressed first and foremost to the prosecution’s conduct pretrial. Brady proscribes withholding evidence “favorable to an accused” and “material to [his] guilt or to punishment.” Cone v. Bell, 556 U. S. ___, ___ (2009) (slip op., at 1). To establish that a Brady violation undermines a conviction, a convicted defendant must make each of three showings: (1) the evidence at issue is “favorable to the accused, either because it is exculpatory, or because it is impeaching”; (2) the State suppressed the evidence, “either willfully or inadvertently”; and (3) “prejudice … ensued.” Strickler v. Greene, 527 U. S. 263, 281–282 (1999); see Banks v. Dretke, 540 U. S. 668, 691 (2004). Unlike DNA testing, which may yield exculpatory, incriminating, or inconclusive results, a Brady claim, when successful postconviction, necessarily yields evidence undermining a conviction: Brady evidence is, by definition, always favorable to the defendant and material to his guilt or punishment. See Strickler, 527 U. S., at 296. And parties asserting Brady violations postconviction generally do seek a judgment qualifying them for “immediate or speedier release” from imprisonment. See Dotson, 544 U. S., at 82. Accordingly, Brady claims have ranked within the traditional core of habeas corpus and outside the province of §1983. See Heck, 512 U. S., at 479, 490 (claim that prosecutors and an investigator had “ ‘knowingly destroyed’ evidence ‘which was exculpatory in nature and could have proved [petitioner’s] innocence’ ” cannot be maintained under §1983); Amaker v. Weiner, 179 F. 3d 48, 51 (CA2 1999) (“claim [that] sounds under Brady v. Maryland … does indeed call into question the validity of [the] conviction”); Beck v. Muskogee Police Dept., 195 F. 3d 553, 560 (CA10 1999) (same). III Finally, Switzer presents several reasons why Skinner’s complaint should fail for lack of merit. Those arguments, unaddressed by the courts below, are ripe for consideration on remand. “[M]indful that we are a court of review, not of first view,” Cutter v. Wilkinson, 544 U. S. 709, 718, n. 7 (2005), we confine this opinion to the matter on which we granted certiorari and express no opinion on the ultimate disposition of Skinner’s federal action. * * * For the reasons stated, the judgment of the Court of Appeals for the Fifth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Footnote 1 At trial, a defense witness testified that, on the evening of the killings, Busby had spurned Donnell’s “rude sexual advances.” Skinner v. State, 956 S. W. 2d 532, 535 (Tex. Crim. App. 1997). A neighbor related at a federal postconviction hearing that she observed Donnell, a day or two after the murders, thoroughly cleaning the carpets and inside of his pickup truck. See Skinner v. Quarterman, 528 F. 3d 336, 345 (CA5 2008). Footnote 2 After Skinner’s conviction, the State performed DNA tests on certain additional materials, but Skinner took no part in the selection of those materials or their testing. Skinner maintains that these ex parte tests were inconclusive. See Complaint ¶19, App. 12 (this “testing raised more questions than it answered”). But see Skinner v. State, 122 S. W. 3d 808, 811 (Tex. Crim. App. 2003) (some findings were “inculpatory”). Footnote 3 Skinner’s trial counsel, although aware that biological evidence remained untested, did not request further testing. Postconviction, Skinner sought DNA testing of vaginal swabs and finger nail clippings taken from Busby, blood and hairs on a jacket found next to Busby’s body, and biological material on knives and a dish towel recovered at the crime scene. Complaint ¶22, App. 14–15. Footnote 4 On the basis of discovery in a federal postconviction proceeding, an expert retained by Skinner concluded that Skinner, Busby, and her two sons could be excluded as sources of a hair collected from Busby’s right hand after the killings. See Record 190. See also Complaint ¶27, App. 18. Footnote 5 The District Attorney, in response to Skinner’s second motion, informed the Texas district court that “[t]o the best of the State’s information, knowledge, and belief, the items sought to be tested are still available for testing, the chain of custody is intact, and the items are in a condition to be tested although the State has not sought expert opinion in that regard.” Record 202. See also Complaint ¶29, App. 19. Footnote 6 The State of Texas scheduled Skinner’s execution for March 24, 2010. We granted Skinner’s application to stay his execution until further action of this Court. 559 U. S. ___ (2010). Footnote 7 He also persistently sought the State’s voluntary testing of the materials he identified. See Complaint ¶31, App. 20. Footnote 8 Unlike the petitioner in District Attorney’s Office for Third Judicial Dist. v. Osborne, 557 U. S. ___ (2009), who “attempt[ed] to sidestep state process through … a federal lawsuit,” id., at ___ (slip op., at 17), Skinner first resorted to state court, see supra, at 5–6. In this respect, Skinner is better positioned to urge in federal court “the inadequacy of the state-law procedures available to him in state postconviction relief.” Osborne, 557 U. S., at ___ (slip op., at 18). Footnote 9 The judgment assailed in Feldman was rendered by the District of Columbia Court of Appeals, equivalent for this purpose to a state’s highest court. Footnote 10 The Court further observed in Exxon Mobil Corp. v. Saudi Basic Industries Corp., 544 U. S. 280, 292–293 (2005), that “[w]hen there is parallel state and federal litigation,” state preclusion law may become decisive, but “[p]reclusion … is not a jurisdictional matter.” Footnote 11 Switzer asserts that Skinner could have raised his federal claim in the Article 64 proceeding. See Tr. of Oral Arg. 48. Even if that were so, “Rooker-Feldman is not simply preclusion by another name,” Lance v. Dennis, 546 U. S. 459, 466 (2006) (per curiam), and questions of preclusion unresolved below are “best left for full airing and decision on remand,” id., at 467 (Ginsburg, J., concurring). Footnote 12 The dissent would muddle the clear line Heck and Dotson drew, and instead would instruct district courts to resort to “first principles” each time a state prisoner files a §1983 claim in federal court. Post, at 2, 7. Footnote 13 Unlike the parole determinations at issue in Wilkinson v. Dotson, 544 U. S. 74 (2005), Switzer urges, claims like Skinner’s require inquiry into the State’s proof at trial and therefore lie at “the core of the criminal proceeding itself.” Tr. of Oral 41; see id., at 33–34. Dotson declared, however, in no uncertain terms, that when a prisoner’s claim would not “necessarily spell speedier release,” that claim does not lie at “the core of habeas corpus,” and may be brought, if at all, under §1983. 544 U. S., at 82 (majority opinion) (internal quotation marks omitted); see id., at 85–86 (Scalia, J., concurring). Whatever might be said of Switzer’s argument were we to recast our doctrine, Switzer’s position cannot be reconciled with the line our precedent currently draws. Nor can the dissent’s advocacy of a “retur[n] to first principles.” Post, at 7. Given the importance of providing clear guidance to the lower courts, “we again see no reason for moving the line our cases draw.” Dotson, 544 U. S., at 84.

564.US.2010_09-1205

Respondent (Bayer) moved in Federal District Court for an injunction ordering a West Virginia state court not to consider a motion for class certification filed by petitioners (Smith), who were plaintiffs in the state-court action. Bayer thought such an injunction warranted because, in a separate case, Bayer had persuaded the same Federal District Court to deny a similar class-certification motion that had been filed against Bayer by a different plaintiff, George McCollins. The District Court had denied McCollins’ certification motion under Fed. Rule Civ. Proc. 23. The court granted Bayer’s requested injunction against the state court proceedings, holding that its denial of certification in McCollins’ case precluded litigation of the certification issue in Smith’s case. The Court of Appeals for the Eighth Circuit affirmed. It first noted that the Anti-Injunction Act (Act) generally prohibits federal courts from enjoining state court proceedings. But it found that the Act’s relitigation exception authorized this injunction because ordinary rules of issue preclusion barred Smith from seeking certification of his proposed class. In so doing, the court concluded that Smith was invoking a State Rule, W. Va. Rule Civ. Proc. 23, that was sufficiently similar to the Federal Rule McCollins had invoked, such that the certification issues presented in the two cases were the same. The court further held that Smith, as an unnamed member of McCollins’ putative class action, could be bound by the judgment in McCollins’ case. Held: In enjoining the state court from considering Smith’s class certification request, the federal court exceeded its authority under the “relitigation exception” to the Act. Pp. 5–18. (a) Under that Act, a federal court “may not grant an injunction to stay proceedings in a State court except” in rare cases, when necessary to “protect or effectuate [the federal court’s] judgments.” 28 U. S. C. §2283. The Act’s “specifically defined exceptions,” Atlantic Coast Line R. Co. v. Locomotive Engineers, 398 U. S. 281, 286, “are narrow and are ‘not [to] be enlarged by loose statutory construction,’ ” Chick Kam Choo v. Exxon Corp., 486 U. S. 140, 146. Indeed, “[a]ny doubts as to the propriety of a federal injunction against state court proceedings should be resolved in favor of permitting the state courts to proceed.” Atlantic Coast Line R. Co., 398 U. S., at 297. The exception at issue in this case, known as the “relitigation exception,” authorizes an injunction to prevent state litigation of a claim or issue “that previously was presented to and decided by the federal court.” Chick Kam Choo, 486 U. S., at 147. This exception is designed to implement “well-recognized concepts” of claim and issue preclusion. Ibid. Because deciding whether and how prior litigation has preclusive effect is usually the bailiwick of the second court—here, the West Virginia court—every benefit of the doubt goes toward the state court, see Atlantic Coast Line, 398 U. S., at 287, 297; an injunction can issue only if preclusion is clear beyond peradventure. For the federal court’s class-action determination to preclude the state court’s adjudication of Smith’s motion, at least two conditions must be met. First, the issue the federal court decided must be the same as the one presented in the state tribunal. And second, Smith must have been a party to the federal suit or must fall within one of a few discrete exceptions to the general rule against binding nonparties. Pp. 5–7. (b) The issue the federal court decided was not the same as the one presented in the state tribunal. This case is little more than a rerun of Chick Kam Choo. There, a federal court dismissed a suit involving Singapore law on forum non conveniens grounds and then enjoined the plaintiff from pursuing the “same” claim in Texas state court. However, because the legal standards for forum non conveniens differed in the two courts, the issues before those courts differed, making an injunction unwarranted. Here, Smith’s proposed class mirrored McCollins’, and the two suits’ substantive claims broadly overlapped. But the federal court adjudicated McCollins’ certification motion under Federal Rule 23, whereas the state court was poised to consider Smith’s proposed class under W. Va. Rule 23. And the State Supreme Court has generally stated that it will not necessarily interpret its Rule 23 as coterminous with the Federal Rule. Absent clear evidence that the state courts had adopted an approach to State Rule 23 tracking the federal court’s analysis in McCollins’ case, this Court could not conclude that they would interpret their Rule the same way and, thus, could not tell whether the certification issues in the two courts were the same. That uncertainty would preclude an injunction. And indeed, the case against an injunction here is even stronger, because the State Supreme Court has expressly disapproved the approach to Rule 23(b)(3)’s predominance requirement embraced by the Federal District Court. Pp. 8–12. (c) The District Court’s injunction was independently improper because Smith was not a party to the federal suit and was not covered by any exception to the rule against nonparty preclusion. Generally, a party “is ‘[o]ne by or against whom a lawsuit is brought,’ ” United States ex rel. Eisenstein v. City of New York, 556 U. S. ___, ___, or who “become[s] a party by intervention, substitution, or third-party practice,” Karcher v. May, 484 U. S. 72, 77. The definition of “party” cannot be stretched so far as to cover a person like Smith, whom McCollins was denied leave to represent. The only exception to the rule against nonparty preclusion potentially relevant here is the exception that binds non-named members of “properly conducted class actions” to judgments entered in such proceedings. Taylor v. Sturgell, 553 U. S. 880, 894. But McCollins’ suit was not a proper class action. Indeed, the very ruling that Bayer argues should have preclusive effect is the District Court’s decision not to certify a class. Absent certification of a class under Federal Rule 23, the precondition for binding Smith was not met. Neither a proposed, nor a rejected, class action may bind nonparties. See id., at 901. Bayer claims that this Court’s approach to class actions would permit class counsel to try repeatedly to certify the same class simply by changing plaintiffs. But principles of stare decisis and comity among courts generally suffice to mitigate the sometimes substantial costs of similar litigation brought by different plaintiffs. The right approach does not lie in binding nonparties to a judgment. And to the extent class actions raise special relitigation problems, the federal Class Action Fairness Act of 2005 provides a remedy that does not involve departing from the usual preclusion rules. Pp. 12–18. 593 F. 3d 716, reversed. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Ginsburg, Breyer, Alito, and Sotomayor, JJ., joined, and in which Thomas, J., joined as to Parts I and II–A.

* In this case, a Federal District Court enjoined a state court from considering a plaintiff’s request to approve a class action. The District Court did so because it had earlier denied a motion to certify a class in a related case, brought by a different plaintiff against the same defendant alleging similar claims. The federal court thought its injunction appropriate to prevent relitigation of the issue it had decided. We hold to the contrary. In issuing this order to a state court, the federal court exceeded its authority under the “relitigation exception” to the Anti-Injunction Act. That statutory provision permits a federal court to enjoin a state proceeding only in rare cases, when necessary to “pro-tect or effectuate [the federal court’s] judgments.” 28 U. S. C. §2283. Here, that standard was not met for two reasons. First, the issue presented in the state court was not identical to the one decided in the federal tribunal. And second, the plaintiff in the state court did not have the requisite connection to the federal suit to be bound by the District Court’s judgment. I Because the question before us involves the effect of a former adjudication on this case, we begin our statement of the facts not with this lawsuit, but with another. In August 2001, George McCollins sued respondent Bayer Corporation in the Circuit Court of Cabell County, West Virginia, asserting various state-law claims arising from Bayer’s sale of an allegedly hazardous prescription drug called Baycol (which Bayer withdrew from the market that same month). McCollins contended that Bayer had violated West Virginia’s consumer-protection statute and the company’s express and implied warranties by selling him a defective product. And pursuant to West Virginia Rule of Civil Procedure 23 (2011), McCollins asked the state court to certify a class of West Virginia residents who had also purchased Baycol, so that the case could proceed as a class action. Approximately one month later, the suit now before us began in a different part of West Virginia. Petitioners Keith Smith and Shirley Sperlazza (Smith for short) filed state-law claims against Bayer, similar to those raised in McCollins’ suit, in the Circuit Court of Brooke County, West Virginia. And like McCollins, Smith asked the court to certify under West Virginia’s Rule 23 a class of Baycol purchasers residing in the State. Neither Smith nor McCollins knew about the other’s suit. In January 2002, Bayer removed McCollins’ case to the United States District Court for the Southern District of West Virginia on the basis of diversity jurisdiction. See 28 U. S. C. §§1332, 1441. The case was then transferred to the District of Minnesota pursuant to a preexisting order of the Judicial Panel on Multi-District Litigation, which had consolidated all federal suits involving Baycol (numbering in the tens of thousands) before a single District Court Judge. See §1407. Bayer, however, could not remove Smith’s case to federal court because Smith had sued several West Virginia defendants in addition to Bayer, and so the suit lacked complete diversity. See §1441(b).[Footnote 1] Smith’s suit thus remained in the state courthouse in Brooke County. Over the next six years, the two cases proceeded along their separate pretrial paths at roughly the same pace. By 2008, both courts were preparing to turn to their respective plaintiffs’ motions for class certification. The Federal District Court was the first to reach a decision. Applying Federal Rule of Civil Procedure 23,[Footnote 2] the District Court declined to certify McCollins’ proposed class of West Virginia Baycol purchasers. The District Court’s reasoning proceeded in two steps. The court first ruled that, under West Virginia law, each plaintiff would have to prove “actual injury” from his use of Baycol to recover. App. to Pet. for Cert. 44a. The court then held that because the necessary showing of harm would vary from plaintiff to plaintiff, “individual issues of fact predominate[d]” over issues common to all members of the proposed class, and so the case was not suitable for class treatment. Id., at 45a. In the same order, the District Court also dismissed McCollins’ claims on the merits in light of his failure to demonstrate physical injury from his use of Baycol. McCollins chose not to appeal. Although McCollins’ suit was now concluded, Bayer asked the District Court for another order based upon it, this one affecting Smith’s case in West Virginia. In a motion—receipt of which first apprised Smith of McCollins’ suit—Bayer explained that the proposed class in Smith’s case was identical to the one the federal court had just rejected. Bayer therefore requested that the federal court enjoin the West Virginia state court from hearing Smith’s motion to certify a class. According to Bayer, that order was appropriate to protect the District Court’s judgment in McCollins’ suit denying class certification. The District Court agreed and granted the injunction. The Court of Appeals for the Eighth Circuit affirmed. In re Baycol Prods. Litigation, 593 F. 3d 716 (2010). The court noted that the Anti-Injunction Act generally prohibits federal courts from enjoining state court proceedings. But the court held that the Act’s relitigation exception authorized the injunction here because ordinary rules of issue preclusion barred Smith from seeking certification of his proposed class. According to the court, Smith was invoking a similar class action rule as McCollins had used to seek certification “of the same class” in a suit alleging “the same legal theories,” id., at 724; the issue in the state court therefore was “sufficiently identical” to the one the federal court had decided to warrant preclusion, ibid. In addition, the court held, the parties in the two proceedings were sufficiently alike: Because Smith was an unnamed member of the class McCollins had proposed, and because their “interests were aligned,” Smith was appropriately bound by the federal court’s judgment. Ibid. We granted certiorari, 561 U. S. __ (2010), because the order issued here implicates two circuit splits arising from application of the Anti-Injunction Act’s relitigation exception. The first involves the requirement of preclusion law that a subsequent suit raise the “same issue” as a previous case.[Footnote 3] The second concerns the scope of the rule that a court’s judgment cannot bind nonparties.[Footnote 4] We think the District Court erred on both grounds when it granted the injunction, and we now reverse. II The Anti-Injunction Act, first enacted in 1793, provides that “A court of the United States may not grant an injunction to stay proceedings in a State court except as expressly authorized by Act of Congress, or where necessary in aid of its jurisdiction, or to protect or effectuate its judgments.” 28 U. S. C. §2283. The statute, we have recognized, “is a necessary concomitant of the Framers’ decision to authorize, and Congress’ decision to implement, a dual system of federal and state courts.” Chick Kam Choo v. Exxon Corp., 486 U. S. 140, 146 (1988). And the Act’s core message is one of respect for state courts. The Act broadly commands that those tribunals “shall remain free from interference by federal courts.” Atlantic Coast Line R. Co. v. Locomotive Engineers, 398 U. S. 281, 282 (1970). That edict is subject to only “three specifically defined exceptions.” Id., at 286. And those exceptions, though designed for important purposes, “are narrow and are ‘not [to] be enlarged by loose statutory construction.’ ” Chick Kam Choo, 486 U. S., at 146 (quoting Atlantic Coast Line, 398 U. S., at 287; alteration in original). Indeed, “[a]ny doubts as to the propriety of a federal injunction against state court proceedings should be resolved in favor of permitting the state courts to proceed.” Id., at 297. This case involves the last of the Act’s three exceptions, known as the relitigation exception. That exception is de-signed to implement “well-recognized concepts” of claim and issue preclusion. Chick Kam Choo, 486 U. S., at 147. The provision authorizes an injunction to prevent state litigation of a claim or issue “that previously was presented to and decided by the federal court.” Ibid. But in applying this exception, we have taken special care to keep it “strict and narrow.” Id., at 148. After all, a court does not usually “get to dictate to other courts the preclusion consequences of its own judgment.” 18 C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §4405, p. 82 (2d ed. 2002) (hereinafter Wright & Miller). Deciding whether and how prior litigation has preclusive effect is usually the bailiwick of the second court (here, the one in West Virginia). So issuing an injunction under the relitigation exception is resorting to heavy artillery.[Footnote 5] For that reason, every benefit of the doubt goes toward the state court, see Atlantic Coast Line, 398 U. S., at 287, 297; an injunction can issue only if preclusion is clear beyond peradventure. The question here is whether the federal court’s rejection of McCollins’ proposed class precluded a later adjudication in state court of Smith’s certification motion. For the federal court’s determination of the class issue to have this preclusive effect, at least two conditions must be met.[Footnote 6] First, the issue the federal court decided must be the same as the one presented in the state tribunal. See 18 Wright & Miller §4417, at 412. And second, Smith must have been a party to the federal suit, or else must fall within one of a few discrete exceptions to the general rule against binding nonparties. See 18A id., §4449, at 330. In fact, as we will explain, the issues before the two courts were not the same, and Smith was neither a party nor the ex-ceptional kind of nonparty who can be bound. So the courts below erred in finding the certification issue precluded, and erred all the more in thinking an injunction appropriate.[Footnote 7] A In our most recent case on the relitigation exception, Chick Kam Choo v. Exxon, we applied the “same issue” requirement of preclusion law to invalidate a federal court’s injunction. 486 U. S., at 151. The federal court had dismissed a suit involving Singapore law on grounds of forum non conveniens. After the plaintiff brought the same claim in Texas state court, the federal court issued an injunction barring the plaintiff from pursuing relief in that alternate forum. We held that the District Court had gone too far. “[A]n essential prerequisite for applying the relitigation exception,” we explained, “is that the … issues which the federal injunction insulates from litigation in state proceedings actually have been decided by the federal court.” Id., at 148. That prerequisite, we thought, was not satisfied because the issue to be adjudicated in state court was not the one the federal court had resolved. The federal court had considered the permissibility of the claim under federal forum non conveniens principles. But the Texas courts, we thought, “would apply a significantly different forum non conveniens analysis,” id., at 149; they had in prior cases rejected the strictness of the federal doctrine. Our conclusion followed: “[W]hether the Texas state courts are an appropriate forum for [the plaintiff’s] Singapore law claims has not yet been litigated.” Ibid. Because the legal standards in the two courts differed, the issues before the courts differed, and an injunction was unwarranted. The question here closely resembles the one in Chick Kam Choo. The class Smith proposed in state court mirrored the class McCollins sought to certify in federal court: Both included all Baycol purchasers resident in West Virginia. Moreover, the substantive claims in the two suits broadly overlapped: Both complaints alleged that Bayer had sold a defective product in violation of the State’s consumer protection law and the company’s warranties. So far, so good for preclusion. But not so fast: a critical question—the question of the applicable legal standard—remains. The District Court ruled that the proposed class did not meet the requirements of Federal Rule 23 (because individualized issues would predominate over common ones). But the state court was poised to consider whether the proposed class satisfied West Virginia Rule 23. If those two legal standards differ (as federal and state forum non conveniens law differed in Chick Kam Choo)—then the federal court resolved an issue not before the state court. In that event, much like in Chick Kam Choo, “whether the [West Virginia] state cour[t]” should certify the proposed class action “has not yet been litigated.” 486 U. S., at 149. The Court of Appeals and Smith offer us two competing ways of deciding whether the West Virginia and Federal Rules differ, but we think the right path lies somewhere in the middle. The Eighth Circuit relied almost exclusively on the near-identity of the two Rules’ texts. See 593 F. 3d, at 723. That was the right place to start, but not to end. Federal and state courts, after all, can and do apply identically worded procedural provisions in widely varying ways. If a State’s procedural provision tracks the language of a Federal Rule, but a state court interprets that provision in a manner federal courts have not, then the state court is using a different standard and thus deciding a different issue. See 18 Wright & Miller §4417, at 454 (stating that preclusion is “inappropriate” when “different legal standards … masquerad[e] behind similar legal labels”). At the other extreme, Smith contends that the source of law is all that matters: a different sovereign must in each and every case “have the opportunity, if it chooses, to construe its procedural rule differently.” Brief for Petitioners 22 (quoting ALI, Principles of the Law, Aggregate Litigation §2.11, Reporters’ Notes, cmt. b, p. 181 (2010)). But if state courts have made crystal clear that they follow the same approach as the federal court applied, we see no need to ignore that determination; in that event, the issues in the two cases would indeed be the same. So a federal court considering whether the relitigation exception applies should examine whether state law parallels its federal counterpart. But as suggested earlier, see supra, at 6, the federal court must resolve any uncertainty on that score by leaving the question of preclusion to the state courts. Under this approach, the West Virginia Supreme Court has gone some way toward resolving the matter before us by declaring its independence from federal courts’ interpretation of the Federal Rules—and particularly of Rule 23. In In re W. Va. Rezulin Litigation, 214 W. Va. 52, 585 S. E. 2d 52 (2003) (In re Rezulin), the West Virginia high court considered a plaintiff’s motion to certify a class—coincidentally enough, in a suit about an allegedly defective pharmaceutical product. The court made a point of complaining about the parties’ and lower court’s near-exclusive reliance on federal cases about Federal Rule 23 to decide the certification question. Such cases, the court cautioned, “ ‘may be persuasive, but [they are] not binding or controlling.’ ” Id., at 61, 585 S. E. 2d, at 61. And lest anyone mistake the import of this message, the court went on: The aim of “this rule is to avoid having our legal analysis of our Rules ‘amount to nothing more than Pavlovian responses to federal decisional law.’ ” Ibid. (italics omitted). Of course, the state courts might still have adopted an approach to their Rule 23 that tracked the analysis the federal court used in McCollins’ case. But absent clear evidence that the state courts had done so, we could not conclude that they would interpret their Rule in the same way. And if that is so, we could not tell whether the certification issues in the state and federal courts were the same. That uncertainty would preclude an injunction. But here the case against an injunction is even stronger, because the West Virginia Supreme Court has disapproved the approach to Rule 23(b)(3)’s predominance requirement that the Federal District Court embraced. Recall that the federal court held that the presence of a single individualized issue—injury from the use of Baycol—prevented class certification. See supra, at 3. The court did not identify the common issues in the case; nor did it balance these common issues against the need to prove individual injury to determine which predominated. The court instead applied a strict test barring class treatment when proof of each plaintiff’s injury is necessary.[Footnote 8] By contrast, the West Virginia Supreme Court in In re Rezulin adopted an all-things-considered, balancing inquiry in interpreting its Rule 23. Rejecting any “rigid test,” the state court opined that the predominance requirement “contemplates a review of many factors.” 214 W. Va., at 72, 585 S. E. 2d, at 72. Indeed, the court noted, a “ ‘single common issue’ ” in a case could outweigh “ ‘numerous … individual questions.’ ” Ibid. That meant, the court further explained (quoting what it termed the “leading treatise” on the subject), that even objections to certification “ ‘based on … causation, or reliance’ ”—which typically involve showings of individual injury—“ ‘will not bar predominance satisfaction.’ ” Ibid. (quoting 2 A. Conte & H. Newberg, Newberg on Class Actions §4.26, p. 241 (4th ed. 2002)). So point for point, the analysis set out in In re Rezulin diverged from the District Court’s interpretation of Federal Rule 23. A state court using the In re Rezulin standard would decide a different question than the one the federal court had earlier resolved.[Footnote 9] This case, indeed, is little more than a rerun of Chick Kam Choo. A federal court and a state court apply different law. That means they decide distinct questions. The federal court’s resolution of one issue does not preclude the state court’s determination of another. It then goes without saying that the federal court may not issue an injunction. The Anti-Injunction Act’s re-litigation exception does not extend nearly so far. B The injunction issued here runs into another basic premise of preclusion law: A court’s judgment binds only the parties to a suit, subject to a handful of discrete and limited exceptions. See, e.g., 18A Wright & Miller §4449, at 330. The importance of this rule and the narrowness of its exceptions go hand in hand. We have repeatedly “emphasize[d] the fundamental nature of the general rule” that only parties can be bound by prior judgments; accordingly, we have taken a “constrained approach to nonparty preclusion.” Taylor v. Sturgell, 553 U. S. 880, 898 (2008). Against this backdrop, Bayer defends the decision below by arguing that Smith—an unnamed member of a proposed but uncertified class—qualifies as a party to the McCollins litigation. See Brief for Respondent 32–34. Alternatively, Bayer claims that the District Court’s judgment binds Smith under the recognized exception to the rule against nonparty preclusion for members of class actions. See id., at 34–39. We think neither contention has merit. Bayer’s first claim ill-comports with any proper understanding of what a “party” is. In general, “[a] ‘party’ to litigation is ‘[o]ne by or against whom a lawsuit is brought,’ ” United States ex rel. Eisenstein v. City of New York, 556 U. S. ___, ___ (2009) (slip op., at 4), or one who “become[s] a party by intervention, substitution, or third-party practice,” Karcher v. May, 484 U. S. 72, 77 (1987). And we have further held that an unnamed member of a certified class may be “considered a ‘party’ for the [particular] purpos[e] of appealing” an adverse judgment. Devlin v. Scardelletti, 536 U. S. 1, 7 (2002). But as the dissent in Devlin noted, no one in that case was “willing to advance the novel and surely erroneous argument that a nonnamed class member is a party to the class-action litigation before the class is certified.” Id., at 16, n. 1 (opinion of Scalia, J.). Still less does that argument make sense once certification is denied. The definition of the term “party” can on no account be stretched so far as to cover a person like Smith, whom the plaintiff in a lawsuit was denied leave to represent.[Footnote 10] If the judgment in the McCollins litigation can indeed bind Smith, it must do so under principles of nonparty preclusion. As Bayer notes, see Brief for Respondent 37, one such principle allows unnamed members of a class action to be bound, even though they are not parties to the suit. See Cooper v. Federal Reserve Bank of Richmond, 467 U. S. 867, 874 (1984) (“[U]nder elementary principles of prior adjudication a judgment in a properly entertained class action is binding on class members in any subsequent litigation”); see also Taylor, 553 U. S., at 894 (stating that nonparties can be bound in “properly conducted class actions”). But here Bayer faces a conundrum. If we know one thing about the McCollins suit, we know that it was not a class action. Indeed, the very ruling that Bayer argues ought to be given preclusive effect is the District Court’s decision that a class could not properly be certified. So Bayer wants to bind Smith as a member of a class action (because it is only as such that a nonparty in Smith’s situation can be bound) to a determination that there could not be a class action. And if the logic of that position is not immediately transparent, here is Bayer’s attempt to clarify: “[U]ntil the moment when class certi-fication was denied, the McCollins case was a properly conducted class action.” Brief for Respondent 37. That is true, according to Bayer, because McCollins’ interests were aligned with the members of the class he proposed and he “act[ed] in a representative capacity when he sought class certification.” Id., at 36. But wishing does not make it so. McCollins sought class certification, but he failed to obtain that result. Because the District Court found that individual issues predominated, it held that the action did not satisfy Federal Rule 23’s requirements for class proceedings. In these circumstances, we cannot say that a properly conducted class action existed at any time in the litigation. Federal Rule 23 determines what is and is not a class action in federal court, where McCollins brought his suit. So in the absence of a certification under that Rule, the precondition for binding Smith was not met. Neither a proposed class action nor a rejected class action may bind nonparties. What does have this effect is a class action approved under Rule 23. But McCollins’ lawsuit was never that. We made essentially these same points in Taylor v. Sturgell just a few Terms ago. The question there concerned the propriety of binding nonparties under a theory of “virtual representation” based on “identity of interests and some kind of relationship between parties and nonparties.” 553 U. S., at 901. We rejected the theory unanimously, explaining that it “would ‘recogniz[e], in effect, a common-law kind of class action.’ ” Ibid. Such a device, we objected, would authorize preclusion “shorn of [Rule 23’s] procedural protections.” Ibid. Or as otherwise stated in the opinion: We could not allow “circumvent[ion]” of Rule 23’s protections through a “virtual representation doctrine that allowed courts to ‘create de facto class actions at will.’ ” Ibid. We could hardly have been more clear that a “properly conducted class action,” with binding effect on nonparties, can come about in federal courts in just one way—through the procedure set out in Rule 23. Bayer attempts to distinguish Taylor by noting that the party in the prior litigation there did not propose a class action. But we do not see why that difference matters. Yes, McCollins wished to represent a class, and made a motion to that effect. But it did not come to pass. To allow McCollins’ suit to bind nonparties would be to adopt the very theory Taylor rejected.[Footnote 11] Bayer’s strongest argument comes not from established principles of preclusion, but instead from policy concerns relating to use of the class action device. Bayer warns that under our approach class counsel can repeatedly try to certify the same class “by the simple expedient of changing the named plaintiff in the caption of the complaint.” Brief for Respondent 47–48. And in this world of “serial relitigation of class certification,” Bayer contends, defendants “would be forced in effect to buy litigation peace by settling.” Id., at 2, 12; see also In re Bridgestone/Firestone, Inc., Tires Prods. Liability Litigation, 333 F. 3d 763, 767 (CA7 2003) (objecting to an “an asymmetric system in which class counsel can win but never lose” because of their ability to relitigate the issue of certification). But this form of argument flies in the face of the rule against nonparty preclusion. That rule perforce leads to relitigation of many issues, as plaintiff after plaintiff after plaintiff (none precluded by the last judgment because none a party to the last suit) tries his hand at establishing some legal principle or obtaining some grant of relief. We confronted a similar policy concern in Taylor, which involved litigation brought under the Freedom of Infor-mation Act (FOIA). The Government there cautioned that unless we bound nonparties a “ ‘potentially limitless’ ” number of plaintiffs, perhaps coordinating with each other, could “mount a series of repetitive lawsuits” demanding the selfsame documents. 553 U. S., at 903. But we rejected this argument, even though the payoff in a single successful FOIA suit—disclosure of documents to the public—could “trum[p]” or “subsum[e]” all prior losses, just as a single successful class certification motion could do. In re Bridgestone/Firestone, 333 F. 3d, at 766, 767. As that response suggests, our legal system generally relies on principles of stare decisis and comity among courts to mitigate the sometimes substantial costs of similar litigation brought by different plaintiffs. We have not thought that the right approach (except in the discrete categories of cases we have recognized) lies in binding nonparties to a judgment. And to the extent class actions raise special problems of relitigation, Congress has provided a remedy that does not involve departing from the usual rules of preclusion. In the Class Action Fairness Act of 2005 (CAFA), 28 U. S. C. §§1332(d), 1453 (2006 ed. and Supp. III), Congress enabled defendants to remove to federal court any sizable class action involving minimal diversity of citizenship. Once removal takes place, Federal Rule 23 governs certification. And federal courts may consolidate multiple overlapping suits against a single defendant in one court (as the Judicial Panel on Multi-District Litigation did for the many actions involving Baycol). See §1407. Finally, we would expect federal courts to apply principles of comity to each other’s class certification decisions when addressing a common dispute. See, e.g., Cortez Byrd Chips, Inc. v. Bill Harbert Constr. Co., 529 U. S. 193, 198 (2000) (citing Landis v. North American Co., 299 U. S. 248, 254 (1936)). CAFA may be cold comfort to Bayer with respect to suits like this one beginning before its enactment. But Congress’s decision to address the relitigation concerns associated with class actions through the mechanism of removal provides yet another reason for federal courts to adhere in this context to longstanding principles of preclusion.[Footnote 12] And once again, that is especially so when the federal court is deciding whether to go so far as to enjoin a state proceeding. * * * The Anti-Injunction Act prohibits the order the District Court entered here. The Act’s relitigation exception authorizes injunctions only when a former federal adjudication clearly precludes a state-court decision. As we said more than 40 years ago, and have consistently maintained since that time, “[a]ny doubts … should be resolved in favor of permitting the state courts to proceed.” Atlantic Coast Line, 398 U. S., at 297. Under this approach, close cases have easy answers: The federal court should not issue an injunction, and the state court should decide the preclusion question. But this case does not even strike us as close. The issues in the federal and state lawsuits differed because the relevant legal standards differed. And the mere proposal of a class in the federal action could not bind persons who were not parties there. For these reasons, the judgment of the Court of Appeals is Reversed. * Justice Thomas joins Parts I and II–A of this opinion. Footnote 1 The Class Action Fairness Act of 2005, 119 Stat. 4, which postdates and therefore does not govern this lawsuit, now enables a defendant to remove to federal court certain class actions involving nondiverse parties. See 28 U. S. C. §§1332(d), 1453(b); see also infra, at 17. Footnote 2 Although McCollins had originally sought certification under West Virginia Rule of Civil Procedure 23 (2011), federal procedural rules govern a case that has been removed to federal court. See Shady Grove Orthopedic Associates, P. A. v. Allstate Ins. Co., 559 U. S. ___ (2010). Footnote 3 Compare In re Baycol Prods. Litigation, 593 F. 3d 716, 723 (CA8 2010) (case below) (holding that two cases involve the same issue when “[t]he state and federal [class] certification rules … are not significantly different”), with J. R. Clearwater Inc. v. Ashland Chemical Co., 93 F. 3d 176, 180 (CA5 1996) (holding that two cases implicate different issues even when “[the state rule] is modeled on … the Federal Rules” because a “[state] court might well exercise [its] discretion in a different manner”). Footnote 4 Compare 593 F. 3d, at 724 (“[T]he denial of class certification is binding on unnamed [putative] class members” because they are “in privity to [the parties] in the prior action”) and In re Bridgestone/Firestone, Inc., Tires Prods. Liability Litigation, 333 F. 3d 763, 768–769 (CA7 2003) (same), with In re Ford Motor Co., 471 F. 3d 1233, 1245 (CA11 2006) (holding that “[t]he denial of class certification” prevents a court from “binding” anyone other than “the parties appearing before it”) and In re General Motors Corp. Pick-Up Truck Fuel Tank Prods. Liability Litigation, 134 F. 3d 133, 141 (CA3 1998) (holding that putative “class members are not parties” and so cannot be bound by a court’s ruling when “there is no class pending”). Footnote 5 That is especially so because an injunction is not the only way to correct a state trial court’s erroneous refusal to give preclusive effect to a federal judgment. As we have noted before, “the state appellate courts and ultimately this Court” can review and reverse such a ruling. See Atlantic Coast Line R. Co. v. Locomotive Engineers, 398 U. S. 281, 287 (1970). Footnote 6 We have held that federal common law governs the preclusive effect of a decision of a federal court sitting in diversity. See Semtek Int’l Inc. v. Lockheed Martin Corp., 531 U. S. 497, 508 (2001). Smith assumes that federal common law should here incorporate West Virginia’s preclusion law, see Brief for Petitioners 15–16, whereas Bayer favors looking only to federal rules of preclusion because of the federal interests at stake in this case, see Brief for Respondent 18. We do not think the question matters here. Neither party identifies any way in which federal and state principles of preclusion law differ in any relevant respect. Nor have we found any such divergence. Compare, e.g., Montana v. United States, 440 U. S. 147, 153–154 (1979) (describing elements of issue preclusion), with State v. Miller, 194 W. Va. 3, 9, 459 S. E. 2d 114, 120 (1995) (same). We therefore need not decide whether, in general, federal common law ought to incorporate state law in situations such as this. Footnote 7 Because we rest our decision on the Anti-Injunction Act and the principles of issue preclusion that inform it, we do not consider Smith’s argument, based on Phillips Petroleum Co. v. Shutts, 472 U. S. 797 (1985), that the District Court’s action violated the Due Process Clause. Footnote 8 The District Court’s approach to the predominance inquiry is consistent with the approach employed by the Eighth Circuit. See In re St. Jude Medical, Inc., 522 F. 3d 836, 837–840 (2008) (holding that most commercial misrepresentation cases are “unsuitable for class treatment” because individual issues of reliance necessarily predominate). We express no opinion as to the correctness of this approach. Footnote 9 Bayer argues that In re Rezulin does not preclude an injunction in this case because the West Virginia court there decided that common issues predominated over individual issues of damages, not over individual issues of liability (as exist here). See Brief for Respondent 25–26. We think Bayer is right about this distinction, but wrong about its consequence. Our point is not that In re Rezulin dictates the answer to the class certification question here; the two cases are indeed too dissimilar for that to be true. The point instead is that In re Rezulin articulated a general approach to the predominance requirement that differs markedly from the one the federal court used. Minor variations in the application of what is in essence the same legal standard do not defeat preclusion; but where, as here, the State’s courts “would apply a significantly different … analysis,” Chick Kam Choo v. Exxon Corp., 486 U. S. 140, 149 (1988), the federal and state courts decide different issues. Footnote 10 In support of its claim that Smith counts as a party, Bayer cites two cases in which we held that a putative member of an uncertified class may wait until after the court rules on the certification motion to file an individual claim or move to intervene in the suit. See Brief for Respondent 32–33 (citing United Airlines, Inc. v. McDonald, 432 U. S. 385 (1977); American Pipe & Constr. Co. v. Utah, 414 U. S. 538 (1974)). But these cases, which were specifically grounded in policies of judicial administration, demonstrate only that a person not a party to a class suit may receive certain benefits (such as the tolling of a limitations period) related to that proceeding. See id., at 553; McDonald, 432 U. S., at 394, n. 15. That result is consistent with a commonplace of preclusion law—that nonparties sometimes may benefit from, even though they cannot be bound by, former litigation. See Parklane Hosiery Co. v. Shore, 439 U. S. 322, 326–333 (1979); Blonder-Tongue Laboratories, Inc. v. University of Ill. Foundation, 402 U. S. 313 (1971). Footnote 11 The great weight of scholarly authority—from the Restatement of Judgments to the American Law Institute to Wright and Miller—agrees that an uncertified class action cannot bind proposed class members. See Restatement (Second) of Judgments §41(1), p. 393 (1980) (A nonparty may be bound only when his interests are adequately represented by “[t]he representative of a class of persons similarly situated, designated as such with the approval of the court”); ALI, Principles of the Law Aggregate Litigation §2.11, Reporters’ Notes, cmt. b, p. 181 (2010) (“[N]one of [the exceptions to the rule against nonparty preclusion] extend generally to the situation of a would-be absent class member with respect to a denial of class certification”); 18A Wright & Miller §4455, at 457–458 (“[A]bsent certification there is no basis for precluding a nonparty” under the class-action exception). Footnote 12 By the same token, nothing in our holding today forecloses legislation to modify established principles of preclusion should Congress decide that CAFA does not sufficiently prevent relitigation of class certification motions. Nor does this opinion at all address the permissibility of a change in the Federal Rules of Civil Procedure pertaining to this question. Cf. n. 7, supra (declining to reach Smith’s due process claim).

562.US.443

For the past 20 years, the congregation of the Westboro Baptist Church has picketed military funerals to communicate its belief that God hates the United States for its tolerance of homosexuality, particularly in America’s military. The church’s picketing has also condemned the Catholic Church for scandals involving its clergy. Fred Phelps, who founded the church, and six Westboro Baptist parishioners (all relatives of Phelps) traveled to Maryland to picket the funeral of Marine Lance Corporal Matthew Snyder, who was killed in Iraq in the line of duty. The picketing took place on public land approximately 1,000 feet from the church where the funeral was held, in accordance with guidance from local law enforcement officers. The picketers peacefully displayed their signs—stating, e.g., “Thank God for Dead Soldiers,” “Fags Doom Nations,” “America is Doomed,” “Priests Rape Boys,” and “You’re Going to Hell”—for about 30 minutes before the funeral began. Matthew Snyder’s father (Snyder), petitioner here, saw the tops of the picketers’ signs when driving to the funeral, but did not learn what was written on the signs until watching a news broadcast later that night. Snyder filed a diversity action against Phelps, his daughters—who participated in the picketing—and the church (collectively Westboro) alleging, as relevant here, state tort claims of intentional infliction of emotional distress, intrusion upon seclusion, and civil conspiracy. A jury held Westboro liable for millions of dollars in compensatory and punitive damages. Westboro challenged the verdict as grossly excessive and sought judgment as a matter of law on the ground that the First Amendment fully protected its speech. The District Court reduced the punitive damages award, but left the verdict otherwise intact. The Fourth Circuit reversed, concluding that Westboro’s statements were entitled to First Amendment protection because those statements were on matters of public concern, were not provably false, and were expressed solely through hyperbolic rhetoric. Held: The First Amendment shields Westboro from tort liability for its picketing in this case. Pp. 5–15. (a) The Free Speech Clause of the First Amendment can serve as a defense in state tort suits, including suits for intentional infliction of emotional distress. Hustler Magazine, Inc. v. Falwell, 485 U. S. 46, 50-51. Whether the First Amendment prohibits holding Westboro liable for its speech in this case turns largely on whether that speech is of public or private concern, as determined by all the circumstances of the case. “[S]peech on public issues occupies the ‘ “highest rung of the hierarchy of First Amendment values” ’ and is entitled to special protection.” Connick v. Myers, 461 U. S. 138, 145. Although the boundaries of what constitutes speech on matters of public concern are not well defined, this Court has said that speech is of public concern when it can “be fairly considered as relating to any matter of political, social, or other concern to the community,” id., at 146, or when it “is a subject of general interest and of value and concern to the public,” San Diego v. Roe, 543 U. S. 77, 83–84. A statement’s arguably “inappropriate or controversial character … is irrelevant to the question whether it deals with a matter of public concern.” Rankin v. McPherson, 483 U. S. 378, 387. Pp. 5–7. To determine whether speech is of public or private concern, this Court must independently examine the “ ‘content, form, and context,’ ” of the speech “ ‘as revealed by the whole record.’ ” Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U. S. 749, 761. In considering content, form, and context, no factor is dispositive, and it is necessary to evaluate all aspects of the speech. Pp. 7–8. The “content” of Westboro’s signs plainly relates to public, rather than private, matters. The placards highlighted issues of public import—the political and moral conduct of the United States and its citizens, the fate of the Nation, homosexuality in the military, and scandals involving the Catholic clergy—and Westboro conveyed its views on those issues in a manner designed to reach as broad a public audience as possible. Even if a few of the signs were viewed as containing messages related to a particular individual, that would not change the fact that the dominant theme of Westboro’s demonstration spoke to broader public issues. P. 8. The “context” of the speech—its connection with Matthew Snyder’s funeral—cannot by itself transform the nature of Westboro’s speech. The signs reflected Westboro’s condemnation of much in modern society, and it cannot be argued that Westboro’s use of speech on public issues was in any way contrived to insulate a personal attack on Snyder from liability. Westboro had been actively engaged in speaking on the subjects addressed in its picketing long before it became aware of Matthew Snyder, and there can be no serious claim that the picketing did not represent Westboro’s honestly held beliefs on public issues. Westboro may have chosen the picket location to increase publicity for its views, and its speech may have been particularly hurtful to Snyder. That does not mean that its speech should be afforded less than full First Amendment protection under the circumstances of this case. Pp. 8–10. That said, “ ‘[e]ven protected speech is not equally permissible in all places and at all times.’ ” Frisby v. Schultz, 487 U. S. 474, 479. Westboro’s choice of where and when to conduct its picketing is not beyond the Government’s regulatory reach—it is “subject to reasonable time, place, or manner restrictions.” Clark v. Community for Creative Non-Violence, 468 U. S. 288, 293. The facts here are quite different, however, both with respect to the activity being regulated and the means of restricting those activities, from the few limited situations where the Court has concluded that the location of targeted picketing can be properly regulated under provisions deemed content neutral. Frisby, supra, at 477; Madsen v. Women’s Health Center, Inc., 512 U. S. 753, 768, distinguished. Maryland now has a law restricting funeral picketing but that law was not in effect at the time of these events, so this Court has no occasion to consider whether that law is a “reasonable time, place, or manner restrictio[n]” under the standards announced by this Court. Clark, supra, at 293. Pp. 10–12. The “special protection” afforded to what Westboro said, in the whole context of how and where it chose to say it, cannot be overcome by a jury finding that the picketing was “outrageous” for purposes of applying the state law tort of intentional infliction of emotional distress. That would pose too great a danger that the jury would punish Westboro for its views on matters of public concern. For all these reasons, the jury verdict imposing tort liability on Westboro for intentional infliction of emotional distress must be set aside. Pp. 12–13. (b) Snyder also may not recover for the tort of intrusion upon seclusion. He argues that he was a member of a captive audience at his son’s funeral, but the captive audience doctrine—which has been applied sparingly, see Rowan v. Post Office Dept., 397 U. S. 728, 736–738; Frisby, supra, at 484–485—should not be expanded to the circumstances here. Westboro stayed well away from the memorial service, Snyder could see no more than the tops of the picketers’ signs, and there is no indication that the picketing interfered with the funeral service itself. Pp. 13–14. (c) Because the First Amendment bars Snyder from recovery for intentional infliction of emotional distress or intrusion upon seclusion—the allegedly unlawful activity Westboro conspired to accomplish—Snyder also cannot recover for civil conspiracy based on those torts. P. 14. (d) Westboro addressed matters of public import on public property, in a peaceful manner, in full compliance with the guidance of local officials. It did not disrupt Mathew Snyder’s funeral, and its choice to picket at that time and place did not alter the nature of its speech. Because this Nation has chosen to protect even hurtful speech on public issues to ensure that public debate is not stifled, Westboro must be shielded from tort liability for its picketing in this case. Pp. 14–15. 580 F. 3d 206, affirmed. Roberts, C. J., delivered the opinion of the Court, in which Scalia, Kennedy, Thomas, Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined. Breyer, J., filed a concurring opinion. Alito, J., filed a dissenting opinion.

A jury held members of the Westboro Baptist Church liable for millions of dollars in damages for picketing near a soldier’s funeral service. The picket signs reflected the church’s view that the United States is overly tolerant of sin and that God kills American soldiers as punishment. The question presented is whether the First Amendment shields the church members from tort liability for their speech in this case. I A Fred Phelps founded the Westboro Baptist Church in Topeka, Kansas, in 1955. The church’s congregation believes that God hates and punishes the United States for its tolerance of homosexuality, particularly in America’s military. The church frequently communicates its views by picketing, often at military funerals. In the more than 20 years that the members of Westboro Baptist have publicized their message, they have picketed nearly 600 funerals. Brief for Rutherford Institute as Amicus Curiae 7, n. 14. Marine Lance Corporal Matthew Snyder was killed in Iraq in the line of duty. Lance Corporal Snyder’s father selected the Catholic church in the Snyders’ hometown of Westminster, Maryland, as the site for his son’s funeral. Local newspapers provided notice of the time and location of the service. Phelps became aware of Matthew Snyder’s funeral and decided to travel to Maryland with six other Westboro Baptist parishioners (two of his daughters and four of his grandchildren) to picket. On the day of the memorial service, the Westboro congregation members picketed on public land adjacent to public streets near the Maryland State House, the United States Naval Academy, and Matthew Snyder’s funeral. The Westboro picketers carried signs that were largely the same at all three locations. They stated, for instance: “God Hates the USA/Thank God for 9/11,” “America is Doomed,” “Don’t Pray for the USA,” “Thank God for IEDs,” “Thank God for Dead Soldiers,” “Pope in Hell,” “Priests Rape Boys,” “God Hates Fags,” “You’re Going to Hell,” and “God Hates You.” The church had notified the authorities in advance of its intent to picket at the time of the funeral, and the picketers complied with police instructions in staging their demonstration. The picketing took place within a 10- by 25-foot plot of public land adjacent to a public street, behind a temporary fence. App. to Brief for Appellants in No. 08–1026 (CA4), pp. 2282–2285 (hereinafter App.). That plot was approximately 1,000 feet from the church where the funeral was held. Several buildings separated the picket site from the church. Id., at 3758. The Westboro picketers displayed their signs for about 30 minutes before the funeral began and sang hymns and recited Bible verses. None of the picketers entered church property or went to the cemetery. They did not yell or use profanity, and there was no violence associated with the picketing. Id., at 2168, 2371, 2286, 2293. The funeral procession passed within 200 to 300 feet of the picket site. Although Snyder testified that he could see the tops of the picket signs as he drove to the funeral, he did not see what was written on the signs until later that night, while watching a news broadcast covering the event. Id., at 2084–2086.[Footnote 1] B Snyder filed suit against Phelps, Phelps’s daughters, and the Westboro Baptist Church (collectively Westboro or the church) in the United States District Court for the District of Maryland under that court’s diversity jurisdiction. Snyder alleged five state tort law claims: defamation, publicity given to private life, intentional infliction of emotional distress, intrusion upon seclusion, and civil conspiracy. Westboro moved for summary judgment contending, in part, that the church’s speech was insulated from liability by the First Amendment. See 533 F. Supp. 2d 567, 570 (Md. 2008). The District Court awarded Westboro summary judgment on Snyder’s claims for defamation and publicity given to private life, concluding that Snyder could not prove the necessary elements of those torts. Id., at 572–573. A trial was held on the remaining claims. At trial, Snyder described the severity of his emotional injuries. He testified that he is unable to separate the thought of his dead son from his thoughts of Westboro’s picketing, and that he often becomes tearful, angry, and physically ill when he thinks about it. Id., at 588–589. Expert witnesses testified that Snyder’s emotional anguish had resulted in severe depression and had exacerbated pre-existing health conditions. A jury found for Snyder on the intentional infliction of emotional distress, intrusion upon seclusion, and civil conspiracy claims, and held Westboro liable for $2.9 million in compensatory damages and $8 million in punitive damages. Westboro filed several post-trial motions, including a motion contending that the jury verdict was grossly excessive and a motion seeking judgment as a matter of law on all claims on First Amendment grounds. The District Court remitted the punitive damages award to $2.1 million, but left the jury verdict otherwise intact. Id., at 597. In the Court of Appeals, Westboro’s primary argument was that the church was entitled to judgment as a matter of law because the First Amendment fully protected Westboro’s speech. The Court of Appeals agreed. 580 F. 3d 206, 221 (CA4 2009). The court reviewed the picket signs and concluded that Westboro’s statements were entitled to First Amendment protection because those statements were on matters of public concern, were not provably false, and were expressed solely through hyperbolic rhetoric. Id., at 222–224.[Footnote 2] We granted certiorari. 559 U. S. ___ (2010). II To succeed on a claim for intentional infliction of emotional distress in Maryland, a plaintiff must demonstrate that the defendant intentionally or recklessly engaged in extreme and outrageous conduct that caused the plaintiff to suffer severe emotional distress. See Harris v. Jones, 281 Md. 560, 565–566, 380 A. 2d 611, 614 (1977). The Free Speech Clause of the First Amendment—“Congress shall make no law … abridging the freedom of speech”—can serve as a defense in state tort suits, including suits for intentional infliction of emotional distress. See, e.g., Hustler Magazine, Inc. v. Falwell, 485 U. S. 46, 50–51 (1988).[Footnote 3] Whether the First Amendment prohibits holding Westboro liable for its speech in this case turns largely on whether that speech is of public or private concern, as determined by all the circumstances of the case. “[S]peech on ‘matters of public concern’ … is ‘at the heart of the First Amendment’s protection.’ ” Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U. S. 749, 758–759 (1985) (opinion of Powell, J.) (quoting First Nat. Bank of Boston v. Bellotti, 435 U. S. 765, 776 (1978)). The First Amendment reflects “a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open.” New York Times Co. v. Sullivan, 376 U. S. 254, 270 (1964). That is because “speech concerning public affairs is more than self-expression; it is the essence of self-government.” Garrison v. Louisiana, 379 U. S. 64, 74–75 (1964). Accordingly, “speech on public issues occupies the highest rung of the hierarchy of First Amendment values, and is entitled to special protection.” Connick v. Myers, 461 U. S. 138, 145 (1983) (internal quotation marks omitted). “ ‘[N]ot all speech is of equal First Amendment importance,’ ” however, and where matters of purely private significance are at issue, First Amendment protections are often less rigorous. Hustler, supra, at 56 (quoting Dun & Bradstreet, supra, at 758); see Connick, supra, at 145–147. That is because restricting speech on purely private matters does not implicate the same constitutional concerns as limiting speech on matters of public interest: “[T]here is no threat to the free and robust debate of public issues; there is no potential interference with a meaningful dialogue of ideas”; and the “threat of liability” does not pose the risk of “a reaction of self-censorship” on matters of public import. Dun & Bradstreet, supra, at 760 (internal quotation marks omitted). We noted a short time ago, in considering whether public employee speech addressed a matter of public concern, that “the boundaries of the public concern test are not well defined.” San Diego v. Roe, 543 U. S. 77, 83 (2004) (per curiam). Although that remains true today, we have articulated some guiding principles, principles that accord broad protection to speech to ensure that courts themselves do not become inadvertent censors. Speech deals with matters of public concern when it can “be fairly considered as relating to any matter of politi- cal, social, or other concern to the community,” Connick, supra, at 146, or when it “is a subject of legitimate news interest; that is, a subject of general interest and of value and concern to the public,” San Diego, supra, at 83–84. See Cox Broadcasting Corp. v. Cohn, 420 U. S. 469, 492–494 (1975); Time, Inc. v. Hill, 385 U. S. 374, 387– 388 (1967). The arguably “inappropriate or controversial character of a statement is irrelevant to the question whether it deals with a matter of public concern.” Rankin v. McPherson, 483 U. S. 378, 387 (1987). Our opinion in Dun & Bradstreet, on the other hand, provides an example of speech of only private concern. In that case we held, as a general matter, that information about a particular individual’s credit report “concerns no public issue.” 472 U. S., at 762. The content of the report, we explained, “was speech solely in the individual interest of the speaker and its specific business audience.” Ibid. That was confirmed by the fact that the particular report was sent to only five subscribers to the reporting service, who were bound not to disseminate it further. Ibid. To cite another example, we concluded in San Diego v. Roe that, in the context of a government employer regulating the speech of its employees, videos of an employee engaging in sexually explicit acts did not address a public concern; the videos “did nothing to inform the public about any aspect of the [employing agency’s] functioning or operation.” 543 U. S., at 84. Deciding whether speech is of public or private concern requires us to examine the “ ‘content, form, and context’ ” of that speech, “ ‘as revealed by the whole record.’ ” Dun & Bradstreet, supra, at 761 (quoting Connick, supra, at 147–148). As in other First Amendment cases, the court is obligated “to ‘make an independent examination of the whole record’ in order to make sure that ‘the judgment does not constitute a forbidden intrusion on the field of free expression.’ ” Bose Corp. v. Consumers Union of United States, Inc., 466 U. S. 485, 499 (1984) (quoting New York Times, supra, at 284–286). In considering content, form, and context, no factor is dispositive, and it is necessary to evaluate all the circumstances of the speech, including what was said, where it was said, and how it was said. The “content” of Westboro’s signs plainly relates to broad issues of interest to society at large, rather than matters of “purely private concern.” Dun & Bradstreet, supra, at 759. The placards read “God Hates the USA/Thank God for 9/11,” “America is Doomed,” “Don’t Pray for the USA,” “Thank God for IEDs,” “Fag Troops,” “Semper Fi Fags,” “God Hates Fags,” “Maryland Taliban,” “Fags Doom Nations,” “Not Blessed Just Cursed,” “Thank God for Dead Soldiers,” “Pope in Hell,” “Priests Rape Boys,” “You’re Going to Hell,” and “God Hates You.” App. 3781–3787. While these messages may fall short of refined social or political commentary, the issues they highlight—the political and moral conduct of the United States and its citizens, the fate of our Nation, homosexuality in the military, and scandals involving the Catholic clergy—are matters of public import. The signs certainly convey Westboro’s position on those issues, in a manner designed, unlike the private speech in Dun & Bradstreet, to reach as broad a public audience as possible. And even if a few of the signs—such as “You’re Going to Hell” and “God Hates You”—were viewed as containing messages related to Matthew Snyder or the Snyders specifically, that would not change the fact that the overall thrust and dominant theme of Westboro’s demonstration spoke to broader public issues. Apart from the content of Westboro’s signs, Snyder contends that the “context” of the speech—its connection with his son’s funeral—makes the speech a matter of private rather than public concern. The fact that Westboro spoke in connection with a funeral, however, cannot by itself transform the nature of Westboro’s speech. Westboro’s signs, displayed on public land next to a public street, reflect the fact that the church finds much to condemn in modern society. Its speech is “fairly characterized as constituting speech on a matter of public concern,” Connick, 461 U. S., at 146, and the funeral setting does not alter that conclusion. Snyder argues that the church members in fact mounted a personal attack on Snyder and his family, and then attempted to “immunize their conduct by claiming that they were actually protesting the United States’ tolerance of homosexuality or the supposed evils of the Catholic Church.” Reply Brief for Petitioner 10. We are not concerned in this case that Westboro’s speech on public matters was in any way contrived to insulate speech on a private matter from liability. Westboro had been actively engaged in speaking on the subjects addressed in its picketing long before it became aware of Matthew Snyder, and there can be no serious claim that Westboro’s picketing did not represent its “honestly believed” views on public issues. Garrison, 379 U. S., at 73. There was no pre-existing relationship or conflict between Westboro and Snyder that might suggest Westboro’s speech on public matters was intended to mask an attack on Snyder over a private matter. Contrast Connick, supra, at 153 (finding public employee speech a matter of private concern when it was “no coincidence that [the speech] followed upon the heels of [a] transfer notice” affecting the employee). Snyder goes on to argue that Westboro’s speech should be afforded less than full First Amendment protection “not only because of the words” but also because the church members exploited the funeral “as a platform to bring their message to a broader audience.” Brief for Petitioner 44, 40. There is no doubt that Westboro chose to stage its picketing at the Naval Academy, the Maryland State House, and Matthew Snyder’s funeral to increase publicity for its views and because of the relation between those sites and its views—in the case of the military funeral, because Westboro believes that God is killing American soldiers as punishment for the Nation’s sinful policies. Westboro’s choice to convey its views in conjunction with Matthew Snyder’s funeral made the expression of those views particularly hurtful to many, especially to Matthew’s father. The record makes clear that the applicable legal term—“emotional distress”—fails to capture fully the anguish Westboro’s choice added to Mr. Snyder’s already incalculable grief. But Westboro conducted its picketing peacefully on matters of public concern at a public place adjacent to a public street. Such space occupies a “special position in terms of First Amendment protection.” United States v. Grace, 461 U. S. 171, 180 (1983). “[W]e have repeatedly referred to public streets as the archetype of a traditional public forum,” noting that “ ‘[t]ime out of mind’ public streets and sidewalks have been used for public assembly and debate.” Frisby v. Schultz, 487 U. S. 474, 480 (1988).[Footnote 4] That said, “[e]ven protected speech is not equally permissible in all places and at all times.” Id., at 479 (quoting Cornelius v. NAACP Legal Defense & Ed. Fund, Inc., 473 U. S. 788, 799 (1985)). Westboro’s choice of where and when to conduct its picketing is not beyond the Government’s regulatory reach—it is “subject to reasonable time, place, or manner restrictions” that are consistent with the standards announced in this Court’s precedents. Clark v. Community for Creative Non-Violence, 468 U. S. 288, 293 (1984). Maryland now has a law imposing restrictions on funeral picketing, Md. Crim. Law Code Ann. §10–205 (Lexis Supp. 2010), as do 43 other States and the Federal Government. See Brief for American Legion as Amicus Curiae 18–19, n. 2 (listing statutes). To the extent these laws are content neutral, they raise very different questions from the tort verdict at issue in this case. Maryland’s law, however, was not in effect at the time of the events at issue here, so we have no occasion to consider how it might apply to facts such as those before us, or whether it or other similar regulations are constitutional.[Footnote 5] We have identified a few limited situations where the location of targeted picketing can be regulated under provisions that the Court has determined to be content neutral. In Frisby, for example, we upheld a ban on such picketing “before or about” a particular residence, 487 U. S., at 477. In Madsen v. Women’s Health Center, Inc., we approved an injunction requiring a buffer zone between protesters and an abortion clinic entrance. 512 U. S. 753, 768 (1994). The facts here are obviously quite different, both with respect to the activity being regulated and the means of restricting those activities. Simply put, the church members had the right to be where they were. Westboro alerted local authorities to its funeral protest and fully complied with police guidance on where the picketing could be staged. The picketing was conducted under police supervision some 1,000 feet from the church, out of the sight of those at the church. The protest was not unruly; there was no shouting, profanity, or violence. The record confirms that any distress occasioned by Westboro’s picketing turned on the content and viewpoint of the message conveyed, rather than any interference with the funeral itself. A group of parishioners standing at the very spot where Westboro stood, holding signs that said “God Bless America” and “God Loves You,” would not have been subjected to liability. It was what Westboro said that exposed it to tort damages. Given that Westboro’s speech was at a public place on a matter of public concern, that speech is entitled to “special protection” under the First Amendment. Such speech cannot be restricted simply because it is upsetting or arouses contempt. “If there is a bedrock principle underlying the First Amendment, it is that the government may not prohibit the expression of an idea simply because society finds the idea itself offensive or disagreeable.” Texas v. Johnson, 491 U. S. 397, 414 (1989). Indeed, “the point of all speech protection … is to shield just those choices of content that in someone’s eyes are misguided, or even hurtful.” Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, Inc., 515 U. S. 557, 574 (1995). The jury here was instructed that it could hold Westboro liable for intentional infliction of emotional distress based on a finding that Westboro’s picketing was “outrageous.” “Outrageousness,” however, is a highly malleable standard with “an inherent subjectiveness about it which would allow a jury to impose liability on the basis of the jurors’ tastes or views, or perhaps on the basis of their dislike of a particular expression.” Hustler, 485 U. S., at 55 (internal quotation marks omitted). In a case such as this, a jury is “unlikely to be neutral with respect to the content of [the] speech,” posing “a real danger of becoming an instrument for the suppression of … ‘vehement, caustic, and sometimes unpleasan[t]’ ” expression. Bose Corp., 466 U. S., at 510 (quoting New York Times, 376 U. S., at 270). Such a risk is unacceptable; “in public debate [we] must tolerate insulting, and even outrageous, speech in order to provide adequate ‘breathing space’ to the freedoms protected by the First Amendment.” Boos v. Barry, 485 U. S. 312, 322 (1988) (some internal quotation marks omitted). What Westboro said, in the whole context of how and where it chose to say it, is entitled to “special protection” under the First Amendment, and that protection cannot be overcome by a jury finding that the picketing was outrageous. For all these reasons, the jury verdict imposing tort liability on Westboro for intentional infliction of emotional distress must be set aside. III The jury also found Westboro liable for the state law torts of intrusion upon seclusion and civil conspiracy. The Court of Appeals did not examine these torts independently of the intentional infliction of emotional distress tort. Instead, the Court of Appeals reversed the District Court wholesale, holding that the judgment wrongly “attache[d] tort liability to constitutionally protected speech.” 580 F. 3d, at 226. Snyder argues that even assuming Westboro’s speech is entitled to First Amendment protection generally, the church is not immunized from liability for intrusion upon seclusion because Snyder was a member of a captive audience at his son’s funeral. Brief for Petitioner 45–46. We do not agree. In most circumstances, “the Constitution does not permit the government to decide which types of otherwise protected speech are sufficiently offensive to require protection for the unwilling listener or viewer. Rather, … the burden normally falls upon the viewer to avoid further bombardment of [his] sensibilities simply by averting [his] eyes.” Erznoznik v. Jacksonville, 422 U. S. 205, 210–211 (1975) (internal quotation marks omitted). As a result, “[t]he ability of government, consonant with the Constitution, to shut off discourse solely to protect others from hearing it is … dependent upon a showing that substantial privacy interests are being invaded in an essentially intolerable manner.” Cohen v. California, 403 U. S. 15, 21 (1971). As a general matter, we have applied the captive audience doctrine only sparingly to protect unwilling listeners from protected speech. For example, we have upheld a statute allowing a homeowner to restrict the delivery of offensive mail to his home, see Rowan v. Post Office Dept., 397 U. S. 728, 736–738 (1970), and an ordinance prohibiting picketing “before or about” any individual’s residence, Frisby, 487 U. S., at 484–485. Here, Westboro stayed well away from the memorial service. Snyder could see no more than the tops of the signs when driving to the funeral. And there is no indication that the picketing in any way interfered with the funeral service itself. We decline to expand the captive audience doctrine to the circumstances presented here. Because we find that the First Amendment bars Snyder from recovery for intentional infliction of emotional distress or intrusion upon seclusion—the alleged unlawful activity Westboro conspired to accomplish—we must likewise hold that Snyder cannot recover for civil conspiracy based on those torts. IV Our holding today is narrow. We are required in First Amendment cases to carefully review the record, and the reach of our opinion here is limited by the particular facts before us. As we have noted, “the sensitivity and significance of the interests presented in clashes between First Amendment and [state law] rights counsel relying on limited principles that sweep no more broadly than the appropriate context of the instant case.” Florida Star v. B. J. F., 491 U. S. 524, 533 (1989). Westboro believes that America is morally flawed; many Americans might feel the same about Westboro. Westboro’s funeral picketing is certainly hurtful and its con-tribution to public discourse may be negligible. But Westboro addressed matters of public import on public property, in a peaceful manner, in full compliance with the guidance of local officials. The speech was indeed planned to coincide with Matthew Snyder’s funeral, but did not itself disrupt that funeral, and Westboro’s choice to conduct its picketing at that time and place did not alter the nature of its speech. Speech is powerful. It can stir people to action, move them to tears of both joy and sorrow, and—as it did here—inflict great pain. On the facts before us, we cannot react to that pain by punishing the speaker. As a Nation we have chosen a different course—to protect even hurtful speech on public issues to ensure that we do not stifle public debate. That choice requires that we shield Westboro from tort liability for its picketing in this case. The judgment of the United States Court of Appeals for the Fourth Circuit is affirmed. It is so ordered. Footnote 1 A few weeks after the funeral, one of the picketers posted a message on Westboro’s Web site discussing the picketing and containing religiously oriented denunciations of the Snyders, interspersed among lengthy Bible quotations. Snyder discovered the posting, referred to by the parties as the “epic,” during an Internet search for his son’s name. The epic is not properly before us and does not factor in our analysis. Although the epic was submitted to the jury and discussed in the courts below, Snyder never mentioned it in his petition for certiorari. See Pet. for Cert. i (“Snyder’s claim arose out of Phelps’ intentional acts at Snyder’s son’s funeral” (emphasis added)); this Court’s Rule 14.1(g) (petition must contain statement “setting out the facts material to consideration of the question presented”). Nor did Snyder respond to the statement in the opposition to certiorari that “[t]hough the epic was asserted as a basis for the claims at trial, the petition … appears to be addressing only claims based on the picketing.” Brief in Opposition 9. Snyder devoted only one paragraph in the argument section of his opening merits brief to the epic. Given the foregoing and the fact that an Internet posting may raise distinct issues in this context, we decline to consider the epic in deciding this case. See Ontario v. Quon, 560 U. S. ___, ___ – ___ (2010) (slip op., at 10–12). Footnote 2 One judge concurred in the judgment on the ground that Snyder had failed to introduce sufficient evidence at trial to support a jury verdict on any of his tort claims. 580 F. 3d, at 227 (opinion of Shedd, J.). The Court of Appeals majority determined that the picketers had “voluntarily waived” any such contention on appeal. Id., at 216. Like the court below, we proceed on the unexamined premise that respondents’ speech was tortious. Footnote 3 The dissent attempts to draw parallels between this case and hy-pothetical cases involving defamation or fighting words. Post, at 10–11 (opinion of Alito, J.). But, as the court below noted, there is “no suggestion that the speech at issue falls within one of the categorical exclusions from First Amendment protection, such as those for obscenity or ‘fighting words.’ ” 580 F. 3d, at 218, n. 12; see United States v. Stevens, 559 U. S. ___ , ___ (2010) (slip op., at 5). Footnote 4 The dissent is wrong to suggest that the Court considers a public street “a free-fire zone in which otherwise actionable verbal attacks are shielded from liability.” Post, at 10–11. The fact that Westboro conducted its picketing adjacent to a public street does not insulate the speech from liability, but instead heightens concerns that what is at issue is an effort to communicate to the public the church’s views on matters of public concern. That is why our precedents so clearly recognize the special significance of this traditional public forum. Footnote 5 The Maryland law prohibits picketing within 100 feet of a funeral service or funeral procession; Westboro’s picketing would have complied with that restriction.

564.US.2010_10-779

Pharmaceutical manufacturers promote their drugs to doctors through a process called “detailing.” Pharmacies receive “prescriber-identifying information” when processing prescriptions and sell the information to “data miners,” who produce reports on prescriber behavior and lease their reports to pharmaceutical manufacturers. “Detailers” employed by pharmaceutical manufacturers then use the reports to refine their marketing tactics and increase sales to doctors. Vermont’s Prescription Confidentiality Law provides that, absent the prescriber’s consent, prescriber-identifying information may not be sold by pharmacies and similar entities, disclosed by those entities for marketing purposes, or used for marketing by pharmaceutical manufacturers. Vt. Stat. Ann., Tit. 18, §4631(d). The prohibitions are subject to exceptions that permit the prescriber-identifying information to be disseminated and used for a number of purposes, e.g., “health care research.” §4631(e). Respondents, Vermont data miners and an association of brand-name drug manufacturers, sought declaratory and injunctive relief against state officials (hereinafter Vermont), contending that §4631(d) violates their rights under the Free Speech Clause of the First Amendment. The District Court denied relief, but the Second Circuit reversed, holding that §4631(d) unconstitutionally burdens the speech of pharmaceutical marketers and data miners without adequate justification. Held: 1. Vermont’s statute, which imposes content- and speaker-based burdens on protected expression, is subject to heightened judicial scrutiny. Pp. 6–15. (a) On its face, the law enacts a content- and speaker-based restriction on the sale, disclosure, and use of prescriber-identifying information. The law first forbids sale subject to exceptions based in large part on the content of a purchaser’s speech. It then bars pharmacies from disclosing the information when recipient speakers will use that information for marketing. Finally, it prohibits pharmaceutical manufacturers from using the information for marketing. The statute thus disfavors marketing, i.e., speech with a particular content, as well as particular speakers, i.e., detailers engaged in marketing on behalf of pharmaceutical manufacturers. Cincinnati v. Discovery Network, Inc., 507 U. S. 410, 426; Turner Broadcasting System, Inc. v. FCC, 512 U. S. 622, 658. Yet the law allows prescriber-identifying information to be purchased, acquired, and used for other types of speech and by other speakers. The record and formal legislative findings of purpose confirm that §4631(d) imposes an aimed, content-based burden on detailers, in particular detailers who promote brand-name drugs. In practical operation, Vermont’s law “goes even beyond mere content discrimination, to actual viewpoint discrimination.” R. A. V. v. St. Paul, 505 U. S. 377, 391. Heightened judicial scrutiny is warranted. Pp. 8–11. (b) Vermont errs in arguing that heightened scrutiny is unwarranted. The State contends that its law is a mere commercial regulation. Far from having only an incidental effect on speech, however, §4631(d) imposes a burden based on the content of speech and the identity of the speaker. The State next argues that, because prescriber-identifying information was generated in compliance with a legal mandate, §4631(d) is akin to a restriction on access to government-held information. That argument finds some support in Los Angeles Police Dept. v. United Reporting Publishing Corp., 528 U. S. 32, but that case is distinguishable. Vermont has imposed a restriction on access to information in private hands. United Reporting reserved that situation—i.e., “a case in which the government is prohibiting a speaker from conveying information that the speaker already possesses.” Id., at 40. In addition, the United Reporting plaintiff was presumed to have suffered no personal First Amendment injury, while respondents claim that §4631(d) burdens their own speech. That circumstance warrants heightened scrutiny. Vermont also argues that heightened judicial scrutiny is unwarranted because sales, transfer, and use of prescriber-identifying information are conduct, not speech. However, the creation and dissemination of information are speech for First Amendment purposes. See, e.g., Bartnicki v. Vopper, 532 U. S. 514, 527. There is no need to consider Vermont’s request for an exception to that rule. Section 4631(d) imposes a speaker- and content-based burden on protected expression, and that circumstance is sufficient to justify applying heightened scrutiny, even assuming that prescriber-identifying information is a mere commodity. Pp. 11–15. 2. Vermont’s justifications for §4631(d) do not withstand heightened scrutiny. Pp. 15–24. (a) The outcome here is the same whether a special commercial speech inquiry or a stricter form of judicial scrutiny is applied, see, e.g., Greater New Orleans Broadcasting Assn., Inc. v. United States, 527 U. S. 173, 184. To sustain §4631(d)’s targeted, content-based burden on protected expression, Vermont must show at least that the statute directly advances a substantial governmental interest and that the measure is drawn to achieve that interest. See Board of Trustees of State Univ. of N. Y. v. Fox, 492 U. S. 469, 480–481. Vermont contends that its law (1) is necessary to protect medical privacy, including physician confidentiality, avoidance of harassment, and the integrity of the doctor-patient relationship, and (2) is integral to the achievement of the policy objectives of improving public health and reducing healthcare costs. Pp. 15–17. (b) Assuming that physicians have an interest in keeping their prescription decisions confidential, §4631(d) is not drawn to serve that interest. Pharmacies may share prescriber-identifying information with anyone for any reason except for marketing. Vermont might have addressed physician confidentiality through “a more coherent policy,” Greater New Orleans Broadcasting, supra, at 195, such as allowing the information’s sale or disclosure in only a few narrow and well-justified circumstances. But it did not. Given the information’s widespread availability and many permissible uses, Vermont’s asserted interest in physician confidentiality cannot justify the burdens that §4631(d) imposes on protected expression. It is true that doctors can forgo the law’s advantages by consenting to the sale, disclosure, and use of their prescriber-identifying information. But the State has offered only a contrived choice: Either consent, which will allow the doctor’s prescriber-identifying information to be disseminated and used without constraint; or, withhold consent, which will allow the information to be used by those speakers whose message the State supports. Cf. Rowan v. Post Office Dept., 397 U. S. 728. Respondents suggest a further defect lies in §4631(d)’s presumption of applicability absent an individual election to the contrary. Reliance on a prior election, however, would not save a privacy measure that imposed an unjustified burden on protected expression. Vermont also asserts that its broad content-based rule is necessary to avoid harassment, but doctors can simply decline to meet with detailers. Cf. Watchtower Bible & Tract Soc. of N. Y., Inc. v. Village of Stratton, 536 U. S. 150, 168. Vermont further argues that detailers’ use of prescriber-identifying information undermines the doctor-patient relationship by allowing detailers to influence treatment decisions. But if pharmaceutical marketing affects treatment decisions, it can do so only because it is persuasive. Fear that speech might persuade provides no lawful basis for quieting it. Pp. 17–21. (c) While Vermont’s goals of lowering the costs of medical services and promoting public health may be proper, §4631(d) does not advance them in a permissible way. Vermont seeks to achieve those objectives through the indirect means of restraining certain speech by certain speakers—i.e., by diminishing detailers’ ability to influence prescription decisions. But “the fear that people would make bad decisions if given truthful information” cannot justify content-based burdens on speech. Thompson v. Western States Medical Center, 535 U. S. 357, 374. That precept applies with full force when the audience—here, prescribing physicians—consists of “sophisticated and experienced” consumers. Edenfield v. Fane, 507 U. S. 761, 775. The instant law’s defect is made clear by the fact that many listeners find detailing instructive. Vermont may be displeased that detailers with prescriber-indentifying information are effective in promoting brand-name drugs, but the State may not burden protected expression in order to tilt public debate in a preferred direction. Vermont nowhere contends that its law will prevent false or misleading speech within the meaning of this Court’s First Amendment precedents. The State’s interest in burdening detailers’ speech thus turns on nothing more than a difference of opinion. Pp. 21–24. 630 F. 3d 263, affirmed. Kennedy, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Thomas, Alito, and Sotomayor, JJ., joined. Breyer, J., filed a dissenting opinion, in which Ginsburg and Kagan, JJ., joined.

Vermont law restricts the sale, disclosure, and use of pharmacy records that reveal the prescribing practices of individual doctors. Vt. Stat. Ann., Tit. 18, §4631 (Supp. 2010). Subject to certain exceptions, the information may not be sold, disclosed by pharmacies for marketing purposes, or used for marketing by pharmaceutical manufacturers. Vermont argues that its prohibitions safeguard medical privacy and diminish the likelihood that marketing will lead to prescription decisions not in the best interests of patients or the State. It can be assumed that these interests are significant. Speech in aid of pharmaceutical marketing, however, is a form of expression protected by the Free Speech Clause of the First Amendment. As a consequence, Vermont’s statute must be subjected to heightened judicial scrutiny. The law cannot satisfy that standard. I A Pharmaceutical manufacturers promote their drugs to doctors through a process called “detailing.” This often in- volves a scheduled visit to a doctor’s office to persuade the doctor to prescribe a particular pharmaceutical. Detailers bring drug samples as well as medical studies that explain the “details” and potential advantages of various prescription drugs. Interested physicians listen, ask questions, and receive followup data. Salespersons can be more effective when they know the background and purchasing preferences of their clientele, and pharmaceutical salespersons are no exception. Knowledge of a physi- cian’s prescription practices—called “prescriber-identifying information”—enables a detailer better to ascertain which doctors are likely to be interested in a particular drug and how best to present a particular sales message. Detailing is an expensive undertaking, so pharmaceutical companies most often use it to promote high-profit brand-name drugs protected by patent. Once a brand-name drug’s patent expires, less expensive bioequivalent generic alternatives are manufactured and sold. Pharmacies, as a matter of business routine and federal law, receive prescriber-identifying information when processing prescriptions. See 21 U. S. C. §353(b); see also Vt. Bd. of Pharmacy Admin. Rule 9.1 (2009); Rule 9.2. Many pharmacies sell this information to “data miners,” firms that analyze prescriber-identifying information and produce reports on prescriber behavior. Data miners lease these reports to pharmaceutical manufacturers subject to nondisclosure agreements. Detailers, who represent the manufacturers, then use the reports to refine their marketing tactics and increase sales. In 2007, Vermont enacted the Prescription Confidentiality Law. The measure is also referred to as Act 80. It has several components. The central provision of the present case is §4631(d). “A health insurer, a self-insured employer, an electronic transmission intermediary, a pharmacy, or other similar entity shall not sell, license, or exchange for value regulated records containing prescriber-identifiable information, nor permit the use of regulated records containing prescriber-identifiable information for marketing or promoting a prescription drug, unless the prescriber consents … . Pharmaceutical manufacturers and pharmaceutical marketers shall not use prescriber-identifiable information for marketing or promoting a prescription drug unless the prescriber consents … .” The quoted provision has three component parts. The provision begins by prohibiting pharmacies, health insurers, and similar entities from selling prescriber-identifying information, absent the prescriber’s consent. The parties here dispute whether this clause applies to all sales or only to sales for marketing. The provision then goes on to prohibit pharmacies, health insurers, and similar enti- ties from allowing prescriber-identifying information to be used for marketing, unless the prescriber consents. This prohibition in effect bars pharmacies from disclosing the information for marketing purposes. Finally, the provision’s second sentence bars pharmaceutical manufacturers and pharmaceutical marketers from using prescriber-identifying information for marketing, again absent the prescriber’s consent. The Vermont attorney general may pursue civil remedies against violators. §4631(f). Separate statutory provisions elaborate the scope of the prohibitions set out in §4631(d). “Marketing” is defined to include “advertising, promotion, or any activity” that is “used to influence sales or the market share of a prescription drug.” §4631(b)(5). Section 4631(c)(1) further provides that Vermont’s Department of Health must allow “a prescriber to give consent for his or her identifying information to be used for the purposes” identified in §4631(d). Finally, the Act’s prohibitions on sale, disclosure, and use are subject to a list of exceptions. For example, prescriber-identifying information may be disseminated or used for “health care research”; to enforce “compliance” with health insurance formularies, or preferred drug lists; for “care management educational communications provided to” patients on such matters as “treatment options”; for law enforcement operations; and for purposes “otherwise provided by law.” §4631(e). Act 80 also authorized funds for an “evidence-based pre-scription drug education program” designed to provide doctors and others with “information and education on the therapeutic and cost-effective utilization of prescription drugs.” §4622(a)(1). An express aim of the program is to advise prescribers “about commonly used brand-name drugs for which the patent has expired” or will soon expire. §4622(a)(2). Similar efforts to promote the use of generic pharmaceuticals are sometimes referred to as “counter-detailing.” App. 211; see also IMS Health Inc. v. Ayotte, 550 F. 3d 42, 91 (CA1 2008) (Lipez, J., concurring and dissenting). The counterdetailer’s recommended substitute may be an older, less expensive drug and not a bioequivalent of the brand-name drug the physician might otherwise prescribe. Like the pharmaceutical manufacturers whose efforts they hope to resist, counterdetailers in some States use prescriber-identifying information to increase their effectiveness. States themselves may supply the prescriber-identifying information used in these programs. See App. 313; id., at 375 (“[W]e use the data given to us by the State of Pennsylvania … to figure out which physicians to talk to”); see also id., at 427–429 (Director of the Office of Vermont Health Access explaining that the office collects prescriber-identifying information but “does not at this point in time have a counterdetailing or detailing effort”). As first enacted, Act 80 also required detailers to provide information about alternative treatment options. The Vermont Legislature, however, later repealed that provision. 2008 Vt. Laws No. 89, §3. Act 80 was accompanied by legislative findings. Vt. Acts No. 80, §1. Vermont found, for example, that the “goals of marketing programs are often in conflict with the goals of the state” and that the “marketplace for ideas on medicine safety and effectiveness is frequently one-sided in that brand-name companies invest in expensive pharmaceutical marketing campaigns to doctors.” §§1(3), (4). Detailing, in the legislature’s view, caused doctors to make decisions based on “incomplete and biased information.” §1(4). Because they “are unable to take the time to research the quickly changing pharmaceutical market,” Vermont doctors “rely on information provided by pharmaceutical representatives.” §1(13). The legislature further found that detailing increases the cost of health care and health insurance, §1(15); encourages hasty and excessive reliance on brand-name drugs, before the profession has observed their effectiveness as compared with older and less expensive generic alternatives, §1(7); and fosters disruptive and repeated marketing visits tantamount to harassment, §§1(27)–(28). The legislative findings further noted that use of prescriber-identifying information “increase[s] the effect of detailing programs” by allowing detailers to target their visits to particular doctors. §§1(23)–(26). Use of prescriber-identifying data also helps detailers shape their messages by “tailoring” their “presentations to individual prescriber styles, preferences, and attitudes.” §1(25). B The present case involves two consolidated suits. One was brought by three Vermont data miners, the other by an association of pharmaceutical manufacturers that produce brand-name drugs. These entities are the respondents here. Contending that §4631(d) violates their First Amendment rights as incorporated by the Fourteenth Amendment, the respondents sought declaratory and injunctive relief against the petitioners, the Attorney General and other officials of the State of Vermont. After a bench trial, the United States District Court for the District of Vermont denied relief. 631 F. Supp. 2d 434 (2009). The District Court found that “[p]harmaceutical manufacturers are essentially the only paying customers of the data vendor industry” and that, because detailing unpatented generic drugs is not “cost-effective,” pharmaceutical sales representatives “detail only branded drugs.” Id., at 451, 442. As the District Court further con- cluded, “the Legislature’s determination that [prescriber-identifying] data is an effective marketing tool that enables detailers to increase sales of new drugs is supported in the record.” Id., at 451. The United States Court of Appeals for the Second Circuit reversed and remanded. It held that §4631(d) violates the First Amendment by burdening the speech of pharmaceutical marketers and data miners without an adequate justification. 630 F. 3d 263. Judge Livingston dissented. The decision of the Second Circuit is in conflict with de- cisions of the United States Court of Appeals for the First Circuit concerning similar legislation enacted by Maine and New Hampshire. See IMS Health Inc. v. Mills, 616 F. 3d 7 (CA1 2010) (Maine); Ayotte, supra (New Hamp- shire). Recognizing a division of authority regarding the constitutionality of state statutes, this Court granted certiorari. 562 U. S. __ (2011). II The beginning point is the text of §4631(d). In the pro- ceedings below, Vermont stated that the first sentence of §4631(d) prohibits pharmacies and other regulated entities from selling or disseminating prescriber-identifying information for marketing. The information, in other words, could be sold or given away for purposes other than marketing. The District Court and the Court of Appeals accepted the State’s reading. See 630 F. 3d, at 276. At oral argument in this Court, however, the State for the first time advanced an alternative reading of §4631(d)—namely, that pharmacies, health insurers, and similar entities may not sell prescriber-identifying information for any purpose, subject to the statutory exceptions set out at §4631(e). See Tr. of Oral Arg. 19–20. It might be argued that the State’s newfound interpretation comes too late in the day. See Sprietsma v. Mercury Marine, 537 U. S. 51, 56, n. 4 (2002) (waiver); New Hampshire v. Maine, 532 U. S. 742, 749 (2001) (judicial estoppel). The respondents, the District Court, and the Court of Appeals were entitled to rely on the State’s plausible interpretation of the law it is charged with enforcing. For the State to change its position is particularly troubling in a First Amendment case, where plaintiffs have a special interest in obtaining a prompt adjudication of their rights, despite potential ambiguities of state law. See Houston v. Hill, 482 U. S. 451, 467–468, and n. 17 (1987); Zwickler v. Koota, 389 U. S. 241, 252 (1967). In any event, §4631(d) cannot be sustained even under the interpretation the State now adopts. As a consequence this Court can assume that the opening clause of §4631(d) prohibits pharmacies, health insurers, and similar entities from selling prescriber-identifying information, subject to the statutory exceptions set out at §4631(e). Under that reading, pharmacies may sell the information to private or academic researchers, see §4631(e)(1), but not, for example, to pharmaceutical marketers. There is no dispute as to the remainder of §4631(d). It prohibits pharmacies, health insurers, and similar entities from disclosing or otherwise allowing prescriber-identifying information to be used for marketing. And it bars pharmaceutical manufacturers and detailers from using the information for marketing. The questions now are whether §4631(d) must be tested by heightened judicial scrutiny and, if so, whether the State can justify the law. A 1 On its face, Vermont’s law enacts content- and speaker-based restrictions on the sale, disclosure, and use of prescriber-identifying information. The provision first forbids sale subject to exceptions based in large part on the content of a purchaser’s speech. For example, those who wish to engage in certain “educational communications,” §4631(e)(4), may purchase the information. The measure then bars any disclosure when recipient speakers will use the information for marketing. Finally, the provision’s second sentence prohibits pharmaceutical manufacturers from using the information for marketing. The statute thus disfavors marketing, that is, speech with a particular content. More than that, the statute disfavors specific speakers, namely pharmaceutical manufacturers. As a result of these content- and speaker-based rules, detailers cannot obtain prescriber-identifying information, even though the information may be purchased or acquired by other speakers with diverse purposes and viewpoints. Detailers are likewise barred from using the information for marketing, even though the information may be used by a wide range of other speakers. For example, it appears that Vermont could supply academic organizations with prescriber-identifying information to use in countering the messages of brand-name pharmaceutical manufacturers and in promoting the prescription of generic drugs. But §4631(d) leaves detailers no means of purchasing, acquiring, or using prescriber-identifying information. The law on its face burdens disfavored speech by disfavored speakers. Any doubt that §4631(d) imposes an aimed, content-based burden on detailers is dispelled by the record and by formal legislative findings. As the District Court noted, “[p]harmaceutical manufacturers are essentially the only paying customers of the data vendor industry”; and the almost invariable rule is that detailing by pharmaceutical manufacturers is in support of brand-name drugs. 631 F. Supp. 2d, at 451. Vermont’s law thus has the effect of preventing detailers—and only detailers—from communicating with physicians in an effective and informative manner. Cf. Edenfield v. Fane, 507 U. S. 761, 766 (1993) (explaining the “considerable value” of in-person solicitation). Formal legislative findings accompanying §4631(d) confirm that the law’s express purpose and practical effect are to diminish the effectiveness of marketing by manufacturers of brand-name drugs. Just as the “inevitable effect of a statute on its face may render it unconstitutional,” a statute’s stated purposes may also be considered. United States v. O’Brien, 391 U. S. 367, 384 (1968). Here, the Vermont Legislature explained that detailers, in particular those who promote brand-name drugs, convey messages that “are often in conflict with the goals of the state.” 2007 Vt. No. 80, §1(3). The legislature designed §4631(d) to target those speakers and their messages for disfavored treatment. “In its practical operation,” Vermont’s law “goes even beyond mere content discrimination, to actual viewpoint discrimination.” R. A. V. v. St. Paul, 505 U. S. 377, 391 (1992). Given the legislature’s expressed statement of purpose, it is apparent that §4631(d) imposes burdens that are based on the content of speech and that are aimed at a particular viewpoint. Act 80 is designed to impose a specific, content-based burden on protected expression. It follows that heightened judicial scrutiny is warranted. See Cincinnati v. Discovery Network, Inc., 507 U. S. 410, 418 (1993) (applying heightened scrutiny to “a categorical prohibition on the use of newsracks to disseminate commercial messages”); id., at 429 (“[T]he very basis for the regulation is the difference in content between ordinary newspapers and commercial speech” in the form of “commercial handbills … . Thus, by any commonsense understanding of the term, the ban in this case is ‘content based’ ” (some internal quotation marks omitted)); see also Turner Broadcasting System, Inc. v. FCC, 512 U. S. 622, 658 (1994) (explaining that strict scrutiny applies to regulations reflecting “aversion” to what “disfavored speakers” have to say). The Court has recognized that the “distinction between laws burdening and laws banning speech is but a matter of degree” and that the “Government’s content-based burdens must satisfy the same rigorous scrutiny as its content-based bans.” United States v. Playboy Entertainment Group, Inc., 529 U. S. 803, 812 (2000). Lawmakers may no more silence unwanted speech by burdening its utterance than by censoring its content. See Simon & Schuster, Inc. v. Mem- bers of N. Y. State Crime Victims Bd., 502 U. S. 105, 115 (1991) (content-based financial burden); Minneapolis Star & Tribune Co. v. Minnesota Comm’r of Revenue, 460 U. S. 575 (1983) (speaker-based financial burden). The First Amendment requires heightened scrutiny whenever the government creates “a regulation of speech because of disagreement with the message it conveys.” Ward v. Rock Against Racism, 491 U. S. 781, 791 (1989); see also Renton v. Playtime Theatres, Inc., 475 U. S. 41, 48 (1986) (explaining that “ ‘content-neutral’ speech regulations” are “those that are justified without reference to the content of the regulated speech” (internal quotation marks omitted)). A government bent on frustrating an impending demonstration might pass a law demanding two years’ notice before the issuance of parade permits. Even if the hypothetical measure on its face appeared neutral as to content and speaker, its purpose to suppress speech and its unjustified burdens on expression would render it unconstitutional. Ibid. Commercial speech is no exception. See Discovery Network, supra, at 429–430 (commercial speech restriction lacking a “neutral justification” was not content neutral). A “consumer’s concern for the free flow of commercial speech often may be far keener than his concern for urgent political dialogue.” Bates v. State Bar of Ariz., 433 U. S. 350, 364 (1977). That reality has great relevance in the fields of medicine and public health, where information can save lives. 2 The State argues that heightened judicial scrutiny is unwarranted because its law is a mere commercial regulation. It is true that restrictions on protected expression are distinct from restrictions on economic activity or, more generally, on nonexpressive conduct. It is also true that the First Amendment does not prevent restrictions directed at commerce or conduct from imposing inciden- tal burdens on speech. That is why a ban on race-based hiring may require employers to remove “ ‘White Applicants Only’ ” signs, Rumsfeld v. Forum for Academic and Institutional Rights, Inc., 547 U. S. 47, 62 (2006); why “an ordinance against outdoor fires” might forbid “burning a flag,” R. A. V., supra, at 385; and why antitrust laws can prohibit “agreements in restraint of trade,” Giboney v. Empire Storage & Ice Co., 336 U. S. 490, 502 (1949). But §4631(d) imposes more than an incidental burden on protected expression. Both on its face and in its practical operation, Vermont’s law imposes a burden based on the content of speech and the identity of the speaker. See supra, at 8–11. While the burdened speech results from an economic motive, so too does a great deal of vital expression. See Bigelow v. Virginia, 421 U. S. 809, 818 (1975); New York Times Co. v. Sullivan, 376 U. S. 254, 266 (1964); see also United States v. United Foods, Inc., 533 U. S. 405, 410–411 (2001) (applying “First Amendment scrutiny” where speech effects were not incidental and noting that “those whose business and livelihood depend in some way upon the product involved no doubt deem First Amendment protection to be just as important for them as it is for other discrete, little noticed groups”). Vermont’s law does not simply have an effect on speech, but is directed at certain content and is aimed at particular speakers. The Constitution “does not enact Mr. Herbert Spencer’s Social Statics.” Lochner v. New York, 198 U. S. 45, 75 (1905) (Holmes, J., dissenting). It does enact the First Amendment. Vermont further argues that §4631(d) regulates not speech but simply access to information. Prescriber-identifying information was generated in compliance with a legal mandate, the State argues, and so could be considered a kind of governmental information. This argument finds some support in Los Angeles Police Dept. v. United Reporting Publishing Corp., 528 U. S. 32 (1999), where the Court held that a plaintiff could not raise a facial challenge to a content-based restriction on access to government- held information. Because no private party faced a threat of legal punishment, the Court characterized the law at issue as “nothing more than a governmental denial of access to information in its possession.” Id., at 40. Under those circumstances the special reasons for permitting First Amendment plaintiffs to invoke the rights of others did not apply. Id., at 38–39. Having found that the plaintiff could not raise a facial challenge, the Court remanded for consideration of an as-applied challenge. Id., at 41. United Reporting is thus a case about the availability of facial challenges. The Court did not rule on the merits of any First Amendment claim. United Reporting is distinguishable in at least two respects. First, Vermont has imposed a restriction on access to information in private hands. This confronts the Court with a point reserved, and a situation not addressed, in United Reporting. Here, unlike in United Reporting, we do have “a case in which the government is prohibiting a speaker from conveying information that the speaker already possesses.” Id., at 40. The difference is significant. An individual’s right to speak is implicated when information he or she possesses is subjected to “restraints on the way in which the information might be used” or disseminated. Seattle Times Co. v. Rhinehart, 467 U. S. 20, 32 (1984); see also Bartnicki v. Vopper, 532 U. S. 514, 527 (2001); Florida Star v. B. J. F., 491 U. S. 524 (1989); New York Times Co. v. United States, 403 U. S. 713 (1971) (per curiam). In Seattle Times, this Court applied heightened judicial scrutiny before sustaining a trial court order prohibiting a newspaper’s disclosure of information it learned through coercive discovery. It is true that the respondents here, unlike the newspaper in Seattle Times, do not themselves possess information whose disclosure has been curtailed. That information, however, is in the hands of pharmacies and other private entities. There is no question that the “threat of prosecution … hangs over their heads.” United Reporting, 528 U. S., at 41. For that reason United Reporting does not bar respondents’ facial challenge. United Reporting is distinguishable for a second and even more important reason. The plaintiff in United Reporting had neither “attempt[ed] to qualify” for access to the government’s information nor presented an as-applied claim in this Court. Id., at 40. As a result, the Court assumed that the plaintiff had not suffered a personal First Amendment injury and could prevail only by invoking the rights of others through a facial challenge. Here, by contrast, the respondents claim—with good reason—that §4631(d) burdens their own speech. That argument finds support in the separate writings in United Reporting, which were joined by eight Justices. All of those writings recognized that restrictions on the disclosure of government-held information can facilitate or burden the expression of potential recipients and so transgress the First Amendment. See id., at 42 (Scalia, J., concurring) (suggesting that “a restriction upon access that allows access to the press … but at the same time denies access to persons who wish to use the information for certain speech purposes, is in reality a restriction upon speech”); id., at 43 (Ginsburg, J., concurring) (noting that “the provision of [government] information is a kind of subsidy to people who wish to speak” about certain subjects, “and once a State decides to make such a benefit available to the public, there are no doubt limits to its freedom to decide how that benefit will be distributed”); id., at 46 (Stevens, J., dissenting) (concluding that, “because the State’s discrimination is based on its desire to prevent the information from being used for constitutionally protected purposes, [i]t must assume the burden of justifying its conduct”). Vermont’s law imposes a content- and speaker-based burden on respondents’ own speech. That consideration provides a separate basis for distinguishing United Reporting and requires heightened judicial scrutiny. The State also contends that heightened judicial scrutiny is unwarranted in this case because sales, transfer, and use of prescriber-identifying information are conduct, not speech. Consistent with that submission, the United States Court of Appeals for the First Circuit has characterized prescriber-identifying information as a mere “commodity” with no greater entitlement to First Amend- ment protection than “beef jerky.” Ayotte, 550 F. 3d, at 52–53. In contrast the courts below concluded that a prohibition on the sale of prescriber-identifying information is a content-based rule akin to a ban on the sale of cookbooks, laboratory results, or train schedules. See 630 F. 3d, at 271–272 (“The First Amendment protects even dry information, devoid of advocacy, political relevance, or artistic expression” (internal quotation marks and alteration omitted)); 631 F. Supp. 2d, at 445 (“A restriction on disclosure is a regulation of speech, and the ‘sale’ of [information] is simply disclosure for profit”). This Court has held that the creation and dissemination of information are speech within the meaning of the First Amendment. See, e.g., Bartnicki, supra, at 527 (“[I]f the acts of ‘disclosing’ and ‘publishing’ information do not constitute speech, it is hard to imagine what does fall within that category, as distinct from the category of expressive conduct” (some internal quotation marks omitted)); Rubin v. Coors Brewing Co., 514 U. S. 476, 481 (1995) (“information on beer labels” is speech); Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U. S. 749, 759 (1985) (plurality opinion) (credit report is “speech”). Facts, after all, are the beginning point for much of the speech that is most essential to advance human knowledge and to conduct human affairs. There is thus a strong argument that prescriber-identifying information is speech for First Amendment purposes. The State asks for an exception to the rule that information is speech, but there is no need to consider that request in this case. The State has imposed content- and speaker-based restrictions on the availability and use of prescriber-identifying information. So long as they do not engage in marketing, many speakers can obtain and use the information. But detailers cannot. Vermont’s statute could be compared with a law prohibiting trade magazines from purchasing or using ink. Cf. Minneapolis Star, 460 U. S. 575. Like that hypothetical law, §4631(d) imposes a speaker- and content-based burden on protected expression, and that circumstance is sufficient to justify application of heightened scrutiny. As a consequence, this case can be resolved even assuming, as the State argues, that prescriber-identifying information is a mere commodity. B In the ordinary case it is all but dispositive to conclude that a law is content-based and, in practice, viewpoint-discriminatory. See R. A. V., 505 U. S., at 382 (“Content-based regulations are presumptively invalid”); id., at 391–392. The State argues that a different analysis applies here because, assuming §4631(d) burdens speech at all, it at most burdens only commercial speech. As in previous cases, however, the outcome is the same whether a special commercial speech inquiry or a stricter form of judicial scrutiny is applied. See, e.g., Greater New Orleans Broadcasting Assn., Inc. v. United States, 527 U. S. 173, 184 (1999). For the same reason there is no need to determine whether all speech hampered by §4631(d) is commercial, as our cases have used that term. Cf. Board of Trustees of State Univ. of N. Y. v. Fox, 492 U. S. 469, 474 (1989) (discussing whether “pure speech and commercial speech” were inextricably intertwined, so that “the entirety must … be classified as noncommercial”). Under a commercial speech inquiry, it is the State’s burden to justify its content-based law as consistent with the First Amendment. Thompson v. Western States Medical Center, 535 U. S. 357, 373 (2002). To sustain the targeted, content-based burden §4631(d) imposes on protected expression, the State must show at least that the statute directly advances a substantial governmental interest and that the measure is drawn to achieve that interest. See Fox, supra, at 480–481; Central Hudson Gas & Elec. Corp. v. Public Serv. Comm’n of N. Y., 447 U. S. 557, 566 (1980). There must be a “fit between the legislature’s ends and the means chosen to accomplish those ends.” Fox, supra, at 480 (internal quotation marks omitted). As in other contexts, these standards ensure not only that the State’s interests are proportional to the result- ing burdens placed on speech but also that the law does not seek to suppress a disfavored message. See Turner Broadcasting, 512 U. S., at 662–663. The State’s asserted justifications for §4631(d) come under two general headings. First, the State contends that its law is necessary to protect medical privacy, including physician confidentiality, avoidance of harassment, and the integrity of the doctor-patient relationship. Second, the State argues that §4631(d) is integral to the achievement of policy objectives—namely, improved public health and reduced healthcare costs. Neither justification withstands scrutiny. 1 Vermont argues that its physicians have a “reasonable expectation” that their prescriber-identifying information “will not be used for purposes other than … filling and processing” prescriptions. See 2007 Vt. Laws No. 80, §1(29). It may be assumed that, for many reasons, physicians have an interest in keeping their prescription decisions confidential. But §4631(d) is not drawn to serve that interest. Under Vermont’s law, pharmacies may share prescriber-identifying information with anyone for any rea- son save one: They must not allow the information to be used for marketing. Exceptions further allow pharmacies to sell prescriber-identifying information for certain purposes, including “health care research.” §4631(e). And the measure permits insurers, researchers, journalists, the State itself, and others to use the information. See §4631(d); cf. App. 370–372; id., at 211. All but conceding that §4631(d) does not in itself advance confidentiality interests, the State suggests that other laws might impose separate bars on the disclosure of prescriber-identifying information. See Vt. Bd. of Pharmacy Admin. Rule 20.1. But the potential effectiveness of other measures cannot justify the distinctive set of prohibitions and sanctions imposed by §4631(d). Perhaps the State could have addressed physician confidentiality through “a more coherent policy.” Greater New Orleans Broadcasting, supra, at 195; see also Discovery Network, 507 U. S., at 428. For instance, the State might have advanced its asserted privacy interest by allowing the information’s sale or disclosure in only a few narrow and well-justified circumstances. See, e.g., Health Insurance Portability and Accountability Act of 1996, 42 U. S. C. §1320d–2; 45 CFR pts. 160 and 164 (2010). A statute of that type would present quite a different case than the one presented here. But the State did not enact a statute with that purpose or design. Instead, Vermont made prescriber-identifying information available to an almost limitless audience. The explicit structure of the statute allows the information to be studied and used by all but a narrow class of disfavored speakers. Given the information’s widespread availability and many permissible uses, the State’s asserted interest in physician confidentiality does not justify the burden that §4631(d) places on protected expression. The State points out that it allows doctors to forgo the advantages of §4631(d) by consenting to the sale, disclosure, and use of their prescriber-identifying information. See §4631(c)(1). It is true that private decisionmaking can avoid governmental partiality and thus insulate privacy measures from First Amendment challenge. See Rowan v. Post Office Dept., 397 U. S. 728 (1970); cf. Bolger v. Youngs Drug Products Corp., 463 U. S. 60, 72 (1983). But that principle is inapposite here. Vermont has given its doctors a contrived choice: Either consent, which will allow your prescriber-identifying information to be disseminated and used without constraint; or, withhold consent, which will allow your information to be used by those speakers whose message the State supports. Section 4631(d) may offer a limited degree of privacy, but only on terms favorable to the speech the State prefers. Cf. Rowan, supra, at 734, 737, 739, n. 6 (sustaining a law that allowed private parties to make “unfettered,” “unlimited,” and “unreviewable” choices regarding their own privacy). This is not to say that all privacy measures must avoid content-based rules. Here, however, the State has conditioned privacy on acceptance of a content-based rule that is not drawn to serve the State’s asserted interest. To obtain the limited privacy allowed by §4631(d), Vermont physicians are forced to acquiesce in the State’s goal of burdening disfavored speech by disfavored speakers. Respondents suggest that a further defect of §4631(d) lies in its presumption of applicability absent a physician’s election to the contrary. Vermont’s law might burden less speech if it came into operation only after an individual choice, but a revision to that effect would not necessarily save §4631(d). Even reliance on a prior election would not suffice, for instance, if available categories of coverage by design favored speakers of one political persuasion over another. Rules that burden protected expression may not be sustained when the options provided by the State are too narrow to advance legitimate interests or too broad to protect speech. As already explained, §4631(d) permits extensive use of prescriber-identifying information and so does not advance the State’s asserted interest in physician confidentiality. The limited range of available privacy options instead reflects the State’s impermissible purpose to burden disfavored speech. Vermont’s argument accordingly fails, even if the availability and scope of private election might be relevant in other contexts, as when the statute’s design is unrelated to any purpose to advance a preferred message. The State also contends that §4631(d) protects doctors from “harassing sales behaviors.” 2007 Vt. Laws No. 80, §1(28). “Some doctors in Vermont are experiencing an undesired increase in the aggressiveness of pharmaceutical sales representatives,” the Vermont Legislature found, “and a few have reported that they felt coerced and harassed.” §1(20). It is doubtful that concern for “a few” physicians who may have “felt coerced and harassed” by pharmaceutical marketers can sustain a broad content-based rule like §4631(d). Many are those who must endure speech they do not like, but that is a necessary cost of freedom. See Erznoznik v. Jacksonville, 422 U. S. 205, 210–211 (1975); Cohen v. California, 403 U. S. 15, 21 (1971). In any event the State offers no explanation why remedies other than content-based rules would be inadequate. See 44 Liquormart, Inc. v. Rhode Island, 517 U. S. 484, 503 (1996) (opinion of Stevens, J.). Physicians can, and often do, simply decline to meet with detailers, including detailers who use prescriber-identifying information. See, e.g., App. 180, 333–334. Doctors who wish to forgo detailing altogether are free to give “No Solicitation” or “No Detailing” instructions to their office managers or to receptionists at their places of work. Personal privacy even in one’s own home receives “ample protection” from the “resident’s unquestioned right to refuse to engage in conversation with unwelcome visitors.” Watchtower Bible & Tract Soc. of N. Y., Inc. v. Village of Stratton, 536 U. S. 150, 168 (2002); see also Bolger, supra, at 72. A physician’s office is no more private and is entitled to no greater protection. Vermont argues that detailers’ use of prescriber-identifying information undermines the doctor-patient relationship by allowing detailers to influence treatment decisions. According to the State, “unwanted pressure occurs” when doctors learn that their prescription decisions are being “monitored” by detailers. 2007 Vt. Laws No. 80, §1(27). Some physicians accuse detailers of “spying” or of engaging in “underhanded” conduct in order to “subvert” prescription decisions. App. 336, 380, 407–408; see also id., at 326–328. And Vermont claims that detailing makes people “anxious” about whether doctors have their patients’ best interests at heart. Id., at 327. But the State does not explain why detailers’ use of prescriber-identifying information is more likely to prompt these objections than many other uses permitted by §4631(d). In any event, this asserted interest is contrary to basic First Amendment principles. Speech remains protected even when it may “stir people to action,” “move them to tears,” or “inflict great pain.” Snyder v. Phelps, 562 U. S. ___, ___ (2011) (slip op., at 15). The more benign and, many would say, beneficial speech of pharmaceutical marketing is also entitled to the protection of the First Amendment. If pharmaceutical marketing affects treatment decisions, it does so because doctors find it persuasive. Absent circumstances far from those presented here, the fear that speech might persuade provides no lawful basis for quieting it. Brandenburg v. Ohio, 395 U. S. 444, 447 (1969) (per curiam). 2 The State contends that §4631(d) advances impor- tant public policy goals by lowering the costs of medical services and promoting public health. If prescriber-identifying information were available for use by detailers, the State contends, then detailing would be effective in promoting brand-name drugs that are more expensive and less safe than generic alternatives. This logic is set out at length in the legislative findings accompanying §4631(d). Yet at oral argument here, the State declined to acknowledge that §4631(d)’s objective purpose and practical effect were to inhibit detailing and alter doctors’ prescription decisions. See Tr. of Oral Arg. 5–6. The State’s reluctance to embrace its own legislature’s rationale reflects the vulnerability of its position. While Vermont’s stated policy goals may be proper, §4631(d) does not advance them in a permissible way. As the Court of Appeals noted, the “state’s own explanation of how” §4631(d) “advances its interests cannot be said to be direct.” 630 F. 3d, at 277. The State seeks to achieve its policy objectives through the indirect means of restraining certain speech by certain speakers—that is, by diminishing detailers’ ability to influence prescription decisions. Those who seek to censor or burden free expression often assert that disfavored speech has adverse effects. But the “fear that people would make bad decisions if given truthful information” cannot justify content-based burdens on speech. Thompson, 535 U. S., at 374; see also Virginia Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U. S. 748, 769–770 (1976). “The First Amendment directs us to be especially skeptical of regulations that seek to keep people in the dark for what the government perceives to be their own good.” 44 Liquormart, supra, at 503 (opinion of Stevens, J.); see also Linmark Associates, Inc. v. Willingboro, 431 U. S. 85, 97 (1977). These precepts apply with full force when the audience, in this case prescribing physicians, consists of “sophisticated and experienced” consumers. Edenfield, 507 U. S., at 775. As Vermont’s legislative findings acknowledge, the prem- ise of §4631(d) is that the force of speech can justify the government’s attempts to stifle it. Indeed the State defends the law by insisting that “pharmaceutical marketing has a strong influence on doctors’ prescribing practices.” Brief for Petitioners 49–50. This reasoning is incompatible with the First Amendment. In an attempt to reverse a disfavored trend in public opinion, a State could not ban campaigning with slogans, picketing with signs, or marching during the daytime. Likewise the State may not seek to remove a popular but disfavored product from the marketplace by prohibiting truthful, nonmisleading advertisements that contain impressive endorsements or catchy jingles. That the State finds expression too persuasive does not permit it to quiet the speech or to burden its messengers. The defect in Vermont’s law is made clear by the fact that many listeners find detailing instructive. Indeed the record demonstrates that some Vermont doctors view targeted detailing based on prescriber-identifying information as “very helpful” because it allows detailers to shape their messages to each doctor’s practice. App. 274; see also id., at 181, 218, 271–272. Even the United States, which appeared here in support of Vermont, took care to dispute the State’s “unwarranted view that the dangers of [n]ew drugs outweigh their benefits to patients.” Brief for United States as Amicus Curiae 24, n. 4. There are divergent views regarding detailing and the prescription of brand-name drugs. Under the Constitution, resolution of that debate must result from free and uninhibited speech. As one Vermont physician put it: “We have a saying in medicine, information is power. And the more you know, or anyone knows, the better decisions can be made.” App. 279. There are similar sayings in law, including that “information is not in itself harmful, that people will perceive their own best interests if only they are well enough informed, and that the best means to that end is to open the channels of communication rather than to close them.” Virginia Bd., 425 U. S., at 770. The choice “between the dangers of suppressing information, and the dangers of its misuse if it is freely available” is one that “the First Amendment makes for us.” Ibid. Vermont may be displeased that detailers who use prescriber-identifying information are effective in promoting brand-name drugs. The State can express that view through its own speech. See Linmark, 431 U. S., at 97; cf. §4622(a)(1) (establishing a prescription drug educational program). But a State’s failure to persuade does not allow it to hamstring the opposition. The State may not burden the speech of others in order to tilt public debate in a preferred direction. “The commercial marketplace, like other spheres of our social and cultural life, provides a forum where ideas and information flourish. Some of the ideas and information are vital, some of slight worth. But the general rule is that the speaker and the audience, not the government, assess the value of the information presented.” Edenfield, supra, at 767. It is true that content-based restrictions on protected expression are sometimes permissible, and that principle applies to commercial speech. Indeed the government’s legitimate interest in protecting consumers from “commercial harms” explains “why commercial speech can be subject to greater governmental regulation than noncommercial speech.” Discovery Network, 507 U. S., at 426; see also 44 Liquormart, 517 U. S., 502 (opinion of Stevens, J.). The Court has noted, for example, that “a State may choose to regulate price advertising in one industry but not in others, because the risk of fraud … is in its view greater there.” R. A. V., 505 U. S., at 388–389 (citing Virginia Bd., supra, at 771–772). Here, however, Vermont has not shown that its law has a neutral justification. The State nowhere contends that detailing is false or misleading within the meaning of this Court’s First Amendment precedents. See Thompson, 535 U. S., at 373. Nor does the State argue that the provision challenged here will prevent false or misleading speech. Cf. post, at 10–11 (Breyer, J., dissenting) (collecting regulations that the government might defend on this ground). The State’s interest in burdening the speech of detailers instead turns on nothing more than a difference of opinion. See Bolger, 463 U. S., at 69; Thompson, supra, at 376. * * * The capacity of technology to find and publish personal information, including records required by the government, presents serious and unresolved issues with respect to personal privacy and the dignity it seeks to secure. In considering how to protect those interests, however, the State cannot engage in content-based discrimination to advance its own side of a debate. If Vermont’s statute provided that prescriber-identifying information could not be sold or disclosed except in narrow circumstances then the State might have a stronger position. Here, however, the State gives possessors of the information broad discretion and wide latitude in disclosing the information, while at the same time restricting the information’s use by some speakers and for some purposes, even while the State itself can use the information to counter the speech it seeks to suppress. Privacy is a concept too integral to the person and a right too essential to freedom to allow its manipulation to support just those ideas the government prefers. When it enacted §4631(d), the Vermont Legislature found that the “marketplace for ideas on medicine safety and effectiveness is frequently one-sided in that brand-name companies invest in expensive pharmaceutical marketing campaigns to doctors.” 2007 Vt. Laws No. 80, §1(4). “The goals of marketing programs,” the legislature said, “are often in conflict with the goals of the state.” §1(3). The text of §4631(d), associated legislative findings, and the record developed in the District Court establish that Vermont enacted its law for this end. The State has burdened a form of protected expression that it found too persuasive. At the same time, the State has left unburdened those speakers whose messages are in accord with its own views. This the State cannot do. The judgment of the Court of Appeals is affirmed. It is so ordered.

563.US.277

After this Court held that the Religious Freedom Restoration Act of 1993 was unconstitutional as applied to state and local governments because it exceeded Congress’ power under §5 of the Fourteenth Amendment, see City of Boerne v. Flores, 521 U. S. 507, Congress passed the Religious Land Use and Institutionalized Persons Act of 2000 (RLUIPA) pursuant to its Spending Clause and Commerce Clause authority. RLUIPA targets two areas of state and local action: land–use regulation, RLUIPA §2, 42 U. S. C. §2000cc, and restrictions on the religious exercise of institutionalized persons, RLUIPA §3, §2000cc–1. It also provides an express private cause of action for “appropriate relief against a government,” §2000cc–2(a), including, inter alia, States, their instrumentalities and officers, and persons acting under color of state law, §2000cc–5(4)(A). Petitioner Sossamon, a Texas prison inmate, sued respondents, the State and prison officials, seeking injunctive and monetary relief under RLUIPA for prison policies that prevented inmates from attending religious services while on cell restriction for disciplinary infractions and that barred use of the prison chapel for religious worship. Granting respondents summary judgment, the District Court held that sovereign immunity barred Sossamon’s claims for monetary relief. The Fifth Circuit affirmed, holding that the statutory phrase “appropriate relief against a government” did not unambiguously notify Texas that its acceptance of federal funds was conditioned on a waiver of sovereign immunity to claims for monetary relief. Held: States, in accepting federal funding, do not consent to waive their sovereign immunity to private suits for money damages under RLUIPA. Pp. 4–14. (a) Sovereign immunity principles enforce an important constitutional limitation on the power of the federal courts. See Pennhurst State School and Hospital v. Halderman, 465 U. S. 89, 98. This Court has consistently made clear that “federal jurisdiction over suits against unconsenting States ‘was not contemplated by the Constitution when establishing the judicial power of the United States.’ ” Seminole Tribe of Fla. v. Florida, 517 U. S. 44, 54. A State, however, may choose to waive its immunity. Clark v. Barnard, 108 U. S. 436, 447–448. The “ ‘test for determining whether [it has done so] is a stringent one.’ ” College Savings Bank v. Florida Prepaid Postsecondary Ed. Expense Bd., 527 U. S. 666, 675. The State’s consent to suit must be “unequivocally expressed” in the relevant statute’s text. Pennhurst, supra, at 99. A waiver “will be strictly construed, in terms of its scope, in favor of the sovereign.” Lane v. Peńa, 518 U. S. 187, 192. Pp. 4–6. (b) RLUIPA’s authorization of “appropriate relief against a government,” is not an unequivocal expression of state consent. Pp. 6–10. (1) “Appropriate relief” is open-ended and ambiguous about the relief it includes. “Appropriate” is inherently context-dependent. And the context here—where the defendant is a sovereign—suggests, if anything, that monetary damages are not “suitable” or “proper.” See Federal Maritime Comm’n v. South Carolina Ports Authority, 535 U. S. 743, 765. Further, where a statute is susceptible of multiple plausible interpretations, including one preserving immunity, this Court will not consider a State to have waived its sovereign immunity. Sossamon’s and Texas’ conflicting plausible arguments about whether immunity is preserved here demonstrate that “appropriate relief” in RLUIPA is not so free from ambiguity that the Court may conclude that the States, by receiving federal funds, have unequivocally expressed intent to waive their immunity. Pp. 6–9. (2) The Court’s use of the phrase “appropriate relief” in Franklin v. Gwinnett County Public Schools, 503 U. S. 60, and Barnes v. Gorman, 536 U. S. 181, does not compel a contrary conclusion. In those cases, where there was no express congressional intent to limit remedies available against municipal entities under an implied right of action, the Court presumed that compensatory damages were available. Franklin, supra, at 73. But that presumption is irrelevant to construing the scope of an express waiver of sovereign immunity, where the question is not whether Congress has given clear direction that it intends to exclude a damages remedy, but whether it has given clear direction that it intends to include a damages remedy. Pp. 9–10. (c) Sossamon mistakenly contends that Congress’ enactment of RLUIPA §3 pursuant to the Spending Clause put the States on notice that they would be liable for damages because Spending Clause legislation operates as a contract and damages are always available for a breach of contract. While acknowledging the contract-law analogy, this Court has been clear “not [to] imply … that suits under Spending Clause legislation are suits in contract, or that contract-law principles apply to all issues that they raise,” Barnes, supra, at 188, n. 2, or to rely on that analogy to expand liability beyond what would exist under nonspending statutes, much less to extend monetary liability against the States. Applying ordinary contract principles here would also make little sense because contracts with a sovereign are unique: They do not traditionally confer a right of action for damages to enforce compliance. More fundamentally, Sossamon’s implied-contract remedy cannot be squared with the rule that a sovereign immunity waiver must be expressly and unequivocally stated in the relevant statute’s text. Pp. 10–12. (d) Sossamon also errs in arguing that Texas was put on notice that it could be sued for damages under RLUIPA by §1003 of the Rehabilitation Act Amendments of 1986, which expressly waives state sovereign immunity for violations of “section 504 of the Rehabilitation Act of 1973, title IX of the Education Amendments of 1972, the Age Discrimination Act of 1975, title VI of the Civil Rights Act of 1964, or the provisions of any other Federal statute prohibiting discrimination by recipients of Federal financial assistance,” 42 U. S. C. §2000d–7. Even if such a residual clause could constitute an unequivocal textual waiver, RLUIPA §3—which prohibits “substantial burden[s]” on religious exercise—is not unequivocally a “statute prohibiting discrimination” within §1003’s meaning. All the statutory provisions enumerated in §1003 explicitly prohibit discrimination; a State might reasonably conclude that the residual clause, strictly construed, covers only provisions using the term “discrimination.” Pp. 12–14. 560 F. 3d 316, affirmed. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Ginsburg, and Alito, JJ., joined. Sotomayor, J., filed a dissenting opinion, in which Breyer, J., joined. Kagan, J., took no part in the consideration or decision of the case.

This case presents the question whether the States, by accepting federal funds, consent to waive their sovereign immunity to suits for money damages under the Religious Land Use and Institutionalized Persons Act of 2000 (RLUIPA), 114 Stat. 803, 42 U. S. C. §2000cc et seq. We hold that they do not. Sovereign immunity therefore bars this suit for damages against the State of Texas. I A RLUIPA is Congress’ second attempt to accord heightened statutory protection to religious exercise in the wake of this Court’s decision in Employment Division, Department of Human Resources of Oregon v. Smith, 494 U. S. 872 (1990). Congress first enacted the Religious Freedom Restoration Act of 1993 (RFRA), 107 Stat. 1488, 42 U. S. C. §2000bb et seq., with which it intended to “restore the compelling interest test as set forth in Sherbert v. Verner, 374 U. S. 398 (1963) and Wisconsin v. Yoder, 406 U. S. 205 (1972) … in all cases where free exercise of religion is substantially burdened.” §2000bb(b)(1). See generally Gonzales v. O Centro Espírita Beneficente Uniăo do Vegetal, 546 U. S. 418, 424 (2006). We held RFRA unconstitutional as applied to state and local governments because it exceeded Congress’ power under §5 of the Fourteenth Amendment. See City of Boerne v. Flores, 521 U. S. 507 (1997). Congress responded by enacting RLUIPA pursuant to its Spending Clause and Commerce Clause authority. RLUIPA borrows important elements from RFRA—which continues to apply to the Federal Government—but RLUIPA is less sweeping in scope. See Cutter v. Wilkinson, 544 U. S. 709, 715 (2005). It targets two areas of state and local action: land-use regulation, 42 U. S. C. §2000cc (RLUIPA §2), and restrictions on the religious exercise of institutionalized persons, §2000cc–1 (RLUIPA §3). Section 3 of RLUIPA provides that “[n]o government shall impose a substantial burden on the religious exercise” of an institutionalized person unless, as in RFRA, the government demonstrates that the burden “is in furtherance of a compelling governmental interest” and “is the least restrictive means of furthering” that interest. §2000cc–1(a); cf. §§2000bb–1(a), (b). As relevant here, §3 applies “in any case” in which “the substantial burden is imposed in a program or activity that receives Federal financial assistance.”[Footnote 1] §2000cc–1(b)(1). RLUIPA also includes an express private cause of action that is taken from RFRA: “A person may assert a violation of [RLUIPA] as a claim or defense in a judicial proceeding and obtain appropriate relief against a government.” §2000cc–2(a); cf. §2000bb–1(c). For purposes of this provision, “government” includes, inter alia, States, counties, municipalities, their instrumentalities and officers, and persons acting under color of state law. §2000cc–5(4)(A). B Petitioner Harvey Leroy Sossamon III is an inmate in the Robertson Unit of the Texas Department of Criminal Justice, Correctional Institutions Division. In 2006, Sossamon sued the State of Texas and various prison officials in their official capacities under RLUIPA’s private cause of action, seeking injunctive and monetary relief. Sossamon alleged that two prison policies violated RLUIPA: (1) a policy preventing inmates from attending religious services while on cell restriction for disciplinary infractions; and (2) a policy barring use of the prison chapel for religious worship. The District Court granted summary judgment in favor of respondents and held, as relevant here, that sovereign immunity barred Sossamon’s claims for monetary relief.[Footnote 2] See 713 F. Supp. 2d 657, 662–663 (WD Tex. 2007). The Court of Appeals for the Fifth Circuit affirmed. 560 F. 3d 316, 329 (2009). Acknowledging that Congress enacted RLUIPA pursuant to the Spending Clause, the court determined that Texas had not waived its sovereign immunity by accepting federal funds. The Court of Appeals strictly construed the text of RLUIPA’s cause of action in favor of the State and concluded that the statutory phrase “appropriate relief against a government” did not “unambiguously notif[y]” Texas that its acceptance of funds was conditioned on a waiver of immunity from claims for money damages. Id., at 330–331. We granted certiorari to resolve a division of authority among the courts of appeals on this question.[Footnote 3] 560 U. S. ___ (2010). II “Dual sovereignty is a defining feature of our Nation’s constitutional blueprint.” Federal Maritime Comm’n v. South Carolina Ports Authority, 535 U. S. 743, 751 (2002). Upon ratification of the Constitution, the States entered the Union “with their sovereignty intact.” Ibid. (internal quotation marks omitted). Immunity from private suits has long been considered “central to sovereign dignity.” Alden v. Maine, 527 U. S. 706, 715 (1999). As was widely understood at the time the Constitution was drafted: “It is inherent in the nature of sovereignty not to be amenable to the suit of an individual without its consent. This is the general sense, and the general practice of mankind; and the exemption, as one of the attributes of sovereignty, is now enjoyed by the government of every State in the Union.” The Federalist No. 81, p. 511 (Wright ed. 1961) (A. Hamilton). Indeed, when this Court threatened state immunity from private suits early in our Nation’s history, the people responded swiftly to reiterate that fundamental principle. See Hans v. Louisiana, 134 U. S. 1, 11 (1890) (discussing Chisholm v. Georgia, 2 Dall. 419 (1793), and the Eleventh Amendment). Sovereign immunity principles enforce an important constitutional limitation on the power of the federal courts. See Pennhurst State School and Hospital v. Halderman, 465 U. S. 89, 98 (1984). For over a century now, this Court has consistently made clear that “federal jurisdiction over suits against unconsenting States ‘was not contemplated by the Constitution when establishing the judicial power of the United States.’ ” Seminole Tribe of Fla. v. Florida, 517 U. S. 44, 54 (1996) (quoting Hans, supra, at 15); see Seminole Tribe, supra, at 54–55, n. 7 (collecting cases). A State, however, may choose to waive its immunity in federal court at its pleasure. Clark v. Barnard, 108 U. S. 436, 447–448 (1883). Accordingly, “our test for determining whether a State has waived its immunity from federal-court jurisdiction is a stringent one.” College Savings Bank v. Florida Prepaid Postsecondary Ed. Expense Bd., 527 U. S. 666, 675 (1999) (internal quotation marks omitted). A State’s consent to suit must be “unequivocally expressed” in the text of the relevant statute. Pennhurst State School and Hospital, supra, at 99; see Atascadero State Hospital v. Scanlon, 473 U. S. 234, 238, n. 1, 239–240 (1985). Only by requiring this “clear declaration” by the State can we be “certain that the State in fact consents to suit.” College Savings Bank, 527 U. S., at 680. Waiver may not be implied. Id., at 682. For these reasons, a waiver of sovereign immunity “will be strictly construed, in terms of its scope, in favor of the sovereign.” Lane v. Peńa, 518 U. S. 187, 192 (1996).[Footnote 4] So, for example, a State’s consent to suit in its own courts is not a waiver of its immunity from suit in federal court. College Savings Bank, supra, at 676. Similarly, a waiver of sovereign immunity to other types of relief does not waive immunity to damages: “[T]he waiver of sovereign immunity must extend unambiguously to such monetary claims.” Lane, supra, at 192; cf. United States v. Nordic Village, 503 U. S. 30, 34 (1992) (construing an ambiguous waiver of sovereign immunity to permit equitable but not monetary claims); Hoffman v. Connecticut Dept. of Income Maintenance, 492 U. S. 96, 101–102 (1989) (construing a statute to authorize injunctive relief but not “monetary recovery from the States” because intent to abrogate immunity to monetary recovery was not “ ‘unmistakably clear in the language of the statute’ ” (quoting Atascadero, supra, at 242)). III A RLUIPA’s authorization of “appropriate relief against a government,” §2000cc–2(a), is not the unequivocal expression of state consent that our precedents require. “Appropriate relief ” does not so clearly and unambiguously waive sovereign immunity to private suits for damages that we can “be certain that the State in fact consents” to such a suit. College Savings Bank, 527 U. S., at 680. 1 “Appropriate relief ” is open-ended and ambiguous about what types of relief it includes, as many lower courts have recognized. See, e.g., 560 F. 3d, at 330–331.[Footnote 5] Far from clearly identifying money damages, the word “appropriate” is inherently context-dependent. See Webster’s Third New International Dictionary 106 (1993) (defining “appropriate” as “specially suitable: fit, proper”). The context here—where the defendant is a sovereign—suggests, if anything, that monetary damages are not “suitable” or “proper.” See Federal Maritime Comm’n, 535 U. S., at 765 (“[S]tate sovereign immunity serves the important function of shielding state treasuries …”). Indeed, both the Court and dissent appeared to agree in West v. Gibson, 527 U. S. 212 (1999), that “appropriate” relief, by itself, does not unambiguously include damages against a sovereign. The question was whether the Equal Employment Opportunity Commission, which has authority to enforce Title VII of the Civil Rights Act against the Federal Government “through appropriate remedies,” could require the Federal Government to pay damages. 42 U. S. C. §2000e–16(b). The dissent argued that the phrase “appropriate remedies” did not authorize damages “in express and unequivocal terms.” Gibson, 527 U. S., at 226 (opinion of Kennedy, J.). The Court apparently did not disagree but reasoned that “appropriate remedies” had a flexible meaning that had expanded to include money damages after a related statute was amended to explicitly allow damages in actions under Title VII. See id., at 217–218. Further, where a statute is susceptible of multiple plausible interpretations, including one preserving immunity, we will not consider a State to have waived its sovereign immunity. See Dellmuth v. Muth, 491 U. S. 223, 232 (1989) (holding that “a permissible inference” is not the necessary “unequivocal declaration” that States were intended to be subject to damages actions); Nordic Village, supra, at 37 (holding that the existence of “plausible” interpretations that would not permit recovery “is enough to establish that a reading imposing monetary liability on the Government is not ‘unambiguous’ and therefore should not be adopted”). That is the case here. Sossamon argues that, because RLUIPA expressly limits the United States to “injunctive or declaratory relief ” to enforce the statute, the phrase “appropriate re-lief ” in the private cause of action necessarily must be broader. 42 U. S. C. §2000cc–2(f). Texas responds that, because the State has no immunity defense to a suit brought by the Federal Government, Congress needed to exclude damages affirmatively in that context but not in the context of private suits. Further, the private cause of action provides that a person may assert a violation of the statute “as a claim or defense.” §2000cc–2(a) (emphasis added). Because an injunction or declaratory judgment is not “appropriate relief ” for a successful defense, Texas explains, explicitly limiting the private cause of action to those forms of relief would make no sense. Sossamon also emphasizes that the statute requires that it be “construed in favor of a broad protection of religious exercise.” §2000cc–3(g). Texas responds that this provision is best read as addressing the substantive standards in the statute, not the scope of “appropriate relief.” Texas also highlights Congress’ choice of the word “relief,” which it argues primarily connotes equitable relief. See Black’s Law Dictionary 1295 (7th ed. 1999) (defining “relief ” as “[t]he redress or benefit, esp. equitable in nature … , that a party asks of a court”). These plausible arguments demonstrate that the phrase “appropriate relief ” in RLUIPA is not so free from ambiguity that we may conclude that the States, by receiving federal funds, have unequivocally expressed intent to waive their sovereign immunity to suits for damages. Strictly construing that phrase in favor of the sovereign—as we must, see Lane, 518 U. S., at 192—we conclude that it does not include suits for damages against a State. 2 The Court’s use of the phrase “appropriate relief ” in Franklin v. Gwinnett County Public Schools, 503 U. S. 60 (1992), and Barnes v. Gorman, 536 U. S. 181 (2002), does not compel a contrary conclusion. In those cases, the Court addressed what remedies are available against municipal entities under the implied right of action to enforce Title IX of the Education Amendments of 1972, §202 of the Americans with Disabilities Act of 1990, and §504 of the Rehabilitation Act of 1973. With no statutory text to interpret, the Court “presume[d] the availability of all appropriate remedies unless Congress ha[d] expressly indicated otherwise.” Franklin, 503 U. S., at 66. The Court described the presumption as “[t]he general rule” that “the federal courts have the power to award any appropriate relief in a cognizable cause of action brought pursuant to a federal statute.” Id., at 70–71 (emphasis added); see Barnes, supra, at 185 (quoting Franklin, supra, at 73). Finding no express congressional intent to limit the remedies available under the implied right of action, the Court held that compensatory damages were available. Franklin, supra, at 73. The presumption in Franklin and Barnes is irrelevant to construing the scope of an express waiver of sovereign immunity. See Lane, supra, at 196 (“[R]eliance on Franklin … is misplaced” in determining whether damages are available against the Federal Government). The question here is not whether Congress has given clear direction that it intends to exclude a damages remedy, see Franklin, supra, at 70–71, but whether Congress has given clear direction that it intends to include a damages remedy. The text must “establish unambiguously that the waiver extends to monetary claims.” Nordic Village, 503 U. S., at 34. In Franklin and Barnes, congressional silence had an entirely different implication than it does here. Whatever “appropriate relief ” might have meant in those cases does not translate to this context.[Footnote 6] B Sossamon contends that, because Congress enacted §3 of RLUIPA pursuant to the Spending Clause, the States were necessarily on notice that they would be liable for damages. He argues that Spending Clause legislation operates as a contract and damages are always available relief for a breach of contract, whether the contract explicitly includes a damages remedy or not. Relying on Barnes and Franklin, he asserts that all recipients of federal funding are “ ‘generally on notice that [they are] subject … to those remedies traditionally available in suits for breach of contract,’ ” including compensatory damages. Brief for Petitioner 27 (quoting Barnes, 536 U. S., at 187). We have acknowledged the contract-law analogy, but we have been clear “not [to] imply … that suits under Spending Clause legislation are suits in contract, or that contract-law principles apply to all issues that they raise.” Barnes, supra, at 189, n. 2. We have not relied on the Spending Clause contract analogy to expand liability beyond what would exist under non-spending statutes, much less to extend monetary liability against the States, as Sossamon would have us do. In fact, in Barnes and Franklin, the Court discussed the Spending Clause context only as a potential limitation on liability. See Barnes, supra, at 187–188; Franklin, supra, at 74–75. In any event, applying ordinary contract principles here would make little sense because contracts with a sovereign are unique. They do not traditionally confer a right of action for damages to enforce compliance: “ ‘The contracts between a Nation and an individual are only binding on the conscience of the sovereign and have no pretensions to compulsive force. They confer no right of action independent of the sovereign will.’ ” Lynch v. United States, 292 U. S. 571, 580–581 (1934) (quoting The Federalist, No. 81, at 511 (A. Hamilton)).[Footnote 7] More fundamentally, Sossamon’s implied-contract-remedies proposal cannot be squared with our longstanding rule that a waiver of sovereign immunity must be expressly and unequivocally stated in the text of the relevant statute. It would be bizarre to create an “unequivocal statement” rule and then find that every Spending Clause enactment, no matter what its text, satisfies that rule because it includes unexpressed, implied remedies against the States. The requirement of a clear statement in the text of the statute ensures that Congress has specifically considered state sovereign immunity and has intentionally legislated on the matter. Cf. Spector v. Norwegian Cruise Line Ltd., 545 U. S. 119, 139 (2005) (plurality opinion) (“[C]lear statement rules ensure Congress does not, by broad or general language, legislate on a sensitive topic inadvertently or without due deliberation”). Without such a clear statement from Congress and notice to the States, federal courts may not step in and abrogate state sovereign immunity.[Footnote 8] IV Sossamon also argues that §1003 of the Rehabilitation Act Amendments of 1986, 42 U. S. C. §2000d–7, independently put the State on notice that it could be sued for damages under RLUIPA. That provision expressly waives state sovereign immunity for violations of “section 504 of the Rehabilitation Act of 1973, title IX of the Education Amendments of 1972, the Age Discrimination Act of 1975, title VI of the Civil Rights Act of 1964, or the provisions of any other Federal statute prohibiting discrimination by recipients of Federal financial assistance.” §2000d–7(a)(1) (emphasis added). Section 1003 makes “remedies (including remedies both at law and in equity) … available for such a violation to the same extent as such remedies are available for such a violation in the suit against any public or private entity other than a State.” §2000d–7(a)(2). Sossamon contends that §3 of RLUIPA falls within the residual clause of §1003 and therefore §1003 waives Texas’ sovereign immunity to RLUIPA suits for damages. Even assuming that a residual clause like the one in §1003 could constitute an unequivocal textual waiver, §3 is not unequivocally a “statute prohibiting discrimination” within the meaning of §1003.[Footnote 9] The text of §3 does not prohibit “discrimination”; rather, it prohibits “substantial burden[s]” on religious exercise. This distinction is especially conspicuous in light of §2 of RLUIPA, in which Congress expressly prohibited “land use regulation[s] that discriminat[e] … on the basis of religion.” §2000cc(b)(2). A waiver of sovereign immunity must be “strictly construed, in terms of its scope, in favor of the sovereign.” Lane, 518 U. S., at 192. We cannot say that the residual clause clearly extends to §3; a State might reasonably conclude that the clause covers only provisions using the term “discrimination.” The statutory provisions specifically listed in §1003 confirm that §3 does not unequivocally come within the scope of the residual clause. “[G]eneral words,” such as the residual clause here, “are construed to embrace only objects similar in nature to those objects enumerated by the preceding specific words.” Washington State Dept. of Social and Health Servs. v. Guardianship Estate of Keffeler, 537 U. S. 371, 384 (2003) (internal quotation marks omitted); see also Jarecki v. G. D. Searle & Co., 367 U. S. 303, 307 (1961) (noting that this maxim “is often wisely applied where a word is capable of many meanings in order to avoid the giving of unintended breadth to the Acts of Congress”). Unlike §3, each of the statutes specifically enumerated in §1003 explicitly prohibits “discrimination.” See 29 U. S. C. §794(a); 20 U. S. C. §1681(a); 42 U. S. C. §§6101, 6102; 42 U. S. C. §2000d.[Footnote 10] * * * We conclude that States, in accepting federal funding, do not consent to waive their sovereign immunity to private suits for money damages under RLUIPA because no statute expressly and unequivocally includes such a waiver. The judgment of the United States Court of Appeals for the Fifth Circuit is affirmed. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 No party contends that the Commerce Clause permitted Congress to address the alleged burden on religious exercise at issue in this case. See 42 U. S. C. §2000cc–1(b)(2). Nor is Congress’ authority to enact RLUIPA under the Spending Clause challenged here. We therefore do not address those issues. Footnote 2 The District Court also denied injunctive relief. 713 F. Supp. 2d 657, 668 (WD Tex. 2007). The Court of Appeals subsequently held that Sossamon’s claim for injunctive relief with respect to the cell-restriction policy was moot because the State had abandoned that policy after Sossamon filed a prison grievance. 560 F. 3d 316, 326 (CA5 2009). The Court of Appeals reversed the District Court with respect to Sossamon’s chapel-use policy claim, id., at 331–335, although the Robertson Unit later amended that policy also and now permits inmates to attend scheduled worship services in the chapel subject to certain safety precautions. Footnote 3 Compare Madison v. Virginia, 474 F. 3d 118, 131 (CA4 2006); 560 F. 3d, at 331 (case below); Cardinal v. Metrish, 564 F. 3d 794, 801 (CA6 2009); Nelson v. Miller, 570 F. 3d 868, 885 (CA7 2009); Van Wyhe v. Reisch, 581 F. 3d 639, 655 (CA8 2009); and Holley v. California Dept. of Corrections, 599 F. 3d 1108, 1112 (CA9 2010), with Smith v. Allen, 502 F. 3d 1255, 1276, n. 12 (CA11 2007) (citing Benning v. Georgia, 391 F. 3d 1299, 1305–1306 (CA11 2004)). Footnote 4 Although Lane concerned the Federal Government, the strict construction principle, which flows logically from the requirement that consent be “unequivocally expressed,” applies to the sovereign immunity of the States as well. Cf. United States v. Nordic Village, Inc., 503 U. S. 30, 37 (1992) (equating the “unequivocal expression” principle from “the Eleventh Amendment context” with the principle applicable to federal sovereign immunity); College Savings Bank v. Florida Prepaid Postsecondary Ed. Expense Bd., 527 U. S. 666, 682 (1999) (noting the “clos[e] analogy” between federal and state sovereign immunity); Belknap v. Schild, 161 U. S. 10, 18 (1896) (“[A] State … is as exempt as the United States [is] from private suit”). Footnote 5 See also Holley, supra, at 1112; Nelson, supra, at 884; Van Wyhe, supra, at 654; Cardinal, supra, at 801; Madison, supra, at 131–132; cf. Webman v. Federal Bur. of Prisons, 441 F. 3d 1022, 1023 (CADC 2006) (interpreting the “appropriate relief ” provision of RFRA). Footnote 6 Nor can it be said that this Court’s use of the phrase “appropriate relief ” in Franklin and Barnes somehow put the States on notice that the same phrase in RLUIPA subjected them to suits for monetary relief. Those cases did not involve sovereign defendants, so the Court had no occasion to consider sovereign immunity. Liability against nonsovereigns could not put the States on notice that they would be liable in the same manner, absent an unequivocal textual waiver. Moreover, the same phrase in RFRA had been interpreted not to include damages relief against the Federal Government or the States and so could have signaled to the States that damages are not “appropriate relief ” under RLUIPA. See, e.g., Tinsley v. Pittari, 952 F. Supp. 384, 389 (ND Tex. 1996); Commack Self-Service Kosher Meats Inc. v. New York, 954 F. Supp. 65, 69 (EDNY 1997). Footnote 7 Of course, the Federal Government has, by statute, waived its sovereign immunity to damages for breach of contract in certain contexts. See, e.g., 28 U. S. C. §1491(a)(1). Footnote 8 The dissent finds our decision “difficult to understand,” post, at 6 (opinion of Sotomayor, J.), but it follows naturally from this Court’s precedents regarding waiver of sovereign immunity, which the dissent gives astonishingly short shrift. The dissent instead concerns itself primarily with “general remedies principles.” Post, at 1. The essence of sovereign immunity, however, is that remedies against the government differ from “general remedies principles” applicable to private litigants. See, e.g., Lane v. Peńa, 518 U. S. 187, 196 (1996) (calling it a “crucial point that, when it comes to an award of money damages, sovereign immunity places the … Government on an entirely different footing than private parties”). Footnote 9 Every Court of Appeals to consider the question has so held. See Holley, 599 F. 3d, at 1113–1114; Van Wyhe, 581 F. 3d, at 654–655; Madison, 474 F. 3d, at 132–133. Footnote 10 Sossamon argues that §3 resembles §504 of the Rehabilitation Act, one of the statutes listed in §1003, because both require special accommodations for particular people or activities. By Sossamon’s reasoning, every Spending Clause statute that arguably provides a benefit to a class of people or activities would become a federal statute “prohibiting discrimination,” thereby waiving sovereign immunity. Such an interpretation cannot be squared with the foundational rule that waiver of sovereign immunity must be unequivocally expressed and strictly construed.

562.US.411

While employed as an angiography technician by respondent Proctor Hospital, petitioner Staub was a member of the United States Army Reserve. Both his immediate supervisor (Mulally) and Mulally’s supervisor (Korenchuk) were hostile to his military obligations. Mulally gave Staub disciplinary warning which included a directive requiring Staub to report to her or Korenchuk when his cases were completed. After receiving a report from Korenchuk that Staub had violated the Corrective Action, Proctor’s vice president of human resources (Buck) reviewed Staub’s personnel file and decided to fire him. Staub filed a grievance, claiming that Mulally had fabricated the allegation underlying the warning out of hostility toward his military obligations, but Buck adhered to her decision. Staub sued Proctor under the Uniformed Services Employment and Reemployment Rights Act of 1994 (USERRA), which forbids an employer to deny “employment, reemployment, retention in employment, promotion, or any benefit of employment” based on a person’s “membership” in or “obligation to perform service in a uniformed service,” 38 U. S. C. §4311(a), and provides that liability is established “if the person’s membership … is a motivating factor in the employer’s action,” §4311(c). He contended not that Buck was motivated by hostility to his military obligations, but that Mulally and Korenchuk were, and that their actions influenced Buck’s decision. A jury found Proctor liable and awarded Staub damages, but the Seventh Circuit reversed, holding that Proctor was entitled to judgment as a matter of law because the decisionmaker had relied on more than Mulally’s and Korenchuk’s advice in making her decision. Held: 1. If a supervisor performs an act motivated by antimilitary animus that is intended by the supervisor to cause an adverse employment action, and if that act is a proximate cause of the ultimate employment action, then the employer is liable under USERRA. In construing the phrase “motivating factor in the employer’s action,” this Court starts from the premise that when Congress creates a federal tort it adopts the background of general tort law. See, e.g., Burlington N. & S. F. R. Co. v. United States, 556 U. S. ___, ___. Intentional torts such as the one here “generally require that the actor intend ‘the consequences’ of an act,’ not simply ‘the act itself.’ ” Kawaauhau v. Geiger, 523 U. S. 57, 61–62. However, Proctor errs in contending that an employer is not liable unless the de facto decisionmaker is motivated by discriminatory animus. So long as the earlier agent intended, for discriminatory reasons, that the adverse action occur, he has the scienter required for USERRA liability. Moreover, it is axiomatic under tort law that the decisionmaker’s exercise of judgment does not prevent the earlier agent’s action from being the proximate cause of the harm. See Hemi Group, LLC v. City of New York, 559 U. S. 1, ___. Nor can the ultimate decisionmaker’s judgment be deemed a superseding cause of the harm. See Exxon Co., U. S. A. v. Sofec, Inc., 517 U. S. 830, 837. Proctor’s approach would have an improbable consequence: If an employer isolates a personnel official from its supervisors, vests the decision to take adverse employment actions in that official, and asks that official to review the employee’s personnel file before taking the adverse action, then the employer will be effectively shielded from discriminatory acts and recommendations of supervisors that were designed and intended to produce the adverse action. Proctor also errs in arguing that a decisionmaker’s independent investigation, and rejection, of an employee’s discriminatory animus allegations should negate the effect of the prior discrimination. Pp. 4–10. 2. Applying this analysis here, the Seventh Circuit erred in holding that Proctor was entitled to judgment as a matter of law. Both Mulally and Korenchuk acted within the scope of their employment when they took the actions that allegedly caused Buck to fire Staub. There was also evidence that their actions were motivated by hostility toward Staub’s military obligations, and that those actions were causal factors underlying Buck’s decision. Finally, there was evidence that both Mulally and Korenchuk had the specific intent to cause Staub’s termination. The Seventh Circuit is to consider in the first instance whether the variance between the jury instruction given at trial and the rule adopted here was harmless error or should mandate a new trial. Pp. 11–12. 560 F. 3d 647, reversed and remanded. Scalia, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Ginsburg, Breyer, and Sotomayor, JJ., joined. Alito, J., filed an opinion concurring in the judgment, in which Thomas, J., joined. Kagan, J., took no part in the consideration or decision of the case.

We consider the circumstances under which an employer may be held liable for employment discrimination based on the discriminatory animus of an employee who influenced, but did not make, the ultimate employment decision. I Petitioner Vincent Staub worked as an angiography technician for respondent Proctor Hospital until 2004, when he was fired. Staub and Proctor hotly dispute the facts surrounding the firing, but because a jury found for Staub in his claim of employment discrimination against Proctor, we describe the facts viewed in the light most favorable to him. While employed by Proctor, Staub was a member of the United States Army Reserve, which required him to attend drill one weekend per month and to train full time for two to three weeks a year. Both Janice Mulally, Staub’s immediate supervisor, and Michael Korenchuk, Mulally’s supervisor, were hostile to Staub’s military obligations. Mulally scheduled Staub for additional shifts without notice so that he would “ ‘pa[y] back the department for everyone else having to bend over backwards to cover [his] schedule for the Reserves.’ ” 560 F. 3d 647, 652 (CA7 2009). She also informed Staub’s co-worker, Leslie Sweborg, that Staub’s “ ‘military duty had been a strain on th[e] department,’ ” and asked Sweborg to help her “ ‘get rid of him.’ ” Ibid. Korenchuk referred to Staub’s military obligations as “ ‘a b[u]nch of smoking and joking and [a] waste of taxpayers[’] money.’ ” Ibid. He was also aware that Mulally was “ ‘out to get’ ” Staub. Ibid. In January 2004, Mulally issued Staub a “Corrective Action” disciplinary warning for purportedly violating a company rule requiring him to stay in his work area whenever he was not working with a patient. The Corrective Action included a directive requiring Staub to report to Mulally or Korenchuk “ ‘when [he] ha[d] no patients and [the angio] cases [we]re complete[d].’ ” Id., at 653. According to Staub, Mulally’s justification for the Corrective Action was false for two reasons: First, the company rule invoked by Mulally did not exist; and second, even if it did, Staub did not violate it. On April 2, 2004, Angie Day, Staub’s co-worker, complained to Linda Buck, Proctor’s vice president of human resources, and Garrett McGowan, Proctor’s chief operating officer, about Staub’s frequent unavailability and abruptness. McGowan directed Korenchuk and Buck to create a plan that would solve Staub’s “ ‘availability’ problems.” Id., at 654. But three weeks later, before they had time to do so, Korenchuk informed Buck that Staub had left his desk without informing a supervisor, in violation of the January Corrective Action. Staub now contends this accusation was false: he had left Korenchuk a voice-mail notification that he was leaving his desk. Buck relied on Korenchuk’s accusation, however, and after reviewing Staub’s personnel file, she decided to fire him. The termination notice stated that Staub had ignored the directive issued in the January 2004 Corrective Action. Staub challenged his firing through Proctor’s grievance process, claiming that Mulally had fabricated the allegation underlying the Corrective Action out of hostility toward his military obligations. Buck did not follow up with Mulally about this claim. After discussing the matter with another personnel officer, Buck adhered to her decision. Staub sued Proctor under the Uniformed Services Employment and Reemployment Rights Act of 1994, 38 U. S. C. §4301 et seq., claiming that his discharge was motivated by hostility to his obligations as a military reservist. His contention was not that Buck had any such hostility but that Mulally and Korenchuk did, and that their actions influenced Buck’s ultimate employment decision. A jury found that Staub’s “military status was a motivating factor in [Proctor’s] decision to discharge him,” App. 68a, and awarded $57,640 in damages. The Seventh Circuit reversed, holding that Proctor was entitled to judgment as a matter of law. 560 F. 3d 647. The court observed that Staub had brought a “ ‘cat’s paw’ case,” meaning that he sought to hold his employer liable for the animus of a supervisor who was not charged with making the ultimate employment decision. Id., at 655–656.[Footnote 1] It explained that under Seventh Circuit precedent, a “cat’s paw” case could not succeed unless the nondecisionmaker exercised such “ ‘singular influence’ ” over the decisionmaker that the decision to terminate was the product of “blind reliance.” Id., at 659. It then noted that “Buck looked beyond what Mulally and Korenchuk said,” relying in part on her conversation with Day and her review of Staub’s personnel file. Ibid. The court “admit[ted] that Buck’s investigation could have been more robust,” since it “failed to pursue Staub’s theory that Mulally fabricated the write-up.” Ibid. But the court said that the “ ‘singular influence’ ” rule “does not require the decisionmaker to be a paragon of independence”: “It is enough that the decisionmaker is not wholly dependent on a single source of information and conducts her own investigation into the facts relevant to the decision.” Ibid. (internal quotation marks omitted). Because the undisputed evidence established that Buck was not wholly dependent on the advice of Korenchuk and Mulally, the court held that Proctor was entitled to judgment. Ibid. We granted certiorari. 559 U. S. ___ (2010). II The Uniformed Services Employment and Reemployment Rights Act (USERRA) provides in relevant part as follows: “A person who is a member of … or has an obligation to perform service in a uniformed service shall not be denied initial employment, reemployment, retention in employment, promotion, or any benefit of employment by an employer on the basis of that membership, … or obligation.” 38 U. S. C. §4311(a). It elaborates further: “An employer shall be considered to have engaged in actions prohibited … under subsection (a), if the person’s membership … is a motivating factor in the employer’s action, unless the employer can prove that the action would have been taken in the absence of such membership.” §4311(c). The statute is very similar to Title VII, which prohibits employment discrimination “because of … race, color, religion, sex, or national origin” and states that such discrimination is established when one of those factors “was a motivating factor for any employment practice, even though other factors also motivated the practice.” 42 U. S. C. §§2000e–2(a), (m). The central difficulty in this case is construing the phrase “motivating factor in the employer’s action.” When the company official who makes the decision to take an adverse employment action is personally acting out of hostility to the employee’s membership in or obligation to a uniformed service, a motivating factor obviously exists. The problem we confront arises when that official has no discriminatory animus but is influenced by previous company action that is the product of a like animus in someone else. In approaching this question, we start from the premise that when Congress creates a federal tort it adopts the background of general tort law. See Burlington N. & S. F. R. Co. v. United States, 556 U. S. ___, ___ (2009) (slip op., at 13–14); Safeco Ins. Co. of America v. Burr, 551 U. S. 47, 68–69 (2007); Burlington Industries, Inc. v. Ellerth, 524 U. S. 742, 764 (1998). Intentional torts such as this, “as distinguished from negligent or reckless torts, … generally require that the actor intend ‘the consequences’ of an act,’ not simply ‘the act itself.’ ” Kawaauhau v. Geiger, 523 U. S. 57, 61–62 (1998). Staub contends that the fact that an unfavorable entry on the plaintiff’s personnel record was caused to be put there, with discriminatory animus, by Mulally and Korenchuk, suffices to establish the tort, even if Mulally and Korenchuk did not intend to cause his dismissal. But discrimination was no part of Buck’s reason for the dismissal; and while Korenchuk and Mulally acted with discriminatory animus, the act they committed—the mere making of the reports—was not a denial of “initial employment, reemployment, retention in employment, promotion, or any benefit of employment,” as liability under USERRA requires. If dismissal was not the object of Mulally’s and Korenchuk’s reports, it may have been their result, or even their foreseeable consequence, but that is not enough to render Mulally or Korenchuk responsible. Here, however, Staub is seeking to hold liable not Mulally and Korenchuk, but their employer. Perhaps, therefore, the discriminatory motive of one of the employer’s agents (Mulally or Korenchuk) can be aggregated with the act of another agent (Buck) to impose liability on Proctor. Again we consult general principles of law, agency law, which form the background against which federal tort laws are enacted. See Meyer v. Holley, 537 U. S. 280, 285 (2003); Burlington, supra, at 754–755. Here, however, the answer is not so clear. The Restatement of Agency suggests that the malicious mental state of one agent cannot generally be combined with the harmful action of another agent to hold the principal liable for a tort that requires both. See Restatement (Second) Agency §275, Illustration 4 (1958). Some of the cases involving federal torts apply that rule. See United States v. Science Applications Int’l Corp., 626 F. 3d 1257, 1273–1276 (CADC 2010); Chaney v. Dreyfus Service Corp., 595 F. 3d 219, 241 (CA5 2010); United States v. Philip Morris USA Inc., 566 F. 3d 1095, 1122 (CADC 2009). But another case involving a federal tort, and one involving a federal crime, hold to the contrary. See United States ex rel. Harrison v. Westinghouse Savannah River Co., 352 F. 3d 908, 918–919 (CA4 2003); United States v. Bank of New England, N. A., 821 F. 2d 844, 856 (CA1 1987). Ultimately, we think it unnecessary in this case to decide what the background rule of agency law may be, since the former line of authority is suggested by the governing text, which requires that discrimination be “a motivating factor” in the adverse action. When a decision to fire is made with no unlawful animus on the part of the firing agent, but partly on the basis of a report prompted (unbeknownst to that agent) by discrimination, discrimination might perhaps be called a “factor” or a “causal factor” in the decision; but it seems to us a considerable stretch to call it “a motivating factor.” Proctor, on the other hand, contends that the employer is not liable unless the de facto decisionmaker (the technical decisionmaker or the agent for whom he is the “cat’s paw”) is motivated by discriminatory animus. This avoids the aggregation of animus and adverse action, but it seems to us not the only application of general tort law that can do so. Animus and responsibility for the adverse action can both be attributed to the earlier agent (here, Staub’s supervisors) if the adverse action is the intended consequence of that agent’s discriminatory conduct. So long as the agent intends, for discriminatory reasons, that the adverse action occur, he has the scienter required to be liable under USERRA. And it is axiomatic under tort law that the exercise of judgment by the decisionmaker does not prevent the earlier agent’s action (and hence the earlier agent’s discriminatory animus) from being the proximate cause of the harm. Proximate cause requires only “some direct relation between the injury asserted and the injurious conduct alleged,” and excludes only those “link[s] that are too remote, purely contingent, or indirect.” Hemi Group, LLC v. City of New York, 559 U. S. 1, ___ (2010) (slip op., at 9) (internal quotation marks omitted).[Footnote 2] We do not think that the ultimate decisionmaker’s exercise of judgment automatically renders the link to the supervisor’s bias “remote” or “purely contingent.” The decisionmaker’s exercise of judgment is also a proximate cause of the employment decision, but it is common for injuries to have multiple proximate causes. See Sosa v. Alvarez-Machain, 542 U. S. 692, 704 (2004). Nor can the ultimate decisionmaker’s judgment be deemed a superseding cause of the harm. A cause can be thought “superseding” only if it is a “cause of independent origin that was not foreseeable.” Exxon Co., U. S. A. v. Sofec, Inc., 517 U. S. 830, 837 (1996) (internal quotation marks omitted). Moreover, the approach urged upon us by Proctor gives an unlikely meaning to a provision designed to prevent employer discrimination. An employer’s authority to reward, punish, or dismiss is often allocated among multiple agents. The one who makes the ultimate decision does so on the basis of performance assessments by other supervisors. Proctor’s view would have the improbable consequence that if an employer isolates a personnel official from an employee’s supervisors, vests the decision to take adverse employment actions in that official, and asks that official to review the employee’s personnel file before taking the adverse action, then the employer will be effectively shielded from discriminatory acts and recommendations of supervisors that were designed and intended to produce the adverse action. That seems to us an implausible meaning of the text, and one that is not compelled by its words. Proctor suggests that even if the decisionmaker’s mere exercise of independent judgment does not suffice to negate the effect of the prior discrimination, at least the decisionmaker’s independent investigation (and rejection) of the employee’s allegations of discriminatory animus ought to do so. We decline to adopt such a hard-and-fast rule. As we have already acknowledged, the requirement that the biased supervisor’s action be a causal factor of the ultimate employment action incorporates the traditional tort-law concept of proximate cause. See, e.g., Anza v. Ideal Steel Supply Corp., 547 U. S. 451, 457–458 (2006); Sosa, supra, at 703. Thus, if the employer’s investigation results in an adverse action for reasons unrelated to the supervisor’s original biased action (by the terms of USERRA it is the employer’s burden to establish that), then the employer will not be liable. But the supervisor’s biased report may remain a causal factor if the independent investigation takes it into account without determining that the adverse action was, apart from the supervisor’s recommendation, entirely justified. We are aware of no principle in tort or agency law under which an employer’s mere conduct of an independent investigation has a claim-preclusive effect. Nor do we think the independent investigation somehow relieves the employer of “fault.” The employer is at fault because one of its agents committed an action based on discriminatory animus that was intended to cause, and did in fact cause, an adverse employment decision. Justice Alito claims that our failure to adopt a rule immunizing an employer who performs an independent investigation reflects a “stray[ing] from the statutory text.” Post, at 2 (opinion concurring in judgment). We do not understand this accusation. Since a supervisor is an agent of the employer, when he causes an adverse employment action the employer causes it; and when discrimination is a motivating factor in his doing so, it is a “motivating factor in the employer’s action,” precisely as the text requires. Justice Alito suggests that the employer should be held liable only when it “should be regarded as having delegated part of the decisionmaking power” to the biased supervisor. Ibid. But if the independent investigation relies on facts provided by the biased supervisor—as is necessary in any case of cat’s-paw liability—then the employer (either directly or through the ultimate decisionmaker) will have effectively delegated the factfinding portion of the investigation to the biased supervisor. Contrary to Justice Alito’s suggestion, the biased supervisor is not analogous to a witness at a bench trial. The mere witness is not an actor in the events that are the subject of the trial. The biased supervisor and the ultimate decisionmaker, however, acted as agents of the entity that the plaintiff seeks to hold liable; each of them possessed supervisory authority delegated by their employer and exercised it in the interest of their employer. In sum, we do not see how “fidelity to the statutory text,” ibid., requires the adoption of an independent-investigation defense that appears nowhere in the text. And we find both speculative and implausible Justice Alito’s prediction that our Nation’s employers will systematically disfavor members of the armed services in their hiring decisions to avoid the possibility of cat’s-paw liability, a policy that would violate USERRA in any event. We therefore hold that if a supervisor performs an act motivated by antimilitary animus that is intended by the supervisor to cause an adverse employment action,[Footnote 3] and if that act is a proximate cause of the ultimate employment action, then the employer is liable under USERRA. [Footnote 4] III Applying our analysis to the facts of this case, it is clear that the Seventh Circuit’s judgment must be reversed. Both Mulally and Korenchuk were acting within the scope of their employment when they took the actions that allegedly caused Buck to fire Staub. A “reprimand … for workplace failings” constitutes conduct within the scope of an agent’s employment. Faragher v. Boca Raton, 524 U. S. 775, 798–799 (1998). As the Seventh Circuit recognized, there was evidence that Mulally’s and Korenchuk’s actions were motivated by hostility toward Staub’s military obligations. There was also evidence that Mulally’s and Korenchuk’s actions were causal factors underlying Buck’s decision to fire Staub. Buck’s termination notice expressly stated that Staub was terminated because he had “ignored” the directive in the Corrective Action. Finally, there was evidence that both Mulally and Korenchuk had the specific intent to cause Staub to be terminated. Mulally stated she was trying to “ ‘get rid of ’ ” Staub, and Korenchuk was aware that Mulally was “ ‘out to get’ ” Staub. Moreover, Korenchuk informed Buck, Proctor’s personnel officer responsible for terminating employees, of Staub’s alleged noncompliance with Mulally’s Corrective Action, and Buck fired Staub immediately thereafter; a reasonable jury could infer that Korenchuk intended that Staub be fired. The Seventh Circuit therefore erred in holding that Proctor was entitled to judgment as a matter of law. It is less clear whether the jury’s verdict should be reinstated or whether Proctor is entitled to a new trial. The jury instruction did not hew precisely to the rule we adopt today; it required only that the jury find that “military status was a motivating factor in [Proctor’s] decision to discharge him.” App. 68a. Whether the variance between the instruction and our rule was harmless error or should mandate a new trial is a matter the Seventh Circuit may consider in the first instance. * * * The judgment of the Seventh Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 The term “cat’s paw” derives from a fable conceived by Aesop, put into verse by La Fontaine in 1679, and injected into United States employment discrimination law by Posner in 1990. See Shager v. Upjohn Co., 913 F. 2d 398, 405 (CA7). In the fable, a monkey induces a cat by flattery to extract roasting chestnuts from the fire. After the cat has done so, burning its paws in the process, the monkey makes off with the chestnuts and leaves the cat with nothing. A coda to the fable (relevant only marginally, if at all, to employment law) observes that the cat is similar to princes who, flattered by the king, perform services on the king’s behalf and receive no reward. Footnote 2 Under the traditional doctrine of proximate cause, a tortfeasor is sometimes, but not always, liable when he intends to cause an adverse action and a different adverse action results. See Restatement (Second) Torts §§435, 435B and Comment a (1963 and 1964). That issue is not presented in this case since the record contains no evidence that Mulally or Korenchuk intended any particular adverse action other than Staub’s termination. Footnote 3 Under traditional tort law, “ ‘intent’ … denote[s] that the actor desires to cause consequences of his act, or that he believes that the consequences are substantially certain to result from it.” Id., §8A. Footnote 4 Needless to say, the employer would be liable only when the supervisor acts within the scope of his employment, or when the supervisor acts outside the scope of his employment and liability would be imputed to the employer under traditional agency principles. See Burlington Industries, Inc. v. Ellerth, 524 U. S. 742, 758 (1998). We express no view as to whether the employer would be liable if a co-worker, rather than a supervisor, committed a discriminatory act that influenced the ultimate employment decision. We also observe that Staub took advantage of Proctor’s grievance process, and we express no view as to whether Proctor would have an affirmative defense if he did not. Cf. Pennsylvania State Police v. Suders, 542 U. S. 129, 148–149 (2004).

564.US.2010_10-179

Article III, §1, of the Constitution mandates that “[t]he judicial Power of the United States, shall be vested in one supreme Court, and in such inferior Courts as the Congress may from time to time ordain and establish,” and provides that the judges of those constitutional courts “shall hold their Offices during good Behaviour” and “receive for their Services[ ] a Compensation[ ] [that] shall not be diminished” during their tenure. The questions presented in this case are whether a bankruptcy court judge who did not enjoy such tenure and salary protections had the authority under 28 U. S. C. §157 and Article III to enter final judgment on a counterclaim filed by Vickie Lynn Marshall (whose estate is the petitioner) against Pierce Marshall (whose estate is the respondent) in Vickie’s bankruptcy proceedings. Vickie married J. Howard Marshall II, Pierce’s father, approximately a year before his death. Shortly before J. Howard died, Vickie filed a suit against Pierce in Texas state court, asserting that J. Howard meant to provide for Vickie through a trust, and Pierce tortiously interfered with that gift. After J. Howard died, Vickie filed for bankruptcy in federal court. Pierce filed a proof of claim in that proceeding, asserting that he should be able to recover damages from Vickie’s bankruptcy estate because Vickie had defamed him by inducing her lawyers to tell the press that he had engaged in fraud in controlling his father’s assets. Vickie responded by filing a counterclaim for tortious interference with the gift she expected from J. Howard. The Bankruptcy Court granted Vickie summary judgment on the defamation claim and eventually awarded her hundreds of millions of dollars in damages on her counterclaim. Pierce objected that the Bankruptcy Court lacked jurisdiction to enter a final judgment on that counterclaim because it was not a “core proceeding” as defined by 28 U. S. C. §157(b)(2)(C). As set forth in §157(a), Congress has divided bankruptcy proceedings into three categories: those that “aris[e] under title 11”; those that “aris[e] in” a Title 11 case; and those that are “related to a case under title 11.” District courts may refer all such proceedings to the bankruptcy judges of their district, and bankruptcy courts may enter final judgments in “all core proceedings arising under title 11, or arising in a case under title 11.” §§157(a), (b)(1). In non-core proceedings, by contrast, a bankruptcy judge may only “submit proposed findings of fact and conclusions of law to the district court.” §157(c)(1). Section 157(b)(2) lists 16 categories of core proceedings, including “counterclaims by the estate against persons filing claims against the estate.” §157(b)(2)(C). The Bankruptcy Court concluded that Vickie’s counterclaim was a core proceeding. The District Court reversed, reading this Court’s precedent in Northern Pipeline Constr. Co. v. Marathon Pipe Line Co., 458 U. S. 50, to “suggest[ ] that it would be unconstitutional to hold that any and all counterclaims are core.” The court held that Vickie’s counterclaim was not core because it was only somewhat related to Pierce’s claim, and it accordingly treated the Bankruptcy Court’s judgment as proposed, not final. Although the Texas state court had by that time conducted a jury trial on the merits of the parties’ dispute and entered a judgment in Pierce’s favor, the District Court went on to decide the matter itself, in Vickie’s favor. The Court of Appeals ultimately reversed. It held that the Bankruptcy Court lacked authority to enter final judgment on Vickie’s counterclaim because the claim was not “so closely related to [Pierce’s] proof of claim that the resolution of the counterclaim is necessary to resolve the allowance or disallowance of the claim itself.” Because that holding made the Texas probate court’s judgment the earliest final judgment on matters relevant to the case, the Court of Appeals held that the District Court should have given the state judgment preclusive effect. Held: Although the Bankruptcy Court had the statutory authority to enter judgment on Vickie’s counterclaim, it lacked the constitutional authority to do so. Pp. 6–38. 1. Section 157(b) authorized the Bankruptcy Court to enter final judgment on Vickie’s counterclaim. Pp. 8–16. (a) The Bankruptcy Court had the statutory authority to enter final judgment on Vickie’s counterclaim as a core proceeding under §157(b)(2)(C). Pierce argues that §157(b) authorizes bankruptcy courts to enter final judgments only in those proceedings that are both core and either arise in a Title 11 case or arise under Title 11 itself. But that reading necessarily assumes that there is a category of core proceedings that do not arise in a bankruptcy case or under bankruptcy law, and the structure of §157 makes clear that no such category exists. Pp. 8–11. (b) In the alternative, Pierce argues that the Bankruptcy Court lacked jurisdiction to resolve Vickie’s counterclaim because his defamation claim is a “personal injury tort” that the Bankruptcy Court lacked jurisdiction to hear under §157(b)(5). The Court agrees with Vickie that §157(b)(5) is not jurisdictional, and Pierce consented to the Bankruptcy Court’s resolution of the defamation claim. The Court is not inclined to interpret statutes as creating a jurisdictional bar when they are not framed as such. See generally Henderson v. Shinseki, 562 U. S. ___; Arbaugh v. Y & H Corp., 546 U. S. 500. Section 157(b)(5) does not have the hallmarks of a jurisdictional decree, and the statutory context belies Pierce’s claim that it is jurisdictional. Pierce consented to the Bankruptcy Court’s resolution of the defamation claim by repeatedly advising that court that he was happy to litigate his claim there. Pp. 12–16. 2. Although §157 allowed the Bankruptcy Court to enter final judgment on Vickie’s counterclaim, Article III of the Constitution did not. Pp. 16–38. (a) Article III is “an inseparable element of the constitutional system of checks and balances” that “both defines the power and protects the independence of the Judicial Branch.” Northern Pipeline, 458 U. S., at 58 (plurality opinion). Article III protects liberty not only through its role in implementing the separation of powers, but also by specifying the defining characteristics of Article III judges to protect the integrity of judicial decisionmaking. This is not the first time the Court has faced an Article III challenge to a bankruptcy court’s resolution of a debtor’s suit. In Northern Pipeline, the Court considered whether bankruptcy judges serving under the Bankruptcy Act of 1978—who also lacked the tenure and salary guarantees of Article III—could “constitutionally be vested with jurisdiction to decide [a] state-law contract claim” against an entity that was not otherwise part of the bankruptcy proceedings. Id., at 53, 87, n. 40 (plurality opinion). The plurality in Northern Pipeline recognized that there was a category of cases involving “public rights” that Congress could constitutionally assign to “legislative” courts for resolution. A full majority of the Court, while not agreeing on the scope of that exception, concluded that the doctrine did not encompass adjudication of the state law claim at issue in that case, and rejected the debtor’s argument that the Bankruptcy Court’s exercise of jurisdiction was constitutional because the bankruptcy judge was acting merely as an adjunct of the district court or court of appeals. Id., at 69–72; see id., at 90–91 (Rehnquist, J., concurring in judgment). After the decision in Northern Pipeline, Congress revised the statutes governing bankruptcy jurisdiction and bankruptcy judges. With respect to the “core” proceedings listed in §157(b)(2), however, the bankruptcy courts under the Bankruptcy Amendments and Federal Judgeship Act of 1984 exercise the same powers they wielded under the 1978 Act. The authority exercised by the newly constituted courts over a counterclaim such as Vickie’s exceeds the bounds of Article III. Pp. 16–22. (b) Vickie’s counterclaim does not fall within the public rights exception, however defined. The Court has long recognized that, in general, Congress may not “withdraw from judicial cognizance any matter which, from its nature, is the subject of a suit at the common law, or in equity, or admiralty.” Murray’s Lessee v. Hoboken Land & Improvement Co., 18 How. 272, 284. The Court has also recognized that “[a]t the same time there are matters, involving public rights, … which are susceptible of judicial determination, but which congress may or may not bring within the cognizance of the courts of the United States, as it may deem proper.” Ibid. Several previous decisions have contrasted cases within the reach of the public rights exception—those arising “between the Government and persons subject to its authority in connection with the performance of the constitutional functions of the executive or legislative departments”—and those that are instead matters “of private right, that is, of the liability of one individual to another under the law as defined.” Crowell v. Benson, 285 U. S. 22, 50, 51. Shortly after Northern Pipeline, the Court rejected the limitation of the public rights exception to actions involving the Government as a party. The Court has continued, however, to limit the exception to cases in which the claim at issue derives from a federal regulatory scheme, or in which resolution of the claim by an expert Government agency is deemed essential to a limited regulatory objective within the agency’s authority. In other words, it is still the case that what makes a right “public” rather than private is that the right is integrally related to particular Federal Government action. See United States v. Jicarilla Apache Nation, 564 U. S. ___, ___–___ (slip op., at 10–11); Thomas v. Union Carbide Agricultural Products Co., 473 U. S. 568, 584; Commodity Futures Trading Commission v. Schor, 478 U. S. 833, 844, 856. In Granfinanciera, S. A. v. Nordberg, 492 U. S. 33, the most recent case considering the public rights exception, the Court rejected a bankruptcy trustee’s argument that a fraudulent conveyance action filed on behalf of a bankruptcy estate against a noncreditor in a bankruptcy proceeding fell within the exception. Vickie’s counterclaim is similar. It is not a matter that can be pursued only by grace of the other branches, as in Murray’s Lessee, 18 How., at 284; it does not flow from a federal statutory scheme, as in Thomas, 473 U. S., at 584–585; and it is not “completely dependent upon” adjudication of a claim created by federal law, as in Schor, 478 U. S., at 856. This case involves the most prototypical exercise of judicial power: the entry of a final, binding judgment by a court with broad substantive jurisdiction, on a common law cause of action, when the action neither derives from nor depends upon any agency regulatory regime. If such an exercise of judicial power may nonetheless be taken from the Article III Judiciary simply by deeming it part of some amorphous “public right,” then Article III would be transformed from the guardian of individual liberty and separation of powers the Court has long recognized into mere wishful thinking. Pp. 22–29. (c) The fact that Pierce filed a proof of claim in the bankruptcy proceedings did not give the Bankruptcy Court the authority to adjudicate Vickie’s counterclaim. Initially, Pierce’s defamation claim does not affect the nature of Vickie’s tortious interference counterclaim as one at common law that simply attempts to augment the bankruptcy estate—the type of claim that, under Northern Pipeline and Granfinanciera, must be decided by an Article III court. The cases on which Vickie relies, Katchen v. Landy, 382 U. S. 323, and Langenkamp v. Culp, 498 U. S. 42 (per curiam), are inapposite. Katchen permitted a bankruptcy referee to exercise jurisdiction over a trustee’s voidable preference claim against a creditor only where there was no question that the referee was required to decide whether there had been a voidable preference in determining whether and to what extent to allow the creditor’s claim. The Katchen Court “intimate[d] no opinion concerning whether” the bankruptcy referee would have had “summary jurisdiction to adjudicate a demand by the [bankruptcy] trustee for affirmative relief, all of the substantial factual and legal bases for which ha[d] not been disposed of in passing on objections to the [creditor’s proof of ] claim.” 382 U. S., at 333, n. 9. The per curiam opinion in Langenkamp is to the same effect. In this case, by contrast, the Bankruptcy Court—in order to resolve Vickie’s counterclaim—was required to and did make several factual and legal determinations that were not “disposed of in passing on objections” to Pierce’s proof of claim. In both Katchen and Langenkamp, moreover, the trustee bringing the preference action was asserting a right of recovery created by federal bankruptcy law. Vickie’s claim is instead a state tort action that exists without regard to any bankruptcy proceeding. Pp. 29–34. (d) The bankruptcy courts under the 1984 Act are not “adjuncts” of the district courts. The new bankruptcy courts, like the courts considered in Northern Pipeline, do not “ma[k]e only specialized, narrowly confined factual determinations regarding a particularized area of law” or engage in “statutorily channeled factfinding functions.” 458 U. S., at 85 (plurality opinion). Whereas the adjunct agency in Crowell v. Benson “possessed only a limited power to issue compensation orders … [that] could be enforced only by order of the district court,” ibid., a bankruptcy court resolving a counterclaim under §157(b)(2)(C) has the power to enter “appropriate orders and judgments”—including final judgments—subject to review only if a party chooses to appeal, see §§157(b)(1), 158(a)–(b). Such a court is an adjunct of no one. Pp. 34–36. (e) Finally, Vickie and her amici predict that restrictions on a bankruptcy court’s ability to hear and finally resolve compulsory counterclaims will create significant delays and impose additional costs on the bankruptcy process. It goes without saying that “the fact that a given law or procedure is efficient, convenient, and useful in facilitating functions of government, standing alone, will not save it if it is contrary to the Constitution.” INS v. Chadha, 462 U. S. 919, 944. In addition, the Court is not convinced that the practical consequences of such limitations are as significant as Vickie suggests. The framework Congress adopted in the 1984 Act already contemplates that certain state law matters in bankruptcy cases will be resolved by state courts and district courts, see §§157(c), 1334(c), and the Court does not think the removal of counterclaims such as Vickie’s from core bankruptcy jurisdiction meaningfully changes the division of labor in the statute. Pp. 36–38. 600 F. 3d 1037, affirmed. Roberts, C. J., delivered the opinion of the Court, in which Scalia, Kennedy, Thomas, and Alito, JJ., joined. Scalia, J., filed a concurring opinion. Breyer, J., filed a dissenting opinion, in which Ginsburg, Sotomayor, and Kagan, JJ., joined.

This “suit has, in course of time, become so complicated, that … no two … lawyers can talk about it for five minutes, without coming to a total disagreement as to all the premises. Innumerable children have been born into the cause: innumerable young people have married into it;” and, sadly, the original parties “have died out of it.” A “long procession of [judges] has come in and gone out” dur-ing that time, and still the suit “drags its weary length before the Court.” Those words were not written about this case, see C. Dickens, Bleak House, in 1 Works of Charles Dickens 4–5 (1891), but they could have been. This is the second time we have had occasion to weigh in on this long-running dispute between Vickie Lynn Marshall and E. Pierce Marshall over the fortune of J. Howard Marshall II, a man believed to have been one of the richest people in Texas. The Marshalls’ litigation has worked its way through state and federal courts in Louisiana, Texas, and California, and two of those courts—a Texas state probate court and the Bankruptcy Court for the Central District of California—have reached contrary decisions on its mer-its. The Court of Appeals below held that the Texas state decision controlled, after concluding that the Bankruptcy Court lacked the authority to enter final judgment on a counterclaim that Vickie brought against Pierce in her bankruptcy proceeding.[Footnote 1] To determine whether the Court of Appeals was correct in that regard, we must resolve two issues: (1) whether the Bankruptcy Court had the statutory authority under 28 U. S. C. §157(b) to issue a final judgment on Vickie’s counterclaim; and (2) if so, whether conferring that authority on the Bankruptcy Court is constitutional. Although the history of this litigation is complicated, its resolution ultimately turns on very basic principles. Article III, §1, of the Constitution commands that “[t]he judicial Power of the United States, shall be vested in one supreme Court, and in such inferior Courts as the Congress may from time to time ordain and establish.” That Article further provides that the judges of those courts shall hold their offices during good behavior, without diminution of salary. Ibid. Those requirements of Article III were not honored here. The Bankruptcy Court in this case exercised the judicial power of the United States by entering final judgment on a common law tort claim, even though the judges of such courts enjoy neither tenure during good behavior nor salary protection. We conclude that, although the Bankruptcy Court had the statutory authority to enter judgment on Vickie’s counterclaim, it lacked the constitutional authority to do so. I Because we have already recounted the facts and procedural history of this case in detail, see Marshall v. Marshall, 547 U. S. 293, 300–305 (2006), we do not repeat them in full here. Of current relevance are two claims Vickie filed in an attempt to secure half of J. Howard’s fortune. Known to the public as Anna Nicole Smith, Vickie was J. Howard’s third wife and married him about a year before his death. Id., at 300; see In re Marshall, 392 F. 3d 1118, 1122 (CA9 2004). Although J. Howard bestowed on Vickie many monetary and other gifts during their courtship and marriage, he did not include her in his will. 547 U. S., at 300. Before J. Howard passed away, Vickie filed suit in Texas state probate court, asserting that Pierce—J. Howard’s younger son—fraudulently induced J. Howard to sign a living trust that did not include her, even though J. Howard meant to give her half his property. Pierce denied any fraudulent activity and defended the validity of J. Howard’s trust and, eventually, his will. 392 F. 3d, at 1122–1123, 1125. After J. Howard’s death, Vickie filed a petition for bankruptcy in the Central District of California. Pierce filed a complaint in that bankruptcy proceeding, contending that Vickie had defamed him by inducing her lawyers to tell members of the press that he had engaged in fraud to gain control of his father’s assets. 547 U. S., at 300–301; In re Marshall, 600 F. 3d 1037, 1043–1044 (CA9 2010). The complaint sought a declaration that Pierce’s defamation claim was not dischargeable in the bankruptcy proceedings. Ibid.; see 11 U. S. C. §523(a). Pierce subsequently filed a proof of claim for the defamation action, meaning that he sought to recover damages for it from Vickie’s bankruptcy estate. See §501(a). Vickie responded to Pierce’s initial complaint by asserting truth as a defense to the alleged defamation and by filing a counterclaim for tortious interference with the gift she expected from J. Howard. As she had in state court, Vickie alleged that Pierce had wrongfully prevented J. Howard from taking the legal steps necessary to provide her with half his property. 547 U. S., at 301. On November 5, 1999, the Bankruptcy Court issued an order granting Vickie summary judgment on Pierce’s claim for defamation. On September 27, 2000, after a bench trial, the Bankruptcy Court issued a judgment on Vickie’s counterclaim in her favor. The court later awarded Vickie over $400 million in compensatory damages and $25 million in punitive damages. 600 F. 3d, at 1045; see 253 B. R. 550, 561–562 (Bkrtcy. Ct. CD Cal. 2000); 257 B. R. 35, 39–40 (Bkrtcy. Ct. CD Cal. 2000). In post-trial proceedings, Pierce argued that the Bankruptcy Court lacked jurisdiction over Vickie’s counterclaim. In particular, Pierce renewed a claim he had made earlier in the litigation, asserting that the Bankruptcy Court’s authority over the counterclaim was limited because Vickie’s counterclaim was not a “core proceeding” under 28 U. S. C. §157(b)(2)(C). See 257 B. R., at 39. As explained below, bankruptcy courts may hear and en- ter final judgments in “core proceedings” in a bankruptcy case. In non-core proceedings, the bankruptcy courts instead submit proposed findings of fact and conclusions of law to the district court, for that court’s review and issuance of final judgment. The Bankruptcy Court in this case concluded that Vickie’s counterclaim was “a core proceeding” under §157(b)(2)(C), and the court therefore had the “power to enter judgment” on the counterclaim under §157(b)(1). Id., at 40. The District Court disagreed. It recognized that “Vickie’s counterclaim for tortious interference falls within the literal language” of the statute designating certain proceedings as “core,” see §157(b)(2)(C), but understood this Court’s precedent to “suggest[ ] that it would be unconstitutional to hold that any and all counterclaims are core.” 264 B. R. 609, 629–630 (CD Cal. 2001) (citing Northern Pipeline Constr. Co. v. Marathon Pipe Line Co., 458 U. S. 50, 79, n. 31 (1982) (plurality opinion)). The District Court accordingly concluded that a “counterclaim should not be characterized as core” when it “is only somewhat related to the claim against which it is asserted, and when the unique characteristics and context of the counterclaim place it outside of the normal type of set-off or other counterclaims that customarily arise.” 264 B. R., at 632. Because the District Court concluded that Vickie’s counterclaim was not core, the court determined that it was required to treat the Bankruptcy Court’s judgment as “proposed[,] rather than final,” and engage in an “independent review” of the record. Id., at 633; see 28 U. S. C. §157(c)(1). Although the Texas state court had by that time conducted a jury trial on the merits of the parties’ dispute and entered a judgment in Pierce’s favor, the District Court declined to give that judgment preclusive effect and went on to decide the matter itself. 271 B. R. 858, 862–867 (CD Cal. 2001); see 275 B. R. 5, 56–58 (CD Cal. 2002). Like the Bankruptcy Court, the District Court found that Pierce had tortiously interfered with Vickie’s expectancy of a gift from J. Howard. The District Court awarded Vickie compensatory and punitive damages, each in the amount of $44,292,767.33. Id., at 58. The Court of Appeals reversed the District Court on a different ground, 392 F. 3d, at 1137, and we—in the first visit of the case to this Court—reversed the Court of Appeals on that issue. 547 U. S., at 314–315. On remand from this Court, the Court of Appeals held that §157 mandated “a two-step approach” under which a bankruptcy judge may issue a final judgment in a proceeding only if the matter both “meets Congress’ definition of a core proceeding and arises under or arises in title 11,” the Bankruptcy Code. 600 F. 3d, at 1055. The court also reasoned that allowing a bankruptcy judge to enter final judgments on all counterclaims raised in bankruptcy proceedings “would certainly run afoul” of this Court’s decision in Northern Pipeline. 600 F. 3d, at 1057. With those concerns in mind, the court concluded that “a counterclaim under §157(b)(2)(C) is properly a ‘core’ proceeding ‘arising in a case under’ the [Bankruptcy] Code only if the counterclaim is so closely related to [a creditor’s] proof of claim that the resolution of the counterclaim is necessary to resolve the allowance or disallowance of the claim itself.” Id., at 1058 (internal quotation marks omitted; second brackets added). The court ruled that Vickie’s counterclaim did not meet that test. Id., at 1059. That holding made “the Texas probate court’s judgment … the earliest final judgment entered on matters relevant to this proceeding,” and therefore the Court of Appeals concluded that the District Court should have “afford[ed] preclusive effect” to the Texas “court’s determination of relevant legal and factual issues.” Id., at 1064–1065.[Footnote 2] We again granted certiorari. 561 U. S. __ (2010). II A With certain exceptions not relevant here, the district courts of the United States have “original and exclusive jurisdiction of all cases under title 11.” 28 U. S. C. §1334(a). Congress has divided bankruptcy proceedings into three categories: those that “aris[e] under title 11”; those that “aris[e] in” a Title 11 case; and those that are “related to a case under title 11.” §157(a). District courts may refer any or all such proceedings to the bankruptcy judges of their district, ibid., which is how the Bankruptcy Court in this case came to preside over Vickie’s bankruptcy proceedings. District courts also may withdraw a case or proceeding referred to the bankruptcy court “for cause shown.” §157(d). Since Congress enacted the Bankruptcy Amendments and Federal Judgeship Act of 1984 (the 1984 Act), bankruptcy judges for each district have been appointed to 14-year terms by the courts of appeals for the circuits in which their district is located. §152(a)(1). The manner in which a bankruptcy judge may act on a referred matter depends on the type of proceeding involved. Bankruptcy judges may hear and enter final judgments in “all core proceedings arising under title 11, or arising in a case under title 11.” §157(b)(1). “Core proceedings include, but are not limited to” 16 different types of matters, including “counterclaims by [a debtor’s] estate against persons filing claims against the estate.” §157(b)(2)(C).[Footnote 3] Parties may appeal final judgments of a bankruptcy court in core proceedings to the district court, which reviews them under traditional appellate standards. See §158(a); Fed. Rule Bkrtcy. Proc. 8013. When a bankruptcy judge determines that a referred “pro-ceeding … is not a core proceeding but … is other- wise related to a case under title 11,” the judge may only “submit proposed findings of fact and conclusions of law to the district court.” §157(c)(1). It is the district court that enters final judgment in such cases after reviewing de novo any matter to which a party objects. Ibid. B Vickie’s counterclaim against Pierce for tortious interference is a “core proceeding” under the plain text of §157(b)(2)(C). That provision specifies that core proceedings include “counterclaims by the estate against persons filing claims against the estate.” In past cases, we have suggested that a proceeding’s “core” status alone authorizes a bankruptcy judge, as a statutory matter, to enter final judgment in the proceeding. See, e.g., Granfinanciera, S. A. v. Nordberg, 492 U. S. 33, 50 (1989) (explaining that Congress had designated certain actions as “ ‘core proceedings,’ which bankruptcy judges may adjudicate and in which they may issue final judgments, if a district court has referred the matter to them” (citations omitted)). We have not directly addressed the question, however, and Pierce argues that a bankruptcy judge may enter final judgment on a core proceeding only if that proceeding also “aris[es] in” a Title 11 case or “aris[es] under” Title 11 itself. Brief for Respondent 51 (internal quotation marks omitted). Section 157(b)(1) authorizes bankruptcy courts to “hear and determine all cases under title 11 and all core proceedings arising under title 11, or arising in a case under title 11.” As written, §157(b)(1) is ambiguous. The “arising under” and “arising in” phrases might, as Pierce suggests, be read as referring to a limited category of those core proceedings that are addressed in that section. On the other hand, the phrases might be read as simply describing what core proceedings are: matters arising under Title 11 or in a Title 11 case. In this case the structure and context of §157 contradict Pierce’s interpretation of §157(b)(1). As an initial matter, Pierce’s reading of the statute necessarily assumes that there is a category of core proceedings that neither arise under Title 11 nor arise in a Title 11 case. The manner in which the statute delineates the bankruptcy courts’ authority, however, makes plain that no such category exists. Section 157(b)(1) authorizes bankruptcy judges to enter final judgments in “core proceedings arising under title 11, or arising in a case under title 11.” Section 157(c)(1) instructs bankruptcy judges to instead submit proposed findings in “a proceeding that is not a core proceeding but that is otherwise related to a case under title 11.” Nowhere does §157 specify what bankruptcy courts are to do with respect to the category of matters that Pierce posits—core proceedings that do not arise under Title 11 or in a Title 11 case. To the contrary, §157(b)(3) only instructs a bankruptcy judge to “determine, on the judge’s own motion or on timely motion of a party, whether a proceeding is a core proceeding under this subsection or is a proceeding that is otherwise related to a case under title 11.” Two options. The statute does not suggest that any other distinctions need be made. Under our reading of the statute, core proceedings are those that arise in a bankruptcy case or under Title 11. The detailed list of core proceedings in §157(b)(2) pro- vides courts with ready examples of such matters. Pierce’s reading of §157, in contrast, supposes that some core pro-ceedings will arise in a Title 11 case or under Title 11 and some will not. Under that reading, the statute provides no guidance on how to tell which are which. We think it significant that Congress failed to provide any framework for identifying or adjudicating the asserted category of core but not “arising” proceedings, given the otherwise detailed provisions governing bankruptcy court authority. It is hard to believe that Congress would go to the trouble of cataloging 16 different types of proceedings that should receive “core” treatment, but then fail to specify how to determine whether those matters arise under Title 11 or in a bankruptcy case if—as Pierce asserts—the latter inquiry is determinative of the bankruptcy court’s authority. Pierce argues that we should treat core matters that arise neither under Title 11 nor in a Title 11 case as proceedings “related to” a Title 11 case. Brief for Respondent 60 (internal quotation marks omitted). We think that a contradiction in terms. It does not make sense to describe a “core” bankruptcy proceeding as merely “related to” the bankruptcy case; oxymoron is not a typical feature of congressional drafting. See Northern Pipeline, 458 U. S., at 71 (plurality opinion) (distinguishing “the restructuring of debtor-creditor relations, which is at the core of the federal bankruptcy power, … from the adjudication of state-created private rights”); Collier on Bankruptcy ¶3.02[2], p. 3–26, n. 5 (16th ed. 2010) (“The terms ‘non-core’ and ‘related’ are synonymous”); see also id., at 3–26, (“The phraseology of section 157 leads to the conclusion that there is no such thing as a core matter that is ‘related to’ a case under title 11. Core proceedings are, at most, those that arise in title 11 cases or arise under title 11” (footnote omitted)). And, as already discussed, the statute simply does not provide for a proceeding that is simultaneously core and yet only related to the bankruptcy case. See §157(c)(1) (providing only for “a proceeding that is not a core proceeding but that is otherwise related to a case under title 11”). As we explain in Part III, we agree with Pierce that designating all counterclaims as “core” proceedings raises serious constitutional concerns. Pierce is also correct that we will, where possible, construe federal statutes so as “to avoid serious doubt of their constitutionality.” Commod-ity Futures Trading Comm’n v. Schor, 478 U. S. 833, 841 (1986) (internal quotation marks omitted). But that “canon of construction does not give [us] the prerogative to ignore the legislative will in order to avoid constitutional adjudication.” Ibid. In this case, we do not think the plain text of §157(b)(2)(C) leaves any room for the canon of avoidance. We would have to “rewrit[e]” the statute, not interpret it, to bypass the constitutional issue §157(b)(2)(C) presents. Id., at 841 (internal quotation marks omitted). That we may not do. We agree with Vickie that §157(b)(2)(C) permits the bankruptcy court to enter a final judgment on her tortious interference counterclaim. C Pierce argues, as another alternative to reaching the constitutional question, that the Bankruptcy Court lacked jurisdiction to enter final judgment on his defamation claim. Section 157(b)(5) provides that “[t]he district court shall order that personal injury tort and wrongful death claims shall be tried in the district court in which the bankruptcy case is pending, or in the district court in the district in which the claim arose.” Pierce asserts that his defamation claim is a “personal injury tort,” that the Bankruptcy Court therefore had no jurisdiction over that claim, and that the court therefore necessarily lacked jurisdiction over Vickie’s counterclaim as well. Brief for Respondent 65–66. Vickie objects to Pierce’s statutory analysis across the board. To begin, Vickie contends that §157(b)(5) does not address subject matter jurisdiction at all, but simply specifies the venue in which “personal injury tort and wrongful death claims” should be tried. See Reply Brief for Petitioner 16–17, 19; see also Tr. of Oral Arg. 23 (Deputy Solicitor General) (Section “157(b)(5) is in [the United States’] view not jurisdictional”). Given the limited scope of that provision, Vickie argues, a party may waive or for-feit any objections under §157(b)(5), in the same way that a party may waive or forfeit an objection to the bankruptcy court finally resolving a non-core claim. Reply Brief for Petitioner 17–20; see §157(c)(2) (authorizing the district court, “with the consent of all the parties to the proceeding,” to refer a “related to” matter to the bankruptcy court for final judgment). Vickie asserts that in this case Pierce consented to the Bankruptcy Court’s adjudication of his defamation claim, and forfeited any argument to the contrary, by failing to seek withdrawal of the claim until he had litigated it before the Bankruptcy Court for 27 months. Id., at 20–23. On the merits, Vickie contends that the statutory phrase “personal injury tort and wrongful death claims” does not include non-physical torts such as defamation. Id., at 25–26. We need not determine what constitutes a “personal injury tort” in this case because we agree with Vickie that §157(b)(5) is not jurisdictional, and that Pierce consented to the Bankruptcy Court’s resolution of his defamation claim.[Footnote 4] Because “[b]randing a rule as going to a court’s subject-matter jurisdiction alters the normal operation of our adversarial system,” Henderson v. Shinseki, 562 U. S. ___, ___–___ (2011) (slip op., at 4–5), we are not inclined to interpret statutes as creating a jurisdictional bar when they are not framed as such. See generally Arbaugh v. Y & H Corp., 546 U. S. 500, 516 (2006) (“when Congress does not rank a statutory limitation on coverage as jurisdictional, courts should treat the restriction as nonjurisdictional in character”). Section 157(b)(5) does not have the hallmarks of a jurisdictional decree. To begin, the statutory text does not refer to either district court or bankruptcy court “jurisdiction,” instead addressing only where personal injury tort claims “shall be tried.” The statutory context also belies Pierce’s jurisdictional claim. Section 157 allocates the authority to enter final judgment between the bankruptcy court and the district court. See §§157(b)(1), (c)(1). That allocation does not implicate questions of subject matter jurisdiction. See §157(c)(2) (parties may consent to entry of final judgment by bankruptcy judge in non-core case). By the same token, §157(b)(5) simply specifies where a particular category of cases should be tried. Pierce does not explain why that statutory limitation may not be similarly waived. We agree with Vickie that Pierce not only could but did consent to the Bankruptcy Court’s resolution of his defamation claim. Before the Bankruptcy Court, Vickie objected to Pierce’s proof of claim for defamation, arguing that Pierce’s claim was unenforceable and that Pierce should not receive any amount for it. See 29 Court of Appeals Supplemental Excerpts of Record 6031, 6035 (hereinafter Supplemental Record). Vickie also noted that the Bankruptcy Court could defer ruling on her objection, given the litigation posture of Pierce’s claim before the Bankruptcy Court. See id., at 6031. Vickie’s filing prompted Pierce to advise the Bankruptcy Court that “[a]ll parties are in agreement that the amount of the contingent Proof of Claim filed by [Pierce] shall be determined by the adversary proceedings” that had been commenced in the Bankruptcy Court. 31 Supplemental Record 6801. Pierce asserted that Vickie’s objection should be overruled or, alternatively, that any ruling on the objection “should be continued until the resolution of the pending adversary proceeding litigation.” Ibid. Pierce identifies no point in the record where he argued to the Bankruptcy Court that it lacked the authority to adjudicate his proof of claim because the claim sought recompense for a personal injury tort. Indeed, Pierce apparently did not object to any court that §157(b)(5) prohibited the Bankruptcy Court from resolving his defamation claim until over two years—and several adverse discovery rulings—after he filed that claim in June 1996. The first filing Pierce cites as rais- ing that objection is his September 22, 1998 motion to the District Court to withdraw the reference of the case to the Bankruptcy Court. See Brief for Respondent 26–27. The District Court did initially withdraw the reference as requested, but it then returned the proceeding to the Bankruptcy Court, observing that Pierce “implicated the jurisdiction of that bankruptcy court. He chose to be a party to that litigation.” App. 129. Although Pierce had objected in July 1996 to the Bankruptcy Court’s exercise of jurisdiction over Vickie’s counterclaim, he advised the court at that time that he was “happy to litigate [his] claim” there. 29 Supplemental Record 6101. Counsel stated that even though Pierce thought it was “probably cheaper for th[e] estate if [Pierce’s claim] were sent back or joined back with the State Court litigation,” Pierce “did choose” the Bankruptcy Court forum and “would be more than pleased to do it [t]here.” Id., at 6101–6102; see also App. to Pet. for Cert. 266, n. 17 (District Court referring to these statements). Given Pierce’s course of conduct before the Bankruptcy Court, we conclude that he consented to that court’s resolution of his defamation claim (and forfeited any argument to the contrary). We have recognized “the value of waiver and forfeiture rules” in “complex” cases, Exxon Shipping Co. v. Baker, 554 U. S. 471, 487–488, n. 6 (2008), and this case is no exception. In such cases, as here, the consequences of “a litigant … ‘sandbagging’ the court—remaining silent about his objection and belatedly raising the error only if the case does not conclude in his favor,” Puckett v. United States, 556 U. S. ___, ___ (2009) (slip op., at 5) (some internal quotation marks omitted)—can be particularly severe. If Pierce believed that the Bankruptcy Court lacked the authority to decide his claim for defamation, then he should have said so—and said so promptly. See United States v. Olano, 507 U. S. 725, 731 (1993) (“ ‘No procedural principle is more familiar to this Court than that a constitutional right,’ or a right of any other sort, ‘may be forfeited … by the failure to make timely assertion of the right before a tribunal having jurisdiction to determine it’ ” (quoting Yakus v. United States, 321 U. S. 414, 444 (1944))). Instead, Pierce repeatedly stated to the Bankruptcy Court that he was happy to litigate there. We will not consider his claim to the contrary, now that he is sad. III Although we conclude that §157(b)(2)(C) permits the Bankruptcy Court to enter final judgment on Vickie’s counterclaim, Article III of the Constitution does not. A Article III, §1, of the Constitution mandates that “[t]he judicial Power of the United States, shall be vested in one supreme Court, and in such inferior Courts as the Congress may from time to time ordain and establish.” The same section provides that the judges of those constitutional courts “shall hold their Offices during good Behaviour” and “receive for their Services[ ] a Compensation[ ] [that] shall not be diminished” during their tenure. As its text and our precedent confirm, Article III is “an inseparable element of the constitutional system of checks and balances” that “both defines the power and protects the independence of the Judicial Branch.” Northern Pipeline, 458 U. S., at 58 (plurality opinion). Under “the basic concept of separation of powers … that flow[s] from the scheme of a tripartite government” adopted in the Constitution, “the ‘judicial Power of the United States’ … can no more be shared” with another branch than “the Chief Executive, for example, can share with the Judiciary the veto power, or the Congress share with the Judiciary the power to override a Presidential veto.” United States v. Nixon, 418 U. S. 683, 704 (1974) (quoting U. S. Const., Art. III, §1). In establishing the system of divided power in the Constitution, the Framers considered it essential that “the judiciary remain[ ] truly distinct from both the legisla- ture and the executive.” The Federalist No. 78, p. 466 (C. Rossiter ed. 1961) (A. Hamilton). As Hamilton put it, quoting Montesquieu, “ ‘there is no liberty if the power of judging be not separated from the legislative and executive powers.’ ” Ibid. (quoting 1 Montesquieu, Spirit of Laws 181). We have recognized that the three branches are not hermetically sealed from one another, see Nixon v. Administrator of General Services, 433 U. S. 425, 443 (1977), but it remains true that Article III imposes some basic limitations that the other branches may not transgress. Those limitations serve two related purposes. “Separation-of-powers principles are intended, in part, to protect each branch of government from incursion by the others. Yet the dynamic between and among the branches is not the only object of the Constitution’s concern. The structural principles secured by the separation of powers protect the individual as well.” Bond v. United States, 564 U. S. ___, ___ (2011) (slip op., at 10). Article III protects liberty not only through its role in implementing the separation of powers, but also by specifying the defining characteristics of Article III judges. The colonists had been subjected to judicial abuses at the hand of the Crown, and the Framers knew the main reasons why: because the King of Great Britain “made Judges dependent on his Will alone, for the tenure of their offices, and the amount and payment of their salaries.” The Declaration of Independence ¶11. The Framers undertook in Article III to protect citizens subject to the judicial power of the new Federal Government from a repeat of those abuses. By appointing judges to serve without term limits, and restricting the ability of the other branches to remove judges or diminish their salaries, the Framers sought to ensure that each judicial decision would be rendered, not with an eye toward currying favor with Congress or the Executive, but rather with the “[c]lear heads … and honest hearts” deemed “essential to good judges.” 1 Works of James Wilson 363 (J. Andrews ed. 1896). Article III could neither serve its purpose in the system of checks and balances nor preserve the integrity of judicial decisionmaking if the other branches of the Federal Government could confer the Government’s “judicial Power” on entities outside Article III. That is why we have long recognized that, in general, Congress may not “withdraw from judicial cognizance any matter which, from its nature, is the subject of a suit at the common law, or in equity, or admiralty.” Murray’s Lessee v. Hoboken Land & Improvement Co., 18 How. 272, 284 (1856). When a suit is made of “the stuff of the traditional actions at common law tried by the courts at Westminster in 1789,” Northern Pipeline, 458 U. S., at 90 (Rehnquist, J., concurring in judgment), and is brought within the bounds of federal jurisdiction, the responsibility for deciding that suit rests with Article III judges in Article III courts. The Constitution assigns that job—resolution of “the mundane as well as the glamorous, matters of common law and statute as well as constitutional law, issues of fact as well as issues of law”—to the Judiciary. Id., at 86–87, n. 39 (plurality opinion). B This is not the first time we have faced an Article III challenge to a bankruptcy court’s resolution of a debtor’s suit. In Northern Pipeline, we considered whether bankruptcy judges serving under the Bankruptcy Act of 1978—appointed by the President and confirmed by the Senate, but lacking the tenure and salary guarantees of Article III—could “constitutionally be vested with jurisdiction to decide [a] state-law contract claim” against an entity that was not otherwise part of the bankruptcy proceedings. 458 U. S., at 53, 87, n. 40 (plurality opinion); see id., at 89–92 (Rehnquist, J., concurring in judgment). The Court concluded that assignment of such state law claims for resolution by those judges “violates Art. III of the Con-stitution.” Id., at 52, 87 (plurality opinion); id., at 91 (Rehnquist, J., concurring in judgment). The plurality in Northern Pipeline recognized that there was a category of cases involving “public rights” that Congress could constitutionally assign to “legislative” courts for resolution. That opinion concluded that this “public rights” exception extended “only to matters arising between” individuals and the Government “in connection with the performance of the constitutional functions of the executive or legislative departments … that historically could have been determined exclusively by those” branches. Id., at 67–68 (internal quotation marks omitted). A full majority of the Court, while not agreeing on the scope of the exception, concluded that the doctrine did not encompass adjudication of the state law claim at issue in that case. Id., at 69–72; see id., at 90–91 (Rehnquist, J., concurring in judgment) (“None of the [previous cases addressing Article III power] has gone so far as to sanction the type of adjudication to which Marathon will be subjected … . To whatever extent different powers granted under [the 1978] Act might be sustained under the ‘public rights’ doctrine of Murray’s Lessee … and succeeding cases, I am satisfied that the adjudication of Northern’s lawsuit cannot be so sustained”).[Footnote 5] A full majority of Justices in Northern Pipeline also rejected the debtor’s argument that the bankruptcy court’s exercise of jurisdiction was constitutional because the bankruptcy judge was acting merely as an adjunct of the district court or court of appeals. Id., at 71–72, 81–86 (plurality opinion); id., at 91 (Rehnquist, J., concurring in judgment) (“the bankruptcy court is not an ‘adjunct’ of either the district court or the court of appeals”). After our decision in Northern Pipeline, Congress revised the statutes governing bankruptcy jurisdiction and bankruptcy judges. In the 1984 Act, Congress provided that the judges of the new bankruptcy courts would be appointed by the courts of appeals for the circuits in which their districts are located. 28 U. S. C. §152(a). And, as we have explained, Congress permitted the newly constituted bankruptcy courts to enter final judgments only in “core” proceedings. See supra, at 7–8. With respect to such “core” matters, however, the bankruptcy courts under the 1984 Act exercise the same powers they wielded under the Bankruptcy Act of 1978 (1978 Act), 92 Stat. 2549. As in Northern Pipeline, for example, the newly constituted bankruptcy courts are charged under §157(b)(2)(C) with resolving “[a]ll matters of fact and law in whatever domains of the law to which” a counterclaim may lead. 458 U. S., at 91 (Rehnquist, J., concurring in judgment); see, e.g., 275 B. R., at 50–51 (noting that Vickie’s counterclaim required the bankruptcy court to determine whether Texas recognized a cause of ac- tion for tortious interference with an inter vivos gift—something the Supreme Court of Texas had yet to do). As in Northern Pipeline, the new courts in core proceedings “issue final judgments, which are binding and enforceable even in the absence of an appeal.” 458 U. S., at 85–86 (plurality opinion). And, as in Northern Pipeline, the district courts review the judgments of the bankruptcy courts in core proceedings only under the usual limited appellate standards. That requires marked deference to, among other things, the bankruptcy judges’ findings of fact. See §158(a); Fed. Rule Bkrtcy. Proc. 8013 (findings of fact “shall not be set aside unless clearly erroneous”). C Vickie and the dissent argue that the Bankruptcy Court’s entry of final judgment on her state common law counterclaim was constitutional, despite the similarities between the bankruptcy courts under the 1978 Act and those exercising core jurisdiction under the 1984 Act. We disagree. It is clear that the Bankruptcy Court in this case exercised the “judicial Power of the United States” in purporting to resolve and enter final judgment on a state common law claim, just as the court did in Northern Pipeline. No “public right” exception excuses the failure to comply with Article III in doing so, any more than in Northern Pipeline. Vickie argues that this case is different because the defendant is a creditor in the bankruptcy. But the debtors’ claims in the cases on which she relies were themselves federal claims under bankruptcy law, which would be completely resolved in the bankruptcy process of allowing or disallowing claims. Here Vickie’s claim is a state law action independent of the federal bankruptcy law and not necessarily resolvable by a ruling on the creditor’s proof of claim in bankruptcy. Northern Pipeline and our subsequent decision in Granfinanciera, 492 U. S. 33, rejected the application of the “public rights” exception in such cases. Nor can the bankruptcy courts under the 1984 Act be dismissed as mere adjuncts of Article III courts, any more than could the bankruptcy courts under the 1978 Act. The judicial powers the courts exercise in cases such as this remain the same, and a court exercising such broad powers is no mere adjunct of anyone. 1 Vickie’s counterclaim cannot be deemed a matter of “public right” that can be decided outside the Judicial Branch. As explained above, in Northern Pipeline we rejected the argument that the public rights doctrine permitted a bankruptcy court to adjudicate a state law suit brought by a debtor against a company that had not filed a claim against the estate. See 458 U. S., at 69–72 (plurality opinion); id., at 90–91 (Rehnquist, J., concurring in judgment). Although our discussion of the public rights exception since that time has not been entirely consistent, and the exception has been the subject of some debate, this case does not fall within any of the various formulations of the concept that appear in this Court’s opinions. We first recognized the category of public rights in Mur-ray’s Lessee v. Hoboken Land & Improvement Co., 18 How. 272 (1856). That case involved the Treasury Department’s sale of property belonging to a customs collector who had failed to transfer payments to the Federal Government that he had collected on its behalf. Id., at 274, 275. The plaintiff, who claimed title to the same land through a different transfer, objected that the Treasury Department’s calculation of the deficiency and sale of the property was void, because it was a judicial act that could not be assigned to the Executive under Article III. Id., at 274–275, 282–283. “To avoid misconstruction upon so grave a subject,” the Court laid out the principles guiding its analysis. Id., at 284. It confirmed that Congress cannot “withdraw from judicial cognizance any matter which, from its nature, is the subject of a suit at the common law, or in equity, or admiralty.” Ibid. The Court also recognized that “[a]t the same time there are matters, involving public rights, which may be presented in such form that the judicial power is capable of acting on them, and which are susceptible of judicial determination, but which congress may or may not bring within the cognizance of the courts of the United States, as it may deem proper.” Ibid. As an example of such matters, the Court referred to “[e]quitable claims to land by the inhabitants of ceded territories” and cited cases in which land issues were conclusively resolved by Executive Branch officials. Ibid. (citing Foley v. Harrison, 15 How. 433 (1854); Burgess v. Gray, 16 How. 48 (1854)). In those cases “it depends upon the will of congress whether a remedy in the courts shall be allowed at all,” so Congress could limit the extent to which a judicial forum was available. Murray’s Lessee, 18 How., at 284. The challenge in Murray’s Lessee to the Treasury Department’s sale of the collector’s land likewise fell within the “public rights” category of cases, because it could only be brought if the Federal Government chose to allow it by waiving sovereign immunity. Id., at 283–284. The point of Murray’s Lessee was simply that Congress may set the terms of adjudicating a suit when the suit could not otherwise proceed at all. Subsequent decisions from this Court contrasted cases within the reach of the public rights exception—those arising “between the Government and persons subject to its authority in connection with the performance of the constitutional functions of the executive or legislative departments”—and those that were instead matters “of private right, that is, of the liability of one individual to another under the law as defined.” Crowell v. Benson, 285 U. S. 22, 50, 51 (1932).[Footnote 6] See Atlas Roofing Co. v. Occupational Safety and Health Review Comm’n, 430 U. S. 442, 458 (1977) (Exception extends to cases “where the Government is involved in its sovereign capacity under … [a] statute creating enforceable public rights,” while “[w]holly private tort, contract, and property cases, as well as a vast range of other cases … are not at all implicated”); Ex parte Bakelite Corp., 279 U. S. 438, 451–452 (1929). See also Northern Pipeline, supra, at 68 (plurality opinion) (citing Ex parte Bakelite Corp. for the proposition that the doctrine extended “only to matters that historically could have been determined exclusively by” the Executive and Legislative Branches). Shortly after Northern Pipeline, the Court rejected the limitation of the public rights exception to actions involving the Government as a party. The Court has continued, however, to limit the exception to cases in which the claim at issue derives from a federal regulatory scheme, or in which resolution of the claim by an expert government agency is deemed essential to a limited regulatory objective within the agency’s authority. In other words, it is still the case that what makes a right “public” rather than private is that the right is integrally related to particular federal government action. See United States v. Jicarilla Apache Nation, 564 U. S. ___, ___–___ (2011) (slip op., at 10–11) (“The distinction between ‘public rights’ against the Government and ‘private rights’ between private parties is well established,” citing Murray’s Lessee and Crowell). Our decision in Thomas v. Union Carbide Agricultural Products Co., for example, involved a data-sharing arrangement between companies under a federal statute pro-viding that disputes about compensation between the companies would be decided by binding arbitration. 473 U. S. 568, 571–575 (1985). This Court held that the scheme did not violate Article III, explaining that “[a]ny right to compensation … results from [the statute] and does not depend on or replace a right to such compensation under state law.” Id., at 584. Commodity Futures Trading Commission v. Schor concerned a statutory scheme that created a procedure for customers injured by a broker’s violation of the federal commodities law to seek reparations from the broker before the Commodity Futures Trading Commission (CFTC). 478 U. S. 833, 836 (1986). A customer filed such a claim to recover a debit balance in his account, while the broker filed a lawsuit in Federal District Court to recover the same amount as lawfully due from the customer. The broker later submitted its claim to the CFTC, but after that agency ruled against the customer, the customer argued that agency jurisdiction over the broker’s counterclaim violated Article III. Id., at 837–838. This Court disagreed, but only after observing that (1) the claim and the counterclaim concerned a “single dispute”—the same account balance; (2) the CFTC’s assertion of authority involved only “a narrow class of common law claims” in a “ ‘particularized area of law’ ”; (3) the area of law in question was governed by “a specific and limited federal regulatory scheme” as to which the agency had “obvious expertise”; (4) the parties had freely elected to resolve their differences before the CFTC; and (5) CFTC orders were “enforceable only by order of the district court.” Id., at 844, 852–855 (quoting Northern Pipeline, 458 U. S., at 85); see 478 U. S., at 843–844; 849–857. Most significantly, given that the customer’s reparations claim before the agency and the broker’s counterclaim were competing claims to the same amount, the Court repeatedly emphasized that it was “necessary” to allow the agency to exercise jurisdiction over the broker’s claim, or else “the reparations procedure would have been confounded.” Id., at 856. The most recent case in which we considered application of the public rights exception—and the only case in which we have considered that doctrine in the bankruptcy context since Northern Pipeline—is Granfinanciera, S. A. v. Nordberg, 492 U. S. 33 (1989). In Granfinanciera we rejected a bankruptcy trustee’s argument that a fraudulent conveyance action filed on behalf of a bankruptcy estate against a noncreditor in a bankruptcy proceeding fell within the “public rights” exception. We explained that, “[i]f a statutory right is not closely intertwined with a federal regulatory program Congress has power to enact, and if that right neither belongs to nor exists against the Federal Government, then it must be adjudicated by an Article III court.” Id., at 54–55. We reasoned that fraudulent conveyance suits were “quintessentially suits at common law that more nearly resemble state law contract claims brought by a bankrupt corporation to augment the bankruptcy estate than they do creditors’ hierarchically ordered claims to a pro rata share of the bankruptcy res.” Id., at 56. As a consequence, we concluded that fraudulent conveyance actions were “more accurately characterized as a private rather than a public right as we have used those terms in our Article III decisions.” Id., at 55.[Footnote 7] Vickie’s counterclaim—like the fraudulent conveyance claim at issue in Granfinanciera—does not fall within any of the varied formulations of the public rights exception in this Court’s cases. It is not a matter that can be pursued only by grace of the other branches, as in Murray’s Lessee, 18 How., at 284, or one that “historically could have been determined exclusively by” those branches, Northern Pipeline, supra, at 68 (citing Ex parte Bakelite Corp., 279 U. S., at 458). The claim is instead one under state common law between two private parties. It does not “depend[ ] on the will of congress,” Murray’s Lessee, supra, at 284; Congress has nothing to do with it. In addition, Vickie’s claimed right to relief does not flow from a federal statutory scheme, as in Thomas, 473 U. S., at 584–585, or Atlas Roofing, 430 U. S., at 458. It is not “completely dependent upon” adjudication of a claim created by federal law, as in Schor, 478 U. S., at 856. And in contrast to the objecting party in Schor, id., at 855–856, Pierce did not truly consent to resolution of Vickie’s claim in the bankruptcy court proceedings. He had nowhere else to go if he wished to recover from Vickie’s estate. See Granfinanciera, supra, at 59, n. 14 (noting that “[p]arallel reasoning [to Schor] is unavailable in the context of bankruptcy proceedings, because creditors lack an alternative forum to the bankruptcy court in which to pursue their claims”).[Footnote 8] Furthermore, the asserted authority to decide Vickie’s claim is not limited to a “particularized area of the law,” as in Crowell, Thomas, and Schor. Northern Pipeline, 458 U. S., at 85 (plurality opinion). We deal here not with an agency but with a court, with substantive jurisdiction reaching any area of the corpus juris. See ibid.; id., at 91 (Rehnquist, J., concurring in judgment). This is not a situation in which Congress devised an “expert and inexpensive method for dealing with a class of questions of fact which are particularly suited to examination and determination by an administrative agency specially assigned to that task.” Crowell, 285 U. S., at 46; see Schor, supra, at 855–856. The “experts” in the federal system at resolving common law counterclaims such as Vickie’s are the Article III courts, and it is with those courts that her claim must stay. The dissent reads our cases differently, and in particular contends that more recent cases view Northern Pipeline as “ ‘establish[ing] only that Congress may not vest in a non-Article III court the power to adjudicate, render final judgment, and issue binding orders in a traditional contract action arising under state law, without consent of the litigants, and subject only to ordinary appellate review.’ ” Post, at 6 (quoting Thomas, supra, at 584). Just so: Substitute “tort” for “contract,” and that statement directly covers this case. We recognize that there may be instances in which the distinction between public and private rights—at least as framed by some of our recent cases—fails to provide concrete guidance as to whether, for example, a particular agency can adjudicate legal issues under a substantive regulatory scheme. Given the extent to which this case is so markedly distinct from the agency cases discussing the public rights exception in the context of such a regime, however, we do not in this opinion express any view on how the doctrine might apply in that different context. What is plain here is that this case involves the most prototypical exercise of judicial power: the entry of a final, binding judgment by a court with broad substantive jurisdiction, on a common law cause of action, when the action neither derives from nor depends upon any agency regulatory regime. If such an exercise of judicial power may nonetheless be taken from the Article III Judiciary simply by deeming it part of some amorphous “public right,” then Article III would be transformed from the guardian of individual liberty and separation of powers we have long recognized into mere wishful thinking. 2 Vickie and the dissent next attempt to distinguish Northern Pipeline and Granfinanciera on the ground that Pierce, unlike the defendants in those cases, had filed a proof of claim in the bankruptcy proceedings. Given Pierce’s participation in those proceedings, Vickie argues, the Bankruptcy Court had the authority to adjudicate her counterclaim under our decisions in Katchen v. Landy, 382 U. S. 323 (1966), and Langenkamp v. Culp, 498 U. S. 42 (1990) (per curiam). We do not agree. As an initial matter, it is hard to see why Pierce’s decision to file a claim should make any difference with respect to the characterization of Vickie’s counterclaim. “ ‘[P]roperty interests are created and defined by state law,’ and ‘[u]nless some federal interest requires a different result, there is no reason why such interests should be analyzed differently simply because an interested party is involved in a bankruptcy proceeding.” Travelers Casualty & Surety Co. of America v. Pacific Gas & Elec. Co., 549 U. S. 443, 451 (2007) (quoting Butner v. United States, 440 U. S. 48, 55 (1979)). Pierce’s claim for defamation in no way affects the nature of Vickie’s counterclaim for tortious interference as one at common law that simply attempts to augment the bankruptcy estate—the very type of claim that we held in Northern Pipeline and Granfinanciera must be decided by an Article III court. Contrary to Vickie’s contention, moreover, our decisions in Katchen and Langenkamp do not suggest a different result. Katchen permitted a bankruptcy referee acting under the Bankruptcy Acts of 1898 and 1938 (akin to a bankruptcy court today) to exercise what was known as “summary jurisdiction” over a voidable preference claim brought by the bankruptcy trustee against a creditor who had filed a proof of claim in the bankruptcy proceeding. See 382 U. S., at 325, 327–328. A voidable preference claim asserts that a debtor made a payment to a particular creditor in anticipation of bankruptcy, to in effect increase that creditor’s proportionate share of the estate. The preferred creditor’s claim in bankruptcy can be disallowed as a result of the preference, and the amounts paid to that creditor can be recovered by the trustee. See id., at 330; see also 11 U. S. C. §§502(d), 547(b). Although the creditor in Katchen objected that the preference issue should be resolved through a “plenary suit” in an Article III court, this Court concluded that summary adjudication in bankruptcy was appropriate, because it was not possible for the referee to rule on the creditor’s proof of claim without first resolving the voidable preference issue. 382 U. S., at 329–330, 332–333, and n. 9, 334. There was no question that the bankruptcy referee could decide whether there had been a voidable preference in determining whether and to what extent to allow the creditor’s claim. Once the referee did that, “nothing remains for adjudication in a plenary suit”; such a suit “would be a meaningless gesture.” Id., at 334. The plenary proceeding the creditor sought could be brought into the bankruptcy court because “the same issue [arose] as part of the process of allowance and disallowance of claims.” Id., at 336. It was in that sense that the Court stated that “he who invokes the aid of the bankruptcy court by offering a proof of claim and demanding its allowance must abide the consequences of that procedure.” Id., at 333, n. 9. In Katchen one of those consequences was resolution of the preference issue as part of the process of allowing or disallowing claims, and accordingly there was no basis for the creditor to insist that the issue be resolved in an Article III court. See id., at 334. Indeed, the Katchen Court expressly noted that it “intimate[d] no opinion concerning whether” the bankruptcy referee would have had “summary jurisdiction to adjudicate a demand by the [bankruptcy] trustee for affirmative relief, all of the substantial factual and legal bases for which ha[d] not been disposed of in passing on objections to the [creditor’s proof of] claim.” Id., at 333, n. 9. Our per curiam opinion in Langenkamp is to the same effect. We explained there that a preferential transfer claim can be heard in bankruptcy when the allegedly favored creditor has filed a claim, because then “the ensuing preference action by the trustee become[s] integral to the restructuring of the debtor-creditor relationship.” 498 U. S., at 44. If, in contrast, the creditor has not filed a proof of claim, the trustee’s preference action does not “become[ ] part of the claims-allowance process” subject to resolution by the bankruptcy court. Ibid.; see id., at 45. In ruling on Vickie’s counterclaim, the Bankruptcy Court was required to and did make several factual and legal determinations that were not “disposed of in pass- ing on objections” to Pierce’s proof of claim for defama- tion, which the court had denied almost a year earlier. Katchen, supra, at 332, n. 9. There was some overlap between Vickie’s counterclaim and Pierce’s defamation claim that led the courts below to conclude that the counterclaim was compulsory, 600 F. 3d, at 1057, or at least in an “attenuated” sense related to Pierce’s claim, 264 B. R., at 631. But there was never any reason to believe that the process of adjudicating Pierce’s proof of claim would necessarily resolve Vickie’s counterclaim. See id., at 631, 632 (explaining that “the primary facts at issue on Pierce’s claim were the relationship between Vickie and her attorneys and her knowledge or approval of their statements,” and “the counterclaim raises issues of law entirely dif-ferent from those raise[d] on the defamation claim”). The United States acknowledges the point. See Brief for United States as Amicus Curiae, p. (I) (question presented concerns authority of a bankruptcy court to enter final judgment on a compulsory counterclaim “when adjudication of the counterclaim requires resolution of issues that are not implicated by the claim against the estate”); id., at 26. The only overlap between the two claims in this case was the question whether Pierce had in fact tortiously taken control of his father’s estate in the manner alleged by Vickie in her counterclaim and described in the allegedly defamatory statements. From the outset, it was clear that, even assuming the Bankruptcy Court would (as it did) rule in Vickie’s favor on that question, the court could not enter judgment for Vickie unless the court additionally ruled on the questions whether Texas recognized tortious interference with an expected gift as a valid cause of action, what the elements of that action were, and whether those elements were met in this case. 275 B. R., at 50–53. Assuming Texas accepted the elements adopted by other jurisdictions, that meant Vickie would need to prove, above and beyond Pierce’s tortious interference, (1) the existence of an expectancy of a gift; (2) a reasonable certainty that the expectancy would have been realized but for the interference; and (3) damages. Id., at 51; see 253 B. R., at 558–561. Also, because Vickie sought punitive damages in connection with her counterclaim, the Bankruptcy Court could not finally dispose of the case in Vickie’s favor without determining whether to subject Pierce to the sort of “retribution,” “punishment[,] and deterrence,” Exxon Shipping Co., 554 U. S., at 492, 504 (internal quotation marks omitted), those damages are designed to impose. There thus was never reason to believe that the process of ruling on Pierce’s proof of claim would necessarily result in the resolution of Vickie’s counterclaim. In both Katchen and Langenkamp, moreover, the trustee bringing the preference action was asserting a right of recovery created by federal bankruptcy law. In Langen-kamp, we noted that “the trustee instituted adversary proceedings under 11 U. S. C. §547(b) to recover, as avoidable preferences,” payments respondents received from the debtor before the bankruptcy filings. 498 U. S., at 43; see, e.g., §547(b)(1) (“the trustee may avoid any transfer of an interest of the debtor in property—(1) to or for the benefit of a creditor”). In Katchen, “[t]he Trustee … [asserted] that the payments made [to the creditor] were preferences inhibited by Section 60a of the Bankruptcy Act.” Memorandum Opinion (Feb. 8, 1963), Tr. of Record in O. T. 1965, No. 28, p. 3; see 382 U. S., at 334 (considering impact of the claims allowance process on “action by the trustee under §60 to recover the preference”); 11 U. S. C. §96(b) (1964 ed.) (§60(b) of the then-applicable Bankruptcy Act) (“preference may be avoided by the trustee if the creditor receiving it or to be benefited thereby … has, at the time when the transfer is made, reasonable cause to believe that the debtor is insolvent”). Vickie’s claim, in contrast, is in no way derived from or dependent upon bankruptcy law; it is a state tort action that exists without regard to any bankruptcy proceeding. In light of all the foregoing, we disagree with the dissent that there are no “relevant distinction[s]” between Pierce’s claim in this case and the claim at issue in Langenkamp. Post, at 14. We see no reason to treat Vickie’s counterclaim any differently from the fraudulent conveyance action in Granfinanciera. 492 U. S., at 56. Granfinanciera’s distinction between actions that seek “to augment the bankruptcy estate” and those that seek “a pro rata share of the bankruptcy res,” ibid., reaffirms that Congress may not bypass Article III simply because a proceeding may have some bearing on a bankruptcy case; the question is whether the action at issue stems from the bank-ruptcy itself or would necessarily be resolved in the claims allowance process. Vickie has failed to demonstrate that her counterclaim falls within one of the “limited circumstances” covered by the public rights exception, particularly given our conclusion that, “even with respect to matters that arguably fall within the scope of the ‘public rights’ doctrine, the presumption is in favor of Art. III courts.” Northern Pipeline, 458 U. S., at 69, n. 23, 77, n. 29 (plurality opinion). 3 Vickie additionally argues that the Bankruptcy Court’s final judgment was constitutional because bankruptcy courts under the 1984 Act are properly deemed “adjuncts” of the district courts. Brief for Petitioner 61–64. We rejected a similar argument in Northern Pipeline, see 458 U. S., at 84–86 (plurality opinion); id., at 91 (Rehnquist, J., concurring in judgment), and our reasoning there holds true today. To begin, as explained above, it is still the bankruptcy court itself that exercises the essential attributes of judicial power over a matter such as Vickie’s counterclaim. See supra, at 20. The new bankruptcy courts, like the old, do not “ma[k]e only specialized, narrowly confined factual determinations regarding a particularized area of law” or engage in “statutorily channeled factfinding functions.” Northern Pipeline, 458 U. S., at 85 (plurality opinion). Instead, bankruptcy courts under the 1984 Act resolve “[a]ll matters of fact and law in whatever domains of the law to which” the parties’ counterclaims might lead. Id., at 91 (Rehnquist, J., concurring in judgment). In addition, whereas the adjunct agency in Crowell v. Benson “possessed only a limited power to issue compensation orders … [that] could be enforced only by order of the district court,” Northern Pipeline, supra, at 85, a bankruptcy court resolving a counterclaim under 28 U. S. C. §157(b)(2)(C) has the power to enter “appropriate orders and judgments”—including final judgments—subject to review only if a party chooses to appeal, see §§157(b)(1), 158(a)–(b). It is thus no less the case here than it was in Northern Pipeline that “[t]he authority—and the respon-sibility—to make an informed, final determination … remains with” the bankruptcy judge, not the district court. 458 U. S., at 81 (plurality opinion) (internal quotation marks omitted). Given that authority, a bankruptcy court can no more be deemed a mere “adjunct” of the district court than a district court can be deemed such an “adjunct” of the court of appeals. We certainly cannot accept the dissent’s notion that judges who have the power to enter final, binding orders are the “functional[ ]” equivalent of “law clerks[ ] and the Judiciary’s administrative officials.” Post, at 11. And even were we wrong in this regard, that would only confirm that such judges should not be in the business of entering final judgments in the first place. It does not affect our analysis that, as Vickie notes, bankruptcy judges under the current Act are appointed by the Article III courts, rather than the President. See Brief for Petitioner 59. If—as we have concluded—the bankruptcy court itself exercises “the essential attributes of judicial power [that] are reserved to Article III courts,” Schor, 478 U. S., at 851 (internal quotation marks omitted), it does not matter who appointed the bankruptcy judge or authorized the judge to render final judgments in such proceedings. The constitutional bar remains. See The Federalist No. 78, at 471 (“Periodical appointments, however regulated, or by whomsoever made, would, in some way or other, be fatal to [a judge’s] necessary independence”). D Finally, Vickie and her amici predict as a practical matter that restrictions on a bankruptcy court’s ability to hear and finally resolve compulsory counterclaims will create significant delays and impose additional costs on the bankruptcy process. See, e.g., Brief for Petitioner 34–36, 57–58; Brief for United States as Amicus Curiae 29–30. It goes without saying that “the fact that a given law or procedure is efficient, convenient, and useful in facilitating functions of government, standing alone, will not save it if it is contrary to the Constitution.” INS v. Chadha, 462 U. S. 919, 944 (1983). In addition, we are not convinced that the practical consequences of such limitations on the authority of bankruptcy courts to enter final judgments are as significant as Vickie and the dissent suggest. See post, at 16–17. The dissent asserts that it is important that counterclaims such as Vickie’s be resolved “in a bankruptcy court,” and that, “to be effective, a single tribunal must have broad authority to restructure [debtor-creditor] relations.” Post, at 14, 15 (emphasis deleted). But the framework Congress adopted in the 1984 Act already contemplates that certain state law matters in bankruptcy cases will be resolved by judges other than those of the bankruptcy courts. Section 1334(c)(2), for example, requires that bankruptcy courts abstain from hearing specified non-core, state law claims that “can be timely adjudicated[ ] in a State forum of appropriate jurisdiction.” Section 1334(c)(1) similarly provides that bankruptcy courts may abstain from hearing any proceeding, including core matters, “in the interest of comity with State courts or respect for State law.” As described above, the current bankruptcy system also requires the district court to review de novo and enter final judgment on any matters that are “related to” the bankruptcy proceedings, §157(c)(1), and permits the district court to withdraw from the bankruptcy court any referred case, proceeding, or part thereof, §157(d). Pierce has not argued that the bankruptcy courts “are barred from ‘hearing’ all counterclaims” or proposing findings of fact and conclusions of law on those matters, but rather that it must be the district court that “finally decide[s]” them. Brief for Respondent 61. We do not think the removal of counterclaims such as Vickie’s from core bankruptcy jurisdiction meaningfully changes the division of labor in the current statute; we agree with the United States that the question presented here is a “narrow” one. Brief for United States as Amicus Curiae 23. If our decision today does not change all that much, then why the fuss? Is there really a threat to the separation of powers where Congress has conferred the judicial power outside Article III only over certain counterclaims in bankruptcy? The short but emphatic answer is yes. A statute may no more lawfully chip away at the authority of the Judicial Branch than it may eliminate it entirely. “Slight encroachments create new boundaries from which legions of power can seek new territory to capture.” Reid v. Covert, 354 U. S. 1, 39 (1957) (plurality opinion). Although “[i]t may be that it is the obnoxious thing in its mildest and least repulsive form,” we cannot overlook the intrusion: “illegitimate and unconstitutional practices get their first footing in that way, namely, by silent approaches and slight deviations from legal modes of procedure.” Boyd v. United States, 116 U. S. 616, 635 (1886). We cannot compromise the integrity of the system of separated powers and the role of the Judiciary in that system, even with respect to challenges that may seem innocuous at first blush. * * * Article III of the Constitution provides that the judicial power of the United States may be vested only in courts whose judges enjoy the protections set forth in that Article. We conclude today that Congress, in one isolated respect, exceeded that limitation in the Bankruptcy Act of 1984. The Bankruptcy Court below lacked the constitutional authority to enter a final judgment on a state law counterclaim that is not resolved in the process of ruling on a creditor’s proof of claim. Accordingly, the judgment of the Court of Appeals is affirmed. It is so ordered. Footnote 1 Because both Vickie and Pierce passed away during this litigation, the parties in this case are Vickie’s estate and Pierce’s estate. We continue to refer to them as “Vickie” and “Pierce.” Footnote 2 One judge wrote a separate concurring opinion. He concluded that “Vickie’s counterclaim … [wa]s not a core proceeding, so the Texas probate court judgment preceded the district court judgment and controls.” 600 F. 3d, at 1065 (Kleinfeld, J.). The concurring judge also “offer[ed] additional grounds” that he believed required judgment in Pierce’s favor. Ibid. Pierce presses only one of those additional grounds here; it is discussed below, in Part II–C. Footnote 3 In full, §§157(b)(1)–(2) provides: “(1) Bankruptcy judges may hear and determine all cases under title 11 and all core proceedings arising under title 11, or arising in a case under title 11, referred under subsection (a) of this section, and may enter appropriate orders and judgments, subject to review under section 158 of this title. “(2) Core proceedings include, but are not limited to— “(A) matters concerning the administration of the estate; “(B) allowance or disallowance of claims against the estate or exemptions from property of the estate, and estimation of claims or interests for the purposes of confirming a plan under chapter 11, 12, or 13 of title 11 but not the liquidation or estimation of contingent or unliquidated personal injury tort or wrongful death claims against the estate for purposes of distribution in a case under title 11; “(C) counterclaims by the estate against persons filing claims against the estate; “(D) orders in respect to obtaining credit; “(E) orders to turn over property of the estate; “(F) proceedings to determine, avoid, or recover preferences; “(G) motions to terminate, annul, or modify the automatic stay; “(H) proceedings to determine, avoid, or recover fraudulent conveyances; “(I) determinations as to the dischargeability of particular debts; “(J) objections to discharges; “(K) determinations of the validity, extent, or priority of liens; “(L) confirmations of plans; “(M) orders approving the use or lease of property, including the use of cash collateral; “(N) orders approving the sale of property other than property resulting from claims brought by the estate against persons who have not filed claims against the estate; “(O) other proceedings affecting the liquidation of the assets of the estate or the adjustment of the debtor-creditor or the equity security holder relationship, except personal injury tort or wrongful death claims; and “(P) recognition of foreign proceedings and other matters under chapter 15 of title 11.” Footnote 4 Although Pierce suggests that consideration of “the 157(b)(5) issue” would facilitate an “easy” resolution of the case, Tr. of Oral Arg. 47–48, he is mistaken. Had Pierce preserved his argument under that provision, we would have been confronted with several questions on which there is little consensus or precedent. Those issues include: (1) the scope of the phrase “personal injury tort”—a question over which there is at least a three-way divide, see In re Arnold, 407 B. R. 849, 851–853 (Bkrtcy. Ct. MDNC 2009); (2) whether, as Vickie argued in the Court of Appeals, the requirement that a personal injury tort claim be “tried” in the district court nonetheless permits the bankruptcy court to resolve the claim short of trial, see Appellee’s/Cross-Appellant’s Supplemental Brief in No. 02–56002 etc. (CA9), p. 24; see also In re Dow Corning Corp., 215 B. R. 346, 349–351 (Bkrtcy. Ct. ED Mich. 1997) (noting divide over whether, and on what grounds, a bankruptcy court may resolve a claim pretrial); and (3) even if Pierce’s defamation claim could be considered only by the District Court, whether the Bankruptcy Court might retain jurisdiction over the counterclaim, cf. Arbaugh v. Y & H Corp., 546 U. S. 500, 514 (2006) (“when a court grants a motion to dismiss for failure to state a federal claim, the court generally retains discretion to exercise supplemental jurisdiction, pursuant to 28 U. S. C. §1367, over pendent state-law claims”). We express no opinion on any of these issues and simply note that the §157(b)(5) question is not as straightforward as Pierce would have it. Footnote 5 The dissent is thus wrong in suggesting that less than a full Court agreed on the points pertinent to this case. Post, at 2 (opinion of Breyer, J.). Footnote 6 Although the Court in Crowell went on to decide that the facts of the private dispute before it could be determined by a non-Article III tribunal in the first instance, subject to judicial review, the Court did so only after observing that the administrative adjudicator had only limited authority to make specialized, narrowly confined factual determinations regarding a particularized area of law and to issue orders that could be enforced only by action of the District Court. 285 U. S., at 38, 44–45, 54; see Northern Pipeline Constr. Co. v. Marathon Pipe Line Co., 458 U. S. 50, 78 (1982) (plurality opinion). In other words, the agency in Crowell functioned as a true “adjunct” of the District Court. That is not the case here. See infra, at 34–36. Although the dissent suggests that we understate the import of Crowell in this regard, the dissent itself recognizes—repeatedly—that Crowell by its terms addresses the determination of facts outside Article III. See post, at 4 (Crowell “upheld Congress’ delegation of primary factfinding authority to the agency”); post, at 12 (quoting Crowell, 285 U. S., at 51, for the proposition that “ ‘there is no requirement that, in order to maintain the essential attributes of the judicial power, all determinations of fact in constitutional courts shall be made by judges’ ”). Crowell may well have additional significance in the context of expert administrative agencies that oversee particular substantive federal regimes, but we have no occasion to and do not address those issues today. See infra, at 29. The United States apparently agrees that any broader significance of Crowell is not pertinent in this case, citing to Crowell in its brief only once, in the last footnote, again for the limited proposition discussed above. Brief for United States as Amicus Curiae 32, n. 5. Footnote 7 We noted that we did not mean to “suggest that the restructuring of debtor-creditor relations is in fact a public right.” 492 U. S., at 56, n. 11. Our conclusion was that, “even if one accepts this thesis,” Congress could not constitutionally assign resolution of the fraudulent conveyance action to a non-Article III court. Ibid. Because neither party asks us to reconsider the public rights framework for bankruptcy, we follow the same approach here. Footnote 8 Contrary to the claims of the dissent, see post, at 12–13, Pierce did not have another forum in which to pursue his claim to recover from Vickie’s pre-bankruptcy assets, rather than take his chances with whatever funds might remain after the Title 11 proceedings. Creditors who possess claims that do not satisfy the requirements for nondischargeability under 11 U. S. C. §523 have no choice but to file their claims in bankruptcy proceedings if they want to pursue the claims at all. That is why, as we recognized in Granfinanciera, the notion of “consent” does not apply in bankruptcy proceedings as it might in other contexts.

564.US.2010_09-11311

When he pleaded guilty to being a felon in possession of a firearm, see 8 U. S. C. §922(g)(1), petitioner Sykes had prior convictions for at least three felonies, including the state-law crime of “us[ing] a vehicle” to “knowingly or intentionally” “fle[e] from a law enforcement officer” after being ordered to stop, Ind. Code §35–44–3–3(b)(1)(A) (2004). The Federal District Court decided that the prior convictions subjected Sykes to the 15-year mandatory minimum prison term that the Armed Career Criminal Act (ACCA), 18 U. S. C. §924(e), provides for an armed defendant who has three prior “violent felony” convictions. Rejecting Sykes’ argument that his vehicle flight felony was not “violent” under ACCA, the Seventh Circuit affirmed. Held: Felony vehicle flight, as proscribed by Indiana law, is a violent felony for purposes of ACCA. Pp. 5–14. (a) The “categorical approach” used to determine if a particular crime is a violent felony “consider[s] whether the elements of the offense are of the type that would justify its inclusion within the residual provision [of 18 U. S. C. §924(e)(2)(B)], without inquiring into the specific conduct of th[e] particular offender.” James v. United States, 550 U. S. 192, 202 (emphasis deleted). When punishable by more than one year in prison, burglary, arson, extortion, and crimes that involve use of explosives are violet felonies. Under the residual clause in question so too is a crime that “otherwise involves conduct that presents a serious potential risk of physical injury to another,” §924(e)(2)(B)(ii), i.e., a risk “comparable to that posed by its closest analog among” the statute’s enumerated offenses. Id., at 203. When a perpetrator flees police in a car, his determination to elude capture makes a lack of concern for the safety of others an inherent part of the offense. Even if he drives without going full speed or the wrong way, he creates the possibility that police will, in a legitimate and lawful manner, exceed or almost match his speed or use force to bring him within their custody. His indifference to these collateral consequences has violent—even lethal—potential for others. A fleeing criminal who creates a risk of this dimension takes action similar in degree of danger to that involved in arson, which also entails intentional release of a destructive force dangerous to others. Also telling is a comparison to burglary, which is dangerous because it can end in confrontation leading to violence. In fact, the risks associated with vehicle flight may outstrip the dangers of both burglary and arson. While statistics are not dispositive, studies show that the risk of personal injuries is about 20% lower for each of those enumerated crimes than for vehicle pursuits. Thus, Indiana’s prohibition on vehicle flight falls within §924(e)(2)(B)(ii)’s residual clause because, as a categorical matter, it presents a serious potential risk of physical injury to another. Pp. 5–9. (b) Sykes’ argument—that Begay v. United States, 553 U. S. 137, and Chambers v. United States, 555 U. S. 122, require ACCA predicate crimes to be purposeful, violent, and aggressive in ways that vehicle flight is not—overreads those opinions. In general, levels of risk divide crimes that qualify as violent felonies from those that do not. Chambers is no exception: It explained that failure to report does not qualify because the typical offender is not “significantly more likely than others to attack, or physically to resist, an apprehender.” 555 U. S., at ___–___. Begay, which held that driving under the influence (DUI) is not an ACCA predicate and stated that it is not purposeful, violent, and aggressive, 553 U. S., at 145–148, is the Court’s sole residual clause decision in which risk was not the dispositive factor. But Begay also gave a more specific reason for its holding: DUI “need not be purposeful or deliberate,” id., at 145, and is analogous to strict-liability, negligence, and recklessness crimes. Begay’s “purposeful, violent, and aggressive” phrase is an addition to the statutory text that has no precise link to the residual clause. Because vehicle flight is not a strict-liability, negligence, or recklessness crime and is, as a categorical matter, similar in risk to the crimes listed in the residual clause, it is a violent felony. Pp. 10–11 (c) Sykes contends that the fact that Ind. Code §35–44–3–3(b)(1)(B) criminalizes flight by an offender who “operates a vehicle in a manner that creates a substantial risk of bodily injury to another person” indicates that Indiana did not intend for §35–44–3–3(b)(1)(A), under which he was convicted, to encompass the particular class of vehicle flights reached by subsection (b)(1)(B). This argument is unconvincing. Indiana treats the two subsections as felonies of the same magnitude carrying similar prison terms, suggesting that subsection (b)(1)(A) is roughly equivalent to one type of subsection (b)(1)(B) violation. Pp. 11–13. (d) Congress framed ACCA in general and qualitative, rather than encyclopedic, terms. The residual clause imposes enhanced punishment for unlawful firearm possession when the relevant prior offenses involved a potential risk of physical injury similar to that presented by several enumerated offenses. It instructs potential recidivists regarding the applicable sentencing regime if they again transgress. This intelligible principle provides guidance, allowing a person to conform his conduct to the law. While this approach may at times be more difficult for courts to implement, it is within congressional power to enact. Pp. 13–14. 598 F. 3d 334, affirmed Kennedy, J., delivered the opinion of the Court, in which Roberts, C. J., and Breyer, Alito, and Sotomayor, JJ., joined. Thomas, J., filed an opinion concurring in the judgment. Scalia, J., filed a dissenting opinion. Kagan, J., filed a dissenting opinion, in which Ginsburg, J., joined.

It is a federal crime for a convicted felon to be in unlawful possession of a firearm. 18 U. S. C. §922(g)(1). The or-dinary maximum sentence for that crime is 10 years of imprisonment. §924(a)(2). If, however, when the unlawful possession occurred, the felon had three previous convictions for a violent felony or serious drug offense, the punishment is increased to a minimum term of 15 years. §924(e). The instant case is another in a series in which the Court is called upon to interpret §924(e) to determine if a particular previous conviction was for a “violent felony,” as that term is used in the punishment enhancement statute. See James v. United States, 550 U. S. 192 (2007); Begay v. United States, 553 U. S. 137 (2008); Chambers v. United States, 555 U. S. 122 (2009). In this case the previous conviction in question is under an Indiana statute that makes it a criminal offense whenever the driver of a vehicle knowingly or intentionally “flees from a law enforcement officer.” Ind. Code §35–44–3–3 (2004). The relevant text of the statute is set out in the discussion below. For the reasons explained, the vehicle flight that the statute proscribes is a violent felony as the federal statute uses that term. I Petitioner Marcus Sykes pleaded guilty to being a felon in possession of a firearm, 18 U. S. C. §922(g)(1), in connection with an attempted robbery of two people at gunpoint. Sykes had previous convictions for at least three felonies. On two separate occasions Sykes used a firearm to commit robbery, in one case to rob a man of his $200 wristwatch and in another to rob a woman of her purse. His third prior felony is the one of concern here. Sykes was convicted for vehicle flight, in violation of Indiana’s “resisting law enforcement” law. Ind. Code §35–44–3–3. That law provides: “(a) A person who knowingly or intentionally: “(1) forcibly resists, obstructs, or interferes with a law enforcement officer or a person assisting the officer while the officer is lawfully engaged in the execution of his duties as an officer; “(2) forcibly resists, obstructs, or interferes with the authorized service or execution of a civil or criminal process or order of a court; or “(3) flees from a law enforcement officer after the officer has, by visible or audible means, identified himself and ordered the person to stop; “commits resisting law enforcement, a Class A misdemeanor, except as provided in subsection (b). “(b) The offense under subsection (a) is a: “(1) Class D felony if: “(A) the offense is described in subsection (a)(3) and the person uses a vehicle to commit the offense; or “(B) while committing any offense described in subsection (a), the person draws or uses a deadly weapon, inflicts bodily injury on another person, or operates a vehicle in a manner that creates a substantial risk of bodily injury to another person; (2) Class C felony if, while committing any offense described in subsection (a), the person operates a vehicle in a manner that causes serious bodily injury to another person; and (3) Class B felony if, while committing any offense described in subsection (a), the person operates a vehicle in a manner that causes the death of another person.” Here, as will be further explained, Sykes used a vehicle to flee after an officer ordered him to stop, which was, as the statute provides, a class D felony. The Court of Appeals of Indiana has interpreted the crime of vehicle flight to require “a knowing attempt to escape law enforcement.” Woodward v. State, 770 N. E. 2d 897, 901 (2002) (internal quotation marks omitted). Woodward involved a driver who repeatedly flashed his bright lights and failed to obey traffic signals. Id., at 898. When an officer activated his emergency equipment, the defendant became “aware … that [the officer] wanted him to pull his vehicle over,” but instead drove for a mile without “stopping, slowing, or otherwise acknowledging” the officer because, he later testified, he “was ‘trying to rationalize why [he] would be pulled over.’ ” Id., at 898, 901. Though the defendant later claimed that he was also seeking a “well-lighted place to stop where there would be someone who knew him,” id., at 901, his actions suggested otherwise. He passed two gas stations, a food outlet store, and a McDonald’s be- fore pulling over. When he got out of the car, he began to shout profanities at the pursuing officer. Ibid. By that time, the officer had called for backup and exited his own vehicle with his gun drawn. Id., at 898. In answering the defendant’s challenge to the sufficiency of the above evidence, the Indiana court held that because he knew that a police officer sought to stop him, the defendant could not “choose the location of the stop” and insist on completing the stop “on his own terms,” as he had done, “without adequate justification,” which he lacked. Id., at 901–902. In the instant case a report prepared for Sykes’ federal sentencing describes the details of the Indiana crime. After observing Sykes driving without using needed headlights, police activated their emergency equipment for a traffic stop. Sykes did not stop. A chase ensued. Sykes wove through traffic, drove on the wrong side of the road and through yards containing bystanders, passed through a fence, and struck the rear of a house. Then he fled on foot. He was found only with the aid of a police dog. The District Court decided that his three prior convictions, including the one for violating the prohibition on vehicle flight in subsection (b)(1)(A) of the Indiana statute just discussed, were violent felonies for purposes of §924(e) and sentenced Sykes to 188 months of imprisonment. On appeal Sykes conceded that his two prior robbery con-victions were violent felonies. He did not dispute that his vehicle flight offense was a felony, but he did argue that it was not violent. The Court of Appeals for the Seventh Circuit affirmed. 598 F. 3d 334 (2010). The court’s opinion was consistent with the rulings of the Courts of Appeals in the First, Fifth, Sixth, and Tenth Circuits. Powell v. United States, 430 F. 3d 490 (CA1 2005) (per curiam); United States v. Harrimon, 568 F. 3d 531, 534–537 (CA5 2009); United States v. LaCasse, 567 F. 3d 763, 765–767 (CA6 2009); United States v. McConnell, 605 F. 3d 822, 827–830 (CA10 2010) (finding the flight to be a “crime of violence” under the “nearly identical” §4B1.2(a)(2) of the United States Sentencing Guidelines). It was in conflict with a ruling by a Court of Appeals for the Eleventh Circuit in United States v. Harrison, 558 F. 3d 1280, 1291–1296 (2009), and at least in tension, if not in conflict, with the reasoning of the Court of Appeals for the Eighth Circuit in United States v. Tyler, 580 F. 3d 722, 724–726 (2009), and for the Ninth Circuit in United States v. Kelly, 422 F. 3d 889, 892–897 (2005), United States v. Jennings, 515 F. 3d 980, 992–993 (2008), and United States v. Peterson, No. 07–30465, 2009 WL 3437834, *1 (Oct. 27, 2009). The writ of certiorari, 561 U. S. ___ (2010), allows this Court to address the conflict. II In determining whether an offense is a violent felony, this Court has explained, “we employ the categorical approach … . Under this approach, we look only to the fact of conviction and the statutory definition of the prior offense, and do not generally consider the particular facts disclosed by the record of conviction. That is, we consider whether the elements of the offense are of the type that would justify its inclusion within the residual provision, without inquiring into the specific conduct of this particular offender.” James, 550 U. S., at 202 (internal quotation marks and citations omitted); see also Taylor v. United States, 495 U. S. 575, 599–602 (1990). So while there may be little doubt that the circumstances of the flight in Sykes’ own case were violent, the question is whether §35–44–3–3 of the Indiana Code, as a categorical matter, is a violent felony. Under 18 U. S. C. §924(e)(2)(B), an offense is deemed a violent felony if it is a crime punishable by more than one year of imprisonment that “(i) has as an element the use, attempted use, or threatened use of physical force against the person of another; or “(ii) is burglary, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.” Resisting law enforcement through felonious vehicle flight does not meet the requirements of clause (i), and it is not among the specific offenses named in clause (ii). Thus, it is violent under this statutory scheme only if it fits within the so-called residual provision of clause (ii). To be a vio-lent crime, it must be an offense that “otherwise in- volves conduct that presents a serious potential risk of physical injury to another.” The question, then, is whether Indiana’s prohibition on flight from an officer by driving a vehicle—the violation of Indiana law for which Sykes sustained his earlier conviction—falls within the residual clause because, as a categorical matter, it presents a serious potential risk of physical injury to another. The offenses enumerated in §924(e)(2)(B)(ii)—burglary, extortion, arson, and crimes involving use of explosives—provide guidance in making this determination. For instance, a crime involves the requisite risk when “the risk posed by [the crime in question] is comparable to that posed by its closest analog among the enumerated offenses.” James, 550 U. S., at 203 (explaining that attempted burglary poses risks akin to that of completed burglary). When a perpetrator defies a law enforcement command by fleeing in a car, the determination to elude capture makes a lack of concern for the safety of property and persons of pedestrians and other drivers an inherent part of the offense. Even if the criminal attempting to elude capture drives without going at full speed or going the wrong way, he creates the possibility that police will, in a legitimate and lawful manner, exceed or almost match his speed or use force to bring him within their custody. A perpetrator’s indifference to these collateral consequences has violent—even lethal—potential for others. A criminal who takes flight and creates a risk of this dimension takes action similar in degree of danger to that involved in arson, which also entails intentional release of a destructive force dangerous to others. This similarity is a beginning point in establishing that vehicle flight presents a serious potential risk of physical injury to another. Another consideration is a comparison to the crime of burglary. Burglary is dangerous because it can end in confrontation leading to violence. Id., at 200. The same is true of vehicle flight, but to an even greater degree. The attempt to elude capture is a direct challenge to an officer’s authority. It is a provocative and dangerous act that dares, and in a typical case requires, the officer to give chase. The felon’s conduct gives the officer reason to believe that the defendant has something more serious than a traffic violation to hide. In Sykes’ case, officers pursued a man with two prior violent felony convictions and marijuana in his possession. In other cases officers may discover more about the violent potential of the fleeing suspect by running a check on the license plate or by recognizing the fugitive as a convicted felon. See, e.g., Arizona v. Gant, 556 U. S. ___, ___ (2009) (slip op., at 2). Because an accepted way to restrain a driver who poses dangers to others is through seizure, officers pursuing fleeing drivers may deem themselves duty bound to escalate their response to ensure the felon is apprehended. Scott v. Harris, 550 U. S. 372, 385 (2007), rejected the possibility that police could eliminate the danger from a vehicle flight by giving up the chase because the perpetrator “might have been just as likely to respond by continuing to drive recklessly as by slowing down and wiping his brow.” And once the pursued vehicle is stopped, it is sometimes necessary for officers to approach with guns drawn to effect arrest. Confrontation with police is the expected result of vehicle flight. It places property and persons at serious risk of injury. Risk of violence is inherent to vehicle flight. Between the confrontations that initiate and terminate the incident, the intervening pursuit creates high risks of crashes. It presents more certain risk as a categorical matter than burglary. It is well known that when offenders use motor vehicles as their means of escape they create serious po-tential risks of physical injury to others. Flight from a law enforcement officer invites, even demands, pursuit. As that pursuit continues, the risk of an accident accumulates. And having chosen to flee, and thereby commit a crime, the perpetrator has all the more reason to seek to avoid capture. Unlike burglaries, vehicle flights from an officer by definitional necessity occur when police are present, are flights in defiance of their instructions, and are effected with a vehicle that can be used in a way to cause serious potential risk of physical injury to another. See post, at 5–6 (Thomas, J., concurring in judgment); see also post, at 6–7 (listing Indiana cases addressing ordinary intentional vehicle flight and noting the high-risk conduct that those convictions involved). Although statistics are not dispositive, here they confirm the commonsense conclusion that Indiana’s vehicular flight crime is a violent felony. See Chambers, 555 U. S., at 129 (explaining that statistical evidence sometimes “helps provide a conclusive … answer” concerning the risks that crimes present). As Justice Thomas explains, chase-related crashes kill more than 100 nonsuspects every year. See post, at 4–5. Injury rates are much higher. Studies show that between 18% and 41% of chases involve crashes, which always carry a risk of injury, and that between 4% and 17% of all chases end in injury. See ibid. A 2008 International Association of Chiefs of Police (IACP) study examined 7,737 police pursuits reported by 56 agencies in 30 States during 2001–2007. C. Lum & G. Fachner, Police Pursuits in an Age of Innovation and Reform 54. Those pursuits, the study found, resulted in 313 injuries to police and bystanders, a rate of slightly over 4 injuries to these nonsuspects per 100 pursuits. Id., at 57. Given that police may be least likely to pursue suspects where the dangers to bystanders are greatest—i.e., when flights occur at extraordinarily high speeds— it is possible that risks associated with vehicle flight are even higher. Those risks may outstrip the dangers of at least two offenses enumerated in 18 U. S. C. §924(e)(2)(B)(ii). According to a study by the Department of Justice, approximately 3.7 million burglaries occurred on average each year in the United States between 2003 and 2007. Bureau of Justice Statistics, S. Catalano, Victimization during Household Burglary 1 (Sept. 2010). Those burglaries resulted in an annual average of approximately 118,000 injuries, or 3.2 injuries for every 100 burglaries. Id., at 9–10. That risk level is 20% lower than that which the IACP found for vehicle pursuits. The U. S. Fire Administration (USFA) maintains the world’s largest data bank on fires. It secures participation from over one-third of U. S. fire departments. It reports an estimated 38,400 arsons in 2008. Those fires resulted in an estimated 1,255 injuries, or 3.3 injuries per 100 arsons. USFA, Methodology Used in the Development of the Topical Fire Research Series, http://www.usfa.dhs.gov/ downloads/pdf/tfrs/methodology.pdf (all Internet materials as visited June 3, 2011, and available in Clerk of Court’s case file); USFA, Nonresidential Building Intentional Fire Trends (Dec. 2010), http://www.usfa.dhs.gov/downloads/ pdf/statistics/nonres_bldg_intentional_fire_trends.pdf; USFA, Residential Building Causes, http://www.usfa.dhs.gov/ downloads/xls/estimates/res_bldg_fire_cause.xlsx; USFA Residential and Nonresidential Fire Estimate Summaries, 2003–2009, http://www.usfa.dhs.gov/statistics/estimates/ index.shtm. That risk level is about 20% lower than that reported by the IACP for vehicle flight. III Sykes argues that, regardless of risk level, typical vehicle flights do not involve the kinds of dangers that the Armed Career Criminal Act’s (ACCA) residual clause demands. In his view this Court’s decisions in Begay and Chambers require ACCA predicates to be purposeful, violent, and aggressive in ways that vehicle flight is not. Sykes, in taking this position, overreads the opinions of this Court. ACCA limits the residual clause to crimes “typically committed by those whom one normally labels ‘armed career criminals,’ ” that is, crimes that “show an increased likelihood that the offender is the kind of person who might deliberately point the gun and pull the trigger.” Begay, 553 U. S., at 146. In general, levels of risk divide crimes that qualify from those that do not. See, e.g., James, 550 U. S. 192 (finding attempted burglary risky enough to qualify). Chambers is no exception. 555 U. S., at ___–___ (slip op., at 5–6) (explaining that failure to report does not qualify because the typical offender is not “significantly more likely than others to attack, or physically to resist, an apprehender”). The sole decision of this Court concerning the reach of ACCA’s residual clause in which risk was not the dis-positive factor is Begay, which held that driving under the influence (DUI) is not an ACCA predicate. There, the Court stated that DUI is not purposeful, violent, and aggressive. 553 U. S., at 145–148. But the Court also gave a more specific reason for its holding. “[T]he conduct for which the drunk driver is convicted (driving under the influence) need not be purposeful or deliberate,” id., at 145 (analogizing DUI to strict-liability, negligence, and recklessness crimes). By contrast, the Indiana statute at issue here has a stringent mens rea requirement. Violators must act “knowingly or intentionally.” Ind. Code §35–44–3–3(a); see Woodward, 770 N. E. 2d, at 901 (construing the statute to require “a knowing attempt to escape law enforcement” (internal quotation marks omitted)). The phrase “purposeful, violent, and aggressive” has no precise textual link to the residual clause, which requires that an ACCA predicate “otherwise involv[e] conduct that presents a serious potential risk of physical injury to another.” §924(e)(2)(B)(ii). The Begay phrase is an addition to the statutory text. In many cases the purposeful, violent, and aggressive inquiry will be redundant with the inquiry into risk, for crimes that fall within the former formulation and those that present serious potential risks of physical injury to others tend to be one and the same. As between the two inquiries, risk levels provide a categorical and manageable standard that suffices to resolve the case before us. Begay involved a crime akin to strict liability, negligence, and recklessness crimes; and the purposeful, violent, and aggressive formulation was used in that case to explain the result. The felony at issue here is not a strict liability, negligence, or recklessness crime and because it is, for the reasons stated and as a categorical matter, similar in risk to the listed crimes, it is a crime that “otherwise involves conduct that presents a serious potential risk of physical injury to another.” §924(e)(2)(B)(ii). IV Sykes finds it significant that his flight conviction was not under the Indiana provision that criminalizes flight in which the offender “operates a vehicle in a manner that creates a substantial risk of bodily injury to another person.” Ind. Code §35–44–3–3(b)(1)(B). In structuring its laws in this way, Sykes contends, Indiana confirmed that it did not intend subsection (b)(1)(A)’s general prohibition on vehicle flight to encompass the particular class of vehicle flights that subsection (b)(1)(B) reaches. Sykes’ argument is unconvincing. Indiana treats violations of subsections (b)(1)(A) and (b)(1)(B) as crimes of the same magnitude. They are both class D felonies, and both carry terms of between six months and three years, Ind. Code §35–50–2–7(a). The distinction between the provisions is their relationship to subsection (a), which prohibits, among other acts, much conduct in which a person “(1) forcibly resists, obstructs, or interferes with a law en-forcement officer … ; (2) forcibly resists, obstructs, or interferes with the authorized service or execution of … process … ; or (3) flees from a law enforcement officer.” §35–44–3–3(a). Subsection (b)(1)(A) only involves the conduct barred by subsection (a)(3)—flight—which, it states, is a felony whenever committed with a vehicle. Under subsection (b)(1)(B), by contrast, any of the offenses in subsection (a) is a felony if the offender commits it while using a vehicle to create a substantial risk of bodily injury to another. Taken together, the statutory incentives always favor prosecuting vehicle flights under subsection (b)(1)(A) rather than subsection (b)(1)(B). They reflect a judgment that some offenses in subsection (a) can be committed without a vehicle or without creating substantial risks. They reflect the further judgment that this is not so for vehicle flights. Serious and substantial risks are an inherent part of vehicle flight. Under subsection (b)(1)(A), they need not be proved separately to secure a conviction equal in magnitude to those available for other forms of resisting law enforcement with a vehicle that involve similar risks. In other words, the “similarity in punishment for these related, overlapping offenses suggests that [subsection (b)(1)(A)] is the rough equivalent of one type of [subsection (b)(1)(B)] violation.” Post, at 10 (Thomas, J., concurring in judgment); see also ibid., n. 2. By adding subsection (b)(1)(A) in 1998, the Indiana Legislature determined that subsection (b)(1)(A) by itself sufficed as a basis for the punishments available under subsection (b)(1)(B). Post, at 10–11; see also post, at 10, n. 2 (identifying reckless endangerment statutes with similar structures); cf. post, at 12 (explaining that because in most cases Indiana does not “specify what additional punishment is warranted when [a] crime kills or injures,” its provisions creating higher penalties for vehicle flights that do so reflect a judgment that these flights are “inherently risky”). The Government would go further and deem it irrelevant under the residual clause whether a crime is a lesser included offense even in cases where that offense carries a less severe penalty than the offense that includes it. As the above discussion indicates, however, the case at hand does not present the occasion to decide that question. V Congress chose to frame ACCA in general and qualitative, rather than encyclopedic, terms. It could have defined violent felonies by compiling a list of specific covered offenses. Under the principle that all are deemed to know the law, every armed felon would then be assumed to know which of his prior felonies could serve to increase his sentence. Given that ACCA “requires judges to make sometimes difficult evaluations of the risks posed by different offenses,” this approach could simplify adjudications for judges in some cases. James, 550 U. S., at 210, n. 6. Congress instead stated a normative principle. The re-sidual clause imposes enhanced punishment for unlaw- ful possession of the firearm when the relevant prior offenses involved a potential risk of physical injury similar to that presented by burglary, extortion, arson, and crimes involving use of explosives. The provision instructs potential recidivists regarding the applicable sentencing regime if they again transgress. It states an intelligible principle and provides guidance that allows a person to “conform his or her conduct to the law.” Chicago v. Morales, 527 U. S. 41, 58 (1999) (plurality opinion). Although this approach may at times be more difficult for courts to implement, it is within congressional power to enact. See James, supra, at 210, n. 6 (giving examples of federal laws similar to ACCA’s residual clause); see also 18 U. S. C. §1031(b)(2) (“conscious or reckless risk of serious personal injury”); §2118(e)(3) (“risk of death, significant physical pain …”); §2246(4) (“substantial risk of death, unconsciousness, extreme physical pain …”); §2258B(b)(2)(B) (2006 ed., Supp. III) (“substantial risk of causing physical injury”); §3286(b) (2006 ed.) (“forseeable risk of … death or seri-ous bodily injury to another person” (footnote omitted)); §4243(d) (“substantial risk of bodily injury to another person”); §§4246(a), (d), (d)(2), (e), (e)(1), (e)(2), (f), (g) (same); §4247(c)(4)(C) (same). VI Felony vehicle flight is a violent felony for purposes of ACCA. The judgment of the Court of Appeals is Affirmed.

564.US.2010_10-313

The Telecommunications Act of 1996 requires incumbent local exchange carriers (LECs)—i.e., providers of local telephone service—to share their physical networks with competitive LECs at cost-based rates in two ways relevant here. First, 47 U. S. C. §251(c)(3) requires an incumbent LEC to lease “on an unbundled basis”—i.e., a la carte—network elements specified by the Federal Communications Commission (FCC) to allow a competitor to create its own network without having to build every element from scratch. In identifying those elements, the FCC must consider whether access is “necessary” and whether failing to provide it would “impair” the competitor’s provision of service. §251(d)(2). Second, §251(c)(2) mandates that incumbent LECs “provide … interconnection” between their networks and competitive LECs’ to ensure that a competitor’s customers can call the incumbent’s customers, and vice versa. The interconnection duty is independent of the unbundling rules and not subject to impairment analysis. In 2003, the FCC issued its Triennial Review Order deciding, contrary to previous orders, that §251(c)(3) did not require an incumbent LEC to provide a competitive LEC with cost-based unbundled access to existing “entrance facilities”—i.e., transmission facilities (typically wires or cables) that connect the two LECs’ networks—because such facilities are not network elements at all. The FCC noted, however, that entrance facilities are used for both interconnection and backhauling, and it emphasized that its order did not alter incumbent LECs’ §251(c)(2) obligation to provide for interconnection. Thus, the practical effect of the order was only that incumbent LECs were not obligated to unbundle entrance facilities for backhauling purposes. In 2005, following D. C. Circuit review, the FCC issued its Triennial Review Remand Order. The FCC retreated from the view that entrance facilities are not network elements, but adhered to its previous position that cost-based unbundled access to such facilities need not be provided under §251(c)(3). Treating entrance facilities as network elements, the FCC concluded that competitive LECs are not impaired without access to such facilities. The FCC again emphasized that competitive LECs’ §251(c)(2) right to obtain interconnection had not been altered. In the Remand Order’s wake, respondent AT&T notified competitive LECs that it would no longer provide entrance facilities at cost-based rates for either backhauling or interconnection, but would instead charge higher rates. Competitive LECs complained to the Michigan Public Service Commission that AT&T was unlawfully abrogating their §251(c)(2) right to cost-based interconnection. The Michigan Public Service Commission agreed and ordered AT&T to continue providing entrance facilities for interconnection at cost-based rates. AT&T challenged the ruling. Relying on the Remand Order, the Federal District Court ruled in AT&T’s favor. The Sixth Circuit affirmed, declining to defer to the FCC’s argument that the order did not change incumbent LECs’ interconnection obligations, including the obligation to lease entrance facilities for interconnection. Held: The FCC has advanced a reasonable interpretation of its regulations—i.e., that to satisfy its duty under §251(c)(2), an incumbent LEC must make its existing entrance facilities available to competitors at cost-based rates if the facilities are to be used for interconnection—and this Court defers to the FCC’s views. Pp. 6–16. (a) No statute or regulation squarely addresses the question. Pp. 6–7. (b) Absent an unambiguous statute or regulation, the Court turns to the FCC’s interpretation of its regulations in its amicus brief. See, e.g., Chase Bank USA, N. A. v. McCoy, 562 U. S. ___, ___. The FCC proffers a three-step argument why its regulations require AT&T to provide access at cost-based rates to existing entrance facilities for interconnection purposes. Pp. 7–10. (1) Interpreting 47 CFR §51.321(a), the FCC first contends that an incumbent LEC must lease “technically feasible” facilities for interconnection. Pp. 8–9. (2) The FCC contends, second, that existing entrance facilities are part of an incumbent LEC’s network, 47 CFR §51.319(e), and therefore are among the facilities that an incumbent LEC must lease for interconnection, if technically feasible. P. 9. (3) Third, says the FCC, it is technically feasible to provide access to the particular entrance facilities at issue in these cases—a point AT&T does not dispute. P. 10. (c) Contrary to AT&T’s arguments, the FCC’s interpretation is not “plainly erroneous or inconsistent with the regulation[s]. ” Auer v. Robbins, 519 U. S. 452, 461. First, it is perfectly sensible to read the FCC’s regulations to include entrance facilities as part of incumbent LECs’ networks. Second, the FCC’s views do not conflict with 47 CFR §51.5’s definition of interconnection as “the linking of two networks for the mutual exchange of traffic[, but not] the transport and termination of traffic.” Pp. 10–12. (d) Nor is there any other “reason to suspect that the [FCC’s] in-terpretation does not reflect the agency’s fair and considered judgment on the matter in question.” Auer, supra, at 462. AT&T incorrectly suggests that the FCC is attempting to require under §251(c)(2) what courts have prevented it from requiring under §251(c)(3) and what the FCC itself said was not required in the Remand Order. Pp. 12–16. 597 F. 3d 370, reversed. Thomas, J., delivered the opinion of the Court, in which all other Members joined, except Kagan, J., who took no part in the consideration or decision of the cases. Scalia, J., filed a concurring opinion. Together with No. 10–329, Isiogu et al. v. Michigan Bell Telephone Co. dba AT&T Michigan, also on certiorari to the same court.

In these cases, we consider whether an incumbent provider of local telephone service must make certain transmission facilities available to competitors at cost-based rates. The Federal Communications Commission (FCC or Commission) as amicus curiae[Footnote 1] contends that its regulations require the incumbent provider to do so if the facilities are to be used for interconnection: to link the incumbent provider’s telephone network with the competitor’s network for the mutual exchange of traffic. We defer to the Commission’s views and reverse the judgment below. I The Telecommunications Act of 1996 (1996 Act), 110 Stat. 56, imposed a number of duties on incumbent providers of local telephone service in order to facilitate market entry by competitors. AT&T Corp. v. Iowa Utilities Bd., 525 U. S. 366, 371 (1999). The incumbent local ex-change carriers (LECs) owned the local exchange networks: the physical equipment necessary to receive, properly route, and deliver phone calls among customers. Verizon Communications Inc. v. FCC, 535 U. S. 467, 490 (2002). Before the 1996 Act, a new, competitive LEC could not compete with an incumbent carrier without basically replicating the incumbent’s entire existing network. Ibid. The 1996 Act addressed that barrier to market entry by requiring incumbent LECs to share their networks with competitive LECs in several ways, two of which are relevant here. First, 47 U. S. C. §251(c)(3) requires incumbent LECs to lease “on an unbundled basis”—i.e., a la carte—network elements specified by the Commission. This makes it easier for a competitor to create its own network without having to build every element from scratch. In identifying which network elements must be available for unbundled lease under §251(c)(3), the Commission is required to consider whether access is “necessary” and whether failing to provide access would “impair” a competitor’s provision of service. §251(d)(2). Second, §251(c)(2) mandates that incumbent LECs “provide … interconnection” between their networks and competitive LECs’ facilities. This ensures that customers on a competitor’s network can call customers on the incumbent’s network, and vice versa. The interconnection duty is independent of the unbundling rules and not subject to impairment analysis. It is undisputed that both un-bundled network elements and interconnection must be provided at cost-based rates. See §252(d)(1); Brief for Petitioner in No. 10–313, p. 28; Brief for Petitioners in No. 10–329, p. 7; Brief for Respondent 4. These cases concern incumbent LECs’ obligation to share existing “entrance facilities” with competitive LECs. Entrance facilities are the transmission facilities (typically wires or cables) that connect competitive LECs’ networks with incumbent LECs’ networks. The FCC recently adopted a regulation specifying that entrance facilities are not among the network elements that §251(c)(3) requires incumbents to lease to competitors on an unbundled basis at cost-based rates. See 47 CFR §51.319(e)(2)(i) (2005). The Commission noted, however, that it “d[id] not alter the right of competitive LECs to obtain interconnection facilities pursuant to section 251(c)(2).” In re Unbundled Access to Network Elements, 20 FCC Rcd. 2533, 2611, ¶140 (2005) (Triennial Review Remand Order). The specific issue here is whether respondent, Michigan Bell Telephone Company d/b/a AT&T Michigan (AT&T), must lease existing entrance facilities to competitive LECs at cost-based rates. The FCC interprets its regulations to require AT&T to do so for the purpose of interconnection. We begin by reviewing the Commission’s recent actions regarding entrance facilities and then explain the particular dispute that is before us today. A In 2003, the FCC decided, contrary to its previous orders, that incumbent LECs were not obligated to provide cost-based unbundled access to entrance facilities under §251(c)(3). In re Review of Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers, 18 FCC Rcd. 16978, 17202–17205, ¶¶365–367 (2003) (Triennial Review Order). Explaining that its previous approach had been “misguided” and “overly broad,” id., ¶¶366, 365, the Commission concluded that entrance facilities were not subject to the unbundling requirement because they are not network elements at all. See id., ¶366 (entrance facilities “exist outside the incumbent LEC’s local network”). The Commission therefore did not conduct an impairment analysis. The FCC emphasized, however, the limits of this ruling. Entrance facilities are used for two purposes: interconnection and backhauling.[Footnote 2] It expressly “d[id] not alter” an incumbent LEC’s obligation under §251(c)(2) to provide “facilities in order to ‘interconnect with the incumbent LEC’s network.’ ” Id., ¶366 (brackets omitted). Thus, al-though the Commission specified that §251(c)(3) did not require any unbundled leasing of entrance facilities, it determined in practical effect only that “incumbent LECs [were not obligated] to unbundle [entrance facilities] for the purpose of backhauling traffic.” Id., ¶365. On direct review, the D. C. Circuit questioned the Commission’s determination that entrance facilities are not network elements under §251(c)(3), but found the agency rulemaking record insufficient and remanded to the Commission for further consideration. See United States Telecom Assn. v. FCC, 359 F. 3d 554, 586, cert. denied, 543 U. S. 925 (2004). The court noted that if entrance facilities were in fact “ ‘network elements,’ ” then “an analysis of impairment would presumably follow.” 359 F. 3d, at 586. In 2005, the Commission responded. See Triennial Review Remand Order ¶¶136–141. The Commission re-treated from its view that entrance facilities are not network elements but adhered to its previous position that cost-based unbundled access to them need not be provided under §251(c)(3). Id., ¶¶137–138. Treating entrance facilities as network elements, the Commission concluded that competitive LECs are not impaired without access to them. Ibid. The Commission again emphasized that it “d[id] not alter the right of competitive LECs to obtain interconnection facilities pursuant to section 251(c)(2).” Id., ¶140. B In the wake of the Triennial Review Remand Order, AT&T notified competitive LECs that it would no longer provide entrance facilities at cost-based rates for either backhauling or interconnection, but would instead charge higher rates. Competitive LECs complained to the Michigan Public Service Commission (PSC) that AT&T was unlawfully abrogating their right to cost-based interconnection under §251(c)(2). The Michigan PSC agreed with the competitive LECs and ordered AT&T to continue providing entrance facilities for interconnection at cost-based rates. AT&T challenged the Michigan PSC’s ruling in the District Court, which, relying on the Triennial Review Remand Order, ruled in AT&T’s favor. The Michigan PSC and several competitive LECs, including petitioner Talk America, Inc., appealed. The Court of Appeals for the Sixth Circuit affirmed over a dissent. Michigan Bell Telephone Co. v. Covad Communications Co., 597 F. 3d 370 (2010). At the court’s invitation, the FCC filed a brief as amicus curiae, arguing that the Triennial Review Remand Order did not change incumbent LECs’ interconnection obligations, including the obligation to lease entrance facilities for interconnection. The Sixth Circuit declined to defer to the FCC’s views, 597 F. 3d, at 375, n. 6, and also expressly disagreed with the Seventh and Eighth Circuits, id., at 384–386 (discussing Illinois Bell Tel. Co. v. Box, 526 F. 3d 1069 (2008), and Southwestern Bell Tel., L. P. v. Missouri Pub. Serv. Comm’n, 530 F. 3d 676 (2008)).[Footnote 3] We granted certiorari, 562 U. S. ___ (2010), and now reverse. II Petitioners contend that AT&T must lease its existing entrance facilities for interconnection at cost-based rates. We agree. A No statute or regulation squarely addresses whether an incumbent LEC must provide access to entrance facilities at cost-based rates as part of its interconnection duty under §251(c)(2). According to the statute, each incumbent LEC has: “The duty to provide, for the facilities and equipment of any requesting telecommunications carrier, interconnection with the local exchange carrier’s network— “(A) for the transmission and routing of telephone exchange service and exchange access; “(B) at any technically feasible point within the carrier’s network; “(C) that is at least equal in quality to that provided by the local exchange carrier to itself or to any subsidiary, affiliate, or any other party to which the carrier provides interconnection; and “(D) on rates, terms, and conditions that are just, reasonable, and nondiscriminatory, in accordance with the terms and conditions of the agreement and the requirements of this section and section 252 of this title.” Nothing in that language expressly addresses entrance facilities. Nor does any regulation do so. See Brief for United States as Amicus Curiae 22, n. 6. AT&T contends that the statute makes clear that an incumbent LEC need not provide access to any facilities—much less entrance facilities—to provide interconnection. The company points out that §251(c)(2) does not mention incumbent LECs’ facilities, but rather mandates only that incumbent LECs provide interconnection “for the facilities and equipment of any [competing] carrier.” In contrast, AT&T notes, §251(c)(3) requires that incumbent LECs provide unbundled “access to [their] network elements.” We do not find the statute so clear. Although §251(c)(2) does not expressly require that incumbent LECs lease facilities to provide interconnection, it also does not expressly excuse them from doing so. The statute says nothing about what an incumbent LEC must do to “provide … interconnection.” §251(c)(2). “[T]he facilities and equipment of any [competing] carrier” identifies the equipment that an incumbent LEC must allow to interconnect, but it does not specify what the incumbent LEC must do to make the interconnection possible. Ibid. B In the absence of any unambiguous statute or regulation, we turn to the FCC’s interpretation of its regulations in its amicus brief. See, e.g., Chase Bank USA, N. A. v. McCoy, 562 U. S. ___, ___ (2011) (slip op., at 12). As we reaffirmed earlier this Term, we defer to an agency’s interpretation of its regulations, even in a legal brief, unless the interpretation is “ ‘plainly erroneous or inconsistent with the regulation[s]’ ” or there is any other “ ‘reason to suspect that the interpretation does not reflect the agency’s fair and considered judgment on the matter in question.’ ” Id., at ___, ___ (slip op., at 12, 14) (quoting Auer v. Robbins, 519 U. S. 452, 461, 462 (1997)). The Commission contends that its regulations require AT&T to provide access at cost-based rates to its exist- ing entrance facilities for the purpose of interconnection. The Commission’s interpretation proceeds in three steps. First, an incumbent LEC must lease “technically feasible” facilities for interconnection. Second, entrance facili- ties are among the facilities that an incumbent must make available for interconnection, if technically feasible. Third, it is technically feasible to provide access to the particular entrance facilities at issue in these cases. 1 The Commission first contends that an incumbent LEC must lease, at cost-based rates, any requested facilities for obtaining interconnection with the incumbent LEC’s network, unless it is technically infeasible to do so. Section 251(c)(2) mandates that an incumbent LEC provide interconnection, at cost-based rates, “at any technically feasible point within the carrier’s network.” The FCC has long construed §251(c)(2) to require incumbent LECs to provide, at cost-based rates, “any technically feasible method of obtaining interconnection … at a particular point.” 47 CFR §51.321(a) (2010). The requirement in §51.321(a) to provide a “method of obtaining interconnection,” the Commission argues, encompasses a duty to lease an existing facility to a competing LEC. When the Commission originally promulgated §51.321(a), it explained that incumbent LECs would be required to “adapt their facilities to interconnection” and to “accept the novel use of, and modification to, [their] network facilities.” In re Implementation of Local Competition Provisions in the Telecommunications Act of 1996, 11 FCC Rcd. 15499, 15605, ¶202 (1996) (Local Competition Order). Since then, as AT&T and its amici concede, incumbent LECs have commonly leased certain facilities at cost-based prices to accommodate interconnection. See Brief for Respondent 28–29; Brief for United States Telecom Association et al. as Amici Curiae 33–35. As additional support for its assertion that incumbent LECs are obligated to lease facilities, the FCC highlights the examples in §51.321(b) of “[t]echnically feasible methods of obtaining interconnection,” which include “[m]eet point interconnection arrangements.” In a meet-point arrangement, an incumbent LEC “accommodat[es]” interconnection by building a transmission facility from its network to a designated point, where it connects with the competitor’s corresponding transmission facility. Local Competition Order ¶553. Compared to that requirement, the Commission argues, the obligation to lease existing facilities for interconnection is quite modest. 2 Next, the Commission contends that existing entrance facilities are among the facilities that an incumbent LEC must lease for interconnection. According to the FCC, the Triennial Review Remand Order adopted a regulatory def-inition that reestablished that entrance facilities are part of an incumbent LEC’s network. See ¶137; see also 47 CFR §51.319(e) (2005). The end of every entrance facility is therefore a “point within [an incumbent] carrier’s network” at which a competing LEC could request interconnection, 47 U. S. C. §251(c)(2), and each entrance facility potentially provides a “technically feasible method of obtaining interconnection,” 47 CFR §51.321(a) (2010). 3 Finally, the FCC contends that providing access to the entrance facilities here for interconnection purposes is technically feasible. Under the Commission’s regulations, an incumbent LEC bears the burden of showing that a requested method or point of interconnection is technically infeasible. See 47 CFR §§51.305(e), 51.321(d); see also §§51.305(d), 51.321(c) (previously successful intercon-nection is “substantial evidence” of technical feasibility). AT&T does not dispute technical feasibility here.[Footnote 4] C The FCC’s interpretation is not “plainly erroneous or inconsistent with the regulation[s]. ” Auer, supra, at 461 (internal quotation marks omitted). First, we disagree with AT&T’s argument that entrance facilities are not a part of incumbent LECs’ networks. Indeed, the Commission’s view on this question is more than reasonable; it is certainly not plainly erroneous. The Triennial Review Remand Order responded to the D. C. Circuit’s decision questioning the Commission’s earlier finding that entrance facilities are not network elements. It revised the definition of dedicated transport—a type of network element—to include entrance facilities. Triennial Review Remand Order ¶¶136–137; see 47 CFR §51.319(e)(1) (defining dedicated transport to include “incumbent LEC transmission facilities … between wire centers or switches owned by incumbent LECs and switches owned by [competing] carriers”). Given that revised definition, it is perfectly sensible to conclude that entrance facilities are a part of incumbent LECs’ networks. Second, we are not persuaded by AT&T’s argument that the Commission’s views conflict with the definition of interconnection in §51.5. That regulation provides: “Interconnection is the linking of two networks for the mutual exchange of traffic. This term does not include the transport and termination of traffic.” AT&T focuses on the definition’s exclusion of “transport and termination of traffic.” An entrance facility is a transport facility, AT&T argues, and it makes no sense to require an incumbent LEC to furnish a transport facility for interconnection when the definition of interconnection expressly excludes transport. We think AT&T reads too much into the exclusion of “transport.” The regulation cannot possibly mean that no transport can occur across an interconnection facility, as that would directly conflict with the statutory language. See §251(c)(2) (requiring “interconnection … for the transmission and routing of [local] telephone exchange service”). The very reason for interconnection is the “mutual exchange of traffic.” 47 CFR §51.5; see also Competitive Telecommunications Assn. v. FCC, 117 F. 3d 1068, 1071–1072 (CA8 1997) (“[T]he transmission and routing of telephone exchange service” is “what the interconnection, the physical link, would be used for” (internal quotation marks omitted)). The better reading of the regulation is that it merely reflects that the “transport and termination of traffic” is subject to different regulatory treatment than interconnection. Compensation for transport and termination—that is, for delivering local telephone calls placed by another carrier’s customer—is governed by separate stat-utory provisions and regulations. See 47 U. S. C. §§251(b)(5), 252(d)(2); 47 CFR §51.701. The Commission explains that a competitive LEC typically pays one fee for interconnection—“just for having the link”—and then an additional fee for the transport and termination of telephone calls. Tr. of Oral Arg. 28; see also Brief for United States as Amicus Curiae 3, n. 1. Entrance facilities, at least when used for the mutual exchange of traffic, seem to us to fall comfortably within the definition of interconnection. See 597 F. 3d, at 388 (Sutton, J., dissenting) (noting that entrance facilities are “designed for the very purpose of linking two carriers’ networks” (internal quotation marks omitted)). In sum, the Commission’s interpretation of its regulations is neither plainly erroneous nor inconsistent with the regulatory text. Contrary to AT&T’s assertion, there is no danger that deferring to the Commission would effectively “permit the agency, under the guise of interpreting a regulation, to create de facto a new regulation.”[Footnote 5] Christensen v. Harris County, 529 U. S. 576, 588 (2000). D Nor is there any other “reason to suspect that the interpretation does not reflect the agency’s fair and considered judgment on the matter in question.” Auer, 519 U. S., at 462. We are not faced with a post-hoc rationalization by Commission counsel of agency action that is under judicial review. See ibid.; see also Burlington Truck Lines, Inc. v. United States, 371 U. S. 156, 168–169 (1962) (“The courts may not accept appellate counsel’s post hoc rationalizations for agency action; [SEC v.] Chenery[ Corp., 332 U. S. 194 (1947),] requires that an agency’s discretionary order be upheld, if at all, on the same basis articulated in the order by the agency itself”). And although the FCC concedes that it is advancing a novel interpretation of its longstanding interconnection regulations, novelty alone is not a reason to refuse deference. The Commission ex-plains that the issue in these cases did not arise until recently—when it initially eliminated unbundled access to entrance facilities in the Triennial Review Order. Until then, the Commission says, a competitive LEC typically would elect to lease a cost-priced entrance facility under §251(c)(3) since entrance facilities leased under §251(c)(3) could be used for any purpose—i.e., both interconnection and backhauling—but entrance facilities leased under §251(c)(2) can be used only for interconnection. We see no reason to doubt this explanation. AT&T suggests that the Commission is attempting to require under §251(c)(2) what courts have prevented it from requiring under §251(c)(3) and what the Commission itself said was not required in the Triennial Review Remand Order. Tr. of Oral Arg. 50 (“[T]his is a rear guard effort to preserve [cost-based] pricing for things that the [C]ommission has said should no longer be available … at [such] pricing”). We do not think that AT&T is correct. 1 To begin with, AT&T’s accusation does not square with the regulatory history. The Commission was not compelled to eliminate the obligation to lease unbundled entrance facilities at cost-based rates. It is true that, prior to the Triennial Review orders, the Commission twice unsuccessfully attempted to impose sweeping unbundling requirements on incumbent LECs. See Local Competition Order ¶278; In re Implementation of Local Competition Provisions of the Telecommunications Act of 1996, 15 FCC Rcd. 3696, 3771–3904, ¶¶162–464 (1999); see also 47 CFR §51.319 (1997); §51.319 (2000). Each time, the Commission’s efforts were rejected for taking an unreasonably broad view of “impair[ment]” under §251(d)(2). See Iowa Utilities Bd., 525 U. S., at 392; United States Telecom Assn. v. FCC, 290 F. 3d 415, 421–428 (2002), cert. denied, 538 U. S. 940 (2003). In the Triennial Review Order, the Commission once again reinterpreted the “impair” standard and revised the list of network elements that incumbents must provide unbundled to competitors. The Commission’s initial decision to eliminate the obligation to unbundle entrance facilities, however, was not a result of the narrower view of impairment mandated by this Court and the D. C. Circuit. Instead, the Commission determined that entrance facilities need not be provided on an unbundled basis under §251(c)(3) on the novel ground that they are not network elements at all—something no court had ever suggested. Moreover, since its initial decision to eliminate the unbundling obligation for entrance facilities, the Commission has been committed to that position. When the D. C. Circuit questioned the Commission’s finding that entrance facilities are not network elements, the Commission responded by observing that the court “did not reject our conclusion that incumbent LECs need not unbundle entrance facilities, only the analysis through which we reached that conclusion.” Triennial Review Remand Order ¶137. The Commission then found another way to support that same conclusion. 2 More importantly, AT&T’s characterization of what the Commission has done, and is doing, is inaccurate. The Triennial Review orders eliminated incumbent LECs’ obli-gation under §251(c)(3) to provide unbundled access to entrance facilities. But the FCC emphasized in both orders that it “d[id] not alter” the obligation on incumbent LECs under §251(c)(2) to provide facilities for interconnection purposes. Triennial Review Order ¶366; Triennial Review Remand Order ¶140. Because entrance facilities are used for backhauling and interconnection purposes, the FCC effectively eliminated only unbundled access to entrance facilities for backhauling purposes—a nuance it expressly noted in the first Triennial Review order. Triennial Review Order ¶365. That distinction is neither unusual nor ambiguous.[Footnote 6] In these cases, the Commission is simply explaining the interconnection obligation that it left undisturbed in the Triennial Review orders. We see no conflict between the Triennial Review orders and the Commission’s views expressed here.[Footnote 7] We are not concerned that the Triennial Review Remand Order did not expressly distinguish between backhauling and interconnection, though AT&T makes much of that fact. AT&T argues that the Commission’s holding in the Triennial Review Remand Order is broader than that in the Triennial Review Order. In AT&T’s view, the Commission concluded in the Triennial Review Remand Order that competitors are not impaired if they lack cost-based access to entrance facilities for backhauling or interconnection. There are two flaws with AT&T’s reasoning. First, as we have discussed, the Triennial Review Remand Order reinstated the ultimate conclusion of the Triennial Review Order and changed only “the analysis through which [it] reached that conclusion.” Triennial Review Remand Order ¶137. Second, unlike §251(c)(3)’s unbundling obligation, §251(c)(2)’s interconnection obligation does not require the Commission to consider impairment. As the dissent below observed, it would be surprising indeed if the FCC had taken the novel step of incorporating impairment into interconnection without comment. 597 F. 3d, at 389 (opinion of Sutton, J.). * * * The FCC as amicus curiae has advanced a reasonable interpretation of its regulations, and we defer to its views. The judgment of the United States Court of Appeals for the Sixth Circuit is reversed. It is so ordered. Justice Kagan took no part in the consideration or decision of these cases. Footnote 1 The Solicitor General, joined by counsel for the FCC, represents that the amicus brief for the United States filed in this Court reflects the Commission’s considered interpretation of its own rules and orders. Brief for United States as Amicus Curiae 31. We thus refer to the Government’s arguments in these cases as those of the agency. See, e.g., Chase Bank USA, N. A. v. McCoy, 562 U. S. ___, ___ (2011) (slip op., at 8). Footnote 2 Although the parties and their amici disagree over the precise definition of backhauling, they all appear to agree that backhauling is important to competitive LECs and occurs when a competitive LEC uses an entrance facility to transport traffic from a leased portion of an incumbent network to the competitor’s own facilities. Backhauling does not involve the exchange of traffic between incumbent and competitive networks. See, e.g., Brief for Petitioners in No. 10–329, p. 25; Brief for United States Telecom Association et al. as Amici Curiae 32. It thus differs from interconnection—“the linking of two networks for the mutual exchange of traffic.” 47 CFR §51.5 (2010). Footnote 3 The Ninth Circuit has since joined the Seventh and Eighth Circuits. Pacific Bell Tel. Co. v. California Pub. Util. Comm’n, 621 F. 3d 836 (2010). Footnote 4 These cases concern only existing entrance facilities, and the Commission expressly declines to address whether it reads its regulations to require incumbent LECs to build new entrance facilities for interconnection. Brief for United States as Amicus Curiae 25, n. 7. The Commission suggests here, as it has before, that additional considerations of cost or reasonableness might be appropriate if a competitive LEC were to request that an incumbent LEC build new entrance facilities for interconnection. Ibid. (noting that the Commission’s Wireline Competition Bureau has declined to require an incumbent LEC to bear the entire cost of building new entrance facilities); see also Local Competition Order ¶553 (explaining with respect to meet-point arrangements that “the parties and state commissions are in a better position than the Commission to determine the appropriate distance that would constitute the required reasonable accommodation of interconnection”). We express no view on the matter. Footnote 5 There is no merit to AT&T’s assertion that the FCC is improperly amending the list of “[t]echnically feasible methods of obtaining interconnection” set forth in 47 CFR §51.321(b). By its own terms, that list is nonexhaustive. See §51.321(b) (“[t]echnically feasible methods of obtaining interconnection … include, but are not limited to” the listed examples); see also §51.321(a) (“[A]n incumbent LEC shall provide … any technically feasible method of obtaining interconnection” (emphasis added)). Footnote 6 The Commission has long recognized that a single facility can be used for different functions and that its regulatory treatment can vary depending on its use. Unbundled network elements, for example, may not be used for the exclusive provision of mobile wireless or long-distance services. 47 CFR §51.309(b) (2010). Similarly, interconnection arrangements may be used for local telephone service but not for long-distance services. §51.305(b). Footnote 7 The parties and their amici dispute whether an incumbent LEC has any way of knowing how a competitive LEC is using an entrance facility. This technical factual dispute simply underscores the appropriateness of deferring to the FCC. So long as the Commission is acting within the scope of its delegated authority and in accordance with prescribed procedures, it has greater expertise and stands in a better position than this Court to make the technical and policy judgments necessary to administer the complex regulatory program at issue here.

564.US.2010_10-5400

Petitioner Tapia was convicted of, inter alia, smuggling unauthorized aliens into the United States. The District Court imposed a 51-month prison term, reasoning that Tapia should serve that long in order to qualify for and complete the Bureau of Prisons’ Residential Drug Abuse Program (RDAP). On appeal, Tapia argued that lengthening her prison term to make her eligible for RDAP violated 18 U. S. C. §3582(a), which instructs sentencing courts to “recogniz[e] that imprisonment is not an appropriate means of promoting correction and rehabilitation.” The Ninth Circuit disagreed. Relying on Circuit precedent, it held that a sentencing court cannot impose a prison term to assist a defendant’s rehabilitation, but once imprisonment is chosen, the court may consider the defendant’s rehabilitation needs in setting the sentence’s length. Held: Section 3582(a) does not permit a sentencing court to impose or lengthen a prison term in order to foster a defendant’s rehabilitation. Pp. 3–15. (a) For nearly a century, the Federal Government used an indeterminate sentencing system premised on faith in rehabilitation. Mistretta v. United States, 488 U. S. 361, 363. Because that system produced “serious disparities in [the] sentences” imposed on similarly situated defendants, id., at 365, and failed to “achieve rehabilitation,” id., at 366, Congress enacted the Sentencing Reform Act of 1984 (SRA), replacing the system with one in which Sentencing Guidelines would provide courts with “a range of determinate sentences,” id., at 368. Under the SRA, a sentencing judge must impose at least imprisonment, probation, or a fine. See §3551(b). In determining the appropriate sentence, judges must consider retribution, deterrence, incapacitation, and rehabilitation, §3553(a)(2), but a particular purpose may apply differently, or not at all, depending on the kind of sentence under consideration. As relevant here, a court ordering imprisonment must “recogniz[e] that imprisonment is not an appropriate means of promoting correction and rehabilitation.” §3582(a). A similar provision instructs the Sentencing Commission, as the Sentencing Guidelines’ author, to “insure that the guidelines reflect the inappropriateness of imposing a sentence to a term of imprisonment for the purpose of rehabilitating the defendant.” 28 U. S. C. §994(k). Pp. 3–6. (b) Consideration of Tapia’s claim starts with §3582(a)’s clear text. Putting together the most natural definitions of “recognize”—“to acknowledge or treat as valid”—and not “appropriate”—not “suitable or fitting for a particular purpose”—§3582(a) tells courts to acknowledge that imprisonment is not suitable for the purpose of promoting rehabilitation. It also instructs courts to make that acknowledgment when “determining whether to impose a term of imprisonment, and … [when] determining the length of the term.” Amicus, appointed to defend the judgment below, argues that the “recognizing” clause is merely a caution for judges not to put too much faith in the capacity of prisons to rehabilitate. But his alternative interpretation is unpersuasive, as Congress expressed itself clearly in §3582(a). Amicus also errs in echoing the Ninth Circuit’s reasoning that §3582’s term “imprisonment” relates to the decision whether to incarcerate, not the determination of the sentence’s length. Because “imprisonment” most naturally means “the state of being confined” or “a period of confinement,” it does not distinguish between the defendant’s initial placement behind bars and his continued stay there. Section 3582(a)’s context supports this textual conclusion. By restating §3582(a)’s message to the Sentencing Commission, Congress ensured that all sentencing officials would work in tandem to implement the statutory determination to “reject imprisonment as a means of promoting rehabilitation.” Mistretta, 488 U. S., at 367. Equally illuminating is the absence of any provision authorizing courts to ensure that offenders participate in prison rehabilitation programs. When Congress wanted sentencing courts to take account of rehabilitative needs, it gave them authority to do so. See, e.g., §3563(b)(9). In fact, although a sentencing court can recommend that an offender be placed in a particular facility or program, see §3582(a), the authority to make the placement rests with the Bureau of Prisons, see, e.g., §3621(e). The point is well illustrated here, where the District Court’s strong recommendations that Tapia participate in RDAP and be placed in a particular facility went unfulfilled. Finally, for those who consider legislative history useful, the key Senate Report on the SRA provides corroborating evidence. Pp. 6–12. (c) Amicus’ attempts to recast what the SRA says about rehabilitation are unavailing. Pp. 12–14. (d) Here, the sentencing transcript suggests that Tapia’s sentence may have been lengthened in light of her rehabilitative needs. A court does not err by discussing the opportunities for rehabilitation within prison or the benefits of specific treatment or training programs. But the record indicates that the District Court may have increased the length of Tapia’s sentence to ensure her completion of RDAP, something a court may not do. The Ninth Circuit is left to consider on remand the effect of Tapia’s failure to object to the sentence when imposed. Pp. 14–15. 376 Fed. Appx. 707, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court. Sotomayor, J., filed a concurring opinion, in which Alito, J., joined.

We consider here whether the Sentencing Reform Act pre-cludes federal courts from imposing or lengthening a prison term in order to promote a criminal defendant’s rehabilitation. We hold that it does. I Petitioner Alejandra Tapia was convicted of, inter alia, smuggling unauthorized aliens into the United States, in violation of 8 U. S. C. §§1324(a)(2)(B)(ii) and (iii). At sen-tencing, the District Court determined that the United States Sentencing Guidelines recommended a prison term of between 41 and 51 months for Tapia’s offenses. The court decided to impose a 51-month term, followed by three years of supervised release. In explaining its reasons, the court referred several times to Tapia’s need for drug treatment, citing in particular the Bureau of Prison’s Residential Drug Abuse Program (known as RDAP or the 500 Hour Drug Program). The court indicated that Tapia should serve a prison term long enough to qualify for and complete that program: “The sentence has to be sufficient to provide needed correctional treatment, and here I think the needed correctional treatment is the 500 Hour Drug Program. . . . . . “Here I have to say that one of the factors that—I am going to impose a 51-month sentence, … and one of the factors that affects this is the need to provide treatment. In other words, so she is in long enough to get the 500 Hour Drug Program, number one.” App. 27. (“Number two” was “to deter her from committing other criminal offenses.” Ibid.) The court “strongly recommend[ed]” to the Bureau of Prisons (BOP) that Tapia “participate in [RDAP] and that she serve her sentence at” the Federal Correctional Institution in Dublin, California (FCI Dublin), where “they have the appropriate tools … to help her, to start to make a recovery.” Id., at 29. Tapia did not object to the sentence at that time. Id., at 31. On appeal, however, Tapia argued that the District Court had erred in lengthening her prison term to make her eligible for RDAP. App. to Pet. for Cert. 2. In Tapia’s view, this action violated 18 U. S. C. §3582(a), which instructs sentencing courts to “recogniz[e] that imprisonment is not an appropriate means of promoting correction and rehabilitation.” The United States Court of Appeals for the Ninth Circuit disagreed, 376 Fed. Appx. 707 (2010), relying on its prior decision in United States v. Duran, 37 F. 3d 557 (1994). The Ninth Circuit had held there that §3582(a) distinguishes between deciding to impose a term of imprisonment and determining its length. See id., at 561. According to Duran, a sentencing court cannot impose a prison term to assist a defendant’s rehabilitation. But “[o]nce imprisonment is chosen as a punishment,” the court may consider the defendant’s need for rehabilitation in setting the length of the sentence. Ibid. We granted certiorari to consider whether §3582(a) permits a sentencing court to impose or lengthen a prison term in order to foster a defendant’s rehabilitation. 562 U. S. ___ (2010). That question has divided the Courts of Appeals.[Footnote 1] Because the United States agrees with Tapia’s interpretation of the statute, we appointed an amicus curiae to defend the judgment below.[Footnote 2] We now reverse. II We begin with statutory background—how the relevant sentencing provisions came about and what they say. Aficionados of our sentencing decisions will recognize much of the story line. “For almost a century, the Federal Government employed in criminal cases a system of indeterminate sentencing.” Mistretta v. United States, 488 U. S. 361, 363 (1989). Within “customarily wide” outer boundaries set by Congress, trial judges exercised “almost unfettered discretion” to select prison sentences for federal offenders. Id., at 364. In the usual case, a judge also could reject prison time altogether, by imposing a “suspended” sentence. If the judge decided to impose a prison term, discretionary authority shifted to parole officials: Once the defendant had spent a third of his term behind bars, they could order his release. See K. Stith & J. Cabranes, Fear of Judging: Sentencing Guidelines in the Federal Courts 18–20 (1998). This system was premised on a faith in rehabilitation. Discretion allowed “the judge and the parole officer to [base] their respective sentencing and release decisions upon their own assessments of the offender’s amenability to rehabilitation.” Mistretta, 488 U. S., at 363. A convict, the theory went, should generally remain in prison only until he was able to reenter society safely. His release therefore often coincided with “the successful completion of certain vocational, educational, and counseling programs within the prisons.” S. Rep. No. 98–225, p. 40 (1983) (hereinafter S. Rep.). At that point, parole officials could “determin[e] that [the] prisoner had become rehabilitated and should be released from confinement.” Stith & Cabranes, supra, at 18.[Footnote 3] But this model of indeterminate sentencing eventually fell into disfavor. One concern was that it produced “[s]eri-ous disparities in [the] sentences” imposed on simi- larly situated defendants. Mistretta, 488 U. S., at 365. Another was that the system’s attempt to “achieve rehabilitation of offenders had failed.” Id., at 366. Lawmakers and others increasingly doubted that prison programs could “rehabilitate individuals on a routine basis”—or that parole officers could “determine accurately whether or when a particular prisoner ha[d] been rehabilitated.” S. Rep., at 40. Congress accordingly enacted the Sentencing Reform Act of 1984, 98 Stat. 1987 (SRA or Act), to overhaul federal sentencing practices. The Act abandoned indeterminate sentencing and parole in favor of a system in which Sentencing Guidelines, promulgated by a new Sentencing Commission, would provide courts with “a range of determinate sentences for categories of offenses and defendants.” Mistretta, 488 U. S., at 368. And the Act further channeled judges’ discretion by establishing a framework to govern their consideration and imposition of sentences. Under the SRA, a judge sentencing a federal offender must impose at least one of the following sanctions: imprisonment (often followed by supervised release), probation, or a fine. See §3551(b). In determining the appropriate sentence from among these options, §3553(a)(2) requires the judge to consider specified factors, including: “the need for the sentence imposed— “(A) to reflect the seriousness of the offense, to promote respect for the law, and to provide just punishment for the offense; “(B) to afford adequate deterrence to criminal conduct; “(C) to protect the public from further crimes of the defendant; and “(D) to provide the defendant with needed educational or vocational training, medical care, or other correctional treatment in the most effective manner.” These four considerations—retribution, deterrence, incapacitation, and rehabilitation—are the four purposes of sentencing generally, and a court must fashion a sentence “to achieve the[se] purposes … to the extent that they are applicable” in a given case. §3551(a). The SRA then provides additional guidance about how the considerations listed in §3553(a)(2) pertain to each of the Act’s main sentencing options—imprisonment, supervised release, probation, and fines. See §3582(a); §3583; §3562(a); §3572(a). These provisions make clear that a particular purpose may apply differently, or even not at all, depending on the kind of sentence under consideration. For example, a court may not take account of retribution (the first purpose listed in §3553(a)(2)) when imposing a term of supervised release. See §3583(c). Section 3582(a), the provision at issue here, specifies the “factors to be considered” when a court orders imprisonment. That section provides: “The court, in determining whether to impose a term of imprisonment, and, if a term of imprisonment is to be imposed, in determining the length of the term, shall consider the factors set forth in section 3553(a) to the extent that they are applicable, recognizing that imprisonment is not an appropriate means of promoting correction and rehabilitation.” A similar provision addresses the Sentencing Commission in its capacity as author of the Sentencing Guidelines. The SRA instructs the Commission to: “insure that the guidelines reflect the inappropriateness of imposing a sentence to a term of imprisonment for the purpose of rehabilitating the defendant or providing the defendant with needed educational or vocational training, medical care, or other correctional treatment.” 28 U. S. C. §994(k). With this statutory background established, we turn to the matter of interpretation. III A Our consideration of Tapia’s claim starts with the text of 18 U. S. C. §3582(a)—and given the clarity of that provi- sion’s language, could end there as well. As just noted, that section instructs courts to “recogniz[e] that imprisonment is not an appropriate means of promoting correction and rehabilitation.” A common—and in context the most natural—definition of the word “recognize” is “to ac-knowledge or treat as valid.” Random House Dictionary of the English Language 1611 (2d ed. 1987). And a thing that is not “appropriate” is not “suitable or fitting for a particular purpose.” Id., at 103. Putting these two definitions together, §3582(a) tells courts that they should acknowledge that imprisonment is not suitable for the purpose of promoting rehabilitation. And when should courts acknowledge this? Section §3582(a) answers: when “determining whether to impose a term of imprisonment, and, if a term of imprisonment is to be imposed, [when] determining the length of the term.” So a court making these decisions should consider the specified rationales of punishment except for rehabilitation, which it should ac-knowledge as an unsuitable justification for a prison term. As against this understanding, amicus argues that §3582(a)’s “recognizing” clause is not a flat prohibition but only a “reminder” or a “guide [for] sentencing judges’ cognitive processes.” Brief for Court-Appointed Amicus Curiae in Support of Judgment Below 23–24 (hereinafter Amicus Brief) (emphasis deleted). Amicus supports this view by offering a string of other definitions of the word “recognize”: “ ‘recall to mind,’ ‘realize,’ or ‘perceive clearly.’ ” Id., at 24 (quoting dictionary definitions). Once these are plugged in, amicus suggests, §3582(a) reveals itself as a kind of loosey-goosey caution not to put too much faith in the capacity of prisons to rehabilitate. But we do not see how these alternative meanings of “recognize” help amicus’s cause. A judge who “perceives clearly” that imprisonment is not an appropriate means of promoting rehabilitation would hardly incarcerate someone for that purpose. Ditto for a judge who “realizes” or “recalls” that imprisonment is not a way to rehabilitate an offender. To be sure, the drafters of the “recognizing” clause could have used still more commanding language: Congress could have inserted a “thou shalt not” or equivalent phrase to convey that a sentencing judge may never, ever, under any circumstances consider rehabilitation in imposing a prison term. But when we interpret a statute, we cannot allow the perfect to be the enemy of the merely excellent. Congress expressed itself clearly in §3582(a), even if armchair legislators might come up with something even better. And what Congress said was that when sentencing an offender to prison, the court shall consider all the purposes of punishment except rehabilitation—because imprisonment is not an appropriate means of pursuing that goal. Amicus also claims, echoing the Ninth Circuit’s reasoning in Duran, that §3582(a)’s “recognizing” clause bars courts from considering rehabilitation only when imposing a prison term, and not when deciding on its length. The argument goes as follows. Section 3582(a) refers to two decisions: “The court, [1] in determining whether to impose a term of imprisonment, and, if a term of imprisonment is to be imposed, [2] in determining the length of the term” must consider the purposes of punishment listed in §3553(a)(2), subject to the caveat of the “recognizing” clause. But that clause says only that “imprisonment” is not an appropriate means of rehabilitation. Because the “primary meaning of ‘imprisonment’ is ‘the act of confining a person,’ ” amicus argues, the clause relates only to [1] the decision to incarcerate, and not to [2] the separate determination of the sentence’s length. Amicus Brief 52. We again disagree. Under standard rules of grammar, §3582(a) says: A sentencing judge shall recognize that imprisonment is not appropriate to promote rehabilita-tion when the court considers the applicable factors of §3553(a)(2); and a court considers these factors when determining both whether to imprison an offender and what length of term to give him. The use of the word “imprisonment” in the “recognizing” clause does not destroy—but instead fits neatly into—this construction. “Imprisonment” as used in the clause most naturally means “[t]he state of being confined” or “a period of confinement.” Black’s Law Dictionary 825 (9th ed. 2009); see also Webster’s Third New International Dictionary 1137 (1993) (the “state of being imprisoned”). So the word does not distinguish between the defendant’s initial placement behind bars and his continued stay there. As the D. C. Circuit noted in rejecting an identical argument, “[a] sentencing court deciding to keep a defendant locked up for an additional month is, as to that month, in fact choosing imprisonment over release.” In re Sealed Case, 573 F. 3d 844, 850 (2009).[Footnote 4] Accordingly, the word “imprisonment” does not change the function of the “recognizing” clause—to constrain a sentencing court’s decision both to impose and to lengthen a prison term.[Footnote 5] The context of §3582(a) puts an exclamation point on this textual conclusion. As noted earlier, supra, at 6, another provision of the SRA restates §3582(a)’s message, but to a different audience. That provision, 28 U. S. C. §994(k), directs the Sentencing Commission to ensure that the Guidelines “reflect the inappropriateness of imposing a sentence to a term of imprisonment for the purpose of rehabilitating the defendant or providing the defendant with needed educational or vocational training, medical care, or other correctional treatment.” In this way, Congress ensured that all sentencing officials would work in tandem to implement the statutory determination to “rejec[t] imprisonment as a means of promoting rehabilitation.” Mistretta, 488 U. S., at 367 (citing 28 U. S. C. §994(k)). Section 994(k) bars the Commission from recommending a “term of imprisonment”—a phrase that again refers both to the fact and to the length of incarceration—based on a defendant’s rehabilitative needs. And §3582(a) prohibits a court from considering those needs to impose or lengthen a period of confinement when selecting a sentence from within, or choosing to depart from, the Guidelines range. Each actor at each stage in the sentencing process receives the same message: Do not think about prison as a way to rehabilitate an offender. Equally illuminating here is a statutory silence—the absence of any provision granting courts the power to ensure that offenders participate in prison rehabilitation programs. For when Congress wanted sentencing courts to take account of rehabilitative needs, it gave courts the authority to direct appropriate treatment for offenders. Thus, the SRA instructs courts, in deciding whether to impose probation or supervised release, to consider whether an offender could benefit from training and treatment programs. See 18 U. S. C. §3562(a); §3583(c). And so the SRA also authorizes courts, when imposing those sentences, to order an offender’s participation in cer- tain programs and facilities. §3563(b)(9); §3563(b)(11); §3563(a)(4); §3583(d). As a condition of probation, for example, the court may require the offender to “undergo available medical, psychiatric, or psychological treatment, including treatment for drug or alcohol dependency, as specified by the court, and [to] remain in a specified institution if required for that purpose.” §3563(b)(9). If Congress had similarly meant to allow courts to base prison terms on offenders’ rehabilitative needs, it would have given courts the capacity to ensure that offenders participate in prison correctional programs. But in fact, courts do not have this authority. When a court sentences a federal offender, the BOP has plenary control, subject to statutory constraints, over “the place of the prisoner’s imprisonment,” §3621(b), and the treatment programs (if any) in which he may participate, §§3621(e), (f); §3624(f). See also 28 CFR pt. 544 (2010) (BOP regulations for administering inmate educational, recreational, and vocational programs); 28 CFR pt. 550, subpart F (drug abuse treatment programs). A sentencing court can recommend that the BOP place an offender in a particular facility or program. See §3582(a). But decisionmaking authority rests with the BOP. This case well illustrates the point. As noted earlier, the District Court “strongly recommend[ed]” that Tapia participate in RDAP, App. 29, and serve her sentence at FCI Dublin, “where they have the facilities to really help her,” id., at 28. But the court’s recommendations were only recommendations—and in the end they had no effect. See Amicus Brief 42 (“[Tapia] was not admitted to RDAP, nor even placed in the prison recommended by the district court”); Reply Brief for United States 8, n. 1 (“According to BOP records, [Tapia] was encouraged to enroll [in RDAP] during her psychology intake screening at [the federal prison], but she stated that she was not interested, and she has not volunteered for the program”). The sentencing court may have had plans for Tapia’s rehabilitation, but it lacked the power to implement them. That incapacity speaks volumes. It indicates that Congress did not intend that courts consider offenders’ rehabilitative needs when imposing prison sentences. Finally, for those who consider legislative history useful, the key Senate Report concerning the SRA provides one last piece of corroborating evidence. According to that Report, decades of experience with indeterminate sentencing, resulting in the release of many inmates after they completed correctional programs, had left Congress skeptical that “rehabilitation can be induced reliably in a prison setting.” S. Rep., at 38. Although some critics argued that “rehabilitation should be eliminated completely as a purpose of sentencing,” Congress declined to adopt that categorical position. Id., at 76. Instead, the Report explains, Congress barred courts from considering rehabilitation in imposing prison terms, ibid., and n. 165, but not in ordering other kinds of sentences, ibid., and n. 164. “[T]he purpose of rehabilitation,” the Report stated, “is still important in determining whether a sanction other than a term of imprisonment is appropriate in a particular case.” See id., at 76–77 (emphasis added). And so this is a case in which text, context, and history point to the same bottom line: Section 3582(a) precludes sentencing courts from imposing or lengthening a prison term to promote an offender’s rehabilitation. B With all these sources of statutory meaning stacked against him, amicus understandably tries to put the SRA’s view of rehabilitation in a wholly different frame. Amicus begins by conceding that Congress, in enacting the SRA, rejected the old “[r]ehabilitation [m]odel.” Amicus Brief 1. But according to amicus, that model had a very limited focus: It was the belief that “isolation and prison routine” could alone produce “penitence and spiritual renewal.” Id., at 1, 11. What the rehabilitation model did not include—and the SRA therefore did not reject—was prison treatment programs (including for drug addiction) targeted to offenders’ particular needs. See id., at 21, 25, 27–28. So even after the passage of §3582(a), amicus argues, a court may impose or lengthen a prison sentence to promote an offender’s participation in a targeted treatment program. The only thing the court may not do is to impose a prison term on the ground that confinement itself—its inherent solitude and routine—will lead to rehabilitation. We think this reading of the SRA is too narrow. For one thing, the relevant history shows that at the time of the SRA’s enactment, prison rehabilitation efforts focused on treatment, counseling, and training programs, not on seclusion and regimentation. See Rotman, The Failure of Reform: United States, 1865–1965, in Oxford History of the Prison: The Practice of Punishment in Western Society 169, 189–190 (N. Morris & D. Rothman eds. 1995) (describing the pre-SRA “therapeutic model of rehabilitation” as characterized by “individualized treatment” and “vocational training and group counseling programs”); see also n. 3, supra (noting pre-SRA statutes linking the confinement of drug addicts to the completion of treatment programs). Indeed, Congress had in mind precisely these programs when it prohibited consideration of rehabilitation in imposing a prison term. See 28 U. S. C. §994(k) (instructing the Sentencing Commission to prevent the use of imprisonment to “provid[e] the defendant with needed educational or vocational training … or other correctional treatment”); S. Rep., at 40 (rejecting the “model of ‘coercive’ rehabilitation—the theory of correction that ties prison release dates to the successful completion of certain vocational, educational, and counseling programs within the prisons”). Far from falling outside the “rehabilitation model,” these programs practically defined it. It is hardly surprising, then, that amicus’s argument finds little support in the statutory text. Read most naturally, 18 U. S. C. §3582(a)’s prohibition on “promoting correction and rehabilitation” covers efforts to place offenders in rehabilitation programs. Indeed, §3582(a)’s lan-guage recalls the SRA’s description of the rehabilitative purpose of sentencing—“provid[ing] the defendant with needed educational or vocational training, medical care, or other correctional treatment.” §3553(a)(2)(D). That description makes clear that, under the SRA, treatment, training, and like programs are rehabilitation’s sum and substance. So amicus’s efforts to exclude rehabilitation programs from the “recognizing” clause’s reach do not succeed. That section prevents a sentencing court from imposing or lengthening a prison term because the court thinks an offender will benefit from a prison treatment program. IV In this case, the sentencing transcript suggests the possibility that Tapia’s sentence was based on her rehabilitative needs. We note first what we do not disapprove about Tapia’s sentencing. A court commits no error by discussing the opportunities for rehabilitation within prison or the benefits of specific treatment or training programs. To the contrary, a court properly may address a person who is about to begin a prison term about these important matters. And as noted earlier, a court may urge the BOP to place an offender in a prison treatment program. See supra, at 11. Section 3582(a) itself provides, just after the clause at issue here, that a court may “make a recommendation concerning the type of prison facility appropriate for the defendant”; and in this calculus, the presence of a rehabilitation program may make one facility more appropriate than another. So the sentencing court here did nothing wrong—and probably something very right—in trying to get Tapia into an effective drug treatment program. But the record indicates that the court may have done more—that it may have selected the length of the sentence to ensure that Tapia could complete the 500 Hour Drug Program. “The sentence has to be sufficient,” the court explained, “to provide needed correctional treatment, and here I think the needed correctional treatment is the 500 Hour Drug Program.” App. 27; see supra, at 1–2. Or again: The “number one” thing “is the need to provide treatment. In other words, so she is in long enough to get the 500 Hour Drug Program.” App. 27; see supra, at 2. These statements suggest that the court may have calculated the length of Tapia’s sentence to ensure that she receive certain rehabilitative services. And that a sentencing court may not do. As we have held, a court may not impose or lengthen a prison sentence to enable an offender to complete a treatment program or otherwise to promote rehabilitation. For the reasons stated, we reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. Consistent with our practice, see, e.g., United States v. Marcus, 560 U. S. ___, ___ (2010) (slip op., at 8), we leave it to the Court of Appeals to consider the effect of Tapia’s failure to object to the sentence when imposed. See Fed. Rule Crim. Proc. 52(b); United States v. Olano, 507 U. S. 725, 731 (1993). It is so ordered. Footnote 1 Three Circuits have held that §3582(a) allows a court to lengthen, although not to impose, a prison term based on the need for rehabilitation. See United States v. Duran, 37 F. 3d 557 (CA9 1994); United States v. Hawk Wing, 433 F. 3d 622 (CA8 2006); United States v. Jimenez, 605 F. 3d 415 (CA6 2010). Two Courts of Appeals have ruled that §3582(a) bars a court from either imposing or increasing a period of confinement for rehabilitative reasons. See United States v. Manzella, 475 F. 3d 152 (CA3 2007); In re Sealed Case, 573 F. 3d 844 (CADC 2009). Footnote 2 We appointed Stephanos Bibas to brief and argue the case, 562 U. S. ___ (2011), and he has ably discharged his responsibilities. Footnote 3 The statutes governing punishment of drug-addicted offenders (like Tapia) provide an example of this system at work. If a court concluded that such an offender was “likely to be rehabilitated through treatment,” it could order confinement “for treatment … for an indeterminate period of time” not to exceed the lesser of 10 years or the statutory maximum for the offender’s crime. 18 U. S. C. §4253(a) (1982 ed.); see also §4251(c) (“ ‘Treatment’ includes confinement and treatment in an institution … and includes, but is not limited to, medical, educational, social, psychological, and vocational services, corrective and preventive guidance and training, and other rehabilitative services”). Once the offender had undergone treatment for six months, the Attorney General could recommend that the Board of Parole release him from custody, and the Board could then order release “in its discretion.” §4254. Footnote 4 Indeed, we can scarcely imagine a reason why Congress would have wanted to draw the distinction that amicus urges on us. That distinction would prevent a court from considering rehabilitative needs in imposing a 1-month sentence rather than probation, but not in choosing a 60-month sentence over a 1-month term. The only policy argument amicus can offer in favor of this result is that “[t]he effects of imprisonment plateau a short while after the incarceration” and “ ‘[t]he dif-ference in harm between longer and shorter prison terms is smaller than typically assumed.’ ” Amicus Brief 56. But nothing in the SRA indicates that Congress is so indifferent to the length of prison terms. Footnote 5 The Government argues that “Congress did not intend to prohibit courts from imposing less imprisonment in order to promote a defendant’s rehabilitation.” Brief for United States 40 (emphasis added). This case does not require us to address that question, and nothing in our decision expresses any views on it.

562.US.170

After petitioner Thompson’s fiancée, Miriam Regalado, filed a sex discrimination charge with the Equal Employment Opportunity Commission (EEOC) against their employer, respondent North American Stainless (NAS), NAS fired Thompson. He filed his own charge and a subsequent suit under Title VII of the Civil Rights Act, claiming that NAS fired him to retaliate against Regalado for filing her charge. The District Court granted NAS summary judgment on the ground that third-party retaliation claims were not permitted by Title VII, which prohibits discrimination against an employee “because he has made a [Title VII] charge,” 42 U. S. C. §2000e–3(a), and which permits, inter alia, a “person claiming to be aggrieved … by [an] alleged employment practice” to file a civil action, §2000e–5(f)(1). The en banc Sixth Circuit affirmed, reasoning that Thompson was not entitled to sue NAS for retaliation because he had not engaged in any activity protected by the statute. Held: 1. If the facts Thompson alleges are true, his firing by NAS constituted unlawful retaliation. Title VII’s antiretaliation provision must be construed to cover a broad range of employer conduct. Burlington N. & S. F. R. Co. v. White, 548 U. S. 53. It prohibits any employer action that “ ‘well might have “dissuaded a reasonable worker from making or supporting a [discrimination] charge,” ’ ” id., at 68. That test must be applied in an objective fashion, to “avoi[d] the uncertainties and unfair discrepancies that can plague a judicial effort to determine a plaintiff’s unusual subjective feelings.” Id., at 68–69. A reasonable worker obviously might be dissuaded from engaging in protected activity if she knew that her fiancé would be fired. Pp. 2–4. 2. Title VII grants Thompson a cause of action. Pp. 4–7. (a) For Title VII standing purposes, the term “person aggrieved” must be construed more narrowly than the outer boundaries of Article III. Dictum in Trafficante v. Metropolitan Life Ins. Co., 409 U. S. 205, suggesting that Title VII’s aggrievement requirement reaches as far as Article III permits, is too expansive and the Court declines to follow it. At the other extreme, limiting “person aggrieved” to the person who was the subject of unlawful retaliation is an artificially narrow reading. A common usage of the term “person aggrieved” avoids both of these extremes. The Administrative Procedure Act, which authorizes suit to challenge a federal agency by any “person … adversely affected or aggrieved … within the meaning of a relevant statute,” 5 U. S. C. §702, establishes a regime under which a plaintiff may not sue unless he “falls within the ‘zone of interests’ sought to be protected by the statutory provision whose violation forms the legal basis for his complaint,” Lujan v. National Wildlife Federation, 497 U. S. 871, 883. Title VII’s term “aggrieved” incorporates that test, enabling suit by any plaintiff with an interest “ ‘arguably [sought] to be protected’ by the statutes,” National Credit Union Admin. v. First Nat. Bank & Trust Co., 522 U. S. 479, 495, while excluding plaintiffs who might technically be injured in an Article III sense but whose interests are unrelated to Title VII’s statutory prohibitions. Pp. 4–7. (b) Applying that test here, Thompson falls within the zone of interests protected by Title VII. He was an employee of NAS, and Title VII’s purpose is to protect employees from their employers’ unlawful actions. Moreover, accepting the facts as alleged, Thompson is not an accidental victim of the retaliation. Hurting him was the unlawful act by which NAS punished Regalado. Thus, Thompson is a person aggrieved with standing to sue under Title VII. P. 7. 567 F. 3d 804, reversed and remanded. Scalia, J., delivered the opinion of the Court, in which all other Members joined, except Kagan, J., who took no part in the consideration or decision of the case. Ginsburg, J., filed a concurring opinion, in which Breyer, J., joined.

Until 2003, both petitioner Eric Thompson and his fiancée, Miriam Regalado, were employees of respondent North American Stainless (NAS). In February 2003, the Equal Employment Opportunity Commission (EEOC) notified NAS that Regalado had filed a charge alleging sex discrimination. Three weeks later, NAS fired Thompson. Thompson then filed a charge with the EEOC. After conciliation efforts proved unsuccessful, he sued NAS in the United States District Court for the Eastern District of Kentucky under Title VII of the Civil Rights Act of 1964, 78 Stat. 253, 42 U. S. C. §2000e et seq., claiming that NAS had fired him in order to retaliate against Regalado for filing her charge with the EEOC. The District Court granted summary judgment to NAS, concluding that Title VII “does not permit third party retaliation claims.” 435 F. Supp. 2d 633, 639 (ED Ky. 2006). After a panel of the Sixth Circuit reversed the District Court, the Sixth Circuit granted rehearing en banc and affirmed by a 10-to-6 vote. 567 F. 3d 804 (2009). The court reasoned that because Thompson did not “engag[e] in any statutorily protected activity, either on his own behalf or on behalf of Miriam Regalado,” he “is not included in the class of persons for whom Congress created a retaliation cause of action.” Id., at 807–808. We granted certiorari. 561 U. S. ___ (2010). I Title VII provides that “[i]t shall be an unlawful employment practice for an employer to discriminate against any of his employees … because he has made a charge” under Title VII. 42 U. S. C. §2000e–3(a). The statute permits “a person claiming to be aggrieved” to file a charge with the EEOC alleging that the employer committed an unlawful employment practice, and, if the EEOC declines to sue the employer, it permits a civil action to “be brought … by the person claiming to be aggrieved … by the alleged unlawful employment practice.” §2000e–5(b), (f)(1). It is undisputed that Regalado’s filing of a charge with the EEOC was protected conduct under Title VII. In the procedural posture of this case, we are also required to assume that NAS fired Thompson in order to retaliate against Regalado for filing a charge of discrimination. This case therefore presents two questions: First, did NAS’s firing of Thompson constitute unlawful retaliation? And second, if it did, does Title VII grant Thompson a cause of action? II With regard to the first question, we have little difficulty concluding that if the facts alleged by Thompson are true, then NAS’s firing of Thompson violated Title VII. In Burlington N. & S. F. R. Co. v. White, 548 U. S. 53 (2006), we held that Title VII’s antiretaliation provision must be construed to cover a broad range of employer conduct. We reached that conclusion by contrasting the text of Title VII’s antiretaliation provision with its substantive antidiscrimination provision. Title VII prohibits discrimination on the basis of race, color, religion, sex, and national origin “ ‘with respect to … compensation, terms, conditions, or privileges of employment,’ ” and discriminatory practices that would “ ‘deprive any individual of employment opportunities or otherwise adversely affect his status as an employee.’ ” Id., at 62 (quoting 42 U. S. C. §2000e–2(a) (emphasis deleted)). In contrast, Title VII’s antiretaliation provision prohibits an employer from “ ‘discriminat[ing] against any of his employees’ ” for engaging in protected conduct, without specifying the employer acts that are prohibited. 548 U. S., at 62 (quoting §2000e–3(a) (emphasis deleted)). Based on this textual distinction and our understanding of the antiretaliation provision’s purpose, we held that “the antiretaliation provision, unlike the substantive provision, is not limited to discriminatory actions that affect the terms and conditions of employment.” Id., at 64. Rather, Title VII’s antiretaliation provision prohibits any employer action that “well might have dissuaded a reasonable worker from making or supporting a charge of discrimination.” Id., at 68 (internal quotation marks omitted). We think it obvious that a reasonable worker might be dissuaded from engaging in protected activity if she knew that her fiancé would be fired. Indeed, NAS does not dispute that Thompson’s firing meets the standard set forth in Burlington. Tr. of Oral Arg. 30. NAS raises the concern, however, that prohibiting reprisals against third parties will lead to difficult line-drawing problems concerning the types of relationships entitled to protection. Perhaps retaliating against an employee by firing his fiancée would dissuade the employee from engaging in protected activity, but what about firing an employee’s girlfriend, close friend, or trusted co-worker? Applying the Burlington standard to third-party reprisals, NAS argues, will place the employer at risk any time it fires any employee who happens to have a connection to a different employee who filed a charge with the EEOC. Although we acknowledge the force of this point, we do not think it justifies a categorical rule that third-party reprisals do not violate Title VII. As explained above, we adopted a broad standard in Burlington because Title VII’s antiretaliation provision is worded broadly. We think there is no textual basis for making an exception to it for third-party reprisals, and a preference for clear rules cannot justify departing from statutory text. We must also decline to identify a fixed class of relationships for which third-party reprisals are unlawful. We expect that firing a close family member will almost always meet the Burlington standard, and inflicting a milder reprisal on a mere acquaintance will almost never do so, but beyond that we are reluctant to generalize. As we explained in Burlington, 548 U. S., at 69, “the significance of any given act of retaliation will often depend upon the particular circumstances.” Given the broad statutory text and the variety of workplace contexts in which retaliation may occur, Title VII’s antiretaliation provision is simply not reducible to a comprehensive set of clear rules. We emphasize, however, that “the provision’s standard for judging harm must be objective,” so as to “avoi[d] the uncertainties and unfair discrepancies that can plague a judicial effort to determine a plaintiff’s unusual subjective feelings.” Id., at 68–69. III The more difficult question in this case is whether Thompson may sue NAS for its alleged violation of Title VII. The statute provides that “a civil action may be brought … by the person claiming to be aggrieved.” 42 U. S. C. §2000e–5(f)(1). The Sixth Circuit concluded that this provision was merely a reiteration of the requirement that the plaintiff have Article III standing. 567 F. 3d, at 808, n. 1. We do not understand how that can be. The provision unquestionably permits a person “claiming to be aggrieved” to bring “a civil action.” It is arguable that the aggrievement referred to is nothing more than the minimal Article III standing, which consists of injury in fact caused by the defendant and remediable by the court. See Lujan v. Defenders of Wildlife, 504 U. S. 555, 560–561 (1992). But Thompson’s claim undoubtedly meets those requirements, so if that is indeed all that aggrievement consists of, he may sue. We have suggested in dictum that the Title VII aggrievement requirement conferred a right to sue on all who satisfied Article III standing. Trafficante v. Metropolitan Life Ins. Co., 409 U. S. 205 (1972), involved the “person aggrieved” provision of Title VIII (the Fair Housing Act) rather than Title VII. In deciding the case, however, we relied upon, and cited with approval, a Third Circuit opinion involving Title VII, which, we said, “concluded that the words used showed ‘a congressional intention to define standing as broadly as is permitted by Article III of the Constitution.’ ” Id., at 209 (quoting Hackett v. McGuire Bros., Inc., 445 F. 2d 442, 446 (1971)). We think that dictum regarding Title VII was too expansive. Indeed, the Trafficante opinion did not adhere to it in expressing its Title VIII holding that residents of an apartment complex could sue the owner for his racial discrimination against prospective tenants. The opinion said that the “person aggrieved” of Title VIII was coextensive with Article III “insofar as tenants of the same housing unit that is charged with discrimination are concerned.” 409 U. S., at 209 (emphasis added). Later opinions, we must acknowledge, reiterate that the term “aggrieved” in Title VIII reaches as far as Article III permits, see Bennett v. Spear, 520 U. S. 154, 165–166 (1997); Gladstone, Realtors v. Village of Bellwood, 441 U. S. 91, 109 (1979), though the holdings of those cases are compatible with the “zone of interests” limitation that we discuss below. In any event, it is Title VII rather than Title VIII that is before us here, and as to that we are surely not bound by the Trafficante dictum. We now find that this dictum was ill-considered, and we decline to follow it. If any person injured in the Article III sense by a Title VII violation could sue, absurd consequences would follow. For example, a shareholder would be able to sue a company for firing a valuable employee for racially discriminatory reasons, so long as he could show that the value of his stock decreased as a consequence. At oral argument Thompson acknowledged that such a suit would not lie, Tr. of Oral Arg. 5–6. We agree, and therefore conclude that the term “aggrieved” must be construed more narrowly than the outer boundaries of Article III. At the other extreme from the position that “person aggrieved” means anyone with Article III standing, NAS argues that it is a term of art that refers only to the employee who engaged in the protected activity. We know of no other context in which the words carry this artificially narrow meaning, and if that is what Congress intended it would more naturally have said “person claiming to have been discriminated against” rather than “person claiming to be aggrieved.” We see no basis in text or prior practice for limiting the latter phrase to the person who was the subject of unlawful retaliation. Moreover, such a reading contradicts the very holding of Trafficante, which was that residents of an apartment complex were “person[s] aggrieved” by discrimination against prospective tenants. We see no reason why the same phrase in Title VII should be given a narrower meaning. In our view there is a common usage of the term “person aggrieved” that avoids the extremity of equating it with Article III and yet is fully consistent with our application of the term in Trafficante. The Administrative Procedure Act, 5 U. S. C. §551 et seq., authorizes suit to challenge a federal agency by any “person … adversely affected or aggrieved … within the meaning of a relevant statute.” §702. We have held that this language establishes a regime under which a plaintiff may not sue unless he “falls within the ‘zone of interests’ sought to be protected by the statutory provision whose violation forms the legal basis for his complaint.” Lujan v. National Wildlife Federation, 497 U. S. 871, 883 (1990). We have described the “zone of interests” test as denying a right of review “if the plaintiff’s interests are so marginally related to or inconsistent with the purposes implicit in the statute that it cannot reasonably be assumed that Congress intended to permit the suit.” Clarke v. Securities Industry Assn., 479 U. S. 388, 399–400 (1987). We hold that the term “aggrieved” in Title VII incorporates this test, enabling suit by any plaintiff with an interest “arguably [sought] to be protected by the statutes,” National Credit Union Admin. v. First Nat. Bank & Trust Co., 522 U. S. 479, 495 (1998) (internal quotation marks omitted), while excluding plaintiffs who might technically be injured in an Article III sense but whose interests are unrelated to the statutory prohibitions in Title VII. Applying that test here, we conclude that Thompson falls within the zone of interests protected by Title VII. Thompson was an employee of NAS, and the purpose of Title VII is to protect employees from their employers’ unlawful actions. Moreover, accepting the facts as alleged, Thompson is not an accidental victim of the retaliation—collateral damage, so to speak, of the employer’s unlawful act. To the contrary, injuring him was the employer’s intended means of harming Regalado. Hurting him was the unlawful act by which the employer punished her. In those circumstances, we think Thompson well within the zone of interests sought to be protected by Title VII. He is a person aggrieved with standing to sue. * * * The judgment of the Sixth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

564.US.2010_10-10

After a South Carolina family court ordered petitioner Turner to pay $51.73 per week to respondent Rogers to help support their child, Turner repeatedly failed to pay the amount due and was held in contempt five times. For the first four, he was sentenced to 90 days’ imprisonment, but he ultimately paid what he owed (twice without being jailed, twice after spending a few days in custody). The fifth time he did not pay but completed a 6-month sentence. After his release, the family court clerk issued a new “show cause” order against Turner because he was $5728.76 in arrears. Both he and Rogers were unrepresented by counsel at his brief civil contempt hearing. The judge found Turner in willful contempt and sentenced him to 12 months in prison without making any finding as to his ability to pay or indicating on the contempt order form whether he was able to make support payments. After Turner completed his sentence, the South Carolina Supreme Court rejected his claim that the Federal Constitution entitled him to counsel at his contempt hearing, declaring that civil contempt does not require all the constitutional safeguards applicable in criminal contempt proceedings. Held: 1. Even though Turner has completed his 12-month sentence, and there are not alleged to be collateral consequences of the contempt determination that might keep the dispute alive, this case is not moot, because it is “capable of repetition” while “evading review,” Southern Pac. Terminal Co. v. Interstate Commerce Comm’n, 219 U. S. 498, 515. A case remains live if “(1) the challenged action [is] in its duration too short to be fully litigated prior to its cessation or expiration, and (2) there [is] a reasonable expectation that the same complaining party [will] be subjected to the same action again.” Weinstein v. Bradford, 423 U. S. 147, 149. Here, the “challenged action,” Turner’s imprisonment for up to 12 months, is “in its duration too short to be fully litigated” through the state courts (and arrive here) prior to its “expiration.” First Nat’l Bank of Boston v. Bellotti, 435 U. S. 765, 774. And there is a more than “reasonable” likelihood that Turner will again be “subjected to the same action” because he has frequently failed to make his support payments, has been the subject of several civil contempt proceedings, has been imprisoned several times, and is, once again, the subject of civil contempt proceedings for failure to pay. DeFunis v. Odegaard, 416 U. S. 312, and St. Pierre v. United States, 319 U. S. 41, distinguished. Pp. 5–7. 2. The Fourteenth Amendment’s Due Process Clause does not automatically require the State to provide counsel at civil contempt proceedings to an indigent noncustodial parent who is subject to a child support order, even if that individual faces incarceration. In particular, that Clause does not require that counsel be provided where the opposing parent or other custodian is not represented by counsel and the State provides alternative procedural safeguards equivalent to adequate notice of the importance of the ability to pay, a fair opportunity to present, and to dispute, relevant information, and express court findings as to the supporting parent’s ability to comply with the support order. Pp. 7–16. (a) This Court’s precedents provide no definitive answer to the question whether counsel must be provided. The Sixth Amendment grants an indigent criminal defendant the right to counsel, see, e.g., United States v. Dixon, 509 U. S. 688, 696, but does not govern civil cases. Civil and criminal contempt differ. A court may not impose punishment “in a civil contempt proceeding when it is clearly established that the alleged contemnor is unable to comply with the terms of the order.” Hicks v. Feiock, 485 U. S. 624, 638, n. 9. And once a civil contemnor complies with the underlying order, he is purged of the contempt and is free. Id., at 633. The Due Process Clause allows a State to provide fewer procedural protections in civil contempt proceedings than in a criminal case. Id., at 637–641. Cases directly concerning a right to counsel in civil cases have found a presumption of such a right “only” in cases involving incarceration, but have not held that a right to counsel exists in all such cases. See In re Gault, 387 U. S. 1; Vitek v. Jones, 445 U. S. 480; and Lassiter v. Department of Social Servs. of Durham Cty., 452 U. S. 18. Pp. 7–10. (b) Because a contempt proceeding to compel support payments is civil, the question whether the “specific dictates of due process” require appointed counsel is determined by examining the “distinct factors” this Court has used to decide what specific safeguards are needed to make a civil proceeding fundamentally fair. Mathews v. Eldridge, 424 U. S. 319, 335. As relevant here those factors include (1) the nature of “the private interest that will be affected,” (2) the comparative “risk” of an “erroneous deprivation” of that interest with and without “additional or substitute procedural safeguards,” and (3) the nature and magnitude of any countervailing interest in not providing “additional or substitute procedural requirement[s].” Ibid. The “private interest that will be affected” argues strongly for the right to counsel here. That interest consists of an indigent defendant’s loss of personal liberty through imprisonment. Freedom “from bodily restraint” lies “at the core of the liberty protected by the Due Process Clause.” Foucha v. Louisiana, 504 U. S. 71, 80. Thus, accurate decisionmaking as to the “ability to pay”—which marks a dividing line between civil and criminal contempt, Hicks, supra, at 635, n. 7—must be assured because an incorrect decision can result in a wrongful incarceration. And because ability to comply divides civil and criminal contempt proceedings, an erroneous determination would also deprive a defendant of the procedural protections a criminal proceeding would demand. Questions about ability to pay are likely to arise frequently in child custody cases. On the other hand, due process does not always require the provision of counsel in civil proceedings where incarceration is threatened. See Gagnon v. Scarpelli, 411 U. S. 778. To determine whether a right to counsel is required here, opposing interests and the probable value of “additional or substitute procedural safeguards” must be taken into account. Mathews, supra, at 335. Doing so reveals three related considerations that, taken together, argue strongly against requiring counsel in every proceeding of the present kind. First, the likely critical question in these cases is the defendant’s ability to pay, which is often closely related to his indigence and relatively straightforward. Second, sometimes, as here, the person opposing the defendant at the hearing is not the government represented by counsel but the custodial parent unrepresented by counsel. A requirement that the State provide counsel to the noncustodial parent in these cases could create an asymmetry of representation that would “alter significantly the nature of the proceeding,” Gagnon, supra, at 787, creating a degree of formality or delay that would unduly slow payment to those immediately in need and make the proceedings less fair overall. Third, as the Federal Government points out, an available set of “substitute procedural safeguards,” Mathews, supra, at 335, if employed together, can significantly reduce the risk of an erroneous deprivation of liberty. These include (1) notice to the defendant that his “ability to pay” is a critical issue in the contempt proceeding; (2) the use of a form (or the equivalent) to elicit relevant financial information from him; (3) an opportunity at the hearing for him to respond to statements and questions about his financial status; and (4) an express finding by the court that the defendant has the ability to pay. This decision does not address civil contempt proceedings where the underlying support payment is owed to the State, e.g., for reimbursement of welfare funds paid to the custodial parent, or the question what due process requires in an unusually complex case where a defendant “can fairly be represented only by a trained advocate,” Gagnon, supra, at 788. Pp. 10–16. 3. Under the circumstances, Turner’s incarceration violated due process because he received neither counsel nor the benefit of alternative procedures like those the Court describes. He did not have clear notice that his ability to pay would constitute the critical question in his civil contempt proceeding. No one provided him with a form (or the equivalent) designed to elicit information about his financial circumstances. And the trial court did not find that he was able to pay his arrearage, but nonetheless found him in civil contempt and ordered him incarcerated. P. 16. 387 S. C. 142, 691 S. E. 2d 470, vacated and remanded. Breyer, J., delivered the opinion of the Court, in which Kennedy, Ginsburg, Sotomayor, and Kagan, JJ., joined. Thomas, J., filed a dissenting opinion, in which Scalia, J., joined, and in which Roberts, C. J., and Alito, JJ., joined as to Parts I–B and II.

South Carolina’s Family Court enforces its child support orders by threatening with incarceration for civil contempt those who are (1) subject to a child support order, (2) able to comply with that order, but (3) fail to do so. We must decide whether the Fourteenth Amendment’s Due Process Clause requires the State to provide counsel (at a civil contempt hearing) to an indigent person potentially faced with such incarceration. We conclude that where as here the custodial parent (entitled to receive the support) is unrepresented by counsel, the State need not provide counsel to the noncustodial parent (required to provide the support). But we attach an important caveat, namely, that the State must nonetheless have in place alternative procedures that assure a fundamentally fair determination of the critical incarceration-related question, whether the supporting parent is able to comply with the support order. I A South Carolina family courts enforce their child support orders in part through civil contempt proceedings. Each month the family court clerk reviews outstanding child support orders, identifies those in which the supporting parent has fallen more than five days behind, and sends that parent an order to “show cause” why he should not be held in contempt. S. C. Rule Family Ct. 24 (2011). The “show cause” order and attached affidavit refer to the relevant child support order, identify the amount of the arrearage, and set a date for a court hearing. At the hearing that parent may demonstrate that he is not in contempt, say, by showing that he is not able to make the required payments. See Moseley v. Mosier, 279 S. C. 348, 351, 306 S. E. 2d 624, 626 (1983) (“When the parent is unable to make the required payments, he is not in contempt”). If he fails to make the required showing, the court may hold him in civil contempt. And it may require that he be imprisoned unless and until he purges himself of contempt by making the required child support payments (but not for more than one year regardless). See S. C. Code Ann. §63–3–620 (Supp. 2010) (imprisonment for up to one year of “adult who wilfully violates” a court order); Price v. Turner, 387 S. C. 142, 145, 691 S. E. 2d 470, 472 (2010) (civil contempt order must permit purging of contempt through compliance). B In June 2003 a South Carolina family court entered an order, which (as amended) required petitioner, Michael Turner, to pay $51.73 per week to respondent, Rebecca Rogers, to help support their child. (Rogers’ father, Larry Price, currently has custody of the child and is also a respondent before this Court.) Over the next three years, Turner repeatedly failed to pay the amount due and was held in contempt on five occasions. The first four times he was sentenced to 90 days’ imprisonment, but he ultimately paid the amount due (twice without being jailed, twice after spending two or three days in custody). The fifth time he did not pay but completed a 6-month sentence. After his release in 2006 Turner remained in arrears. On March 27, 2006, the clerk issued a new “show cause” order. And after an initial postponement due to Turner’s failure to appear, Turner’s civil contempt hearing took place on January 3, 2008. Turner and Rogers were present, each without representation by counsel. The hearing was brief. The court clerk said that Turner was $5,728.76 behind in his payments. The judge asked Turner if there was “anything you want to say.” Turner replied, “Well, when I first got out, I got back on dope. I done meth, smoked pot and everything else, and I paid a little bit here and there. And, when I finally did get to working, I broke my back, back in September. I filed for disability and SSI. And, I didn’t get straightened out off the dope until I broke my back and laid up for two months. And, now I’m off the dope and everything. I just hope that you give me a chance. I don’t know what else to say. I mean, I know I done wrong, and I should have been paying and helping her, and I’m sorry. I mean, dope had a hold to me.” App. to Pet. for Cert. 17a. The judge then said, “[o]kay,” and asked Rogers if she had anything to say. Ibid. After a brief discussion of federal benefits, the judge stated, “If there’s nothing else, this will be the Order of the Court. I find the Defendant in willful contempt. I’m [going to] sentence him to twelve months in the Oconee County Detention Center. He may purge himself of the contempt and avoid the sentence by having a zero balance on or before his release. I’ve also placed a lien on any SSI or other benefits.” Id., at 18a. The judge added that Turner would not receive good-time or work credits, but “[i]f you’ve got a job, I’ll make you eligible for work release.” Ibid. When Turner asked why he could not receive good-time or work credits, the judge said, “[b]ecause that’s my ruling.” Ibid. The court made no express finding concerning Turner’s ability to pay his arrearage (though Turner’s wife had voluntarily submitted a copy of Turner’s application for disability benefits, cf. post, at 7, n. 3 (Thomas, J., dissenting); App. 135a–136a). Nor did the judge ask any followup questions or otherwise address the ability-to-pay issue. After the hearing, the judge filled out a prewritten form titled “Order for Contempt of Court,” which included the statement: “Defendant (was) (was not) gainfully employed and/or (had) (did not have) the ability to make these support payments when due.” Id., at 60a, 61a. But the judge left this statement as is without indicating whether Turner was able to make support payments. C While serving his 12-month sentence, Turner, with the help of pro bono counsel, appealed. He claimed that the Federal Constitution entitled him to counsel at his con- tempt hearing. The South Carolina Supreme Court decided Turner’s appeal after he had completed his sentence. And it rejected his “right to counsel” claim. The court pointed out that civil contempt differs significantly from criminal contempt. The former does not require all the “constitutional safeguards” applicable in criminal proceedings. 387 S. C., at 145, 691 S. E. 2d, at 472. And the right to government-paid counsel, the Supreme Court held, was one of the “safeguards” not required. Ibid. Turner sought certiorari. In light of differences among state courts (and some federal courts) on the applicability of a “right to counsel” in civil contempt proceedings enforcing child support orders, we granted the writ. Compare, e.g., Pasqua v. Council, 186 N. J. 127, 141–146, 892 A. 2d 663, 671–674 (2006); Black v. Division of Child Support Enforcement, 686 A. 2d 164, 167–168 (Del. 1996); Mead v. Batchlor, 435 Mich. 480, 488–505, 460 N. W. 2d 493, 496–504 (1990); Ridgway v. Baker, 720 F. 2d 1409, 1413–1415 (CA5 1983) (all finding a federal constitutional right to counsel for indigents facing imprisonment in a child support civil contempt proceeding), with Rodriguez v. Eighth Judicial Dist. Ct., County of Clark, 120 Nev. 798, 808–813, 102 P. 3d 41, 48–51 (2004) (no right to counsel in civil contempt hearing for nonsupport, except in “rarest of cases”); Andrews v. Walton, 428 So. 2d 663, 666 (Fla. 1983) (“no circumstances in which a parent is entitled to court-appointed counsel in a civil contempt proceeding for failure to pay child support”). Compare also In re Grand Jury Proceedings, 468 F. 2d 1368, 1369 (CA9 1972) (per curiam) (general right to counsel in civil contempt proceedings), with Duval v. Duval, 114 N. H. 422, 425–427, 322 A. 2d 1, 3–4 (1974) (no general right, but counsel may be required on case-by-case basis). II Respondents argue that this case is moot. See Massachusetts v. Mellon, 262 U. S. 447, 480 (1923) (Article III judicial power extends only to actual “cases” and “controversies”); Alvarez v. Smith, 558 U. S. __, __ (2009) (slip op., at 4) (“An actual controversy must be extant at all stages of review” (internal quotation marks omitted)). They point out that Turner completed his 12-month prison sentence in 2009. And they add that there are no “collateral consequences” of that particular contempt determination that might keep the dispute alive. Compare Sibron v. New York, 392 U. S. 40, 55–56 (1968) (release from prison does not moot a criminal case because “collateral consequences” are presumed to continue), with Spencer v. Kemna, 523 U. S. 1, 14 (1998) (declining to extend the presumption to parole revocation). The short, conclusive answer to respondents’ mootness claim, however, is that this case is not moot because it falls within a special category of disputes that are “capable of repetition” while “evading review.” Southern Pacific Terminal Co. v. ICC, 219 U. S. 498, 515 (1911). A dispute falls into that category, and a case based on that dispute remains live, if “(1) the challenged action [is] in its duration too short to be fully litigated prior to its cessation or expiration, and (2) there [is] a reasonable expectation that the same complaining party [will] be subjected to the same action again.” Weinstein v. Bradford, 423 U. S. 147, 149 (1975) (per curiam). Our precedent makes clear that the “challenged action,” Turner’s imprisonment for up to 12 months, is “in its duration too short to be fully litigated” through the state courts (and arrive here) prior to its “expiration.” See, e.g., First Nat. Bank of Boston v. Bellotti, 435 U. S. 765, 774 (1978) (internal quotation marks omitted) (18-month period too short); Southern Pacific Terminal Co., supra, at 514–516 (2-year period too short). At the same time, there is a more than “reasonable” likelihood that Turner will again be “subjected to the same action.” As we have pointed out, supra, at 2–3, Turner has frequently failed to make his child support payments. He has been the subject of several civil contempt proceedings. He has been imprisoned on several of those occasions. Within months of his release from the imprisonment here at issue he was again the subject of civil contempt proceedings. And he was again imprisoned, this time for six months. As of December 9, 2010, Turner was $13,814.72 in arrears, and another contempt hearing was scheduled for May 4, 2011. App. 104a; Reply Brief for Petitioner 3, n. 1. These facts bring this case squarely within the special category of cases that are not moot because the underlying dispute is “capable of repetition, yet evading review.” See, e.g., Nebraska Press Assn. v. Stuart, 427 U. S. 539, 546–547 (1976) (internal quotation marks omitted). Moreover, the underlying facts make this case unlike DeFunis v. Odegaard, 416 U. S. 312 (1974) (per curiam), and St. Pierre v. United States, 319 U. S. 41 (1943) (per curiam), two cases that respondents believe require us to find this case moot regardless. DeFunis was moot, but that is because the plaintiff himself was unlikely to again suffer the conduct of which he complained (and others likely to suffer from that conduct could bring their own lawsuits). Here petitioner himself is likely to suffer future imprisonment. St. Pierre was moot because the petitioner (a witness held in contempt and sentenced to five months’ imprisonment) had failed to “apply to this Court for a stay” of the federal-court order imposing imprisonment. 319 U. S., at 42–43. And, like the witness in St. Pierre, Turner did not seek a stay of the contempt order requiring his imprisonment. But this case, unlike St. Pierre, arises out of a state-court proceeding. And respondents give us no reason to believe that we would have (or that we could have) granted a timely request for a stay had one been made. Cf. 28 U. S. C. §1257 (granting this Court jurisdiction to review final state-court judgments). In Sibron, we rejected a similar “mootness” argument for just that reason. 392 U. S., at 53, n. 13. And we find this case similar in this respect to Sibron, not to St. Pierre. III A We must decide whether the Due Process Clause grants an indigent defendant, such as Turner, a right to state-appointed counsel at a civil contempt proceeding, which may lead to his incarceration. This Court’s precedents provide no definitive answer to that question. This Court has long held that the Sixth Amendment grants an indigent defendant the right to state-appointed counsel in a criminal case. Gideon v. Wainwright, 372 U. S. 335 (1963). And we have held that this same rule applies to criminal contempt proceedings (other than summary proceedings). United States v. Dixon, 509 U. S. 688, 696 (1993); Cooke v. United States, 267 U. S. 517, 537 (1925). But the Sixth Amendment does not govern civil cases. Civil contempt differs from criminal contempt in that it seeks only to “coerc[e] the defendant to do” what a court had previously ordered him to do. Gompers v. Bucks Stove & Range Co., 221 U. S. 418, 442 (1911). A court may not impose punishment “in a civil contempt proceeding when it is clearly established that the alleged contemnor is unable to comply with the terms of the order.” Hicks v. Feiock, 485 U. S. 624, 638, n. 9 (1988). And once a civil contemnor complies with the underlying order, he is purged of the contempt and is free. Id., at 633 (he “carr[ies] the keys of [his] prison in [his] own pockets” (internal quotation marks omitted)). Consequently, the Court has made clear (in a case not involving the right to counsel) that, where civil contempt is at issue, the Fourteenth Amendment’s Due Process Clause allows a State to provide fewer procedural protections than in a criminal case. Id., at 637–641 (State may place the burden of proving inability to pay on the defendant). This Court has decided only a handful of cases that more directly concern a right to counsel in civil matters. And the application of those decisions to the present case is not clear. On the one hand, the Court has held that the Fourteenth Amendment requires the State to pay for representation by counsel in a civil “juvenile delinquency” proceeding (which could lead to incarceration). In re Gault, 387 U. S. 1, 35–42 (1967). Moreover, in Vitek v. Jones, 445 U. S. 480, 496–497 (1980), a plurality of four Members of this Court would have held that the Fourteenth Amendment requires representation by counsel in a proceeding to transfer a prison inmate to a state hospital for the mentally ill. Further, in Lassiter v. Department of Social Servs. of Durham Cty., 452 U. S. 18 (1981), a case that focused upon civil proceedings leading to loss of parental rights, the Court wrote that the “pre-eminent generalization that emerges from this Court’s precedents on an indigent’s right to appointed counsel is that such a right has been recognized to exist only where the litigant may lose his physical liberty if he loses the litigation.” Id., at 25. And the Court then drew from these precedents “the presumption that an indigent litigant has a right to appointed counsel only when, if he loses, he may be deprived of his physical liberty.” Id., at 26–27. On the other hand, the Court has held that a criminal offender facing revocation of probation and imprisonment does not ordinarily have a right to counsel at a probation revocation hearing. Gagnon v. Scarpelli, 411 U. S. 778 (1973); see also Middendorf v. Henry, 425 U. S. 25 (1976) (no due process right to counsel in summary court-martial proceedings). And, at the same time, Gault, Vitek, and Lassiter are readily distinguishable. The civil juvenile delinquency proceeding at issue in Gault was “little different” from, and “comparable in seriousness” to, a criminal prosecution. 387 U. S., at 28, 36. In Vitek, the controlling opinion found no right to counsel. 445 U. S., at 499–500 (Powell, J., concurring in part) (assistance of mental health professionals sufficient). And the Court’s statements in Lassiter constitute part of its rationale for denying a right to counsel in that case. We believe those statements are best read as pointing out that the Court previously had found a right to counsel “only” in cases involving incarceration, not that a right to counsel exists in all such cases (a position that would have been difficult to reconcile with Gagnon). B Civil contempt proceedings in child support cases con- stitute one part of a highly complex system designed to assure a noncustodial parent’s regular payment of funds typically necessary for the support of his children. Often the family receives welfare support from a state-administered federal program, and the State then seeks reimbursement from the noncustodial parent. See 42 U. S. C. §§608(a)(3) (2006 ed., Supp. III), 656(a)(1) (2006 ed.); S. C. Code Ann. §§43–5–65(a)(1), (2) (2010 Cum. Supp.). Other times the custodial parent (often the mother, but sometimes the father, a grandparent, or another person with custody) does not receive government benefits and is entitled to receive the support payments herself. The Federal Government has created an elaborate procedural mechanism designed to help both the government and custodial parents to secure the payments to which they are entitled. See generally Blessing v. Freestone, 520 U. S. 329, 333 (1997) (describing the “interlocking set of cooperative federal-state welfare programs” as they relate to child support enforcement); 45 CFR pt. 303 (2010) (prescribing standards for state child support agencies). These systems often rely upon wage withholding, expedited procedures for modifying and enforcing child support orders, and automated data processing. 42 U. S. C. §§666(a), (b), 654(24). But sometimes States will use contempt orders to ensure that the custodial parent receives support payments or the government receives reimbursement. Although some experts have criticized this last-mentioned procedure, and the Federal Government believes that “the routine use of contempt for non-payment of child support is likely to be an ineffective strategy,” the Government also tells us that “coercive enforcement remedies, such as contempt, have a role to play.” Brief for United States as Amicus Curiae 21–22, and n. 8 (citing Dept. of Health and Human Services, National Child Support Enforcement, Strategic Plan: FY 2005–2009, pp. 2, 10). South Carolina, which relies heavily on contempt proceedings, agrees that they are an important tool. We here consider an indigent’s right to paid counsel at such a contempt proceeding. It is a civil proceeding. And we consequently determine the “specific dictates of due process” by examining the “distinct factors” that this Court has previously found useful in deciding what specific safeguards the Constitution’s Due Process Clause re-quires in order to make a civil proceeding fundamentally fair. Mathews v. Eldridge, 424 U. S. 319, 335 (1976) (considering fairness of an administrative proceeding). As relevant here those factors include (1) the nature of “the private interest that will be affected,” (2) the comparative “risk” of an “erroneous deprivation” of that interest with and without “additional or substitute procedural safeguards,” and (3) the nature and magnitude of any countervailing interest in not providing “additional or substitute procedural requirement[s].” Ibid. See also Lassiter, 452 U. S., at 27–31 (applying the Mathews framework). The “private interest that will be affected” argues strongly for the right to counsel that Turner advocates. That interest consists of an indigent defendant’s loss of personal liberty through imprisonment. The interest in securing that freedom, the freedom “from bodily restraint,” lies “at the core of the liberty protected by the Due Process Clause.” Foucha v. Louisiana, 504 U. S. 71, 80 (1992). And we have made clear that its threatened loss through legal proceedings demands “due process protection.” Addington v. Texas, 441 U. S. 418, 425 (1979). Given the importance of the interest at stake, it is ob-viously important to assure accurate decisionmaking in respect to the key “ability to pay” question. Moreover, the fact that ability to comply marks a dividing line between civil and criminal contempt, Hicks, 485 U. S., at 635, n. 7, reinforces the need for accuracy. That is because an incorrect decision (wrongly classifying the contempt proceeding as civil) can increase the risk of wrongful incarceration by depriving the defendant of the procedural protections (including counsel) that the Constitution would demand in a criminal proceeding. See, e.g., Dixon, 509 U. S., at 696 (proof beyond a reasonable doubt, protection from double jeopardy); Codispoti v. Pennsylvania, 418 U. S. 506, 512–513, 517 (1974) (jury trial where the result is more than six months’ imprisonment). And since 70% of child support arrears nationwide are owed by parents with either no reported income or income of $10,000 per year or less, the issue of ability to pay may arise fairly often. See E. Sorensen, L. Sousa, & S. Schaner, Assessing Child Support Arrears in Nine Large States and the Nation 22 (2007) (prepared by The Urban Institute), online at http://aspe.hhs.gov/hsp/07/assessing-CS-debt/report.pdf (as visited June 16, 2011, and available in Clerk of Court’s case file); id., at 23 (“research suggests that many obligors who do not have reported quarterly wages have relatively limited resources”); Patterson, Civil Contempt and the Indigent Child Support Obligor: The Silent Return of Debtor’s Prison, 18 Cornell J. L. & Pub. Pol’y 95, 117 (2008). See also, e.g., McBride v. McBride, 334 N. C. 124, 131, n. 4, 431 S. E. 2d 14, 19, n. 4 (1993) (surveying North Carolina contempt orders and finding that the “failure of trial courts to make a determination of a contemnor’s ability to comply is not altogether infrequent”). On the other hand, the Due Process Clause does not always require the provision of counsel in civil proceedings where incarceration is threatened. See Gagnon, 411 U. S. 778. And in determining whether the Clause requires a right to counsel here, we must take account of opposing interests, as well as consider the probable value of “additional or substitute procedural safeguards.” Mathews, supra, at 335. Doing so, we find three related considerations that, when taken together, argue strongly against the Due Process Clause requiring the State to provide indigents with counsel in every proceeding of the kind before us. First, the critical question likely at issue in these cases concerns, as we have said, the defendant’s ability to pay. That question is often closely related to the question of the defendant’s indigence. But when the right procedures are in place, indigence can be a question that in many—but not all—cases is sufficiently straightforward to warrant determination prior to providing a defendant with counsel, even in a criminal case. Federal law, for example, requires a criminal defendant to provide information showing that he is indigent, and therefore entitled to state-funded counsel, before he can receive that assistance. See 18 U. S. C. §3006A(b). Second, sometimes, as here, the person opposing the defendant at the hearing is not the government represented by counsel but the custodial parent unrepresented by counsel. See Dept. of Health and Human Services, Office of Child Support Enforcement, Understanding Child Support Debt: A Guide to Exploring Child Support Debt in Your State 5, 6 (2004) (51% of nationwide arrears, and 58% in South Carolina, are not owed to the government). The custodial parent, perhaps a woman with custody of one or more children, may be relatively poor, unemployed, and unable to afford counsel. Yet she may have encouraged the court to enforce its order through contempt. Cf. Tr. Contempt Proceedings (Sept. 14, 2005), App. 44a–45a (Rogers asks court, in light of pattern of nonpayment, to confine Turner). She may be able to provide the court with significant information. Cf. id., at 41a–43a (Rogers describes where Turner lived and worked). And the proceeding is ultimately for her benefit. A requirement that the State provide counsel to the noncustodial parent in these cases could create an asymmetry of representation that would “alter significantly the nature of the proceeding.” Gagnon, supra, at 787. Doing so could mean a degree of formality or delay that would unduly slow payment to those immediately in need. And, perhaps more important for present purposes, doing so could make the proceedings less fair overall, increasing the risk of a decision that would erroneously deprive a family of the support it is entitled to receive. The needs of such families play an important role in our analysis. Cf. post, at 10–12 (opinion of Thomas, J.). Third, as the Solicitor General points out, there is available a set of “substitute procedural safeguards,” Mathews, 424 U. S., at 335, which, if employed together, can significantly reduce the risk of an erroneous deprivation of liberty. They can do so, moreover, without incurring some of the drawbacks inherent in recognizing an automatic right to counsel. Those safeguards include (1) notice to the de-fendant that his “ability to pay” is a critical issue in the contempt proceeding; (2) the use of a form (or the equivalent) to elicit relevant financial information; (3) an opportunity at the hearing for the defendant to respond to statements and questions about his financial status, (e.g., those triggered by his responses on the form); and (4) an express finding by the court that the defendant has the ability to pay. See Tr. of Oral Arg. 26–27; Brief for United States as Amicus Curiae 23–25. In presenting these alternatives, the Government draws upon considerable experience in helping to manage statutorily mandated federal-state efforts to enforce child support orders. See supra, at 10. It does not claim that they are the only possible alternatives, and this Court’s cases suggest, for example, that sometimes assistance other than purely legal assistance (here, say, that of a neutral social worker) can prove constitutionally sufficient. Cf. Vitek, 445 U. S., at 499–500 (Powell, J., concurring in part) (provision of mental health professional). But the Government does claim that these alternatives can assure the “fundamental fairness” of the proceeding even where the State does not pay for counsel for an indigent defendant. While recognizing the strength of Turner’s arguments, we ultimately believe that the three considerations we have just discussed must carry the day. In our view, a categorical right to counsel in proceedings of the kind before us would carry with it disadvantages (in the form of unfairness and delay) that, in terms of ultimate fairness, would deprive it of significant superiority over the alternatives that we have mentioned. We consequently hold that the Due Process Clause does not automatically require the provision of counsel at civil contempt proceedings to an indigent individual who is subject to a child support order, even if that individual faces incarceration (for up to a year). In particular, that Clause does not require the provision of counsel where the opposing parent or other custodian (to whom support funds are owed) is not represented by counsel and the State provides alternative procedural safeguards equivalent to those we have mentioned (adequate notice of the importance of ability to pay, fair opportunity to present, and to dispute, relevant information, and court findings). We do not address civil contempt proceedings where the underlying child support payment is owed to the State, for example, for reimbursement of welfare funds paid to the parent with custody. See supra, at 10. Those proceedings more closely resemble debt-collection proceedings. The government is likely to have counsel or some other competent representative. Cf. Johnson v. Zerbst, 304 U. S. 458, 462–463 (1938) (“[T]he average defendant does not have the professional legal skill to protect himself when brought before a tribunal with power to take his life or liberty, wherein the prosecution is presented by experienced and learned counsel” (emphasis added)). And this kind of proceeding is not before us. Neither do we address what due process requires in an unusually complex case where a defendant “can fairly be represented only by a trained advocate.” Gagnon, 411 U. S., at 788; see also Reply Brief for Petitioner 18–20 (not claiming that Turner’s case is especially complex). IV The record indicates that Turner received neither counsel nor the benefit of alternative procedures like those we have described. He did not receive clear notice that his ability to pay would constitute the critical question in his civil contempt proceeding. No one provided him with a form (or the equivalent) designed to elicit information about his financial circumstances. The court did not find that Turner was able to pay his arrearage, but instead left the relevant “finding” section of the contempt order blank. The court nonetheless found Turner in contempt and ordered him incarcerated. Under these circumstances Turner’s incarceration violated the Due Process Clause. We vacate the judgment of the South Carolina Supreme Court and remand the case for further proceedings not inconsistent with this opinion. It is so ordered.

563.US.647

The Speedy Trial Act of 1974 (Act) provides, inter alia, that in “any case in which a plea of not guilty is entered, the trial … shall commence within seventy days” after the arraignment, 18 U. S. C. §3161(c)(1), but lists a number of exclusions from the 70-day period, including “delay resulting from any pretrial motion, from the filing of the motion through the conclusion of the hearing on, or other prompt disposition of, such motion,” §3161(h)(1)(D). Respondent Tinklenberg’s trial on federal drug and gun charges began 287 days after his arraignment. The District Court denied his motion to dismiss the indictment on the ground that the trial violated the Act’s 70-day requirement, finding that 218 of the days fell within various of the Act’s exclusions, leaving 69 nonexcludable days, thus making the trial timely. On Tinklenberg’s appeal from his conviction, the Sixth Circuit agreed that many of the 287 days were excludable, but concluded that 9 days during which three pretrial motions were pending were not, because the motions did not actually cause a delay, or the expectation of delay, of trial. Since these 9 days were sufficient to bring the number of nonexcludable days above 70, the court found a violation of the Act. And given that Tinklenberg had already served his prison sentence, it ordered the indictment dismissed with prejudice. Held: 1. The Act contains no requirement that the filing of a pretrial motion actually caused, or was expected to cause, delay of a trial. Rather, §3161(h)(1)(D) stops the Speedy Trial clock from running automatically upon the filing of a pretrial motion irrespective of whether the motion has any impact on when the trial begins. Pp. 3–12. (a) The Sixth Circuit reasoned that subparagraph (D)’s “delay resulting from” phrase, read most naturally, requires a court to apply the exclusion provision only to motions that actually cause a trial delay, or the expectation of such a delay. While such a reading is linguistically reasonable, it is not the only reasonable interpretation. The subparagraph falls within a general set of provisions introduced by the phrase: “The following periods of delay shall be excluded.” §3161(h). That phrase is followed by a list that includes “[a]ny period of delay resulting from other proceedings concerning the defendant, including… .” §3161(h)(1). This latter list is followed by a sublist, each member (but one) of which is introduced by the phrase “delay resulting from … .” Ibid. Those words are followed by a more specific description, such as “any pretrial motion” from its “filing” “through the conclusion of the hearing on, or other prompt disposition of, such motion.” §3161(h)(1)(D). The whole paragraph can be read as requiring the automatic exclusion of the members of that specific sublist, while referring to those members in general as “periods of delay” and as causing that delay, not because Congress intended the judge to determine causation, but because, in a close to definitional way, the words embody Congress’ own view of the matter. Thus, language alone cannot resolve the basic question presented. Pp. 4–7. (c) Several considerations, taken together, compel the conclusion that Congress intended subparagraph (D) to apply automatically. First, subparagraph (D) and neighboring subparagraphs (F) and (H) contain language that instructs courts to measure the time actually consumed by the specified pretrial occurrence, but those subparagraphs do not mention the date on which the trial begins or was expected to begin. Second, during the 37 years since Congress enacted the statute, every other Court of Appeals has rejected the Sixth Circuit’s interpretation. Third, the Sixth Circuit’s interpretation would make the subparagraph (D) exclusion significantly more difficult to administer, thereby hindering the Act’s efforts to secure fair and efficient trials. Fourth, the Court’s conclusion is reinforced by the difficulty of squaring the Sixth Circuit’s interpretation with the “automatic application” rule expressed in, e.g., Henderson v. United States, 476 U. S. 321, 327. Fifth, the legislative history also supports the Court’s conclusion. Sixth, because all the subparagraphs but one under paragraph (1) begin with the phrase “delay resulting from,” the Sixth Circuit’s interpretation would potentially extend well beyond pretrial motions and encompass such matters as mental and physical competency examinations, interlocutory appeals, consideration of plea agreements, and the absence of essential witnesses. Pp. 7–12. 2. The Sixth Circuit also misinterpreted §3161(h)(1)(F), which excludes from the 70-day calculation “delay resulting from transportation of any defendant … to and from places of examination … , except that any time consumed in excess of ten days … shall be presumed to be unreasonable.” The lower courts agreed that a total of 20 transportation days had elapsed when Tinklenberg was evaluated for competency, and that because the Government provided no justification, all days in excess of the 10 days specified in the statute were unreasonable. However, the Sixth Circuit exempted 8 weekend days and holidays from the count on the theory that subparagraph (F) incorporated Federal Rule of Criminal Procedure 45(a), which, at the time, excluded such days when computing any period specified in “rules” and “court order[s]” that was less than 11 days. Thus, the Circuit considered only two transportation days excessive, and the parties concede that the eight extra days were enough to make the difference between compliance with, and violation of, the Act. This Court exercises its discretion to consider the subsidiary subparagraph (F) question because doing so is fairer to Tinklenberg, who has already served his sentence. In the Court’s view, subparagraph (F) does not incorporate Rule 45. The Act does not say that it does so, the Government gives no good reason for such a reading, and the Rule itself, as it existed at the relevant time, stated it applied to rules and court orders, but said nothing about statutes. The fact that Rule 45 is revised from time to time also argues against its direct application to subparagraph (F) because such changes, likely reflecting considerations other than those related to the Act, may well leave courts treating similar defendants differently. The better reading includes weekend days and holidays in subparagraph (F)’s 10-day period under the common-law rule that such days are included when counting a statutory time period of 10 days unless a statute specifically excludes them. Many courts have treated statutory time periods this way, and Congress has tended specifically to exclude weekend days and holidays from statutory time periods of 10 days when it intended that result. Indeed, Rule 45 has been recently modified to require a similar result. Pp. 12–14. 3. Although the Sixth Circuit’s interpretations of subparagraphs (D) and (F) are both mistaken, the conclusions the court drew from its interpretations in relevant part cancel each other out, such that the court’s ultimate conclusion that Tinklenberg’s trial failed to comply with the Act’s deadline is correct. Pp. 14–15. 579 F. 3d 589, affirmed. Breyer, J., delivered the opinion of the Court, in which Kennedy, Ginsburg, Alito, and Sotomayor, JJ., joined, and in which Roberts, C. J., and Scalia and Thomas, JJ., joined as to Parts I and III. Scalia, J., filed an opinion concurring in part and concurring in the judgment, in which Roberts, C. J., and Thomas, J., joined. Kagan, J., took no part in the consideration or decision of the case.

The Speedy Trial Act of 1974, 18 U. S. C. §3161 et seq., provides that in “any case in which a plea of not guilty is entered, the trial … shall commence within seventy days” from the later of (1) the “filing date” of the information or indictment or (2) the defendant’s initial appearance before a judicial officer (i.e., the arraignment). §3161(c)(1). The Act goes on to list a set of exclusions from the 70-day period, including “delay resulting from any pretrial motion, from the filing of the motion through the conclusion of the hearing on, or other prompt disposition of, such motion.” §3161(h)(1)(D) (2006 ed., Supp. III) (emphasis added). The United States Court of Appeals for the Sixth Circuit held in this case that a pretrial motion falls within this exclusion only if it “actually cause[s] a delay, or the expectation of a delay, of trial.” 579 F. 3d 589, 598 (2009). In our view, however, the statutory exclusion does not contain this kind of causation requirement. Rather, the filing of a pretrial motion falls within this provision irrespective of whether it actually causes, or is expected to cause, delay in starting a trial. I Jason Louis Tinklenberg, the respondent, was convicted of violating federal drug and gun laws. 18 U. S. C. §922(g)(1) (felon in possession of a firearm); 21 U. S. C. §843(a)(6) (possession of items used to manufacture a controlled substance). He made his initial appearance before a judicial officer on October 31, 2005, and the Speedy Trial clock then began to run. His trial began on August 14, 2006, 287 days later. Just before trial, Tinklenberg asked the District Court to dismiss the indictment on the ground that the trial came too late, vio-lating the Speedy Trial Act’s 70-day requirement. The District Court denied the motion after finding that 218 of the 287 days fell within various Speedy Trial Act exclusions, leaving 69 nonexcludable days, thereby making the trial timely. On appeal the Sixth Circuit agreed with the District Court that many of the 287 days were excludable. But it disagreed with the District Court about the excludability of time related to three pretrial motions. The Government filed the first motion, an unopposed motion to conduct a video deposition of a witness, on August 1, 2006; the District Court disposed of the motion on August 3, 2006. The Government filed the second motion, an unopposed motion to bring seized firearms into the courtroom as evidence at trial, on August 8, 2006; the District Court disposed of the motion on August 10, 2006. Tinklenberg filed the third motion, a motion to dismiss the indictment under the Speedy Trial Act, on August 11, 2006; the District Court denied that motion on August 14, 2006. In the Sixth Circuit’s view, the nine days during which the three motions were pending were not excludable because the motions did not “actually cause a delay, or the expectation of delay, of trial.” 579 F. 3d, at 598. Because these 9 days were sufficient to bring the number of nonexcludable days above 70, the Court of Appeals found a violation of the Act. And given the fact that Tinklenberg had already served his prison sentence, it ordered the District Court to dismiss the indictment with prejudice. We granted certiorari at the Government’s request in order to review the Sixth Circuit’s motion-by-motion causation test. We now reverse its determination. But because we agree with the defendant about a subsidiary matter, namely, the exclusion of certain holidays and weekend days during the period in which he was transported for a competency examination, id., at 597, we affirm the Court of Appeals’ ultimate conclusion. II A In relevant part the Speedy Trial Act sets forth a basic rule: “In any case in which a plea of not guilty is entered, the trial of a defendant … shall commence within seventy days from [the later of (1)] the filing date … of the information or indictment, or . . . [(2)] the date the defendant has appeared before a judicial officer of the court in which such charge is pending … .” §3161(c)(1) (2006 ed.). The Act then says that the “following periods of delay shall be excluded in computing … the time within which the trial . . . must commence.” §3161(h) (2006 ed., Supp. III). It lists seven such “periods of delay.” It describes the first of these seven excludable periods as “(1) Any period of delay resulting from other proceedings concerning the defendant including but not limited to— “(A) delay resulting from any proceeding … to determine the mental competency or physical capacity of the defendant; “(B) delay resulting from trial with respect to other charges … ; “(C) delay resulting from any interlocutory appeal; “(D) delay resulting from any pretrial motion, from the filing of the motion through the conclusion of the hearing on, or other prompt disposition of, such motion; “(E) delay resulting from any proceeding relating to the transfer of a case [or defendant] . . . from another district … ; “(F) delay resulting from transportation of any defendant from another district, or to and from places of examination or hospitalization, except that any time consumed in excess of ten days … shall be presumed to be unreasonable; “(G) delay resulting from consideration by the court of a proposed plea agreement …; “(H) delay reasonably attributable to any period, not to exceed thirty days, during which any proceeding concerning the defendant is actually under advisement by the court.” Ibid. (2006 ed. and Supp. III) (emphasis added). B The particular provision before us, subparagraph (D), excludes from the Speedy Trial period “delay resulting from any pretrial motion, from the filing of the motion through the conclusion of the hearing on, or other prompt disposition of, such motion.” §3161(h)(1)(D). The question is whether this provision stops the Speedy Trial clock from running automatically upon the filing of a pretrial motion irrespective of whether the motion has any impact on when the trial begins. Unlike the Sixth Circuit, we believe the answer to this question is yes. We begin with the Act’s language. The Sixth Circuit based its answer primarily upon that language. It argued that the phrase “delay resulting from,” read most naturally, requires a court to apply the exclusion provision only to those “motion[s]” that “actually cause a delay, or the expectation of a delay, of trial.” 579 F. 3d, at 598. We agree that such a reading is linguistically reasonable, but the Court of Appeals wrote that there “is no conceivable way to read this language other than to require a delay to result from any pretrial motion before excludable time occurs.” Ibid. See also ibid. (“[T]he statute is clear”). And here we disagree. When the Court of Appeals says that its reading is the only way any reasonable person could read this language, it overstates its claim. For one thing, even though the word “delay” ordinarily indicates a postponement, it need not inevitably do so. Compare The American Heritage Dictionary 480 (4th ed. 2000) (“[t]o postpone until a later time” or “[t]o cause to be later or slower than expected or desired”) with ibid. (“[t]he interval of time between two events”). In any event, terms must be read in their statutory context in order to determine how the provision in question should be applied in an individual case. Statutory language that describes a particular circumstance, for example, might require a judge to examine each individual case to see if that circumstance is present. But, alternatively, it might ask a judge instead to look at more general matters, such as when a statute requires a judge to increase the sentence of one convicted of a “crime of violence” without requiring the judge to determine whether the particular crime at issue in a particular case was committed in a violent manner. See Taylor v. United States, 495 U. S. 575, 602 (1990) (“crime of violence” characterizes the generic crime, not the particular act committed). Similarly a statute that forbids the importation of “wild birds” need not require a court to decide whether a particular parrot is, in fact, wild or domesticated. It may intend to place the entire species within that definition without investigation of the characteristics of an individual specimen. See United States v. Fifty-Three (53) Eclectus Parrots, 685 F. 2d 1131, 1137 (CA9 1982). More than that, statutory language can sometimes specify that a set of circumstances exhibits a certain characteristic virtually as a matter of definition and irrespective of how a court may view it in a particular case. A statute that describes “extortion” as a “crime of violence” makes that fact so by definition, without asking a court to second-guess Congress about the matter. 18 U. S. C. §924(e)(2)(B)(ii) (2006 ed.) (defining “violent felony” to include extortion for purposes of the Armed Career Criminal Act). The statute before us, though more complex, can be read similarly. The pretrial motion subparagraph falls within a general set of provisions introduced by the phrase: “The following periods of delay shall be excluded.” §3161(h) (2006 ed., Supp. III). That phrase is then followed by a list that includes “[a]ny period of delay resulting from other proceedings concerning the defendant, including … .” §3161(h)(1). This latter list is followed by a sublist, each member (but one) of which is introduced by the phrase “delay resulting from … ,” ibid. (2006 ed. and Supp. III), which words are followed by a more specific description, such as “any pretrial motion” from its “filing” “through the conclusion of the hearing on, or other prompt disposition of, such motion.” §3161(h)(1)(D) (2006 ed., Supp. III). The whole paragraph can be read as requiring the automatic exclusion of the members of that specific sublist, while referring to those members in general as “periods of delay” and as causing that delay, not because Congress intended the judge to determine causation, but because, in a close to definitional way, the words embody Congress’ own view of the matter. It is not farfetched to describe the members of the specific sublist in the statute before us in this definitional sense—as “periods of delay” or as bringing about delay. After all, the exclusion of any of the specific periods described always delays the expiration of the 70-day Speedy Trial deadline. Or Congress might have described the specific periods listed in paragraph (1) as “periods of delay” and “delay[s] resulting from” simply because periods of the type described often do cause a delay in the start of trial. Both explanations show that, linguistically speaking, one can read the statutory exclusion as automatically applying to the specific periods described without leaving to the district court the task of determining whether the period described would or did actually cause a postponement of the trial in the particular case. Thus, language alone cannot resolve the basic question presented in this case. But when read in context and in light of the statute’s structure and purpose, we think it clear that Congress intended subparagraph (D) to apply automatically. C We now turn to several considerations, which, taken together, convince us that the subparagraphs that specifically list common pretrial occurrences apply automatically in the way we have just described. First, subparagraph (D) clarifies that the trial court should measure the period of excludable delay for a pretrial motion “from the filing of the motion through the conclusion of the hearing on, or other prompt disposition of such motion,” but nowhere does it mention the date on which the trial begins or was expected to begin. §3161(h)(1)(D) (2006 ed., Supp. III). Thus, it is best read to instruct measurement of the time actually consumed by consideration of the pretrial mo-tion. Two other related subparagraphs contain clarifying language that contemplates measurement of the time actually consumed by the specified pretrial occurrence without regard to the commencement of the trial. See §3161(h)(1)(F) (“Any time consumed in excess of ten days from the date an order of removal or an order directing such transportation, and the defendant’s arrival at the destination shall be presumed to be unreasonable”); §3161(h)(1)(H) (“delay reasonably attributable to any period, not to exceed thirty days, during which any proceeding concerning the defendant is actually under advisement by the court”). If “delay” truly referred to the postponement of trial, then presumably those subparagraphs would instruct that excludable periods should be measured from the date that trial was otherwise scheduled to begin. Second, we are impressed that during the 37 years since Congress enacted the Speedy Trial Act, every Court of Appeals has considered the question before us now, and every Court of Appeals, implicitly or explicitly, has rejected the interpretation that the Sixth Circuit adopted in this case. See United States v. Wilson, 835 F. 2d 1440, 1443 (CADC 1987) (explicit), abrogated on other grounds by Bloate v. United States, 559 U. S. ___ (2010); United States v. Hood, 469 F. 3d 7, 10 (CA1 2006) (explicit); United States v. Cobb, 697 F. 2d 38, 42 (CA2 1982) (explicit), abrogated on other grounds by Henderson v. United States, 476 U. S. 321 (1986); United States v. Novak, 715 F. 2d 810, 813 (CA3 1983) (explicit) abrogated on other grounds by Henderson v. United States, 476 U. S. 321 (1986); United States v. Dorlouis, 107 F. 3d 248, 253–254 (CA4 1997) (explicit); United States v. Green, 508 F. 3d 195, 200 (CA5 2007) (explicit); United States v. Montoya, 827 F. 2d 143, 151 (CA7 1987) (explicit); United States v. Titlbach, 339 F. 3d 692, 698 (CA8 2003) (implicit); United States v. Van Brandy, 726 F. 2d 548, 551 (CA9 1984) (explicit); United States v. Vogl, 374 F. 3d 976, 985–986 (CA10 2004) (explicit); United States v. Stafford, 697 F. 2d 1368, 1371–1372 (CA11 1983) (explicit). This unanimity among the lower courts about the meaning of a statute of great practical administrative importance in the daily working lives of busy trial judges is itself entitled to strong consideration, particularly when those courts have maintained that interpretation consistently over a long a period of time. See General Dynamics Land Systems, Inc. v. Cline, 540 U. S. 581, 593–594 (2004). Third, the Sixth Circuit’s interpretation would make the subparagraph (D) exclusion significantly more difficult to administer. And in doing so, it would significantly hinder the Speedy Trial Act’s efforts to secure fair and efficient criminal trial proceedings. See Zedner v. United States, 547 U. S. 489, 497 (2006) (noting that the Act’s exceptions provide “necessary flexibility”); H. R. Rep. No. 93–1508, p. 15 (1974) (the Act seeks to achieve “efficiency in the processing of cases which is commensurate with due process”); S. Rep. No. 93–1021, p. 21 (1974). Trial judges may, for example, set trial dates beyond 70 days in light of other commitments. And in doing so, a trial judge may well be aware, based on his or her experience, that pretrial motions will likely consume the extra time—even though the judge may know little about which specific motions will be filed, when, and how many. How is that judge to apply the Sixth Circuit’s approach, particularly when several, including unanticipated, pretrial proceedings did consume the time in question? Moreover, what is to happen if several excludable and several nonexcludable potential causes of delay (e.g., pre-trial motions to take depositions, potential scheduling conflicts, various health examinations, etc.) coincide, particularly in multidefendant cases? Can the judge, motion by motion, decide which motions were responsible and which were not responsible for postponing what otherwise might have been an earlier trial date? And how is a defendant or his attorney to predict whether or when a judge will later find a particular motion to have caused a postponement of trial? And if the matter is difficult to predict, how is the attorney to know when or whether he or she should seek further postponement of the 70-day deadline? With considerable time and judicial effort, perhaps through the use of various presumptions, courts could find methods for overcoming these and other administrative difficulties. In some instances, the judge may know at the time of filing that a given motion is easily resolved or that its complexity will almost certainly postpone the trial. Judges could note on the record their predictions about whether the motion will postpone trial at the time that the motion is filed. Parties could also stipulate as to whether a given motion would be excluded from the Speedy Trial clock. But those theoretical strategies would not prevent all or even most mistakes, needless dismissals of indictments, and potential retrials after appeal—all of which exact a toll in terms of the fairness of and confidence in the criminal justice system. And any such future strategies for administering the Sixth Circuit’s rule cannot provide a present justification for turning the federal judicial system away from the far less obstacle-strewn path that the system has long traveled. Fourth, we are reinforced in our conclusion by the difficulty of squaring the Sixth Circuit’s interpretation with this Court’s precedent. In Henderson v. United States, 476 U. S. 321 (1986), the Court rejected the contention that the exclusion provision for pretrial motions governs only reasonable delays. The Court there concluded (as the Court of Appeals had held) that the exclusion “was intended to be automatic.” Id., at 327 (quoting United States v. Henderson, 746 F. 2d 619, 622 (CA9 1984); internal quotation marks omitted). See also Bloate v. United States, 559 U. S. ___ (2010) (holding based in part on the view that the exclusion applies “automatically” to the specified period of delay). Henderson did not consider whether a trial court must determine whether the pretrial motion actually caused postponement of the trial in each individual case. But the Sixth Circuit’s interpretation would nonetheless significantly limit the premise of “auto-matic application” upon which the case rests. Fifth, for those who find legislative history useful, it is worthwhile noting, (as this Court noted in Henderson) that the Senate Report concerning the reenactment of the provision in 1979 described it, along with the other provisions in §3161(h)(1), as referring to “specific and recurring periods of time often found in criminal cases,” and characterized them as “automatically excludable delay,” S. Rep. No. 96–212, p. 9 (1979). See H. R. Rep. No. 93–1508, at 21 (“The time limits would be tolled by hearings, proceedings and necessary delay which normally occur prior to the trial of criminal cases” (emphasis added)); S. Rep. No. 93–1021, at 21 (“[The Act] has carefully constructed exclusions and exceptions which permit normal pre-trial preparation in the ordinary noncomplex cases which represent the bulk of business in the Federal courts”). But cf. id., at 35 (paragraph (h)(1) excludes “[d]elays caused by proceedings relating to the defendant” (emphasis added)). Sixth, because all the subparagraphs but one under paragraph (1) begin with the phrase “delay resulting from,” the Sixth Circuit’s interpretation would potentially extend well beyond pretrial motions and encompass such matters as mental and physical competency examinations, interlocutory appeals, consideration of plea agreements, and the absence of essential witnesses. See §3161(h)(1) (2006 ed., Supp. III); §3161(h)(3)(A) (2006 ed.). Given the administrative complexity the causation requirement would bring about in all these areas, those Circuits that have considered a causation requirement in respect to these other matters have rejected it. See, e.g., United States v. Pete, 525 F. 3d 844, 852 (CA9 2008) (interlocutory appeal); United States v. Miles, 290 F. 3d 1341, 1350 (CA11 2002) (unavailability of essential witnesses); United States v. Robinson, 887 F. 2d 651, 656–657 (CA6 1989) (trial on other charges). That further complexity, along with these lower court holdings, reinforce our conclusion. We consequently disagree with the Sixth Circuit that the Act’s exclusion requires a court to find that the event the exclusion specifically describes, here the filing of the pretrial motion, actually caused or was expected to cause delay of a trial. We hold that the Act contains no such requirement. III Tinklenberg also argues that the Sixth Circuit wrongly interpreted a different exclusion provision, this time the provision excluding “delay resulting from transportation of any defendant from another district, or to and from places of examination or hospitalization, except that any time consumed in excess of ten days from the date an order of removal or an order directing such transportation, and the defendant’s arrival at the destination shall be presumed to be unreasonable.” §3161(h)(1)(F) (2006 ed., Supp. III) (emphasis added). The District Court granted Tinklenberg’s request for a competency evaluation and he was transported to a medical facility for examination. The lower courts agreed that a total of 20 transportation days elapsed and that since the Government provided no justification, all days in excess of the 10 days specified in the statute were unreasonable. But in counting those excess days, the court exempted weekend days and holidays. Since Veterans Day, Thanksgiving Day, and three weekends all fell within the 20-day period, only 2 days, not 10 days, were considered excessive, during which the 70-day Speedy Trial Act clock continued to tick. Tinklenberg argues that subparagraph (F) does not exempt weekend days and holidays; hence the court should have considered 10, not 2, days to be excessive. And the parties concede that those 8 extra ticking days are enough to make the difference between compliance with, and violation of, the Act. As the Solicitor General notes, we may consider, or “decline to entertain,” alternative grounds for affirmance. See United States v. Nobles, 422 U. S. 225, 242, n. 16 (1975). In this case, we believe it treats Tinklenberg, who has already served his sentence, more fairly to consider the alternative ground and thereby more fully to dispose of the case. The Sixth Circuit exempted weekend days and holidays because it believed that subparagraph (F) incorporated Federal Rule of Criminal Procedure 45(a). At the relevant time, that Rule excluded weekend days and holidays when computing any period of time specified in the “rules,” in “any local rule,” or in “any court order” that was less than 11 days. Fed. Rule Crim. Proc. 45(a) (2005). But in our view subparagraph (F) does not incorporate Rule 45. The Act does not say that it incorporates Rule 45. The Government has given us no good reason for reading it as incorporating the Rule. And the Rule itself, as it existed at the relevant time, said that it applied to “rules” and to “orders,” but it said nothing about statutes. Other things being equal, the fact that Rule 45 is revised from time to time also argues against its direct application to subparagraph (F). That is because those changes, likely reflecting considerations other than those related to the Speedy Trial Act, may well leave courts treating similar defendants differently. Without relying upon a cross-reference to Rule 45, we believe the better reading of subparagraph (F) would include weekend days and holidays in its 10-day time period. Under the common-law rule, weekend days and holidays are included when counting a statutory time period of 10 days unless the statute specifically excludes them. See 74 Am. Jur. 2d, Time §22, p. 589 (2001) (in calculating time periods expressed in statutes, “when the time stipulated must necessarily include one or more Saturdays, Sundays, or holidays, those days will not be excluded, in the absence of an express proviso for their exclusion”). Many courts have treated statutory time periods this way. See, e.g., Howeisen v. Chapman, 195 Ind. 381, 383–384, 145 N. E. 487, 488 (1924); American Tobacco Co. v. Strickling, 88 Md. 500, 508–511, 41 A. 1083, 1086 (1898). And Congress has tended specifically to exclude weekend days and holidays from statutory time periods of 10 days when it intended that result. Compare 18 U. S. C. §3142(d)(2) (permitting the temporary detention of certain defendants “for a period of not more than ten days, excluding Saturdays, Sundays, and holidays”) and 5 U. S. C. §552a(d)(2)(A) (requiring an agency to acknowledge receipt of a request to amend agency records within “10 days (excluding Saturdays, Sundays, and legal public holidays)”) with 18 U. S. C. §2518(9) (establishing a 10-day period for disclosing applications for and court orders authorizing wiretaps without specifically excluding weekends and holidays) and §4244(a) (providing a 10-day period after conviction for filing a motion to request mental health treatment without specifically excluding weekends and holidays). Indeed, Rule 45 has been recently modified so that now (though not at the time of Tinklenberg’s proceedings) it requires a similar result. Fed. Rule Crim. Proc. 45(a)(1) (2010) (instructing that weekend days and holidays are to be counted when calculating all time periods, including statutory time periods for which no alternative method of computing time is specified). * * * We disagree with the Sixth Circuit’s interpretation of both subparagraph (D) and subparagraph (F), and now hold that its interpretations of those two provisions are mistaken. Nonetheless the conclusions the court drew from those two interpretations in relevant part cancel each other out such that the court’s ultimate conclusion that Tinklenberg’s trial failed to comply with the Speedy Trial Act’s deadline is correct. Therefore, the Sixth Circuit’s judgment ordering dismissal of the indictment on remand is Affirmed. Justice Kagan took no part in the consideration or decision of this case.

564.US.2010_10-382

Respondent Jicarilla Apache Nation’s (Tribe) reservation contains natural resources that are developed pursuant to statutes administered by the Interior Department. Proceeds from these resources are held by the United States in trust for the Tribe. The Tribe filed a breach-of-trust action in the Court of Federal Claims (CFC), seeking monetary damages for the Government’s alleged mismanagement of the Tribe’s trust funds in violation of 25 U. S. C. §§161–162a and other laws. During discovery, the Tribe moved to compel production of certain documents. The Government agreed to release some of the documents, but asserted that others were protected by, inter alia, the attorney-client privilege. The CFC granted the motion in part, holding that departmental communications relating to the management of trust funds fall within a “fiduciary exception” to the attorney-client privilege. Under that exception, which courts have applied to common-law trusts, a trustee who obtains legal advice related to trust administration is precluded from asserting the attorney-client privilege against trust beneficiaries. Denying the Government’s petition for a writ of mandamus directing the CFC to vacate its production order, the Federal Circuit agreed with the CFC that the trust relationship between the United States and the Indian tribes is sufficiently similar to a private trust to justify applying the fiduciary exception. The appeals court held that the United States cannot deny a tribe’s request to discover communications between the Government and its attorneys based on the attorney-client privilege when those communications concern management of an Indian trust and the Government has not claimed that it or its attorneys considered a specific competing interest in those communications. Held: The fiduciary exception to the attorney-client privilege does not apply to the general trust relationship between the United States and the Indian tribes. Pp. 5–24. (a) The Court considers the bounds of the fiduciary exception and the nature of the Indian trust relationship. Pp. 5–14. (1) Under English common law, when a trustee obtained legal advice to guide his trust administration and not for his own defense in litigation, the beneficiaries were entitled to the production of documents related to that advice on the rationale that the advice was sought for their benefit and obtained at their expense in that trust funds were used to pay the attorney. In the leading American case, Riggs Nat. Bank of Washington, D. C. v. Zimmer, 355 A. 2d 709, the Delaware Chancery Court applied the fiduciary exception to hold that trust beneficiaries could compel trustees to produce a legal memorandum related to the trust’s administration because: (1) the trustees had obtained the legal advice as “mere representative[s]” of the beneficiaries, who were the “real clients” of the attorney, id., at 711–712, and (2) the fiduciary duty to furnish trust-related information to the beneficiaries outweighed the trustees’ interest in the attorney-client privilege, id., at 714. The Federal Courts of Appeals apply the fiduciary exception based on the same two criteria. Pp. 6–9. (2) The Federal Circuit analogized the Government to a private trustee. While the United States’ responsibilities with respect to the management of tribal funds bear some resemblance to those of a private trustee, this analogy cannot be taken too far. The Government’s trust obligations to the tribes are established and governed by statute, not the common law, see, e.g., United States v. Navajo Nation, 537 U. S. 488, 506 (Navajo I), and in fulfilling its statutory duties, the Government acts not as a private trustee, but pursuant to its sovereign interest in the execution of federal law, see, e.g., Heckman v. United States, 224 U. S. 413, 437. Once federal law imposes fiduciary obligations on the Government, the common law “could play a role,” United States v. Navajo Nation, 556 U. S. ___, ___ (Navajo II); e.g., to inform the interpretation of statutes, see United States v. White Mountain Apache Tribe, 537 U. S. 465, 475–476. But the applicable statutes and regulations control. When “the Tribe cannot identify a specific, applicable, trust-creating statute or regulation that the Government violated … neither the Government’s ‘control’ over [Indian assets] nor common-law trust principles matter.” Navajo II, supra, at ___. Pp. 9–14. (b) The two criteria justifying the fiduciary exception are absent in the trust relationship between the United States and Indian tribes. Pp. 14–23. (1) In cases applying the fiduciary exception, courts identify the “real client” based on whether the advice was bought by the trust corpus, whether the trustee had reason to seek advice in a personal rather than a fiduciary capacity, and whether the advice could have been intended for any purpose other than to benefit the trust. Riggs, 355 A. 2d, at 711–712. Applying these factors, the Court concludes that the United States does not obtain legal advice as a “mere representative” of the Tribe; nor is the Tribe the “real client” for whom that advice is intended. See id., at 711. Here, the Government attorneys are paid out of congressional appropriations at no cost to the Tribe. The Government also seeks legal advice in its sovereign capacity rather than as a conventional fiduciary of the Tribe. Because its sovereign interest is distinct from the beneficiaries’ private interests, the Government seeks legal advice in a personal, not a fiduciary, capacity. Moreover, the Government has too many competing legal concerns to allow a case-by-case inquiry into each communication’s purpose. In addition to its duty to the Tribe, the Government may need to comply with other statutory duties, such as environmental and conservation obligations. It may also face conflicting duties to different tribes or individual Indians. It may seek the advice of counsel for guidance in balancing these competing interests or to help determine whether there are conflicting interests at all. For the attorney-client privilege to be effective, it must be predictable. See, e.g., Jaffee v. Redmond, 518 U. S. 1, 18. The Government will not always be able to predict what considerations qualify as competing interests, especially before receiving counsel’s advice. If the Government were required to identify the specific interests it considered in each communication, its ability to receive confidential legal advice would be substantially compromised. See Upjohn Co. v. United States, 449 U. S. 383, 393. Pp. 15–20. (2) The Federal Circuit also decided that the fiduciary exception properly applied here because of the fiduciary’s duty to disclose all trust-management-related information to the beneficiary. The Government, however, does not have the same common-law disclosure obligations as a private trustee. In this case, 25 U. S. C. §162a(d) delineates the Government’s “trust responsibilities.” It identifies the Interior Secretary’s obligation to supply tribal account holders “with periodic statements of their account performance” and to make “available on a daily basis” their account balances, §162a(d)(5). The Secretary has complied with these requirements in regulations mandating that each tribe be provided with a detailed quarterly statement of performance. 25 CFR pt. 115.8. The common law of trusts does not override these specific trust-creating statutes and regulations. A statutory clause labeling the enumerated trust responsibilities as nonexhaustive, see §162a(d), cannot be read to include a general common-law duty to disclose all information related to the administration of Indian trusts, since that would vitiate Congress’ specification of narrowly defined disclosure obligations, see, e.g., Mackey v. Lanier Collection Riggs Agency & Service, Inc., 486 U. S. 825, 837. By law and regulation, moreover, the documents at issue are classed “the property of the United States” while other records are “the property of the tribe.” 25 CFR §115.1000. This Court considers ownership of records to be a significant factor in deciding who “ought to have access to the document,” Riggs, supra, at 712. Here, that privilege belongs to the United States. Pp. 20–23. 590 F. 3d 1305, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, and Thomas, JJ., joined. Ginsburg, J., filed an opinion concurring in the judgment, in which Breyer, J., joined. Sotomayor, J., filed a dissenting opinion. Kagan, J., took no part in the consideration or decision of the case.

The attorney-client privilege ranks among the oldest and most established evidentiary privileges known to our law. The common law, however, has recognized an exception to the privilege when a trustee obtains legal advice related to the exercise of fiduciary duties. In such cases, courts have held, the trustee cannot withhold attorney-client communications from the beneficiary of the trust. In this case, we consider whether the fiduciary exception applies to the general trust relationship between the United States and the Indian tribes. We hold that it does not. Although the Government’s responsibilities with re-spect to the management of funds belonging to Indian tribes bear some resemblance to those of a private trustee, this analogy cannot be taken too far. The trust obligations of the United States to the Indian tribes are established and governed by statute rather than the common law, and in fulfilling its statutory duties, the Government acts not as a private trustee but pursuant to its sovereign interest in the execution of federal law. The reasons for the fiduciary exception—that the trustee has no independent interest in trust administration, and that the trustee is subject to a general common-law duty of disclosure—do not apply in this context. I The Jicarilla Apache Nation (Tribe) occupies a 900,000-acre reservation in northern New Mexico that was established by Executive Order in 1887. The land contains timber, gravel, and oil and gas reserves, which are developed pursuant to statutes administered by the Department of the Interior. Proceeds derived from these natural resources are held by the United States in trust for the Tribe pursuant to the American Indian Trust Fund Management Reform Act of 1994, 108 Stat. 4239, and other statutes. In 2002, the Tribe commenced a breach-of-trust action against the United States in the Court of Federal Claims (CFC). The Tribe sued under the Tucker Act, 28 U. S. C. §1491 (2006 ed. and Supp. III), and the Indian Tucker Act, §1505, which vest the CFC with jurisdiction over claims against the Government that are founded on the Constitution, laws, treaties, or contracts of the United States. The complaint seeks monetary damages for the Government’s alleged mismanagement of funds held in trust for the Tribe. The Tribe argues that the Government violated various laws, including 25 U. S. C. §§161a and 162a, that govern the management of funds held in trust for Indian tribes. See 88 Fed. Cl. 1, 3 (2009). From December 2002 to June 2008, the Government and the Tribe participated in alternative dispute resolution in order to resolve the claim. During that time, the Government turned over thousands of documents but withheld 226 potentially relevant documents as protected by the attorney-client privilege, the attorney work-product doctrine, or the deliberative-process privilege. In 2008, at the request of the Tribe, the case was restored to the active litigation docket. The CFC divided the case into phases for trial and set a discovery schedule. The first phase, relevant here, concerns the Government’s management of the Tribe’s trust accounts from 1972 to 1992. The Tribe alleges that during this period the Government failed to invest its trust funds properly. Among other things, the Tribe claims the Government failed to maximize returns on its trust funds, invested too heavily in short-term maturities, and failed to pool its trust funds with other tribal trusts. During discovery, the Tribe moved to compel the Government to produce the 226 withheld documents. In response, the Government agreed to withdraw its claims of deliberative-process privilege and, accordingly, to produce 71 of the documents. But the Government continued to assert the attorney-client privilege and attorney work-product doctrine with respect to the remaining 155 documents. The CFC reviewed those documents in camera and classified them into five cate-gories: (1) requests for legal advice relating to trust administration sent by personnel at the Department of the Interior to the Office of the Solicitor, which directs legal affairs for the Department, (2) legal advice sent from the Solicitor’s Office to personnel at the Interior and Treas- ury Departments, (3) documents generated under contracts between Interior and an accounting firm, (4) Interior documents concerning litigation with other tribes, and (5) miscellaneous documents not falling into the other categories. The CFC granted the Tribe’s motion to compel in part. The CFC held that communications relating to the management of trust funds fall within a “fiduciary exception” to the attorney-client privilege. Under that exception, which courts have applied in the context of common-law trusts, a trustee who obtains legal advice related to the execution of fiduciary obligations is precluded from asserting the attorney-client privilege against beneficiaries of the trust. The CFC concluded that the trust relationship between the United States and the Indian tribes is sufficiently analogous to a common-law trust relationship that the exception should apply. Accordingly, the CFC held, the United States may not shield from the Tribe communications with attorneys relating to trust matters. The CFC ordered disclosure of almost all documents in the first two categories because those documents “involve matters regarding the administration of tribal trusts, either directly or indirectly implicating the investments that benefit Jicarilla” and contain “legal advice relating to trust administration.” Id., at 14–15. The CFC allowed the Government to withhold most of the documents in the remaining categories as attorney work product,[Footnote 1] but the court identified some individual documents that it determined were also subject to the fiduciary exception. Id., at 18–19. The Government sought to prevent disclosure of the documents by petitioning the Court of Appeals for the Federal Circuit for a writ of mandamus directing the CFC to vacate its production order. The Court of Appeals denied the petition because, in its view, the CFC correctly applied the fiduciary exception. The court held that “the United States cannot deny an Indian tribe’s request to discover communications between the United States and its attorneys based on the attorney-client privilege when those communications concern management of an Indian trust and the United States has not claimed that the government or its attorneys considered a specific com-peting interest in those communications.” In re United States, 590 F. 3d 1305, 1313 (CA Fed. 2009). In qualifying its holding, the court recognized that sometimes the Government may have other statutory obligations that clash with its fiduciary duties to the Indian tribes. But because the Government had not alleged that the legal advice in this case related to such conflicting interests, the court reserved judgment on how the fiduciary exception might apply in that situation. The court rejected the Government’s argument that, because its duties to the Indian tribes were governed by statute rather than the common law, it had no general duty of disclosure that would override the attorney-client privilege. The court also disagreed with the Government’s contention that a case-by-case approach made the attorney-client privilege too unpredictable and would impair the Government’s ability to obtain confidential legal advice. We granted certiorari, 562 U. S. ___ (2011),[Footnote 2] and now reverse and remand for further proceedings. II The Federal Rules of Evidence provide that evidentiary privileges “shall be governed by the principles of the common law … in the light of reason and experience.” Fed. Rule Evid. 501. The attorney-client privilege “is the oldest of the privileges for confidential communications known to the common law.” Upjohn Co. v. United States, 449 U. S. 383, 389 (1981) (citing 8 J. Wigmore, Evidence §2290 (J. McNaughton rev. 1961)). Its aim is “to encourage full and frank communication between attorneys and their clients and thereby promote broader public interests in the observance of law and administration of justice.” 449 U. S., at 389; Hunt v. Blackburn, 128 U. S. 464, 470 (1888). The objectives of the attorney-client privilege apply to governmental clients. “The privilege aids government en-tities and employees in obtaining legal advice founded on a complete and accurate factual picture.” 1 Restatement (Third) of the Law Governing Lawyers §74, Comment b, pp. 573–574 (1998). Unless applicable law provides other-wise, the Government may invoke the attorney-client privilege in civil litigation to protect confidential communications between Government officials and Government attorneys. Id., at 574 (“[G]overnmental agencies and employees enjoy the same privilege as nongovernmental counterparts”). The Tribe argues, however, that the common law also recognizes a fiduciary exception to the attorney-client privilege and that, by virtue of the trust relationship between the Government and the Tribe, documents that would otherwise be privileged must be disclosed. As preliminary matters, we consider the bounds of the fiduciary exception and the nature of the trust relationship between the United States and the Indian tribes. A English courts first developed the fiduciary exception as a principle of trust law in the 19th century. The rule was that when a trustee obtained legal advice to guide the administration of the trust, and not for the trustee’s own defense in litigation, the beneficiaries were entitled to the production of documents related to that advice. Wynne v. Humberston, 27 Beav. 421, 423–424, 54 Eng. Rep. 165, 166 (1858); Talbot v. Marshfield 2 Dr. & Sm. 549, 550–551, 62 Eng. Rep. 728, 729 (1865). The courts reasoned that the normal attorney-client privilege did not apply in this situation because the legal advice was sought for the beneficiaries’ benefit and was obtained at the beneficiaries’ expense by using trust funds to pay the attorney’s fees. Ibid.; Wynne, supra, at 423–424, 54 Eng. Rep., at 166. The fiduciary exception quickly became an established feature of English common law, see, e.g., In re Mason, 22 Ch. D. 609 (1883), but it did not appear in this country until the following century. American courts seem first to have expressed skepticism. See In re Prudence-Bonds Corp., 76 F. Supp. 643, 647 (EDNY 1948) (declining to apply the fiduciary exception to the trustee of a bondholding corporation because of the “important right of such a corporate trustee … to seek legal advice and nevertheless act in accordance with its own judgment”). By the 1970’s, however, American courts began to adopt the English common-law rule. See Garner v. Wolfinbarger, 430 F. 2d 1093, 1103–1104 (CA5 1970) (allowing shareholders, upon a showing of “good cause,” to discover legal advice given to corporate management).[Footnote 3] The leading American case on the fiduciary exception is Riggs Nat. Bank of Washington, D. C. v. Zimmer, 355 A. 2d 709 (Del. Ch. 1976). In that case, the beneficiaries of a trust estate sought to compel the trustees to reimburse the estate for alleged breaches of trust. The beneficiaries moved to compel the trustees to produce a legal memorandum related to the administration of the trust that the trustees withheld on the basis of attorney-client privilege. The Delaware Chancery Court, observing that “American case law is practically nonexistent on the duty of a trustee in this context,” looked to the English cases. Id., at 712. Applying the common-law fiduciary exception, the court held that the memorandum was discoverable. It identified two reasons for applying the exception. First, the court explained, the trustees had obtained the legal advice as “mere representative[s]” of the beneficiaries because the trustees had a fiduciary obligation to act in the beneficiaries’ interest when administering the trust. Ibid. For that reason, the beneficiaries were the “real clients” of the attorney who had advised the trustee on trust-related matters, and therefore the attorney-client privilege properly belonged to the beneficiaries rather than the trustees. Id., at 711–712. The court based its “real client” determination on several factors: (1) when the advice was sought, no adversarial proceedings between the trustees and beneficiaries had been pending, and therefore there was no reason for the trustees to seek legal advice in a personal rather than a fiduciary capacity; (2) the court saw no indication that the memorandum was intended for any purpose other than to benefit the trust; and (3) the law firm had been paid out of trust assets. That the advice was obtained at the beneficiaries’ expense was not only a “significant factor” entitling the beneficiaries to see the document but also “a strong indication of precisely who the real clients were.” Id., at 712. The court distinguished between “legal advice procured at the trustee’s own expense and for his own protection,” which would remain privileged, “and the situation where the trust itself is assessed for obtaining opinions of counsel where interests of the beneficiaries are presently at stake.” Ibid. In the latter case, the fiduciary exception applied, and the trustees could not withhold those attorney-client communications from the beneficiaries. Second, the court concluded that the trustees’ fiduciary duty to furnish trust-related information to the benefi-ciaries outweighed their interest in the attorney-client privilege. “The policy of preserving the full disclosure necessary in the trustee-beneficiary relationship,” the court explained, “is here ultimately more important than the protection of the trustees’ confidence in the attorney for the trust.” Id., at 714. Because more information helped the beneficiaries to police the trustees’ management of the trust, disclosure was, in the court’s judgment, “a weight-ier public policy than the preservation of confidential attorney-client communications.” Ibid. The Federal Courts of Appeals apply the fiduciary exception based on the same two criteria. See, e.g., In re Long Island Lighting Co., 129 F. 3d 268, 272 (CA2 1997); Wachtel v. Health Net, Inc., 482 F. 3d 225, 233–234 (CA3 2007); Solis v. Food Employers Labor Relations Assn., 2011 U. S. App. LEXIS 9110, *12 (CA4, May 4, 2011); Wildbur v. Arco Chemical Co., 974 F. 2d 631, 645 (CA5 1992); United States v. Evans, 796 F. 2d 264, 265–266 (CA9 1986) (per curiam). Not until the decision below had a federal appellate court held the exception to apply to the United States as trustee for the Indian tribes. B In order to apply the fiduciary exception in this case, the Court of Appeals analogized the Government to a private trustee. 590 F. 3d, at 1313. We have applied that analogy in limited contexts, see, e.g., United States v. Mitchell, 463 U. S. 206, 226 (1983) (Mitchell II), but that does not mean the Government resembles a private trustee in every respect. On the contrary, this Court has previously noted that the relationship between the United States and the Indian tribes is distinctive, “different from that existing between individuals whether dealing at arm’s length, as trustees and beneficiaries, or otherwise.” Klamath and Moadoc Tribes v. United States, 296 U. S. 244, 254 (1935) (emphasis added). “The general relationship between the United States and the Indian tribes is not comparable to a private trust relationship.” Cherokee Nation of Okla. v. United States, 21 Cl. Ct. 565, 573 (1990) (emphasis added). The Government, of course, is not a private trustee. Though the relevant statutes denominate the relationship between the Government and the Indians a “trust,” see, e.g., 25 U. S. C. §162a, that trust is defined and governed by statutes rather than the common law. See United States v. Navajo Nation, 537 U. S. 488, 506 (2003) (Navajo I) (“[T]he analysis must train on specific rights-creating or duty-imposing statutory or regulatory prescriptions”). As we have recognized in prior cases, Congress may style its relations with the Indians a “trust” without assuming all the fiduciary duties of a private trustee, creating a trust relationship that is “limited” or “bare” compared to a trust relationship between private parties at common law. United States v. Mitchell, 445 U. S. 535, 542 (1980) (Mitchell I); Mitchell II, supra, at 224.[Footnote 4] The difference between a private common-law trust and the statutory Indian trust follows from the unique position of the Government as sovereign. The distinction between “public rights” against the Government and “private rights” between private parties is well established. The Government consents to be liable to private parties “and may yield this consent upon such terms and under such restrictions as it may think just.” Murray’s Lessee v. Hoboken Land & Improvement Co., 18 How. 272, 283 (1856). This creates an important distinction “between cases of private right and those which arise between the Government and persons subject to its authority in connection with the performance of the constitutional functions of the executive or legislative departments.” Crowell v. Benson, 285 U. S. 22, 50 (1932). Throughout the history of the Indian trust relationship, we have recognized that the organization and management of the trust is a sovereign function subject to the plenary authority of Congress. See Merrion v. Jicarilla Apache Tribe, 455 U. S. 130, 169, n. 18 (1982) (“The United States retains plenary authority to divest the tribes of any attributes of sovereignty”); United States v. Wheeler, 435 U. S. 313, 319 (1978) (“Congress has plenary authority to legislate for the Indian tribes in all matters, including their form of government”); Winton v. Amos, 255 U. S. 373, 391 (1921) (“Congress has plenary authority over the Indians and all their tribal relations, and full power to legislate concerning their tribal property”); Lone Wolf v. Hitchcock, 187 U. S. 553, 565 (1903) (“Plenary authority over the tribal relations of the Indians has been exercised by Congress from the beginning, and the power has always been deemed a political one, not subject to be controlled by the judicial department of the government”); Cherokee Nation v. Hitchcock, 187 U. S. 294, 308 (1902) (“The power existing in Congress to administer upon and guard the tribal property, and the power being political and administrative in its nature, the manner of its exercise is a question within the province of the legislative branch to determine, and is not one for the courts”); see also United States v. Candelaria, 271 U. S. 432, 439 (1926); Tiger v. Western Investment Co., 221 U. S. 286, 315 (1911). Because the Indian trust relationship represents an exercise of that authority, we have explained that the Government “has a real and direct interest” in the guardianship it exercises over the Indian tribes; “the interest is one which is vested in it as a sovereign.” United States v. Minnesota, 270 U. S. 181, 194 (1926). This is especially so because the Government has often structured the trust relationship to pursue its own policy goals. Thus, while trust administration “relat[es] to the welfare of the Indians, the maintenance of the limitations which Congress has prescribed as a part of its plan of distribution is distinctly an interest of the United States.” Heckman v. United States, 224 U. S. 413, 437 (1912); see also Candelaria, supra, at 443–444. In Heckman, the Government brought suit to cancel certain conveyances of allotted lands by members of an Indian tribe because the conveyances violated restrictions on alienation imposed by Congress. This Court explained that the Government brought suit as the representative of the very Indian grantors whose conveyances it sought to cancel, and those Indians were thereby bound by the judgment. 224 U. S., at 445–446. But while it was formally acting as a trustee, the Government was in fact asserting its own sovereign interest in the disposition of Indian lands, and the Indians were precluded from intervening in the litigation to advance a position contrary to that of the Government. Id., at 445. Such a result was possible because the Government assumed a fiduciary role over the Indians not as a common-law trustee but as the governing authority enforcing statutory law. We do not question “the undisputed existence of a general trust relationship between the United States and the Indian people.” Mitchell II, 463 U. S., at 225. The Government, following “a humane and self imposed policy … has charged itself with moral obligations of the highest responsibility and trust,” Seminole Nation v. United States, 316 U. S. 286, 296–297 (1942), obligations “to the fulfillment of which the national honor has been committed,” Heckman, supra, at 437. Congress has expressed this policy in a series of statutes that have defined and redefined the trust relationship between the United States and the Indian tribes. In some cases, Congress established only a limited trust relationship to serve a narrow purpose. See Mitchell I, 445 U. S., at 544 (Congress intended the United States to hold land “ ‘in trust’ ” under the General Allotment Act “simply because it wished to prevent alienation of the land and to ensure that allottees would be immune from state taxation”); Navajo I, 537 U. S., at 507–508 (Indian Mineral Leasing Act imposes no “detailed fiduciary responsibilities” nor is the Government “expressly invested with responsibility to secure ‘the needs and best interests of the Indian owner’ ”). In other cases, we have found that particular “statutes and regulations … clearly establish fiduciary obligations of the Government” in some areas. Mitchell II, supra, at 226; see also United States v. White Mountain Apache Tribe, 537 U. S. 465, 475 (2003). Once federal law imposes such duties, the common law “could play a role.” United States v. Navajo Nation, 556 U. S. ___, ___ (2009) (Navajo II) (slip op., at 14). We have looked to common-law principles to inform our interpretation of statutes and to determine the scope of liability that Congress has imposed. See White Mountain Apache Tribe, supra, at 475–476. But the applicable statutes and regulations “establish [the] fiduciary relationship and define the contours of the United States’ fiduciary responsibilities.” Mitchell II, supra, at 224. When “the Tribe cannot identify a specific, applicable, trust-creating statute or regulation that the Government violated, … neither the Government’s ‘control’ over [Indian assets] nor common-law trust principles matter.” Navajo II, supra, at ___ (slip op., at 14).[Footnote 5] The Government assumes Indian trust responsibilities only to the extent it expressly accepts those responsibilities by statute.[Footnote 6] Over the years, we have described the federal relationship with the Indian tribes using various formulations. The Indian tribes have been called “domestic dependent nations,” Cherokee Nation v. Georgia, 5 Pet. 1, 17 (1831), under the “tutelage” of the United States, Heckman, supra, at 444, and subject to “the exercise of the Government’s guardianship over . . . their affairs,” United States v. Sandoval, 231 U. S. 28, 48 (1913). These concepts do not necessarily correspond to a common-law trust relationship. See, e.g., Restatement 2d, §7 (“A guardianship is not a trust”). That is because Congress has chosen to structure the Indian trust relationship in different ways. We will apply common-law trust principles where Congress has indicated it is appropriate to do so. For that reason, the Tribe must point to a right conferred by statute or regulation in order to obtain otherwise privileged information from the Government against its wishes. III In this case, the Tribe’s claim arises from 25 U. S. C. §§161–162a and the American Indian Trust Fund Management Reform Act of 1994, §4001 et seq. These pro-visions define “the trust responsibilities of the United States” with respect to tribal funds. §162a(d). The Court of Appeals concluded that the trust relationship between the United States and the Indian tribes, outlined in these and other statutes, is “sufficiently similar to a private trust to justify applying the fiduciary exception.” 590 F. 3d, at 1313. We disagree. As we have discussed, the Government exercises its carefully delimited trust responsibilities in a sovereign capacity to implement national policy respecting the Indian tribes. The two features justifying the fiduciary exception—the beneficiary’s status as the “real client” and the trustee’s common-law duty to disclose information about the trust—are notably absent in the trust relationship Congress has established between the United States and the Tribe. A The Court of Appeals applied the fiduciary exception based on its determination that the Tribe rather than the Government was the “real client” with respect to the Government attorneys’ advice. Ibid. In cases applying the fiduciary exception, courts identify the “real client” based on whether the advice was bought by the trust corpus, whether the trustee had reason to seek advice in a personal rather than a fiduciary capacity, and whether the advice could have been intended for any purpose other than to benefit the trust. Riggs, 355 A. 2d, at 711–712. Applying these factors, we conclude that the United States does not obtain legal advice as a “mere representative” of the Tribe; nor is the Tribe the “real client” for whom that advice is intended. See ibid. Here, the Government attorneys are paid out of congressional appropriations at no cost to the Tribe. Courts look to the source of funds as a “strong indicator of precisely who the real clients were” and a “significant factor” in determining who ought to have access to the legal advice. Id., at 712. We similarly find it significant that the attorneys were paid by the Government for advice regarding the Government’s statutory obligations. The payment structure confirms our view that the Government seeks legal advice in its sovereign capacity rather than as a conventional fiduciary of the Tribe. Undoubtedly, Congress intends the Indian tribes to benefit from the Government’s management of tribal trusts. That intention represents “a humane and self imposed policy” based on felt “moral obligations.” Seminole Nation, 316 U. S., at 296–297. This statutory purpose does not imply a full common-law trust, however. Cf. Restatement 2d, §25, Comment b (“No trust is created if the settlor manifests an intention to impose merely a moral obligation”). Congress makes such policy judgments pursuant to its sovereign governing authority, and the implementation of federal policy remains “distinctly an interest of the United States.” Heckman, 224 U. S., at 437.[Footnote 7] We have said that “the United States continue[s] as trustee to have an active interest” in the disposition of Indian assets because the terms of the trust relationship embody policy goals of the United States. McKay v. Kalyton, 204 U. S. 458, 469 (1907). In some prior cases, we have found that the Government had established the trust relationship in order to impose its own policy on Indian lands. See Mitchell I, 445 U. S., at 544 (Congress “intended that the United States ‘hold the land … in trust’ … because it wished to prevent alienation of the land”). In other cases, the Government has invoked its trust relationship to prevent state interference with its policy toward the Indian tribes. See Minnesota v. United States, 305 U. S. 382, 386 (1939); Candelaria, 271 U. S., at 442–444; United States v. Kagama, 118 U. S. 375, 382–384 (1886). And the exercise of federal authority thereby established has often been “left under the acts of Congress to the discretion of the Executive Department.” Heckman, supra, at 446. In this way, Congress has designed the trust relationship to serve the interests of the United States as well as to benefit the Indian tribes. See United States v. Rickert, 188 U. S. 432, 443 (1903) (trust relationship “ ‘authorizes the adoption on the part of the United States of such policy as their own public interests may dictate’ ” (quoting Choctaw Nation v. United States, 119 U. S. 1, 28 (1886))).[Footnote 8] We cannot agree with the Tribe and its amici that “[t]he government and its officials who obtained the advice have no stake in [the] substance of the advice, beyond their trustee role,” Brief for Respondent 9, or that “the United States’ interests in trust administration were identical to the interests of the tribal trust fund beneficiaries,” Brief for National Congress of American Indians et al. as Amici Curiae 5. The United States has a sovereign interest in the administration of Indian trusts distinct from the private interests of those who may benefit from its administration. Courts apply the fiduciary exception on the ground that “management does not manage for itself.” Garner, 430 F. 2d, at 1101; Wachtel, 482 F. 3d, at 232 (“[O]f central importance in both Garner and Riggs was the fiduciary’s lack of a legitimate personal interest in the legal advice obtained”). But the Government is never in that position. While one purpose of the Indian trust relationship is to benefit the tribes, the Government has its own independent interest in the implementation of federal Indian policy. For that reason, when the Government seeks legal advice related to the administration of tribal trusts, it establishes an attorney-client relationship related to its sovereign interest in the execution of federal law. In other words, the Government seeks legal advice in a “personal” rather than a fiduciary capacity. See Riggs, 355 A. 2d, at 711. Moreover, the Government has too many competing legal concerns to allow a case-by-case inquiry into the purpose of each communication. When “multiple interests” are involved in a trust relationship, the equivalence between the interests of the beneficiary and the trustee breaks down. Id., at 714. That principle applies with particular force to the Government. Because of the multiple interests it must represent, “the Government cannot follow the fastidious standards of a private fiduciary, who would breach his duties to his single beneficiary solely by representing potentially conflicting interests without the beneficiary’s consent.” Nevada v. United States, 463 U. S. 110, 128 (1983). As the Court of Appeals acknowledged, the Government may be obliged “to balance competing interests” when it administers a tribal trust. 590 F. 3d, at 1315. The Government may need to comply with other statutory duties, such as the environmental and conservation obligations that the Court of Appeals discussed. See id., at 1314–1315. The Government may also face conflicting obligations to different tribes or individual Indians. See, e.g., Nance v. EPA, 645 F. 2d 701, 711 (CA9 1981) (Federal Government has “conflicting fiduciary responsibilities” to the Northern Cheyenne and Crow Tribes); Hoopa Valley Tribe v. Christie, 812 F. 2d 1097, 1102 (CA9 1986) (“No trust relation exists which can be discharged to the plaintiff here at the expense of other Indians”). Within the bounds of its “general trust relationship” with the Indian people, we have recognized that the Government has “discretion to reorder its priorities from serving a subgroup of beneficiaries to serving the broader class of all Indians nationwide.” Lincoln v. Vigil, 508 U. S. 182, 195 (1993); see also ibid. (“Federal Government ‘does have a fiduciary obligation to the Indians; but it is a fiduciary obligation that is owed to all Indian tribes’ ” (quoting Hoopa Valley Tribe, supra, at 1102)). And sometimes, we have seen, the Government has enforced the trust statutes to dispose of Indian property contrary to the wishes of those for whom it was nominally kept in trust. The Government may seek the advice of counsel for guidance in balancing these competing interests. Indeed, the point of consulting counsel may be to determine whether conflicting interests are at stake. The Court of Appeals sought to accommodate the Government’s multiple obligations by suggesting that the Government may invoke the attorney-client privilege if it identifies “a specific competing interest” that was considered in the particular communications it seeks to withhold. 590 F. 3d, at 1313. But the conflicting interests the Government must consider are too pervasive for such a case-by-case approach to be workable. We have said that for the attorney-client privilege to be effective, it must be predictable. See Jaffee v. Redmond, 518 U. S. 1, 18 (1996); Upjohn, 449 U. S., at 393. If the Government were required to identify the specific interests it considered in each communication, its ability to receive confidential legal advice would be substantially compromised. The Government will not always be able to predict what considerations qualify as a “specific competing interest,” especially in advance of receiving counsel’s advice. Forcing the Government to monitor all the considerations contained in each communication with counsel would render its attorney-client privilege “little better than no privilege at all.” Ibid. B The Court of Appeals also decided the fiduciary exception properly applied to the Government because “the fiduciary has a duty to disclose all information related to trust management to the beneficiary.” 590 F. 3d, at 1312. In general, the common-law trustee of an irrevocable trust must produce trust-related information to the beneficiary on a reasonable basis, though this duty is sometimes limited and may be modified by the settlor. Restatement (Third) of Trusts §82 (2005) (hereinafter Restatement 3d); Bogert §§962, 965.[Footnote 9] The fiduciary exception applies where this duty of disclosure overrides the attorney-client privilege. United States v. Mett, 178 F. 3d 1058, 1063 (CA9 1999) (“[T]he fiduciary exception can be understood as an instance of the attorney-client privilege giving way in the face of a competing legal principle”). The United States, however, does not have the same common-law disclosure obligations as a private trustee. As we have previously said, common-law principles are relevant only when applied to a “specific, applicable, trust-creating statute or regulation.” Navajo II, 556 U. S., at ___ (slip op., at 14). The relevant statute in this case is 25 U. S. C. §162a(d), which delineates “trust responsibilities of the United States” that the Secretary of the Interior must discharge. The enumerated responsibilities include a provision identifying the Secretary’s obligation to provide specific information to tribal account holders: The Secretary must “suppl[y] account holders with periodic statements of their account performance” and must make “available on a daily basis” the “balances of their account.” §162a(d)(5). The Secretary has complied with these requirements by adopting regulations that instruct the Office of Trust Fund Management to provide each tribe with a quarterly statement of performance, 25 CFR §115.801 (2010), that identifies “the source, type, and status of the trust funds deposited and held in a trust account; the beginning balance; the gains and losses; receipts and disbursements; and the ending account balance of the quarterly statement period,” §115.803. Tribes may request more frequent statements or further “information about account transactions and balances.” §115.802. The common law of trusts does not override the specific trust-creating statute and regulations that apply here. Those provisions define the Government’s disclosure ob-ligation to the Tribe. The Tribe emphasizes, Brief for Respondent 34, that the statute identifies the list of trust responsibilities as nonexhaustive. See §162a(d) (trust responsibilities “are not limited to” those enumerated). The Government replies that this clause “is best read to refer to other statutory and regulatory requirements” rather than to common-law duties. Brief for United States 38. Whatever Congress intended, we cannot read the clause to include a general common-law duty to disclose all information related to the administration of Indian trusts. When Congress provides specific statutory obligations, we will not read a “catchall” provision to impose general obligations that would include those specifically enumerated. Massachusetts Mut. Life Ins. Co. v. Russell, 473 U. S. 134, 141–142 (1985). “As our cases have noted in the past, we are hesitant to adopt an interpretation of a congressional enactment which renders superfluous another portion of that same law.” Mackey v. Lanier Collection Agency & Service, Inc., 486 U. S. 825, 837 (1988). Reading the statute to incorporate the full duties of a private, common-law fiduciary would vitiate Congress’ specification of narrowly defined disclosure obligations.[Footnote 10] By law and regulation, moreover, the documents at issue in this case are classed “the property of the United States” while other records are “the property of the tribe.” 25 CFR §115.1000 (2010); see also §§15.502, 162.111, 166.1000. Just as the source of the funds used to pay for legal advice is highly relevant in identifying the “real client” for purposes of the fiduciary exception, we consider ownership of the resulting records to be a significant factor in deciding who “ought to have access to the document.” See Riggs, 355 A. 2d, at 712. In this case, that privilege belongs to the United States.[Footnote 11] * * * Courts and commentators have long recognized that “[n]ot every aspect of private trust law can properly govern the unique relationship of tribes and the federal government.” Cohen §5.02[2], at 434–435. The fiduciary exception to the attorney-client privilege ranks among those aspects inapplicable to the Government’s administration of Indian trusts. The Court of Appeals denied the Government’s petition for a writ of mandamus based on its erroneous view to the contrary. We leave it for that court to determine whether the standards for granting the writ are met in light of our opinion.[Footnote 12] We therefore reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 The CFC held that there is no fiduciary exception to the work-product doctrine. 88 Fed. Cl. 1, 12 (2009). The Court of Appeals did not address that issue, In re United States, 590 F. 3d 1305, 1313 (CA Fed. 2009), and it is not before us. Footnote 2 After the Federal Circuit denied the Government’s mandamus petition, the Government produced the documents under a protective order that prevents disclosure to third parties until the case is resolved by this Court. App. to Pet. for Cert. 93a–97a. The Government’s compliance with the production order does not affect our review. Our decision may still provide effective relief by preventing further disclosure and by excluding the evidence from trial. See Mohawk Industries, Inc. v. Carpenter, 558 U. S. ___, ___ (2009) (slip op., at 8). Footnote 3 Today, “[c]ourts differ on whether the [attorney-client] privilege is available for communications between the trustee and counsel regarding the administration of the trust.” A. Newman, G. Bogert & G. Bogert, Law of Trusts and Trustees §962, p. 68 (3d ed. 2010) (hereinafter Bogert). Some state courts have altogether rejected the notion that the attorney-client privilege is subject to a fiduciary exception. See, e.g., Huie v. DeShazo, 922 S. W. 2d 920, 924 (Tex. 1996) (“The attorney-client privilege serves the same important purpose in the trustee-attorney relationship as it does in other attorney-client relationships”); Wells Fargo Bank v. Superior Ct., 22 Cal. 4th 201, 208–209, 990 P. 2d 591, 595 (2000) (“[T]he attorney for the trustee of a trust is not, by virtue of this relationship, also the attorney for the beneficiaries of the trust” (internal quotation marks omitted)). Neither party before this Court disputes the existence of a common-law fiduciary exception, however, so in deciding this case we assume such an exception exists. Footnote 4 “There are a number of widely varying relationships which more or less closely resemble trusts, but which are not trusts, although the term ‘trust’ is sometimes used loosely to cover such relationships. It is important to differentiate trusts from these other relationships, since many of the rules applicable to trusts are not applicable to them.” Restatement (Second) of Trusts §4, Introductory Note, p. 15 (1957) (hereinafter Restatement 2d); see also Begay v. United States, 16 Cl. Ct. 107, 127, n. 17 (1987) (“[T]he provisions relating to private trustees and fiduciaries, while useful as analogies, cannot be regarded as finally dispositive in a government-Indian trustee-fiduciary relationship”). Footnote 5 Thus, the dissent’s reliance on the Government’s “managerial control,” post, at 8 (opinion of Sotomayor, J.), is misplaced. Footnote 6 Cf. Restatement 2d, §25, Comment a (“[A]lthough the settlor has called the transaction a trust[,] no trust is created unless he manifests an intention to impose duties which are enforceable in the courts”). Footnote 7 Chief Justice Hughes, writing for a unanimous Court, insisted that the “national interest” in the management of Indian affairs “is not to be expressed in terms of property, or to be limited to the assertion of rights incident to the ownership of a reversion or to the holding of a technical title in trust.” Heckman, 224 U. S., at 437. Footnote 8 Congress has structured the trust relationship to reflect its considered judgment about how the Indians ought to be governed. For example, the Indian General Allotment Act of 1887, 24 Stat. 388, was “a comprehensive congressional attempt to change the role of Indians in American society.” F. Cohen, Handbook of Federal Indian Law §1.04, p. 77 (2005) (hereinafter Cohen). Congress aimed to promote the assimilation of Indians by dividing Indian lands into individually owned allotments. The federal policy aimed “to substitute a new in-dividual way of life for the older Indian communal way.” Id., at 79. The Indian Reorganization Act of 1934, 48 Stat. 984, marked a shift away “from assimilation policies and toward more tolerance and respect for traditional aspects of Indian culture.” Cohen §1.05, at 84. The Act prohibited further allotment and restored tribal ownership. Id., at 86. The Indian Self-Determination and Education Assistance Act of 1975, 88 Stat. 2203, and the Tribal Self-Governance Act of 1994, 108 Stat. 4270, enabled tribes to run health, education, economic development, and social programs for themselves. Cohen §1.07, at 103. This strengthened self-government supported Congress’ decision to authorize tribes to withdraw trust funds from Federal Government control and place the funds under tribal control. American Indian Trust Fund Management Reform Act of 1994, 108 Stat. 4239, 4242–4244; see 25 U. S. C. §§4021–4029 (2006 ed. and Supp. III). The control over the Indian tribes that has been exercised by the United States pursuant to the trust relationship—forcing the division of tribal lands, restraining alienation—does not correspond to the fiduciary duties of a common-law trustee. Rather, the trust relationship has been altered and administered as an instrument of federal policy. Footnote 9 We assume for the sake of argument that an Indian trust is properly analogized to an irrevocable trust rather than to a revocable trust. A revocable trust imposes no duty of the trustee to disclose information to the beneficiary. “[W]hile a trust is revocable, only the person who may revoke it is entitled to receive information about it from the trustee.” Bogert §962, at 25, §964; Restatement 3d, §74, Comment e, at 31 (“[T]he trustee of a revocable trust is not to provide reports or accountings or other information concerning the terms or administration of the trust to other beneficiaries without authorization either by the settlor or in the terms of the trust or a statute”). In many respects, Indian trusts resemble revocable trusts at common law because Congress has acted as the settlor in establishing the trust and retains the right to alter the terms of the trust by statute, even in derogation of tribal property interests. See Winton v. Amos, 255 U. S. 373, 391 (1921) (“It is thoroughly established that Congress has plenary authority over the Indians … and full power to legislate concerning their tribal property”); Cohen §5.02[4], at 401–403. The Government has not advanced the argument that the relationship here is similar to a revocable trust, and the point need not be addressed to resolve this case. Footnote 10 Our reading of 25 U. S. C. §162a(d) receives additional support from another statute in which Congress expressed its understanding that the Government retains evidentiary privileges allowing it to withhold information related to trust property from Indian tribes. The Indian Claims Limitation Act of 1982, 96 Stat. 1976, addressed Indian claims that the claimants desired to have litigated by the United States. If the Secretary of the Interior decided to reject a claim for litigation, he was required to furnish a report to the affected Indian claimants and, upon their request, to provide “any nonprivileged research materials or evidence gathered by the United States in the documentation of such claim.” Id., at 1978. That Congress authorized the withholding of information on grounds of privilege makes us doubt that Congress understood the Government’s trust obligations to override so basic a privilege as that between attorney and client. Footnote 11 The dissent tells us that applying the fiduciary exception is even more important against the Government than against a private trustee because of a “history of governmental mismanagement.” Post, at 21. While it is not necessary to our decision, we note that the Indian tribes are not required to keep their funds in federal trust. See 25 U. S. C. §4022 (authorizing tribes to withdraw funds held in trust by the United States); 25 CFR pt. 1200(B). If the Tribe wishes to have its funds managed by a “conventional fiduciary,” post, at 10, it may seek to do so. Footnote 12 If the Court of Appeals declines to issue the writ, we assume that the CFC on remand will follow our holding here regarding the applicability of the fiduciary exception in the present context.

563.US.307

Respondent Tohono O’odham Nation (Nation) filed suit in Federal District Court against federal officials who managed tribal assets held in trust by the Federal Government, alleging violations of fiduciary duty and requesting equitable relief. The next day, the Nation filed this action against the United States in the Court of Federal Claims (CFC), alleging almost identical violations and requesting money damages. The CFC case was dismissed under 28 U. S. C. §1500, which bars CFC jurisdiction over a claim if the plaintiff has another suit “for or in respect to” that claim pending against the United States or its agents in another court. The Federal Circuit reversed, finding that the two suits were not for or in respect to the same claim because, although they shared operative facts, they did not seek overlapping relief. Held: 1. Two suits are for or in respect to the same claim, precluding CFC jurisdiction, if they are based on substantially the same operative facts, regardless of the relief sought in each suit. Pp. 2–9. (a) Since 1868, Congress has restricted the jurisdiction of the CFC and its predecessors when related actions are pending elsewhere. Keene Corp. v. United States, 508 U. S. 200, 212, held that two suits are for or in respect to the same claim when they are “based on substantially the same operative facts … , at least if there [is] some overlap in the relief requested,” but it reserved the question whether the jurisdictional bar operates if suits based on the same operative facts do not seek overlapping relief. The rule now codified in §1500 was first enacted to curb duplicate lawsuits by residents of the Confederacy who, in seeking to recover for cotton taken by the Federal Government, sued the Government in the Court of Claims and, at the same time, sued federal officials in other courts, seeking tort relief for the same actions. Section 1500’s robust response to this problem bars CFC jurisdiction not only if the plaintiff sues on an identical claim elsewhere, but also if the other action is related but not identical. The phrase “in respect to” does not resolve all doubt as to the bar’s scope, but it suggests a broad prohibition, regardless of whether “claim” carries a special or limited meaning. Pp. 2–4. (b) Keene permits two constructions of “for or in respect to” the same claim, one based on facts alone and the other on factual plus remedial overlap. The former is the more reasonable interpretation in light of the statute’s use of a similar phrase in a way consistent only with factual overlap. The CFC bar applies where the other action is against a “person who, … when the cause of action … arose, was, in respect thereto, acting” under color of federal law. But at the time that a cause of action arose, the person could not act in respect to the relief requested, for no complaint was yet filed. Although the phrase at issue involves a “claim” rather than a cause of action, there is reason to think that both phrases refer to facts alone and not to relief. As Keene explained, “ ‘claim’ is used here synonymously with ‘cause of action,’ ” 508 U. S., at 210. And if the phrase that uses “cause of action,” the more technical term, does not embrace the concept of remedy, it is reasonable to conclude that neither phrase does. Pp. 4–5. (c) This reading also makes sense in light of the CFC’s unique remedial powers. Because the CFC is the only judicial forum for most nontort requests for significant monetary relief against the United States and because it has no general power to provide equitable relief against the Government or its officers, a statute aimed at precluding duplicate CFC suits would be unlikely to require remedial overlap. Remedial overlap was even more unusual when §1500’s rule was first enacted in 1868. The Federal Circuit could identify no purpose the statute served in light of that court’s precedent. But courts should not render statutes nugatory through construction. The statute’s purpose is clear from its origins—the need to save the Government from redundant litigation—and the conclusion that two suits are for or in respect to the same claim when they share substantially the same operative facts allows the statute to achieve that aim. Concentrating on operative facts is also consistent with the doctrine of claim preclusion, or res judicata. The Nation errs in arguing that this Court’s interpretation unjustly forces plaintiffs to choose between partial remedies available in different courts. The Nation could have recovered any losses in the CFC alone. Even if some hardship were shown, this Court “enjoy[s] no ‘liberty to add an exception … to remove apparent hardship.’ ” Keene, supra, at 217–218. Pp. 5–9. 2. The substantial overlap in operative facts between the Nation’s District Court and CFC suits precludes jurisdiction in the CFC. Both actions allege that the United States holds the same assets in trust for the Nation’s benefit, and they describe almost identical breaches of fiduciary duty. Pp. 9–10. 559 F. 3d 1284, reversed and remanded. Kennedy, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Thomas, and Alito, JJ., joined. Sotomayor, J., filed an opinion concurring in the judgment, in which Breyer, J., joined. Ginsburg, J., filed a dissenting opinion. Kagan, J., took no part in the consideration or decision of the case.

The Tohono O’odham Nation is an Indian Tribe with federal recognition. The Nation’s main reservation is in the Sonoran desert of southern Arizona. Counting this and other reservation lands, the Nation’s landholdings are approximately 3 million acres. The Nation brought two actions based on the same alleged violations of fiduciary duty with respect to the Nation’s lands and other assets. One action was filed against federal officials in district court and the other against the United States in the Court of Federal Claims (CFC). The Court of Appeals for the Federal Circuit held that the CFC suit was not barred by the rule that the CFC lacks jurisdiction over an action “for or in respect to” a claim that is also the subject of an action pending in another court. 28 U. S. C. §1500. The question presented is whether a common factual basis like the one apparent in the Nation’s suits suffices to bar jurisdiction under §1500. I The case turns on the relationship between the two suits the Nation filed. The first suit was filed in the United States District Court for the District of Columbia against federal officials responsible for managing tribal assets held in trust by the Federal Government. The complaint alleged various violations of fiduciary duty with respect to those assets. The Nation claimed, for example, that the officials failed to provide an accurate accounting of trust property; to refrain from self-dealing; or to use reasonable skill in investing trust assets. The complaint requested equitable relief, including an accounting. The next day the Nation filed the instant action against the United States in the CFC. The CFC complaint described the same trust assets and the same fiduciary duties that were the subject of the District Court complaint. And it alleged almost identical violations of fiduciary duty, for which it requested money damages. The CFC case was dismissed under §1500 for want of jurisdiction. A divided panel of the Court of Appeals for the Federal Circuit reversed. 559 F. 3d 1284 (2009). Two suits are for or in respect to the same claim, it reasoned, only if they share operative facts and also seek overlapping relief. Finding no overlap in the relief requested, the court held that the two suits at issue were not for or in respect to the same claim. This Court granted certiorari. 559 U. S. ___ (2010). II Since 1868, Congress has restricted the jurisdiction of the CFC and its predecessors when related actions are pending elsewhere. Section 1500, identical in most respects to the original statute, provides: “The United States Court of Federal Claims shall not have jurisdiction of any claim for or in respect to which the plaintiff or his assignee has pending in any other court any suit or process against the United States or any person who, at the time when the cause of action alleged in such suit or process arose, was, in respect thereto, acting or professing to act, directly or indirectly under the authority of the United States.” The rule is more straightforward than its complex wording suggests. The CFC has no jurisdiction over a claim if the plaintiff has another suit for or in respect to that claim pending against the United States or its agents. The question to be resolved is what it means for two suits to be “for or in respect to” the same claim. Keene Corp. v. United States, 508 U. S. 200 (1993), provided a partial answer. It held that two suits are for or in respect to the same claim when they are “based on substantially the same operative facts … , at least if there [is] some overlap in the relief requested.” Id., at 212. The Keene case did not decide whether the jurisdictional bar also operates if the suits are based on the same operative facts but do not seek overlapping relief. Still, Keene narrows the permissible constructions of “for or in respect to” a claim to one of two interpretations. Either it requires substantial factual and some remedial overlap, or it requires substantial factual overlap without more. Congress first enacted the jurisdictional bar now codified in §1500 to curb duplicate lawsuits brought by residents of the Confederacy following the Civil War. The so-called “cotton claimants”—named for their suits to recover for cotton taken by the Federal Government—sued the United States in the Court of Claims under the Abandoned Property Collection Act, 12 Stat. 820, while at the same time suing federal officials in other courts, seeking relief under tort law for the same alleged actions. See Keene, supra, at 206–207; Schwartz, Section 1500 of the Judicial Code and Duplicate Suits Against the Government and Its Agents, 55 Geo. L. J. 573, 574–580 (1967). Although the rule embodied in §1500 originated long ago, Congress reenacted the statute at various times, most recently in 1948. See Act of June 25, 1948, 62 Stat. 942; Keene, 508 U. S., at 206–207. The text of §1500 reflects a robust response to the problem first presented by the cotton claimants. It bars jurisdiction in the CFC not only if the plaintiff sues on an identical claim elsewhere—a suit “for” the same claim—but also if the plaintiff’s other action is related although not identical—a suit “in respect to” the same claim. The phrase “in respect to” does not resolve all doubt as to the scope of the jurisdictional bar, but “it does make it clear that Congress did not intend the statute to be rendered useless by a narrow concept of identity.” Id., at 213. It suggests a broad prohibition, regardless of whether “claim” carries a special or limited meaning. Cf. United States v. Jones, 131 U. S. 1 (1889) (“claim” in the Little Tucker Act refers only to requests for money). Of the two constructions of “for or in respect to” the same claim that Keene permits—one based on facts alone and the other on factual plus remedial overlap—the former is the more reasonable interpretation in light of the statute’s use of a similar phrase in a way consistent only with factual overlap. The CFC bar applies even where the other action is not against the Government but instead against a “person who, at the time when the cause of action alleged in such suit or process arose, was, in respect thereto, acting or professing to act, directly or indirectly under the authority of the United States.” The statute refers to a person who acts under color of federal law in respect to a cause of action at the time it arose. But at that time, the person could not act in respect to the relief requested, for no complaint was yet filed. This use of the phrase “in respect to a cause of action” must refer to operative facts and not whatever remedies an aggrieved party might later request. A person acts under color of federal law in respect to a cause of action by claiming or wielding federal authority in the relevant factual context. Although the two phrases are not identical—one is in respect to a claim, the other a cause of action—they are almost so, and there is reason to think that both phrases refer to facts alone and not to relief. As the Keene Court explained, “the term ‘claim’ is used here synonymously with ‘cause of action.’ ” 508 U. S., at 210. And if either of the two phrases were to include both operative facts and a specific remedy, it would be the one that uses the term “cause of action” rather than “claim.” “Cause of action” is the more technical term, while “claim” is often used in a commonsense way to mean a right or demand. Here, for the reasons stated in the preceding paragraph, “in respect to a cause of action” refers simply to facts without regard to judicial remedies. So, if the phrase with the more technical of the two terms does not embrace the concept of remedy, it is reasonable to conclude that neither phrase does. Even if the terms “claim” or “cause of action” include the request for relief, the phrase “for or in respect to” gives the statutory bar a broader scope. Reading the statute to require only factual and not also remedial overlap makes sense in light of the unique remedial powers of the CFC. The CFC is the only judicial forum for most non-tort requests for significant monetary relief against the United States. See 28 U. S. C. §1491 (2006 ed. and Supp. III); §1346(a)(2) (2006 ed.). Unlike the district courts, however, the CFC has no general power to provide equitable relief against the Government or its officers. Compare United States v. King, 395 U. S. 1, 2–3 (1969), with 5 U. S. C. §702; see also United States v. Alire, 6 Wall. 573, 575 (1868) (“[T]he only judgments which the Court of Claims are authorized to render against the government … are judgments for money found due from the government to the petitioner”). The distinct jurisdiction of the CFC makes overlapping relief the exception and distinct relief the norm. For that reason, a statute aimed at precluding suits in the CFC that duplicate suits elsewhere would be unlikely to require remedial overlap. Remedial overlap between CFC actions and those in other courts was even more unusual when §1500’s rule was first enacted in 1868. At that time the CFC had a more limited jurisdiction than it does now, for the Tucker Act’s general waiver of sovereign immunity for non-tort claims for monetary relief had not yet been enacted. See 24 Stat. 505. And while the district courts can today adjudicate suits against the United States for money damages under the Little Tucker Act, 28 U. S. C. §1346(a)(2), and the Federal Tort Claims Act §1346(b), in 1868 the United States could only be sued in the Court of Claims. United States v. Mitchell, 463 U. S. 206, 212–214 (1983); G. Sisk, Litigation with the Federal Government §4.02(a)(1) (4th ed. 2006). Because the kinds of suits and forms of relief available against the United States were few and constrained, remedial overlap between CFC suits and those in other courts was even less common then than now. If the statute were to require remedial as well as factual overlap, it would have had very limited application in 1868 despite its broad language that bars not only identical but also related claims. The rule in §1500 effects a significant jurisdictional limitation, and Congress reenacted it even as changes in the structure of the courts made suits on the same facts more likely to arise. Doing so reaffirmed the force of the bar and thus the commitment to curtailing redundant litigation. The panel of the Court of Appeals could not identify “any purpose that §1500 serves today,” 559 F. 3d, at 1292, in large part because it was bound by Circuit precedent that left the statute without meaningful force. For example, the panel cited Tecon Engineers, Inc. v. United States, 170 Ct. Cl. 389, 343 F. 2d 943 (1965), which held that §1500 does not prohibit two identical suits from proceeding so long as the action in the CFC, or at that time the Court of Claims, is filed first. The Tecon holding is not presented in this case because the CFC action here was filed after the District Court suit. Still, the Court of Appeals was wrong to allow its precedent to suppress the statute’s aims. Courts should not render statutes nugatory through construction. In fact the statute’s purpose is clear from its origins with the cotton claimants—the need to save the Government from burdens of redundant litigation—and that purpose is no less significant today. The conclusion that two suits are for or in respect to the same claim when they are based on substantially the same operative facts allows the statute to achieve its aim. Keene, supra, at 206. Developing a factual record is responsible for much of the cost of litigation. Discovery is a conspicuous example, and the preparation and examination of witnesses at trial is another. The form of relief requested matters less, except insofar as it affects what facts parties must prove. An interpretation of §1500 focused on the facts rather than the relief a party seeks preserves the provision as it was meant to function, and it keeps the provision from becoming a mere pleading rule, to be circumvented by carving up a single transaction into overlapping pieces seeking different relief. Cf. Casman v. United States, 135 Ct. Cl. 647 (1956) (CFC had jurisdiction notwithstanding common facts in district court suit because the plaintiff sought different relief in each forum). Concentrating on operative facts is also consistent with the doctrine of claim preclusion, or res judicata, which bars “repetitious suits involving the same cause of action” once “a court of competent jurisdiction has entered a final judgment on the merits.” Commissioner v. Sunnen, 333 U. S. 591, 597 (1948). The jurisdictional bar in §1500 was enacted in part to address the problem that judgments in suits against officers were not preclusive in suits against the United States. Matson Nav. Co. v. United States, 284 U. S. 352, 355–356 (1932). So it is no surprise that the statute would operate in similar fashion. The now-accepted test in preclusion law for determining whether two suits involve the same claim or cause of action depends on factual overlap, barring “claims arising from the same transaction.” Kremer v. Chemical Constr. Corp., 456 U. S. 461, 482, n. 22 (1982); see also Restatement (Second) of Judgments §24 (1980). The transactional test is of course much younger than the rule embodied in §1500, but even in the 19th century it was not uncommon to identify a claim for preclusion purposes based on facts rather than relief. See J. Wells, Res Adjudicata and Stare Decisis §241, p. 208 (1878) (“The true distinction between demands or rights of action which are single and entire, and those which are several and distinct, is, that the former immediately arise out of one and the same act or contract, and the latter out of different acts or contracts” (internal quotation marks omitted)); 2 H. Black, Law of Judgments §726, p. 866 (1891) (The test for identity is: “Would the same evidence support and establish both the present and the former cause of action?”). Reading §1500 to depend on the underlying facts and not also on the relief requested gives effect to the principles of preclusion law embodied in the statute. There is no merit to the Nation’s assertion that the interpretation adopted here cannot prevail because it is unjust, forcing plaintiffs to choose between partial remedies available in different courts. The hardship in this case is far from clear. The Nation could have filed in the CFC alone and if successful obtained monetary relief to compensate for any losses caused by the Government’s breach of duty. It also seems likely that Indian tribes in the Nation’s position could go to district court first without losing the chance to later file in the CFC, for Congress has provided in every appropriations Act for the Department of Interior since 1990 that the statute of limitations on Indian trust mismanagement claims shall not run until the affected tribe has been given an appropriate accounting. See, e.g., 123 Stat. 2922; 104 Stat. 1930. Even were some hardship to be shown, considerations of policy divorced from the statute’s text and purpose could not override its meaning. Although Congress has permitted claims against the United States for monetary relief in the CFC, that relief is available by grace and not by right. See Beers v. Arkansas, 20 How. 527, 529 (1858) (“[A]s this permission is altogether voluntary on the part of the sovereignty, it follows that it may prescribe the terms and conditions on which it consents to be sued, and the manner in which the suit shall be conducted”). If indeed the statute leads to incomplete relief, and if plaintiffs like the Nation are dissatisfied, they are free to direct their complaints to Congress. This Court “enjoy[s] no ‘liberty to add an exception … to remove apparent hardship.’ ” Keene, 508 U. S., at 217–218 (quoting Corona Coal Co. v. United States, 263 U. S. 537, 540 (1924)). Keene reserved the question whether common facts are sufficient to bar a CFC action where a similar case is pending elsewhere. To continue to reserve the question would force the CFC to engage in an unnecessary and complicated remedial inquiry, and it would increase the expense and duration of litigation. The question thus demands an answer, and the answer is yes. Two suits are for or in respect to the same claim, precluding jurisdiction in the CFC, if they are based on substantially the same operative facts, regardless of the relief sought in each suit. III The remaining question is whether the Nation’s two suits have sufficient factual overlap to trigger the jurisdictional bar. The CFC dismissed the action here in part because it concluded that the facts in the Nation’s two suits were, “for all practical purposes, identical.” 79 Fed. Cl. 645, 656 (2007). It was correct to do so. The two actions both allege that the United States holds the same assets in trust for the Nation’s benefit. They describe almost identical breaches of fiduciary duty—that the United States engaged in self-dealing and imprudent investment, and failed to provide an accurate accounting of the assets held in trust, for example. Indeed, it appears that the Nation could have filed two identical complaints, save the caption and prayer for relief, without changing either suit in any significant respect. Under §1500, the substantial overlap in operative facts between the Nation’s District Court and CFC suits precludes jurisdiction in the CFC. The Court of Appeals erred when it concluded otherwise. IV The holding here precludes the CFC from exercising jurisdiction over the Nation’s suit while the District Court case is pending. Should the Nation choose to dismiss the latter action, or upon that action’s completion, the Nation is free to file suit again in the CFC if the statute of limitations is no bar. In the meantime, and in light of the substantial overlap in operative facts between them, the two suits are “for or in respect to” the same claim under §1500, and the CFC case must be dismissed. The contrary judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

563.US.247

Together, the Developmental Disabilities Assistance and Bill of Rights Act of 2000 (DD Act) and the Protection and Advocacy for Individuals with Mental Illness Act (PAIMI Act) offer States federal money to improve, inter alia, medical care for persons with developmental disabilities or mental illness. As a condition of funding, a State must establish a protection and advocacy (P&A) system “to protect and advocate [those individuals’] rights.” 42 U. S. C. §15043(a)(1). A participating State may appoint either a state agency or a private nonprofit entity as its P&A system, but if a state agency it must have authority to litigate and freedom from the control of other state agencies or officers. Virginia has appointed an independent state agency, petitioner Virginia Office for Protection and Advocacy (VOPA), authorizing it to litigate to secure disabled individuals’ rights, free of executive-branch oversight; to operate independently of Virginia’s attorney general; and to employ its own lawyers to sue on its behalf. While investigating patient deaths and injuries at state mental hospitals, VOPA asked respondents—state officials in charge of those hospitals—to produce relevant patient records. Respondents refused, asserting that a state-law privilege shielded the records from disclosure. VOPA then filed suit in Federal District Court, seeking a declaration that respondents’ refusal to produce the records violated the DD and PAIMI Acts and an injunction requiring respondents to produce the records and refrain in the future from interfering with VOPA’s right of access. Respondents moved to dismiss on the ground that they are immune from suit under the Eleventh Amendment, but the court held that the suit was permitted by the doctrine of Ex parte Young, 209 U. S. 123, which normally allows federal courts to award prospective relief against state officials for violations of federal law. The Fourth Circuit reversed, finding that Ex parte Young did not apply because the suit was brought by a state agency. Held: Ex parte Young allows a federal court to hear a lawsuit for prospective relief against state officials brought by another agency of the same State. Pp. 4–13. (a) Absent a waiver of sovereign immunity by a State itself or a valid abrogation by Congress, federal courts may not entertain a private person’s suit against a State. Pp. 4–5. (b) The doctrine of Ex parte Young, which establishes an important limitation on the sovereign-immunity principle, is accepted as necessary to “permit the federal courts to vindicate federal rights.” Pennhurst State School and Hospital v. Halderman, 465 U. S. 89. It rests on the premise that when a federal court commands a state official to do nothing more than refrain from violating federal law, he is not the State for sovereign-immunity purposes. It does not apply “when ‘the state is the … party in interest.’ ” Id., at 101. Pp. 5–6. (c) Entertaining VOPA’s action is consistent with precedent and does not offend the distinctive interests protected by sovereign immunity. Pp. 6–13. (1) Verizon Md. Inc. v. Public Serv. Comm’n of Md., 535 U. S. 635, held that, in determining the Ex parte Young doctrine’s applicability, “a court need only conduct a ‘straightforward inquiry into whether [the] complaint alleges an ongoing violation of federal law and seeks relief properly characterized as prospective.’ ” Id., at 645. VOPA’s suit satisfies that inquiry. Respondents concede that the action would be proper were VOPA a private organization rather than a state agency. The “general criterion for determining when a suit is in fact against the sovereign is the effect of the relief sought,” Pennhurst, supra, at 107, not who is bringing the lawsuit. This Court applied that criterion in Idaho v. Coeur d’Alene Tribe of Idaho, 521 U. S. 261, which held that an Indian Tribe could not invoke Ex parte Young to bring what was essentially a quiet title suit that would “extinguish [Idaho’s] control over … lands and waters long deemed … an integral part of its territory.” Id., at 282. Respondents have advanced no argument that the relief sought here threatens a similar invasion of Virginia’s sovereignty. Pp. 7–9. (2) Respondents claim that a State’s dignity is diminished when a federal court adjudicates a dispute between its components. But a State’s stature is not diminished to any greater degree when its own agency sues to enforce its officers’ compliance with federal law than when a private person does so. Moreover, VOPA’s power to sue state officials is a consequence of Virginia’s own decision to establish a public P&A system. Not every offense to a State’s dignity constitutes a denial of sovereign immunity. The specific indignity against which sovereign immunity protects is the insult to a State of being haled into court without its consent; that does not occur just because a suit happens to be brought by another state agency. Pp. 9–11. (3) The apparent novelty of this suit is not likely a consequence of past constitutional doubts. In order to invoke the Ex parte Young exception, a state agency needs both a federal right that it possesses against its parent State and authority to sue state officials to enforce that right, free from any internal state-government veto; such conditions rarely coincide. In any event, the principles undergirding the Ex parte Young doctrine support its extension to actions of this kind. Pp. 12–13. 568 F. 3d 110, reversed and remanded. Scalia, J., delivered the opinion of the Court, in which Kennedy, Thomas, Ginsburg, Breyer, and Sotomayor, JJ., joined. Kennedy, J., filed a concurring opinion, in which Thomas, J., joined. Roberts, C. J., filed a dissenting opinion, in which Alito, J., joined. Kagan, J., took no part in the consideration or decision of the case.

We consider whether Ex parte Young, 209 U. S. 123 (1908), allows a federal court to hear a lawsuit for prospective relief against state officials brought by another agency of the same State. I A The Developmental Disabilities Assistance and Bill of Rights Act of 2000 (DD Act), 114 Stat. 1677, 42 U. S. C. §15001 et seq., offers States federal money to improve community services, such as medical care and job training, for individuals with developmental disabilities. See §§15023(a), 15024. As a condition of that funding, a State must establish a protection and advocacy (P&A) system “to protect and advocate the rights of individuals with developmental disabilities.” §15043(a)(1). The P&A system receives separate federal funds, paid to it directly. §15042(a) and (b). A second federal law, the Protection and Advocacy for Individuals with Mental Illness Act (PAIMI Act), 100 Stat. 478, 42 U. S. C. §10801 et seq., increases that separate funding and extends the mission of P&A systems to include the mentally ill. §§10802(2), 10803, 10827. At present, every State accepts funds under these statutes. Under the DD and PAIMI Acts, a P&A system must have certain powers. The system “shall … have the authority to investigate incidents of abuse and neglect … if the incidents are reported to the system or if there is probable cause to believe that the incidents occurred.” §15043(a)(2)(B); §10805(a)(1)(A). Subject to certain statutory requirements, it must be given access to “all records” of individuals who may have been abused, see §15043(a)(2)(I)(iii)(II); §10805(a)(4)(B)(iii), as well as “other records that are relevant to conducting an investigation,” §15043(a)(2)(J)(i). The Acts also require that a P&A system have authority to “pursue legal, administrative, and other appropriate remedies or approaches to ensure the protection of” its charges. §15043(a)(2)(A)(i); see §10805(a)(1)(B). And in addition to pressing its own rights, a P&A system may “pursue administrative, legal, and other remedies on behalf of” those it protects. §10805(a)(1)(C); see §15044(b). A participating State is free to appoint either a state agency or a private nonprofit entity as its P&A system. §15044(a); §10805(c)(1)(B). But in either case, the designated entity must have certain structural features that ensure its independence from the State’s government. The DD Act prohibits the Governor from appointing more than one-third of the members of the system’s governing board, §15044(a)(2), and restricts the State’s ability to impose hiring freezes or other measures that would impair the system’s ability to carry out its mission, §15043(a)(2)(K). Once a State designates an entity as its P&A system, it may not change its selection without “good cause.” §15043(a)(4)(A). Virginia is one of just eight States that have designated a government entity as their P&A system. The Virginia Office for Protection and Advocacy (VOPA) is an “independent state agency.” Va. Code Ann. §51.5–39.2(A) (Lexis 2009). Its board consists of eleven “nonlegislative citizen members,” of whom only three are appointed by the Governor. §51.5–39.2(B). The remaining eight are appointed by components of the legislature: five by the Speaker of the House of Delegates, and three by the Senate Committee on Rules. Ibid. VOPA itself nominates candidates for consideration, and the statute instructs the appointing officials that they “shall seriously consider the persons nominated and appoint such persons whenever feasible.” Ibid. Board members serve for fixed terms and are removable only by a court and only for specified reasons. See §51.5–39.2(C) and (F); §24.2–233 and 234 (Lexis 2006). VOPA enjoys authority to litigate free of executive-branch oversight. It operates independently of the Attorney General of Virginia and employs its own lawyers, who are statutorily authorized to sue on VOPA’s behalf. §51.5–39.2(A); §2.2–510(5) (Lexis 2008). And Virginia law specifically empowers VOPA to “initiate any proceedings to secure the rights” of disabled individuals. §51.5–39.2(A). B In 2006, VOPA opened an investigation into the deaths of two patients and injuries to a third at state-run mental hospitals. It asked respondents—state officials in charge of those institutions—to produce any records related to risk-management or mortality reviews conducted by the hospitals with respect to those patients. Respondents refused, asserting that the records were protected by a state-law privilege shielding medical peer-review materials from disclosure. VOPA then brought this action in the United States District Court for the Eastern District of Virginia, alleging that the DD and PAIMI Acts entitled it to the peer-review records, notwithstanding any state-law privilege that might apply. It sought a declaration that respondents’ refusal to produce the records violated the DD and PAIMI Acts, along with an injunction requiring respondents to provide access to the records and refrain in the future from interfering with VOPA’s right of access to them. Respondents moved to dismiss the action on the grounds that they are immune from suit under the Eleventh Amendment. The District Court denied the motion. In its view, the suit was permitted by the doctrine of Ex parte Young, which normally allows federal courts to award prospective relief against state officials for violations of federal law. Virginia v. Reinhard, 2008 WL 2795940, *6 (ED Va., July 18, 2008). The Court of Appeals reversed. Virginia v. Reinhard, 568 F. 3d 110 (CA4 2009). Believing VOPA’s lawsuit to be an “intramural contest” that “encroaches more severely on the dignity and sovereignty of the states than an Ex parte Young action brought by a private plaintiff,” the Court of Appeals concluded it was not authorized by that case. Id., at 119–120 (internal quotation marks omitted). We granted certiorari. 561 U. S. ____ (2010). II A Sovereign immunity is the privilege of the sovereign not to be sued without its consent. The language of the Eleventh Amendment[Footnote 1] only eliminates the basis for our judgment in the famous case of Chisholm v. Georgia, 2 Dall. 419 (1793), which involved a suit against a State by a noncitizen of the State. Since Hans v. Louisiana, 134 U. S. 1 (1890), however, we have understood the Eleventh Amendment to confirm the structural understanding that States entered the Union with their sovereign immunity intact, unlimited by Article III’s jurisdictional grant. Blatchford v. Native Village of Noatak, 501 U. S. 775, 779 (1991); see Pennhurst State School and Hospital v. Halderman, 465 U. S. 89, 98 (1984). Our cases hold that the States have retained their traditional immunity from suit, “except as altered by the plan of the Convention or certain constitutional amendments.” Alden v. Maine, 527 U. S. 706, 713 (1999). A State may waive its sovereign immunity at its pleasure, College Savings Bank v. Florida Prepaid Postsecondary Ed. Expense Bd., 527 U. S. 666, 675–676 (1999), and in some circumstances Congress may abrogate it by appropriate legislation.[Footnote 2] But absent waiver or valid abrogation, federal courts may not entertain a private person’s suit against a State. B In Ex parte Young, 209 U. S. 123, we established an important limit on the sovereign-immunity principle. That case involved a challenge to a Minnesota law reducing the freight rates that railroads could charge. A railroad shareholder claimed that the new rates were un-constitutionally confiscatory, and obtained a federal injunction against Edward Young, the Attorney General of Minnesota, forbidding him in his official capacity to enforce the state law. Perkins v. Northern Pacific R. Co., 155 F. 445 (CC Minn. 1907). When Young violated the injunction by initiating an enforcement action in state court, the Circuit Court held him in contempt and committed him to federal custody. In his habeas corpus application in this Court, Young challenged his confinement by arguing that Minnesota’s sovereign immunity deprived the federal court of jurisdiction to enjoin him from performing his official duties. We disagreed. We explained that because an unconstitutional legislative enactment is “void,” a state official who enforces that law “comes into conflict with the superior authority of [the] Constitution,” and therefore is “stripped of his official or representative character and is subjected in his person to the consequences of his individual conduct. The State has no power to impart to him any immunity from responsibility to the supreme authority of the United States.” 209 U. S., at 159–160. This doctrine has existed alongside our sovereign-immunity jurisprudence for more than a century, accepted as necessary to “permit the federal courts to vindicate federal rights.” Pennhurst, 465 U. S., at 105. It rests on the premise—less delicately called a “fiction,” id., at 114, n. 25—that when a federal court commands a state official to do nothing more than refrain from violating federal law, he is not the State for sovereign-immunity purposes. The doctrine is limited to that precise situation, and does not apply “when ‘the state is the real, substantial party in interest,’ ” id., at 101 (quoting Ford Motor Co. v. Department of Treasury of Ind., 323 U. S. 459, 464 (1945)), as when the “ ‘judgment sought would expend itself on the public treasury or domain, or interfere with public administration,’ ” 465 U. S., at 101, n. 11 (quoting Dugan v. Rank, 372 U. S. 609, 620 (1963)). C This case requires us to decide how to apply the Ex parte Young doctrine to a suit brought by an independent state agency claiming to possess federal rights. Although we have never encountered such a suit before, we are satisfied that entertaining VOPA’s action is consistent with our precedents and does not offend the distinctive interests protected by sovereign immunity. 1 In Verizon Md. Inc. v. Public Serv. Comm’n of Md., 535 U. S. 635 (2002), we held that “[i]n determining whether the doctrine of Ex parte Young avoids an Eleventh Amendment bar to suit, a court need only conduct a ‘straightforward inquiry into whether [the] complaint alleges an ongoing violation of federal law and seeks relief properly characterized as prospective.’ ” Id., at 645 (quoting Idaho v. Coeur d’Alene Tribe of Idaho, 521 U. S. 261, 296 (1997) (O’Connor, J., concurring in part and concurring in judgment)). There is no doubt VOPA’s suit satisfies that straightforward inquiry. It alleges that respondents’ refusal to produce the requested medical records violates federal law; and it seeks an injunction requiring the production of the records, which would prospectively abate the alleged violation. Respondents concede that were VOPA a private organization rather than a state agency, the doctrine would permit this action to proceed.[Footnote 3] We see no reason for a different result here. Although respondents argue that VOPA’s status as a state agency changes the calculus, there is no warrant in our cases for making the validity of an Ex parte Young action turn on the identity of the plaintiff. To be sure, we have been willing to police abuses of the doctrine that threaten to evade sovereign immunity. To do otherwise “would be to adhere to an empty formalism.” Coeur d’Alene Tribe, supra, at 270. But (as the dissent concedes, post, at 8 (opinion of Roberts, C. J.)) the limits we have recognized reflect the principle that the “general criterion for determining when a suit is in fact against the sovereign is the effect of the relief sought,” Pennhurst, supra, at 107, not who is bringing the lawsuit. Thus, Ex parte Young cannot be used to obtain an injunction requiring the payment of funds from the State’s treasury, see Edelman v. Jordan, 415 U. S. 651, 666 (1974); or an order for specific performance of a State’s contract, see id., at 666–667; In re Ayers, 123 U. S. 443 (1887). Coeur d’Alene Tribe, on which respondents heavily rely, is an application of this principle. There we refused to allow an Indian Tribe to use Ex parte Young to obtain injunctive and declaratory relief establishing its exclusive right to the use and enjoyment of certain submerged lands in Idaho and the invalidity of all state statutes and regulations governing that land. 521 U. S., at 265. We determined that the suit was “the functional equivalent of a quiet title suit against Idaho,” would “extinguish … the State’s control over a vast reach of lands and waters long deemed by the State to be an integral part of its territory,” and thus was barred by sovereign immunity. Id., at 282. Respondents have advanced no argument that the relief sought in this case threatens any similar invasion of Virginia’s sovereignty. Indeed, they concede that the very injunction VOPA requests could properly be awarded by a federal court at the instance of a private P&A system. 2 Respondents and the dissent argue that entertaining VOPA’s lawsuit in a federal forum would nevertheless infringe Virginia’s sovereign interests because it diminishes the dignity of a State for a federal court to adjudicate a dispute between its components. See Brief for Respondents 23–26; post, at 4–8 (arguing that “ ‘special sovereignty interests’ ” bar VOPA’s lawsuit (quoting Coeur d’Alene Tribe, supra, at 281)). We disagree. As an initial matter, we do not understand how a State’s stature could be diminished to any greater degree when its own agency polices its officers’ compliance with their federal obligations, than when a private person hales those officers into federal court for that same purpose—something everyone agrees is proper.[Footnote 4] And in this case, of course, VOPA’s power to sue state officials is a consequence of Virginia’s own decision to establish a public, rather than a private, P&A system. We fail to perceive what Eleventh Amendment indignity is visited on the Commonwealth when, by operation of its own laws, VOPA is admitted to federal court as a plaintiff.[Footnote 5] But even if it were true that the State’s dignity were offended in some way by the maintenance of this action in federal court, that would not prove respondents’ case. Denial of sovereign immunity, to be sure, offends the dignity of a State; but not every offense to the dignity of a State constitutes a denial of sovereign immunity. The specific indignity against which sovereign immunity protects is the insult to a State of being haled into court without its consent. That effectively occurs, our cases reasonably conclude, when (for example) the object of the suit against a state officer is to reach funds in the state treasury or acquire state lands; it does not occur just because the suit happens to be brought by another state agency. Respondents’ asserted dignitary harm is simply unconnected to the sovereign-immunity interest. The dissent complains that applying Ex parte Young to this lawsuit divides Virginia against itself, since the opposing parties are both creatures of the Commonwealth. Post, at 7. Even if that were a distinctive consequence of letting this suit proceed in federal court, it would have nothing to do with the concern of sovereign-immunity—whether the suit is against an unconsenting State, rather than against its officers. But it is not a consequence of the federal nature of the forum. The same result will follow if the federal claim is sued upon in state court, as the dissent would require. There also, “[w]hatever the decision in the litigation, … [t]he Commonwealth will win[, a]nd the Commonwealth will lose.” Ibid. Nor would sending the matter to state court even avoid the prospect that “a federal judge will resolve which part of the Commonwealth will prevail,” ibid., since the state-court loser could always ask this Court to review the matter by certiorari. (Or is that appeal also to be disallowed on grounds of sovereign immunity? But see Cohens v. Virginia, 6 Wheat. 264 (1821).)[Footnote 6] And of course precisely the same thing would happen if respondents specifically waived their sovereign-immunity objections in this very case. Yet no one would contend that despite the waiver, sovereign immunity forbade the suit. So also here: If, by reason of Ex parte Young, there has been no violation of sovereign immunity, the prospect of a federal judge’s resolving VOPA’s dispute with respondents does not make it so. We do not doubt, of course, that there are limits on the Federal Government’s power to affect the internal operations of a State. See, e.g., Printz v. United States, 521 U. S. 898 (1997) (Congress may not commandeer state officers); Coyle v. Smith, 221 U. S. 559, 579 (1911) (Congress may not dictate a State’s capital). But those limits must be found in some textual provision or structural premise of the Constitution. Additional limits cannot be smuggled in under the Eleventh Amendment by barring a suit in federal court that does not violate the State’s sovereign immunity. [Footnote 7] 3 A weightier objection, perhaps, is the relative novelty of this lawsuit. Respondents rightly observe that federal courts have not often encountered lawsuits brought by state agencies against other state officials. That does give us pause. Lack of historical precedent can indicate a constitutional infirmity, see, e.g., Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U. S. ____, ____ (2010) (slip op., at 25), and our sovereign-immunity decisions have traditionally warned against “ ‘anomalous and unheard-of proceedings or suits,’ ” Alden, 527 U. S., at 727 (quoting Hans, 134 U. S., at 18). Novelty, however, is often the consequence of past constitutional doubts, but we have no reason to believe that is the case here. In order to invoke the Ex parte Young exception to sovereign immunity, a state agency needs two things: first, a federal right that it possesses against its parent State; and second, authority to sue other state officials to enforce that right, free from any internal veto wielded by the state government. These conditions will rarely coincide—and at least the latter of them cannot exist without the consent of the State that created the agency and defined its powers. See post, at 3–4 (Kennedy, J., concurring). We are unaware that the necessary conditions have ever presented themselves except in connection with the DD and PAIMI Acts, and the parties have referred us to no examples.[Footnote 8] Thus, the apparent novelty of this sort of suit does not at all suggest its unconstitutionality. In any event, we are satisfied, for the reasons we have explained, that—novelty notwithstanding—the principles undergirding the Ex parte Young doc-trine support its application to actions of this kind. * * * Like the Court of Appeals, we are mindful of the central role autonomous States play in our federal system, and wary of approving new encroachments on their sovereignty. But we conclude no such encroachment is occasioned by straightforwardly applying Ex parte Young to allow this suit. It was Virginia law that created VOPA and gave it the power to sue state officials. In that circumstance, the Eleventh Amendment presents no obstacle to VOPA’s ability to invoke federal jurisdiction on the same terms as any other litigant. We reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Footnote 1 The Eleventh Amendment reads as follows: “The Judicial power of the United States shall not be construed to extend to any suit in law or equity, commenced or prosecuted against one of the United States by Citizens of another State, or by Citizens or Subjects of any Foreign State.” Footnote 2 We have recognized that Congress may abrogate a State’s immunity when it acts under §5 of the Fourteenth Amendment, Seminole Tribe of Fla. v. Florida, 517 U. S. 44, 59 (1996), but not when it acts under its original Article I authority to regulate commerce, id., at 65–66. Footnote 3 The dissent is mistaken when it claims that applying the Verizon Maryland test would mean two of our cases were “wrongly decided.” Post, at 4 (opinion of Roberts, C. J.). We discuss the first of those cases, Coeur d’Alene Tribe, below. Infra, at 8. As for the second, Seminole Tribe, supra, it is inapposite. The reason we refused to permit suit to proceed in that case was that the Indian Gaming Regulatory Act created an alternative remedial scheme that would be undermined by permitting Ex parte Young suits; Congress, we said, had foreclosed recourse to the doctrine. See Seminole Tribe, supra, at 73–76. Respondents now argue—for the first time in this litigation—that the DD and PAIMI Acts have the same effect here. We reject that suggestion. The fact that the Federal Government can exercise oversight of a federal spending program and even withhold or withdraw funds—which are the chief statutory features respondents point to—does not demonstrate that Congress has “displayed an intent not to provide the ‘more complete and more immediate relief’ that would otherwise be available under Ex parte Young.” Verizon Maryland, 535 U. S., at 647 (quoting Seminole Tribe, 517 U. S., at 75). Footnote 4 The dissent compares VOPA’s lawsuit to such indignities as “cannibalism” and “patricide,” since it is a greater “affront to someone’s dignity to be sued by a brother than to be sued by a stranger.” Post, at 9. We think the dissent’s principle of familial affront less than universally applicable, even with respect to real families, never mind governmental siblings. Most of us would probably prefer contesting a testamentary disposition with a relative to contesting it with a stranger. And confining one’s child to his room is called grounding, while confining a stranger’s child is called kidnaping. Jurisdiction over this case does not depend on which is the most apt comparison. Footnote 5 The dissent accuses us of circular reasoning, because we “wrongly assum[e] [that] Virginia knew in advance the answer to the question presented in this case.” Post, at 10. That would be true if we were relying on the Commonwealth’s waiver of sovereign immunity. We are not. We rely upon Ex parte Young. We say that Virginia has only itself to blame for the position in which it finds itself, not because it consented to suit, but because it created a state entity to sue, instead of leaving the task to a private entity. It did not have to know that this would allow suit in federal court. Know or not know, Ex parte Young produces that result. Footnote 6 The dissent agrees that because of the “ ‘constitutional plan,’ ” post, at 8, n. 3 (quoting McKesson Corp. v. Division of Alcoholic Beverages and Tobacco, Fla. Dept. of Business Regulation, 496 U. S. 18, 30 (1990), this Court can adjudicate disputes between state agencies without offending sovereign immunity. But explaining away exceptions to its theory does not advance the ball. It has not demonstrated that sovereign immunity has anything at all to say about federal courts’ adjudicating interagency disputes. Footnote 7 We have no occasion to pass on other questions of federalism lurk-ing in this case, such as whether the DD or PAIMI Acts are a proper exercise of Congress’s enumerated powers. As Justice Kennedy ob-serves, whether the Acts run afoul of some other constitutional pro- vision (i.e., besides the Eleventh Amendment) “cannot be permitted to distort the antecedent question of jurisdiction.” Post, at 5 (concurring opinion). Footnote 8 We think greatly exaggerated the dissent’s concern that, “[g]iven the number of state agencies across the country that enjoy independent litigating authority,” today’s decision “could potentially lead to all sorts of litigation in federal courts addressing internal state government disputes.” Post, at 11. Such litigation cannot occur unless the state agency has been given a federal right of its own to vindicate (as VOPA alleges it has been given under the highly unusual statute at issue here).

564.US.2010_10-277

Respondents, current or former employees of petitioner Wal-Mart, sought judgment against the company for injunctive and declaratory relief, punitive damages, and backpay, on behalf of themselves and a nationwide class of some 1.5 million female employees, because of Wal-Mart’s alleged discrimination against women in violation of Title VII of the Civil Rights Act of 1964. They claim that local managers exercise their discretion over pay and promotions disproportionately in favor of men, which has an unlawful disparate impact on female employees; and that Wal-Mart’s refusal to cabin its managers’ authority amounts to disparate treatment. The District Court certified the class, finding that respondents satisfied Federal Rule of Civil Procedure 23(a), and Rule 23(b)(2)’s requirement of showing that “the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole.” The Ninth Circuit substantially affirmed, concluding, inter alia, that respondents met Rule 23(a)(2)’s commonality requirement and that their backpay claims could be certified as part of a (b)(2) class because those claims did not predominate over the declaratory and injunctive relief requests. It also ruled that the class action could be manageably tried without depriving Wal-Mart of its right to present its statutory defenses if the District Court selected a random set of claims for valuation and then extrapolated the validity and value of the untested claims from the sample set. Held: 1. The certification of the plaintiff class was not consistent with Rule 23(a). Pp. 8–20. (a) Rule 23(a)(2) requires a party seeking class certification to prove that the class has common “questions of law or fact.” Their claims must depend upon a common contention of such a nature that it is capable of classwide resolution—which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke. Here, proof of commonality necessarily overlaps with respondents’ merits contention that Wal-Mart engages in a pattern or practice of discrimination. The crux of a Title VII inquiry is “the reason for a particular employment decision,” Cooper v. Federal Reserve Bank of Richmond, 467 U. S. 867, 876, and respondents wish to sue for millions of employment decisions at once. Without some glue holding together the alleged reasons for those decisions, it will be impossible to say that examination of all the class members’ claims will produce a common answer to the crucial discrimination question. Pp. 8–12. (b) General Telephone Co. of Southwest v. Falcon, 457 U. S. 147, describes the proper approach to commonality. On the facts of this case, the conceptual gap between an individual’s discrimination claim and “the existence of a class of persons who have suffered the same injury,” id., at 157–158, must be bridged by “[s]ignificant proof that an employer operated under a general policy of discrimination,” id., at 159, n. 15. Such proof is absent here. Wal-Mart’s announced policy forbids sex discrimination, and the company has penalties for denials of equal opportunity. Respondents’ only evidence of a general discrimination policy was a sociologist’s analysis asserting that Wal-Mart’s corporate culture made it vulnerable to gender bias. But because he could not estimate what percent of Wal-Mart employment decisions might be determined by stereotypical thinking, his testimony was worlds away from “significant proof” that Wal-Mart “operated under a general policy of discrimination.” Pp. 12–14. (c) The only corporate policy that the plaintiffs’ evidence convincingly establishes is Wal-Mart’s “policy” of giving local supervisors discretion over employment matters. While such a policy could be the basis of a Title VII disparate-impact claim, recognizing that a claim “can” exist does not mean that every employee in a company with that policy has a common claim. In a company of Wal-Mart’s size and geographical scope, it is unlikely that all managers would exercise their discretion in a common way without some common direction. Respondents’ attempt to show such direction by means of statistical and anecdotal evidence falls well short. Pp. 14–20. 2. Respondents’ backpay claims were improperly certified under Rule 23(b)(2). Pp. 20–27. (a) Claims for monetary relief may not be certified under Rule 23(b)(2), at least where the monetary relief is not incidental to the requested injunctive or declaratory relief. It is unnecessary to decide whether monetary claims can ever be certified under the Rule because, at a minimum, claims for individualized relief, like backpay, are excluded. Rule 23(b)(2) applies only when a single, indivisible remedy would provide relief to each class member. The Rule’s history and structure indicate that individualized monetary claims belong instead in Rule 23(b)(3), with its procedural protections of predominance, superiority, mandatory notice, and the right to opt out. Pp. 20–23. (b) Respondents nonetheless argue that their backpay claims were appropriately certified under Rule 23(b)(2) because those claims do not “predominate” over their injunctive and declaratory relief requests. That interpretation has no basis in the Rule’s text and does obvious violence to the Rule’s structural features. The mere “predominance” of a proper (b)(2) injunctive claim does nothing to justify eliminating Rule 23(b)(3)’s procedural protections, and creates incentives for class representatives to place at risk potentially valid monetary relief claims. Moreover, a district court would have to reevaluate the roster of class members continuously to excise those who leave their employment and become ineligible for classwide injunctive or declaratory relief. By contrast, in a properly certified (b)(3) class action for backpay, it would be irrelevant whether the plaintiffs are still employed at Wal-Mart. It follows that backpay claims should not be certified under Rule 23(b)(2). Pp. 23–26. (c) It is unnecessary to decide whether there are any forms of “incidental” monetary relief that are consistent with the above interpretation of Rule 23(b)(2) and the Due Process Clause because respondents’ backpay claims are not incidental to their requested injunction. Wal-Mart is entitled to individualized determinations of each employee’s eligibility for backpay. Once a plaintiff establishes a pattern or practice of discrimination, a district court must usually conduct “additional proceedings … to determine the scope of individual relief.” Teamsters v. United States, 431 U. S. 324, 361. The company can then raise individual affirmative defenses and demonstrate that its action was lawful. Id., at 362. The Ninth Circuit erred in trying to replace such proceedings with Trial by Formula. Because Rule 23 cannot be interpreted to “abridge, enlarge or modify any substantive right,” 28 U. S. C. §2072(b), a class cannot be certified on the premise that Wal-Mart will not be entitled to litigate its statutory defenses to individual claims. Pp. 26–27. 603 F. 3d 571, reversed. Scalia, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, and Alito, JJ., joined, and in which Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined as to Parts I and III. Ginsburg, J., filed an opinion concurring in part and dissenting in part, in which Breyer, Sotomayor, and Kagan, JJ., joined.

We are presented with one of the most expansive class actions ever. The District Court and the Court of Appeals approved the certification of a class comprising about one and a half million plaintiffs, current and former female employees of petitioner Wal-Mart who allege that the discretion exercised by their local supervisors over pay and promotion matters violates Title VII by discriminat-ing against women. In addition to injunctive and declaratory relief, the plaintiffs seek an award of backpay. We consider whether the certification of the plaintiff class was consistent with Federal Rules of Civil Procedure 23(a) and (b)(2). I A Petitioner Wal-Mart is the Nation’s largest private employer. It operates four types of retail stores throughout the country: Discount Stores, Supercenters, Neighborhood Markets, and Sam’s Clubs. Those stores are divided into seven nationwide divisions, which in turn comprise 41 regions of 80 to 85 stores apiece. Each store has between 40 and 53 separate departments and 80 to 500 staff positions. In all, Wal-Mart operates approximately 3,400 stores and employs more than one million people. Pay and promotion decisions at Wal-Mart are generally committed to local managers’ broad discretion, which is exercised “in a largely subjective manner.” 222 F. R. D. 137, 145 (ND Cal. 2004). Local store managers may increase the wages of hourly employees (within limits) with only limited corporate oversight. As for salaried employees, such as store managers and their deputies, higher corporate authorities have discretion to set their pay with-in preestablished ranges. Promotions work in a similar fashion. Wal-Mart permits store managers to apply their own subjective criteria when selecting candidates as “support managers,” which is the first step on the path to management. Admission to Wal-Mart’s management training program, however, does require that a candidate meet certain objective criteria, including an above-average performance rating, at least one year’s tenure in the applicant’s current position, and a willingness to relocate. But except for those requirements, regional and district managers have discretion to use their own judgment when selecting candidates for management training. Promotion to higher office—e.g., assistant manager, co-manager, or store manager—is similarly at the discretion of the employee’s superiors after prescribed objective factors are satisfied. B The named plaintiffs in this lawsuit, representing the 1.5 million members of the certified class, are three current or former Wal-Mart employees who allege that the company discriminated against them on the basis of their sex by denying them equal pay or promotions, in violation of Title VII of the Civil Rights Act of 1964, 78 Stat. 253, as amended, 42 U. S. C. §2000e–1 et seq.[Footnote 1] Betty Dukes began working at a Pittsburgh, California, Wal-Mart in 1994. She started as a cashier, but later sought and received a promotion to customer service manager. After a series of disciplinary violations, however, Dukes was demoted back to cashier and then to greeter. Dukes concedes she violated company policy, but contends that the disciplinary actions were in fact retaliation for invoking internal complaint procedures and that male employees have not been disciplined for similar infractions. Dukes also claims two male greeters in the Pittsburgh store are paid more than she is. Christine Kwapnoski has worked at Sam’s Club stores in Missouri and California for most of her adult life. She has held a number of positions, including a supervisory position. She claims that a male manager yelled at her frequently and screamed at female employees, but not at men. The manager in question “told her to ‘doll up,’ to wear some makeup, and to dress a little better.” App. 1003a. The final named plaintiff, Edith Arana, worked at a Wal-Mart store in Duarte, California, from 1995 to 2001. In 2000, she approached the store manager on more than one occasion about management training, but was brushed off. Arana concluded she was being denied opportunity for advancement because of her sex. She initiated internal complaint procedures, whereupon she was told to apply directly to the district manager if she thought her store manager was being unfair. Arana, however, decided against that and never applied for management training again. In 2001, she was fired for failure to comply with Wal-Mart’s timekeeping policy. These plaintiffs, respondents here, do not allege that Wal-Mart has any express corporate policy against the advancement of women. Rather, they claim that their local managers’ discretion over pay and promotions is exercised disproportionately in favor of men, leading to an unlawful disparate impact on female employees, see 42 U. S. C. §2000e–2(k). And, respondents say, because Wal-Mart is aware of this effect, its refusal to cabin its managers’ authority amounts to disparate treatment, see §2000e–2(a). Their complaint seeks injunctive and declaratory relief, punitive damages, and backpay. It does not ask for compensatory damages. Importantly for our purposes, respondents claim that the discrimination to which they have been subjected is common to all Wal-Mart’s female employees. The basic theory of their case is that a strong and uniform “corporate culture” permits bias against women to infect, perhaps subconsciously, the discretionary decisionmaking of each one of Wal-Mart’s thousands of managers—thereby making every woman at the company the victim of one common discriminatory practice. Respondents therefore wish to litigate the Title VII claims of all female employees at Wal-Mart’s stores in a nationwide class action. C Class certification is governed by Federal Rule of Civil Procedure 23. Under Rule 23(a), the party seeking certification must demonstrate, first, that: “(1) the class is so numerous that joinder of all mem- bers is impracticable, “(2) there are questions of law or fact common to the class, “(3) the claims or defenses of the representative parties are typical of the claims or defenses of the class, and “(4) the representative parties will fairly and adequately protect the interests of the class” (paragraph breaks added). Second, the proposed class must satisfy at least one of the three requirements listed in Rule 23(b). Respondents rely on Rule 23(b)(2), which applies when “the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole.”[Footnote 2] Invoking these provisions, respondents moved the District Court to certify a plaintiff class consisting of “ ‘[a]ll women employed at any Wal-Mart domestic retail store at any time since December 26, 1998, who have been or may be subjected to Wal-Mart’s challenged pay and management track promotions policies and practices.’ ” 222 F. R. D., at 141–142 (quoting Plaintiff ’s Motion for Class Certification in case No. 3:01–cv–02252–CRB (ND Cal.), Doc. 99, p. 37). As evidence that there were indeed “questions of law or fact common to” all the women of Wal-Mart, as Rule 23(a)(2) requires, respondents relied chiefly on three forms of proof: statistical evidence about pay and promotion disparities between men and women at the company, anecdotal reports of discrimination from about 120 of Wal-Mart’s female employees, and the testimony of a sociologist, Dr. William Bielby, who conducted a “social framework analysis” of Wal-Mart’s “culture” and personnel practices, and concluded that the company was “vulnerable” to gender discrimination. 603 F. 3d 571, 601 (CA9 2010) (en banc). Wal-Mart unsuccessfully moved to strike much of this evidence. It also offered its own countervailing statistical and other proof in an effort to defeat Rule 23(a)’s requirements of commonality, typicality, and adequate representation. Wal-Mart further contended that respondents’ monetary claims for backpay could not be certified under Rule 23(b)(2), first because that Rule refers only to injunctive and declaratory relief, and second because the backpay claims could not be manageably tried as a class without depriving Wal-Mart of its right to present certain statutory defenses. With one limitation not relevant here, the District Court granted respondents’ motion and certified their proposed class.[Footnote 3] D A divided en banc Court of Appeals substantially affirmed the District Court’s certification order. 603 F. 3d 571. The majority concluded that respondents’ evidence of commonality was sufficient to “raise the common question whether Wal-Mart’s female employees nationwide were subjected to a single set of corporate policies (not merely a number of independent discriminatory acts) that may have worked to unlawfully discriminate against them in violation of Title VII.” Id., at 612 (emphasis deleted). It also agreed with the District Court that the named plaintiffs’ claims were sufficiently typical of the class as a whole to satisfy Rule 23(a)(3), and that they could serve as adequate class representatives, see Rule 23(a)(4). Id., at 614–615. With respect to the Rule 23(b)(2) question, the Ninth Circuit held that respondents’ backpay claims could be certified as part of a (b)(2) class because they did not “predominat[e]” over the requests for declaratory and injunctive relief, meaning they were not “superior in strength, influence, or authority” to the nonmonetary claims. Id., at 616 (internal quotation marks omitted).[Footnote 4] Finally, the Court of Appeals determined that the action could be manageably tried as a class action because the District Court could adopt the approach the Ninth Circuit approved in Hilao v. Estate of Marcos, 103 F. 3d 767, 782–787 (1996). There compensatory damages for some 9,541 class members were calculated by selecting 137 claims at random, referring those claims to a special master for valuation, and then extrapolating the validity and value of the untested claims from the sample set. See 603 F. 3d, at 625–626. The Court of Appeals “s[aw] no reason why a similar procedure to that used in Hilao could not be employed in this case.” Id., at 627. It would allow Wal-Mart “to present individual defenses in the randomly selected ‘sample cases,’ thus revealing the approximate percentage of class members whose unequal pay or nonpromotion was due to something other than gender discrimination.” Ibid., n. 56 (emphasis deleted). We granted certiorari. 562 U. S. ___ (2010). II The class action is “an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only.” Califano v. Yamasaki, 442 U. S. 682, 700–701 (1979). In order to justify a departure from that rule, “a class representative must be part of the class and ‘possess the same interest and suffer the same injury’ as the class members.” East Tex. Motor Freight System, Inc. v. Rodriguez, 431 U. S. 395, 403 (1977) (quoting Schlesinger v. Reservists Comm. to Stop the War, 418 U. S. 208, 216 (1974)). Rule 23(a) ensures that the named plaintiffs are appropriate representatives of the class whose claims they wish to litigate. The Rule’s four requirements—numerosity, commonality, typicality, and adequate representation—“effectively ‘limit the class claims to those fairly encompassed by the named plaintiff ’s claims.’ ” General Telephone Co. of Southwest v. Falcon, 457 U. S. 147, 156 (1982) (quoting General Telephone Co. of Northwest v. EEOC, 446 U. S. 318, 330 (1980)). A The crux of this case is commonality—the rule requiring a plaintiff to show that “there are questions of law or fact common to the class.” Rule 23(a)(2).[Footnote 5] That language is easy to misread, since “[a]ny competently crafted class complaint literally raises common ‘questions.’ ” Nagareda, Class Certification in the Age of Aggregate Proof, 84 N. Y. U. L. Rev. 97, 131–132 (2009). For example: Do all of us plaintiffs indeed work for Wal-Mart? Do our managers have discretion over pay? Is that an unlawful employment practice? What remedies should we get? Reciting these questions is not sufficient to obtain class certification. Commonality requires the plaintiff to demonstrate that the class members “have suffered the same injury,” Falcon, supra, at 157. This does not mean merely that they have all suffered a violation of the same pro-vision of law. Title VII, for example, can be violated in many ways—by intentional discrimination, or by hiring and promotion criteria that result in disparate impact, and by the use of these practices on the part of many different superiors in a single company. Quite obviously, the mere claim by employees of the same company that they have suffered a Title VII injury, or even a disparate-impact Title VII injury, gives no cause to believe that all their claims can productively be litigated at once. Their claims must depend upon a common contention—for example, the assertion of discriminatory bias on the part of the same supervisor. That common contention, moreover, must be of such a nature that it is capable of classwide resolution—which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke. “What matters to class certification … is not the raising of common ‘questions’—even in droves—but, rather the capacity of a classwide proceeding to generate common answers apt to drive the resolution of the litigation. Dissimilarities within the proposed class are what have the potential to impede the generation of common answers.” Nagareda, supra, at 132. Rule 23 does not set forth a mere pleading standard. A party seeking class certification must affirmatively demonstrate his compliance with the Rule—that is, he must be prepared to prove that there are in fact sufficiently numerous parties, common questions of law or fact, etc. We recognized in Falcon that “sometimes it may be necessary for the court to probe behind the pleadings before coming to rest on the certification question,” 457 U. S., at 160, and that certification is proper only if “the trial court is satisfied, after a rigorous analysis, that the prerequisites of Rule 23(a) have been satisfied,” id., at 161; see id., at 160 (“[A]ctual, not presumed, conformance with Rule 23(a) remains … indispensable”). Frequently that “rigorous analysis” will entail some overlap with the merits of the plaintiff ’s underlying claim. That cannot be helped. “ ‘[T]he class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiff ’s cause of action.’ ” Falcon, supra, at 160 (quoting Coopers & Lybrand v. Livesay, 437 U. S. 463, 469 (1978); some internal quotation marks omitted).[Footnote 6] Nor is there anything unusual about that consequence: The necessity of touching aspects of the merits in order to resolve preliminary matters, e.g., jurisdiction and venue, is a familiar feature of litigation. See Szabo v. Bridgeport Machines, Inc., 249 F. 3d 672, 676–677 (CA7 2001) (Easterbrook, J.). In this case, proof of commonality necessarily overlaps with respondents’ merits contention that Wal-Mart engages in a pattern or practice of discrimination.[Footnote 7] That is so because, in resolving an individual’s Title VII claim, the crux of the inquiry is “the reason for a particular employment decision,” Cooper v. Federal Reserve Bank of Richmond, 467 U. S. 867, 876 (1984). Here respondents wish to sue about literally millions of employment decisions at once. Without some glue holding the alleged reasons for all those decisions together, it will be impossible to say that examination of all the class members’ claims for relief will produce a common answer to the crucial question why was I disfavored. B This Court’s opinion in Falcon describes how the commonality issue must be approached. There an employee who claimed that he was deliberately denied a promotion on account of race obtained certification of a class comprising all employees wrongfully denied promotions and all applicants wrongfully denied jobs. 457 U. S., at 152. We rejected that composite class for lack of commonality and typicality, explaining: “Conceptually, there is a wide gap between (a) an individual’s claim that he has been denied a promotion [or higher pay] on discriminatory grounds, and his otherwise unsupported allegation that the company has a policy of discrimination, and (b) the existence of a class of persons who have suffered the same injury as that individual, such that the individual’s claim and the class claim will share common questions of law or fact and that the individual’s claim will be typical of the class claims.” Id., at 157–158. Falcon suggested two ways in which that conceptual gap might be bridged. First, if the employer “used a biased testing procedure to evaluate both applicants for employment and incumbent employees, a class action on behalf of every applicant or employee who might have been prejudiced by the test clearly would satisfy the commonality and typicality requirements of Rule 23(a).” Id., at 159, n. 15. Second, “[s]ignificant proof that an employer operated under a general policy of discrimination conceivably could justify a class of both applicants and employees if the discrimination manifested itself in hiring and promotion practices in the same general fashion, such as through entirely subjective decisionmaking processes.” Ibid. We think that statement precisely describes respondents’ burden in this case. The first manner of bridging the gap obviously has no application here; Wal-Mart has no testing procedure or other companywide evaluation method that can be charged with bias. The whole point of permitting discretionary decisionmaking is to avoid evaluating employees under a common standard. The second manner of bridging the gap requires “significant proof ” that Wal-Mart “operated under a general policy of discrimination.” That is entirely absent here. Wal-Mart’s announced policy forbids sex discrimination, see App. 1567a–1596a, and as the District Court recognized the company imposes penalties for denials of equal employment opportunity, 222 F. R. D., at 154. The only evidence of a “general policy of discrimination” respondents produced was the testimony of Dr. William Bielby, their sociological expert. Relying on “social framework” analysis, Bielby testified that Wal-Mart has a “strong corporate culture,” that makes it “ ‘vulnerable’ ” to “gender bias.” Id., at 152. He could not, however, “determine with any specificity how regularly stereotypes play a meaningful role in employment decisions at Wal-Mart. At his deposition … Dr. Bielby conceded that he could not calculate whether 0.5 percent or 95 percent of the employment decisions at Wal-Mart might be determined by stereotyped thinking.” 222 F. R. D. 189, 192 (ND Cal. 2004). The parties dispute whether Bielby’s testimony even met the standards for the admission of expert testimony under Federal Rule of Civil Procedure 702 and our Daubert case, see Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U. S. 579 (1993).[Footnote 8] The District Court concluded that Daubert did not apply to expert testimony at the certification stage of class-action proceedings. 222 F. R. D., at 191. We doubt that is so, but even if properly considered, Bielby’s testimony does nothing to advance respondents’ case. “[W]hether 0.5 percent or 95 percent of the employment decisions at Wal-Mart might be determined by stereotyped thinking” is the essential question on which respondents’ theory of commonality depends. If Bielby admittedly has no answer to that question, we can safely disregard what he has to say. It is worlds away from “significant proof ” that Wal-Mart “operated under a general policy of discrimination.” C The only corporate policy that the plaintiffs’ evidence convincingly establishes is Wal-Mart’s “policy” of allowing discretion by local supervisors over employment matters. On its face, of course, that is just the opposite of a uniform employment practice that would provide the commonality needed for a class action; it is a policy against having uniform employment practices. It is also a very common and presumptively reasonable way of doing business—one that we have said “should itself raise no inference of discriminatory conduct,” Watson v. Fort Worth Bank & Trust, 487 U. S. 977, 990 (1988). To be sure, we have recognized that, “in appropriate cases,” giving discretion to lower-level supervisors can be the basis of Title VII liability under a disparate-impact theory—since “an employer’s undisciplined system of subjective decisionmaking [can have] precisely the same effects as a system pervaded by impermissible intentional discrimination.” Id., at 990–991. But the recognition that this type of Title VII claim “can” exist does not lead to the conclusion that every employee in a company using a system of discretion has such a claim in common. To the contrary, left to their own devices most managers in any corporation—and surely most managers in a corporation that forbids sex discrimination—would select sex-neutral, performance-based criteria for hiring and promotion that produce no actionable disparity at all. Others may choose to reward various attributes that produce disparate impact—such as scores on general aptitude tests or educational achievements, see Griggs v. Duke Power Co., 401 U. S. 424, 431–432 (1971). And still other managers may be guilty of intentional discrimination that produces a sex-based disparity. In such a company, demonstrating the invalidity of one manager’s use of discretion will do nothing to demonstrate the invalidity of another’s. A party seeking to certify a nationwide class will be unable to show that all the employees’ Title VII claims will in fact depend on the answers to common questions. Respondents have not identified a common mode of exercising discretion that pervades the entire company—aside from their reliance on Dr. Bielby’s social frameworks analysis that we have rejected. In a company of Wal-Mart’s size and geographical scope, it is quite unbelievable that all managers would exercise their discretion in a common way without some common direction. Respondents attempt to make that showing by means of statistical and anecdotal evidence, but their evidence falls well short. The statistical evidence consists primarily of regression analyses performed by Dr. Richard Drogin, a statistician, and Dr. Marc Bendick, a labor economist. Drogin conducted his analysis region-by-region, comparing the number of women promoted into management positions with the percentage of women in the available pool of hourly workers. After considering regional and national data, Drogin concluded that “there are statistically significant disparities between men and women at Wal-Mart . . . [and] these disparities … can be explained only by gender discrimination.” 603 F. 3d, at 604 (internal quotation marks omitted). Bendick compared work-force data from Wal-Mart and competitive retailers and concluded that Wal-Mart “promotes a lower percentage of women than its competitors.” Ibid. Even if they are taken at face value, these studies are insufficient to establish that respondents’ theory can be proved on a classwide basis. In Falcon, we held that one named plaintiff ’s experience of discrimination was insufficient to infer that “discriminatory treatment is typical of [the employer’s employment] practices.” 457 U. S., at 158. A similar failure of inference arises here. As Judge Ikuta observed in her dissent, “[i]nformation about disparities at the regional and national level does not establish the existence of disparities at individual stores, let alone raise the inference that a company-wide policy of discrimination is implemented by discretionary decisions at the store and district level.” 603 F. 3d, at 637. A regional pay disparity, for example, may be attributable to only a small set of Wal-Mart stores, and cannot by itself establish the uniform, store-by-store disparity upon which the plaintiffs’ theory of commonality depends. There is another, more fundamental, respect in which respondents’ statistical proof fails. Even if it established (as it does not) a pay or promotion pattern that differs from the nationwide figures or the regional figures in all of Wal-Mart’s 3,400 stores, that would still not demonstrate that commonality of issue exists. Some managers will claim that the availability of women, or qualified women, or interested women, in their stores’ area does not mirror the national or regional statistics. And almost all of them will claim to have been applying some sex-neutral, performance-based criteria—whose nature and effects will differ from store to store. In the landmark case of ours which held that giving discretion to lower-level supervisors can be the basis of Title VII liability under a disparate-impact theory, the plurality opinion conditioned that holding on the corollary that merely proving that the discretionary system has produced a racial or sexual disparity is not enough. “[T]he plaintiff must begin by identifying the specific employment practice that is challenged.” Watson, 487 U. S., at 994; accord, Wards Cove Packing Co. v. Atonio, 490 U. S. 642, 656 (1989) (approving that statement), superseded by statute on other grounds, 42 U. S. C. §2000e–2(k). That is all the more necessary when a class of plaintiffs is sought to be certified. Other than the bare existence of delegated discretion, respondents have identified no “specific employment practice”—much less one that ties all their 1.5 million claims together. Merely showing that Wal-Mart’s policy of discretion has produced an overall sex-based disparity does not suffice. Respondents’ anecdotal evidence suffers from the same defects, and in addition is too weak to raise any inference that all the individual, discretionary personnel decisions are discriminatory. In Teamsters v. United States, 431 U. S. 324 (1977), in addition to substantial statistical evidence of company-wide discrimination, the Government (as plaintiff) produced about 40 specific accounts of racial discrimination from particular individuals. See id., at 338. That number was significant because the company involved had only 6,472 employees, of whom 571 were minorities, id., at 337, and the class itself consisted of around 334 persons, United States v. T.I.M.E.-D. C., Inc., 517 F. 2d 299, 308 (CA5 1975), overruled on other grounds, Teamsters, supra. The 40 anecdotes thus represented roughly one account for every eight members of the class. Moreover, the Court of Appeals noted that the anecdotes came from individuals “spread throughout” the company who “for the most part” worked at the company’s operational centers that employed the largest numbers of the class members. 517 F. 2d, at 315, and n. 30. Here, by contrast, respondents filed some 120 affidavits reporting experiences of discrimination—about 1 for every 12,500 class members—relating to only some 235 out of Wal-Mart’s 3,400 stores. 603 F. 3d, at 634 (Ikuta, J., dissenting). More than half of these reports are concentrated in only six States (Alabama, California, Florida, Missouri, Texas, and Wisconsin); half of all States have only one or two anecdotes; and 14 States have no anecdotes about Wal-Mart’s operations at all. Id., at 634–635, and n. 10. Even if every single one of these accounts is true, that would not demonstrate that the entire company “operate[s] under a general policy of discrimination,” Falcon, supra, at 159, n. 15, which is what respondents must show to certify a companywide class.[Footnote 9] The dissent misunderstands the nature of the foregoing analysis. It criticizes our focus on the dissimilarities be-tween the putative class members on the ground that we have “blend[ed]” Rule 23(a)(2)’s commonality requirement with Rule 23(b)(3)’s inquiry into whether common questions “predominate” over individual ones. See post, at 8–10 (Ginsburg, J., concurring in part and dissenting in part). That is not so. We quite agree that for purposes of Rule 23(a)(2) “ ‘[e]ven a single [common] question’ ” will do, post, at 10, n. 9 (quoting Nagareda, The Preexistence Principle and the Structure of the Class Action, 103 Colum. L. Rev. 149, 176, n. 110 (2003)). We consider dissimilarities not in order to determine (as Rule 23(b)(3) requires) whether common questions predominate, but in order to determine (as Rule 23(a)(2) requires) whether there is “[e]ven a single [common] question.” And there is not here. Because respondents provide no convincing proof of a companywide discriminatory pay and promotion policy, we have concluded that they have not established the existence of any common question.[Footnote 10] In sum, we agree with Chief Judge Kozinski that the members of the class: “held a multitude of different jobs, at different levels of Wal-Mart’s hierarchy, for variable lengths of time, in 3,400 stores, sprinkled across 50 states, with a kaleidoscope of supervisors (male and female), subject to a variety of regional policies that all differed… . Some thrived while others did poorly. They have little in common but their sex and this lawsuit.” 603 F. 3d, at 652 (dissenting opinion). III We also conclude that respondents’ claims for backpay were improperly certified under Federal Rule of Civil Procedure 23(b)(2). Our opinion in Ticor Title Ins. Co. v. Brown, 511 U. S. 117, 121 (1994) (per curiam) expressed serious doubt about whether claims for monetary relief may be certified under that provision. We now hold that they may not, at least where (as here) the monetary relief is not incidental to the injunctive or declaratory relief. A Rule 23(b)(2) allows class treatment when “the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole.” One possible reading of this provision is that it applies only to requests for such injunctive or declaratory relief and does not authorize the class certification of monetary claims at all. We need not reach that broader question in this case, because we think that, at a minimum, claims for individualized relief (like the backpay at issue here) do not satisfy the Rule. The key to the (b)(2) class is “the indivisible nature of the injunctive or declaratory remedy warranted—the notion that the conduct is such that it can be enjoined or declared unlawful only as to all of the class members or as to none of them.” Nagareda, 84 N. Y. U. L. Rev., at 132. In other words, Rule 23(b)(2) applies only when a single injunction or declaratory judgment would provide relief to each member of the class. It does not authorize class certification when each individual class member would be entitled to a different injunction or declaratory judgment against the defendant. Similarly, it does not authorize class certification when each class member would be entitled to an individualized award of monetary damages. That interpretation accords with the history of the Rule. Because Rule 23 “stems from equity practice” that predated its codification, Amchem Products, Inc. v. Windsor, 521 U. S. 591, 613 (1997), in determining its meaning we have previously looked to the historical models on which the Rule was based, Ortiz v. Fibreboard Corp., 527 U. S. 815, 841–845 (1999). As we observed in Amchem, “[c]ivil rights cases against parties charged with unlawful, class-based discrimination are prime examples” of what (b)(2) is meant to capture. 521 U. S., at 614. In particular, the Rule reflects a series of decisions involving challenges to racial segregation—conduct that was remedied by a single classwide order. In none of the cases cited by the Advisory Committee as examples of (b)(2)’s antecedents did the plaintiffs combine any claim for individualized relief with their classwide injunction. See Advisory Committee’s Note, 39 F. R. D. 69, 102 (1966) (citing cases); e.g., Potts v. Flax, 313 F. 2d 284, 289, n. 5 (CA5 1963); Brunson v. Board of Trustees of Univ. of School Dist. No. 1, Clarendon Cty., 311 F. 2d 107, 109 (CA4 1962) (per curiam); Frasier v. Board of Trustees of N.C., 134 F. Supp. 589, 593 (NC 1955) (three-judge court), aff’d, 350 U. S. 979 (1956). Permitting the combination of individualized and classwide relief in a (b)(2) class is also inconsistent with the structure of Rule 23(b). Classes certified under (b)(1) and (b)(2) share the most traditional justifications for class treatment—that individual adjudications would be impossible or unworkable, as in a (b)(1) class,[Footnote 11] or that the relief sought must perforce affect the entire class at once, as in a (b)(2) class. For that reason these are also mandatory classes: The Rule provides no opportunity for (b)(1) or (b)(2) class members to opt out, and does not even oblige the District Court to afford them notice of the action. Rule 23(b)(3), by contrast, is an “adventuresome innovation” of the 1966 amendments, Amchem, 521 U. S., at 614 (internal quotation marks omitted), framed for situations “in which ‘class-action treatment is not as clearly called for’,” id., at 615 (quoting Advisory Committee’s Notes, 28 U. S. C. App., p. 697 (1994 ed.)). It allows class certification in a much wider set of circumstances but with greater procedural protections. Its only prerequisites are that “the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” Rule 23(b)(3). And unlike (b)(1) and (b)(2) classes, the (b)(3) class is not mandatory; class members are entitled to receive “the best notice that is practicable under the circumstances” and to withdraw from the class at their option. See Rule 23(c)(2)(B). Given that structure, we think it clear that individ-ualized monetary claims belong in Rule 23(b)(3). The procedural protections attending the (b)(3) class—predominance, superiority, mandatory notice, and the right to opt out—are missing from (b)(2) not because the Rule considers them unnecessary, but because it considers them unnecessary to a (b)(2) class. When a class seeks an indivisible injunction benefitting all its members at once, there is no reason to undertake a case-specific inquiry into whether class issues predominate or whether class action is a superior method of adjudicating the dispute. Predominance and superiority are self-evident. But with respect to each class member’s individualized claim for money, that is not so—which is precisely why (b)(3) requires the judge to make findings about predominance and superiority before allowing the class. Similarly, (b)(2) does not require that class members be given notice and opt- out rights, presumably because it is thought (rightly or wrongly) that notice has no purpose when the class is mandatory, and that depriving people of their right to sue in this manner complies with the Due Process Clause. In the context of a class action predominantly for money damages we have held that absence of notice and opt-out violates due process. See Phillips Petroleum Co. v. Shutts, 472 U. S. 797, 812 (1985). While we have never held that to be so where the monetary claims do not predominate, the serious possibility that it may be so provides an additional reason not to read Rule 23(b)(2) to include the monetary claims here. B Against that conclusion, respondents argue that their claims for backpay were appropriately certified as part of a class under Rule 23(b)(2) because those claims do not “predominate” over their requests for injunctive and declaratory relief. They rely upon the Advisory Committee’s statement that Rule 23(b)(2) “does not extend to cases in which the appropriate final relief relates exclusively or predominantly to money damages.” 39 F. R. D., at 102 (emphasis added). The negative implication, they argue, is that it does extend to cases in which the appropriate final relief relates only partially and nonpredominantly to money damages. Of course it is the Rule itself, not the Advisory Committee’s description of it, that governs. And a mere negative inference does not in our view suffice to establish a disposition that has no basis in the Rule’s text, and that does obvious violence to the Rule’s structural features. The mere “predominance” of a proper (b)(2) injunctive claim does nothing to justify elimination of Rule 23(b)(3)’s procedural protections: It neither establishes the superiority of class adjudication over individual adjudication nor cures the notice and opt-out problems. We fail to see why the Rule should be read to nullify these protections whenever a plaintiff class, at its option, combines its monetary claims with a request—even a “predominating request”—for an injunction. Respondents’ predominance test, moreover, creates perverse incentives for class representatives to place at risk potentially valid claims for monetary relief. In this case, for example, the named plaintiffs declined to include employees’ claims for compensatory damages in their complaint. That strategy of including only backpay claims made it more likely that monetary relief would not “predominate.” But it also created the possibility (if the predominance test were correct) that individual class members’ compensatory-damages claims would be precluded by litigation they had no power to hold themselves apart from. If it were determined, for example, that a particular class member is not entitled to backpay because her denial of increased pay or a promotion was not the product of discrimination, that employee might be collaterally estopped from independently seeking compensatory damages based on that same denial. That possibility underscores the need for plaintiffs with individual monetary claims to decide for themselves whether to tie their fates to the class representatives’ or go it alone—a choice Rule 23(b)(2) does not ensure that they have. The predominance test would also require the District Court to reevaluate the roster of class members continually. The Ninth Circuit recognized the necessity for this when it concluded that those plaintiffs no longer employed by Wal-Mart lack standing to seek injunctive or declaratory relief against its employment practices. The Court of Appeals’ response to that difficulty, however, was not to eliminate all former employees from the certified class, but to eliminate only those who had left the company’s employ by the date the complaint was filed. That solution has no logical connection to the problem, since those who have left their Wal-Mart jobs since the complaint was filed have no more need for prospective relief than those who left beforehand. As a consequence, even though the validity of a (b)(2) class depends on whether “final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole,” Rule 23(b)(2) (emphasis added), about half the members of the class approved by the Ninth Circuit have no claim for injunctive or declaratory relief at all. Of course, the alternative (and logical) solution of excising plaintiffs from the class as they leave their employment may have struck the Court of Appeals as wasteful of the District Court’s time. Which indeed it is, since if a backpay action were properly certified for class treatment under (b)(3), the ability to litigate a plaintiff ’s backpay claim as part of the class would not turn on the irrelevant question whether she is still employed at Wal-Mart. What follows from this, however, is not that some arbitrary limitation on class membership should be imposed but that the backpay claims should not be certified under Rule 23(b)(2) at all. Finally, respondents argue that their backpay claims are appropriate for a (b)(2) class action because a backpay award is equitable in nature. The latter may be true, but it is irrelevant. The Rule does not speak of “equitable” remedies generally but of injunctions and declaratory judgments. As Title VII itself makes pellucidly clear, backpay is neither. See 42 U. S. C. §2000e–5(g)(2)(B)(i) and (ii) (distinguishing between declaratory and injunc-tive relief and the payment of “backpay,” see §2000e–5(g)(2)(A)). C In Allison v. Citgo Petroleum Corp., 151 F. 3d 402, 415 (CA5 1998), the Fifth Circuit held that a (b)(2) class would permit the certification of monetary relief that is “incidental to requested injunctive or declaratory relief,” which it defined as “damages that flow directly from liability to the class as a whole on the claims forming the basis of the injunctive or declaratory relief.” In that court’s view, such “incidental damage should not require additional hearings to resolve the disparate merits of each individual’s case; it should neither introduce new substantial legal or factual issues, nor entail complex individualized determinations.” Ibid. We need not decide in this case whether there are any forms of “incidental” monetary relief that are consistent with the interpretation of Rule 23(b)(2) we have announced and that comply with the Due Process Clause. Respondents do not argue that they can satisfy this standard, and in any event they cannot. Contrary to the Ninth Circuit’s view, Wal-Mart is entitled to individualized determinations of each employee’s eligibility for backpay. Title VII includes a detailed remedial scheme. If a plaintiff prevails in showing that an employer has discriminated against him in violation of the statute, the court “may enjoin the respondent from en-gaging in such unlawful employment practice, and order such affirmative action as may be appropriate, [including] reinstatement or hiring of employees, with or without backpay … or any other equitable relief as the court deems appropriate.” §2000e–5(g)(1). But if the employer can show that it took an adverse employment action against an employee for any reason other than discrimination, the court cannot order the “hiring, reinstatement, or promotion of an individual as an employee, or the payment to him of any backpay.” §2000e–5(g)(2)(A). We have established a procedure for trying pattern-or-practice cases that gives effect to these statutory requirements. When the plaintiff seeks individual relief such as reinstatement or backpay after establishing a pattern or practice of discrimination, “a district court must usually conduct additional proceedings … to determine the scope of individual relief.” Teamsters, 431 U. S., at 361. At this phase, the burden of proof will shift to the company, but it will have the right to raise any individual affirmative defenses it may have, and to “demonstrate that the individual applicant was denied an employment opportunity for lawful reasons.” Id., at 362. The Court of Appeals believed that it was possible to replace such proceedings with Trial by Formula. A sample set of the class members would be selected, as to whom liability for sex discrimination and the backpay owing as a result would be determined in depositions supervised by a master. The percentage of claims determined to be valid would then be applied to the entire remaining class, and the number of (presumptively) valid claims thus derived would be multiplied by the average backpay award in the sample set to arrive at the entire class recovery—without further individualized proceedings. 603 F. 3d, at 625–627. We disapprove that novel project. Because the Rules Enabling Act forbids interpreting Rule 23 to “abridge, enlarge or modify any substantive right,” 28 U. S. C. §2072(b); see Ortiz, 527 U. S., at 845, a class cannot be certified on the premise that Wal-Mart will not be entitled to litigate its statutory defenses to individual claims. And because the necessity of that litigation will prevent backpay from being “incidental” to the classwide injunction, respondents’ class could not be certified even assuming, arguendo, that “incidental” monetary relief can be awarded to a 23(b)(2) class. * * * The judgment of the Court of Appeals is Reversed. Footnote 1 The complaint included seven named plaintiffs, but only three remain part of the certified class as narrowed by the Court of Appeals. Footnote 2 Rule 23(b)(1) allows a class to be maintained where “prosecuting separate actions by or against individual class members would create a risk of ” either “(A) inconsistent or varying adjudications,” or “(B) adjudications … that, as a practical matter, would be dispositive of the interests of the other members not parties to the individual adjudications or would substantially impair or impeded their ability to protect their interests.” Rule 23(b)(3) states that a class may be maintained where “questions of law or fact common to class members predominate over any questions affecting only individual members,” and a class action would be “superior to other available methods for fairly and efficiently adjudicating the controversy.” The applicability of these provisions to the plaintiff class is not before us. Footnote 3 The District Court excluded backpay claims based on promotion opportunities that had not been publicly posted, for the reason that no applicant data could exist for such positions. 222 F. R. D. 137, 182 (ND Cal. 2004). It also decided to afford class members notice of the ac- tion and the right to opt-out of the class with respect to respondents’ punitive-damages claim. Id., at 173. Footnote 4 To enable that result, the Court of Appeals trimmed the (b)(2) class in two ways: First, it remanded that part of the certification order which included respondents’ punitive-damages claim in the (b)(2) class, so that the District Court might consider whether that might cause the monetary relief to predominate. 603 F. 3d, at 621. Second, it accepted in part Wal-Mart’s argument that since class members whom it no longer employed had no standing to seek injunctive or declaratory relief, as to them monetary claims must predominate. It excluded from the certified class “those putative class members who were no longer Wal-Mart employees at the time Plaintiffs’ complaint was filed,” id., at 623 (emphasis added). Footnote 5 We have previously stated in this context that “[t]he commonality and typicality requirements of Rule 23(a) tend to merge. Both serve as guideposts for determining whether under the particular circumstances maintenance of a class action is economical and whether the named plaintiff’s claim and the class claims are so interrelated that the interests of the class members will be fairly and adequately protected in their absence. Those requirements therefore also tend to merge with the adequacy-of-representation requirement, although the latter requirement also raises concerns about the competency of class counsel and conflicts of interest.” General Telephone Co. of Southwest v. Falcon, 457 U. S. 147, 157–158, n. 13 (1982). In light of our disposition of the commonality question, however, it is unnecessary to resolve whether respondents have satisfied the typicality and adequate-representation requirements of Rule 23(a). Footnote 6 A statement in one of our prior cases, Eisen v. Carlisle & Jacquelin, 417 U. S. 156, 177 (1974), is sometimes mistakenly cited to the contrary: “We find nothing in either the language or history of Rule 23 that gives a court any authority to conduct a preliminary inquiry into the merits of a suit in order to determine whether it may be maintained as a class action.” But in that case, the judge had conducted a preliminary inquiry into the merits of a suit, not in order to determine the propriety of certification under Rules 23(a) and (b) (he had already done that, see id., at 165), but in order to shift the cost of notice required by Rule 23(c)(2) from the plaintiff to the defendants. To the extent the quoted statement goes beyond the permissibility of a merits inquiry for any other pretrial purpose, it is the purest dictum and is contradicted by our other cases. Perhaps the most common example of considering a merits question at the Rule 23 stage arises in class-action suits for securities fraud. Rule 23(b)(3)’s requirement that “questions of law or fact common to class members predominate over any questions affecting only individual members” would often be an insuperable barrier to class certification, since each of the individual investors would have to prove reliance on the alleged misrepresentation. But the problem dissipates if the plaintiffs can establish the applicability of the so-called “fraud on the market” presumption, which says that all traders who purchase stock in an efficient market are presumed to have relied on the accuracy of a company’s public statements. To invoke this presumption, the plaintiffs seeking 23(b)(3) certification must prove that their shares were traded on an efficient market, Erica P. John Fund, Inc. v. Halliburton Co., 563 U. S. ___, ___ (2011) (slip op., at 5), an issue they will surely have to prove again at trial in order to make out their case on the merits. Footnote 7 In a pattern-or-practice case, the plaintiff tries to “establish by a preponderance of the evidence that … discrimination was the company’s standard operating procedure[,] the regular rather than the unusual practice.” Teamsters v. United States, 431 U. S. 324, 358 (1977); see also Franks v. Bowman Transp. Co., 424 U. S. 747, 772 (1976). If he succeeds, that showing will support a rebuttable inference that all class members were victims of the discriminatory practice, and will justify “an award of prospective relief,” such as “an injunctive order against the continuation of the discriminatory practice.” Teamsters, supra, at 361. Footnote 8 Bielby’s conclusions in this case have elicited criticism from the very scholars on whose conclusions he relies for his social-framework analysis. See Monahan, Walker, & Mitchell, Contextual Evidence of Gender Discrimination: The Ascendance of “Social Frameworks,” 94 Va. L. Rev. 1715, 1747 (2008) (“[Bielby’s] research into conditions and be-havior at Wal-Mart did not meet the standards expected of social scientific research into stereotyping and discrimination”); id., at 1745, 1747 (“[A] social framework necessarily contains only general statements about reliable patterns of relations among variables … and goes no further… . Dr. Bielby claimed to present a social framework, but he testified about social facts specific to Wal-Mart”); id., at 1747–1748 (“Dr. Bielby’s report provides no verifiable method for measuring and testing any of the variables that were crucial to his conclusions and reflects nothing more than Dr. Bielby’s ‘expert judgment’ about how general stereotyping research applied to all managers across all of Wal-Mart’s stores nationwide for the multi-year class period”). Footnote 9 The dissent says that we have adopted “a rule that a discrimination claim, if accompanied by anecdotes, must supply them in numbers proportionate to the size of the class.” Post, at 5, n. 4 (Ginsburg, J., concurring in part and dissenting in part). That is not quite accurate. A discrimination claimant is free to supply as few anecdotes as he wishes. But when the claim is that a company operates under a general policy of discrimination, a few anecdotes selected from literally millions of employment decisions prove nothing at all. Footnote 10 For this reason, there is no force to the dissent’s attempt to distinguish Falcon on the ground that in that case there were “ ‘no common questions of law or fact’ between the claims of the lead plaintiff and the applicant class ” post, at 9, n. 7 (quoting Falcon, 457 U. S., at 162 (Burger, C. J., concurring in part and dissenting in part)). Here also there is nothing to unite all of the plaintiffs’ claims, since (contrary to the dissent’s contention, post, at 9, n. 7), the same employment practices do not “touch and concern all members of the class.” Footnote 11 Rule 23(b)(1) applies where separate actions by or against individual class members would create a risk of “establish[ing] incompatible standards of conduct for the party opposing the class,” Rule 23(b)(1)(A), such as “where the party is obliged by law to treat the members of the class alike,” Amchem Products, Inc. v. Windsor, 521 U. S. 591, 614 (1997), or where individual adjudications “as a practical matter, would be dispositive of the interests of the other members not parties to the individual adjudications or would substantially impair or impede their ability to protect their interests,” Rule 23(b)(1)(B), such as in “ ‘limited fund’ cases, … in which numerous persons make claims against a fund insufficient to satisfy all claims,” Amchem, supra, at 614.

562.US.2010_09-996

While most States set determinate time limits for collateral relief applications, California courts “appl[y] a general ‘reasonableness’ standard” to judge whether a habeas petition is timely filed, Carey v. Saffold, 536 U. S. 214, 222. Under that standard, “a [habeas] petition should be filed as promptly as the circumstances allow … ,” In re Clark, 5 Cal. 4th 750, 765, n. 5, 855 P. 2d 729, 738, n. 5. Three decisions, Clark, In re Robbins, 18 Cal. 4th 770, 959 P. 2d 311, and In re Gallego, 18 Cal. 4th 825, 959 P. 2d 290, describe California’s timeliness requirement. A prisoner must seek habeas relief without “substantial delay,” e.g., Robbins, 18 Cal. 4th, at 780, 959 P. 3d, at 317, as “measured from the time the petitioner or counsel knew, or reasonably should have known, of the information offered in support of the claim and the legal basis of the claim,” id., at 787, 959 P. 2d, at 322. All California courts have “original jurisdiction in habeas corpus proceedings.” Cal. Const., Art. VI, §10. Because a habeas petitioner may skip over the lower courts and file directly in the California Supreme Court, that court rules on a staggering number of habeas petitions each year. A summary denial citing Clark and Robbins means that the petition is rejected as untimely. California courts, however, have discretion to bypass a timeliness issue and, instead, summarily reject the petition for want of merit. Respondent Martin was convicted of murder and robbery, and was sentenced to life in prison without parole. After the California Supreme Court denied Martin’s first state habeas petition, he filed a federal habeas petition. The District Court ordered a stay to permit Martin to return to state court to raise ineffective-assistance-of-counsel claims he had not previously aired. Martin raised those claims in his second habeas petition in the California Supreme Court, but gave no reason for his failure to assert the additional claims until nearly five years after his sentence and conviction became final. The California Supreme Court denied the petition, citing Clark and Robbins. Having exhausted his state-court remedies, Martin filed an amended federal habeas petition. The District Court dismissed his belatedly asserted claims as untimely under California law. The Ninth Circuit vacated that order and directed the District Court to determine the “adequacy” of the State’s time bar. Again rejecting Martin’s petition, the District Court found California’s bar an adequate state ground for denying Martin’s new pleas. Concluding that the time bar was not firmly defined or consistently applied, the Ninth Circuit remanded for a determination of the merits of Martin’s claims. Held: California’s timeliness requirement qualifies as an independent state ground adequate to bar habeas corpus relief in federal court. Pp. 7–13. (a) Absent showings of “cause” and “prejudice,” see Wainwright v. Sykes, 433 U. S. 72, 84–85, federal habeas relief will be unavailable when (1) “a state court [has] declined to address a prisoner’s federal claims because the prisoner had failed to meet a state procedural requirement,” and (2) “the state judgment rests on independent and adequate state procedural grounds,” Coleman v. Thompson, 501 U. S. 722, 729–730. P. 7. (b) A “rule can be ‘firmly established’ and ‘regularly followed,’ ” and therefore adequate, “even if the appropriate exercise of discretion may permit consideration of a federal claim in some cases but not others.” Beard v. Kindler, 558 U. S. ___, ___. California’s time rule, although discretionary, meets this “firmly established” criterion. The California Supreme Court framed the requirement in a trilogy of cases, instructing habeas petitioners to “alleg[e] with specificity” the absence of substantial delay, good cause for delay, or eligibility for one of four exceptions to the time bar. Gallego, 18 Cal. 4th, at 838, 959 P. 2d, at 299. And California’s case law made it plain that Martin’s nearly five-year delay was “substantial.” See, e.g., id., at 829–831, 838, and n. 13, 959 P. 2d, at 293–294, 299, and n. 13. The Court finds unpersuasive Martin’s argument that the terms “reasonable time” period and “substantial delay” make California’s rule too vague to be regarded as “firmly established.” While indeterminate language is typical of discretionary rules, application of those rules in particular circumstances can supply the requisite clarity. Congressional statutes and this Court’s decisions have employed time limitations that are not stated in precise, numerical terms. For example, current federal habeas prescriptions limit the time for filing a petition to one year. The clock runs from “the date on which the [supporting] facts … could have been discovered through … due diligence.” 28 U. S. C. §2255(f)(4). Although “ ‘due diligence’ is an inexact measure of how much delay is too much,” Johnson v. United States, 544 U. S. 295, 309, n. 7, “use of an imprecise standard is no justification for depriving [a rule’s] language of any meaning,” ibid. Nor is California’s time rule vulnerable on the ground that it is not regularly followed. Each year, the California State Supreme Court summarily denies hundreds of habeas petitions by citing Clark and Robbins. Contrary to Martin’s argument, California’s time bar is not infirm simply because a court may opt to bypass the Clark/Robbins assessment and summarily dismiss a petition on the merits, if that is the easier path. Nor should a discretionary rule be disregarded automatically upon a showing that outcomes under the rule vary from case to case. Discretion enables a court to home in on case-specific considerations and to avoid the harsh results that may attend consistent application of an unyielding rule. A state ground may be found inadequate when a court has exercised its discretion in a surprising or unfair manner, but Martin makes no such contention here. Pp. 7–12. (c) This decision leaves unaltered the Court’s repeated recognition that federal courts must carefully examine state procedural requirements to ensure that they do not operate to discriminate against claims of federal rights. See, e.g., Brown v. Western R. Co. of Ala., 338 U. S. 294, 298–299. On the record here, however, there is no basis for concluding that California’s rule operates in such a discriminatory manner. P. 13. 357 Fed. Appx. 793, reversed. Ginsburg, J., delivered the opinion for a unanimous Court.

This case concerns California’s time limitation on applications for postconviction (habeas corpus) relief. The question presented: Does California’s timeliness requirement qualify as an independent state ground adequate to bar habeas corpus relief in federal court? California does not employ fixed statutory deadlines to determine the timeliness of a state prisoner’s petition for habeas corpus. Instead, California directs petitioners to file known claims “as promptly as the circumstances allow.” In re Clark, 5 Cal. 4th 750, 765, n. 5, 855 P. 2d 729, 738, n. 5 (1993). Petitioners are further instructed to state when they first learned of the asserted claims and to explain why they did not seek postconviction relief sooner. In re Robbins, 18 Cal. 4th 770, 780, 959 P. 2d 311, 317–318 (1998). Claims substantially delayed without justification may be denied as untimely. Ibid.; Clark, 5 Cal. 4th, at 765, n. 5, 855 P. 2d, at 738, n. 5. California courts signal that a habeas petition is denied as untimely by citing the controlling decisions, i.e., Clark and Robbins. A spare order denying a petition without explanation or citation ordinarily ranks as a disposition on the merits. Tr. of Oral Arg. 7; see Harrington v. Richter, ante, at 9–10. California courts may elect to pretermit the question whether a petition is timely and simply deny the petition, thereby signaling that the petition lacks merit. Petitioner below, respondent here, Charles W. Martin, presented the claims at issue—all alleging ineffective assistance of counsel—in a habeas petition filed in the California Supreme Court nearly five years after his conviction became final. He stated no reason for the long delay. Citing Clark and Robbins, the court denied Martin’s petition. In turn, the U. S. District Court for the Eastern District of California dismissed Martin’s federal habeas petition raising the same ineffective assistance claims. Denial of Martin’s state-court petition as untimely, the District Court held, rested on an adequate and independent state ground, i.e., Martin’s failure to seek relief in state court “without substantial delay.” See Robbins, 18 Cal. 4th, at 787, 959 P. 2d, at 322. The U. S. Court of Appeals for the Ninth Circuit reversed the District Court’s decision. Contrasting the precision of “fixed statutory deadlines” with California’s proscription of “substantial delay,” the appeals court held that California’s standard lacked the clarity and certainty necessary to constitute an adequate state bar. 357 Fed. Appx. 793, 794 (2009) (relying on Townsend v. Knowles, 562 F. 3d 1200 (CA9 2009)). In a recent decision, Beard v. Kindler, 558 U. S. ___ (2009), this Court clarified that a state procedural bar may count as an adequate and independent ground for denying a federal habeas petition even if the state court had discretion to reach the merits despite the default. Guided by that decision, we hold that California is not put to the choice of imposing a specific deadline for habeas petitions (which would almost certainly rule out Martin’s nearly five-year delay) or preserving the flexibility of current practice, “but only at the cost of undermining the finality of state court judgments.” Id., at ___ (slip op., at 7). In so ruling, we stress that Martin has not alleged that California’s time bar, either by design or in operation, discriminates against federal claims or claimants. I A While most States set determinate time limits for collateral relief applications, in California, neither statute nor rule of court does so. Instead, California courts “appl[y] a general ‘reasonableness’ standard” to judge whether a habeas petition is timely filed. Carey v. Saffold, 536 U. S. 214, 222 (2002). The basic instruction provided by the California Supreme Court is simply that “a [habeas] petition should be filed as promptly as the circumstances allow … .” Clark, 5 Cal. 4th, at 765, n. 5, 855 P. 2d, at 738, n. 5. Three leading decisions describe California’s timeliness requirement: Robbins, Clark, and In re Gallego, 18 Cal. 4th 825, 959 P. 2d 290 (1998). A prisoner must seek habeas relief without “substantial delay,” Robbins, 18 Cal. 4th, at 780, 959 P. 2d, at 317; Gallego, 18 Cal. 4th, at 833, 959 P. 2d, at 296; Clark, 5 Cal. 4th, at 783, 855 P. 2d, at 750, as “measured from the time the petitioner or counsel knew, or reasonably should have known, of the information offered in support of the claim and the legal basis for the claim,” Robbins, 18 Cal. 4th, at 787, 959 P. 2d, at 322. Petitioners in noncapital cases have “the burden of establishing (i) absence of substantial delay, (ii) good cause for the delay, or (iii) that the claim falls within an exception to the bar of untimeliness.” Id., at 780, 959 P. 2d, at 317.[Footnote 1] California’s collateral review regime differs from that of other States in a second notable respect: All California courts “have original jurisdiction in habeas corpus proceedings,” Cal. Const., Art. VI, §10, thus “no appeal lies from the denial of a petition for writ of habeas corpus,” Clark, 5 Cal. 4th, at 767, n. 7, 855 P. 2d, at 740, n. 7. “[A] prisoner whose petition has been denied by the superior court can obtain review of his claims only by the filing of a new petition in the Court of Appeal.” Ibid. The new petition, however, must be confined to claims raised in the initial petition. See In re Martinez, 46 Cal. 4th 945, 956, 209 P. 3d 908, 915 (2009). Because a habeas petitioner may skip over the lower courts and file directly in the California Supreme Court, In re Kler, 188 Cal. App. 4th 1399, 1403, 115 Cal. Rptr. 3d 889, 891–892 (2010), that court rules on a staggering number of habeas petitions each year.[Footnote 2] The court issues generally unelaborated “summary denials” of petitions that “d[o] not state a prima facie case for relief” or that contain “claims [that] are all procedurally barred.” People v. Romero, 8 Cal. 4th 728, 737, 883 P. 2d 388, 391 (1994) (internal quotation marks omitted). A summary denial citing Clark and Robbins means that the petition is rejected as untimely. See, e.g., Brief for Habeas Corpus Resource Center as Amicus Curiae 20, and n. 23. California courts have discretion, however, to bypass a timeliness issue and, instead, summarily reject the petition for want of merit. See Robbins, 18 Cal. 4th, at 778, n. 1, 959 P. 2d, at 316, n. 1. See also Saffold, 536 U. S., at 225–226. B In December 1986, Charles Martin participated in a robbery and murder in California. Martin fled the State, but eight years later he was extradited to California to stand trial. Convicted in state court of murder and robbery, Martin was sentenced to life in prison without the possibility of parole. In 1997, the California Court of Appeal affirmed his conviction and sentence, and the California Supreme Court denied review. Martin initiated his first round of state habeas proceedings in 1998, and the next year, the California Supreme Court denied his petition. He then filed a habeas petition in the appropriate U. S. District Court. Finding that Martin’s federal petition included ineffective-assistance-of-counsel claims he had not aired in state court, the District Court stayed the federal proceedings pending Martin’s return to state court to exhaust his remedies there.[Footnote 3] In March 2002, Martin filed his second habeas petition in the California Supreme Court, raising the federal ineffective assistance claims his earlier filing omitted. He gave no reason for his failure to assert the additional claims until nearly five years after his sentence and conviction became final. Tr. of Oral Arg. 36, 39. In September 2002, the California Supreme Court denied Martin’s petition in an order typical of that court’s summary dispositions for failure to file “as promptly as the circumstances allow.” Clark, 5 Cal. 4th, at 765, n. 5, 855 P. 2d, at 738, n. 5. The order read in its entirety: “Petition for writ of habeas corpus is DENIED. (See In re Clark (1993) 5 Cal. 4th 750, In re Robbins (1998) 18 Cal. 4th 770, 780.).” See App. to Pet. for Cert. 60. Having exhausted state postconviction remedies, Martin returned to federal court and filed an amended petition. Based upon the California Supreme Court’s time-bar disposition, the District Court dismissed Martin’s belatedly asserted claims as procedurally precluded. Id., at 27, 57. The Ninth Circuit vacated the dismissal order and remanded the case, directing the District Court to determine the “adequacy” of the State’s time bar. Martin v. Hubbard, 192 Fed. Appx. 616, 618 (2006). The District Court again rejected Martin’s petition, stating that “[t]he California timeliness bar as set forth in … Clark/Robbins is clearly defined, well established and consistently applied.” App. to Pet. for Cert. 4. The Ninth Circuit again disagreed. Controlled by its prior decision in Townsend, 562 F. 3d, at 1207–1208, the Court of Appeals held that California’s time bar “has yet to be firmly defined” and was not shown by the State to be “consistently applied.” 357 Fed. Appx., at 794. The remand order directed the District Court to determine the merits of the claims Martin asserted in his second petition to the California Supreme Court. We granted certiorari, 561 U. S. ___ (2010), to determine the “adequacy” of California’s practice under which a prisoner may be barred from collaterally attacking his conviction when he has “substantially delayed” filing his habeas petition. Martin does not here dispute that the time limitation is an “independent” state ground. See Brief in Opposition 5–6. See also Bennett v. Mueller, 322 F. 3d 573, 582–583 (CA9 2003). Nor does he contend that he established “cause” and “prejudice,” i.e., cause for the delay in asserting his claims and actual prejudice resulting from the State’s alleged violation of his constitutional rights. See Wainwright v. Sykes, 433 U. S. 72, 87–91 (1977). II A “A federal habeas court will not review a claim rejected by a state court ‘if the decision of [the state] court rests on a state law ground that is independent of the federal question and adequate to support the judgment.’ ” Kind-ler, 558 U. S., at ___ (slip op., at 1) (quoting Coleman v. Thompson, 501 U. S. 722, 729 (1991)). The state-law ground may be a substantive rule dispositive of the case, or a procedural barrier to adjudication of the claim on the merits. See Sykes, 433 U. S., at 81–82, 90. Ordinarily, a state prisoner seeking federal habeas relief must first “exhaus[t] the remedies available in the courts of the State,” 28 U. S. C. §2254(b)(1)(A), thereby affording those courts “the first opportunity to address and correct alleged violations of [the] prisoner’s federal rights,” Coleman, 501 U. S., at 731. The adequate and independent state ground doctrine furthers that objective, for without it, “habeas petitioners would be able to avoid the exhaustion requirement by defaulting their federal claims in state court.” Id., at 732. Accordingly, absent showings of “cause” and “prejudice,” see Sykes, 433 U. S., at 84–85, federal habeas relief will be unavailable when (1) “a state court [has] declined to address a prisoner’s federal claims because the prisoner had failed to meet a state procedural requirement,” and (2) “the state judgment rests on independent and adequate state procedural grounds.” Coleman, 501 U. S., at 729–730. B To qualify as an “adequate” procedural ground, a state rule must be “firmly established and regularly followed. ” Kindler, 558 U. S., at ___ (slip op., at 7) (internal quotation marks omitted).[Footnote 4] “[A] discretionary state procedural rule,” we held in Kindler, “can serve as an adequate ground to bar federal habeas review.” Ibid. A “rule can be ‘firmly established’ and ‘regularly followed,’ ” Kindler observed, “even if the appropriate exercise of discretion may permit consideration of a federal claim in some cases but not others.” Ibid. California’s time rule, although discretionary, meets the “firmly established” criterion, as Kindler comprehended that requirement. The California Supreme Court, as earlier noted, framed the timeliness requirement for habeas petitioners in a trilogy of cases. See supra, at 3. Those decisions instruct habeas petitioners to “alleg[e] with specificity” the absence of substantial delay, good cause for delay, or eligibility for one of four exceptions to the time bar. Gallego, 18 Cal. 4th, at 838, 959 P. 2d, at 299; see Robbins, 18 Cal. 4th, at 780, 959 P. 2d, at 317.[Footnote 5] And California’s case law made it altogether plain that Martin’s delay of nearly five years ranked as “substantial.” See Gallego, 18 Cal. 4th, at 829–831, 838, and n. 13, 959 P. 2d, at 293–294, 299, and n. 13 (delay of four years barred claim); In re Tsaturyan, No. B156012, 2002 WL 1614107, *3 (Cal. App., July 23, 2002) (delay of 16 months barred claim). See also In re Miller, No. B186447, 2006 WL 1980385, *2–3 (Cal. App., July 17, 2006) (delay of two years and six months barred claim). Martin nevertheless urges that California’s rule is too vague to be regarded as “firmly established.” “[R]eason-able time” period and “substantial delay,” he maintains, are “meaningless terms.” Brief for Respondent 48 (internal quotation marks omitted). We disagree. Indeterminate language is typical of discretionary rules. Application of those rules in particular circumstances, however, can supply the requisite clarity. Congressional statutes and this Court’s decisions, we note, have employed time limitations that are not stated in precise, numerical terms. Former Federal Habeas Corpus Rule 9(a), for example, set no fixed time limit on submission of habeas petitions. The Rule permitted dismissal of a state prisoner’s petition when it appeared that delay in commencing litigation “prejudiced [the State] in its ability to respond.” 28 U. S. C. §2254 Rule 9(a) (1994 ed.). To stave off dismissal, the petitioner had to show that he could not earlier have known, “by the exercise of reasonable diligence,” the grounds on which he based the petition. Ibid. In Rhines v. Weber, 544 U. S. 269 (2005), we instructed district courts, when employing stay and abeyance procedure, see supra, at 5, n. 3, to “place reasonable time limits on a petitioner’s trip to state court and back.” 544 U. S., at 278. Current federal habeas prescriptions limit the time for filing a petition to one year. The clock runs from “the date on which the [supporting] facts … could have been discovered through the exercise of due diligence.” 28 U. S. C. §2255(f)(4) (2006 ed., Supp. III) (applicable to federal prisoners); see §2244(d)(1)(D) (2006 ed.) (similar provision applicable to state prisoners). “[D]ue diligence,” we have observed, “ is an inexact measure of how much delay is too much.” Johnson v. United States, 544 U. S. 295, 309, n. 7 (2005) (internal quotation marks omitted). But “use of an imprecise standard,” we immediately added, “is no justification for depriving [a rule’s] language of any meaning.” Ibid. “[I]t would seem particularly strange to disregard state procedural rules that are substantially similar to those to which we give full force in our own courts.” Kind- ler, 558 U. S., at __ (slip op., at 8). Nor is California’s time rule vulnerable on the ground that it is not regularly followed. Each year, the California Supreme Court summarily denies hundreds of habeas petitions by citing Clark and Robbins. Brief for Appellant in No. 08–15752 (CA9), pp. 31–32. On the same day the court denied Martin’s petition, it issued 21 other Clark/Robbins summary denials. See Brief for Habeas Corpus Resource Center as Amicus Curiae 20. In reasoned opinions, too, California courts regularly invoke Clark, Robbins, and Gallego to determine whether a habeas petition is time barred.[Footnote 6] Martin argued below that California’s time bar is not regularly followed in this sense: Use of summary denials makes it “impossible to tell” why the California Supreme Court “decides some delayed petitions on the merits and rejects others as untimely.” Brief for Appellant in No. 08–15752 (CA9), pp. 37–38. We see no reason to reject California’s time bar simply because a court may opt to bypass the Clark/Robbins assessment and summarily dismiss a petition on the merits, if that is the easier path. See, e.g., Strickland v. Washington, 466 U. S. 668, 697 (1984) (“[A] court need not determine whether counsel’s performance was deficient … [i]f it is easier to dispose of an ineffectiveness claim on the ground of lack of sufficient prejudice … .”); cf. Ruhrgas AG v. Marathon Oil Co., 526 U. S. 574, 585 (1999) (“It is hardly novel for a federal court to choose among threshold grounds for denying audience to a case on the merits.”). The Ninth Circuit concluded that California’s time bar is not consistently applied because outcomes under the rule vary from case to case. See 357 Fed. Appx., at 794. For example, in People v. Fairbanks, No. C047810, 2006 WL 950267, *2–*3 (Cal. App., Apr. 11, 2006), a one-year delay was found substantial, while in In re Little, No. D047468, 2008 WL 142832, *4, n. 6 (Cal. App., Jan. 16, 2008), a delay of 14 months was determined to be insubstantial. A discretionary rule ought not be disregarded automatically upon a showing of seeming inconsistencies.[Footnote 7] Discretion enables a court to home in on case-specific considerations and to avoid the harsh results that sometimes attend consistent application of an unyielding rule. See Prihoda v. McCaughtry, 910 F. 2d 1379, 1385 (CA7 1990) (“Uncertainty is not enough to disqualify a state’s procedural ground as one ‘adequate’ under federal law. If it were, states would be induced to make their rules draconian … .”). A state ground, no doubt, may be found inadequate when “discretion has been exercised to impose novel and unforeseeable requirements without fair or substantial support in prior state law … .” 16B C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §4026, p. 386 (2d ed. 1996) (hereinafter Wright & Miller); see Prihoda, 910 F. 2d, at 1383 (state ground “applied infrequently, unexpectedly, or freakishly” may “discriminat[e] against the federal rights asserted” and therefore rank as “inadequate”). Martin does not contend, however, that in his case, the California Supreme Court exercised its discretion in a surprising or unfair manner. “[S]ound procedure often requires discretion to exact or excuse compliance with strict rules,” 16B Wright & Miller §4028, p. 403, and we have no cause to discourage standards allowing courts to exercise such discretion. As this Court observed in Kindler, if forced to choose between mandatory rules certain to be found “adequate,” or more supple prescriptions that federal courts may disregard as “inadequate,” “many States [might] opt for mandatory rules to avoid the high costs that come with plenary federal review.” 558 U. S., at ___ (slip op., at 7). “Th[at] result would be particularly unfortunate for [habeas petitioners], who would lose the opportunity to argue that a procedural default should be excused through the exercise of judicial discretion.” Id., at ___ (slip op., at 8).[Footnote 8] C Today’s decision, trained on California’s timeliness rule for habeas petitions, leaves unaltered this Court’s repeated recognition that federal courts must carefully examine state procedural requirements to ensure that they do not operate to discriminate against claims of federal rights. See Brown v. Western R. Co. of Ala., 338 U. S. 294, 298–299 (1949); Davis v. Wechsler, 263 U. S. 22, 24–25 (1923); 16B Wright & Miller §4026, p. 386 (noting “risk that discretionary procedural sanctions may be invoked more harshly against disfavored federal rights, … deny[ing] [litigants] a fair opportunity to present federal claims”). See also Kindler, 558 U. S., at ___ (Kennedy, J., concurring) (slip op., at 3) (a state procedural ground would be inadequate if the challenger shows a “purpose or pattern to evade constitutional guarantees”). On the record before us, however, there is no basis for concluding that California’s timeliness rule operates to the particular disadvantage of petitioners asserting federal rights. * * * For the reasons stated, we find no inadequacy in California’s timeliness rule generally or as applied in Martin’s case. The judgment of the United States Court of Appeals for the Ninth Circuit is therefore Reversed. Footnote 1 A petition for habeas relief in a capital case is “presumed to be filed without substantial delay if it is filed within 180 days after the final due date for the filing of [an] appellant’s reply brief on the direct appeal … .” California Supreme Court Policies Regarding Cases Arising From Judgments of Death, Policy 3, Standard 1–1.1 (2010). Footnote 2 In fiscal year 2008–2009, the California Supreme Court issued dispositions in 3,258 original habeas actions. Judicial Council of California, 2010 Court Statistics Report, Statewide Caseload Trends, 1999–2000 Through 2008–2009, p. 6, http://www.courtinfo.ca.gov/reference/ documents/csr2010.pdf (as visited Feb. 15, 2011, and in Clerk of Court’s case file). During a similar time period, a total of 2,210 habeas cases were on this Court’s docket. See October Term 2008 Filings by Case Type (available in Clerk of Court’s case file). Footnote 3 Rather than dismiss a petition containing both exhausted and unexhausted claims, “a district court might stay the petition and hold it in abeyance while the petitioner returns to state court to exhaust his previously unexhausted claims. Once the petitioner exhausts his state remedies, the district court will lift the stay and allow the petitioner to proceed in federal court.” Rhines v. Weber, 544 U. S. 269, 275–276 (2005). Footnote 4 We have also recognized a “limited category” of “exceptional cases in which exorbitant application of a generally sound rule renders the state ground inadequate to stop consideration of a federal question.” Lee v. Kemna, 534 U. S. 362, 376 (2002). In Lee, for example, the defendant unsuccessfully moved for a continuance when, for reasons unknown to him, his alibi witnesses left the courthouse the day they were scheduled to testify. This Court held inadequate to bar federal review a state court’s persnickety application of a rule detailing formal requirements for continuance motions. The defendant had substantially complied with the rule’s key requirement and flawless compliance would have been unavailing given the trial court’s reason for denying the motion. See id., at 381–382. Martin does not suggest that the application of California’s timeliness rule in his case falls within the exceptional category Lee described and illustrated. See Brief for Respondent 28, 29, 54. Footnote 5 An untimely petition “will be entertained on the merits if the petitioner demonstrates (i) that error of constitutional magnitude led to a trial that was so fundamentally unfair that absent the error no reasonable judge or jury would have convicted the petitioner; (ii) that the petitioner is actually innocent of the crime or crimes of which he or she was convicted; (iii) that the death penalty was imposed by a sentencing authority that had such a grossly misleading profile of the petitioner before it that, absent the trial error or omission, no reasonable judge or jury would have imposed a sentence of death; or (iv) that the petitioner was convicted or sentenced under an invalid statute.” In re Robbins, 18 Cal. 4th 770, 780–781, 959 P. 2d 311, 318 (1998). Footnote 6 See, e.g., In re Sanders, 21 Cal. 4th 697, 703, 981 P. 2d 1038, 1042 (1999); In re Hamilton, 20 Cal. 4th 273, 283, n. 5, 975 P. 2d 600, 605, n. 5 (1999); In re Watson, 104 Cal. Rptr. 3d 403, 407 (App. 2010) (officially depublished); In re Nunez, 173 Cal. App. 4th 709, 723, 93 Cal. Rptr. 3d 242, 252 (2009). Footnote 7 Closer inspection may reveal that “seeming ‘inconsistencie[s]’ … are not necessarily … arbitrar[y] or irrationa[l].” Thornburgh v. Abbott, 490 U. S. 401, 417, n. 15 (1989). Fairbanks and Little are illustrative. In Fairbanks, the court found that petitioner did not act diligently when she waited to withdraw her guilty plea until one year after learning that revocation of her driver’s license was irreversible. 2006 WL 950267, *2–*3. In Little, a pro se prisoner claimed that his trial counsel should have raised a posttraumatic stress disorder defense. Although the filing delay was 14 months, the court entertained it on the merits. 2008 WL 142832, *4, *14. Given the discrete context in which each case arose, the two decisions present no square conflict. Footnote 8 See also 16B Wright & Miller §4026, pp. 385–386 (“Precisely defined rules cannot take account of the gravity of a procedural failure, the strength of the excuses offered, or the importance of the procedural and substantive consequences of excusing or punishing the failure.”).

562.US.2010_09-868

Respondent was convicted in Rhode Island Superior Court on 10 counts of first-degree sexual assault and sentenced to consecutive life terms. His conviction became final on direct review on May 29, 1996. In addition to his direct appeal, he filed two relevant state motions. One, a May 16, 1996, motion to reduce his sentence under Rhode Island Superior Court Rule of Criminal Procedure 35, was denied. The State Supreme Court affirmed on January 16, 1998. The second, a state postconviction relief motion, was also denied. That decision was affirmed on December 14, 2006. When respondent filed his federal habeas petition, his conviction had been final for over 11 years. The Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) generally requires a federal petition to be filed within one year of the date on which a judgment became final, 28 U. S. C. §2244(d)(1)(A), but “a properly filed application for State post-conviction or other collateral review” tolls that period, §2244(d)(2). Respondent’s postconviction relief motion tolled the period for over nine years, but his Rule 35 motion must also trigger the tolling provision for his habeas petition to be timely. The District Court dismissed the petition as untimely, adopting the Magistrate Judge’s conclusion that the Rule 35 motion was not “a properly filed application for … collateral review” under §2244(d)(2). The First Circuit reversed. Held: 1. The phrase “collateral review” in §2244(d)(2) means judicial review of a judgment in a proceeding that is not part of direct review. Pp. 4–8. (a) The parties agree that the answer to the question whether a motion to reduce sentence is an “application for State post-conviction or other collateral review” turns on the meaning of “collateral review,” but they disagree about what that meaning should be. Pp. 4–5. (b) Because “collateral review” is not defined in AEDPA, the Court begins with the ordinary understanding of that phrase. By definition, “collateral” describes something that is “indirect,” not direct. 3 Oxford English Dictionary 473. This suggests that “collateral” review is not part of direct review. This conclusion is supported by the definition of the related phrase “collateral attack” and by the Court’s prior use of the term “collateral” to describe proceedings that are separate from the direct review process. Pp. 5–7. (c) The term “review” is best understood as a “judicial reexamination.” Webster’s Third New International Dictionary 1944. Pp. 7–8. 2. A Rule 35 motion to reduce sentence under Rhode Island law is an application for “collateral review” that triggers AEDPA’s tolling provision. Pp. 8–15. (a) Rhode Island’s Rule 35 is similar to the version of Federal Rule of Criminal Procedure 35 in effect before the federal Sentencing Reform Act of 1984. The Rule permits a court to provide relief, as relevant here, to “reduce any sentence,” and it is generally addressed to the sound discretion of the sentencing justice. Under the limited review available, an appellate court may disturb the trial justice’s decision if the sentence imposed is without justification and is grossly disparate when compared to sentences for similar offenses. Pp. 8–9. (b) Keeping these principles in mind, a Rule 35 sentence reduction proceeding is “collateral.” The parties agree that the motion is not part of the direct review process, and both this Court and lower federal courts have described a motion to reduce sentence under old Federal Rule 35 as invoking a “collateral” remedy. Therefore, it is not difficult to conclude that Rhode Island’s motion to reduce sentence is “collateral.” A Rule 35 motion also calls for “review” of the sentence within §2244(d)(2)’s meaning. The decision to reduce a sentence involves judicial reexamination of the sentence to determine whether a more lenient sentence is proper. The trial justice is guided by several sentencing factors in making that decision. And those factors are also used by the State Supreme Court in evaluating the trial justice’s justifications for the sentence. Pp. 9–11. (c) Rhode Island’s arguments in support of its opposing view that “collateral review” includes only “legal” challenges to a conviction or sentence, and thus excludes motions for a discretionary sentence reduction, are unpersuasive. Nor does “collateral review” turn on whether a motion is part of the same criminal case. Pp. 11–15. 582 F. 3d 147, affirmed. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined, and in which Scalia, J., joined, except as to footnote 3. Scalia, J., filed an opinion concurring in part.

Under the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), “a properly filed application for State post-conviction or other collateral review with respect to the pertinent judgment or claim” tolls the 1-year limitation period for filing a federal habeas petition. 28 U. S. C. §2244(d)(2). The question in this case is whether a motion to reduce sentence under Rhode Island law tolls the limitation period, thereby rendering respondent Khalil Kholi’s federal habeas petition timely. We hold that the phrase “collateral review” in §2244(d)(2) means judicial review of a judgment in a proceeding that is not part of direct review. Because the parties agree that a motion to reduce sentence under Rhode Island law is not part of the direct review process, we hold that respondent’s motion tolled the AEDPA limitation period and that his federal habeas petition was therefore timely. I A In 1993, respondent was convicted in Rhode Island Superior Court on 10 counts of first-degree sexual assault, and he was sentenced to consecutive terms of life im-prisonment. Respondent raised various challenges to his conviction on direct appeal, but the Supreme Court of Rhode Island affirmed his conviction. See State v. Kholi, 672 A. 2d 429, 431 (1996). The parties agree that respondent’s conviction became final on direct review when his time expired for filing a petition for a writ of certiorari in this Court. Brief for Petitioner 7, n. 4; Brief for Respondent 3, n. 1; 582 F. 3d 147, 150 (CA1 2009); see generally Jimenez v. Quarterman, 555 U. S. 113, ___ (2009) (slip op., at 6). That date was May 29, 1996. See this Court’s Rules 13.1, 13.3, 30.1. In addition to taking a direct appeal, respondent filed two state motions that are relevant to our decision. The first, filed on May 16, 1996, was a motion to reduce sentence under Rule 35 of the Rhode Island Superior Court Rules of Criminal Procedure.[Footnote 1] App. 8. In that motion, respondent asked the trial court to “reconsider its prior determination” and “order that his life sentences run concurrently.” State v. Kholi, 706 A. 2d 1326 (R. I. 1998) (order). Concluding that “the sentence imposed was appropriate,” the hearing justice denied the Rule 35 motion. Ibid. On January 16, 1998, the State Supreme Court affirmed and observed that the facts clearly justified the sentence. Id., at 1326–1327. On May 23, 1997, while the Rule 35 motion was pending, respondent also filed an application for state postconviction relief, see R. I. Gen. Laws 10–9.1–1 et seq. (Lexis 1997) (titled “Post Conviction Remedy”), which challenged his conviction. The trial court denied this motion as well, and the State Supreme Court affirmed that decision on December 14, 2006. See Kholi v. Wall, 911 A. 2d 262, 263–264 (R. I. 2006). B Respondent filed a federal habeas petition in the District of Rhode Island on September 5, 2007. App. 3. By that time, his conviction had been final for over 11 years. AEDPA generally requires a federal habeas petition to be filed within one year of the date on which the judgment became final by the conclusion of direct review. 28 U. S. C. §2244(d)(1)(A). But the 1-year limitation period is tolled during the pendency of “a properly filed application for State post-conviction or other collateral review with respect to the pertinent judgment or claim.” §2244(d)(2). There is no dispute that respondent’s application for postconviction relief tolled the limitation period for over nine years—from May 23, 1997, through December 14, 2006. 582 F. 3d, at 151. Even after subtracting that stretch of time from the 11-year period, however, the period between the conclusion of direct review and the filing of the federal habeas petition still exceeds one year. Thus, in order for respondent’s petition to be timely, the Rule 35 motion to reduce sentence must also trigger the tolling provision. Respondent’s federal habeas petition was referred to a Magistrate Judge for a report and recommendation, and the Magistrate Judge concluded that the Rule 35 motion was not a “ ‘properly filed application for post-conviction or other collateral review’ ” under §2244(d)(2) because it was “a ‘plea of leniency,’ and not a motion challenging the legal sufficiency of his sentence.” No. CA 07–346S, 2008 WL 60194, *4 (R. I., Jan. 3, 2008). The District Court adopted the Magistrate Judge’s report and recommendation and therefore dismissed the federal habeas petition as untimely. See id., at *1. On appeal, the First Circuit reversed. 582 F. 3d 147. The Courts of Appeals are divided over the question whether a motion to reduce sentence tolls the period of limitation under §2244(d)(2).[Footnote 2] We granted certiorari to answer this question with respect to a motion to reduce sentence under Rhode Island law. 560 U. S. ___ (2010). II A AEDPA establishes a 1-year period of limitation for a state prisoner to file a federal application for a writ of habeas corpus. §2244(d)(1). This period runs “from the latest of” four specified dates, including “the date on which the judgment became final by the conclusion of direct review or the expiration of the time for seeking such review.” §2244(d)(1)(A); see also Jimenez, supra, at ___ (slip op., at 6) (explaining when “the conclusion of direct review occurs”). The limitation period is tolled, however, during the pendency of “a properly filed application for State post-conviction or other collateral review with respect to the pertinent judgment or claim.” §2244(d)(2). The question in this case is whether a motion for reduction of sentence under Rhode Island’s Rule 35 is an “application for State post-conviction or other collateral review.” The parties agree that the answer to this question turns on the meaning of the phrase “collateral review,” see Brief for Petitioner 19; Brief for Respondent 12–13, but they disagree about the definition of that term. Rhode Island argues that “collateral review” includes only “legal” challenges to a conviction or sentence and thus excludes motions seeking a discretionary sentence reduction. Respondent, on the other hand, maintains that “collateral review” is “review other than review of a judgment in the direct appeal process” and thus includes motions to reduce sentence. Brief for Respondent 17. We agree with respondent’s understanding of “collateral review.” B “Collateral review” is not defined in AEDPA, and we have never provided a comprehensive definition of that term. See Duncan v. Walker, 533 U. S. 167, 175–178 (2001). We therefore begin by considering the ordinary understanding of the phrase “collateral review.” See Williams v. Taylor, 529 U. S. 420, 431 (2000) (“We give the words of a statute their ordinary, contemporary, common meaning, absent an indication Congress intended them to bear some different import” (internal quotation marks omitted)); see also Carey v. Saffold, 536 U. S. 214, 219 (2002) (considering the ordinary meaning of the word “pending” in §2244(d)(2)). The term “collateral,” in its “customary and preferred sense,” Williams, supra, at 431, means “[l]ying aside from the main subject, line of action, issue, purpose, etc.; … subordinate, indirect,” 3 Oxford English Dictionary 473 (2d ed. 1989) (hereinafter OED); see also Webster’s Third New International Dictionary 444 (1993) (hereinafter Webster’s) (“accompanying as … secondary,” “indirect,” or “ancillary”). By definition, something that is “collateral” is “indirect,” not direct. 3 OED 473. This suggests that “collateral” review is review that is “[l]ying aside from the main” review, i.e., that is not part of direct review. See ibid. The definition of the related phrase “collateral attack” points in the same direction. A “collateral attack” is “[a]n attack on a judgment in a proceeding other than a direct appeal.” Black’s Law Dictionary 298 (9th ed. 2009) (emphasis added); cf. Wash. Rev. Code §10.73.090(2) (2008) (defining “collateral attack” as “any form of postconviction relief other than a direct appeal”). This usage buttresses the conclusion that “collateral review” means a form of review that is not part of the direct appeal process. C Our prior usage of the term “collateral” also supports this understanding. We have previously described a variety of proceedings as “collateral,” and all of these proceedings share the characteristic that we have identified, i.e., they stand apart from the process of direct review. For example, our cases make it clear that habeas corpus is a form of collateral review. We have used the terms habeas corpus and “collateral review” interchangeably, see, e.g., Murray v. Carrier, 477 U. S. 478, 482–483 (1986), and it is well accepted that state petitions for habeas corpus toll the limitation period, e.g., Rhines v. Weber, 544 U. S. 269, 272 (2005) (“[T]he 1-year statute of limitations … was tolled while Rhines’ state habeas corpus petition was pending”). We have also described coram nobis as a means of “collateral attack,” see, e.g., United States v. Morgan, 346 U. S. 502, 510–511 (1954) (internal quotation marks omitted), and we have used the term “collateral” to describe proceedings under 28 U. S. C. §2255 and a prior version of Rule 35 of the Federal Rules of Criminal Procedure. In United States v. Robinson, 361 U. S. 220 (1960), we distinguished between the process of direct appeal and “a number of collateral remedies,” including Federal Rule 35 motions, §2255 motions, and coram nobis. Id., at 230, n. 14. Similarly, in Bartone v. United States, 375 U. S. 52 (1963) (per curiam), we drew a distinction between a “[d]irect attack” on a criminal judgment and “collateral proceedings,” such as Rule 35, habeas corpus, and §2255 proceedings. Id., at 53–54. All of the proceedings identified in these prior opinions as “collateral” are separate from the direct review process, and thus our prior usage of the term “collateral” buttresses the conclusion that “collateral review” means a form of review that is not direct. D Of course, to trigger the tolling provision, a “collateral” proceeding must also involve a form of “review,” but the meaning of that term seems clear. “Review” is best understood as an “act of inspecting or examining” or a “judicial reexamination.” Webster’s 1944; see also Black’s, supra, at 1434 (“[c]onsideration, inspection, or reexamination of a subject or thing”); 13 OED 831 (“[t]o submit (a decree, act, etc.) to examination or revision”). We thus agree with the First Circuit that “ ‘review’ commonly denotes ‘a looking over or examination with a view to amendment or improvement.’ ” 582 F. 3d, at 153 (quoting Webster’s 1944 (2002)). Viewed as a whole, then, “collateral review” of a judgment or claim means a judicial reexamination of a judgment or claim in a proceeding outside of the direct review process. III We now apply this definition of “collateral review” to a Rule 35 motion to reduce sentence under Rhode Island law. A Rule 35 of the Rhode Island Rules of Criminal Procedure is much like the version of Federal Rule of Criminal Procedure 35 that was in force prior to the enactment of the federal Sentencing Reform Act of 1984 and the promulgation of the Federal Sentencing Guidelines. See State v. Byrnes, 456 A. 2d 742, 744 (R. I. 1983) (per curiam); Reporter’s Notes following R. I. Super. Ct. Rule Crim. Proc. 35, R. I. Court Rules Ann., p. 620 (Lexis 2010). Under the Rhode Island Rules, a Rule 35 motion permits a court to provide relief from a sentence in three ways: A court “may” “correct an illegal sentence,” “correct a sentence imposed in an illegal manner,” and “reduce any sentence.” R. I. Super. Ct. Rule Crim. Proc. 35(a); see n. 1, supra. In this case, respondent filed a motion to reduce his sentence, which permits a trial justice to decide “ ‘ “on reflection or on the basis of changed circumstances that the sentence originally imposed was, for any reason, unduly severe.” ’ ” State v. Ruffner, 5 A. 3d 864, 867 (R. I. 2010) (quoting State v. Mendoza, 958 A. 2d 1159, 1161 (R. I. 2008)); see also Reporter’s Notes following R. I. Super. Ct. Rule Crim. Proc. 35, R. I. Court Rules Ann., at 620–621. Rhode Island courts have, at times, referred to such a motion as a “ ‘plea for leniency.’ ” Ruffner, supra, at 867 (quoting Mendoza, supra, at 1161). A Rule 35 motion is made in the Superior Court, and it is generally heard by the same trial justice who sentenced the defendant. Byrnes, supra, at 745. The Rhode Island Supreme Court has explained that a motion to reduce sentence is “ ‘addressed to the sound discretion of the trial justice’ ” and that appellate review of the trial justice’s decision is limited. Ruffner, supra, at 867 (quoting Mendoza, supra, at 1161). An appellate court may nevertheless disturb the trial justice’s decision “when the trial justice has imposed a sentence that is without justification and is grossly disparate from other sentences generally imposed for similar offenses.” Ruffner, supra, at 867 (quoting State v. Coleman, 984 A. 2d 650, 654 (R. I. 2009); internal quotation marks omitted); see also Ruffner, supra, at 867 (asking whether trial justice “abuse[d] his discretion”). B With these principles in mind, we consider whether Rhode Island’s Rule 35 motion to reduce sentence is an application for “collateral review.” The first—and the critical—question is whether a Rhode Island Rule 35 sentence reduction proceeding is “collateral.” Respondent and Rhode Island agree that such a motion is not part of the direct review process. Moreover, we have previously referred to a motion to reduce sentence under old Rule 35 of the Federal Rules of Criminal Procedure as invoking a “collateral” remedy, see Robinson, supra, at 230, n. 14, and Rhode Island’s Rule 35 motion to reduce sentence is “substantially similar” to former Federal Rule 35, Byrnes, supra, at 744. Lower courts have also referred to Federal Rule 35 sentence reduction motions as “collateral.” See, e.g., Fernandez v. United States, 941 F. 2d 1488, 1492 (CA11 1991) (“Fernandez initiated a collateral attack on his sentence with a Rule 35(b) motion to reduce his sentence” under the old Federal Rule). We thus have little difficulty concluding that a Rhode Island sentence reduction proceeding is “collateral.”[Footnote 3] Not only is a motion to reduce sentence under Rhode Island law “collateral,” but it also undoubtedly calls for “review” of the sentence. The decision to reduce a sentence, while largely within the discretion of the trial justice, involves judicial reexamination of the sentence to determine whether a more lenient sentence is proper.[Footnote 4] When ruling on such a motion, a trial justice is guided by several factors, including “(1) the severity of the crime, (2) the defendant’s personal, educational, and employment background, (3) the potential for rehabilitation, (4) the element of societal deterrence, and (5) the appropriateness of the punishment.” State v. Mollicone, 746 A. 2d 135, 138 (R. I. 2000) (per curiam) (internal quotation marks omitted); see also Ruffner, supra, at 867; Coleman, supra, at 655. On appeal from a trial justice’s decision on a motion to reduce sentence, the Supreme Court of Rhode Island evaluates the trial justice’s justifications in light of the relevant sentencing factors to determine whether a sentence is “without justification” and “grossly disparate from other sentences.” Ruffner, supra, at 867 (internal quotation marks omitted).[Footnote 5] This process surely qualifies as “review” of a sentence within the meaning of §2244(d)(2). We thus hold that a motion to reduce sentence under Rhode Island law is an application for “collateral review” that triggers AEDPA’s tolling provision. IV In resisting this interpretation, Rhode Island advances several arguments that we find unpersuasive. The first of these arguments begins by observing that, whenever our opinions have used the precise phrase “collateral review,” the proceeding in question was one challenging the “lawfulness” of a prior judgment, Brief for Petitioner 21–22, such as a §2254 or §2255 action, see id., at 25. Rhode Island argues that Congress, in enacting AEDPA, must be presumed to have been aware of this usage and must have intended the phrase to carry this narrow meaning. This argument reads far too much into these prior references to “collateral review.” While our opinions have used the phrase “collateral review” to refer to proceedings that challenge the lawfulness of a prior judgment, we have never suggested that the phrase may properly be used to describe only proceedings of this type. In addition, Rhode Island overlooks opinions describing a motion to reduce sentence as “collateral.” E.g., Robinson, 361 U. S., at 230, n. 14; Fernandez, supra, at 1492; see also 1 D. Wilkes, State Postconviction Remedies and Relief Handbook §§1:2, 1:7, pp. 2, 15 (2010) (characterizing a motion to reduce sentence as a “collateral” or “postconviction” remedy). In a related argument, Rhode Island notes that several other AEDPA provisions use the term “collateral review” to refer to proceedings that involve a challenge to the lawfulness of a state-court judgment, see 28 U. S. C. §§2244(b)(2)(A), (d)(1)(C), 2254(e)(2)(A)(i),[Footnote 6] and Rhode Island reasons that the phrase “collateral review” in §2244(d)(2) should be limited to proceedings of this nature. This argument has the same flaw as the argument just discussed. Just because the phrase “collateral review” encompasses proceedings that challenge the lawfulness of a prior judgment, it does not follow that other proceedings may not also be described as involving “collateral review.” Finally, Rhode Island contends that the purpose of the tolling provision is to allow a state prisoner to exhaust state remedies and that this purpose is not served when a prisoner’s state application merely seeks sentencing leniency, a matter that cannot be raised in a federal habeas petition. This argument is based on an excessively narrow understanding of §2244(d)(2)’s role. It is certainly true that a purpose—and perhaps the chief purpose—of tolling under §2244(d)(2) is to permit the exhaustion of state remedies, see Duncan, 533 U. S., at 178–179, but that is not §2244(d)(2)’s only role. The tolling provision “provides a powerful incentive for litigants to exhaust all available state remedies before proceeding in the lower federal courts.” Id., at 180 (emphasis added). Tolling the limitation period for all “collateral review” motions provides both litigants and States with an opportunity to resolve objections at the state level, potentially obviating the need for a litigant to resort to federal court. If, for example, a litigant obtains relief on state-law grounds, there may be no need for federal habeas. The same dynamic may be present to a degree with respect to motions that do not challenge the lawfulness of a judgment. If a defendant receives relief in state court, the need for federal habeas review may be narrowed or even obviated, and this furthers principles of “comity, finality, and federalism.” Williams, 529 U. S., at 436. Rhode Island’s interpretation of §2244(d)(2) would also greatly complicate the work of federal habeas courts. Rhode Island would require those courts to separate motions for a reduced sentence into two categories: those that challenge a sentence on legal grounds and those that merely ask for leniency. But this taxonomy is problematic. Even if a jurisdiction allows sentencing judges to exercise a high degree of discretion in selecting a sentence from within a prescribed range, it does not necessarily follow that the judge’s choice is insulated from challenge on legal grounds. “[D]iscretionary choices are not left to a court’s ‘inclination, but to its judgment; and its judgment is to be guided by sound legal principles.’ ” Albemarle Paper Co. v. Moody, 422 U. S. 405, 416 (1975) (quoting United States v. Burr, 25 F. Cas. 30, 35 (No. 14,692d) (CC Va. 1807) (Marshall, C. J.)). If the law of a jurisdiction provides criteria to guide a trial judge’s exercise of sentencing discretion, a motion to reduce sentence may argue that a sentence is inconsistent with those criteria. In that sense, the motion argues that the sentence is contrary to sentencing law. See, e.g., Ruffner, 5 A. 3d, at 867 (“A trial justice considers a number of factors when determining a fair sentence[,] including the defendant’s potential for rehabilitation. The defendant asserts that the trial justice did not consider defendant’s participation in rehabilitative programs” (citations omitted)). We do not think that §2244(d)(2) was meant to require federal habeas courts to draw the sort of difficult distinction that Rhode Island’s interpretation would demand. We also reject the argument that the meaning of the phrase “collateral review” should turn on whether the motion or application that triggers that review is captioned as a part of the criminal case or as a separate proceeding. See Walkowiak v. Haines, 272 F. 3d 234, 237 (CA4 2001). This interpretation of §2244(d)(2) would produce confusion and inconsistency. For one thing, some “collateral” proceedings are often regarded as part of the criminal case. We have said, for example, that a writ of coram nobis “is a step in the criminal case and not … a separate case and record, the beginning of a separate civil proceeding.” Morgan, 346 U. S., at 505, n. 4; see also United States v. Denedo, 556 U. S. ___, ___ (2009) (slip op., at 8) (“[A]n application for the writ is properly viewed as a belated extension of the original proceeding during which the error allegedly transpired”). But we have nonetheless suggested that coram nobis is a means of “collateral attack.” Morgan, supra, at 510–511 (internal quotation marks omitted); see also Robinson, 361 U. S., at 230, n. 14. Similarly, a motion under 28 U. S. C. §2255 (2006 ed., Supp. III) is entered on the docket of the original criminal case and is typically referred to the judge who originally presided over the challenged proceedings, see §2255 Rules 3(b), 4(a), but there is no dispute that §2255 proceedings are “collateral,” see, e.g., Massaro v. United States, 538 U. S. 500, 504 (2003) (describing §2255 proceedings as “collateral”); Daniels v. United States, 532 U. S. 374, 379 (2001) (same).[Footnote 7] Moreover, the methods of filing for postconviction or collateral review vary among the States. In the District of Columbia and fourteen States, the principal postconviction remedy is part of the original case; in other States, it is not. 1 Wilkes, State Postconviction Remedies and Relief Handbook §1:3, at 6–7. Given the States’ “different forms of collateral review,” Duncan, 533 U. S., at 177, the ap-plication of AEDPA’s tolling provision should not turn on such formalities. See ibid. (“Congress may have refrained from exclusive reliance on the term ‘post-conviction’ so as to leave no doubt that the tolling provision applies to all types of state collateral review available after a conviction”). We thus define “collateral review” according to its ordinary meaning: It refers to judicial review that occurs in a proceeding outside of the direct review process. * * * For these reasons, the judgment of the Court of Appeals is affirmed. It is so ordered. Footnote 1 This Rule provides in relevant part: “The court may correct an illegal sentence at any time. The court may correct a sentence imposed in an illegal manner and it may reduce any sentence when a motion is filed within one hundred and twenty (120) days after the sentence is imposed, or within one hundred and twenty (120) days after receipt by the court of a mandate of the Supreme Court of Rhode Island issued upon affirmance of the judgment or dismissal of the appeal, or within one hundred and twenty (120) days after receipt by the court of a mandate or order of the Supreme Court of the United States issued upon affirmance of the judgment, dismissal of the appeal, or denial of a writ of certiorari. The court shall act on the motion within a reasonable time, provided that any delay by the court in ruling on the motion shall not prejudice the movant. The court may reduce a sentence, the execution of which has been suspended, upon revocation of probation.” R. I. Super. Ct. Rule Crim. Proc. 35(a) (2010) (emphasis added). Footnote 2 Compare Alexander v. Secretary, Dept. of Corrections, 523 F. 3d 1291, 1297 (CA11 2008) (motion to reduce sentence does not toll limitation period); Hartmann v. Carroll, 492 F. 3d 478, 484 (CA3 2007) (same); Walkowiak v. Haines, 272 F. 3d 234, 239 (CA4 2001) (same), with 582 F. 3d, at 156 (case below) (motion to reduce sentence tolls); Robinson v. Golder, 443 F. 3d 718, 720–721 (CA10 2006) (per curiam) (same). Footnote 3 We can imagine an argument that a Rhode Island Rule 35 proceeding is in fact part of direct review under §2244(d)(1) because, according to the parties, defendants in Rhode Island cannot raise any challenge to their sentences on direct appeal; instead, they must bring a Rule 35 motion. See, e.g., State v. Day, 925 A. 2d 962, 985 (R. I. 2007) (“It is well settled in this jurisdiction that a challenge to a criminal sentence must begin with the filing of a [Rule 35] motion … . [W]e will not consider the validity or legality of a sentence on direct appeal unless extraordinary circumstances exist” (internal quotation marks omitted)); State v. McManus, 990 A. 2d 1229, 1238 (R. I. 2010) (refusing to consider Eighth Amendment challenge on direct review because “[t]o challenge a criminal sentence, the defendant must first file a motion to reduce in accordance with Rule 35”); see also Jimenez v. Quarterman, 555 U. S. 113, ___ (2009) (slip op., at 6–7). That issue has not been briefed or argued by the parties, however, and we express no opinion as to the merit of such an argument. Even if we were to assume that a Rhode Island Rule 35 motion is part of direct review, our disposition of this case would not change: Respondent’s habeas petition still would be timely, because the limitation period would not have begun to run until after the Rule 35 proceedings concluded. Footnote 4 A motion to reduce sentence is unlike a motion for postconviction discovery or a motion for appointment of counsel, which generally are not direct requests for judicial review of a judgment and do not provide a state court with authority to order relief from a judgment. Footnote 5 E.g., State v. Coleman, 984 A. 2d 650, 657 (R. I. 2009) (“Given these factors, and the trial justice’s exhaustive explanation of her reasoning in sentencing Mr. Coleman, we hold it was not an abuse of her discretion to order Mr. Coleman to serve consecutive sentences”); State v. Ferrara, 818 A. 2d 642, 645 (R. I. 2003) (per curiam) (“[M]itigating circumstances clearly are not present in this case”); State v. Rossi, 771 A. 2d 906, 908 (R. I. 2001) (order) (“Based upon [the court’s] review of the record,” the sentence “was not excessive and was justified under the circumstances,” namely, “the abhorrent conduct of [the] defendant” and “the permissible penalty range” under the statute); State v. Mollicone, 746 A. 2d 135, 138 (R. I. 2000) (per curiam) (“[T]he trial justice was aware of these factors and applied them correctly”). Footnote 6 All of these provisions refer to a new rule of constitutional law made retroactively applicable by this Court to “cases on collateral review.” Footnote 7 In other contexts not relevant here, there has been some confusion over whether §2255 proceedings are civil or criminal in nature. See, e.g., Postconviction Remedies §3:5, p. 251 (2010) (“[T]here is a dispute over whether the [§2255] motion initiates an independent civil action or, instead, is merely a further step in the criminal prosecution”); 3 C. Wright & S. Welling, Federal Practice and Procedure §622 (4th ed. 2011). We express no opinion on this question.

562.US.2010_08-1314

The 1989 version of Federal Motor Vehicle Safety Standard 208 (FMVSS 208) requires, as relevant here, auto manufacturers to install seatbelts on the rear seats of passenger vehicles. They must install lap-and-shoulder belts on seats next to a vehicle’s doors or frames, but may install either those belts or simple lap belts on rear inner seats, e.g., those next to a minivan’s aisle. The Williamson family and Thanh Williamson’s estate brought this California tort suit, claiming that Thanh died in an accident because the rear aisle seat of the Mazda minivan in which she was riding had a lap belt instead of lap-and-shoulder belts. The state trial court dismissed their claim on the pleadings. The State Court of Appeal affirmed, relying on Geier v. American Honda Motor Co., 529 U. S. 861, in which this Court found that an earlier (1984) version of FMVSS 208—which required installation of passive restraint devices—pre-empted a state tort suit against an auto manufacturer on a failure to install airbags. Held: FMVSS 208 does not pre-empt state tort suits claiming that manufacturers should have installed lap-and-shoulder belts, instead of lap belts, on rear inner seats. Pp. 3–12. (a) Because this case involves (1) the same statute as Geier, (2) a later version of the same regulation, and (3) a somewhat similar claim that a state tort action conflicts with the federal regulation, the answers to two of the subsidiary questions posed in Geier apply directly here. Thus, the statute’s express pre-emption clause cannot pre-empt the common-law tort action here; but neither can its saving clause foreclose or limit the operation of ordinary conflict pre-emption principles. The Court consequently turns to Geier’s third subsidiary question, whether, in fact, the state tort action conflicts with the federal regulation. Pp. 3–5. (b) Under ordinary conflict pre-emption principles a state law that “stands as an obstacle to the accomplishment” of a federal law is pre-empted. Hines v. Davidowitz, 312 U. S. 52, 67. In Geier, the state law stood as an obstacle to the accomplishment of a significant federal regulatory objective, namely, giving manufacturers a choice among different kinds of passive restraint systems. This conclusion was supported by the regulation’s history, the agency’s contemporaneous explanation, and the Government’s current understanding of the regulation. The history showed that the Department of Transportation (DOT) had long thought it important to leave manufacturers with a choice of systems. DOT’s contemporaneous explanation of the regulation made clear that manufacturer choice was an important means for achieving DOT’s basic objectives. It phased in passive restraint requirements to give manufacturers time to improve airbag technology and develop better systems; it worried that requiring airbags would cause a public backlash; and it was concerned about airbag safety and cost. Finally, the Government’s current understanding was that a tort suit insisting upon airbag use would “ ‘ “stan[d] as an obstacle to the accomplishment and execution of these objectives.” ’ ” 529 U. S.em>., at 883. Pp. 5–8. (c) Like the regulation in Geier, the instant regulation leaves the manufacturer with a choice, and the tort suit here would restrict that choice. But in contrast to Geier, the choice here is not a significant regulatory objective. The regulation’s history resembles the history of airbags to some degree. DOT rejected a regulation requiring lap-and-shoulder belts in rear seats in 1984. But by 1989, changed circumstances led DOT to require manufacturers to install lap-and-shoulder belts for rear outer seats but to retain a manufacturer choice for rear inner seats. Its reasons for doing so differed considerably from its 1984 reasons for permitting a choice of passive restraint. It was not concerned about consumer acceptance; it thought that lap-and-shoulder belts would increase safety and did not pose additional safety risks; and it was not seeking to use the regulation to spur development of alternative safety devices. Instead, DOT thought that the requirement would not be cost effective. That fact alone cannot show that DOT sought to forbid common-law tort suits. For one thing, DOT did not believe that costs would remain frozen. For another, many federal safety regulations embody a cost-effectiveness judgment. To infer pre-emptive intent from the mere existence of such a cost-effectiveness judgment would eliminate the possibility that the agency seeks only to set forth a minimum standard. Finally, the Solicitor General represents that DOT’s regulation does not pre-empt this tort suit. As in Geier, “the agency’s own views should make a difference,” 529 U. S., at 883, and DOT has not expressed inconsistent views on this subject. Pp. 8–12. 167 Cal. App. 4th 905, 84 Cal. Rptr. 3d 545, reversed. Breyer, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Ginsburg, Alito, and Sotomayor, JJ., joined. Sotomayor, J., filed a concurring opinion. Thomas, J., filed an opinion concurring in the judgment. Kagan, J., took no part in the consideration or decision of the case.

Federal Motor Vehicle Safety Standard 208 (1989 version) requires, among other things, that auto manufacturers install seatbelts on the rear seats of passenger vehicles. They must install lap-and-shoulder belts on seats next to a vehicle’s doors or frames. But they have a choice about what to install on rear inner seats (say, middle seats or those next to a minivan’s aisle). There they can install either (1) simple lap belts or (2) lap-and-shoulder belts. 54 Fed. Reg. 46257–46258 (1989); 49 CFR §571.208 (1993), promulgated pursuant to the National Traffic and Motor Vehicle Safety Act of 1966 (Act), 80 Stat. 718, 15 U. S. C. §1381 et seq. (1988 ed.) (recodified without substantive change at 49 U. S. C. §30101 et seq. (2006 ed.)). The question presented here is whether this federal regulation pre-empts a state tort suit that, if successful, would deny manufacturers a choice of belts for rear inner seats by imposing tort liability upon those who choose to install a simple lap belt. We conclude that providing manufacturers with this seatbelt choice is not a significant objective of the federal regulation. Consequently, the regulation does not pre-empt the state tort suit. I In 2002, the Williamson family, riding in their 1993 Mazda minivan, was struck head on by another vehicle. Thanh Williamson was sitting in a rear aisle seat, wearing a lap belt; she died in the accident. Delbert and Alexa Williamson were wearing lap-and-shoulder belts; they survived. They, along with Thanh’s estate, subsequently brought this California tort suit against Mazda. They claimed that Mazda should have installed lap-and-shoulder belts on rear aisle seats, and that Thanh died because Mazda equipped her seat with a lap belt instead. The California trial court dismissed this tort claim on the basis of the pleadings. And the California Court of Appeal affirmed. The appeals court noted that in Geier v. American Honda Motor Co., 529 U. S. 861 (2000), this Court considered whether a different portion of (an older version of) Federal Motor Vehicle Safety Standard 208 (FMVSS 208)—a portion that required installation of passive restraint devices—pre-empted a state tort suit that sought to hold an auto manufacturer liable for failure to install a particular kind of passive restraint, namely, airbags. We found that the federal regulation intended to assure manufacturers that they would retain a choice of installing any of several different passive restraint devices. And the regulation sought to assure them that they would not have to exercise this choice in favor of airbags. For that reason we thought that the federal regulation pre-empted a state tort suit that, by premising tort liability on a failure to install airbags, would have deprived the manufacturers of the choice that the federal regulation had assured them. Id., at 874–875. The court saw considerable similarity between this case and Geier. The federal regulation at issue here gives manufacturers a choice among two different kinds of seatbelts for rear inner seats. And a state lawsuit that premises tort liability on a failure to install a particular kind of seatbelt, namely, lap-and-shoulder belts, would in effect deprive the manufacturer of that choice. The court concluded that, as in Geier, the federal regulation pre-empts the state tort suit. 167 Cal. App. 4th 905, 84 Cal. Rptr. 3d 545 (2008). The Williamsons sought certiorari. And we granted certiorari in light of the fact that several courts have interpreted Geier as indicating that FMVSS 208 pre-empts state tort suits claiming that manufacturers should have installed lap-and-shoulder belts, not lap belts, on rear inner seats. Carden v. General Motors Corp., 509 F. 3d 227 (CA5 2007); Roland v. General Motors Corp., 881 N. E. 2d 722 (Ind. App. 2008); Heinricher v. Volvo Car Corp., 61 Mass. App. 313, 809 N. E. 2d 1094 (2004). II In Geier, we considered a portion of an earlier (1984) version of FMVSS 208. That regulation required manufacturers to equip their vehicles with passive restraint systems, thereby providing occupants with automatic accident protection. 49 Fed. Reg. 28983 (1984). But that regulation also gave manufacturers a choice among several different passive restraint systems, including airbags and automatic seatbelts. Id., at 28996. The question before the Court was whether the Act, together with the regulation, pre-empted a state tort suit that would have held a manufacturer liable for not installing airbags. 529 U. S., at 865. By requiring manufacturers to install airbags (in order to avoid tort liability) the tort suit would have deprived the manufacturers of the choice among passive restraint systems that the federal regulation gave them. See Hillsborough County v. Automated Medical Laboratories, Inc., 471 U. S. 707, 713 (1985) (“[S]tate laws can be pre-empted by federal regulations as well as by federal statutes”). We divided this basic pre-emption question into three subsidiary questions. 529 U. S., at 867. First, we asked whether the statute’s express pre-emption provision preempted the state tort suit. That statutory clause says that “no State” may “establish, or . . . continue in effect . . . any safety standard applicable to the same aspect of performance” of a motor vehicle or item of equipment “which is not identical to the Federal standard.” 15 U. S. C. §1392(d) (1988 ed.) (emphasis added). We had previously held that a word somewhat similar to “standard,” namely, “requirements” (found in a similar statute) included within its scope state “common-law duties,” such as duties created by state tort law. Medtronic, Inc. v. Lohr, 518 U. S. 470, 502–503 (1996) (plurality opinion); id., at 503–505 (Breyer, J., concurring in part and concurring in judgment); id., at 509–512 (O’Connor, J., concurring in part and dissenting in part). But we nonetheless held that the state tort suit in question fell outside the scope of this particular pre-emption clause. That is primarily because the statute also contains a saving clause, which says that “[c]ompliance with” a federal safety standard “does not exempt any person from any liability under common law.” 15 U. S. C. §1397(k) (emphasis added). Since tort law is ordinarily “common law,” we held that “the presence of the saving clause,” makes clear that Congress intended state tort suits to fall outside the scope of the express pre-emption clause. Geier, 529 U. S., at 868. Second, we asked the converse question: The saving clause at least removes tort actions from the scope of the express pre-emption clause. Id., at 869. But does it do more? Does it foreclose or limit “the operation of ordinary pre-emption principles insofar as those principles instruct us to read” federal statutes as pre-empting state laws (including state common-law standards) that “actually conflict” with the federal statutes (or related regulations)? Ibid. (internal quotation marks omitted). We concluded that the saving clause does not foreclose or limit the operation of “ordinary pre-emption principles, grounded in longstanding precedent.” Id., at 874. These two holdings apply directly to the case before us. We here consider (1) the same statute, 15 U. S. C. §1381 et seq.; (2) a later version of the same regulation, FMVSS 208; and (3) a somewhat similar claim that a state tort action conflicts with the federal regulation. In light of Geier, the statute’s express pre-emption clause cannot pre-empt the common-law tort action; but neither can the statute’s saving clause foreclose or limit the operation of ordinary conflict pre-emption principles. We consequently turn our attention to Geier’s third subsidiary question, whether, in fact, the state tort action conflicts with the federal regulation. III Under ordinary conflict pre-emption principles a state law that “stands as an obstacle to the accomplishment and execution of the full purposes and objectives” of a federal law is pre-empted. Hines v. Davidowitz, 312 U. S. 52, 67 (1941). See ibid. (federal statute can pre-empt a state statute); Cipollone v. Liggett Group, Inc., 505 U. S. 504 (1992) (federal statute can pre-empt a state tort suit); Fidelity Fed. Sav. & Loan Assn. v. De la Cuesta, 458 U. S. 141 (1982) (federal regulation can pre-empt a state statute); Geier, supra (federal regulation can pre-empt a state tort suit). In Geier we found that the state law stood as an “ ‘obstacle’ to the accomplishment” of a significant federal regulatory objective, namely, the maintenance of manufacturer choice. 529 U. S., at 886. We must decide whether the same is true here. A At the heart of Geier lies our determination that giving auto manufacturers a choice among different kinds of passive restraint devices was a significant objective of the federal regulation. We reached this conclusion on the basis of our examination of the regulation, including its history, the promulgating agency’s contemporaneous explanation of its objectives, and the agency’s current views of the regulation’s pre-emptive effect. The history showed that the Department of Transportation (DOT) had long thought it important to leave manufacturers with a choice. In 1967 DOT required manufacturers to install manual seat belts. Geier, supra, at 875; 32 Fed. Reg. 2408, 2415 (1967). Because many car occupants did not “buckle up,” DOT began to require passive protection, such as airbags or automatic seatbelts, but without “favor[ing] or “expect[ing]” the use of airbags. Geier, supra, at 875 (internal quotation marks omitted); 35 Fed. Reg. 16927 (1970). DOT subsequently approved the use of ignition interlocks, which froze the ignition until the occupant buckled the belt, as a substitute for passive restraints. Geier, supra, at 876; 37 Fed. Reg. 3911 (1972). But the interlock devices were unpopular with the public, and Congress soon forbade the agency to make them a means of compliance. Geier, supra, at 876; Motor Vehicle and Schoolbus Safety Amendments of 1974, §109, 88 Stat. 1482 (previously codified at 15 U. S. C. §1410(b) (1988 ed.)). DOT then temporarily switched to the use of demonstration projects, but later it returned to mandating passive restraints, again leaving manufacturers with a choice of systems. Geier, supra, at 876–877; see 49 Fed. Reg. 28962 (1984). DOT’s contemporaneous explanation of its 1984 regulation made clear that manufacturer choice was an important means for achieving its basic objectives. The 1984 regulation gradually phased in passive restraint requirements, initially requiring manufacturers to equip only 10% of their new fleets with passive restraints. DOT explained that it intended its phasing period partly to give manufacturers time to improve airbag technology and to develop “other, better” passive restraint systems. Geier, 529 U. S., at 879. DOT further explained that it had rejected an “ ‘all airbag’ ” system. Ibid. It was worried that requiring airbags in most or all vehicles would cause a public backlash, like the backlash against interlock devices. Ibid. DOT also had concerns about the safety of airbags, for they could injure out-of-place occupants, particularly children. Id., at 877–878. And, given the cost of airbags, vehicle owners might not replace them when necessary, leaving occupants without passive protection. Ibid. The regulation therefore “deliberately sought variety—a mix of several different passive restraint systems.” Id., at 878. DOT hoped that this mix would lead to better information about the devices’ comparative effectiveness and to the eventual development of “alternative, cheaper, and safer passive restraint systems.” Id., at 879. Finally, the Solicitor General told us that a tort suit that insisted upon use of airbags, as opposed to other federally permissible passive restraint systems, would “stan[d] as an obstacle to the accomplishment and execution of these objectives.” Id., at 883 (quoting Brief for United States as Amicus Curiae in Geier v. American Honda Motor Co., O. T. 1999, No. 98–1811, pp. 25–26 (hereinafter United States Brief in Geier) (internal quotation marks omitted)). And we gave weight to the Solicitor General’s view in light of the fact that it “ ‘embodie[d] the Secretary’s policy judgment that safety would best be promoted if manufacturers installed alternative protection systems in their fleets rather than one particular system in every car.’ ” 529 U. S., at 881 (quoting United States Brief in Geier 25–26). Taken together, this history, the agency’s contemporaneous explanation, and the Government’s current understanding of the regulation convinced us that manufacturer choice was an important regulatory objective. And since the tort suit stood as an obstacle to the accomplishment of that objective, we found the tort suit pre-empted. B We turn now to the present case. Like the regulation in Geier, the regulation here leaves the manufacturer with a choice. And, like the tort suit in Geier, the tort suit here would restrict that choice. But unlike Geier, we do not believe here that choice is a significant regulatory objective. We concede that the history of the regulation before us resembles the history of airbags to some degree. In 1984, DOT rejected a regulation that would have required the use of lap-and-shoulder belts in rear seats. 49 Fed. Reg. 15241. Nonetheless, by 1989 when DOT promulgated the present regulation, it had “concluded that several factors had changed.” 54 Fed. Reg. 46258. DOT then required manufacturers to install a particular kind of belt, namely, lap-and-shoulder belts, for rear outer seats. In respect to rear inner seats, it retained manufacturer choice as to which kind of belt to install. But its 1989 reasons for retaining that choice differed considerably from its 1984 reasons for permitting manufacturers a choice in respect to airbags. DOT here was not concerned about consumer acceptance; it was convinced that lap-and-shoulder belts would increase safety; it did not fear additional safety risks arising from use of those belts; it had no interest in assuring a mix of devices; and, though it was concerned about additional costs, that concern was diminishing. In respect to consumer acceptance, DOT wrote that if “people who are familiar with and in the habit of wearing lap/shoulder belts in the front seat find lap/shoulder belts in the rear seat, it stands to reason that they would be more likely to wear those belts when riding in the rear seat.” 53 Fed. Reg. 47983 (1988). In respect to safety, DOT wrote that, because an increasing number of rear seat passengers wore seatbelts, rear seat lap-and-shoulder belts would have “progressively greater actual safety benefits.” 54 Fed. Reg. 46257. It added: “[s]tudies of occupant protection from 1968 forward show that the lap-only safety belts installed in rear seating positions are effective in reducing the risk of death and injury. . . . However, the agency believes that rear-seat lap/shoulder safety belts would be even more effective.” Ibid. Five years earlier, DOT had expressed concern that lap-and-shoulder belts might negatively impact child safety by interfering with the use of certain child car seats that relied upon a tether. But by 1989, DOT found that car-seat designs “had shifted away” from tethers. 53 Fed. Reg. 47983. And rear lap-and-shoulder belts could therefore offer safety benefits for children old enough to use them without diminishing the safety of smaller children in car seats. Id., at 47988–47989 (“[T]he agency believes that this proposal for rear seat lap/shoulder belts would offer benefits for children riding in some types of booster seats, would have no positive or negative effects on children riding in most designs of car seats and children that are too small to use shoulder belts, and would offer older children the same incremental safety protection [as adults]”). Nor did DOT seek to use its regulation to spur the development of alternative kinds of rear aisle or middle seat safety devices. See 54 Fed. Reg. 46257. Why then did DOT not require lap-and-shoulder belts in these seats? We have found some indication that it thought use of lap-and-shoulder belts in rear aisle seats could cause “entry and exit problems for occupants of seating positions to the rear” by “stretch[ing] the shoulder belt across the aisleway,” id., at 46258. However, DOT encouraged manufacturers to address this issue through innovation: “[I]n those cases where manufacturers are able to design and install lap/shoulder belts at seating positions adjacent to aisleways without interfering with the aisleway’s purpose of allowing access to more rearward seating positions[, the agency] encourages the manufacturers to do so.” 54 Fed. Reg. 46258. And there is little indication that DOT considered this matter a significant safety concern. Cf. Letter from Philip R. Recht, Chief Counsel, National Highway Traffic Safety Admin., to Roger Matoba (Dec. 28, 1994), App. to Reply Brief for Petitioners 2 (“With respect to your concerns about the safety of shoulder safety belts which cross an aisle, I note that such belts do not in fact prevent rearward passengers from exiting the vehicle. Such passengers may … g[o] under or over the belt. They may also move the belt aside”). The more important reason why DOT did not require lap-and-shoulder belts for rear inner seats was that it thought that this requirement would not be cost-effective. The agency explained that it would be significantly more expensive for manufacturers to install lap-and-shoulder belts in rear middle and aisle seats than in seats next to the car doors. Ibid. But that fact—the fact that DOT made a negative judgment about cost effectiveness—cannot by itself show that DOT sought to forbid common-law tort suits in which a judge or jury might reach a different conclusion. For one thing, DOT did not believe that costs would remain frozen. Rather it pointed out that costs were falling as manufacturers were “voluntarily equipping more and more of their vehicles with rear seat lap/shoulder belts.” Ibid. For another thing, many, perhaps most, federal safety regulations embody some kind of cost-effectiveness judgment. While an agency could base a decision to pre-empt on its cost-effectiveness judgment, we are satisfied that the rulemaking record at issue here discloses no such pre-emptive intent. And to infer from the mere existence of such a cost-effectiveness judgment that the federal agency intends to bar States from imposing stricter standards would treat all such federal standards as if they were maximum standards, eliminating the possibility that the federal agency seeks only to set forth a minimum standard potentially supplemented through state tort law. We cannot reconcile this consequence with a statutory saving clause that foresees the likelihood of a continued meaningful role for state tort law. Supra, at 4. Finally, the Solicitor General tells us that DOT’s regulation does not pre-empt this tort suit. As in Geier, “the agency’s own views should make a difference.” 529 U. S., at 883. “Congress has delegated to DOT authority to implement the statute; the subject matter is technical; and the relevant history and background are complex and extensive. The agency is likely to have a thorough understanding of its own regulation and its objectives and is ‘uniquely qualified’ to comprehend the likely impact of state requirements.” Ibid. There is “no reason to suspect that the Solicitor General’s representation of DOT’s views reflects anything other than ‘the agency’s fair and considered judgment on the matter.’ ” Id., at 884 (quoting Auer v. Robbins, 519 U. S. 452, 462 (1997)). Neither has DOT expressed inconsistent views on this subject. In Geier, the Solicitor General pointed out that “state tort law does not conflict with a federal ‘minimum standard’ merely because state law imposes a more stringent requirement.” United States Brief in Geier 21 (citation omitted). And the Solicitor General explained that a standard giving manufacturers “multiple options for the design of” a device would not pre-empt a suit claiming that a manufacturer should have chosen one particular option, where “the Secretary did not determine that the availability of options was necessary to promote safety.” Id., at 22; see Brief for United States as Amicus Curiae in Wood v. General Motors Corp., O. T. 1989, No. 89–46, p. 15. This last statement describes the present case. In Geier, then, the regulation’s history, the agency’s contemporaneous explanation, and its consistently held interpretive views indicated that the regulation sought to maintain manufacturer choice in order to further significant regulatory objectives. Here, these same considerations indicate the contrary. We consequently conclude that, even though the state tort suit may restrict the manufacturer’s choice, it does not “stan[d] as an obstacle to the accomplishment . . . of the full purposes and objectives” of federal law. Hines, 312 U. S., at 67. Thus, the regulation does not pre-empt this tort action. The judgment of the California Court of Appeal is reversed. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

567.US.2011_11-182

An Arizona statute known as S. B. 1070 was enacted in 2010 to address pressing issues related to the large number of unlawful aliens in the State. The United States sought to enjoin the law as preempted. The District Court issued a preliminary injunction preventing four of its provisions from taking effect. Section 3 makes failure to comply with federal alien-registration requirements a state misdemeanor; §5(C) makes it a misdemeanor for an unauthorized alien to seek or engage in work in the State; §6 authorizes state and local officers to arrest without a warrant a person “the officer has probable cause to believe . . . has committed any public offense that makes the person removable from the United States”; and §2(B) requires officers conducting a stop, detention, or arrest to make efforts, in some circumstances, to verify the person’s immigration status with the Federal Government. The Ninth Circuit affirmed, agreeing that the United States had established a likelihood of success on its preemption claims. Held: 1. The Federal Government’s broad, undoubted power over immigration and alien status rests, in part, on its constitutional power to “establish an uniform Rule of Naturalization,” Art. I, §8, cl. 4, and on its inherent sovereign power to control and conduct foreign relations, see Toll v. Moreno, 458 U.S. 1, 10. Federal governance is extensive and complex. Among other things, federal law specifies categories of aliens who are ineligible to be admitted to the United States, 8 U. S. C. §1182; requires aliens to register with the Federal Government and to carry proof of status, §§1304(e), 1306(a); imposes sanctions on employers who hire unauthorized workers, §1324a; and specifies which aliens may be removed and the procedures for doing so, see §1227. Removal is a civil matter, and one of its principal features is the broad discretion exercised by immigration officials, who must decide whether to pursue removal at all. Immigration and Customs Enforcement (ICE), an agency within the Department of Homeland Security, is responsible for identifying, apprehending, and removing illegal aliens. It also operates the Law Enforcement Support Center, which provides immigration status information to federal, state, and local officials around the clock. Pp. 2–7. 2. The Supremacy Clause gives Congress the power to preempt state law. A statute may contain an express preemption provision, see, e.g., Chamber of Commerce of United States of America v. Whiting, 563 U. S. ___, ___, but state law must also give way to federal law in at least two other circumstances. First, States are precluded from regulating conduct in a field that Congress has determined must be regulated by its exclusive governance. See Gade v. National Solid Wastes Management Assn., 505 U.S. 88, 115. Intent can be inferred from a framework of regulation “so pervasive . . . that Congress left no room for the States to supplement it” or where a “federal interest is so dominant that the federal system will be assumed to preclude enforcement of state laws on the same subject.” Rice v. Santa Fe Elevator Corp., 331 U.S. 218, 230. Second, state laws are preempted when they conflict with federal law, including when they stand “as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress.” Hines v. Davidowitz, 312 U.S. 52, 67. Pp. 7–8. 3. Sections 3, 5(C), and 6 of S. B. 1070 are preempted by federal law. Pp. 8–19. (a) Section 3 intrudes on the field of alien registration, a field in which Congress has left no room for States to regulate. In Hines, a state alien-registration program was struck down on the ground that Congress intended its “complete” federal registration plan to be a “single integrated and all-embracing system.” 312 U. S., at 74. That scheme did not allow the States to “curtail or complement” federal law or “enforce additional or auxiliary regulations.” Id., at 66–67. The federal registration framework remains comprehensive. Because Congress has occupied the field, even complementary state regulation is impermissible. Pp. 8–11. (b) Section 5(C)’s criminal penalty stands as an obstacle to the federal regulatory system. The Immigration Reform and Control Act of 1986 (IRCA), a comprehensive framework for “combating the employment of illegal aliens,” Hoffman Plastic Compounds, Inc. v. NLRB, 535 U.S. 137, 147, makes it illegal for employers to knowingly hire, recruit, refer, or continue to employ unauthorized workers, 8 U. S. C. §§1324a(a)(1)(A), (a)(2), and requires employers to verify prospective employees’ employment authorization status, §§1324a(a)(1)(B), (b). It imposes criminal and civil penalties on employers, §§1324a(e)(4), (f), but only civil penalties on aliens who seek, or engage in, unauthorized employment, e.g., §§1255(c)(2), (c)(8). IRCA’s express preemption provision, though silent about whether additional penalties may be imposed against employees, “does not bar the ordinary working of conflict pre-emption principles” or impose a “special burden” making it more difficult to establish the preemption of laws falling outside the clause. Geier v. American Honda Motor Co., 529 U.S. 861, 869–872. The correct instruction to draw from the text, structure, and history of IRCA is that Congress decided it would be inappropriate to impose criminal penalties on unauthorized employees. It follows that a state law to the contrary is an obstacle to the regulatory system Congress chose. Pp. 12–15. (c) By authorizing state and local officers to make warrantless arrests of certain aliens suspected of being removable, §6 too creates an obstacle to federal law. As a general rule, it is not a crime for a removable alien to remain in the United States. The federal scheme instructs when it is appropriate to arrest an alien during the removal process. The Attorney General in some circumstances will issue a warrant for trained federal immigration officers to execute. If no federal warrant has been issued, these officers have more limited authority. They may arrest an alien for being “in the United States in violation of any [immigration] law or regulation,” for example, but only where the alien “is likely to escape before a warrant can be obtained.” §1357(a)(2). Section 6 attempts to provide state officers with even greater arrest authority, which they could exercise with no instruction from the Federal Government. This is not the system Congress created. Federal law specifies limited circumstances in which state officers may perform an immigration officer’s functions. This includes instances where the Attorney General has granted that authority in a formal agreement with a state or local government. See, e.g., §1357(g)(1). Although federal law permits state officers to “cooperate with the Attorney General in the identification, apprehension, detention, or removal of aliens not lawfully present in the United States,” §1357(g)(10)(B), this does not encompass the unilateral decision to detain authorized by §6. Pp. 15–19. 4. It was improper to enjoin §2(B) before the state courts had an opportunity to construe it and without some showing that §2(B)’s enforcement in fact conflicts with federal immigration law and its objectives. Pp. 19–24. (a) The state provision has three limitations: A detainee is presumed not to be an illegal alien if he or she provides a valid Arizona driver’s license or similar identification; officers may not consider race, color, or national origin “except to the extent permitted by the United States [and] Arizona Constitution[s]”; and §2(B) must be “implemented in a manner consistent with federal law regulating immigration, protecting the civil rights of all persons and respecting the privileges and immunities of United States citizens.” P. 20. (b) This Court finds unpersuasive the argument that, even with those limits, §2(B) must be held preempted at this stage. Pp. 20–24. (1) The mandatory nature of the status checks does not interfere with the federal immigration scheme. Consultation between federal and state officials is an important feature of the immigration system. In fact, Congress has encouraged the sharing of information about possible immigration violations. See §§1357(g)(10)(A), 1373(c). The federal scheme thus leaves room for a policy requiring state officials to contact ICE as a routine matter. Cf. Whiting, 563 U. S., at ___. Pp. 20–21. (2) It is not clear at this stage and on this record that §2(B), in practice, will require state officers to delay the release of detainees for no reason other than to verify their immigration status. This would raise constitutional concerns. And it would disrupt the federal framework to put state officers in the position of holding aliens in custody for possible unlawful presence without federal direction and supervision. But §2(B) could be read to avoid these concerns. If the law only requires state officers to conduct a status check during the course of an authorized, lawful detention or after a detainee has been released, the provision would likely survive preemption—at least absent some showing that it has other consequences that are adverse to federal law and its objectives. Without the benefit of a definitive interpretation from the state courts, it would be inappropriate to assume §2(B) will be construed in a way that conflicts with federal law. Cf. Fox v. Washington, 236 U.S. 273, 277. This opinion does not foreclose other preemption and constitutional challenges to the law as interpreted and applied after it goes into effect. Pp. 22–24. 641 F.3d 339, affirmed in part, reversed in part, and remanded. Kennedy, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, and Sotomayor, JJ., joined. Scalia, J., Thomas, J., and Alito, J., filed opinions concurring in part and dissenting in part. Kagan, J., took no part in the consideration or decision of the case.

To address pressing issues related to the large number of aliens within its borders who do not have a lawful right to be in this country, the State of Arizona in 2010 enacted a statute called the Support Our Law Enforcement and Safe Neighborhoods Act. The law is often referred to as S. B. 1070, the version introduced in the state senate. See also H. 2162 (2010) (amending S. 1070). Its stated purpose is to “discourage and deter the unlawful entry and presence of aliens and economic activity by persons unlawfully present in the United States.” Note following Ariz. Rev. Stat. Ann. §11–1051 (West 2012). The law’s provisions establish an official state policy of “attrition through enforcement.” Ibid. The question before the Court is whether federal law preempts and renders invalid four separate provisions of the state law. I The United States filed this suit against Arizona, seeking to enjoin S. B. 1070 as preempted. Four provisions of the law are at issue here. Two create new state offenses. Section 3 makes failure to comply with federal alien-registration requirements a state misdemeanor. Ariz. Rev. Stat. Ann. §13–1509 (West Supp. 2011). Section 5, in relevant part, makes it a misdemeanor for an unauthorized alien to seek or engage in work in the State; this provision is referred to as §5(C). See §13–2928(C). Two other provisions give specific arrest authority and inves- tigative duties with respect to certain aliens to state and local law enforcement officers. Section 6 authorizes officers to arrest without a warrant a person “the officer has probable cause to believe . . . has committed any public offense that makes the person removable from the United States.” §13–3883(A)(5). Section 2(B) provides that officers who conduct a stop, detention, or arrest must in some circumstances make efforts to verify the person’s immi- gration status with the Federal Government. See §11–1051(B) (West 2012). The United States District Court for the District of Arizona issued a preliminary injunction preventing the four provisions at issue from taking effect. 703 F. Supp. 2d 980, 1008 (2010). The Court of Appeals for the Ninth Circuit affirmed. 641 F.3d 339, 366 (2011). It agreed that the United States had established a likelihood of success on its preemption claims. The Court of Appeals was unanimous in its conclusion that §§3 and 5(C) were likely preempted. Judge Bea dissented from the decision to uphold the preliminary injunction against §§2(B) and 6. This Court granted certiorari to resolve important questions concerning the interaction of state and federal power with respect to the law of immigration and alien status. 565 U. S. ___ (2011). II A The Government of the United States has broad, undoubted power over the subject of immigration and the status of aliens. See Toll v. Moreno, 458 U.S. 1, 10 (1982); see generally S. Legomsky & C. Rodríguez, Immigration and Refugee Law and Policy 115–132 (5th ed. 2009). This authority rests, in part, on the National Government’s constitutional power to “establish an uniform Rule of Nat- uralization,” U. S. Const., Art. I, §8, cl. 4, and its inher- ent power as sovereign to control and conduct relations with foreign nations, see Toll, supra, at 10 (citing United States v. Curtiss-Wright Export Corp., 299 U.S. 304, 318 (1936)). The federal power to determine immigration policy is well settled. Immigration policy can affect trade, investment, tourism, and diplomatic relations for the entire Nation, as well as the perceptions and expectations of aliens in this country who seek the full protection of its laws. See, e.g., Brief for Argentina et al. as Amici Curiae; see also Harisiades v. Shaughnessy, 342 U.S. 580, 588–589 (1952). Perceived mistreatment of aliens in the United States may lead to harmful reciprocal treatment of American citizens abroad. See Brief for Madeleine K. Albright et al. as Amici Curiae 24–30. It is fundamental that foreign countries concerned about the status, safety, and security of their nationals in the United States must be able to confer and communicate on this subject with one national sovereign, not the 50 separate States. See Chy Lung v. Freeman, 92 U.S. 275, 279–280 (1876); see also The Federalist No. 3, p. 39 (C. Rossiter ed. 2003) (J. Jay) (observing that federal power would be necessary in part because “bordering States . . . under the impulse of sudden irritation, and a quick sense of apparent interest or injury” might take action that would undermine foreign relations). This Court has reaffirmed that “[o]ne of the most important and delicate of all international relationships . . . has to do with the protection of the just rights of a country’s own nationals when those nationals are in another country.” Hines v. Davidowitz, 312 U.S. 52, 64 (1941). Federal governance of immigration and alien status is extensive and complex. Congress has specified catego- ries of aliens who may not be admitted to the United States. See 8 U. S. C. §1182. Unlawful entry and unlawful reentry into the country are federal offenses. §§1325, 1326. Once here, aliens are required to register with the Federal Government and to carry proof of status on their person. See §§1301–1306. Failure to do so is a federal misdemeanor. §§1304(e), 1306(a). Federal law also authorizes States to deny noncitizens a range of public benefits, §1622; and it imposes sanctions on employers who hire unauthorized workers, §1324a. Congress has specified which aliens may be removed from the United States and the procedures for doing so. Aliens may be removed if they were inadmissible at the time of entry, have been convicted of certain crimes, or meet other criteria set by federal law. See §1227. Re- moval is a civil, not criminal, matter. A principal feature of the removal system is the broad discretion exercised by immigration officials. See Brief for Former Commissioners of the United States Immigration and Naturalization Service as Amici Curiae 8–13 (hereinafter Brief for Former INS Commissioners). Federal officials, as an initial matter, must decide whether it makes sense to pursue removal at all. If removal proceedings commence, aliens may seek asylum and other discretionary relief allowing them to remain in the country or at least to leave without formal removal. See §1229a(c)(4); see also, e.g., §§1158 (asylum), 1229b (cancellation of removal), 1229c (voluntary departure). Discretion in the enforcement of immigration law embraces immediate human concerns. Unauthorized workers trying to support their families, for example, likely pose less danger than alien smugglers or aliens who commit a serious crime. The equities of an individual case may turn on many factors, including whether the alien has children born in the United States, long ties to the community, or a record of distinguished military service. Some discretionary decisions involve policy choices that bear on this Nation’s international relations. Returning an alien to his own country may be deemed inappropriate even where he has committed a removable offense or fails to meet the criteria for admission. The foreign state may be mired in civil war, complicit in political persecution, or enduring conditions that create a real risk that the alien or his family will be harmed upon return. The dynamic nature of relations with other countries requires the Executive Branch to ensure that enforcement policies are con- sistent with this Nation’s foreign policy with respect to these and other realities. Agencies in the Department of Homeland Security play a major role in enforcing the country’s immigration laws. United States Customs and Border Protection (CBP) is re- sponsible for determining the admissibility of aliens and securing the country’s borders. See Dept. of Homeland Security, Office of Immigration Statistics, Immigration Enforcement Actions: 2010, p. 1 (2011). In 2010, CBP’s Border Patrol apprehended almost half a million people. Id., at 3. Immigration and Customs Enforcement (ICE), a second agency, “conducts criminal investigations involving the enforcement of immigration-related statutes.” Id., at 2. ICE also operates the Law Enforcement Support Center. LESC, as the Center is known, provides immigra- tion status information to federal, state, and local officials around the clock. See App. 91. ICE officers are respon- sible “for the identification, apprehension, and removal of illegal aliens from the United States.” Immigration Enforcement Actions, supra, at 2. Hundreds of thousands of aliens are removed by the Federal Government every year. See id., at 4 (reporting there were 387,242 removals, and 476,405 returns without a removal order, in 2010). B The pervasiveness of federal regulation does not di- minish the importance of immigration policy to the States. Arizona bears many of the consequences of unlawful immigration. Hundreds of thousands of deportable aliens are apprehended in Arizona each year. Dept. of Homeland Security, Office of Immigration Statistics, 2010 Yearbook of Immigration Statistics 93 (2011) (Table 35). Unauthorized aliens who remain in the State comprise, by one es- timate, almost six percent of the population. See Passel & Cohn, Pew Hispanic Center, U. S. Unauthorized Im- migration Flows Are Down Sharply Since Mid-Decade 3 (2010). And in the State’s most populous county, these aliens are reported to be responsible for a disproportionate share of serious crime. See, e.g., Camarota & Vaughan, Center for Immigration Studies, Immigration and Crime: Assessing a Conflicted Situation 16 (2009) (Table 3) (estimating that unauthorized aliens comprise 8.9% of the population and are responsible for 21.8% of the felonies in Maricopa County, which includes Phoenix). Statistics alone do not capture the full extent of Arizona’s concerns. Accounts in the record suggest there is an “epidemic of crime, safety risks, serious property damage, and environmental problems” associated with the influx of illegal migration across private land near the Mexican border. Brief for Petitioners 6. Phoenix is a major city of the United States, yet signs along an interstate highway 30 miles to the south warn the public to stay away. One reads, “DANGER—PUBLIC WARNING—TRAVEL NOT RECOMMENDED / Active Drug and Human Smuggling Area / Visitors May Encounter Armed Criminals and Smuggling Vehicles Traveling at High Rates of Speed.” App. 170; see also Brief for Petitioners 5–6. The problems posed to the State by illegal immigration must not be underestimated. These concerns are the background for the formal legal analysis that follows. The issue is whether, under pre- emption principles, federal law permits Arizona to implement the state-law provisions in dispute. III Federalism, central to the constitutional design, adopts the principle that both the National and State Governments have elements of sovereignty the other is bound to respect. See Gregory v. Ashcroft, 501 U.S. 452, 457 (1991); U. S. Term Limits, Inc. v. Thornton, 514 U.S. 779, 838 (1995) (Kennedy, J., concurring). From the existence of two sovereigns follows the possibility that laws can be in conflict or at cross-purposes. The Supremacy Clause provides a clear rule that federal law “shall be the supreme Law of the Land; and the Judges in every State shall be bound thereby, any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.” Art. VI, cl. 2. Under this principle, Congress has the power to preempt state law. See Crosby v. National Foreign Trade Council, 530 U.S. 363, 372 (2000); Gibbons v. Ogden, 9 Wheat. 1, 210–211 (1824). There is no doubt that Congress may withdraw specified powers from the States by enacting a statute containing an express preemption provision. See, e.g., Chamber of Commerce of United States of America v. Whiting, 563 U. S. ___, ___ (2011) (slip op., at 4). State law must also give way to federal law in at least two other circumstances. First, the States are precluded from regulating conduct in a field that Congress, acting within its proper authority, has determined must be regulated by its exclusive governance. See Gade v. National Solid Wastes Management Assn., 505 U.S. 88, 115 (1992). The intent to displace state law altogether can be inferred from a framework of regulation “so pervasive . . . that Congress left no room for the States to supplement it” or where there is a “federal interest . . . so dominant that the federal system will be assumed to preclude enforcement of state laws on the same subject.” Rice v. Santa Fe Elevator Corp., 331 U.S. 218, 230 (1947); see English v. General Elec. Co., 496 U.S. 72, 79 (1990). Second, state laws are preempted when they conflict with federal law. Crosby, supra, at 372. This includes cases where “compliance with both federal and state regulations is a physical impossibility,” Florida Lime & Avocado Growers, Inc. v. Paul, 373 U.S. 132, 142–143 (1963), and those instances where the challenged state law “stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress,” Hines, 312 U. S., at 67; see also Crosby, supra, at 373 (“What is a sufficient obstacle is a matter of judgment, to be informed by examining the federal statute as a whole and identifying its purpose and intended effects”). In preemption analysis, courts should assume that “the historic police powers of the States” are not superseded “unless that was the clear and manifest purpose of Congress.” Rice, supra, at 230; see Wyeth v. Levine, 555 U.S. 555, 565 (2009). The four challenged provisions of the state law each must be examined under these preemption principles. IV A Section 3 Section 3 of S. B. 1070 creates a new state misde- meanor. It forbids the “willful failure to complete or carry an alien registration document . . . in violation of 8 United States Code section 1304(e) or 1306(a).” Ariz. Rev. Stat. Ann. §11–1509(A) (West Supp. 2011). In effect, §3 adds a state-law penalty for conduct proscribed by federal law. The United States contends that this state enforcement mechanism intrudes on the field of alien registration, a field in which Congress has left no room for States to regulate. See Brief for United States 27, 31. The Court discussed federal alien-registration requirements in Hines v. Davidowitz, 312 U.S. 52. In 1940, as international conflict spread, Congress added to federal immigration law a “complete system for alien registration.” Id., at 70. The new federal law struck a careful balance. It punished an alien’s willful failure to register but did not require aliens to carry identification cards. There were also limits on the sharing of registration records and fingerprints. The Court found that Congress intended the federal plan for registration to be a “single integrated and all-embracing system.” Id., at 74. Because this “complete scheme . . . for the registration of aliens” touched on foreign relations, it did not allow the States to “curtail or complement” federal law or to “enforce additional or auxiliary regulations.” Id., at 66–67. As a con- sequence, the Court ruled that Pennsylvania could not enforce its own alien-registration program. See id., at 59, 74. The present regime of federal regulation is not identi- cal to the statutory framework considered in Hines, but it remains comprehensive. Federal law now includes a requirement that aliens carry proof of registration. 8 U. S. C. §1304(e). Other aspects, however, have stayed the same. Aliens who remain in the country for more than 30 days must apply for registration and be fingerprinted. Compare §1302(a) with id., §452(a) (1940 ed.). Detailed information is required, and any change of address has to be reported to the Federal Government. Compare §§1304(a), 1305(a) (2006 ed.), with id., §§455(a), 456 (1940 ed.). The statute continues to provide penalties for the willful failure to register. Compare §1306(a) (2006 ed.), with id., §457 (1940 ed.). The framework enacted by Congress leads to the conclusion here, as it did in Hines, that the Federal Government has occupied the field of alien registration. See American Ins. Assn. v. Garamendi, 539 U.S. 396, 419, n. 11 (2003) (characterizing Hines as a field preemption case); Pennsylvania v. Nelson, 350 U.S. 497, 504 (1956) (same); see also Dinh, Reassessing the Law of Preemption, 88 Geo. L. J. 2085, 2098–2099, 2107 (2000) (same). The federal statu- tory directives provide a full set of standards governing alien registration, including the punishment for noncompliance. It was designed as a “ ‘harmonious whole.’ ” Hines, supra, at 72. Where Congress occupies an entire field, as it has in the field of alien registration, even complementary state regulation is impermissible. Field pre- emption reflects a congressional decision to foreclose any state regulation in the area, even if it is parallel to fed- eral standards. See Silkwood v. Kerr-McGee Corp., 464 U.S. 238, 249 (1984). Federal law makes a single sovereign responsible for maintaining a comprehensive and unified system to keep track of aliens within the Nation’s borders. If §3 of the Arizona statute were valid, every State could give itself independent authority to prosecute federal registration violations, “diminish[ing] the [Federal Government]’s control over enforcement” and “detract[ing] from the ‘integrated scheme of regulation’ created by Congress.” Wisconsin Dept. of Industry v. Gould Inc., 475 U.S. 282, 288–289 (1986). Even if a State may make violation of federal law a crime in some instances, it cannot do so in a field (like the field of alien registration) that has been occupied by federal law. See California v. Zook, 336 U.S. 725, 730–731, 733 (1949); see also In re Loney, 134 U.S. 372, 375–376 (1890) (States may not impose their own punishment for perjury in federal courts). Arizona contends that §3 can survive preemption because the provision has the same aim as federal law and adopts its substantive standards. This argument not only ignores the basic premise of field preemption—that States may not enter, in any respect, an area the Federal Government has reserved for itself—but also is unpersuasive on its own terms. Permitting the State to impose its own penalties for the federal offenses here would conflict with the careful framework Congress adopted. Cf. Buckman Co. v. Plaintiffs’ Legal Comm., 531 U.S. 341, 347–348 (2001) (States may not impose their own punishment for fraud on the Food and Drug Administration); Wisconsin Dept., supra, at 288 (States may not impose their own punishment for repeat violations of the National Labor Relations Act). Were §3 to come into force, the State would have the power to bring criminal charges against individuals for violating a federal law even in circumstances where federal officials in charge of the comprehensive scheme determine that prosecution would frustrate federal policies. There is a further intrusion upon the federal scheme. Even where federal authorities believe prosecution is ap- propriate, there is an inconsistency between §3 and fed- eral law with respect to penalties. Under federal law, the failure to carry registration papers is a misdemeanor that may be punished by a fine, imprisonment, or a term of probation. See 8 U. S. C. §1304(e) (2006 ed.); 18 U. S. C. §3561. State law, by contrast, rules out probation as a possible sentence (and also eliminates the possibility of a pardon). See Ariz. Rev. Stat. Ann. §13–1509(D) (West Supp. 2011). This state framework of sanctions creates a conflict with the plan Congress put in place. See Wisconsin Dept., supra, at 286 (“[C]onflict is imminent whenever two separate remedies are brought to bear on the same activity” (internal quotation marks omitted)). These specific conflicts between state and federal law simply underscore the reason for field preemption. As it did in Hines, the Court now concludes that, with respect to the subject of alien registration, Congress intended to preclude States from “complement[ing] the federal law, or enforc[ing] additional or auxiliary regulations.” 312 U. S., at 66–67. Section 3 is preempted by federal law. B Section 5(C) Unlike §3, which replicates federal statutory requirements, §5(C) enacts a state criminal prohibition where no federal counterpart exists. The provision makes it a state misdemeanor for “an unauthorized alien to knowingly ap- ply for work, solicit work in a public place or perform work as an employee or independent contractor” in Ari- zona. Ariz. Rev. Stat. Ann. §13–2928(C) (West Supp. 2011). Violations can be punished by a $2,500 fine and incarceration for up to six months. See §13–2928(F); see also §§13–707(A)(1) (West 2010); 13–802(A); 13–902(A)(5). The United States contends that the provision upsets the bal- ance struck by the Immigration Reform and Control Act of 1986 (IRCA) and must be preempted as an obstacle to the federal plan of regulation and control. When there was no comprehensive federal program regulating the employment of unauthorized aliens, this Court found that a State had authority to pass its own laws on the subject. In 1971, for example, California passed a law imposing civil penalties on the employment of aliens who were “not entitled to lawful residence in the United States if such employment would have an adverse effect on lawful resident workers.” 1971 Cal. Stats. ch. 1442, §1(a). The law was upheld against a preemption challenge in De Canas v. Bica, 424 U.S. 351 (1976). De Canas recognized that “States possess broad authority under their police powers to regulate the employment relationship to protect workers within the State.” Id., at 356. At that point, however, the Federal Government had expressed no more than “a peripheral concern with [the] employment of illegal entrants.” Id., at 360; see Whiting, 563 U. S., at ___ (slip op., at 3). Current federal law is substantially different from the regime that prevailed when De Canas was decided. Congress enacted IRCA as a comprehensive framework for “combating the employment of illegal aliens.” Hoffman Plastic Compounds, Inc. v. NLRB, 535 U.S. 137, 147 (2002). The law makes it illegal for employers to know- ingly hire, recruit, refer, or continue to employ unauthorized workers. See 8 U. S. C. §§1324a(a)(1)(A), (a)(2). It also requires every employer to verify the employment authorization status of prospective employees. See §§1324a(a) (1)(B), (b); 8 CFR §274a.2(b) (2012). These requirements are enforced through criminal penalties and an escalat- ing series of civil penalties tied to the number of times an employer has violated the provisions. See 8 U. S. C. §§1324a(e)(4), (f); 8 CFR §274 A. 10. This comprehensive framework does not impose federal criminal sanctions on the employee side (i.e., penalties on aliens who seek or engage in unauthorized work). Under federal law some civil penalties are imposed instead. With certain exceptions, aliens who accept unlawful employment are not eligible to have their status adjusted to that of a lawful permanent resident. See 8 U. S. C. §§1255(c)(2), (c)(8). Aliens also may be removed from the country for having engaged in unauthorized work. See §1227(a)(1)(C)(i); 8 CFR §214.1(e). In addition to specifying these civil consequences, federal law makes it a crime for unauthorized workers to obtain employment through fraudulent means. See 18 U. S. C. §1546(b). Congress has made clear, however, that any information employees submit to indicate their work status “may not be used” for purposes other than prosecution under specified federal criminal statutes for fraud, perjury, and related conduct. See 8 U. S. C. §§1324a(b)(5), (d)(2)(F)–(G). The legislative background of IRCA underscores the fact that Congress made a deliberate choice not to impose crim- inal penalties on aliens who seek, or engage in, unauthorized employment. A commission established by Congress to study immigration policy and to make recommen- dations concluded these penalties would be “unnecessary and unworkable.” U. S. Immigration Policy and the National Interest: The Final Report and Recommendations of the Select Commission on Immigration and Refugee Policy with Supplemental Views by Commissioners 65–66 (1981); see Pub. L. 95–412, §4, 92Stat. 907. Proposals to make unauthorized work a criminal offense were debated and discussed during the long process of drafting IRCA. See Brief for Service Employees International Union et al. as Amici Curiae 9–12. But Congress rejected them. See, e.g., 119 Cong. Rec. 14184 (1973) (statement of Rep. Dennis). In the end, IRCA’s framework reflects a considered judgment that making criminals out of aliens engaged in unauthorized work—aliens who already face the possibility of employer exploitation because of their removable status—would be inconsistent with federal policy and ob- jectives. See, e.g., Hearings before the Subcommittee No. 1 of the House Committee on the Judiciary, 92d Cong., 1st Sess., pt. 3, pp. 919–920 (1971) (statement of Rep. Rodino, the eventual sponsor of IRCA in the House of Representatives). IRCA’s express preemption provision, which in most instances bars States from imposing penalties on employers of unauthorized aliens, is silent about whether additional penalties may be imposed against the employees themselves. See 8 U. S. C. §1324a(h)(2); Whiting, supra, at ___–___ (slip op., at 1–2). But the existence of an “express pre-emption provisio[n] does not bar the ordinary working of conflict pre-emption principles” or impose a “special burden” that would make it more difficult to establish the preemption of laws falling outside the clause. Geier v. American Honda Motor Co., 529 U.S. 861, 869–872 (2000); see Sprietsma v. Mercury Marine, 537 U.S. 51, 65 (2002). The ordinary principles of preemption include the well-settled proposition that a state law is preempted where it “stands as an obstacle to the accomplishment and exe- cution of the full purposes and objectives of Congress.” Hines, 312 U. S., at 67. Under §5(C) of S. B. 1070, Arizona law would interfere with the careful balance struck by Congress with respect to unauthorized employment of aliens. Although §5(C) attempts to achieve one of the same goals as federal law—the deterrence of unlawful employment—it involves a conflict in the method of enforcement. The Court has recognized that a “[c]onflict in technique can be fully as disruptive to the system Congress enacted as conflict in overt policy.” Motor Coach Employees v. Lockridge, 403 U.S. 274, 287 (1971). The correct instruction to draw from the text, structure, and history of IRCA is that Congress decided it would be inappropriate to impose criminal penalties on aliens who seek or engage in unauthorized employment. It follows that a state law to the contrary is an obstacle to the regulatory system Congress chose. See Puerto Rico Dept. of Con- sumer Affairs v. ISLA Petroleum Corp., 485 U.S. 495, 503 (1988) (“Where a comprehensive federal scheme intentionally leaves a portion of the regulated field without controls, then the pre-emptive inference can be drawn—not from federal inaction alone, but from inaction joined with action”). Section 5(C) is preempted by federal law. C Section 6 Section 6 of S. B. 1070 provides that a state officer, “without a warrant, may arrest a person if the officer has probable cause to believe . . . [the person] has committed any public offense that makes [him] removable from the United States.” Ariz. Rev. Stat. Ann. §13–3883(A)(5) (West Supp. 2011). The United States argues that arrests authorized by this statute would be an obstacle to the removal system Congress created. As a general rule, it is not a crime for a removable alien to remain present in the United States. See INS v. Lopez-Mendoza, 468 U.S. 1032, 1038 (1984). If the police stop someone based on nothing more than possible removability, the usual predicate for an arrest is absent. When an alien is suspected of being removable, a federal official issues an administrative document called a Notice to Ap- pear. See 8 U. S. C. §1229(a); 8 CFR §239.1(a) (2012). The form does not authorize an arrest. Instead, it gives the alien information about the proceedings, including the time and date of the removal hearing. See 8 U. S. C. §1229(a)(1). If an alien fails to appear, an in absentia order may direct removal. §1229a(5)(A). The federal statutory structure instructs when it is ap- propriate to arrest an alien during the removal process. For example, the Attorney General can exercise discretion to issue a warrant for an alien’s arrest and detention “pending a decision on whether the alien is to be removed from the United States.” 8 U. S. C. §1226(a); see Memorandum from John Morton, Director, ICE, to All Field Office Directors et al., Exercising Prosecutorial Discretion Consistent with the Civil Immigration Enforcement Priorities of the Agency for the Apprehension, Detention, and Removal of Aliens (June 17, 2011) (hereinafter 2011 ICE Memorandum) (describing factors informing this and re- lated decisions). And if an alien is ordered removed after a hearing, the Attorney General will issue a warrant. See 8 CFR §241.2(a)(1). In both instances, the warrants are executed by federal officers who have received training in the enforcement of immigration law. See §§241.2(b), 287.5(e)(3). If no federal warrant has been issued, those officers have more limited authority. See 8 U. S. C. §1357(a). They may arrest an alien for being “in the United States in violation of any [immigration] law or regula- tion,” for example, but only where the alien “is likely to escape before a warrant can be obtained.” §1357(a)(2). Section 6 attempts to provide state officers even greater authority to arrest aliens on the basis of possible removability than Congress has given to trained federal immi- gration officers. Under state law, officers who believe an alien is removable by reason of some “public offense” would have the power to conduct an arrest on that basis regardless of whether a federal warrant has issued or the alien is likely to escape. This state authority could be exercised without any input from the Federal Government about whether an arrest is warranted in a particular case. This would allow the State to achieve its own immigra- tion policy. The result could be unnecessary harassment of some aliens (for instance, a veteran, college student, or someone assisting with a criminal investigation) whom federal officials determine should not be removed. This is not the system Congress created. Federal law specifies limited circumstances in which state officers may perform the functions of an immigration officer. A principal example is when the Attorney General has granted that authority to specific officers in a formal agreement with a state or local government. See §1357(g)(1); see also §1103(a)(10) (authority may be extended in the event of an “imminent mass influx of aliens off the coast of the United States”); §1252c (authority to arrest in specific circumstance after consultation with the Federal Government); §1324(c) (authority to arrest for bringing in and harboring certain aliens). Officers covered by these agreements are subject to the Attorney General’s direction and super- vision. §1357(g)(3). There are significant complexities involved in enforcing federal immigration law, including the determination whether a person is removable. See Padilla v. Kentucky, 559 U. S. ___, ___–___ (2010) (Alito, J., concurring in judgment) (slip op., at 4–7). As a result, the agreements reached with the Attorney General must contain written certification that officers have received adequate training to carry out the duties of an immigration officer. See §1357(g)(2); cf. 8 CFR §§287.5(c) (ar- rest power contingent on training), 287.1(g) (defining the training). By authorizing state officers to decide whether an alien should be detained for being removable, §6 violates the principle that the removal process is entrusted to the discretion of the Federal Government. See, e.g., Reno v. American-Arab Anti-Discrimination Comm., 525 U.S. 471, 483–484 (1999); see also Brief for Former INS Commissioners 8–13. A decision on removability requires a determination whether it is appropriate to allow a foreign national to continue living in the United States. Decisions of this nature touch on foreign relations and must be made with one voice. See Jama v. Immigration and Customs Enforcement, 543 U.S. 335, 348 (2005) (“Removal decisions, including the selection of a removed alien’s destination, may implicate [the Nation’s] relations with foreign powers and require consideration of changing political and economic circumstances” (internal quotation marks omitted)); see also Galvan v. Press, 347 U.S. 522, 531 (1954) (“Policies pertaining to the entry of aliens and their right to remain here are . . . entrusted exclusively to Congress . . .”); Truax v. Raich, 239 U.S. 33, 42 (1915) (“The authority to control immigration—to admit or exclude aliens—is vested solely in the Federal Government”). In defense of §6, Arizona notes a federal statute permitting state officers to “cooperate with the Attorney General in the identification, apprehension, detention, or removal of aliens not lawfully present in the United States.” 8 U. S. C. §1357(g)(10)(B). There may be some ambiguity as to what constitutes cooperation under the federal law; but no coherent understanding of the term would incorporate the unilateral decision of state officers to arrest an alien for being removable absent any request, approval, or other instruction from the Federal Government. The Department of Homeland Security gives examples of what would constitute cooperation under federal law. These include situations where States participate in a joint task force with federal officers, provide operational support in executing a warrant, or allow federal immigration officials to gain access to detainees held in state facilities. See Dept. of Homeland Security, Guidance on State and Local Governments’ Assistance in Immigration Enforcement and Related Matters 13–14 (2011), online at http:// www.dhs.gov/files/resources/immigration.shtm (all Internet materials as visited June 21, 2012, and available in Clerk of Court’s case file). State officials can also assist the Federal Government by responding to requests for information about when an alien will be released from their custody. See §1357(d). But the unilateral state action to detain authorized by §6 goes far beyond these measures, defeating any need for real cooperation. Congress has put in place a system in which state officers may not make warrantless arrests of aliens based on possible removability except in specific, limited circumstances. By nonetheless authorizing state and local offi- cers to engage in these enforcement activities as a general matter, §6 creates an obstacle to the full purposes and objectives of Congress. See Hines, 312 U. S., at 67. Section 6 is preempted by federal law. D Section 2(B) Section 2(B) of S. B. 1070 requires state officers to make a “reasonable attempt . . . to determine the immigration status” of any person they stop, detain, or arrest on some other legitimate basis if “reasonable suspicion exists that the person is an alien and is unlawfully present in the United States.” Ariz. Rev. Stat. Ann. §11–1051(B) (West 2012). The law also provides that “[a]ny person who is arrested shall have the person’s immigration status determined before the person is released.” Ibid. The accepted way to perform these status checks is to contact ICE, which maintains a database of immigration records. Three limits are built into the state provision. First, a detainee is presumed not to be an alien unlawfully present in the United States if he or she provides a valid Arizona driver’s license or similar identification. Second, officers “may not consider race, color or national origin . . . except to the extent permitted by the United States [and] Ari- zona Constitution[s].” Ibid. Third, the provisions must be “implemented in a manner consistent with federal law regulating immigration, protecting the civil rights of all persons and respecting the privileges and immunities of United States citizens.” §11–1051(L) (West 2012). The United States and its amici contend that, even with these limits, the State’s verification requirements pose an obstacle to the framework Congress put in place. The first concern is the mandatory nature of the status checks. The second is the possibility of prolonged detention while the checks are being performed. 1 Consultation between federal and state officials is an important feature of the immigration system. Congress has made clear that no formal agreement or special training needs to be in place for state officers to “communicate with the [Federal Government] regarding the immigration status of any individual, including reporting knowledge that a particular alien is not lawfully present in the United States.” 8 U. S. C. §1357(g)(10)(A). And Congress has obligated ICE to respond to any request made by state officials for verification of a person’s citizenship or im- migration status. See §1373(c); see also §1226(d)(1)(A) (requiring a system for determining whether individuals arrested for aggravated felonies are aliens). ICE’s Law Enforcement Support Center operates “24 hours a day, seven days a week, 365 days a year” and provides, among other things, “immigration status, identity information and real-time assistance to local, state and federal law enforcement agencies.” ICE, Fact Sheet: Law Enforcement Support Center (May 29, 2012), online at http:// www.ice.gov/news/library/factsheets/lesc.htm. LESC responded to more than one million requests for information in 2009 alone. App. 93. The United States argues that making status verification mandatory interferes with the federal immigration scheme. It is true that §2(B) does not allow state officers to consider federal enforcement priorities in deciding whether to contact ICE about someone they have detained. See Brief for United States 47–50. In other words, the officers must make an inquiry even in cases where it seems unlikely that the Attorney General would have the alien removed. This might be the case, for example, when an alien is an elderly veteran with significant and longstanding ties to the community. See 2011 ICE Memorandum 4–5 (mentioning these factors as relevant). Congress has done nothing to suggest it is inappropriate to communicate with ICE in these situations, however. Indeed, it has encouraged the sharing of information about possible immigration violations. See 8 U. S. C. §1357(g) (10)(A). A federal statute regulating the public benefits provided to qualified aliens in fact instructs that “no State or local government entity may be prohibited, or in any way restricted, from sending to or receiving from [ICE] information regarding the immigration status, lawful or unlawful, of an alien in the United States.” §1644. The federal scheme thus leaves room for a policy requiring state officials to contact ICE as a routine matter. Cf. Whiting, 563 U. S., at ___–___ (slip op., at 23–24) (rejecting argument that federal law preempted Arizona’s requirement that employers determine whether employees were eligible to work through the federal E-Verify system where the Federal Government had encouraged its use). 2 Some who support the challenge to §2(B) argue that, in practice, state officers will be required to delay the release of some detainees for no reason other than to verify their immigration status. See, e.g., Brief for Former Arizona Attorney General Terry Goddard et al. as Amici Curiae 37, n. 49. Detaining individuals solely to verify their immigration status would raise constitutional concerns. See, e.g., Arizona v. Johnson, 555 U.S. 323, 333 (2009); Illinois v. Caballes, 543 U.S. 405, 407 (2005) (“A seizure that is justified solely by the interest in issuing a warning ticket to the driver can become unlawful if it is prolonged beyond the time reasonably required to complete that mission”). And it would disrupt the federal framework to put state officers in the position of holding aliens in custody for possible unlawful presence without federal direction and supervision. Cf. Part IV–C, supra (concluding that Ari- zona may not authorize warrantless arrests on the basis of removability). The program put in place by Congress does not allow state or local officers to adopt this enforcement mechanism. But §2(B) could be read to avoid these concerns. To take one example, a person might be stopped for jaywalking in Tucson and be unable to produce identification. The first sentence of §2(B) instructs officers to make a “reasonable” attempt to verify his immigration status with ICE if there is reasonable suspicion that his presence in the United States is unlawful. The state courts may conclude that, unless the person continues to be suspected of some crime for which he may be detained by state officers, it would not be reasonable to prolong the stop for the immigration inquiry. See Reply Brief for Petitioners 12, n. 4 (“[Section 2(B)] does not require the verification be completed during the stop or detention if that is not reasonable or practicable”); cf. Muehler v. Mena, 544 U.S. 93, 101 (2005) (finding no Fourth Amendment violation where questioning about immigration status did not prolong a stop). To take another example, a person might be held pending release on a charge of driving under the influence of alcohol. As this goes beyond a mere stop, the arrestee (unlike the jaywalker) would appear to be subject to the categorical requirement in the second sentence of §2(B) that “[a]ny person who is arrested shall have the person’s immigration status determined before [he] is released.” State courts may read this as an instruction to initiate a status check every time someone is arrested, or in some subset of those cases, rather than as a command to hold the person until the check is complete no matter the circumstances. Even if the law is read as an instruction to complete a check while the person is in custody, moreover, it is not clear at this stage and on this record that the verification process would result in prolonged detention. However the law is interpreted, if §2(B) only requires state officers to conduct a status check during the course of an authorized, lawful detention or after a detainee has been released, the provision likely would survive pre- emption—at least absent some showing that it has other consequences that are adverse to federal law and its objectives. There is no need in this case to address whether reasonable suspicion of illegal entry or another immigration crime would be a legitimate basis for prolonging a detention, or whether this too would be preempted by federal law. See, e.g., United States v. Di Re, 332 U.S. 581, 589 (1948) (authority of state officers to make arrests for federal crimes is, absent federal statutory instruction, a matter of state law); Gonzales v. Peoria, 722 F.2d 468, 475–476 (CA9 1983) (concluding that Arizona officers have authority to enforce the criminal provisions of federal immigration law), overruled on other grounds in Hodgers-Durgin v. de la Vina, 199 F.3d 1037 (CA9 1999). The nature and timing of this case counsel caution in evaluating the validity of §2(B). The Federal Government has brought suit against a sovereign State to challenge the provision even before the law has gone into effect. There is a basic uncertainty about what the law means and how it will be enforced. At this stage, without the benefit of a definitive interpretation from the state courts, it would be inappropriate to assume §2(B) will be construed in a way that creates a conflict with federal law. Cf. Fox v. Washington, 236 U.S. 273, 277 (1915) (“So far as statutes fairly may be construed in such a way as to avoid doubtful constitutional questions they should be so construed; and it is to be presumed that state laws will be construed in that way by the state courts” (citation omitted)). As a result, the United States cannot prevail in its current challenge. See Huron Portland Cement Co. v. Detroit, 362 U.S. 440, 446 (1960) (“To hold otherwise would be to ignore the teaching of this Court’s decisions which enjoin seeking out conflicts between state and federal regulation where none clearly exists”). This opinion does not foreclose other preemption and constitutional challenges to the law as interpreted and applied after it goes into effect. V Immigration policy shapes the destiny of the Nation. On May 24, 2012, at one of this Nation’s most distinguished museums of history, a dozen immigrants stood before the tattered flag that inspired Francis Scott Key to write the National Anthem. There they took the oath to become American citizens. The Smithsonian, News Release, Smithsonian Citizenship Ceremony Welcomes a Dozen New Americans (May 24, 2012), online at http://newsdesk.si.edu/releases. These naturalization cere- monies bring together men and women of different ori- gins who now share a common destiny. They swear a common oath to renounce fidelity to foreign princes, to defend the Constitution, and to bear arms on behalf of the country when required by law. 8 CFR §337.1(a) (2012). The history of the United States is in part made of the stories, talents, and lasting contributions of those who crossed oceans and deserts to come here. The National Government has significant power to regulate immigration. With power comes responsibility, and the sound exercise of national power over immigration depends on the Nation’s meeting its responsibility to base its laws on a political will informed by searching, thoughtful, rational civic discourse. Arizona may have understandable frustrations with the problems caused by illegal immigration while that process continues, but the State may not pursue policies that undermine federal law. * * * The United States has established that §§3, 5(C), and 6 of S. B. 1070 are preempted. It was improper, however, to enjoin §2(B) before the state courts had an opportunity to construe it and without some showing that enforcement of the provision in fact conflicts with federal immigration law and its objectives. The judgment of the Court of Appeals for the Ninth Circuit is affirmed in part and reversed in part. The case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case.

566.US.2011_11-161

For decades, Indianapolis (City) funded sewer projects using Indiana’s Barrett Law, which permitted cities to apportion a public improvement project’s costs equally among all abutting lots. Under that system, a city would create an initial assessment, dividing the total estimated cost by the number of lots and making any necessary adjustments. Upon a project’s completion, the city would issue a final lot-by-lot assessment. Lot owners could elect to pay the assessment in a lump sum or over time in installments. After the City completed the Brisbane/Manning Sanitary Sewers Project, it sent affected homeowners formal notice of their payment obligations. Of the 180 affected homeowners, 38 elected to pay the lump sum. The following year, the City abandoned Barrett Law financing and adopted the Septic Tank Elimination Program (STEP), which financed projects in part through bonds, thereby lowering individual owner’s sewer-connection costs. In implementing STEP, the City’s Board of Public Works enacted a resolution forgiving all assessment amounts still owed pursuant to Barrett Law financing. Homeowners who had paid the Brisbane/Manning Project lump sum received no refund, while homeowners who had elected to pay in installments were under no obligation to make further payments. The 38 homeowners who paid the lump sum asked the City for a refund, but the City denied the request. Thirty-one of these homeowners brought suit in Indiana state court claiming, in relevant part, that the City’s refusal violated the Federal Equal Protection Clause. The trial court granted summary judgment to the homeowners, and the State Court of Appeals affirmed. The Indiana Supreme Court reversed, holding that the City’s distinction between those who had already paid and those who had not was rationally related to its legitimate interests in reducing administrative costs, providing financial hardship relief to homeowners, transitioning from the Barrett Law system to STEP, and preserving its limited resources. Held: The City had a rational basis for its distinction and thus did not violate the Equal Protection Clause. Pp. 6–14. (a) The City’s classification does not involve a fundamental right or suspect classification. See Heller v. Doe, 509 U.S. 312, 319–320. Its subject matter is local, economic, social, and commercial. See United States v. Carolene Products Co., 304 U.S. 144, 152. It is a tax classification. See Regan v. Taxation With Representation of Wash., 461 U.S. 540, 547. And no one claims that the City has discriminated against out-of-state commerce or new residents. Cf. Hooper v. Bernalillo County Assessor, 472 U.S. 612. Hence, the City’s distinction does not violate the Equal Protection Clause as long as “there is any reasonably conceivable state of facts that could provide a rational basis for the classification,” FCC v. Beach Communications, Inc., 508 U.S. 307, 313, and the “ ‘burden is on the one attacking the [classification] to negative every conceivable basis which might support it,’ ” Heller, supra, at 320. Pp. 6–7. (b) Administrative concerns can ordinarily justify a tax-related distinction, see, e.g., Carmichael v. Southern Coal & Coke Co., 301 U.S. 495, 511–512, and the City’s decision to stop collecting outstanding Barrett Law debts finds rational support in the City’s administrative concerns. After the City switched to the STEP system, any decision to continue Barrett Law debt collection could have proved complex and expensive. It would have meant maintaining an administrative system for years to come to collect debts arising out of 20-plus different construction projects built over the course of a decade, involving monthly payments as low as $25 per household, with the possible need to maintain credibility by tracking down defaulting debtors and bringing legal action. The rationality of the City’s distinction draws further support from the nature of the line-drawing choices that confronted it. To have added refunds to forgiveness would have meant adding further administrative costs, namely the cost of processing refunds. And limiting refunds only to Brisbane/Manning homeowners would have led to complaints of unfairness, while expanding refunds to the apparently thousands of other Barrett Law project homeowners would have involved an even greater administrative burden. Finally, the rationality of the distinction draws support from the fact that the line that the City drew—distinguishing past payments from future obligations—is well known to the law. See, e.g., 26 U. S. C. §108(a)(1)(E). Pp. 7–10. (c) Petitioners’ contrary arguments are unpersuasive. Whether financial hardship is a factor supporting rationality need not be considered here, since the City’s administrative concerns are sufficient to show a rational basis for its distinction. Petitioners propose other forgiveness systems that they argue are superior to the City’s system, but the Constitution only requires that the line actually drawn by the City be rational. Petitioners further argue that administrative considerations alone should not justify a tax distinction lest a city justify an unfair system through insubstantial administrative considerations. Here it was rational for the City to draw a line that avoided the administrative burden of both collecting and paying out small sums for years to come. Petitioners have not shown that the administrative concerns are too insubstantial to justify the classification. Finally, petitioners argue that precedent makes it more difficult for the City to show a rational basis, but the cases to which they refer involve discrimination based on residence or length of residence. The one exception, Allegheny Pittsburgh Coal Co. v. Commission of Webster Cty., 488 U.S. 336, is distinguishable. Pp. 10–14. 946 N.E.2d 553, affirmed. Breyer, J., delivered the opinion of the Court, in which Kennedy, Thomas, Ginsburg, Sotomayor, and Kagan, JJ., joined. Roberts, C. J., filed a dissenting opinion, in which Scalia and Alito, JJ., joined.

For many years, an Indiana statute, the “Barrett Law,” authorized Indiana’s cities to impose upon benefited lot owners the cost of sewer improvement projects. The Law also permitted those lot owners to pay either immediately in the form of a lump sum or over time in installments. In 2005, the city of Indianapolis (City) adopted a new as- sessment and payment method, the “STEP” plan, and it forgave any Barrett Law installments that lot owners had not yet paid. A group of lot owners who had already paid their entire Barrett Law assessment in a lump sum believe that the City should have provided them with equivalent refunds. And we must decide whether the City’s refusal to do so un- constitutionally discriminates against them in violation of the Equal Protection Clause, Amdt. 14, §1. We hold that the City had a rational basis for distinguishing between those lot owners who had already paid their share of project costs and those who had not. And we conclude that there is no equal protection violation. I A Beginning in 1889 Indiana’s Barrett Law permitted cities to pay for public improvements, such as sewage proj- ects, by “apportion[ing]” the costs of a project “equally among all abutting lands or lots.” Ind. Code §36–9–39–15(b)(3) (2011); see Town Council of New Harmony v. Parker, 726 N.E.2d 1217, 1227, n. 13 (Ind. 2000) (project’s beneficiaries pay its costs). When a city built a Barrett Law project, the city’s public works board would create an initial lot-owner assessment by “dividing the estimated total cost of the sewage works by the total number of lots.” §36–9–39–16(a). It might then adjust an individual assessment downward if the lot would benefit less than would others. §36–9–39–17(b). Upon completion of the project, the board would issue a final lot-by-lot assessment. The Law permitted lot owners to pay the assessment either in a single lump sum or over time in installment payments (with interest). The City would collect installment payments “in the same manner as other taxes.” §36–9–37–6. The Law authorized 10-, 20-, or 30-year installment plans. §36–9–37–8.5(a). Until fully paid, an assessment would constitute a lien against the property, permitting the city to initiate foreclosure proceedings in case of a default. §§36–9–37–9(b), –22. For several decades, Indianapolis used the Barrett Law system to fund sewer projects. See, e.g., Conley v. Brummit, 92 Ind. App. 620, 621, 176 N.E. 880, 881 (1931) (in banc). But in 2005, the City adopted a new system, called the Septic Tank Elimination Program (STEP), which fi- nanced projects in part through bonds, thereby lowering in- dividual lot owners’ sewer-connection costs. By that time, the City had constructed more than 40 Barrett Law projects. App. to Pet. for Cert. 5a. We are told that installment-paying lot owners still owed money in respect to 24 of those projects. See Reply Brief for Petitioners 16–17, n. 3 (citing City’s Response to Plaintiff’s Brief on Damages, Record in Cox v. Indianapolis, No. 1:09–cv–0435 (SD Ind., Doc. 98–1 (Exh. A)). In respect to 21 of the 24, some installment payments had not yet fallen due; in respect to the other 3, those who owed money were in default. Reply Brief for Petitioners 17, n. 3. B This case concerns one of the 24 still-open Barrett Law projects, namely the Brisbane/Manning Sanitary Sewers Project. The Brisbane/Manning Project began in 2001. It connected about 180 homes to the City’s sewage system. Construction was completed in 2003. The Indianapolis Board of Public Works held an assessment hearing in June 2004. And in July 2004 the Board sent the 180 affected homeowners a formal notice of their payment obligations. The notice made clear that each homeowner could pay the entire assessment—$9,278 per property—in a lump sum or in installments, which would include interest at a 3.5% annual rate. Under an installment plan, payments would amount to $77.27 per month for 10 years; $38.66 per month for 20 years; or $25.77 per month for 30 years. In the event, 38 homeowners chose to pay up front; 47 chose the 10-year plan; 27 chose the 20-year plan; and 68 chose the 30-year plan. And in the first year each homeowner paid the amount due ($9,278 upfront; $927.80 under the 10-year plan; $463.90 under the 20-year plan, or $309.27 under the 30-year plan). App. to Pet. for Cert. 48a. The next year, however, the City decided to abandon the Barrett Law method of financing. It thought that the Barrett Law’s lot-by-lot payments had become too burdensome for many homeowners to pay, discouraging changes from less healthy septic tanks to healthier sewer systems. See id., at 4a–5a. (For example, homes helped by the Brisbane/Manning Project, at a cost of more than $9,000 each, were then valued at $120,000 to $270,000. App. 67.) The City’s new STEP method of financing would charge each connecting lot owner a flat $2,500 fee and make up the difference by floating bonds eventually paid for by all lot owners citywide. See App. to Pet. for Cert. 5a, n. 5. On October 31, 2005, the City enacted an ordinance implementing its decision. In December, the City’s Board of Public Works enacted a further resolution, Resolution 101, which, as part of the transition, would “forgive all assessment amounts . . . established pursuant to the Barrett Law Funding for Municipal Sewer programs due and owing from the date of November 1, 2005 forward.” App. 72 (emphasis added). In its preamble, the Resolution said that the Barrett Law “may present financial hardships on many middle to lower income participants who most need sanitary sewer service in lieu of failing septic systems”; it pointed out that the City was transitioning to the new STEP method of financing; and it said that the STEP method was based upon a financial model that had “considered the current assessments being made by participants in active Barrett Law projects” as well as future projects. Id., at 71–72. The upshot was that those who still owed Barrett Law assessments would not have to make further payments but those who had already paid their assessments would not receive refunds. This meant that homeowners who had paid the full $9,278 Brisbane/ Manning Project assessment in a lump sum the preced- ing year would receive no refund, while homeowners who had elected to pay the assessment in installments, and had paid a total of $309.27, $463.90, or $927.80, would be under no obligation to make further payments. In February 2006, the 38 homeowners who had paid the full Brisbane/Manning Project assessment asked the City for a partial refund (in an amount equal to the smallest forgiven Brisbane/Manning installment debt, apparently $8,062). The City denied the request in part because “[r]efunding payments made in your project area, or any portion of the payments, would establish a precedent of unfair and inequitable treatment to all other property owners who have also paid Barrett Law assessments . . . and while [the November 1, 2005, cutoff date] might seem arbitrary to you, it is essential for the City to establish this date and move forward with the new funding approach.” Id., at 50–51. C Thirty-one of the thirty-eight Brisbane/Manning Project lump-sum homeowners brought this lawsuit in Indiana state court seeking a refund of about $8,000 each. They claimed in relevant part that the City’s refusal to provide them with refunds at the same time that the City forgave the outstanding Project debts of other Brisbane/Manning homeowners violated the Federal Constitution’s Equal Pro- tection Clause, Amdt. 14, §1; see also Rev. Stat. §1979, 42 U. S. C. §1983. The trial court granted summary judgment in their favor. The State Court of Appeals affirmed that judgment. 918 N.E.2d 401 (2009). But the Indiana Supreme Court reversed. 946 N.E.2d 553 (2011). In its view, the City’s distinction between those who had already paid their Barrett Law assessments and those who had not was “rationally related to its legitimate interests in reducing its administrative costs, providing relief for property owners experiencing financial hardship, establishing a clear transition from [the] Barrett Law to STEP, and preserving its limited resources.” App. to Pet. for Cert. 19a. We granted certiorari to consider the equal protection question. And we now affirm the Indiana Supreme Court. II A As long as the City’s distinction has a rational basis, that distinction does not violate the Equal Protection Clause. This Court has long held that “a classification neither involving fundamental rights nor proceeding along suspect lines . . . cannot run afoul of the Equal Protection Clause if there is a rational relationship between the dis- parity of treatment and some legitimate governmental purpose.” Heller v. Doe, 509 U.S. 312, 319–320 (1993); cf. Gulf, C. & S. F. R. Co. v. Ellis, 165 U.S. 150, 155, 165–166 (1897). We have made clear in analogous contexts that, where “ordinary commercial transactions” are at issue, ra- tional basis review requires deference to reasonable under- lying legislative judgments. United States v. Carolene Products Co., 304 U.S. 144, 152 (1938) (due process); see also New Orleans v. Dukes, 427 U.S. 297, 303 (1976) (per curiam) (equal protection). And we have repeatedly pointed out that “[l]egislatures have especially broad latitude in creating classifications and distinctions in tax statutes.” Regan v. Taxation With Representation of Wash., 461 U.S. 540, 547 (1983); see also Fitzgerald v. Racing Assn. of Central Iowa, 539 U.S. 103, 107–108 (2003); Nordlinger v. Hahn, 505 U.S. 1, 11 (1992); Lehnhausen v. Lake Shore Auto Parts Co., 410 U.S. 356, 359 (1973); Madden v. Kentucky, 309 U.S. 83, 87–88 (1940); Citizens’ Telephone Co. of Grand Rapids v. Fuller, 229 U.S. 322, 329 (1913). Indianapolis’ classification involves neither a “fundamental right” nor a “suspect” classification. Its subject matter is local, economic, social, and commercial. It is a tax classification. And no one here claims that Indianapolis has discriminated against out-of-state commerce or new residents. Cf. Hooper v. Bernalillo County Assessor, 472 U.S. 612 (1985); Williams v. Vermont, 472 U.S. 14 (1985); Metropolitan Life Ins. Co. v. Ward, 470 U.S. 869 (1985); Zobel v. Williams, 457 U.S. 55 (1982). Hence, this case falls directly within the scope of our precedents holding such a law constitutionally valid if “there is a plausible policy reason for the classification, the legislative facts on which the classification is apparently based rationally may have been considered to be true by the governmental decisionmaker, and the relationship of the classification to its goal is not so attenuated as to render the distinction arbitrary or irrational.” Nordlinger, supra, at 11 (citations omitted). And it falls within the scope of our precedents holding that there is such a plausible reason if “there is any reasonably conceivable state of facts that could provide a rational basis for the classification.” FCC v. Beach Communications, Inc., 508 U.S. 307, 313 (1993); see also Lindsley v. Natural Carbonic Gas Co., 220 U.S. 61, 78 (1911). Moreover, analogous precedent warns us that we are not to “pronounc[e]” this classification “unconstitutional unless in the light of the facts made known or generally assumed it is of such a character as to preclude the assumption that it rests upon some rational basis within the knowledge and experience of the legislators.” Carolene Products Co., supra, at 152 (due process claim). Further, because the classification is presumed constitutional, the “ ‘ burden is on the one attacking the legislative arrangement to negative every conceivable basis which might support it.’ ” Heller, supra, at 320 (quoting Lehnhausen, supra, at 364). B In our view, Indianapolis’ classification has a rational basis. Ordinarily, administrative considerations can jus- tify a tax-related distinction. See, e.g., Carmichael v. Southern Coal & Coke Co., 301 U.S. 495, 511–512 (1937) (tax exemption for businesses with fewer than eight employees rational in light of the “[a]dministrative convenience and expense” involved); see also Lehnhausen, supra, at 365 (comparing administrative cost of taxing corporations versus individuals); Madden, supra, at 90 (comparing administrative cost of taxing deposits in local banks versus those elsewhere). And the City’s decision to stop collecting outstanding Barrett Law debts finds rational support in related administrative concerns. The City had decided to switch to the STEP system. After that change, to continue Barrett Law unpaid-debt collection could have proved complex and expensive. It would have meant maintaining an administrative system that for years to come would have had to collect debts arising out of 20-plus different construction projects built over the course of a decade, involving monthly payments as low as $25 per household, with the possible need to maintain credibility by tracking down defaulting debtors and bringing legal action. The City, for example, would have had to maintain its Barrett Law operation within the City Controller’s Office, keep files on old, small, installment-plan debts, and (a City official says) possibly spend hundreds of thousands of dollars keeping computerized debt-tracking systems current. See Brief for International City/County Management Association et al. as Amici Curiae 13, n. 12 (citing Affidavit of Charles White ¶13, Record in Cox, Doc. No. 57–3). Unlike the collection system prior to abandonment, the City would not have added any new Barrett Law installment-plan debtors. And that fact means that it would have had to spread the fixed administrative costs of collection over an ever-declining number of debtors, thereby continuously increasing the per-debtor cost of collection. Consistent with these facts, the Director of the City’s Department of Public Works later explained that the City decided to forgive outstanding debt in part because “[t]he administrative costs to service and process remaining balances on Barrett Law accounts long past the transition to the STEP program would not benefit the taxpayers” and would defeat the purpose of the transition. App. 76. The four other members of the City’s Board of Public Works have said the same. See Affidavit of Gregory Taylor ¶6, Record in Cox, Doc. No. 57–5; Affidavit of Kipper Tew ¶6, ibid. Doc. No. 57–6; Affidavit of Susan Schalk ¶6, ibid. Doc. No. 57–7; Affidavit of Roger Brown ¶6, ibid. Doc. No. 57–8. The rationality of the City’s distinction draws further support from the nature of the line-drawing choices that confronted it. To have added refunds to forgiveness would have meant adding yet further administrative costs, namely the cost of processing refunds. At the same time, to have tried to limit the City’s costs and lost revenues by limiting forgiveness (or refund) rules to Brisbane/Manning homeowners alone would have led those involved in other Barrett Law projects to have justifiably complained about unfairness. Yet to have granted refunds (as well as pro- viding forgiveness) to all those involved in all Barrett Law projects (there were more than 40 projects) or in all open projects (there were more than 20) would have involved even greater administrative burden. The City could not just “cut . . . checks,” post, at 4 (Roberts, C. J., dissenting), without taking funding from other programs or finding additional revenue. If, instead, the City had tried to keep the amount of revenue it lost constant (a rational goal) but spread it evenly among the apparently thousands of homeowners involved in any of the Barrett Laws projects, the result would have been yet smaller individual payments, even more likely to have been too small to justify the administrative expense. Finally, the rationality of the distinction draws support from the fact that the line that the City drew—distinguishing past payments from future obligations—is a line well known to the law. Sometimes such a line takes the form of an amnesty program, involving, say, mortgage payments, taxes, or parking tickets. E.g., 26 U. S. C. §108(a)(1)(E) (2006 ed., Supp. IV) (federal income tax provision allowing homeowners to omit from gross income newly forgiven home mortgage debt); United States v. Martin, metricconverter523 F.3d 281, 284 (CA4 2008) (tax amnesty program whereby State newly forgave penalties and liabilities if taxpayer satisfied debt); Horn v. Chicago, metricconverter860 F.2d 700, 704, n. 9 (CA7 1988) (city parking ticket amnesty program whereby outstanding tickets could be newly set- tled for a fraction of amount specified). This kind of line is consistent with the distinction that the law often makes between actions previously taken and those yet to come. C Petitioners’ contrary arguments are not sufficient to change our conclusion. Petitioners point out that the Indiana Supreme Court also listed a different consideration, namely “financial hardship,” as one of the factors supporting rationality. App. to Pet. for Cert. 19a. They refer to the City’s resolution that said that the Barrett Law “may present financial hardships on many middle to lower income participants who most need sanitary sewer service in lieu of failing septic systems.” App. 71. And they argue that the tax distinction before us would not necessarily favor low-income homeowners. We need not consider this argument, however, for the administrative considerations we have mentioned are sufficient to show a rational basis for the City’s distinction. The Indiana Supreme Court wrote that the City’s classification was “rationally related” in part “to its legitimate interests in reducing its administrative costs.” App. to Pet. for Cert. 19a (emphasis added). The record of the City’s proceedings is consistent with that determination. See App. 72 (when developing transition, the City “considered the current assessments being made by participants in active Barrett Law projects”). In any event, a legislature need not “actually articulate at any time the purpose or rationale supporting its classification.” Nordlinger, 505 U. S., at 15; see also Fitzgerald, 539 U. S., at 108 (similar). Rather, the “burden is on the one attacking the legislative arrangement to negative every conceivable basis which might support it.” Madden, 309 U. S., at 88; see Heller, 509 U. S., at 320 (same); Lehnhausen, 410 U. S., at 364 (same); see also Allied Stores of Ohio, Inc. v. Bowers, 358 U.S. 522, 530 (1959) (upholding state tax classification resting “upon a state of facts that reasonably can be conceived” as creating a rational distinction). Petitioners have not “negative[d]” the Indiana Supreme Court’s first listed justification, namely the administrative concerns we have discussed. Petitioners go on to propose various other forgiveness systems that would have included refunds for at least some of those who had already paid in full. They argue that those systems are superior to the system that the City chose. We have discussed those, and other possible, systems earlier. Supra, at 8–9. Each has advantages and disadvantages. But even if petitioners have found a superior system, the Constitution does not require the City to draw the perfect line nor even to draw a line superior to some other line it might have drawn. It requires only that the line actually drawn be a rational line. And for the reasons we have set forth in Part II–B, supra, we believe that the line the City drew here is rational. Petitioners further argue that administrative considerations alone should not justify a tax distinction, lest a city arbitrarily allocate taxes among a few citizens while forgiving many similarly situated citizens on the ground that it is cheaper and easier to collect taxes from a few people than from many. Brief for Petitioners 45. Petitioners are right that administrative considerations could not justify such an unfair system. But that is not because administrative considerations can never justify tax differences (any more than they can always do so). The question is whether reducing those expenses, in the particular circumstances, provides a rational basis justifying the tax difference in question. In this case, “in the light of the facts made known or generally assumed,” Carolene Products Co., 304 U. S., at 152, it is reasonable to believe that to graft a refund system onto the City’s forgiveness decision could have (for example) imposed an administrative burden of both collecting and paying out small sums (say, $25 per month) for years. As we have said, supra, at 7–9, it is rational for the City to draw a line that avoids that burden. Petitioners, who are the ones “attacking the legislative arrangement,” have the burden of showing that the circumstances are otherwise, i.e., that the administrative burden is too insubstantial to justify the classification. That they have not done. Finally, petitioners point to precedent that in their view makes it more difficult than we have said for the City to show a “rational basis.” With but one exception, however, the cases to which they refer involve discrimination based on residence or length of residence. E.g., Hooper v. Bernalillo County Assessor, 472 U.S. 612 (state tax preference distinguishing between long-term and short-term resident veterans); Williams v. Vermont, 472 U.S. 14 (state use tax that burdened out-of-state car buyers who moved in-state); Metropolitan Life Ins. Co. v. Ward, 470 U.S. 869 (state law that taxed out-of-state insurance companies at a higher rate than in-state companies); Zobel v. Williams, 457 U.S. 55 (state dividend distribution system that favored long-term residents). But those circumstances are not present here. The exception consists of Allegheny Pittsburgh Coal Co. v. Commission of Webster Cty., 488 U.S. 336 (1989). The Court there took into account a state constitution and related laws that required equal valuation of equally valuable property. Id., at 345. It considered the constitutionality of a county tax assessor’s practice (over a period of many years) of determining property values as of the time of the property’s last sale; that practice meant highly unequal valuations for two identical properties that were sold years or decades apart. Id., at 341. The Court first found that the assessor’s practice was not rationally re- lated to the county’s avowed purpose of assessing properties equally at true current value because of the intentional systemic discrepancies the practice created. Id., at 343–344. The Court then noted that, in light of the state constitution and related laws requiring equal valuation, there could be no other rational basis for the practice. Id., at 344–345. Therefore, the Court held, the assessor’s discriminatory policy violated the Federal Constitution’s insistence upon “equal protection of the law.” Id., at 346. Petitioners argue that the City’s refusal to add refunds to its forgiveness decision is similar, for it constitutes a refusal to apply “equally” an Indiana state law that says that the costs of a Barrett Law project shall be equally “apportioned.” Ind. Code §36–9–39–15(b)(3). In other words, petitioners say that even if the City’s decision might otherwise be related to a rational purpose, state law (as in Allegheny) makes this the rare case where the facts preclude any rational basis for the City’s decision other than to comply with the state mandate of equality. Allegheny, however, involved a clear state law requirement clearly and dramatically violated. Indeed, we have described Allegheny as “the rare case where the facts precluded” any alternative reading of state law and thus any alternative rational basis. Nordlinger, 505 U. S., at 16. Here, the City followed state law by apportioning the cost of its Barrett Law projects equally. State law says nothing about forgiveness, how to design a forgiveness program, or whether or when rational distinctions in doing so are permitted. To adopt petitioners’ view would risk transforming ordinary violations of ordinary state tax law into violations of the Federal Constitution. * * * For these reasons, we conclude that the City has not violated the Federal Equal Protection Clause. And the Indiana Supreme Court’s similar determination is Affirmed.

566.US.2011_11-159

Eighteen months after her husband, Robert Capato, died of cancer, respondent Karen Capato gave birth to twins conceived through in vitro fertilization using her husband’s frozen sperm. Karen applied for Social Security survivors benefits for the twins. The Social Security Administration (SSA) denied her application, and the District Court affirmed. In accord with the SSA’s construction of the Social Security Act (Act), the court determined that the twins would qualify for benefits only if, as 42 U. S. C. §416(h)(2)(A) specifies, they could inherit from the deceased wage earner under state intestacy law. The court then found that Robert was domiciled in Florida at his death, and that under Florida law, posthumously conceived children do not qualify for inheritance through intestate succession. The Third Circuit reversed. It concluded that, under §416(e), which defines child to mean, inter alia, “the child or legally adopted child of an [insured] individual,” the undisputed biological children of an insured and his widow qualify for survivors benefits without regard to state intestacy law. Held: The SSA’s reading is better attuned to the statute’s text and its design to benefit primarily those supported by the deceased wage earner in his or her lifetime. Moreover, even if the SSA’s longstanding interpretation is not the only reasonable one, it is at least a permissible construction entitled to deference under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837. Pp. 4–16. (a) Congress amended the Act in 1939 to provide that, as relevant here, “[e]very child (as defined in section 416(e) of this title)” of a deceased insured individual “shall be entitled to a child’s insurance benefit.” §402(d). Section 416(e), in turn, defines “child” to mean: “(1) the child or legally adopted child of an individual, (2) a stepchild [under certain circumstances], and (3) . . . the grandchild or stepgrandchild of an individual or his spouse [under certain conditions].” Unlike §§416(e)(2) and (3), §416(e)(1) lacks any elaboration of the conditions under which a child qualifies for benefits. Section 416(h)(2)(A), however, further addresses the term “child,” providing: “In determining whether an applicant is the child or parent of [an] insured individual for purposes of this subchapter, the Commissioner of Social Security shall apply [the intestacy law of the insured individual’s domiciliary State].” An applicant who does not meet §416(h)(2)(A)’s intestacy-law criterion may nonetheless qualify for benefits under other criteria set forth in §§416(h)(2)(B) and (h)(3), but respondent does not claim eligibility under those other criteria. Regulations promulgated by the SSA closely track §§416(h)(2) and (3) in defining “[w]ho is the insured’s natural child,” 20 CFR §404.355. As the SSA reads the statute, 42 U. S. C. §416(h) governs the meaning of “child” in §416(e)(1) and serves as a gateway through which all applicants for insurance benefits as a “child” must pass. Pp. 4–7. (b) While the SSA regards §416(h) as completing §416(e)’s sparse definition of “child,” the Third Circuit held, and respondent contends, that §416(e) alone governs whenever the claimant is a married couple’s biological child. There are conspicuous flaws in the Third Circuit’s and respondent’s reading; the SSA offers the more persuasive interpretation. Pp. 7–15. (1) Nothing in §416(e)’s tautological definition suggests that Congress understood the word “child” to refer only to the children of married parents. The dictionary definitions offered by respondent are not so confined. Moreover, elsewhere in the Act, Congress expressly limited the category of children covered to offspring of a marital union, see §402(d)(3)(A), and contemporaneous statutes similarly distinguish child of a marriage from the unmodified term “child.” Nor does §416(e) indicate that Congress intended “biological” parentage to be prerequisite to “child” status. A biological parent is not always a child’s parent under law. Furthermore, marriage does not necessarily make a child’s parentage certain, nor does the absence of marriage necessarily make a child’s parentage uncertain. Finally, it is far from obvious that respondent’s proposed definition would cover her posthumously conceived twins, for under Florida law a marriage ends upon the death of a spouse. Pp. 8–10. (2) The SSA finds a key textual cue in §416(h)(2)(A)’s opening instruction: “In determining whether an applicant is the child . . . of [an] insured individual for purposes of this subchapter,” the Commissioner shall apply state intestacy law. Respondent notes the absence of any cross-reference in §416(e) to §416(h), but she overlooks that §416(h) provides the crucial link: It requires reference to state intestacy law to determine child status not just for §416(h) purposes, but “for purposes of this subchapter,” which includes both §§402(d) and 416(e). Having explicitly complemented §416(e) by the definitional provisions contained in §416(h), Congress had no need to place a redundant cross-reference in §416(e). The Act commonly refers to state law on matters of family status, including an applicant’s status as a wife, widow, husband, or widower. See, e.g., §§416(b), (h)(1)(A). The Act also sets duration-of-relationship limitations, see Weinberger v. Salfi, 422 U.S. 749, 777–782, and time limits qualify the statutes of several States that accord inheritance rights to posthumously conceived children. In contrast, no time constraint attends the Third Circuit’s ruling in this case, under which the biological child of married parents is eligible for survivors benefits, no matter the length of time between the father’s death and the child’s conception and birth. Because a child who may take from a father’s estate is more likely to “be dependent during the parent’s life and at his death,” Mathews v. Lucas, 427 U.S. 495, 514, reliance on state intestacy law to determine who is a “child” serves the Act’s driving objective, which is to “provide . . . dependent members of [a wage earner’s] family with protection against the hardship occasioned by [the] loss of [the insured’s] earnings,” Califano v. Jobst, 434 U.S. 47, 52. Although the Act and regulations set different eligibility requirements for adopted children, stepchildren, grandchildren, and stepgrandchildren, it hardly follows, as respondent argues, that applicants in those categories are treated more advantageously than are children who must meet a §416(h) criterion. Respondent charges that the SSA’s construction of the Act raises serious constitutional concerns under the equal protection component of the Due Process Clause. But under rational-basis review, the appropriate standard here, the regime passed by Congress easily passes inspection. Pp. 10–15. (c) Because the SSA’s interpretation of the relevant provisions, is at least reasonable, the agency’s reading is entitled to this Court’s deference under Chevron, 467 U.S. 837. Chevron deference is appropriate “when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law, and that the agency interpretation claiming deference was promulgated in the exercise of that authority.” United States v. Mead Corp., 533 U.S. 218, 226–227. Here, the SSA’s longstanding interpretation, set forth in regulations published after notice-and-comment rulemaking, is neither “arbitrary or capricious in substance, [n]or manifestly contrary to the statute.” Mayo Foundation for Medical Ed. and Research v. United States, 562 U. S. ___, ___. It therefore warrants the Court’s approbation. Pp. 15–16. 631 F.3d 626, reversed and remanded. Ginsburg, J., delivered the opinion for a unanimous Court.

Karen and Robert Capato married in 1999. Robert died of cancer less than three years later. With the help of in vitro fertilization, Karen gave birth to twins 18 months after her husband’s death. Karen’s application for Social Security survivors benefits for the twins, which the Social Security Administration (SSA) denied, prompted this lit-igation. The technology that made the twins’ conception and birth possible, it is safe to say, was not contemplated by Congress when the relevant provisions of the Social Security Act (Act) originated (1939) or were amended to read as they now do (1965). Karen Capato, respondent here, relies on the Act’s initial definition of “child” in 42 U. S. C. §416(e): “ ‘[C]hild’ means . . . the child or legally adopted child of an [insured] individual.” Robert was an insured individual, and the twins, it is uncontested, are the biological children of Karen and Robert. That satisfies the Act’s terms, and no further inquiry is in order, Karen maintains. The SSA, however, identifies subsequent provisions, §§416(h)(2) and (h)(3)(C), as critical, and reads them to entitle biological children to benefits only if they qualify for inheritance from the decedent under state intestacy law, or satisfy one of the statutory alternatives to that requirement. We conclude that the SSA’s reading is better attuned to the statute’s text and its design to benefit primarily those supported by the deceased wage earner in his or her lifetime. And even if the SSA’s longstanding interpretation is not the only reasonable one, it is at least a permissible construction that garners the Court’s respect under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). I Karen Capato married Robert Capato in May 1999. Shortly thereafter, Robert was diagnosed with esophageal cancer and was told that the chemotherapy he required might render him sterile. Because the couple wanted children, Robert, before undergoing chemotherapy, deposited his semen in a sperm bank, where it was frozen and stored. Despite Robert’s aggressive treatment regime, Karen conceived naturally and gave birth to a son in August 2001. The Capatos, however, wanted their son to have a sibling. Robert’s health deteriorated in late 2001, and he died in Florida, where he and Karen then resided, in March 2002. His will, executed in Florida, named as beneficiaries the son born of his marriage to Karen and two children from a previous marriage. The will made no provision for children conceived after Robert’s death, although the Capatos had told their lawyer they wanted future offspring to be placed on a par with existing children. Shortly after Robert’s death, Karen began in vitro fertilization using her husband’s frozen sperm. She conceived in January 2003 and gave birth to twins in September 2003, 18 months after Robert’s death. Karen Capato claimed survivors insurance benefits on behalf of the twins. The SSA denied her application, and the U. S. District Court for the District of New Jersey affirmed the agency’s decision. See App. to Pet. for Cert. 33a (decision of the Administrative Law Judge); id., at 15a (District Court opinion). In accord with the SSA’s construction of the statute, the District Court determined that the twins would qualify for benefits only if, as §416(h)(2)(A) specifies, they could inherit from the deceased wage earner under state intestacy law. Robert Capato died domiciled in Florida, the court found. Under that State’s law, the court noted, a child born posthu-mously may inherit through intestate succession only if conceived during the decedent’s lifetime. Id., at 27a–28a.[1] The Court of Appeals for the Third Circuit reversed. Under §416(e), the appellate court concluded, “the undisputed biological children of a deceased wage earner and his widow” qualify for survivors benefits without regard to state intestacy law. 631 F.3d 626, 631 (2011).[2] Courts of Appeals have divided on the statutory interpretation question this case presents. Compare ibid. and Gillett-Netting v. Barnhart, 371 F.3d 593, 596–597 (CA9 2004) (biological but posthumously conceived child of insured wage earner and his widow qualifies for benefits), with Beeler v. Astrue, 651 F.3d 954, 960–964 (CA8 2011), and Schafer v. Astrue, 641 F.3d 49, 54–63 (CA4 2011) (posthumously conceived child’s qualification for benefits depends on intestacy law of State in which wage earner was domiciled). To resolve the conflict, we granted the Commissioner’s petition for a writ of certiorari. 565 U. S. ___ (2011). II Congress amended the Social Security Act in 1939 to provide a monthly benefit for designated surviving fam- ily members of a deceased insured wage earner. “Child’s insurance benefits” are among the Act’s family-protective measures. 53Stat. 1364, as amended, 42 U. S. C. §402(d). An applicant qualifies for such benefits if she meets the Act’s definition of “child,” is unmarried, is below specified age limits (18 or 19) or is under a disability which began prior to age 22, and was dependent on the insured at the time of the insured’s death. §402(d)(1).[3] To resolve this case, we must decide whether the Capato twins rank as “child[ren]” under the Act’s definitional provisions. Section 402(d) provides that “[e]very child (as defined in section 416(e) of this title)” of a deceased insured individual “shall be entitled to a child’s insurance benefit.” Section 416(e), in turn, states: “The term ‘child’ means (1) the child or legally adopted child of an individ-ual, (2) a stepchild [under certain circumstances], and (3) . . . the grandchild or stepgrandchild of an individual or his spouse [who meets certain conditions].” The word “child,” we note, appears twice in §416(e)’s opening sentence: initially in the prefatory phrase, “[t]he term ‘child’ means . . . ,” and, immediately thereafter, in subsection (e)(1) (“child or legally adopted child”), deline-ating the first of three beneficiary categories. Unlike §§416(e)(2) and (3), which specify the circumstances under which stepchildren and grandchildren qualify for benefits, §416(e)(1) lacks any elaboration. Compare §416(e)(1) (referring simply to “the child . . . of an individual”) with, e.g., §416(e)(2) (applicant must have been a stepchild for at least nine months before the insured individual’s death). A subsequent definitional provision further addresses the term “child.” Under the heading “Determination of family status,” §416(h)(2)(A) provides: “In determining whether an applicant is the child or parent of [an] insured individual for purposes of this subchapter, the Commissioner of Social Security shall apply [the intestacy law of the insured individual’s domiciliary State].”[4] An applicant for child benefits who does not meet §416(h)(2)(A)’s intestacy-law criterion may nonetheless qualify for benefits under one of several other criteria the Act prescribes. First, an applicant who “is a son or daughter” of an insured individual, but is not determined to be a “child” under the intestacy-law provision, nevertheless ranks as a “child” if the insured and the other parent went through a marriage ceremony that would have been valid but for certain legal impediments. §416(h)(2)(B). Further, an applicant is deemed a “child” if, before death, the insured acknowledged in writing that the applicant is his or her son or daughter, or if the insured had been decreed by a court to be the father or mother of the applicant, or had been ordered to pay child support. §416(h)(3)(C)(i). In addition, an applicant may gain “child” status upon proof that the insured individual was the applicant’s pa- rent and “was living with or contributing to the support of the applicant” when the insured individual died. §416(h)(3)(C)(ii).[5] The SSA has interpreted these provisions in regulations adopted through notice-and-comment rulemaking. The regulations state that an applicant may be entitled to benefits “as a natural child, legally adopted child, stepchild, grandchild, stepgrandchild, or equitably adopted child.” 20 CFR §404.354. Defining “[w]ho is the insured’s natural child,” §404.355, the regulations closely track 42 U. S. C. §§416(h)(2) and (h)(3). They state that an applicant may qualify for insurance benefits as a “natural child” by meeting any of four conditions: (1) the applicant “could inherit the insured’s personal property as his or her natural child under State inheritance laws”; (2) the applicant is “the insured’s natural child and [his or her parents] went through a ceremony which would have resulted in a valid marriage between them except for a legal impediment”; (3) before death, the insured acknowledged in writing his or her parentage of the applicant, was decreed by a court to be the applicant’s parent, or was ordered by a court to contribute to the applicant’s support; or (4) other evidence shows that the insured is the applicant’s “natural father or mother” and was either living with, or contributing to the support of, the applicant. 20 CFR §404.355(a) (internal quotation marks omitted). As the SSA reads the statute, 42 U. S. C. §416(h) governs the meaning of “child” in §416(e)(1). In other words, §416(h) is a gateway through which all applicants for in-surance benefits as a “child” must pass. See Beeler, 651 F. 3d, at 960 (“The regulations make clear that the SSA interprets the Act to mean that the provisions of §416(h) are the exclusive means by which an applicant can establish ‘child’ status under §416(e) as a natural child.”).[6] III Karen Capato argues, and the Third Circuit held, that §416(h), far from supplying the governing law, is irrelevant in this case. Instead, the Court of Appeals determined, §416(e) alone is dispositive of the controversy. 631 F. 3d, at 630–631. Under §416(e), “child” means “child of an [insured] individual,” and the Capato twins, the Third Circuit observed, clearly fit that definition: They are undeniably the children of Robert Capato, the insured wage earner, and his widow, Karen Capato. Section 416(h) comes into play, the court reasoned, only when “a claimant’s status as a deceased wage-earner’s child is in doubt.” Id., at 631. That limitation, the court suggested, is evident from §416(h)’s caption: “Determination of family status.” Here, “there is no family status to determine,” the court said, id., at 630, so §416(h) has no role to play. In short, while the SSA regards §416(h) as completing §416(e)’s sparse definition of “child,” the Third Circuit considered each subsection to control different situations: §416(h) governs when a child’s family status needs to be determined; §416(e), when it does not. When is there no need to determine a child’s family status? The answer that the Third Circuit found plain: whenever the claimant is “the biological child of a married couple.” Id., at 630.[7] We point out, first, some conspicuous flaws in the Third Circuit’s and respondent Karen Capato’s reading of the Act’s provisions, and then explain why we find the SSA’s interpretation persuasive. A Nothing in §416(e)’s tautological definition (“ ‘child’ means . . . the child . . . of an individual”) suggests that Congress understood the word “child” to refer only to the children of married parents. The dictionary definitions offered by respondent are not so confined. See Webster’s New International Dictionary 465 (2d ed. 1934) (defining “child” as, inter alia, “[i]n Law, legitimate offspring; also, sometimes, esp. in wills, an adopted child, or an illegitimate offspring, or any direct descendant, as a grandchild, as the intention may appear”); Merriam-Webster’s Collegiate Dictionary 214 (11th ed. 2003) (“child” means “son or daughter,” or “descendant”). See also Restatement (Third) of Property §2.5(1) (1998) (“[a]n individual is the child of his or her genetic parents,” and that may be so “whether or not [the parents] are married to each other”). More-over, elsewhere in the Act, Congress expressly limited the category of children covered to offspring of a marital union. See §402(d)(3)(A) (referring to the “legitimate . . . child” of an individual). Other contemporaneous statutes similarly differentiate child of a marriage (“legitimate child”) from the unmodified term “child.” See, e.g., Servicemen’s Dependents Allowance Act of 1942, ch. 443, §120, 56Stat. 385 (defining “child” to include “legitimate child,” “child legally adopted,” and, under certain conditions, “stepchild” and “illegitimate child” (internal quotation marks omitted)). Nor does §416(e) indicate that Congress intended “biological” parentage to be prerequisite to “child” status under that provision. As the SSA points out, “[i]n 1939, there was no such thing as a scientifically proven biological relationship between a child and a father, which is . . . part of the reason that the word ‘biological’ appears nowhere in the Act.” Reply Brief 6. Notably, a biological parent is not necessarily a child’s parent under law. Ordinarily, “a parent-child relationship does not exist between an adoptee and the adoptee’s genetic parents.” Uniform Probate Code §2–119(a), 8 U. L. A. 55 (Supp. 2011) (amended 2008). Moreover, laws directly addressing use of today’s assisted reproduction technology do not make biological parentage a universally determinative criterion. See, e.g., Cal. Fam. Code Ann. §7613(b) (West Supp. 2012) (“The donor of semen . . . for use in artificial insemination or in vitro fertilization of a woman other than the donor’s wife is treated in law as if he were not the natural father of a child thereby conceived, unless otherwise agreed to in a writing signed by the donor and the woman prior to the conception of the child.”); Mass. Gen. Laws, ch. 46, §4B (West 2010) (“Any child born to a married woman as a result of artificial insemination with the consent of her husband, shall be considered the legitimate child of the mother and such husband.”). We note, in addition, that marriage does not ever and always make the parentage of a child certain, nor does the absence of marriage necessarily mean that a child’s parentage is uncertain. An unmarried couple can agree that a child is theirs, while the parentage of a child born during a marriage may be uncertain. See Reply Brief 11 (“Respondent errs in treating ‘marital’ and ‘undisputed’ as having the same meaning.”). Finally, it is far from obvious that Karen Capato’s proposed definition—“biological child of married parents,” see Brief for Respondent 9—would cover the posthumously conceived Capato twins. Under Florida law, a marriage ends upon the death of a spouse. See Price v. Price, 114 Fla. 233, 235, 153 So. 904, 905 (1934). If that law applies, rather than a court-declared preemptive federal law, the Capato twins, conceived after the death of their father, would not qualify as “marital” children.[8] B Resisting the importation of words not found in §416(e)—“child” means “the biological child of married parents,” Brief for Respondent 9—the SSA finds a key textual cue in §416(h)(2)(A)’s opening instruction: “In determining whether an applicant is the child . . . of [an] insured individual for purposes of this subchapter,” the Commissioner shall apply state intestacy law. (Emphasis added.) Respondent notes the absence of any cross-reference in §416(e) to §416(h). Brief for Respondent 18. She overlooks, however, that §416(h) provides the crucial link. The “subchapter” to which §416(h) refers is Subchapter II of the Act, which spans §§401 through 434. Section 416(h)’s reference to “this subchapter” thus includes both §§402(d) and 416(e). Having explicitly complemented §416(e) by the definitional provisions contained in §416(h), Congress had no need to place a redundant cross-reference in §416(e). See Schafer, 641 F. 3d, at 54 (Congress, in §416(h)(2)(A), provided “plain and explicit instruction on how the determination of child status should be made”; on this point, the statute’s text “could hardly be more clear.”). The original version of today’s §416(h) was similarly drafted. It provided that, “[i]n determining whether an applicant is the . . . child . . . of [an] insured individual for purposes of sections 401–409 of this title, the Board shall apply [state intestacy law].” 42 U. S. C. §409(m) (1940 ed.) (emphasis added). Sections 401–409 embraced §§402(c) and 409(k), the statutory predecessors of 42 U. S. C. §§402(d) and 416(e) (2006 ed.), respectively. Reference to state law to determine an applicant’s status as a “child” is anything but anomalous. Quite the opposite. The Act commonly refers to state law on matters of family status. For example, the Act initially defines “wife” as “the wife of an [insured] individual,” if certain conditions are satisfied. §416(b). Like §416(e), §416(b) is, at least in part, tautological (“ ‘wife’ means the [insured’s] wife”). One must read on, although there is no ex-press cross-reference, to §416(h) (rules on “[d]etermination of family status”) to complete the definition. Section §416(h)(1)(A) directs that, “for purposes of this subchapter,” the law of the insured’s domicile determines whether “[the] applicant and [the] insured individual were validly married,” and if they were not, whether the applicant would nevertheless have “the same status” as a wife under the State’s intestacy law. (Emphasis added.) The Act similarly defines the terms “widow,” “husband,” and “widower.” See §§416(c), (f), (g), (h)(1)(A). Indeed, as originally enacted, a single provision mandated the use of state intestacy law for “determining whether an applicant is the wife, widow, child, or parent of [an] insured individual.” 42 U. S. C. §409(m) (1940 ed.). All wife, widow, child, and parent applicants thus had to satisfy the same criterion. To be sure, children born during their parents’ marriage would have readily qualified under the 1939 formulation because of their eligibility to inherit under state law. But requiring all “child” ap-plicants to qualify under state intestacy law installed a simple test, one that ensured benefits for persons plainly within the legislators’ contemplation, while avoiding con-gressional entanglement in the traditional state-law realm of family relations. Just as the Act generally refers to state law to determine whether an applicant qualifies as a wife, widow, husband, widower, 42 U. S. C. §416(h)(1) (2006 ed.), child or parent, §416(h)(2)(A), so in several sections (§§416(b), (c), (e)(2), (f), (g)), the Act sets duration-of-relationship limitations. See Weinberger v. Salfi, 422 U.S. 749, 777–782 (1975) (discussing §416(e)(2)’s requirement that, as a check against deathbed marriages, a parent-stepchild relationship must exist “not less than nine months immediately preceding [insured’s death]”). Time limits also qualify the statutes of several States that accord inheritance rights to posthumously conceived children. See Cal. Prob. Code Ann. §249.5(c) (West Supp. 2012) (allowing inheritance if child is in utero within two years of parent’s death); Colo. Rev. Stat. Ann. §15–11–120(11) (2011) (child in utero within three years or born within 45 months); Iowa Code Ann. §633.220A(1) (West Supp. 2012) (child born within two years); La. Rev. Stat. Ann. §9:391.1(A) (West 2008) (child born within three years); N. D. Cent. Code Ann. §30.1–04–19(11) (Lexis 2001) (child in utero within three years or born within 45 months). See also Uniform Probate Code §2–120(k), 8 U. L. A. 58 (Supp. 2011) (treating a posthumously conceived child as “in gestation at the individual’s death,” but only if specified time limits are met). No time constraints attend the Third Circuit’s ruling in this case, under which the biological child of married parents is eligible for survivors benefits, no matter the length of time between the father’s death and the child’s conception and birth. See Tr. of Oral Arg. 36–37 (counsel for Karen Capato acknowledged that, under the preemptive federal rule he advocated, and the Third Circuit adopted, a child born four years after her father’s death would be eligible for benefits). The paths to receipt of benefits laid out in the Act and regulations, we must not forget, proceed from Congress’ perception of the core purpose of the legislation. The aim was not to create a program “generally benefiting needy persons”; it was, more particularly, to “provide . . . dependent members of [a wage earner’s] family with protection against the hardship occasioned by [the] loss of [the insured’s] earnings.” Califano v. Jobst, 434 U.S. 47, 52 (1977). We have recognized that “where state intestacy law provides that a child may take personal property from a father’s estate, it may reasonably be thought that the child will more likely be dependent during the parent’s life and at his death.” Mathews v. Lucas, 427 U.S. 495, 514 (1976). Reliance on state intestacy law to determine who is a “child” thus serves the Act’s driving objective. True, the intestacy criterion yields benefits to some children outside the Act’s central concern. Intestacy laws in a number of States, as just noted, do provide for inheritance by posthumously conceived children, see supra, at 12,[9] and under federal law, a child conceived shortly before her father’s death may be eligible for benefits even though she never actually received her father’s support. It was nonetheless Congress’ prerogative to legislate for the generality of cases. It did so here by employing eligibility to inherit under state intestacy law as a workable substitute for bur-densome case-by-case determinations whether the child was, in fact, dependent on her father’s earnings. Respondent argues that on the SSA’s reading, natural children alone must pass through a §416(h) gateway. Adopted children, stepchildren, grandchildren, and step-grandchildren, it is true, are defined in §416(e), and are not further defined in §416(h). Respondent overlooks, however, that although not touched by §416(h), beneficiaries described in §§416(e)(2) and (e)(3) must meet other statutorily prescribed criteria. In short, the Act and regulations set different eligibility requirements for adopted children, stepchildren, grandchildren, and stepgrandchildren, see 20 CFR §§404.356–404.358, but it hardly follows that applicants in those categories are treated more advantageously than are children who must meet a §416(h) criterion. The SSA’s construction of the Act, respondent charges, raises serious constitutional concerns under the equal pro-tection component of the Due Process Clause. Brief for Respondent 42; see Weinberger v. Wiesenfeld, 420 U.S. 636, 638, n. 2 (1975). She alleges: “Under the government’s interpretation . . . , posthumously conceived children are treated as an inferior subset of natural children who are ineligible for government benefits simply because of their date of birth and method of conception.” Brief for Respondent 42–43. Even the Courts of Appeals that have accepted the reading of the Act respondent advances have rejected this argument. See 631 F. 3d, at 628, n. 1 (citing Vernoff v. Astrue, 568 F.3d 1102, 1112 (CA9 2009)). We have applied an intermediate level of scrutiny to laws “burden[ing] illegitimate children for the sake of punishing the illicit relations of their parents, because ‘visiting this condemnation on the head of an infant is illogical and unjust.’ ” Clark v. Jeter, 486 U.S. 456, 461 (1988) (quoting Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 175 (1972)). No showing has been made that posthumously conceived children share the characteristics that prompted our skepticism of classifications disadvantaging children of unwed parents. We therefore need not decide whether heightened scrutiny would be appropriate were that the case.[10] Under rational-basis review, the regime Congress adopted easily passes inspection. As the Ninth Circuit held, that regime is “reasonably related to the government’s twin interests in [reserving] benefits [for] those children who have lost a parent’s support, and in using reasonable presumptions to minimize the administrative burden of proving dependency on a case-by-case basis.” Vernoff, 568 F. 3d, at 1112 (citing Mathews, 427 U. S., at 509). IV As we have explained, §416(e)(1)’s statement, “[t]he term ‘child’ means . . . the child . . . of an individual,” is a definition of scant utility without aid from neighboring provisions. See Schafer, 641 F. 3d, at 54. That aid is supplied by §416(h)(2)(A), which completes the definition of “child” “for purposes of th[e] subchapter” that includes §416(e)(1). Under the completed definition, which the SSA employs, §416(h)(2)(A) refers to state law to determine the status of a posthumously conceived child. The SSA’s interpretation of the relevant provisions, adhered to without deviation for many decades, is at least reasonable; the agency’s reading is therefore entitled to this Court’s deference under Chevron, 467 U.S. 837. Chevron deference is appropriate “when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law, and that the agency interpretation claiming deference was promulgated in the exercise of that authority.” United States v. Mead Corp., 533 U.S. 218, 226–227 (2001). Here, as already noted, the SSA’s longstanding interpretation is set forth in regu-lations published after notice-and-comment rulemaking. See supra, at 6–7. Congress gave the Commissioner authority to promulgate rules “necessary or appropriate to carry out” the Commissioner’s functions and the relevant statutory provisions. See 42 U. S. C. §§405(a), 902(a)(5). The Commissioner’s regulations are neither “arbitrary or capricious in substance, [n]or manifestly contrary to the statute.” Mayo Foundation for Medical Ed. and Research v. United States, 562 U. S. ___, ___ (2011) (slip op., at 7) (internal quotation marks omitted). They thus warrant the Court’s approbation. See Barnhart v. Walton, 535 U.S. 212, 217–222, 225 (2002) (deferring to the Commissioner’s “considerable authority” to interpret the Social Security Act). V Tragic circumstances—Robert Capato’s death before he and his wife could raise a family—gave rise to this case. But the law Congress enacted calls for resolution of Karen Capato’s application for child’s insurance benefits by reference to state intestacy law. We cannot replace that reference by creating a uniform federal rule the statute’s text scarcely supports. * * * For the reasons stated, the judgment of the Court of Appeals for the Third Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 The District Court observed that Fla. Stat. Ann. §732.106 (West 2010) defines “ ‘afterborn heirs’ ” as “ ‘heirs of the decedent conceived before his or her death, but born thereafter.’ ” App. to Pet. for Cert.27a (emphasis added by District Court). The court also referred to §742.17(4), which provides that a posthumously conceived child “ ‘shall not be eligible for a claim against the decedent’s estate unless the child has been provided for by the decedent’s will.’ ” Id., at 28a. 2 Because the Third Circuit held that posthumously conceived children qualify for survivors benefits as a matter of federal law, it did not definitively determine “where [Robert] Capato was domiciled at his death or . . . delve into the law of intestacy of that state.” 631 F. 3d, at 632, n. 6. These issues, if preserved, may be considered on remand. 3 Applicants not in fact dependent on the insured individual may be “deemed dependent” when the Act so provides. For example, a “legitimate” child, even if she is not living with or receiving support fromher parent, is ordinarily “deemed dependent” on that parent. 42 U. S. C. §402(d)(3). Further, applicants “deemed” the child of an insured individual under §416(h)(2)(B) or (h)(3) are also “deemed legitimate,” hence dependent, even if not living with or receiving support fromthe parent. §402(d)(3). See also Mathews v. Lucas, 427 U.S. 495, 499, n. 2 (1976) (deeming dependent any child who qualifies under §416(h)(2)(A)); Tr. of Oral Arg. 13–14 (counsel for the SSA stated, in response to the Court’s question, that statutory presumptions of dependency are irrebuttable). 4 Section 416(h)(2)(A) also states that persons who, under the law of the insured’s domicile, “would have the same status relative to taking intestate personal property as a child or parent shall be deemed such.” Asked about this prescription, counsel for the SSA responded that it would apply to equitably adopted children. Tr. of Oral Arg. 8–9, 54; see 20 CFR §404.359 (2011) (an equitably adopted child may be eligible for benefits if the agreement to adopt the child would be recognized under state law as enabling the child to inherit upon the intestate death of the adopting parent). 5 Respondent does not invoke any of the alternative criteria as a basis for the twins’ “child” status. 6 The Commissioner of Social Security has acquiesced in the Ninth Circuit’s conflicting interpretation for cases arising in that Circuit. See Social Security Acquiescence Ruling 05–1(9), 70 Fed. Reg. 55656 (2005). 7 Because the Court of Appeals found the statutory language unambiguous, it had no occasion to “determine whether the [SSA’s] interpretation is a permissible construction of the statute.” 631 F. 3d, at 631, n. 5 (citing Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842–843 (1984)). 8 Respondent urges that it would be bizarre to deny benefits to the Capato twins when, under §416(h)(2)(B), they would have gained benefits had their parents gone through a marriage ceremony that would have been valid save for a legal impediment. Brief for Respondent 26, n. 10; see supra, at 5–6. Whether the Capatos’ marriage ceremony was flawed or flawless, the SSA counters, no marital union was extant when the twins were conceived. Reply Brief 11. 9 But see N. Y. Est., Powers & Trusts Law Ann. §4–1.1(c) (West 1998) (“Distributees of the decedent, conceived before his or her death but born alive thereafter, take as if they were born in his or her lifetime.”). Similar provisions are contained in Ga. Code Ann. §53–2–1(b)(1) (2011), Idaho Code §15–2–108 (Lexis 2009), Minn. Stat. Ann. §524.2–120(10) (West Supp. 2012), S. C. Code Ann. §62–2–108 (2009), and S. D. Codified Laws §29A–2–108 (Supp. 2011). 10 Ironically, while drawing an analogy to the “illogical and unjust” discrimination children born out of wedlock encounter, see Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 175–176 (1972), respondent asks us to differentiate between children whose parents were married and children whose parents’ liaisons were not blessed by clergy or the State. She would eliminate the intestacy test only for biological children of married parents.

566.US.2011_10-1320

The State of Arkansas charged petitioner Alex Blueford with capital murder for the death of a one-year-old child. That charge included the lesser offenses of first-degree murder, manslaughter, and negligent homicide. Before the start of deliberations, the trial court instructed the jury to consider the offenses as follows: “If you have a reasonable doubt of the defendant’s guilt on the charge of capital murder, you will consider the charge of murder in the first degree. . . . If you have a reasonable doubt of the defendant’s guilt on the charge of murder in the first degree, you will then consider the charge of manslaughter. . . . If you have a reasonable doubt of the defendant’s guilt on the charge of manslaughter, you will then consider the charge of negligent homicide.” The court also presented the jury with a set of verdict forms, which allowed the jury either to convict Blueford of one of the charged offenses, or to acquit him of all of them. Acquitting on some but not others was not an option. After deliberating for a few hours, the jury reported that it could not reach a verdict. The court inquired about the jury’s progress on each offense. The foreperson disclosed that the jury was unanimous against guilt on the charges of capital murder and first-degree murder, was deadlocked on manslaughter, and had not voted on negligent homicide. The court told the jury to continue to deliberate. The jury did so but still could not reach a verdict, and the court declared a mistrial. When the State subsequently sought to retry Blueford, he moved to dismiss the capital and first-degree murder charges on double jeopardy grounds. The trial court denied the motion, and the Supreme Court of Arkansas affirmed on interlocutory appeal. Held: The Double Jeopardy Clause does not bar retrying Blueford on charges of capital murder and first-degree murder. Pp. 5−10. (a) The jury did not acquit Blueford of capital or first-degree murder. Blueford contends that the foreperson’s report that the jury was unanimous against guilt on the murder offenses represented a resolution of some or all of the elements of those offenses in his favor. But the report was not a final resolution of anything. When the foreperson told the court how the jury had voted on each offense, the jury’s deliberations had not yet concluded. The jurors in fact went back to the jury room to deliberate further, and nothing in the court’s instructions prohibited them from reconsidering their votes on capital and first-degree murder as deliberations continued. The foreperson’s report prior to the end of deliberations therefore lacked the finality necessary to amount to an acquittal on those offenses. That same lack of finality undermines Blueford’s reliance on Green v. United States, 355 U.S. 184, and Price v. Georgia, 398 U.S. 323. In both of those cases, the verdict of the jury was a final decision; here, the report of the foreperson was not. Pp. 5−8. (b) The trial court’s declaration of a mistrial was not improper. A trial can be discontinued without barring a subsequent one for the same offense when “particular circumstances manifest a necessity” to declare a mistrial. Wade v. Hunter, 336 U.S. 684, 690. Blueford contends that there was no necessity for a mistrial on capital and first-degree murder, given the foreperson’s report that the jury had voted unanimously against guilt on those charges. According to Blueford, the court at that time should have taken some action, whether through new partial verdict forms or other means, to allow the jury to give effect to those votes, and then considered a mistrial only as to the remaining charges. Blueford acknowledges, however, that the trial court’s reason for declaring a mistrial here—that the jury was unable to reach a verdict—has long been considered the “classic basis” establishing necessity for doing so. Arizona v. Washington, 434 U.S. 497, 509. And this Court has never required a trial court, before declaring a mistrial because of a hung jury, to consider any particular means of breaking the impasse―let alone to consider giving the jury new options for a verdict. See Renico v. Lett, 559 U. S. ___, ___. As permitted under Arkansas law, the jury’s options in this case were limited to two: either convict on one of the offenses, or acquit on all. The trial court did not abuse its discretion by refusing to add another option—that of acquitting on some offenses but not others. Pp. 9−10. 2011 Ark. 8, ___ S. W. 3d ___, affirmed. Roberts, C. J., delivered the opinion of the Court, in which Scalia, Kennedy, Thomas, Breyer, and Alito, JJ., joined. Sotomayor, J., filed a dissenting opinion, in which Ginsburg and Kagan, JJ., joined.

The Double Jeopardy Clause protects against being tried twice for the same offense. The Clause does not, however, bar a second trial if the first ended in a mistrial. Before the jury concluded deliberations in this case, it re- ported that it was unanimous against guilt on charges of capital murder and first-degree murder, was deadlocked on manslaughter, and had not voted on negligent homicide. The court told the jury to continue to deliberate. The jury did so but still could not reach a verdict, and the court declared a mistrial. All agree that the defendant may be retried on charges of manslaughter and negligent homicide. The question is whether he may also be retried on charges of capital and first-degree murder. I One-year-old Matthew McFadden, Jr., suffered a severe head injury on November 28, 2007, while home with his mother’s boyfriend, Alex Blueford. Despite treatment at a hospital, McFadden died a few days later. The State of Arkansas charged Blueford with capital murder, but waived the death penalty. The State’s theory at trial was that Blueford had injured McFadden intentionally, causing the boy’s death “[u]nder circumstances manifesting extreme indifference to the value of human life.” Ark. Code Ann. §5–10–101(a)(9)(A) (Supp. 2011). The defense, in contrast, portrayed the death as the re- sult of Blueford accidentally knocking McFadden onto the ground. The trial court instructed the jury that the charge of capital murder included three lesser offenses: first-degree murder, manslaughter, and negligent homicide. In addition to describing these offenses, the court addressed the order in which the jury was to consider them: “If you have a reasonable doubt of the defendant’s guilt on the charge of capital murder, you will consider the charge of murder in the first degree. . . . If you have a reasonable doubt of the defendant’s guilt on the charge of murder in the first degree, you will then consider the charge of manslaughter. . . . If you have a reasonable doubt of the defendant’s guilt on the charge of manslaughter, you will then con- sider the charge of negligent homicide.” App. 51–52. The prosecution commented on these instructions in its closing argument. It told the jury, for example, that “before you can consider a lesser included of capital murder, you must first, all 12, vote that this man is not guilty of capital murder.” Id., at 55. The prosecution explained that this was “not a situation where you just lay everything out here and say, well, we have four choices. Which one does it fit the most?” Id., at 59. Rather, the prose- cution emphasized, “unless all 12 of you agree that this man’s actions were not consistent with capital murder, then and only then would you go down to murder in the first degree.” Ibid. After the parties concluded their arguments, the court presented the jury with a set of five verdict forms, each representing a possible verdict. There were four separate forms allowing the jury to convict on each of the charged offenses: capital murder, first-degree murder, manslaughter, and negligent homicide. A fifth form allowed the jury to return a verdict of acquittal, if the jury found Blueford not guilty of any offense. There was no form allowing the jury to acquit on some offenses but not others. As stated in the court’s instructions, the jury could either “find the defendant guilty of one of these offenses” or “acquit him outright.” Id., at 51. Any verdict—whether to convict on one or to acquit on all—had to be unanimous. A few hours after beginning its deliberations, the jury sent the court a note asking “what happens if we cannot agree on a charge at all.” Id., at 62. The court called the jury back into the courtroom and issued a so-called “Allen instruction,” emphasizing the importance of reaching a ver- dict. See Allen v. United States, 164 U.S. 492, 501–502 (1896). The jury then deliberated for a half hour more before sending out a second note, stating that it “cannot agree on any one charge in this case.” App. 64. When the court summoned the jury again, the jury foreperson reported that the jury was “hopelessly” deadlocked. Ibid. The court asked the foreperson to disclose the jury’s votes on each offense: “THE COURT: All right. If you have your numbers together, and I don’t want names, but if you have your numbers I would like to know what your count was on capital murder. “JUROR NUMBER ONE: That was unanimous against that. No. “THE COURT: Okay, on murder in the first degree? “JUROR NUMBER ONE: That was unanimous against that. “THE COURT: Okay. Manslaughter? “JUROR NUMBER ONE: Nine for, three against. “THE COURT: Okay. And negligent homicide? “JUROR NUMBER ONE: We did not vote on that, sir. “THE COURT: Did not vote on that. “JUROR NUMBER ONE: No, sir. We couldn’t get past the manslaughter. Were we supposed to go past that? I thought we were supposed to go one at a time.” Id., at 64–65. Following this exchange, the court gave another Allen instruction and sent the jurors back to the jury room. After deliberations resumed, Blueford’s counsel asked the court to submit new verdict forms to the jurors, to be completed “for those counts that they have reached a verdict on.” Id., at 67. The prosecution objected on the grounds that the jury was “still deliberating” and that a verdict of acquittal had to be “all or nothing.” Id., at 68. The court denied Blueford’s request. To allow for a partial verdict, the court explained, would be “like changing horses in the middle of the stream,” given that the jury had already received instructions and verdict forms. Ibid. The court informed counsel that it would declare a mis- trial “if the jury doesn’t make a decision.” Id., at 69. When the jury returned a half hour later, the foreperson stated that they had not reached a verdict. The court declared a mistrial and discharged the jury. The State subsequently sought to retry Blueford. He moved to dismiss the capital and first-degree murder charges on double jeopardy grounds, citing the foreperson’s report that the jurors had voted unanimously against guilt on those offenses. The trial court denied the motion, and the Supreme Court of Arkansas affirmed on interlocutory appeal. According to the State Supreme Court, the foreperson’s report had no effect on the State’s ability to retry Blueford, because the foreperson “was not making a formal announcement of acquittal” when she disclosed the jury’s votes. 2011 Ark. 8, p. 9, ___ S. W. 3d ___, ___. This was not a case, the court observed, “where a formal verdict was announced or entered of record.” Ibid. The court added that the trial court did not err in denying Blueford’s request for new verdict forms that would have allowed the jury to render a partial verdict on the charges of capital and first-degree murder. Blueford sought review in this Court, and we granted certiorari. 565 U. S. ___ (2011). II The Double Jeopardy Clause provides that no person shall “be subject for the same offence to be twice put in jeopardy of life or limb.” U. S. Const., Amdt. 5. The Clause “guarantees that the State shall not be permitted to make repeated attempts to convict the accused, thereby subjecting him to embarrassment, expense and ordeal and compelling him to live in a continuing state of anxiety and insecurity, as well as enhancing the possibility that even though innocent he may be found guilty.” United States v. Martin Linen Supply Co., 430 U.S. 564, 569 (1977) (internal quotation marks omitted). Blueford contends that the foreperson’s report means that he cannot be tried again on charges of capital and first-degree murder. According to Blueford, the Double Jeopardy Clause prohibits a second trial on those charges, for two reasons. A Blueford’s primary submission is that he cannot be retried for capital and first-degree murder because the jury actually acquitted him of those offenses. See Green v. United States, 355 U.S. 184, 188 (1957). The Arkansas Supreme Court noted—and Blueford acknowledges—that no formal judgment of acquittal was entered in his case. But none was necessary, Blueford maintains, because an acquittal is a matter of substance, not form. Quoting from our decision in Martin Linen, supra, at 571, Blueford contends that despite the absence of a formal verdict, a jury’s announcement constitutes an acquittal if it “ ‘actu- ally represents a resolution . . . of some or all of the factual elements of the offense charged.’ ” Brief for Petitioner 21. Here, according to Blueford, the foreperson’s announcement of the jury’s unanimous votes on capital and first-degree murder represented just that: a resolution of some or all of the elements of those offenses in Blueford’s favor. We disagree. The foreperson’s report was not a final resolution of anything. When the foreperson told the court how the jury had voted on each offense, the jury’s deliberations had not yet concluded. The jurors in fact went back to the jury room to deliberate further, even after the foreperson had delivered her report. When they emerged a half hour later, the foreperson stated only that they were unable to reach a verdict. She gave no indication whether it was still the case that all 12 jurors believed Blueford was not guilty of capital or first-degree murder, that 9 of them believed he was guilty of manslaughter, or that a vote had not been taken on negligent homicide. The fact that deliberations continued after the report deprives that report of the finality necessary to constitute an acquittal on the murder offenses. Blueford maintains, however, that any possibility that the jurors revisited the murder offenses was foreclosed by the instructions given to the jury. Those instructions, he contends, not only required the jury to consider the of- fenses in order, from greater to lesser, but also prevented it from transitioning from one offense to the next with- out unanimously—and definitively—resolving the greater offense in his favor. “A jury is presumed to follow its instructions.” Weeks v. Angelone, 528 U.S. 225, 234 (2000). So, Blueford says, the foreperson’s report that the jury was deadlocked on manslaughter necessarily establishes that the jury had acquitted Blueford of the greater offenses of capital and first-degree murder. But even if we assume that the instructions required a unanimous vote before the jury could consider a lesser offense—as the State assumes for purposes of this case, see Brief for Respondent 25, n. 3—nothing in the instructions prohibited the jury from reconsidering such a vote. The instructions said simply, “If you have a reasonable doubt of the defendant’s guilt on the charge of [the greater offense], you will [then] consider the charge of [the lesser offense].” App. 51–52. The jurors were never told that once they had a reasonable doubt, they could not rethink the issue. The jury was free to reconsider a greater offense, even after considering a lesser one.[1] A simple example illustrates the point. A jury enters the jury room, having just been given these instructions. The foreperson decides that it would make sense to determine the extent of the jurors’ agreement before discussions begin. Accordingly, she conducts a vote on capital murder, and everyone votes against guilt. She does the same for first-degree murder, and again, everyone votes against guilt. She then calls for a vote on manslaughter, and there is disagreement. Only then do the jurors en- gage in a discussion about the circumstances of the crime. While considering the arguments of the other jurors on how the death was caused, one of the jurors starts rethinking his own stance on a greater offense. After reflecting on the evidence, he comes to believe that the defendant did knowingly cause the death—satisfying the definition of first-degree murder. At that point, nothing in the instructions prohibits the jury from doing what juries often do: revisit a prior vote. “The very object of the jury system,” after all, “is to secure unanimity by a comparison of views, and by arguments among the jurors themselves.” Allen, 164 U. S., at 501. A single juror’s change of mind is all it takes to require the jury to reconsider a greater offense. It was therefore possible for Blueford’s jury to revisit the offenses of capital and first-degree murder, notwithstanding its earlier votes. And because of that possibility, the foreperson’s report prior to the end of deliberations lacked the finality necessary to amount to an acquittal on those offenses, quite apart from any requirement that a formal verdict be returned or judgment entered. That same lack of finality undermines Blueford’s reliance on Green v. United States, 355 U.S. 184 (1957), and Price v. Georgia, 398 U.S. 323 (1970). In those cases, we held that the Double Jeopardy Clause is violated when a defendant, tried for a greater offense and convicted of a lesser included offense, is later retried for the greater offense. See Green, supra, at 190; Price, supra, at 329. Blueford argues that the only fact distinguishing his case from Green and Price is that his case involves a deadlock on the lesser included offense, as opposed to a conviction. In his view, that distinction only favors him, because the Double Jeopardy Clause should, if anything, afford greater protection to a defendant who is not found guilty of the lesser included offense. Blueford’s argument assumes, however, that the votes reported by the foreperson did not change, even though the jury deliberated further after that report. That assumption is unjustified, because the reported votes were, for the reasons noted, not final. Blueford thus overlooks the real distinction between the cases: In Green and Price, the verdict of the jury was a final decision; here, the report of the foreperson was not. B Blueford maintains that even if the jury did not acquit him of capital and first-degree murder, a second trial on those offenses would nonetheless violate the Double Jeopardy Clause, because the trial court’s declaration of a mistrial was improper. Blueford acknowledges that a trial can be discontinued without barring a subsequent one for the same offense when “particular circumstances manifest a necessity” to declare a mistrial. Wade v. Hunter, 336 U.S. 684, 690 (1949); see also United States v. Perez, 9 Wheat. 579, 580 (1824). He also acknowledges that the trial court’s reason for declaring a mistrial here—that the jury was unable to reach a verdict—has long been con- sidered the “classic basis” establishing such a necessity. Arizona v. Washington, 434 U.S. 497, 509 (1978). Blueford therefore accepts that a second trial on manslaughter and negligent homicide would pose no double jeopardy problem. He contends, however, that there was no necessity for a mistrial on capital and first-degree murder, given the foreperson’s report that the jury had voted unanimously against guilt on those charges. According to Blueford, the court at that time should have taken “some action,” whether through partial verdict forms or other means, to allow the jury to give effect to those votes, and then considered a mistrial only as to the remaining charges. Reply Brief for Petitioner 11, n. 8. We reject that suggestion. We have never required a trial court, before declaring a mistrial because of a hung jury, to consider any particular means of breaking the impasse—let alone to consider giving the jury new options for a verdict. See Renico v. Lett, 559 U. S. ___, ___ (2010) (slip op., at 8).[2] As permitted under Arkansas law, the jury’s options in this case were limited to two: either convict on one of the offenses, or acquit on all. The instructions explained those options in plain terms, and the verdict forms likewise contemplated no other outcome. There were separate forms to convict on each of the possible offenses, but there was only one form to acquit, and it was to acquit on all of them. When the foreperson disclosed the jury’s votes on capital and first-degree murder, the trial court did not abuse its discretion by refusing to add another option—that of acquitting on some offenses but not others. That, however, is precisely the relief Blueford seeks—relief the Double Jeopardy Clause does not afford him. * * * The jury in this case did not convict Blueford of any offense, but it did not acquit him of any either. When the jury was unable to return a verdict, the trial court prop- erly declared a mistrial and discharged the jury. As a consequence, the Double Jeopardy Clause does not stand in the way of a second trial on the same offenses. The judgment of the Supreme Court of Arkansas is Affirmed. Notes 1 In reaching a contrary conclusion, post, at 6 (opinion of Sotomayor, J.), the dissent construes the jury instructions to “require a jury to complete its deliberations on a greater offense before it may consider a lesser,” post, at 3 (emphasis added). But no such requirement can be found in the text of the instructions themselves. And the dissent’s attempt to glean such a requirement from the Arkansas Supreme Court’s decision in Hughes v. State, 347 Ark. 696, 66 S.W.3d 645 (2002), is unavailing, for that decision nowhere addresses the issue here—whether a jury can reconsider a greater offense after considering a lesser one. 2 Finding our reliance on Renico “perplexing,” the dissent reads that decision to have “little to say about a trial judge’s responsibilities, or this Court’s, on direct review.” Post, at 10–11, n. 4. But Renico’s discussion of the applicable legal principles concerns just that, and the dissent in any event does not dispute that we have never required a trial court to consider any particular means of breaking a jury impasse.

566.US.2011_10-844

The Food and Drug Administration (FDA) regulates the manufacture, sale, and labeling of prescription drugs. A brand-name drug manufacturer seeking FDA approval for a drug submits a new drug application (NDA) containing, among other things, a statement of the drug’s components and proposed labeling describing the uses for which the drug may be marketed. See 21 U. S. C. §§355(b)(1), (d). Once the FDA has approved a brand manufacturer’s drug, another company may seek permission to market a generic version by filing an abbreviated new drug application (ANDA). See §§355(j)(2)(A)(ii), (iv). But the FDA cannot authorize a generic drug that would infringe a brand manufacturer’s patent. To facilitate the approval of generic drugs as soon as patents allow, the Hatch-Waxman Amendments require a brand manufacturer to submit its patent numbers and expiration dates, §355(b)(1); and FDA regulations require a description of any method-of-use patent, known as a use code, see 21 CFR §§314.53(c)(2)(ii)(P)(3), (e). The FDA does not attempt to verify the accuracy of the use codes that brand manufacturers supply. Instead, it simply publishes the codes, patent numbers, and expiration dates in a large volume known as the Orange Book. After consulting the Orange Book, an ANDA applicant enters one of several certifications to assure the FDA that its generic drug will not infringe the brand’s patent. If the patent has not expired, an applicant may fulfill this requirement in one of two ways. First, it may submit a so-called section viii statement asserting that it will market the drug for only those methods of use not covered by the brand’s patent, see 21 U. S. C. §355(j)(2)(A)(viii), and proposing a label that “carves out” the still-patented method(s) of use, see 21 CFR §314.94(a)(8)(iv). The FDA will not approve an ANDA with a section viii statement if the proposed label overlaps at all with the brand’s use code. Second, the ANDA applicant may file a so-called paragraph IV certification stating that the brand’s patent “is invalid or will not be infringed by the [generic drug’s] manufacture, use, or sale.” 21 U. S. C. §355(j)(2)(A)(vii)(IV). Such filing is treated as an act of infringement, giving the brand an immediate right to sue and resulting in a delay in the generic drug’s approval. In 2002, the Federal Trade Commission (FTC) issued a study detailing evidence that brands were submitting inaccurate patent information to the FDA in order to prevent or delay the marketing of generic drugs. In response, Congress created a statutory counterclaim available to generic manufacturers sued for patent infringement. The provision allows a generic manufacturer to “assert a counterclaim seeking an order requiring the [brand] to correct or delete the patent information submitted by the [brand] under subsection (b) or (c) [of 21 U. S. C. §355] on the ground that the patent does not claim . . . an approved method of using the drug.” 21 U. S. C. §355(j)(5)(C)(ii)(I). This case concerns the scope of the counterclaim provision. Respondents (collectively Novo) manufacture the brand-name version of the diabetes drug repaglinide. The FDA has approved three uses of the drug, but Novo’s method-of-use patent claims only one. Petitioners (collectively Caraco) wish to market a generic version of the drug for the other two approved methods of use. Caraco initially filed a paragraph IV certification and, considering this an act of infringement, Novo brought suit. Caraco then submitted a section viii statement and a proposed label carving out Novo’s patented therapy. But before the FDA could approve Caraco’s ANDA, Novo changed its use code to indicate that it held a patent on all three approved methods of using repaglinide. Because Caraco’s proposed label now overlapped with Novo’s use code, the FDA would not permit Caraco to employ section viii to bring its drug to market. Caraco filed a statutory counterclaim in the ongoing infringement action, seeking an order requiring Novo to “correct” its use code because the patent did not claim two of the three approved methods of using repaglinide. The District Court granted Caraco summary judgment, but the Federal Circuit reversed. It read the counterclaim’s phrase “the patent does not claim . . . an approved method of using the drug” as requiring Caraco to demonstrate that Novo’s patent does not claim any approved method of use; because the patent covers one approved method, the counterclaim was unavailable. The court also ruled that the counterclaim provision does not reach use codes because they are not “patent information submitted by the [brand] under subsection (b) or (c)” of §355. That information, the court concluded, consists only of the patent number and expiration date expressly required by the statutory provisions. Held: A generic manufacturer may employ the counterclaim provision to force correction of a use code that inaccurately describes the brand’s patent as covering a particular method of using a drug. Pp. 10–24. (a) The parties first dispute the meaning of “not an” in the phrase “the patent does not claim . . . an approved method of using the drug.” Novo contends that the counterclaim is available only if the patent claims no approved method of use, but Caraco reads this language to permit a counterclaim whenever a patent does not claim the particular method that the ANDA applicant seeks to market. In isolation, either of these readings is plausible, so the meaning of the phrase “not an” turns on statutory context, see Johnson v. United States, 559 U. S. ___, ___. This context favors Caraco: Congress understood that a drug may have multiple methods of use, not all of which a patent covers; and a section viii statement allows the FDA to approve a generic drug for unpatented uses so that it can quickly come to market. The statute thus contemplates that one patented use will not foreclose marketing a generic drug for other unpatented ones. Within this scheme, the counterclaim naturally functions to challenge the brand’s assertion of rights over whichever discrete uses the generic company wishes to pursue; the counterclaim’s availability matches the availability of FDA approval under the statute. Pp. 11–15. (b) The parties further dispute whether use codes qualify as “patent information submitted by the [brand] under subsection (b) or (c)” of §355. A use code, which is a description of the patent, surely qualifies as “patent information.” Novo nonetheless contends that use codes are not “submitted under” subsections (b) and (c) because those provisions expressly require an NDA applicant to provide only “the patent number and the expiration date of any patent” claiming the drug or a method of its use. But §§355(b) and (c) also govern the regulatory process by which brands provide additional patent information to the FDA. The term “under” is broad enough to include patent information, like use codes, that brands submit as required by this scheme. This reading draws support from the Court’s prior decisions in e.g., Eli Lilly & Co. v. Medtronic, Inc., 496 U.S. 661, 665–668, and Ardestani v. INS, 502 U.S. 129, 135; and it is bolstered by Congress’s use of the narrower phrases “described in” and “prescribed by” in neighboring provisions. See §§355(c)(2), (d)(6). Again, the conclusion that use codes are “submitted under” §§355(b) and (c) fits the broader statutory context. Use codes are pivotal to the FDA’s implementation of the Hatch-Waxman Amendments, and so it is unsurprising that the counterclaim provision’s language sweeps widely enough to embrace them. Pp. 15–18. (c) The counterclaim provision’s description of available remedies dispatches whatever remains of Novo’s arguments. The Court’s reading gives content to both remedies: It “delete[s]” a listing from the Orange Book when the brand holds no relevant patent and “correct[s]” the listing when the brand has misdescribed the patent’s scope. By contrast, Novo’s interpretation would all but read “correct” out of the statute. If, as Novo contends, the counterclaim is available only where the patent claims no approved method of use, the remedy will always be to delete the patent information. And if the counterclaim reaches only patent numbers and expiration dates, the Orange Book will include few if any mistakes in need of correction. Pp. 18–20. (d) Novo advances two arguments relating to the counterclaim’s drafting history, but neither overcomes the statutory text and context. The company first points out that Congress failed to pass an earlier bill that would have required brands to file descriptions of method-of-use patents and would have allowed generic companies to bring civil actions to “delete” or “correct” the information filed. Because that bill would have allowed a generic applicant to challenge overbroad descriptions of a patent, Novo contends that this Court cannot read the statute that was eventually enacted as doing the same. But the earlier bill contained numerous items that may have caused its failure. And the limited criticism of its mechanism for challenging brands’ descriptions of their patents focused on the creation of an independent cause of action—stronger medicine than the counterclaim at issue here. Finally, between that bill’s demise and the counterclaim’s enactment, the FDA issued a rule requiring brands to supply use codes. The counterclaim provision’s drafters thus had no need to require this information. Novo next contends that Congress established the counterclaim only to address the impossibility of deleting an improperly listed patent from the Orange Book—a problem that had come to light when the Federal Circuit held in Mylan Pharmaceuticals, Inc. v. Thompson, 268 F.3d 1323, that generics had no cause of action to delist a patent. Novo thus contends that the counterclaim is a mere delisting provision. But this Court thinks Mylan alerted Congress to a broader problem: that generic companies generally had no avenue to challenge the accuracy of brands’ patent listings, and that the FDA therefore could not approve proper applications to bring inexpensive drugs to market. Again, the proof of that lies in the statute itself—its text and context demonstrate that the counterclaim is available not only (as in Mylan) when the patent listing is baseless, but also (as here) when it is overbroad. Moreover, Congress’s equation of the two situations makes perfect sense. In either case, the brand submits misleading patent information to the FDA, delaying or blocking approval of a generic drug that infringes no patent and thus, under the statute, should go to market. Pp. 20–24. 601 F.3d 1359, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court. Sotomayor, J., filed a concurring opinion.

When the Food and Drug Administration (FDA) evaluates an application to market a generic drug, it considers whether the proposed drug would infringe a patent held by the manufacturer of the brand-name version. To assess that matter, the FDA requires brand manufacturers to submit descriptions of the scope of their patents, known as use codes. The FDA does not attempt to determine if that information is accurate. Rather, the FDA assumes that it is so and decides whether to approve a generic drug on that basis. As a result, the breadth of the use code may make the difference between approval and denial of a generic company’s application. In this case, we consider whether Congress has authorized a generic company to challenge a use code’s accuracy by bringing a counterclaim against the brand manufac- turer in a patent infringement suit. The relevant statute provides that a generic company “may assert a counterclaim seeking an order requiring the [brand manufac- turer] to correct or delete the patent information [it] submitted . . . under [two statutory subsections] on the ground that the patent does not claim . . . an approved method of using the drug.” 117Stat. 2452, 21 U. S. C. §355(j)(5)(C)(ii)(I). We hold that a generic manufacturer may employ this provision to force correction of a use code that inaccurately describes the brand’s patent as covering a particular method of using the drug in question. I A The FDA regulates the manufacture, sale, and labeling of prescription drugs under a complex statutory scheme. To begin at the beginning: When a brand manufacturer wishes to market a novel drug, it must submit a new drug application (NDA) to the FDA for approval. The NDA must include, among other things, a statement of the drug’s components, scientific data showing that the drug is safe and effective, and proposed labeling describing the uses for which the drug may be marketed. See §§355(b)(1), (d). The FDA may approve a brand-name drug for multiple methods of use—either to treat different conditions or to treat the same condition in different ways. Once the FDA has approved a brand manufacturer’s drug, another company may seek permission to market a generic version pursuant to legislation known as the Hatch-Waxman Amendments. See Drug Price Competition and Patent Term Restoration Act of 1984, 98Stat. 1585. Those amendments allow a generic competitor to file an abbreviated new drug application (ANDA) piggy-backing on the brand’s NDA. Rather than providing independent evidence of safety and efficacy, the typical ANDA shows that the generic drug has the same active ingredients as, and is biologically equivalent to, the brand-name drug. See §§355(j)(2)(A)(ii), (iv). As we have previously recognized, this process is designed to speed the introduction of low-cost generic drugs to market. See Eli Lilly & Co. v. Medtronic, Inc., 496 U.S. 661, 676 (1990). Because the FDA cannot authorize a generic drug that would infringe a patent, the timing of an ANDA’s approval depends on the scope and duration of the patents cover- ing the brand-name drug. Those patents come in different varieties. One type protects the drug compound itself. Another kind—the one at issue here—gives the brand manufacturer exclusive rights over a particular method of using the drug. In some circumstances, a brand manufacturer may hold such a method-of-use patent even after its patent on the drug compound has expired. To facilitate the approval of generic drugs as soon as patents allow, the Hatch-Waxman Amendments and FDA regulations direct brand manufacturers to file information about their patents. The statute mandates that a brand submit in its NDA “the patent number and the expiration date of any patent which claims the drug for which the [brand] submitted the [NDA] or which claims a method of using such drug.” §§355(b)(1). And the regulations issued under that statute require that, once an NDA is approved, the brand provide a description of any method-of-use patent it holds. See 21 CFR §§314.53(c)(2)(ii)(P)(3), (e) (2011). That description is known as a use code, and the brand submits it on FDA Form 3542. As later discussed, the FDA does not attempt to verify the accuracy of the use codes that brand manufacturers supply. It simply pub- lishes the codes, along with the corresponding patent numbers and expiration dates, in a fat, brightly hued volume called the Orange Book (less colorfully but more officially denominated Approved Drug Products with Therapeutic Equivalence Evaluations). After consulting the Orange Book, a company filing an ANDA must assure the FDA that its proposed generic drug will not infringe the brand’s patents. When no patents are listed in the Orange Book or all listed patents have expired (or will expire prior to the ANDA’s approval), the generic manufacturer simply certifies to that effect. See 21 U. S. C. §§355(j)(2)(A)(vii)(I)–(III). Otherwise, the applicant has two possible ways to obtain approval. One option is to submit a so-called section viii statement, which asserts that the generic manufacturer will market the drug for one or more methods of use not covered by the brand’s patents. See §355(j)(2)(A)(viii). A section viii statement is typically used when the brand’s patent on the drug compound has expired and the brand holds patents on only some approved methods of using the drug. If the ANDA applicant follows this route, it will propose labeling for the generic drug that “carves out” from the brand’s approved label the still-patented methods of use. See 21 CFR §314.94(a)(8)(iv). The FDA may approve such a modified label, see §314.127(a)(7), as an ex- ception to the usual rule that a generic drug must bear the same label as the brand-name product, see 21 U. S. C. §§355(j)(2)(A)(v), (j)(4)(G). FDA acceptance of the carve-out label allows the generic company to place its drug on the market (assuming the ANDA meets other requirements), but only for a subset of approved uses—i.e., those not covered by the brand’s patents. Of particular relevance here, the FDA will not approve such an ANDA if the generic’s proposed carve-out label overlaps at all with the brand’s use code. See 68 Fed. Reg. 36682–36683 (2003). The FDA takes that code as a given: It does not independently assess the patent’s scope or otherwise look behind the description authored by the brand. According to the agency, it lacks “both [the] expertise and [the] authority” to review patent claims; although it will forward questions about the accuracy of a use code to the brand,[1] its own “role with respect to patent listing is ministerial.” Id., at 36683; see ibid. (“A fundamental assumption of the Hatch-Waxman Amendments is that the courts are the appropriate mechanism for the resolution of disputes about the scope and validity of patents”).[2] Thus, whether section viii is available to a generic manufacturer depends on how the brand describes its patent. Only if the use code provides sufficient space for the generic’s proposed label will the FDA approve an ANDA with a section viii statement. The generic manufacturer’s second option is to file a so-called paragraph IV certification, which states that a listed patent “is invalid or will not be infringed by the manufacture, use, or sale of the [generic] drug.” 21 U. S. C. §355(j)(2)(A)(vii)(IV). A generic manufacturer will typically take this path in either of two situations: if it wants to market the drug for all uses, rather than carving out those still allegedly under patent; or if it discovers, as described above, that any carve-out label it is willing to adopt cannot avoid the brand’s use code. Filing a paragraph IV certification means provoking litigation. The patent statute treats such a filing as itself an act of infringement, which gives the brand an immediate right to sue. See 35 U. S. C. §271(e)(2)(A). Assuming the brand does so, the FDA generally may not approve the ANDA until 30 months pass or the court finds the patent invalid or not infringed. See 21 U. S. C. §355(j)(5)(B)(iii). Accordingly, the paragraph IV process is likely to keep the gen- eric drug off the market for a lengthy period, but may eventually enable the generic company to market its drug for all approved uses. In the late 1990’s, evidence mounted that some brands were exploiting this statutory scheme to prevent or delay the marketing of generic drugs, and the Federal Trade Commission (FTC) soon issued a study detailing these anticompetitive practices. See FTC, Generic Drug Entry Prior to Patent Expiration: An FTC Study, pp. iii–vi (July 2002) (hereinafter FTC Study). That report focused at- tention on brands’ submission of inaccurate patent information to the FDA. In one case cited by the FTC, Mylan Pharmaceuticals, Inc. v. Thompson, 268 F.3d 1323 (CA Fed. 2001), a brand whose original patent on a drug was set to expire listed a new patent ostensibly extending its rights over the drug, but in fact covering neither the compound nor any method of using it. The FDA, as was (and is) its wont, accepted the listing at its word and accord- ingly declined to approve a generic product. The generic manufacturer sued to delete the improper listing from the Orange Book, but the Federal Circuit held that the Hatch-Waxman Amendments did not allow such a right of action. See id., at 1330–1333. As the FTC noted, that ruling meant that the only option for generic manufacturers in Mylan’s situation was to file a paragraph IV certification (triggering an infringement suit) and then wait out the usual 30-month period before the FDA could approve an ANDA. See FTC Study 40–45. Congress responded to these abuses by creating a mechanism, in the form of a legal counterclaim, for generic manufacturers to challenge patent information a brand has submitted to the FDA. See Medicare Prescription Drug, Improvement, and Modernization Act of 2003, 117Stat. 2452. The provision authorizes an ANDA applicant sued for patent infringement to “assert a counterclaim seeking an order requiring the [brand] to correct or delete the patent information submitted by the [brand] under subsection (b) or (c) [of §355] on the ground that the patent does not claim either— “(aa) the drug for which the [brand’s NDA] wasapproved; or “(bb) an approved method of using the drug.” 21U. S. C. §355(j)(5)(C)(ii)(I). The counterclaim thus enables a generic competitor to obtain a judgment directing a brand to “correct or delete” certain patent information that is blocking the FDA’s approval of a generic product. This case raises the question whether the counterclaim is available to fix a brand’s use code. B The parties to this case sell or seek to sell the diabetes drug repaglinide. Respondents (collectively Novo) manufacture Prandin, the brand-name version of the drug. The FDA has approved three uses of Prandin to treat diabe- tes: repaglinide by itself; repaglinide in combination with metformin; and repaglinide in combination with thiazolidinediones (TZDs). Petitioners (collectively Caraco) wish to market a generic version of the drug for two of those uses. Novo originally owned a patent for the repaglinide compound, known as the ’035 patent, but it expired in 2009. In 2004, Novo also acquired a method-of-use patent for the drug, called the ’358 patent, which does not expire until 2018. That patent—the one at issue here—claims a “method for treating [diabetes by] administering . . . repaglinide in combination with metformin.” 601 F.3d 1359, 1362 (CA Fed. 2010). Thus, Novo currently holds a patent for one of the three FDA-approved uses of repaglinide—its use with metformin. But Novo holds no patent for the use of repaglinide with TZDs or its use alone. In 2005, Caraco filed an ANDA seeking to market a generic version of repaglinide. At that time, the Orange Book entry for Prandin listed both the ’035 patent (the drug compound) and the ’358 patent (the use of the drug with metformin). Caraco assured the FDA that it would not market its generic drug until the ’035 patent expired, thus making that patent irrelevant to the FDA’s review of the ANDA. Caraco filed a paragraph IV certification for the remaining, ’358 patent, stating that it was “invalid or [would] not be infringed.” §355(j)(2)(A)(vii)(IV); see supra, at 5. In accord with the patent statute, Novo treated this filing as an act of infringement and brought suit. When Caraco filed its ANDA, Novo’s use code for the ’358 patent represented that the patent covered “ ‘[u]se of repaglinide in combination with metformin to lower blood glucose.’ ” 601 F. 3d, at 1362–1363. The FDA therefore advised Caraco that if it did not seek to market repag- linide for use with metformin, it could submit a section viii statement. That would allow Caraco, assuming its ANDA was otherwise in order, to market its generic drug for the other two uses. Caraco took the FDA’s cue and in 2008 submitted a section viii statement, with proposed labeling carving out Novo’s patented metformin therapy. See App. 166–176. Before the FDA took further action, however, Novo changed its use code for the ’358 patent. The new use code describes “ ‘[a] method for improving glycemic control in adults with type 2 diabetes.’ ”[3] 601 F. 3d, at 1363. Because that code indicates that the ’358 patent protects all three approved methods of using repaglinide to treat diabetes, Caraco’s proposed carve-out of metformin ther- apy was no longer sufficient; even with that exclusion, Caraco’s label now overlapped with Novo’s use code on the other two uses. And Caraco could not carve out those uses as well, because at that point nothing would be left for it to market. The FDA has approved repaglinide for only three uses, and Novo’s use code encompassed them all. The FDA accordingly informed Caraco that it could no longer employ section viii to bring its drug to market. Caraco responded to Novo’s new, preclusive use code by filing a statutory counterclaim in the ongoing infringement suit. The counterclaim sought an order requiring Novo to “correct” its use code “on the ground that [the ’358] patent does not claim” two approved methods of using repaglinide—alone and in combination with TZDs. §355(j)(5)(C)(ii)(I); see supra, at 6–7. That order would permit the FDA to accept Caraco’s proposed carve-out label and approve the company’s ANDA. The District Court granted summary judgment to Caraco, enjoining Novo to “correct . . . its inaccurate description of the ’358 patent” by submitting a new Form 3542 to the FDA that would “reinstat[e] its former” use code. App. to Pet. for Cert. 65a–66a. The Court of Appeals reversed, holding that Caraco lacked “a statutory basis to assert a counterclaim.” 601 F. 3d, at 1360. The court first read the statutory phrase “the patent does not claim . . . an approved method of using the drug” to require Caraco to demonstrate that the ’358 patent does not claim any approved method of use. See id., at 1365 (“ ‘[A]n approved method’ means ‘any approved method’ ”). Because the patent covers one approved method of use—repaglinide in combination with metformin—the counterclaim was unavailable. The court further ruled that the counterclaim provision does not reach use codes because they are not “patent information submitted by the [brand] under subsection (b) or (c).” On the Federal Circuit’s view, that information consists only of the patent number and expiration date. See id., at 1366–1367. Judge Dyk dissented. He would have read the phrase “the patent does not claim . . . an approved method of using the drug” to include situations where, as here, the use code wrongly indicates that the patent covers one or more particular approved methods of use. See id., at 1376–1378. And he would have construed “patent in- formation submitted . . . under subsection (b) or (c)” to include use codes. See id., at 1370–1376.[4] We granted certiorari, 564 U. S. ___ (2011), and now reverse. II We begin “where all such inquiries must begin: with the language of the statute itself.” United States v. Ron Pair Enterprises, Inc., 489 U.S. 235, 241 (1989). This case requires us to construe two statutory phrases. First, we must decide when a “patent does not claim . . . an approved method of using” a drug. Second, we must determine the content of “patent information submitted . . . under subsection (b) or (c)” of §355. We consider both of those questions against the backdrop of yet a third statutory phrase, providing that the remedy for a prevailing counterclaimant is an order requiring the brand “to correct or delete” that patent information. And we consider each question in the context of the entire statute. See Robinson v. Shell Oil Co., 519 U.S. 337, 341 (1997) (Statutory interpretation focuses on “the language itself, the specific context in which that language is used, and the broader context of the statute as a whole”). We cannot say that the counterclaim clause is altogether free of ambiguity. But when we consider statutory text and context together, we conclude that a generic manufacturer in Caraco’s position can use the counterclaim.[5] A An ANDA applicant sued for patent infringement may bring a counterclaim “on the ground that the patent does not claim . . . an approved method of using the drug.” 21 U. S. C. §355(j)(5)(C)(ii)(I). The parties debate the meaning of this language. Novo (like the Federal Circuit) reads “not an” to mean “not any,” contending that “the counterclaim is available only if the listed patent does not claim any (or, equivalently, claims no) approved method of using the drug.” Brief for Respondents 29 (internal quotation marks omitted). By that measure, Caraco may not bring a counterclaim because Novo’s ’358 patent claims the use of repaglinide with metformin. In contrast, Caraco reads “not an” to mean “not a particular one,” so that the statute permits a counterclaim whenever the patent does not claim a method of use for which the ANDA applicant seeks to market the drug. On that view, the counterclaim is available here—indeed, is available twice over—because the ’358 patent does not claim the use of repaglinide with TZDs or its use alone. Truth be told, the answer to the general question “What does ‘not an’ mean?” is “It depends”: The meaning of the phrase turns on its context. See Johnson v. United States, 559 U. S. ___, ___ (2010) (slip op., at 5) (“Ultimately, context determines meaning”). “Not an” sometimes means “not any,” in the way Novo claims. If your spouse tells you he is late because he “did not take a cab,” you will infer that he took no cab at all (but took the bus instead). If your child admits that she “did not read a book all summer,” you will surmise that she did not read any book (but went to the movies a lot). And if a sports-fan friend bemoans that “the New York Mets do not have a chance of winning the World Series,” you will gather that the team has no chance whatsoever (because they have no hitting). But now stop a moment. Suppose your spouse tells you that he got lost because he “did not make a turn.” You would understand that he failed to make a particular turn, not that he drove from the outset in a straight line. Suppose your child explains her mediocre grade on a college exam by saying that she “did not read an assigned text.” You would infer that she failed to read a specific book, not that she read nothing at all on the syllabus. And suppose a lawyer friend laments that in her last trial, she “did not prove an element of the offense.” You would grasp that she is speaking not of all the elements, but of a particular one. The examples could go on and on, but the point is simple enough: When it comes to the meaning of “not an,” context matters.[6] And the statutory context here supports Caraco’s position. As described earlier (and as Congress understood), a single drug may have multiple methods of use, only one or some of which a patent covers. See, e.g., 21 U. S. C. §355(b)(1) (requiring that an NDA applicant file information about “any patent which claims the drug . . . or which claims a method of using such drug” (emphasis added)). The Hatch-Waxman Amendments authorize the FDA to approve the marketing of a generic drug for particular unpatented uses; and section viii provides the mechanism for a generic company to identify those uses, so that a product with a label matching them can quickly come to market. The statutory scheme, in other words, contemplates that one patented use will not foreclose marketing a generic drug for other unpatented ones. Within that framework, the counterclaim naturally functions to challenge the brand’s assertion of rights over whichever discrete use (or uses) the generic company wishes to pursue. That assertion, after all, is the thing blocking the generic drug’s entry on the market. The availability of the counterclaim thus matches the availability of FDA approval under the statute: A company may bring a counterclaim to show that a method of use is unpatented because establishing that fact allows the FDA to authorize a generic drug via section viii. Consider the point as applied to this case. Caraco wishes to market a generic version of repaglinide for two (and only two) uses. Under the statute, the FDA could approve Caraco’s application so long as no patent covers those uses, regardless whether a patent protects yet a third method of using the drug. Novo agrees that Caraco could bring a counterclaim if Novo’s assertion of patent protection for repaglinide lacked any basis—for example, if Novo held no patent, yet claimed rights to the pair of uses for which Caraco seeks to market its drug. But because Novo has a valid patent on a different use, Novo argues that Caraco’s counterclaim evaporates. And that is so even though, once again, Caraco has no wish to market its product for that patented use and the FDA stands ready, pursuant to the statute, to approve Caraco’s product for the other two. To put the matter simply, Novo thinks the counterclaim disappears because it has a patent for a method of use in which neither Caraco nor the FDA is interested at all. “It would take strong evidence to persuade us that this is what Congress wrought.” Eli Lilly, 496 U. S., at 673. That “not an” sometimes (but sometimes not) means “not any” is not enough. Novo argues that our reading must be wrong because Congress could have expressly “impose[d] additional . . . qualifications” on the term “an approved method of us[e]”—and indeed did so in another place in the statute. Brief for Respondents 31; 21 U. S. C. §355(j)(5)(C)(ii)(I). Novo points here to section viii itself, which applies when the brand’s patent “does not claim a use for which the [ANDA] applicant is seeking approval.” §355(j)(2)(A)(viii) (emphasis added). But the mere possibility of clearer phrasing cannot defeat the most natural reading of a statute; if it could (with all due respect to Congress), we would interpret a great many statutes differently than we do. Nor does Congress’s use of more detailed language in another provision, enacted years earlier, persuade us to put the counterclaim clause at odds with its statutory context. That is especially so because we can turn this form of argument back around on Novo. Congress, after all, could have more clearly expressed Novo’s proposed meaning in the easiest of ways—by adding a single letter to make clear that “not an” really means “not any.” And indeed, Congress used a “not any” construction in the very next subclause, enacted at the very same time. See §355(j)(5)(C)(ii)(II) (“Subclause (I) does not authorize the assertion of a claim . . . in any [other] civil action”). So if we needed any proof that Congress knew how to say “not any” when it meant “not any,” here we find it. We think that sees, raises, and bests Novo’s argument. Our more essential point, though, has less gamesmanship about it: We think that the “not any” construction does not appear in the relevant counterclaim provision because Congress did not mean what Novo wishes it had. And we think that is so because Congress meant (as it usually does) for the provision it enacted to fit within the statutory scheme—here, by facilitating the approval of non-infringing generic drugs under section viii. B Novo contends that Caraco’s counterclaim must fail for another, independent reason: On its view (as on the Federal Circuit’s), the counterclaim does not provide a way to correct use codes because they are not “patent information submitted by the [brand] under subsection (b) or (c)” of §355. Once again, we disagree. The statute does not define “patent information,” but a use code must qualify. It describes the method of use claimed in a patent. See 21 CFR §§314.53(c)(2)(ii)(P)(3), (e). That fits under any ordinary understanding of the language.[7] The more difficult question arises from the “submitted under” phrase. The subsections mentioned there—(b) and (c) of §355—require an NDA applicant to submit specified information: “the patent number and the expiration date of any patent” claiming the drug or a method of its use. 21 U. S. C. §§355(b)(1), (c)(2). According to Novo, only that information comes within the counterclaim provision. But subsections (b) and (c) as well govern the regulatory process by which brands provide additional patent information to the FDA, both before and after an NDA is approved. In particular, those subsections provide the basis for the regulation requiring brands to submit use codes, see 21 CFR §314.53; in issuing that regulation, the FDA noted that “[o]ur principal legal authority . . . is section 505 of the act [codified at §355], in conjunction with our general rulemaking authority.” 68 Fed. Reg. 36697–36698 (specifically referring to subsections (b) and (c)). And the form (Form 3542) on which brands submit their use codes states that the information appearing there is “provided in accordance with Section [355](b) and (c).” App. 97. So use codes fall within the counterclaim’s ambit if the phrase “submitted under” reaches filings that not only subsections (b) and (c) themselves, but also their implementing regulations require. Several of our cases support giving “under” this broad meaning. For example, in Eli Lilly, 496 U. S., at 665–668, we examined a similar statutory reference to the “submission of information under a Federal law which regulates the manufacture, use, or sale of drugs,” 35 U. S. C. §271(e)(1). We noted there that submitting information “under a Federal law” suggests doing so “in furtherance of or compliance with a comprehensive scheme of regulation.” 496 U. S., at 667. Likewise, in Ardestani v. INS, 502 U.S. 129, 135 (1991), we held that a regulatory proceeding “under section 554,” 5 U. S. C. §504(b)(1)(C)(i), meant any proceeding “subject to,” “governed by,” or conducted “by reason of the authority of” that statutory provision. So too here. “Patent information submitted . . . under subsection (b) or (c)” most naturally refers to patent information provided as part of the “comprehensive scheme of regulation” premised on those subsections. Eli Lilly, 496 U. S., at 667. It includes everything (about patents) that the FDA requires brands to furnish in the proceedings “subject to,” “governed by,” or conducted “by reason of the authority of” §§355(b) and (c). Ardestani, 502 U. S., at 135. The breadth of the term “under” becomes particu- larly clear when compared with other phrases—“described in” and “prescribed by”—appearing in neighboring provisions. See, e.g., 21 U. S. C. §355(c)(2) (“patent information described in subsection (b)”); §355(d)(6) (“patent information prescribed by subsection (b)”). Those phrases denote a patent number and expiration date and nothing more. In contrast, the word “under” naturally reaches beyond that most barebones information to other patent materials the FDA demands in the regulatory process. Once again, that congressional choice fits the broader statutory context. Use codes are pivotal to the FDA’s implementation of the Hatch-Waxman Amendments—and no less so because a regulation, rather than the statute itself, requires their submission. Recall that those Amendments instruct the FDA (assuming other requirements are met) to approve an ANDA filed with a section viii statement when it proposes to market a drug for only unpatented methods of use. To fulfill that charge, the FDA must determine whether any patent covers a particular method of use; and to do that, the agency (which views itself as lacking expertise in patent matters, see supra, at 4–5, and n. 2) relies on the use codes submitted in the regulatory process. See 68 Fed. Reg. 36682–36683. An overbroad use code therefore throws a wrench into the FDA’s ability to approve generic drugs as the statute con- templates. So it is not surprising that the language Congress used in the counterclaim provision sweeps widely enough to embrace that filing. C Another aspect of the counterclaim provision—its description of available remedies—dispatches whatever remains of Novo’s arguments. According to the statute, a successful claimant may obtain an order requiring the brand to “correct or delete” its patent information. §355(j)(5)(C)(ii)(I). Our interpretation of the statute gives content to both those remedies: It deletes a listing from the Orange Book when the brand holds no relevant patent and corrects the listing when the brand has misdescribed the patent’s scope. By contrast, Novo’s two arguments would all but read the term “correct” out of the statute. Consider first how Novo’s an-means-any contention would accomplish that result. Recall that on Novo’s view, a counterclaim can succeed only if the patent challenged does not claim either the drug or any approved method of using it. See supra, at 11. But when a generic manufacturer makes that showing, the remedy must be to “delete” the listing; no correction would be enough. Novo agrees with that proposition; “[a]t bottom,” Novo avers, “the counterclaim is a delisting provision.” Brief for Respondents 20. But that raises the obvious question: Why did Congress also include the term “correct” in the statute? Novo can come up with just one answer: The counterclaim, it proposes, can correct erroneous patent numbers. Imagine, for example, that Novo mistakenly entered the number ’359, instead of ’358, when submitting information about its repaglinide patent for publication in the Orange Book. Then, Novo suggests, Caraco could bring a counterclaim to challenge the inaccurate listing (on the ground that ’359 does not claim any method of use), and the remedy would be “correct[ion]” (substituting an 8 for a 9). But we think Novo’s admission that this scenario would be “unusual,” Tr. of Oral Arg. 41, considerably understates the matter. As Novo concedes, brands have every incentive to provide the right patent number in the first place, and to immediately rectify any error brought to their attention. See id., at 40–41. By doing so, they place both generic companies and the FDA on notice of their patents and thereby prevent infringement. And conversely, generics have little or no incentive to bring a counterclaim that will merely replace one digit in the Orange Book with another. So we doubt Congress created a legal action to “correct” patent information just to fix such scrivener’s errors. See, e.g., TRW Inc. v. Andrews, 534 U.S. 19, 31 (2001) (refusing to adopt an interpretation of a statute that would render a piece of it “insignificant, if not wholly superfluous” (internal quotation marks omitted)). That would have been, in the most literal sense, to make a federal case out of nothing. The same problem afflicts Novo’s alternative contention—that “patent information submitted . . . under subsection (b) or (c)” includes only numbers and expiration dates (and not use codes). Once again, we cannot think Congress included the remedy of “correct[ion]” so that courts could expunge typos in patent numbers. And not even Novo has proffered a way for the counterclaim to “correct” an erroneous expiration date. Suppose, for example, that a brand incorrectly lists the expiration date of a valid patent as 2018 rather than 2015. The counterclaim would be useless: It authorizes a remedy only “on the ground that” the listed patent does not claim the drug or an approved method of using it—and notwithstanding the wrong expiration date, this patent does so. Alternatively, suppose the brand lists a patent as having a 2018 expiration date when in fact the patent has already lapsed. Then, a generic manufacturer could bring a counterclaim alleging that the patent no longer claims the drug or a method of using it—but the appropriate remedy would be deletion, not correction, of the brand’s listing. Novo’s reading of “patent information,” like its reading of “not an,” effectively deletes the term “correct” from the statute. III Novo finally advances two arguments relating to the counterclaim’s drafting history. Neither contention, however, overcomes the statutory text and context. Indeed, consideration of the provision’s background only strengthens our view of its meaning. A Novo first contends that our interpretation of the statute “effectively resurrect[s] the scheme rejected by Congress.” Brief for Respondents 44 (quoting Smith v. United States, 507 U.S. 197, 203, n. 4 (1993)). In 2002, Novo notes, Congress failed to pass a bill that would have required brands to file specified “patent information,” including, for method-of-use patents, a description of “the approved use covered by the [patent] claim.” S. 812, 107th Cong., 2d Sess., §103(a)(1), p. 7 (engrossed bill). That bill would have allowed a generic company to bring its own civil action—not merely a counterclaim in ongoing litigation—to “delete” or “correct” the information filed. Id., at 8. The Senate approved the bill, but the House of Repre- sentatives took no action on it. Novo argues that because this failed legislation would have allowed a generic company to challenge overbroad descriptions of a patent, we cannot read the statute Congress eventually enacted as doing so. We disagree. We see no reason to assume, as Novo does, that Congress rejected S. 812 because it required brands to submit patent information beyond a number and expiration date. Indeed, Novo’s argument highlights the perils of relying on the fate of prior bills to divine the meaning of enacted legislation. “A bill can be proposed for any number of reasons, and it can be rejected for just as many others.” Solid Waste Agency of Northern Cook Cty. v. Army Corps of Engineers, 531 U.S. 159, 170 (2001)). S. 812 contained numerous items, including a title on importing prescription drugs (no controversy there!), that may have caused its failure. See S. 812, Tit. II. Moreover, what criticism there was of the bill’s mechanism for challenging brands’ patent claims focused not on the specification of “patent information,” but instead on the creation of an independent cause of action—stronger medicine than the counterclaim Congress ultimately adopted.[8] And finally, Novo ignores a likely cause for the redrafting of the provision on submitting information. Between S. 812’s demise and the counterclaim’s enactment, the FDA issued a rule requiring brands to supply material concerning method-of-use patents, including use codes. The drafters of the counterclaim provision knew about that rule,[9] and had no need to duplicate its list of mandated filings. So the drafting history does not support Novo’s conclusion. If anything, the statute’s evolution indicates that Congress determined to enforce the FDA’s new listing provisions, including its use-code requirement, through the new counterclaim. B Novo next argues that Congress established the counterclaim only to solve the problem raised by the Federal Circuit’s decision in Mylan, 268 F. 3d 1323—the impossibility of deleting an improperly listed patent from the Orange Book. In Mylan, as earlier described, a generic company alleged that a brand had listed a patent that covered neither the approved drug nor any method of using it, and brought an action seeking delisting. See supra, at 6. The Federal Circuit held that no such action was available, even assuming the allegation was true. Because several legislators saw Mylan as “exemplif[ying]” brands’ “perceived abuse” of the FDA’s patent listing prac- tices, Brief for Respondents 35, Novo contends that we should construe the counterclaim provision to aid only a generic company that “finds itself in the same position as Mylan was in Mylan,” Supp. Brief in Opposition 5–6. Once again, we think not. Maybe Mylan triggered the legislative effort to enact a counterclaim, or maybe it didn’t: By the time Congress acted, it also had at hand an FTC study broadly criticizing brands’ patent listings and an FDA rule designed to address the very same issue. See supra, at 6, 21. But even assuming Mylan “prompted the proposal” of the counterclaim, “whether that alone accounted for its enactment is quite a different question.” Eli Lilly, 496 U. S., at 670, n. 3 (emphasis deleted). Here, we think Mylan alerted Congress to a broader problem—that generic companies generally had no avenue to challenge the accuracy of brands’ patent listings, and that the FDA therefore could not approve proper applications to bring inexpensive drugs to market. The proof of that lies in the statute itself (where the best proof of what Congress means to address almost always resides). As we have de- scribed, the statute’s text and context demonstrate that the counterclaim is available not only (as in Mylan) when the patent listing is baseless, but also (as here) when it is overbroad. See supra, at 10–20. In particular, Congress’s decision to allow a counterclaimant to seek “correct[ion]” of patent information explodes Novo’s theory, because the remedy for a Mylan-type impropriety is complete delisting. And to make matters still easier, Congress’s equation of the two situations—the one in Mylan and the one here—makes perfect sense. Whether a brand lists a patent that covers no use or describes a patent on one use as extending to others, the brand submits misleading patent in- formation to the FDA. In doing so, the brand equally ex- ploits the FDA’s determination that it cannot police patent claims. And the brand’s action may in either case delay or block approval of a generic drug that infringes no patent—and that under the statute should go to market. See su- pra, at 4. That is the danger Caraco faces here, as much as it was the threat in Mylan: Novo seeks to preclude Caraco from selling repaglinide for unpatented uses until 2018, when Novo’s patent on a different use expires. Indeed, the need for the counterclaim is greater here than in Mylan. When a brand lists a patent that covers no use, a generic company has a pathway aside from the counterclaim to challenge the listing. As described ear- lier, the company may make a paragraph IV certification stating that the listed patent “is invalid or will not be infringed” by the generic drug. 21 U. S. C. §355(j)(2)(A)(vii)(IV); see supra, at 5. If the brand sues, the generic company can argue that its product would not infringe the patent. Using the counterclaim may enable a generic manufacturer to obtain delisting more quickly, see Tr. of Oral Arg. 54; but even without it, the company can eventually get a judgment of non-infringement enabling the FDA to approve its ANDA. In contrast, where (as here) a brand files an overbroad use code, a generic company cannot use paragraph IV litigation to that end. A paragraph IV certification (unlike a section viii statement) requires the generic company to propose labeling identical to the brand’s; it cannot carve out any uses. See supra, at 4. And that proposed label will necessarily infringe because it will include the use(s) on which the brand does have a patent. So here, a paragraph IV suit cannot lead to a judgment enabling FDA approval; the counterclaim offers the only route to bring the generic drug to market for non-infringing uses. Novo’s view eliminates the counterclaim where it has the greatest value. IV The statutory counterclaim we have considered enables courts to resolve patent disputes so that the FDA can fulfill its statutory duty to approve generic drugs that do not infringe patent rights. The text and context of the provision demonstrate that a generic company can employ the counterclaim to challenge a brand’s overbroad use code. We accordingly hold that Caraco may bring a counterclaim seeking to “correct” Novo’s use code “on the ground that” the ’358 patent “does not claim . . . an approved method of using the drug”—indeed, does not claim two. The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 Under the FDA’s regulations, any person may dispute the accuracy of patent information listed in the Orange Book by notifying the agency in writing. See 21 CFR §314.53(f). The FDA will then request that the brand verify the information, but will make no changes “[u]nless the [brand] withdraws or amends” the listing. Ibid. 2 Several courts have affirmed the FDA’s view of its ministerial role. See, e.g., Apotex, Inc. v. Thompson, 347 F.3d 1335, 1349 (CA Fed. 2003); aaaiPharma v. Thompson, 296 F.3d 227, 242–243 (CA4 2002). That question is not before us, and we express no view on it. 3 Novo asserts that it made the change so that its use code would mirror its label, which the FDA had just asked it to alter. See Brief for Respondents 14. But the FDA, in calling for new labeling, neither requested nor required Novo to amend its use code. And indeed, Novo’s counsel conceded before the Federal Circuit that Novo modified its use code in part as “ ‘a response to the [FDA’s] section viii’ ” suggestion. 601 F. 3d, at 1380–1381. 4 On remand from the Federal Circuit’s decision, the District Court determined that the ’358 patent was invalid and unenforceable. See 775 F. Supp. 2d 985 (ED Mich. 2011). The Federal Circuit stayed Novo’s appeal from that judgment pending the decision here. 5 Before proceeding to the merits, we dispose of a recently raised jurisdictional argument. Novo now contends that the federal courts lost subject-matter jurisdiction over this infringement action (including the counterclaim) at the moment Caraco filed its section viii statement. On Novo’s theory, such a statement (unlike a paragraph IV certification) does not count as an act of infringement under the patent statute, see 35 U. S. C. §271(e)(2)(A), and so cannot provide a jurisdictional basisfor the suit. But that argument is wrong even assuming (as Novo con-tends) that Caraco’s section viii filing terminated its paragraph IV certification and that a section viii filing is not an act of infringement. The want of an infringing act is a merits problem, not a jurisdictional one. Nothing in the section of the statute defining certain filings as acts of infringement suggests anything to the contrary. And “we are not inclined to interpret statutes as creating a jurisdictional bar when they are not framed as such.” Stern v. Marshall, 564 U. S. ___, ___ (2011) (slip op., at 13). In the absence of such a bar, the federal courts have jurisdiction over this suit for a single, simple reason: It “ar[ose] under a[n] Act of Congress relating to patents.” 28 U. S. C. §1338(a). 6 For this reason, we find Novo’s reliance on the occasional dictionary definition of “a[n]” unconvincing. Although “an” sometimes means “any” when used in negative structures, see, e.g., Microsoft Encarta College Dictionary 1 (2001) (fifth definition), it sometimes does not. Cf. FCC v. AT&T Inc., 562 U. S. ___, ___ (2011) (slip. op., at 3–5) (rejecting a proposed definition of “personal” because it did not always hold in ordinary usage and the statutory context suggested it did not apply). 7 Novo’s only counter is to redefine a use code. Novo argues that a use code need not be tied to the patent at all—that “[t]he relevant regulation requires [NDA] applicants to provide [only] ‘a description of each approved method of use or indication.’ ” Brief for Respondents 48 (quoting 21 CFR §314.53(c)(2)(ii)(P)(1)). Because an “indication” refers generally to what a drug does (here, treat diabetes), see §201.57(c)(2), Novo claims that a use code may sweep more broadly than the patent. But that is incorrect. First, Novo does not cite the regulations that specify the information required for publication—i.e., use codes. See §314.53(c)(2)(ii)(P)(3) (requiring a “description of the patented method of use as required for publication”); §314.53(e) (“[F]or each use patent,” FDA will publish “the approved indications or other conditions of use covered by a patent”). Those provisions (whether referring to methods of use, conditions of use, or indications) all demand a description of the patent. And second, even the provision Novo cites—which mandates the submission of additional material, not listed in the Orange Book—ties information about indications to patent coverage; that regulation requires (when quoted in full) that the brand provide “a description of each approved method of use or indication and related patent claim of the patent being submitted.” §314.53(c)(2)(ii)(P)(1). 8 See, e.g., 148 Cong. Rec. 15424 (2002) (remarks of Sen. Gregg) (“Probably the most significant issue is the fact that it creates a new cause of action”); id., at 15431–15432 (remarks of Sen. Grassley) (similar); id., at 14434 (remarks of Sen. Hatch) (similar). 9 See, e.g., Hearings on Barriers to Entry in the Pharmaceutical Marketplace before the Senate Committee on the Judiciary, 108th Cong., 1st Sess., 5–8 (2003) (statement of Daniel Troy, Chief Counsel to the FDA); id., at 19 (statement of Sen. Schumer) (“The bill provides a critical complement to the work FDA has done in clarifying its regulations on patent listing, but it goes much further”).

567.US.2011_11-204

The Fair Labor Standards Act (FLSA) requires employers to pay employees overtime wages, see 29 U. S. C. §207(a), but this requirement does not apply with respect to workers employed “in the capacity of outside salesman,” §213(a)(1). Congress did not elaborate on the meaning of “outside salesman,” but it delegated authority to the Department of Labor (DOL) to issue regulations to define the term. Three of the DOL’s regulations are relevant to this case. First, 29 CFR §541.500 defines “outside salesman” to mean “any employee . . . [w]hose primary duty is . . . making sales within the meaning of [ 29 U. S. C. §203(k)].” §§541.500(a)(1)–(2). Section 203(k), in turn, states that “ ‘[s]ale’ or ‘sell’ includes any sale, exchange, contract to sell, consignment for sale, shipment for sale, or other disposition.” Second, §541.501 clarifies that “[s]ales within the meaning of [§203(k)] include the transfer of title to tangible property.” §541.501(b). Third, §541.503 provides that promotion work that is “performed incidental to and in conjunction with an employee’s own outside sales or solicitations is exempt work,” whereas promotion work that is “incidental to sales made, or to be made, by someone else is not.” §541.503(a). The DOL provided additional guidance in connection with its promulgation of these regulations, stressing that an employee is an “outside salesman” when the employee “in some sense, has made sales.” 69 Fed. Reg. 22162. The prescription drug industry is subject to extensive federal regulation, including the requirement that prescription drugs be dispensed only upon a physician’s prescription. In light of this requirement, pharmaceutical companies have long focused their direct marketing efforts on physicians. Pharmaceutical companies promote their products to physicians through a process called “detailing,” whereby employees known as “detailers” or “pharmaceutical sales representatives” try to persuade physicians to write prescriptions for the products in appropriate cases. Petitioners were employed by respondent as pharmaceutical sales representatives for roughly four years, and during that time their primary objective was to obtain a nonbinding commitment from physicians to prescribe respondent’s products in appropriate cases. Each week, petitioners spent about 40 hours in the field calling on physicians during normal business hours and an additional 10 to 20 hours attending events and performing other miscellaneous tasks. Petitioners were not required to punch a clock or report their hours, and they were subject to only minimal supervision. Petitioners were well compensated for their efforts, and their gross pay included both a base salary and incentive pay. The amount of incentive pay was determined based on the performance of petitioners’ assigned portfolio of drugs in their assigned sales territories. It is undisputed that petitioners were not paid time-and-a-half wages when they worked more than 40 hours per week. Petitioners filed suit, alleging that respondent violated the FLSA by failing to compensate them for overtime. Respondent moved for summary judgment, arguing that petitioners were “employed in the capacity of outside salesman,” §213(a)(1), and therefore were exempt from the FLSA’s overtime compensation requirement. The District Court agreed and granted summary judgment to respondent. Petitioners filed a motion to alter or amend the judgment, contending that the District Court had erred in failing to accord controlling deference to the DOL’s interpretation of the pertinent regulations, which the DOL had announced in an amicus brief filed in a similar action. The District Court rejected this argument and denied the motion. The Ninth Circuit, agreeing that the DOL’s interpretation was not entitled to controlling deference, affirmed. Held: Petitioners qualify as outside salesmen under the most reasonable interpretation of the DOL’s regulations. Pp. 8–25. (a) The DOL filed amicus briefs in the Second Circuit and the Ninth Circuit in which it took the view that “a ‘sale’ for the purposes of the outside sales exemption requires a consummated transaction directly involving the employee for whom the exemption is sought.” Brief for Secretary of Labor as Amicus Curiae in In re Novartis Wage and Hour Litigation, No. 09–0437 (CA2), p. 11. The DOL changed course after the Court granted certiorari in this case, however, and now maintains that “[a]n employee does not make a ‘sale’ . . . unless he actually transfers title to the property at issue.” Brief for United States as Amicus Curiae 12–13. The DOL’s current interpretation of its regulations is not entitled to deference under Auer v. Robbins, 519 U.S. 452. Although Auer ordinarily calls for deference to an agency’s interpretation of its own ambiguous regulation, even when that interpretation is advanced in a legal brief, see, id., at 461–462, this general rule does not apply in all cases. Deference is inappropriate, for example, when the agency’s interpretation is “ ‘plainly erroneous or inconsistent with the regulation,’ ” id., at 461, or when there is reason to suspect that the interpretation “does not reflect the agency’s fair and considered judgment on the matter,” id., at 462. There are strong reasons for withholding Auer deference in this case. Petitioners invoke the DOL’s interpretation to impose potentially massive liability on respondent for conduct that occurred well before the interpretation was announced. To defer to the DOL’s interpretation would result in precisely the kind of “unfair surprise” against which this Court has long warned. See, e.g., Long Island Care at Home, Ltd. v. Coke, 551 U.S. 158, 170–171. Until 2009, the pharmaceutical industry had little reason to suspect that its longstanding practice of treating detailers as exempt outside salesmen transgressed the FLSA. The statute and regulations do not provide clear notice. Even more important, despite the industry’s decades-long practice, the DOL never initiated any enforcement actions with respect to detailers or otherwise suggested that it thought the industry was acting unlawfully. The only plausible explanation for the DOL’s inaction is acquiescence. Whatever the general merits of Auer deference, it is unwarranted here. The DOL’s interpretation should instead be given a measure of deference proportional to its power to persuade. See United States v. Mead Corp., 533 U.S. 218, 228. Pp. 8–14. (b) The DOL’s current interpretation—that a sale demands a transfer of title—is quite unpersuasive. It plainly lacks the hallmarks of thorough consideration. Because the DOL first announced its view that pharmaceutical sales representatives are not outside salesmen in a series of amicus briefs, there was no opportunity for public comment, and the interpretation that initially emerged from the DOL’s internal decisionmaking process proved to be untenable. The interpretation is also flatly inconsistent with the FLSA. The statute defines “sale” to mean, inter alia, a “consignment for sale,” and a “consignment for sale” does not involve the transfer of title. The DOL relies heavily on 29 CFR §541.501, which provides that “[s]ales . . . include the transfer of title to tangible property,” §541.501(b), but it is apparent that this regulation does not mean that a sale must include a transfer of title, only that transactions involving a transfer of title are included within the term “sale.” The DOL’s “explanation that obtaining a non-binding commitment to prescribe a drug constitutes promotion, and not sales,” Reply Brief for Petitioners 17, is also unconvincing. Since promotion work that is performed incidental to an employee’s own sales is exempt, the DOL’s conclusion that detailers perform only nonexempt promotion work is only as strong as the reasoning underlying its conclusion that those employees do not make sales. Pp. 14–16. (c) Because the DOL’s interpretation is neither entitled to Auer deference nor persuasive in its own right, traditional tools of interpretation must be employed to determine whether petitioners are exempt outside salesmen. Pp. 16–24. (1) The FLSA does not furnish a clear answer to this question, but it provides at least one interpretive clue by exempting anyone “employed . . . in the capacity of [an] outside salesman.” 29 U. S. C. §213(a)(1). The statute’s emphasis on “capacity” counsels in favor of a functional, rather than a formal, inquiry, one that views an employee’s responsibilities in the context of the particular industry in which the employee works. The DOL’s regulations provide additional guidance. Section 541.500 defines an outside salesman as an employee whose primary duty is “making sales” and adopts the statu- tory definition of “sale.” This statutory definition contains at least three important textual clues. First, the definition is introduced with the verb “includes,” which indicates that the examples enumerated in the text are illustrative, not exhaustive. See Burgess v. United States, 553 U.S. 124, 131, n. 3. Second, the list of transactions included in the statutory definition is modified by “any,” which, in the context of §203(k), is best read to mean “ ‘one or some indiscriminately of whatever kind,’ ” United States v. Gonzales, 520 U.S. 1, 5. Third, the definition includes the broad catchall phrase “other disposition.” Under the rule of ejusdem generis, the catchall phrase is most reasonably interpreted as including those arrangements that are tantamount, in a particular industry, to a paradigmatic sale of a commodity. Nothing in the remaining regulations requires a narrower construction. Pp. 16–20. (2) Given this interpretation of “other disposition,” it follows that petitioners made sales under the FLSA and thus are exempt outside salesmen within the meaning of the DOL’s regulations. Petitioners obtain nonbinding commitments from physicians to prescribe respondent’s drugs. This kind of arrangement, in the unique regula- tory environment within which pharmaceutical companies operate, comfortably falls within the catchall category of “other disposition.” That petitioners bear all of the external indicia of salesmen provides further support for this conclusion. And this holding also comports with the apparent purpose of the FLSA’s exemption. The exemption is premised on the belief that exempt employees normally earn salaries well above the minimum wage and perform a kind of work that is difficult to standardize to a particular time frame and that cannot easily be spread to other workers. Petitioners—each of whom earned an average of more than $70,000 per year and spent 10 to 20 hours outside normal business hours each week performing work related to his assigned portfolio of drugs in his assigned sales territory—are hardly the kind of employees that the FLSA was intended to protect. Pp. 20–22. (3) Petitioners’ remaining arguments are also unavailing. Pp. 22–24. 635 F.3d 383, affirmed. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, and Thomas, JJ., joined. Breyer, J., filed a dissenting opinion, in which Ginsburg, Sotomayor, and Kagan, JJ., joined.

The Fair Labor Standards Act (FLSA) imposes minimum wage and maximum hours requirements on employers, see 29 U. S. C. §§206–207 (2006 ed. and Supp. IV), but those requirements do not apply to workers employed “in the capacity of outside salesman,” §213(a)(1). This case requires us to decide whether the term “outside salesman,” as defined by Department of Labor (stocktickerDOL or Department) regulations, encompasses pharmaceutical sales representatives whose primary duty is to obtain nonbinding commitments from physicians to prescribe their employ- er’s prescription drugs in appropriate cases. We conclude that these employees qualify as “outside salesm[e]n.” I A Congress enacted the FLSA in 1938 with the goal of “protect[ing] all covered workers from substandard wages and oppressive working hours.” Barrentine v. Arkansas-Best Freight System, Inc., 450 U.S. 728, 739 (1981); see also 29 U. S. C. §202(a). Among other requirements, the FLSA obligates employers to compensate employees for hours in excess of 40 per week at a rate of 1½ times the employees’ regular wages. See §207(a). The overtime compensation requirement, however, does not apply with respect to all employees. See §213. As relevant here, the statute exempts workers “employed . . . in the capacity of outside salesman.” §213(a)(1).[1] Congress did not define the term “outside salesman,” but it delegated authority to the DOL to issue regulations “from time to time” to “defin[e] and delimi[t]” the term. Ibid. The DOL promulgated such regulations in 1938, 1940, and 1949. In 2004, following notice-and-comment procedures, the DOL reissued the regulations with minor amendments. See 69 Fed. Reg. 22122 (2004). The current regulations are nearly identical in substance to the regulations issued in the years immediately following the FLSA’s enactment. See 29 CFR §§541.500–541.504 (2011). Three of the DOL’s regulations are directly relevant to this case: §§541.500, 541.501, and 541.503. We refer to these three regulations as the “general regulation,” the “sales regulation,” and the “promotion-work regulation,” respectively. The general regulation sets out the definition of the statutory term “employee employed in the capacity of outside salesman.” It defines the term to mean “any employee . . . [w]hose primary duty is . . . making sales within the meaning of [ 29 U. S. C. §203(k)]”[2] and “[w]ho is customarily and regularly engaged away from the employer’s place or places of business in performing such primary duty.”[3] §§541.500(a)(1)–(2). The referenced statutory provision, 29 U. S. C. §203(k), states that “ ‘[s]ale’ or ‘sell’ includes any sale, exchange, contract to sell, consignment for sale, shipment for sale, or other disposition.” Thus, un- der the general regulation, an outside salesman is any employee whose primary duty is making any sale, exchange, contract to sell, consignment for sale, shipment for sale, or other disposition. The sales regulation restates the statutory definition of sale discussed above and clarifies that “[s]ales within the meaning of [ 29 U. S. C. §203(k)] include the transfer of title to tangible property, and in certain cases, of tangible and valuable evidences of intangible property.” 29 CFR §541.501(b). Finally, the promotion-work regulation identifies “[p]romotion work” as “one type of activity often performed by persons who make sales, which may or may not be exempt outside sales work, depending upon the circumstances under which it is performed.” §541.503(a). Promotion work that is “performed incidental to and in conjunction with an employee’s own outside sales or solicitations is exempt work,” whereas promotion work that is “incidental to sales made, or to be made, by someone else is not exempt outside sales work.” Ibid. Additional guidance concerning the scope of the outside salesman exemption can be gleaned from reports issued in connection with the DOL’s promulgation of regulations in 1940 and 1949, and from the preamble to the 2004 regulations. See Dept. of Labor, Wage and Hour Division, Report and Recommendations of the Presiding Officer at Hearings Preliminary to Redefinition (1940) (hereinafter 1940 Report); Dept. of Labor, Wage and Hour Division, Report and Recommendations on Proposed Revisions of Regulations, Part 541 (1949) (hereinafter 1949 Report); 69 Fed. Reg. 22160–22163 (hereinafter Preamble). Although the DOL has rejected proposals to eliminate or dilute the requirement that outside salesmen make their own sales, the Department has stressed that this requirement is met whenever an employee “in some sense make[s] a sale.” 1940 Report 46; see also Preamble 22162 (reiterating that the exemption applies only to an employee who “in some sense, has made sales”). And the DOL has made it clear that “[e]xempt status should not depend” on technicalities, such as “whether it is the sales employee or the customer who types the order into a computer system and hits the return button,” Preamble 22163, or whether “the order is filled by [a] jobber rather than directly by [the employee’s] own employer,” 1949 Report 83. B Respondent SmithKline Beecham Corporation is in the business of developing, manufacturing, and selling prescription drugs. The prescription drug industry is subject to extensive federal regulation, including the now-familiar requirement that prescription drugs be dispensed only upon a physician’s prescription.[4] In light of this requirement, pharmaceutical companies have long focused their direct marketing efforts, not on the retail pharmacies that dispense prescription drugs, but rather on the medi- cal practitioners who possess the authority to prescribe the drugs in the first place. Pharmaceutical companies promote their prescription drugs to physicians through a process called “detailing,” whereby employees known as “detailers” or “pharmaceutical sales representatives” provide information to physicians about the company’s products in hopes of persuading them to write prescriptions for the products in appropriate cases. See Sorrell v. IMS Health Inc., 564 U. S. ___, ___ (2011) (slip op., at 1–2) (describing the process of “detailing”). The position of “detailer” has existed in the pharmaceutical industry in substantially its current form since at least the 1950’s, and in recent years the industry has employed more than 90,000 detailers nationwide. See 635 F.3d 383, 387, and n. 5, 396 (CA9 2011). Respondent hired petitioners Michael Christopher and Frank Buchanan as pharmaceutical sales representatives in 2003. During the roughly four years when petitioners were employed in that capacity,[5] they were responsible for calling on physicians in an assigned sales territory to discuss the features, benefits, and risks of an assigned portfolio of respondent’s prescription drugs. Petitioners’ primary objective was to obtain a nonbinding commitment[6] from the physician to prescribe those drugs in appropriate cases, and the training that petitioners received underscored the importance of that objective. Petitioners spent about 40 hours each week in the field calling on physicians. These visits occurred during normal business hours, from about 8:30 a.m. to 5 p.m. Outside of normal business hours, petitioners spent an additional 10 to 20 hours each week attending events, reviewing product information, returning phone calls, responding to e-mails, and performing other miscellaneous tasks. Petitioners were not required to punch a clock or report their hours, and they were subject to only minimal supervision. Petitioners were well compensated for their efforts. On average, Christopher’s annual gross pay was just over $72,000, and Buchanan’s was just over $76,000.[7] Petitioners’ gross pay included both a base salary and incentive pay. The amount of petitioners’ incentive pay was based on the sales volume or market share of their assigned drugs in their assigned sales territories,[8] and this amount was uncapped. Christopher’s incentive pay exceeded 30 percent of his gross pay during each of his years of employment; Buchanan’s exceeded 25 percent. It is undisputed that respondent did not pay petitioners time-and-a-half wages when they worked in excess of 40 hours per week. C Petitioners brought this action in the United States District Court for the District of Arizona under 29 U. S. C. §216(b). Petitioners alleged that respondent violated the FLSA by failing to compensate them for overtime, and they sought both backpay and liquidated damages as relief. Respondent moved for summary judgment, arguing that petitioners were “employed . . . in the capacity of outside salesman,” §213(a)(1), and therefore were exempt from the FLSA’s overtime compensation requirement.[9] The District Court agreed and granted summary judgment to respondent. See App. to Pet. for Cert. 37a–47a. After the District Court issued its order, petitioners filed a motion to alter or amend the judgment, contending that the District Court had erred in failing to accord control- ling deference to the DOL’s interpretation of the pertinent regulations. That interpretation had been announced in an uninvited amicus brief filed by the DOL in a similar action then pending in the Second Circuit. See Brief for Secretary of Labor as Amicus Curiae in In re Novartis Wage and Hour Litigation, No. 09–0437 (hereinafter Secretary’s Novartis Brief). The District Court rejected this argument and denied the motion. See App. to Pet. for Cert. 48a–52a. The Court of Appeals for the Ninth Circuit affirmed. See 635 F.3d 383. The Court of Appeals agreed that the DOL’s interpretation[10] was not entitled to controlling deference. See id., at 393–395. It held that, because the commitment that petitioners obtained from physicians was the maximum possible under the rules applicable to the pharmaceutical industry, petitioners made sales within the meaning of the regulations. See id., at 395–397. The court found it significant, moreover, that the DOL had previously interpreted the regulations as requiring only that an employee “ ‘in some sense’ ” make a sale, see id., at 395–396 (emphasis deleted), and had “acquiesce[d] in the sales practices of the drug industry for over seventy years,” id., at 399. The Ninth Circuit’s decision conflicts with the Second Circuit’s decision in In re Novartis Wage and Hour Litigation, 611 F.3d 141, 153–155 (2010) (holding that the DOL’s interpretation is entitled to controlling deference). We granted certiorari to resolve this split, 565 U. S. ___ (2011), and we now affirm the judgment of the Ninth Circuit. II We must determine whether pharmaceutical detailers are outside salesmen as the DOL has defined that term in its regulations. The parties agree that the regulations themselves were validly promulgated and are therefore entitled to deference under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). But the parties disagree sharply about whether the DOL’s interpretation of the regulations is owed deference under Auer v. Robbins, 519 U.S. 452 (1997). It is to that question that we now turn. A The DOL first announced its view that pharmaceutical detailers are not exempt outside salesmen in an amicus brief filed in the Second Circuit in 2009, and the Department has subsequently filed similar amicus briefs in other cases, including the case now before us.[11] While the DOL’s ultimate conclusion that detailers are not exempt has remained unchanged since 2009, the same cannot be said of its reasoning. In both the Second Circuit and the Ninth Circuit, the DOL took the view that “a ‘sale’ for the purposes of the outside sales exemption requires a con- summated transaction directly involving the employee for whom the exemption is sought.” Secretary’s Novartis Brief 11; see also Brief for Secretary of Labor as Amicus Curiae in No. 10–15257 (CA9), p. 12. Perhaps because of the nebulous nature of this “consummated transaction” test,[12] the Department changed course after we granted certiorari in this case. The Department now takes the position that “[a]n employee does not make a ‘sale’ for purposes of the ‘outside salesman’ exemption unless he actually transfers title to the property at issue.” Brief for United States as Amicus Curiae 12–13 (hereinafter U. S. Brief).[13] Petitioners and the DOL assert that this new interpretation of the regulations is entitled to controlling deference. See Brief for Petitioners 31–42; U. S. Brief 30–34.[14] Although Auer ordinarily calls for deference to an agency’s interpretation of its own ambiguous regulation, even when that interpretation is advanced in a legal brief, see Chase Bank USA, N. A. v. McCoy, 562 U. S. ___, ___ (2011) (slip op., at 12); Auer, 519 U. S., at 461–462, this general rule does not apply in all cases. Deference is undoubtedly inappropriate, for example, when the agency’s interpretation is “ ‘plainly erroneous or inconsistent with the regulation.’ ” Id., at 461 (quoting Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 359 (1989)). And deference is likewise unwarranted when there is reason to suspect that the agency’s interpretation “does not reflect the agency’s fair and considered judgment on the matter in question.” Auer, supra, at 462; see also, e.g., Chase Bank, supra, at ___ (slip op., at 14). This might occur when the agency’s interpretation conflicts with a prior interpretation, see, e.g., Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 515 (1994), or when it appears that the interpretation is nothing more than a “convenient litigating position,” Bowen v. Georgetown Univ. Hospital, 488 U.S. 204, 213 (1988), or a “ ‘post hoc rationalizatio[n]’ advanced by an agency seeking to defend past agency action against attack,” Auer, supra, at 462 (quoting Bowen, supra, at 212; alteration in original). In this case, there are strong reasons for withholding the deference that Auer generally requires. Petitioners invoke the DOL’s interpretation of ambiguous regulations to impose potentially massive liability on respondent for conduct that occurred well before that interpretation was announced. To defer to the agency’s interpretation in this circumstance would seriously undermine the principle that agencies should provide regulated parties “fair warning of the conduct [a regulation] prohibits or requires.” Gates & Fox Co. v. Occupational Safety and Health Review Comm’n, 790 F.2d 154, 156 (CADC 1986) (Scalia, J.).[15] Indeed, it would result in precisely the kind of “unfair surprise” against which our cases have long warned. See Long Island Care at Home, Ltd. v. Coke, 551 U.S. 158, 170–171 (2007) (deferring to new interpretation that “create[d] no unfair surprise” because agency had pro- ceeded through notice-and-comment rulemaking); Martin v. Occupational Safety and Health Review Comm’n, 499 U.S. 144, 158 (1991) (identifying “adequacy of notice to regulated parties” as one factor relevant to the reasonableness of the agency’s interpretation); NLRB v. Bell Aerospace Co., 416 U.S. 267, 295 (1974) (suggesting that an agency should not change an interpretation in an adjudicative proceeding where doing so would impose “new liability . . . on individuals for past actions which were taken in good-faith reliance on [agency] pronouncements” or in a case involving “fines or damages”). This case well illustrates the point. Until 2009, the pharmaceutical industry had little reason to suspect that its longstanding practice of treating detailers as exempt outside salesmen transgressed the FLSA. The statute and regulations certainly do not provide clear notice of this. The general regulation adopts the broad statutory definition of “sale,” and that definition, in turn, employs the broad catchall phrase “other disposition.” See 29 CFR §541.500(a)(1). This catchall phrase could reasonably be construed to encompass a nonbinding commitment from a physician to prescribe a particular drug, and nothing in the statutory or regulatory text or the DOL’s prior guidance plainly requires a contrary reading. See Preamble 22162 (explaining that an employee must “in some sense” make a sale); 1940 Report 46 (same). Even more important, despite the industry’s decades-long practice of classifying pharmaceutical detailers as exempt employees, the DOL never initiated any enforcement actions with respect to detailers or otherwise suggested that it thought the industry was acting unlawfully.[16] We acknowledge that an agency’s enforcement decisions are informed by a host of factors, some bearing no relation to the agency’s views regarding whether a violation has occurred. See, e.g., Heckler v. Chaney, 470 U.S. 821, 831 (1985) (noting that “an agency decision not to enforce often involves a complicated balancing of a number of factors which are peculiarly within its expertise”). But where, as here, an agency’s announcement of its interpretation is preceded by a very lengthy period of conspicuous inaction, the potential for unfair surprise is acute. As the Seventh Circuit has noted, while it may be “possible for an entire industry to be in violation of the [FLSA] for a long time without the Labor Department noticing,” the “more plausible hypothesis” is that the Department did not think the industry’s practice was un- lawful. Yi v. Sterling Collision Centers, Inc., 480 F.3d 505, 510–511 (2007). There are now approximately 90,000 pharmaceutical sales representatives; the nature of their work has not materially changed for decades and is well known; these employees are well paid; and like quintessential outside salesmen, they do not punch a clock and often work more than 40 hours per week. Other than acquiescence, no explanation for the DOL’s inaction is plausible. Our practice of deferring to an agency’s interpretation of its own ambiguous regulations undoubtedly has important advantages,[17] but this practice also creates a risk that agencies will promulgate vague and open-ended regulations that they can later interpret as they see fit, thereby “frustrat[ing] the notice and predictability purposes of rulemaking.” Talk America, Inc. v. Michigan Bell Telephone Co., 564 U. S. ___, ___ (2011) (Scalia, J., concurring) (slip op., at 3); see also Stephenson & Pogoriler, Seminole Rock’s Domain, 79 Geo. Wash. L. Rev. 1449, 1461–1462 (2011); Manning, Constitutional Structure and Judicial Deference to Agency Interpretations of Agency Rules, 96 Colum. L. Rev. 612, 655–668 (1996). It is one thing to expect regulated parties to conform their conduct to an agency’s interpretations once the agency announces them; it is quite another to require regulated parties to divine the agency’s interpretations in advance or else be held liable when the agency announces its interpretations for the first time in an enforcement proceeding and demands deference. Accordingly, whatever the general merits of Auer deference, it is unwarranted here. We instead accord the Department’s interpretation a measure of deference proportional to the “ ‘thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade.’ ” United States v. Mead Corp., 533 U.S. 218, 228 (2001) (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944)). B We find the DOL’s interpretation of its regulations quite unpersuasive. The interpretation to which we are now asked to defer—that a sale demands a transfer of title—plainly lacks the hallmarks of thorough consideration. Because the DOL first announced its view that pharmaceutical sales representatives do not qualify as outside salesmen in a series of amicus briefs, there was no opportunity for public comment, and the interpretation that initially emerged from the Department’s internal decisionmaking process proved to be untenable. After arguing successfully in the Second Circuit and then unsucess- fully in the Ninth Circuit that a sale for present purposes simply requires a “consummated transaction,” the DOL advanced a different interpretation in this Court. Here, the DOL’s brief states unequivocally that “[a]n employee does not make a ‘sale’ for purposes of the ‘outside salesman’ exemption unless he actually transfers title to the property at issue.” U. S. Brief 12–13. This new interpretation is flatly inconsistent with the FLSA, which defines “sale” to mean, inter alia, a “consignment for sale.” A “consignment for sale” does not involve the transfer of title. See, e.g., Sturm v. Boker, 150 U.S. 312, 330 (1893) (“The agency to sell and return the proceeds, or the specific goods if not sold . . . does not involve a change of title”); Hawkland, Consignment Selling Under the Uniform Commercial Code, 67 Com. L. J. 146, 147 (1962) (explaining that “ ‘[a] consignment of goods for sale does not pass the title at any time, nor does it contemplate that it should be passed’ ” (quoting Rio Grande Oil Co. v. Miller Rubber Co. of N. Y., 31 Ariz. 84, 87, 250 P. 564, 565 (1926))). The DOL cannot salvage its interpretation by arguing that a “consignment for sale” may eventually result in the transfer of title (from the consignor to the ultimate purchaser if the consignee in fact sells the good). Much the same may be said about a physician’s nonbinding commitment to prescribe a particular product in an appropriate case. In that situation, too, agreement may eventually result in the transfer of title (from the manufacturer to a pharmacy and ultimately to the patient for whom the drug is prescribed). In support of its new interpretation, the DOL relies heavily on its sales regulation, which states in part that “[s]ales [for present purposes] include the transfer of title to tangible property,” 29 CFR §541.501(b) (emphasis added). This regulation, however, provides little support for the DOL’s position. The DOL reads the sales regulation to mean that a “sale” necessarily includes the transfer of title, but that is not what the regulation says. And it seems clear that that is not what the regulation means. The sentence just subsequent to the one on which the DOL relies, echoing the terms of the FLSA, makes clear that a “consignment for sale” qualifies as a sale. Since a consignment for sale does not involve a transfer of title, it is apparent that the sales regulation does not mean that a sale must include a transfer of title, only that transactions involving a transfer of title are included within the term “sale.” Petitioners invite us to look past the DOL’s “determination that a sale must involve the transfer of title” and instead defer to the Department’s “explanation that obtaining a non-binding commitment to prescribe a drug constitutes promotion, and not sales.” Reply Brief for Petitioners 17. The problem with the DOL’s interpretation of the promotion-work regulation, however, is that it depends almost entirely on the DOL’s flawed transfer-of-title interpretation. The promotion-work regulation does not distinguish between promotion work and sales; rather, it distinguishes between exempt promotion work and nonexempt promotion work. Since promotion work that is performed incidental to an employee’s own sales is exempt, the DOL’s conclusion that pharmaceutical detailers perform only nonexempt promotion work is only as strong as the reasoning underlying its conclusion that those employees do not make sales. For the reasons already discussed, we find this reasoning wholly unpersuasive. In light of our conclusion that the DOL’s interpretation is neither entitled to Auer deference nor persuasive in its own right, we must employ traditional tools of interpretation to determine whether petitioners are exempt outside salesmen. C 1 We begin with the text of the FLSA. Although the provision that establishes the overtime salesman exemption does not furnish a clear answer to the question before us, it provides at least one interpretive clue: It exempts anyone “employed . . . in the capacity of [an] outside salesman.” 29 U. S. C. §213(a)(1) (emphasis added). “Capacity,” used in this sense, means “[o]utward condition or circumstances; relation; character; position.” Webster’s New International Dictionary 396 (2d ed. 1934); see also 2 Oxford English Dictionary 89 (def. 9) (1933) (“Position, condition, character, relation”). The statute’s emphasis on the “capacity” of the employee counsels in favor of a functional, rather than a formal, inquiry, one that views an employee’s responsibilities in the context of the particular industry in which the employee works. The DOL’s regulations provide additional guidance. The general regulation defines an outside salesman as an employee whose primary duty is “making sales,” and it adopts the statutory definition of “sale.” 29 CFR §541.500(a)(1)(i). This definition contains at least three important textual clues. First, the definition is introduced with the verb “includes” instead of “means.” This word choice is significant because it makes clear that the examples enumerated in the text are intended to be illustrative, not exhaustive. See Burgess v. United States, 553 U.S. 124, 131, n. 3 (2008) (explaining that “[a] term whose statutory definition declares what it ‘includes’ is more susceptible to extension of meaning . . . than where . . . the definition declares what a term ‘means’ ” (alteration in original; some internal quotation marks omitted)). Indeed, Congress used the narrower word “means” in other provisions of the FLSA when it wanted to cabin a definition to a specific list of enumerated items. See, e.g., 29 U. S. C. §203(a) (“ ‘Person’ means an individual, partnership, association, corporation, business trust, legal representative, or any organized group of persons” (emphasis added)). Second, the list of transactions included in the statu- tory definition of sale is modified by the word “any.” We have recognized that the modifier “any” can mean “different things depending upon the setting,” Nixon v. Missouri Municipal League, 541 U.S. 125, 132 (2004), but in the context of 29 U. S. C. §203(k), it is best read to mean “ ‘one or some indiscriminately of whatever kind,’ ” United States v. Gonzales, 520 U.S. 1, 5 (1997) (quoting Webster’s Third New International Dictionary 97 (1976)). That is so because Congress defined “sale” to include both the unmodified word “sale” and transactions that might not be considered sales in a technical sense, including exchanges and consignments for sale.[18] Third, Congress also included a broad catchall phrase: “other disposition.” Neither the statute nor the regulations define “disposition,” but dictionary definitions of the term range from “relinquishment or alienation” to “arrangement.” See Webster’s New International Dictionary 644 (def. 1(b)) (1927) (“[t]he getting rid, or making over, of anything; relinquishment or alienation”); ibid. (def. 1(a)) (“[t]he ordering, regulating, or administering of anything”); 3 Oxford English Dictionary, supra, at 493 (def. 4) (“[t]he action of disposing of, putting away, getting rid of, making over, etc.”); ibid. (def. 1) (“[t]he action of setting in order, or condition of being set in order; arrangement, order”). We agree with the DOL that the rule of ejusdem generis should guide our interpretation of the catchall phrase, since it follows a list of specific items.[19] But the limit the DOL posits, one that would confine the phrase to dispositions involving “contract[s] for the exchange of goods or services in return for value,” see U. S. Brief 20, is much too narrow, as is petitioners’ view that a sale requires a “firm agreement” or “firm commitment” to buy, see Tr. of Oral Arg. 64, 66. These interpretations would defeat Congress’ intent to define “sale” in a broad manner and render the general statutory language meaningless. See United States v. Alpers, 338 U.S. 680, 682 (1950) (instructing that rule of ejusdem generis cannot be employed to “obscure and defeat the intent and purpose of Congress” or “render general words meaningless”). Indeed, we are hard pressed to think of any contract for the exchange of goods or services in return for value or any firm agreement to buy that would not also fall within one of the specifically enumerated categories.[20] The specific list of transactions that precedes the phrase “other disposition” seems to us to represent an attempt to accommodate industry-by-industry variations in methods of selling commodities. Consequently, we think that the catchall phrase “other disposition” is most reasonably interpreted as including those arrangements that are tantamount, in a particular industry, to a paradigmatic sale of a commodity. Nothing in the remaining regulations requires a narrower construction.[21] As discussed above, the sales regulation instructs that sales within the meaning of 29 U. S. C. §203(k) “include the transfer of title to tangible property,” 29 CFR §541.501(b) (emphasis added), but this regulation in no way limits the broad statutory definition of “sale.” And although the promotion-work regulation distinguishes between promotion work that is incidental to an employee’s own sales and work that is incidental to sales made by someone else, see §541.503(a), this distinction tells us nothing about the meaning of “sale.”[22] 2 Given our interpretation of “other disposition,” it follows that petitioners made sales for purposes of the FLSA and therefore are exempt outside salesmen within the meaning of the DOL’s regulations. Obtaining a nonbinding commitment from a physician to prescribe one of respondent’s drugs is the most that petitioners were able to do to ensure the eventual disposition of the products that respondent sells.[23] This kind of arrangement, in the unique regulatory environment within which pharmaceutical companies must operate, comfortably falls within the catch- all category of “other disposition.” That petitioners bear all of the external indicia of salesmen provides further support for our conclusion. Petitioners were hired for their sales experience. They were trained to close each sales call by obtaining the maximum commitment possible from the physician. They worked away from the office, with minimal supervision, and they were rewarded for their efforts with incentive compensation. It would be anomalous to require respondent to compensate petitioners for overtime, while at the same time exempting employees who function identically to petitioners in every respect except that they sell physician-administered drugs, such as vaccines and other inject- able pharmaceuticals, that are ordered by the physician directly rather than purchased by the end user at a pharmacy with a prescription from the physician. Our holding also comports with the apparent purpose of the FLSA’s exemption for outside salesmen. The exemption is premised on the belief that exempt employees “typically earned salaries well above the minimum wage” and enjoyed other benefits that “se[t] them apart from the nonexempt workers entitled to overtime pay.” Preamble 22124. It was also thought that exempt employees performed a kind of work that “was difficult to standardize to any time frame and could not be easily spread to other workers after 40 hours in a week, making compliance with the overtime provisions difficult and generally precluding the potential job expansion intended by the FLSA’s time-and-a-half overtime premium.” Ibid. Petitioners—each of whom earned an average of more than $70,000 per year and spent between 10 and 20 hours outside normal business hours each week performing work related to his assigned portfolio of drugs in his assigned sales territory—are hardly the kind of employees that the FLSA was intended to protect. And it would be challenging, to say the least, for pharmaceutical companies to compensate detailers for overtime going forward without significantly changing the nature of that position. See, e.g., Brief for PhRMA as Amicus Curiae 14–20 (explaining that “key aspects of [detailers’] jobs as they are currently structured are fundamentally incompatible with treating [detailers] as hourly employees”). 3 The remaining arguments advanced by petitioners and the dissent are unavailing. Petitioners contend that detailers are more naturally classified as nonexempt promotional employees who merely stimulate sales made by others than as exempt outside salesmen. They point out that respondent’s prescription drugs are not actually sold until distributors and retail pharmacies order the drugs from other employees. See Reply Brief for Petitioners 7. Those employees,[24] they reason, are the true salesmen in the industry, not detailers. This formalistic argument is inconsistent with the realistic approach that the outside salesman exemption is meant to reflect. Petitioners’ theory seems to be that an employee is properly classified as a nonexempt promotional employee whenever there is another employee who actually makes the sale in a technical sense. But, taken to its extreme, petitioners’ theory would require that we treat as a nonexempt promotional employee a manufacturer’s representative who takes an order from a retailer but then transfers the order to a jobber’s employee to be filled, or a car salesman who receives a commitment to buy but then asks his or her assistant to enter the order into the computer. This formalistic approach would be difficult to reconcile with the broad language of the regulations and the statutory definition of “sale,” and it is in significant ten- sion with the DOL’s past practice. See 1949 Report 83 (explaining that the manufacturer’s representative was clearly “performing sales work regardless of the fact that the order is filled by the jobber rather than directly by his own employer”); Preamble 22162 (noting that “technological changes in how orders are taken and processed should not preclude the exemption for employees who in some sense make the sales”). Petitioners additionally argue that detailers are the functional equivalent of employees who sell a “concept,” and they point to Wage and Hour Division opinion letters, as well as lower court decisions, deeming such employees nonexempt. See Brief for Petitioners 47–48. Two of these opinions, however, concerned employees who were more analogous to buyers than to sellers. See Clements v. Serco, Inc., 530 F.3d 1224, 1229–1230, n. 4 (CA10 2008) (explaining that, although military recruiters “[i]n a loose sense” were “selling the Army’s services,” it was the Army that would “pa[y] for the services of the recruits who enlist”); Opinion Letter from Dept. of Labor, Wage and Hour Division (Aug. 19, 1994), 1994 WL 1004855 (explaining that selling the “concept” of organ donation “is similar to that of outside buyers who in a very loose sense are sometimes described as selling their employer’s ‘service’ to the person for whom they obtain their goods”). And the other two opinions are likewise inapposite. One concerned employees who were not selling a good or service at all, see Opinion Letter from Dept. of Labor, Wage and Hour Division (May 22, 2006), 2006 WL 1698305 (concluding that employees who solicit charitable contributions are not exempt), and the other concerned employees who were incapable of selling any good or service because their employer had yet to extend an offer, see Opinion Letter from Dept. of Labor, Wage and Hour Division (Apr. 20, 1999), 1999 WL 1002391 (concluding that college recruiters are not exempt because they merely induce qualified customers to apply to the college, and the college “in turn decides whether to make a contractual offer of its educational services to the applicant”). Finally, the dissent posits that the “primary duty” of a pharmaceutical detailer is not “to obtain a promise to prescribe a particular drug,” but rather to “provid[e] information so that the doctor will keep the drug in mind with an eye toward using it when appropriate.” Post, at 6. But the record in this case belies that contention. Petitioners’ end goal was not merely to make physicians aware of the medically appropriate uses of a particular drug. Rather, it was to convince physicians actually to prescribe the drug in appropriate cases. See App. to Pet. for Cert. 40a (finding that petitioners’ “primary objective was convincing physicians to prescribe [respondent’s] products to their patients”). * * * For these reasons, we conclude that petitioners qualify as outside salesmen under the most reasonable interpretation of the DOL’s regulations. The judgment of the Court of Appeals is Affirmed. Notes 1 This provision also exempts workers “employed in a bona fide executive, administrative, or professional capacity.” 29 U. S. C. §213(a)(1). 2 The definition also includes any employee “[w]hose primary duty is . . . obtaining orders or contracts for services or for the use of facilities for which a consideration will be paid by the client or customer.” 29 CFR §541.500(a)(1)(ii). That portion of the definition is not at issue in this case. 3 It is undisputed that petitioners were “customarily and regularly engaged away” from respondent’s place of business in performing their responsibilities. 4 Congress imposed this requirement in 1951 when it amended the Federal Food, Drug, and Cosmetic Act (FDCA) to provide that drugs that are “not safe for use except under the supervision of a practitioner” may be dispensed “only . . . upon a . . . prescription of a practitioner licensed by law to administer such drug.” Durham-Humphrey Amendment of 1951, ch. 578, 65Stat. 648–649 (codified at 21 U. S. C. §353(b)). As originally enacted in 1938, the FDCA allowed manufacturers to designate certain drugs as prescription only, but “it did not say which drugs were to be sold by prescription or that there were any drugs that could not be sold without a prescription.” Temin, The Origin of Compulsory Drug Prescriptions, 22 J. Law & Econ. 91, 98 (1979). Prior to Congress’ enactment of the FDCA, a prescription was not needed to obtain any drug other than certain narcotics. See id., at 97. 5 Respondent terminated Christopher’s employment in 2007, and Buchanan left voluntarily the same year to accept a similar position with another pharmaceutical company. 6 The parties agree that the commitment is nonbinding. 7 The median pay for pharmaceutical detailers nationwide exceeds $90,000 per year. See Brief for Respondent 14. 8 The amount of incentive pay is not formally tied to the number of prescriptions written or commitments obtained, but because retail pharmacies are prohibited from dispensing prescription drugs without a physician’s prescription, retail sales of respondent’s products necessarily reflect the number of prescriptions written. 9 Respondent also argued that petitioners were exempt administrative employees. The District Court and the Court of Appeals found it unnecessary to reach that argument, and the question is not before us. 10 The DOL filed an amicus brief in the Ninth Circuit advancing substantially the same interpretation it had advanced in its brief in the Second Circuit. See Brief for Secretary of Labor as Amicus Curiae in No. 10–15257. 11 The DOL invites “interested parties to inform it of private cases involving the misclassification of employees in contravention of the new Part 541 rule” so that it may file amicus briefs “in appropriate cases to share with courts the Department’s view of the proper application of the new Part 541 rule.” See Dept. of Labor, Office of Solicitor, Overtime Security Amicus Program, http://www.dol.gov/sol/541amicus.htm (as visited June 15, 2012, and available in Clerk of Court’s case file). 12 For example, it is unclear why a physician’s nonbinding commitment to prescribe a drug in an appropriate case cannot qualify as a sale under this test. The broad term “transaction” easily encompasses such a commitment. See Webster’s Third New International Dictionary 2425 (2002) (hereinafter Webster’s Third) (defining “transaction” to mean “a communicative action or activity involving two parties or two things reciprocally affecting or influencing each other”). A “consummated transaction” is simply a transaction that has been fully completed. See id., at 490 (defining “consummate” to mean “to bring to completion”). And a pharmaceutical sales representative who obtains such a commitment is “directly involv[ed]” in this transaction. Thus, once a pharmaceutical sales representative and a physician have fully completed their agreement, it may be said that they have entered into a “consummated transaction.” 13 When pressed to clarify its position at oral argument, the DOL suggested that a “transfer of possession in contemplation of a transfer of title” might also suffice. Tr. of Oral Arg. 17. 14 Neither petitioners nor the DOL asks us to accord controlling deference to the “consummated transaction” interpretation the Department advanced in its briefs in the Second Circuit and Ninth Circuit, nor could we given that the Department has now abandoned that interpretation. See Estate of Cowart v. Nicklos Drilling Co., 505 U.S. 469, 480 (1992) (noting that “it would be quite inappropriate to defer to an interpretation which has been abandoned by the policymaking agency itself ”). 15 Accord, Phelps Dodge Corp. v. Federal Mine Safety and Health Review Comm’n, 681 F.2d 1189, 1192 (CA9 1982) (recognizing that “the application of a regulation in a particular situation may be challenged on the ground that it does not give fair warning that the allegedly violative conduct was prohibited”); Kropp Forge Co. v. Secretary of Labor, 657 F.2d 119, 122 (CA7 1981) (refusing to impose sanctions where standard the regulated party allegedly violated “d[id] not provide ‘fair warning’ of what is required or prohibited”); Dravo Corp. v. Occupational Safety and Health Review Comm’n, 613 F.2d 1227, 1232–1233 (CA3 1980) (rejecting agency’s expansive interpretation where agency did not “state with ascertainable certainty what is meant by the standards [it] ha[d] promulgated” (internal quotation marks omitted and emphasis deleted)); Diamond Roofing Co. v. Occupational Safety and Health Review Comm’n, 528 F.2d 645, 649 (CA5 1976) (explaining that “statutes and regulations which allow monetary penalties against those who violate them” must “give an employer fair warning of the conduct [they] prohibi[t] or requir[e]”); 1 R. Pierce, Administrative Law Treatise §6.11, p. 543 (5th ed. 2010) (observing that “[i]n penalty cases, courts will not accord substantial deference to an agency’s interpretation ofan ambiguous rule in circumstances where the rule did not place the individual or firm on notice that the conduct at issue constituted a violation of a rule”). 16 It appears that the DOL only once directly opined on the exempt status of detailers prior to 2009. In 1945, the Wage and Hour Division issued an opinion letter tentatively concluding that “medical detailists” who performed “work . . . aimed at increasing the use of [their employer’s] product in hospitals and through physicians’ recommendations” qualified as administrative employees. Opinion Letter from Dept. of Labor, Wage and Hour Division (May 19, 1945), 1 CCH Labor Law Service, Federal Wage-Hour Guide ¶33,093. But that letter did not address the outside salesman exemption. 17 For instance, it “makes the job of a reviewing court much easier, and since it usually produces affirmance of the agency’s view without conflict in the Circuits, it imparts (once the agency has spoken to clarify the regulation) certainty and predictability to the administrative process.” Talk America, Inc. v. Michigan Bell Telephone Co., 564 U. S. ___, ___ (2011) (Scalia, J., concurring) (slip op., at 3). 18 Given that the FLSA provides its own definition of “sale” that is more expansive than the term’s ordinary meaning, the DOL’s reliance on dictionary definitions of the word “sale” is misplaced. See, e.g., Burgess v. United States, 553 U.S. 124, 130 (2008) (noting that “[w]hen a statute includes an explicit definition, we must follow that definition” (internal quotation marks omitted)). 19 The canon of ejusdem generis “limits general terms [that] follow specific ones to matters similar to those specified.” CSX Transp., Inc. v. Alabama Dept. of Revenue, 562 U. S. ___, ___ (2011) (slip op., at 16) (alteration in original; internal quotation marks omitted). 20 The dissent’s approach suffers from the same flaw. The dissent contends that, in order to make a sale, an employee must at least obtain a “firm commitment to buy.” Post, at 10 (opinion of Breyer, J.). But when an employee who has extended an offer to sell obtains a “firm commitment to buy,” that transaction amounts to a “contract to sell.” Given that a “contract to sell” already falls within the statutory definition of “sale,” the dissent’s interpretation would strip the catchall phrase of independent meaning. 21 In the past, we have stated that exemptions to the FLSA must be “narrowly construed against the employers seeking to assert them and their application limited to those [cases] plainly and unmistakably within their terms and spirit.” Arnold v. Ben Kanowsky, Inc., 361 U.S. 388, 392 (1960). Petitioners and the DOL contend that Arnold requires us to construe the outside salesman exemption narrowly, but Arnold is inapposite where, as here, we are interpreting a general definition that applies throughout the FLSA. 22 The dissent’s view that pharmaceutical detailers are more naturally characterized as nonexempt promotional employees than as exempt outside salesmen relies heavily on the DOL’s explanation in its 1940 Re-port that “sales promotion men” are not salesmen. See post, at 7; see also 1940 Report 46. There, the Department described a “sales promotion man” as an employee who merely “pav[es] the way for salesmen” and who frequently “deals with retailers who are not customers of his own employer but of his employer’s customer” and is “interested in sales by the retailer, not to the retailer.” 1940 Report 46. The dissent asserts that detailers are analogous to “sales promotion men” because they deal with “individuals, namely doctors, ‘who are not customers’ of their own employer” and “are primarily interested in sales authorized by the doctor, not to the doctor.” Post, at 7. But this comparison is inapt. The equivalent of a “sales promotion man” in the pharmaceutical industry would be an employee who promotes a manufacturer’s products to the retail pharmacies that sell the products after purchasing them from a wholesaler or distributor. Detailers, by contrast, obtain nonbinding commitments from the gatekeepers who must prescribe the product if any sale is to take place at all. 23 Our point is not, as the dissent suggests, that any employee who does the most that he or she is able to do in a particular position to ensure the eventual sale of a product should qualify as an exempt outside salesman. See post, at 9 (noting that “the ‘most’ a California firm’s marketing employee may be able ‘to do’ to secure orders from New York customers is to post an advertisement on the Internet”). Rather, our point is that, when an entire industry is constrained by law or regulation from selling its products in the ordinary manner, an employee who functions in all relevant respects as an outside salesman should not be excluded from that category based on technicalities. 24 According to one of respondent’s amici, most pharmaceutical companies “have systems in place to maintain the inventories of wholesalers and retailers of prescription drugs (consisting mainly of periodic restocking pursuant to a general contract), [and] these systems are largely ministerial and require only a few employees to administer them.” Brief for PhRMA as Amicus Curiae 24; see also ibid. (explaining that one of its members employs more than 2,000 pharmaceutical sales representatives but “fewer than ten employees who are responsible for processing orders from retailers and wholesalers, a ratio that is typical of how the industry is structured”).

566.US.2011_10-1016

The Family and Medical Leave Act of 1993 (FMLA) entitles an employee to take up to 12 work weeks of unpaid leave per year for (A) the care of a newborn son or daughter; (B) the adoption or foster-care placement of a child; (C) the care of a spouse, son, daughter, or parent with a serious medical condition; and (D) the employee’s own serious health condition when the condition interferes with the employee’s ability to perform at work. 29 U. S. C. §2612(a)(1). The FMLA also creates a private right of action for equitable relief and damages “against any employer (including a public agency) in any Federal or State court.” §2617(a)(2). For present purposes, subparagraphs (A), (B), and (C) are referred to as the family-care provisions, and subparagraph (D) as the self-care provision. In Nevada Dept. of Human Resources v. Hibbs, 538 U.S. 721, 730−732, this Court held that Congress could subject States to suit for violations of subparagraph (C) based on evidence of family-leave policies that discriminated on the basis of sex. Petitioner filed suit, alleging that his employer, the Maryland Court of Appeals, an instrumentality of the State, violated the FMLA by denying him self-care leave. The Federal District Court dismissed the suit on sovereign immunity grounds. The Fourth Circuit affirmed, holding that unlike the family-care provision in Hibbs, the self-care provision was not directed at an identified pattern of gender-based discrimination and was not congruent and proportional to any pattern of sex-based discrimination on the part of States. Held: The judgment is affirmed. 626 F.3d 187, affirmed. Justice Kennedy, joined by The Chief Justice, Justice Thomas, and Justice Alito, concluded that suits against States under the self-care provision are barred by sovereign immunity. Pp. 3−12. (a) Under the federal system, States, as sovereigns, are immune from damages suits, unless they waive that defense. See, e.g., Kimel v. Florida Bd. of Regents, 528 U.S. 62, 72−73. Congress may also abrogate the States’ immunity pursuant to its powers under §5 of the Fourteenth Amendment, but it must make that intention “unmistakably clear in the language of the statute,” Hibbs, supra, at 726. It did so in the FMLA. Congress also “must tailor” legislation enacted under §5 “to remedy or prevent” “conduct transgressing the Fourteenth Amendment’s substantive provisions.” Florida Prepaid Postsecondary Ed. Expense Bd. v. College Savings Bank, 527 U.S. 627, 639. “There must be a congruence and proportionality between the injury to be prevented or remedied and the means adopted to that end.” City of Boerne v. Flores, 521 U.S. 507, 520. Pp. 3−5. (b) The sex-based discrimination that supported allowing subparagraph (C) suits against States is absent with respect to the self-care provision. Petitioner’s three arguments to the contrary are unpersuasive. Pp. 5–12. (1) Petitioner maintains that the self-care provision addresses sex discrimination and sex stereotyping. But the provision, standing alone, is not a valid abrogation of the States’ immunity from suit. At the time the FMLA was enacted, there was no evidence of such discrimination or stereotyping in sick-leave policies. Congress was concerned about the economic burdens imposed by illness-related job loss on employees and their families and about discrimination based on illness, not sex. Although the self-care provision offers some women a benefit by allowing them to take leave for pregnancy-related illnesses, the provision, as a remedy, is not congruent and proportional to any identified constitutional violations. When the FMLA was enacted, Congress had no evidence that States were excluding pregnancy-related illnesses from their leave policies. Pp. 6–7. (2) Petitioner also argues that the self-care provision is a necessary adjunct to the family-care provision sustained in Hibbs. But his claim—that the provisions work in tandem to ensure the equal availability of total FMLA leave time to women and men despite their different leave-usage patterns―is unconvincing and does not comply with the requirements of City of Boerne. Also, there are no congressional findings of, or evidence on, how the self-care provision is necessary to the family-care provisions or how it reduces employer discrimination against women. Pp. 8–11. (3) Finally, petitioner contends that the self-care provision helps single parents keep their jobs when they get ill. The fact that most single parents happen to be women demonstrates, at most, that the self-care provision was directed at remedying neutral leave restrictions that have a disparate effect on women. However, “[a]lthough disparate impact may be relevant evidence of . . . discrimination . . . such evidence is insufficient [to prove a constitutional violation] even where the Fourteenth Amendment subjects state action to strict scrutiny.” Board of Trustees of Univ. of Ala. v. Garrett, 531 U.S. 356, 373. Because it is unlikely that many of the neutral leave policies affected by the self-care provision are unconstitutional, the scope of the self-care provision is out of proportion to its supposed remedial or preventive objectives. Pp. 11−12. Justice Scalia adhered to his view that the Court should abandon the “congruence and proportionality” approach in favor of one that is properly tied to the text of §5, which grants Congress the power “to enforce, by appropriate legislation,” the other provisions of the Fourteenth Amendment. Outside the context of racial discrimination, Congress’s §5 power should be limited to the regulation of conduct that itself violates the Fourteenth Amendment and thus would not reach a State’s failure to grant self-care leave to its employees. Pp. 1−2. Kennedy, J., announced the judgment of the Court and delivered an opinion, in which Roberts, C. J., and Thomas and Alito JJ., joined. Thomas, J., filed a concurring opinion. Scalia, J., filed an opinion concurring in the judgment. Ginsburg, J., filed a dissenting opinion, in which Breyer, J., joined, and in which Sotomayor and Kagan, JJ., joined as to all but footnote 1.

; Toeller v. Wisconsin Dept. of Corrections, metricconverter461 F.3d 871 (CA7 2006); Touvell v. Ohio Dept. of Mental Retardation & Developmental Disabilities, metricconverter422 F.3d 392 (CA6 2005); Brockman v. Wyoming Dept. of Family Servs., metricconverter342 F.3d 1159 (CA10 2003); Laro v. New Hampshire, metricconverter259 F.3d 1 (CA1 2001). I A The Family and Medical Leave Act of 1993 (FMLA or Act) entitles eligible employees to take up to 12 work weeks of unpaid leave per year. An employee may take leave under the FMLA for: (A) “the birth of a son or daughter . . . in order to care for such son or daughter,” (B) the adoption or foster-care placement of a child with the employee, (C) the care of a “spouse . . . son, daughter, or parent” with “a serious health condition,” and (D) the employee’s own serious health condition when the condition interferes with the employee’s ability to perform at work. 29 U. S. C. §2612(a)(1). The Act creates a private right of action to seek both equitable relief and money damages “against any employer (including a public agency) in any Federal or State court of competent jurisdiction.” §2617(a)(2). As noted, subparagraph (D) is at issue here. This Court considered subparagraph (C) in Nevada Dept. of Human Resources v. Hibbs, 538 U.S. 721 (2003). Subparagraph (C), like (A) and (B), grants leave for reasons related to family care, and those three provisions are referred to here as the family-care provisions. Hibbs held that Congress could subject the States to suit for violations of subparagraph (C), §2612(a)(1)(C). That holding rested on evidence that States had family-leave policies that differentiated on the basis of sex and that States administered even neutral family-leave policies in ways that discriminated on the basis of sex. See id., at 730–732. Subparagraph (D), the self-care provision, was not at issue in Hibbs. B Petitioner Daniel Coleman was employed by the Court of Appeals of the State of Maryland. When Coleman requested sick leave, he was informed he would be terminated if he did not resign. Coleman then sued the state court in the United States District Court for the District of Maryland, alleging, inter alia, that his employer violated the FMLA by failing to provide him with self-care leave. The District Court dismissed the suit on the basis that the Maryland Court of Appeals, as an entity of a sovereign State, was immune from the suit for damages. The parties do not dispute the District Court’s ruling that the Maryland Court of Appeals is an entity or instrumentality of the State for purposes of sovereign immunity. The District Court concluded the FMLA’s self-care provision did not validly abrogate the State’s immunity from suit. App. to Pet. for Cert. 15–20. The Court of Appeals for the Fourth Circuit affirmed, reasoning that, unlike the family-care provision at issue in Hibbs, the self-care provision was not directed at an identified pattern of gender-based discrimination and was not congruent and proportional to any pattern of sex-based discrimination on the part of States. 626 F.3d 187. Certiorari was granted. 564 U. S. ___ (2011). II A A foundational premise of the federal system is that States, as sovereigns, are immune from suits for damages, save as they elect to waive that defense. See Kimel v. Florida Bd. of Regents, 528 U.S. 62, 72–73 (2000); Alden v. Maine, 527 U.S. 706 (1999). As an exception to this principle, Congress may abrogate the States’ immunity from suit pursuant to its powers under §5 of the Fourteenth Amendment. See, e.g., Fitzpatrick v. Bitzer, 427 U.S. 445 (1976). Congress must “mak[e] its intention to abrogate unmistakably clear in the language of the statute.” Hibbs, 538 U. S., at 726. On this point the Act does express the clear purpose to abrogate the States’ immunity. Ibid. (“The clarity of Congress’ intent” to abrogate the States’ immunity from suits for damages under the FMLA “is not fairly debatable”). Congress subjected any “public agency” to suit under the FMLA, 29 U. S. C. §2617(a)(2), and a “public agency” is defined to include both “the government of a State or political subdivision thereof” and “any agency of . . . a State, or a political subdivision of a State,” §§203(x), 2611(4)(A)(iii). The question then becomes whether the self-care provision and its attempt to abrogate the States’ immunity are a valid exercise of congressional power under §5 of the Fourteenth Amendment. Section 5 grants Congress the power “to enforce” the substantive guarantees of §1 of the Amendment by “appropriate legislation.” The power to enforce “ ‘includes the authority both to remedy and to deter violation[s] of rights guaranteed’ ” by §1. See Board of Trustees of Univ. of Ala. v. Garrett, 531 U.S. 356, 365 (2001) (quoting Kimel, supra, at 81). To ensure Congress’ enforcement powers under §5 remain enforcement powers, as envisioned by the ratifiers of the Amendment, rather than powers to redefine the substantive scope of §1, Congress “must tailor” legislation enacted under §5 “ ‘to rem- edy or prevent’ ” “conduct transgressing the Fourteenth Amendment’s substantive provisions.” Florida Prepaid Postsecondary Ed. Expense Bd. v. College Savings Bank, 527 U.S. 627, 639 (1999). Whether a congressional Act passed under §5 can impose monetary liability upon States requires an assessment of both the “ ‘evil’ or ‘wrong’ that Congress intended to remedy,” ibid., and the means Congress adopted to address that evil, see City of Boerne v. Flores, 521 U.S. 507, 520 (1997). Legislation enacted under §5 must be targeted at “conduct transgressing the Fourteenth Amendment’s substantive provisions.” Florida Prepaid, supra, at 639; see Kimel, supra, at 88; City of Boerne, 521 U. S., at 525. And “[t]here must be a congruence and proportionality between the injury to be prevented or remedied and the means adopted to that end.” Id., at 520. Under this analysis Hibbs permitted employees to recover damages from States for violations of subparagraph (C). In enacting the FMLA, Congress relied upon evidence of a well-documented pattern of sex-based discrimination in family-leave policies. States had facially discriminatory leave policies that granted longer periods of leave to women than to men. 538 U. S., at 730–731. States also administered facially neutral family-leave policies in gender-biased ways. Id., at 732. These practices reflected what Congress found to be a “pervasive sex-role stereotype that caring for family members is women’s work,” id., at 731, a stereotype to which even this Court had succumbed in earlier times, id., at 729. Faced with “the States’ record of unconstitutional participation in, and fostering of, gender-based discrimination in the administration of leave benefits,” Hibbs concluded that requiring state employers to give all employees the opportunity to take family-care leave was “narrowly targeted at the faultline between work and family—precisely where sex-based overgeneralization has been and remains strongest.” Id., at 735, 738. B The same cannot be said for requiring the States to give all employees the opportunity to take self-care leave. Petitioner advances three arguments for allowing employees to recover damages from States that violate the FMLA’s self-care provision: The self-care provision standing alone addresses sex discrimination and sex stereotyping; the provision is a necessary adjunct to the family-care provision sustained in Hibbs; and the provision eases the burden on single parents. But what the family-care provisions have to support them, the self-care provision lacks, namely evidence of a pattern of state constitutional violations accompanied by a remedy drawn in narrow terms to address or prevent those violations. 1 Standing alone, the self-care provision is not a valid abrogation of the States’ immunity from suit. When the FMLA was enacted, “ninety-five percent of full-time state- and local-government employees were covered by paid sick leave plans and ninety-six percent of such employees likewise enjoyed short-term disability protection.” Brief for States of Texas et al. as Amici Curiae 13–14 (hereinafter Texas Brief) (citing Bureau of Labor Statistics, U. S. Dept. of Labor, Employee Benefits in State and Local Governments 17–26 (1994) (hereinafter BLS Rept.)). The evidence did not suggest States had facially discriminatory self-care leave policies or that they administered neutral self-care leave policies in a discriminatory way. And there is scant evidence in the legislative history of a purported stereotype harbored by employers that women take self-care leave more often than men. Congress considered evidence that “men and women are out on medical leave approximately equally.” H. R. Rep. No. 101–28, pt. 1, p. 15 (1989) (hereinafter H. R. Rep.). Nothing in the record shows employers formulated self-care leave policies based on a contrary view. Without widespread evidence of sex discrimination or sex stereotyping in the administration of sick leave, it is apparent that the congressional purpose in enacting the self-care provision is unrelated to these supposed wrongs. The legislative history of the self-care provision reveals a concern for the economic burdens on the employee and the employee’s family resulting from illness-related job loss and a concern for discrimination on the basis of illness, not sex. See, e.g., S. Rep. No. 103–3, pp. 11–12 (1993); H. R. Rep., at 23. In the findings pertinent to the self-care provision, the statute makes no reference to any distinction on the basis of sex. See 29 U. S. C. §2601(a)(4) (“[T]here is inadequate job security for employees who have serious health conditions that prevent them from working for temporary periods”). By contrast, with regard to family care Congress invoked concerns related to gender. See §2601(a)(5) (“[D]ue to the nature of the roles of men and women in our society, the primary responsibility for family caretaking often falls on women, and such responsibility affects the working lives of women more than it affects the working lives of men”). It is true the self-care provision offers some women a benefit by allowing them to take leave for pregnancy-related illnesses; but as a remedy, the provision is not congruent and proportional to any identified constitutional violations. At the time of the FMLA’s enactment, “ninety-five percent” of state employees had paid sick-leave plans at work, and “ninety-six percent” had short-term disability protection. Texas Brief 13–14 (citing BLS Rept. 17–26). State employees presumably could take leave for pregnancy-related illnesses under these policies, and Congress did not document any pattern of States excluding pregnancy-related illnesses from sick-leave or disability-leave policies. “Congress . . . said nothing about the existence or adequacy of state remedies.” Florida Prepaid, 527 U. S., at 644. It follows that abrogating the States’ immunity from suits for damages for failure to give self-care leave is not a congruent and proportional remedy if the existing state leave policies would have sufficed. 2 As an alternative justification for the self-care provision, it has been suggested that the provision is a necessary adjunct to the family-care provisions. Petitioner argues that employers may assume women are more likely to take family-care leave than men and that the FMLA therefore offers up to 12 weeks of leave for family care and self care combined. According to petitioner, when the self-care provision is coupled with the family-care provisions, the self-care provision could reduce the difference in the expected number of weeks of FMLA leave that different employees take for different reasons. The fact that self-care leave could have this effect does not mean that it would. If, for example, women are expected to take 20 days of family-care leave per year and men to take 10, and women and men are each expected to take 5 days of self-care leave per year, the difference in the expected number of days of leave and cost to the employer remains the same regardless of the availability of self-care leave. Congress made no findings, and received no specific testimony, to suggest the availability of self-care leave equalizes the expected amount of FMLA leave men and women will take. Even if women take family-care leave more often than men, men do not take self-care leave more often than women; and there is little evidence that employers assume they do. See H. R. Rep., at 15. Petitioner suggests that some women will be expected to take all 12 weeks of leave under the FMLA for family-care purposes, and therefore that any amount of self-care leave taken by men will diminish the difference in the amount of FMLA leave taken by men and women. But there is little evidence to support petitioner’s assumption about the magnitude of women’s expected FMLA leave for family-care purposes. And men are only expected to take five days of sick leave per year, see ibid., so the self-care provision diminishes the difference in expected leave time by a maximum of five days. And that is only to the extent women use all their available FMLA leave for family-care reasons. Petitioner’s overly complicated argument about how the self-care provision works in tandem with the family-care provisions is unconvincing and in the end does not comply with the clear requirements of City of Boerne. In addition petitioner’s first defense of the self-care provision contradicts his second defense of the provision. In the first defense, the Court is told employers assume women take more self-care leave than men. See Tr. of Oral Arg. 10–12. In the second defense, the Court is told the self-care provision provides an incentive to hire women that will counteract the incentives created by the family-care provisions because employers assume women take more family-care leave than men. But if the first defense is correct, the second defense is wrong. In other words, if employers assume women take self-care leave more often than men (the first defense), a self-care provision will not provide an incentive to hire women. To the contrary, the self-care provision would provide an incentive to discriminate against women. There is “little support in the record for the concerns that supposedly animated” the self-care provision. Florida Prepaid, supra, at 639. Only supposition and conjecture support the contention that the self-care provision is necessary to make the family-care provisions effective. The evidence documented in support of the self-care provision is, to a large degree, unrelated to sex discrimination, or to the administration of the family-care provisions. See supra, at 7. Congress made no findings and did not cite specific or detailed evidence to show how the self-care provision is necessary to the family-care provisions or how it reduces an employer’s incentives to discriminate against women. And “Congress . . . said nothing about the existence or adequacy of state” sick-leave policies. Florida Prepaid, supra, at 644; see Garrett, 531 U. S., at 373. Under this Court’s precedents, more is required to sub- ject unconsenting States to suits for damages, particularly where, as here, it is for violations of a provision (the self-care provision) that is a supposedly preventive step in aid of already preventive provisions (the family-care provisions). See Florida Prepaid, 527 U. S., at 642 (“[T]he legislative record still provides little support for the prop- osition that Congress sought to remedy a Fourteenth Amendment violation in enacting the Patent Remedy Act”); Kimel, 528 U. S., at 88 (“One means by which we have made such a determination . . . is by examining the legislative record containing the reasons for Congress’ action”). The “few fleeting references” to how self-care leave is inseparable from family-care leave fall short of what is required for a valid abrogation of States’ immunity from suits for damages. Florida Prepaid, supra, at 644. These “isolated sentences clipped from floor debates” and testimony, Kimel, supra, at 89, are stated as conclusions, unsupported by evidence or findings about how the self-care provision interrelates to the family-care provisions to counteract employers’ incentives to discriminate against women. Congress must rely on more than abstract generalities to subject the States to suits for damages. Otherwise, Congress could choose to combat the purported effects of the family-care provisions by allowing employees to sue States that do not permit employees to take vacation time under the FMLA. There is nothing in particular about self-care leave, as opposed to leave for any personal reason, that connects it to gender discrimination. And when the issue, as here, is whether subparagraph (D) can abrogate a State’s immunity from damages, there is no sufficient nexus, or indeed any demonstrated nexus, between self-care leave and gender discrimination by state employers. Documented discrimination against women in the general workplace is a persistent, unfortunate reality, and, we must assume, a still prevalent wrong. An explicit purpose of the Congress in adopting the FMLA was to improve workplace conditions for women. See 29 U. S. C. §§2601(b)(4), (5). But States may not be subject to suits for damages based on violations of a comprehensive statute unless Congress has identified a specific pattern of constitutional violations by state employers. See City of Boerne, 521 U. S., at 532. 3 The petitioner’s last defense of the self-care provision is that the provision helps single parents retain their jobs when they become ill. This, however, does not explain how the provision remedies or prevents constitutional violations. The fact that most single parents happen to be women, see, e.g., S. Rep. No. 103–3, at 7, demonstrates, at most, that the self-care provision was directed at remedying employers’ neutral leave restrictions which have a disparate effect on women. “Although disparate impact may be relevant evidence of . . . discrimination . . . such evidence alone is insufficient [to prove a constitutional violation] even where the Fourteenth Amendment subjects state action to strict scrutiny.” Garrett, supra, at 372–373; see Tuan Anh Nguyen v. INS, 533 U.S. 53, 82–83 (2001) (O’Connor, J., dissenting); Washington v. Davis, 426 U.S. 229, 239 (1976). To the extent, then, that the self-care provision addresses neutral leave policies with a disparate impact on women, it is not directed at a pattern of constitutional violations. Because, moreover, it is “unlikely that many of the [neutral leave policies] . . . affected by” the self-care provision are unconstitutional, “the scope of the [self-care provision is] out of proportion to its supposed remedial or preventive objectives.” Kimel, supra, at 82; see City of Boerne, supra, at 519. Of course, a State need not assert its Eleventh Amendment immunity from suits for damages. See, e.g., Sossamon v. Texas, 563 U. S. ___, ___ (2011) (slip op., at 5) (“A State . . . may choose to waive its immunity in federal court at its pleasure”). Discrimination against women is contrary to the public policy of the State of Maryland, see, e.g., Maryland’s Fair Employment Practices Act, Md. State Govt. Code Ann. §20–606 (Lexis 2009), and the State has conceded that the Act is good social policy, see Tr. of Oral Arg. 35. If the State agrees with petitioner that damages liability for violations of the self-care provision is necessary to combat discrimination against women, the State may waive its immunity or create a parallel state law cause of action. * * * As a consequence of our constitutional design, money damages are the exception when sovereigns are defendants. See, e.g., Pennhurst State School and Hospital v. Halderman, 451 U.S. 1, 29 (1981). Subjecting States to suits for damages pursuant to §5 requires more than a theory for why abrogating the States’ immunity aids in, or advances, a stated congressional purpose. To abrogate the States’ immunity from suits for damages under §5, Congress must identify a pattern of constitutional violations and tailor a remedy congruent and proportional to the documented violations. It failed to do so when it allowed employees to sue States for violations of the FMLA’s self-care provision. The judgment of the Court of Appeals is affirmed. It is so ordered.

565.US.2011_10-948

Although respondents’ credit card agreement required their claims to be resolved by binding arbitration, they filed a lawsuit against petitioner CompuCredit Corporation and a division of petitioner bank, alleging, inter alia, violations of the Credit Repair Organizations Act (CROA). The Federal District Court denied the defendants’ motion to compel arbitration, concluding that Congress intended CROA claims to be nonarbitrable. The Ninth Circuit affirmed. Held: Because the CROA is silent on whether claims under the Act can proceed in an arbitrable forum, the Federal Arbitration Act (FAA) requires the arbitration agreement to be enforced according to its terms. Pp. 2–10. (a) Section 2 of the FAA establishes “a liberal federal policy favoring arbitration.” Moses H. Cone Memorial Hospital v. Mercury Constr. Corp., 460 U.S. 1, 24. It requires that courts enforce arbitration agreements according to their terms. See Dean Witter Reynolds Inc. v. Byrd, 470 U.S. 213, 221. That is the case even when federal statutory claims are at issue, unless the FAA’s mandate has been “overridden by a contrary congressional command.” Shearson/American Express Inc. v. McMahon, 482 U.S. 220, 226. Pp. 2–3. (b) The CROA provides no such command. Respondents contend that the CROA’s disclosure provision—which requires credit repair organizations to provide consumers with a statement that includes the sentence “ ‘You have a right to sue a credit repair organization that violates the [Act],’ ” 15 U. S. C. §1679c(a)—gives consumers the right to bring an action in a court of law; and that, because the CROA prohibits the waiver of “any right of the consumer under this subchapter,” §1679f(a), the arbitration agreement’s waiver of the “right” to bring a court action cannot be enforced. Respondents’ premise is flawed. The disclosure provision creates only a right for consumers to receive a specific statement describing the consumer protections that the law elsewhere provides, one of which is the right to enforce a credit repair organization’s “liab[ility]” for “fail[ure] to comply with [the Act].” §1679g(a). That provision does not override the FAA’s mandate. Its mere contemplation of judicial enforcement does not demonstrate that the Act provides consumers with a “right” to initial judicial enforcement. Pp. 3–8. (c) At the time of the CROA’s enactment in 1996, arbitration clauses such as the one at issue were no rarity in consumer contracts generally, or in financial services contracts in particular. Had Congress meant to prohibit these very common provisions in the CROA, it would have done so in a manner less obtuse than what respondents suggest. Pp. 8–9. 615 F.3d 1204, reversed and remanded. Scalia, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, Breyer, and Alito, JJ., joined. Sotomayor, J., filed an opinion concurring in the judgment, in which Kagan, J., joined. Ginsburg, J., filed a dissenting opinion.

We consider whether the Credit Repair Organizations Act (CROA), 15 U. S. C. §1679 et seq., precludes enforcement of an arbitration agreement in a lawsuit alleging violations of that Act. I Respondents are individuals who applied for and received an Aspire Visa credit card marketed by petitioner CompuCredit Corporation and issued by Columbus Bank and Trust, now a division of petitioner Synovus Bank. In their applications they agreed to be bound by a provision which read: “Any claim, dispute or controversy (whether in contract, tort, or otherwise) at any time arising from or relating to your Account, any transferred balances or this Agreement (collectively, ‘Claims’), upon the election of you or us, will be resolved by binding arbitration . . . .” App. 62. In 2008, respondents filed a class-action complaint against CompuCredit and Columbus in the United States District Court for the Northern District of California, alleging, as relevant here, violations of the CROA. The claims largely involved the defendants’ allegedly misleading representation that the credit card could be used to rebuild poor credit and their assessment of multiple fees upon opening of the accounts, which greatly reduced the advertised credit limit. The District Court denied the defendants’ motion to compel arbitration of the claims, concluding that “Congress intended claims under the CROA to be non-arbitrable.” 617 F. Supp. 2d 980, 988 (2009). A panel of the United States Court of Appeals for the Ninth Circuit affirmed, Judge Tashima dissenting. 615 F.3d 1204 (2010). We granted certiorari, 563 U. S. ___ (2011). II The background law governing the issue before us is the Federal Arbitration Act (FAA), 9 U. S. C. §1 et seq., enacted in 1925 as a response to judicial hostility to arbitration. AT&T Mobility LLC v. Concepcion, 563 U. S. ___, ___ (2011) (slip op., at 4). As relevant here, the FAA provides: “A written provision in any maritime transaction or a contract evidencing a transaction involving commerce to settle by arbitration a controversy thereafter arising out of such contract or transaction . . . shall be valid, irrevocable, and enforceable, save upon such grounds as exist at law or in equity for the revocation of any contract.” 9 U. S. C. §2. This provision establishes “a liberal federal policy favoring arbitration agreements.” Moses H. Cone Memorial Hospital v. Mercury Constr. Corp., 460 U.S. 1, 24 (1983) . See also, e.g., Concepcion, supra, at __ (slip op., at 4); Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 25 (1991) . It requires courts to enforce agreements to arbitrate according to their terms. See Dean Witter Reynolds Inc. v. Byrd, 470 U.S. 213, 221 (1985) . That is the case even when the claims at issue are federal statutory claims, unless the FAA’s mandate has been “overridden by a contrary congressional command.” Shearson/American Express Inc. v. McMahon, 482 U.S. 220, 226 (1987) . See also Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 628 (1985) . Respondents contend that the CROA contains such a command. That statute regulates the practices of credit repair organizations, defined as certain entities that offer services for the purpose of “(i) improving any consumer’s credit record, credit history, or credit rating; or (ii) providing advice or assistance to any consumer with regard to any activity or service described in clause (i).” [ 1 ] 15 U. S. C. §1679a(3). In its principal substantive provisions, the CROA prohibits certain practices, §1679b, establishes certain requirements for contracts with consumers, §1679d, and gives consumers a right to cancel, §1679e. Enforcement is achieved through the Act’s provision of a private cause of action for violation, §1679g, as well as through federal and state administrative enforcement, §1679h. III Like the District Court and the Ninth Circuit, respondents focus on the CROA’s disclosure and nonwaiver provisions. The former, which is reproduced in full in the Appendix, infra, sets forth a statement that the credit re-pair organization must provide to the consumer before any contract is executed. §1679c(a). One sentence of that required statement reads, “ ‘You have a right to sue a credit repair organization that violates the Credit Repair Organization Act.’ ” The Act’s nonwaiver provision states, “Any waiver by any consumer of any protection provided by or any right of the consumer under this subchapter— (1) shall be treated as void; and (2) may not be enforced by any Federal or State court or any other person.” §1679f(a). The Ninth Circuit adopted the following line of reasoning, urged upon us by respondents here: The disclosure provision gives consumers the “right to sue,” which “clearly involves the right to bring an action in a court of law.” 615 F. 3d, at 1208. Because the nonwaiver provision prohibits the waiver of “any right of the consumer under this subchapter,” the arbitration agreement—which waived the right to bring an action in a court of law—cannot be enforced. Id., at 1214. The flaw in this argument is its premise: that the disclosure provision provides consumers with a right to bring an action in a court of law. It does not. Rather, it imposes an obligation on credit repair organizations to supply consumers with a specific statement set forth (in quotation marks) in the statute. The only consumer right it creates is the right to receive the statement, which is meant to describe the consumer protections that the law elsewhere provides. The statement informs consumers, for instance, that they can dispute the accuracy of information in their credit file and that “ ‘[t]he credit bureau must then reinvestigate and modify or remove inaccurate or incomplete information.’ ” 15 U. S. C. §1679c(a). That description is derived from §1681i(a), which sets out in great detail the procedures to be followed by a credit bureau in the event of challenges to the accuracy of its information. Similarly, the required statement informs consumers that they may “ ‘cancel your contract with any credit repair organization for any reason within 3 business days from the date you signed it’ ”—the right created and set forth in more detail in §1679e. And the “right to sue” language describes the consumer’s right to enforce the credit repair organization’s “liab[ility]” for “fail[ure] to comply with any provision of this subchapter” provided for in §1679g(a). [ 2 ] Thus, con-trary to the dissent’s assertion, our interpretation does not “[r]educ[e] the required disclosure to insignificance,” post, at 6. The disclosure provision informs consumers of their right to enforce liability for any failure to conform to the statute—information they might otherwise not possess. It is the dissent’s interpretation that effectively reduces a portion of the CROA to a nullity. Interpreting the “right to sue” language in §1679c(a) to “create” a right to sue in court not only renders it strikingly out of place in a section that is otherwise devoted to giving the consumer notice of rights created elsewhere; it also renders the creation of the “right to sue” elsewhere superfluous. Respondents suggest that the CROA’s civil-liability pro-vision, §1679g (set forth in full in the Appendix, infra), demonstrates that the Act provides consumers with a “right” to bring an action in court. They cite the provision’s repeated use of the terms “action,” “class action,” and “court”—terms that they say call to mind a judicial proceeding. These references cannot do the heavy lifting that respondents assign them. It is utterly commonplace for statutes that create civil causes of action to describe the details of those causes of action, including the relief available, in the context of a court suit. If the mere formulation of the cause of action in this standard fashion were sufficient to establish the “contrary congressional com- mand” overriding the FAA, McMahon, supra, at 226, valid arbitration agreements covering federal causes of action would be rare indeed. But that is not the law. In Gilmer we enforced an arbitration agreement with respect to a cause of action created by the Age Discrimination in Employment Act of 1967 (ADEA) which read, in part: “Any person aggrieved may bring a civil action in any court of competent jurisdiction for such legal or equitable relief as will effectuate the purposes of this chapter.” 29 U. S. C. §626(c)(1). In McMahon we enforced an arbitration agreement with respect to a cause of action created by a provision of the Racketeer Influenced and Corrupt Organizations Act (RICO) which read, in part: “Any person injured in his business or property by reason of a violation of section 1962 of this chapter may sue therefor in any appropriate United States district court and shall recover threefold the damages he sustains and the cost of the suit . . . .” 18 U. S. C. §1964(c). And in Mitsubishi Motors we enforced an arbitration agreement with respect to a cause of action created by a provision of the Clayton Act which read, in part: “[A]ny person who shall be injured in his business or property by reason of anything forbidden in the antitrust laws may sue therefor in any district court of the United States . . . and shall recover threefold the damages by him sustained, and the cost of suit, including a reasonable attorney’s fee.” 15 U. S. C. §15(a). Thus, we have repeatedly recognized that contractually required arbitration of claims satisfies the statutory prescription of civil liability in court. See Gilmer, 500 U. S., at 28; McMahon, 482 U. S., at 240; Mitsubishi Motors, 473 U. S., at 637. To be sure, none of the statutes described above contained a nonwaiver provision, as the statute before us does. But if a cause-of-action provision mentioning judicial enforcement does not create a right to initial judicial enforcement, the waiver of initial judicial enforcement is not the waiver of a “right of the consumer,” §1679f(a). It takes a considerable stretch to regard the nonwaiver provision as a “congressional command” that the FAA shall not apply. [ 3 ] Moreover, if one believes that §1679g’s contemplation of court suit (combined with §1679f(a)) establishes a nonwaiv-able right to initial judicial enforcement, one must also believe that it establishes a nonwaivable right to initial judicial enforcement in any competent judicial tribunal, since it contains no limitation. We think it clear, however, that this mere “contemplation” of suit in any competent court does not guarantee suit in all competent courts, disabling the parties from adopting a reasonable forum-selection clause. And just as the contemplated availability of all judicial forums may be reduced to a single forum by contractual specification, so also can the contemplated availability of judicial action be limited to judicial action compelling or reviewing initial arbitral adjudication. The parties remain free to specify such matters, so long as the guarantee of §1679g—the guarantee of the legal power to impose liability—is preserved. Respondents and the dissent maintain that if the CROA does not create a right to a judicial forum, then the disclosure provision effectively requires that credit repair organizations mislead consumers. We think not. The disclosure provision is meant to describe the law to consumers in a manner that is concise and comprehensible to the layman—which necessarily means that it will be imprecise. The required statement says, for example, that the CROA “ ‘prohibits deceptive practices by credit repair organiza- tions,’ ” 15 U. S. C. §1679c(a). This is in some respects an overstatement, and in some respects an understatement, of the “Prohibited practices” set forth in §1679b. It would include, for example, deception apart from “the offer or sale of the services of the credit repair organization,” §1679b(a)(4). Yet we would not hold, in order to prevent the required statement from being “misleading,” that a consumer has a right to be protected from deceptive practices beyond those actually covered by §1679b. So also with respect to the statement’s description of a “right to sue.” This is a colloquial method of communicating to consumers that they have the legal right, enforceable in court, to recover damages from credit repair organizations that violate the CROA. We think most consumers would understand it this way, without regard to whether the suit in court has to be preceded by an arbitration proceeding. Leaving that possibility out may be imprecise, but it is not misleading—and certainly not so misleading as to demand, in order to avoid that result, reading the statute to contain a guaranteed right it does not in fact contain. IV At the time of the CROA’s enactment in 1996, arbitration clauses in contracts of the type at issue here were no rarity. Quite the contrary, the early 1990’s saw the increased use of arbitration clauses in consumer contracts generally, and in financial services contracts in particular. See Ware, Arbitration and Unconscionability After Doctor’s Associates, Inc. v. Casarotto, 31 Wake Forest L. Rev. 1001, 1001, and n. 3 (1996); J. Shimabukuro, Congressional Research Service Report for Congress, The Federal Arbitration Act: Background and Recent Developments 1 (2002). Had Congress meant to prohibit these very common provisions in the CROA, it would have done so in a manner less obtuse than what respondents suggest. When it has restricted the use of arbitration in other contexts, it has done so with a clarity that far exceeds the claimed indications in the CROA. See, e.g., 7 U. S. C. §26(n)(2) (2006 ed., Supp. IV) (“No predispute arbitration agreement shall be valid or enforceable, if the agreement requires arbitration of a dispute arising under this section”); 15 U. S. C. §1226(a)(2) (2006 ed.) (“Notwithstanding any other provision of law, whenever a motor vehicle franchise contract provides for the use of arbitration to resolve a controversy arising out of or relating to such contract, arbitration may be used to settle such controversy only if after such controversy arises all parties to such controversy consent in writing to use arbitration to settle such controversy”); cf. 12 U. S. C. §5518(b) (2006 ed., Supp. IV) (granting authority to the newly created Consumer Financial Protection Bureau to regulate predispute arbitration agreements in contracts for consumer financial products or services). [ 4 ] That Congress would have sought to achieve the same result in the CROA through combination of the nonwaiver provision with the “right to sue” phrase in the disclosure provision, and the references to “action” and “court” in the description of damages recoverable, is unlikely. * * * Because the CROA is silent on whether claims under the Act can proceed in an arbitrable forum, the FAA requires the arbitration agreement to be enforced according to its terms. The judgment of the Ninth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. APPENDIX Section 1679c provides: “(a) Disclosure required “Any credit repair organization shall provide any consumer with the following written statement before any contract or agreement between the consumer and the credit repair organization is executed: “ ‘Consumer Credit File Rights Under State and Federal Law “ ‘You have a right to dispute inaccurate information in your credit report by contacting the credit bureau directly. However, neither you nor any ‘credit repair’ company or credit repair organization has the right to have accurate, current, and verifiable information removed from your credit report. The credit bureau must remove accurate, negative information from your report only if it is over 7 years old. Bankruptcy information can be reported for 10 years. “ ‘You have a right to obtain a copy of your credit report from a credit bureau. You may be charged a reasonable fee. There is no fee, however, if you have been turned down for credit, employment, insurance, or a rental dwelling because of information in your credit report within the preceding 60 days. The credit bureau must provide someone to help you interpret the information in your credit file. You are entitled to receive a free copy of your credit report if you are unemployed and intend to apply for employment in the next 60 days, if you are a recipient of public welfare assistance, or if you have reason to believe that there is inaccurate information in your credit report due to fraud. “ ‘You have a right to sue a credit repair organiza- tion that violates the Credit Repair Organization Act. This law prohibits deceptive practices by credit repair organizations. “ ‘You have the right to cancel your contract with any credit repair organization for any reason within 3 business days from the date you signed it. “ ‘Credit bureaus are required to follow reasonable procedures to ensure that the information they report is accurate. However, mistakes may occur. “ ‘You may, on your own, notify a credit bureau in writing that you dispute the accuracy of information in your credit file. The credit bureau must then reinvestigate and modify or remove inaccurate or incomplete information. The credit bureau may not charge any fee for this service. Any pertinent information and copies of all documents you have concerning an error should be given to the credit bureau. “ ‘If the credit bureau’s reinvestigation does not resolve the dispute to your satisfaction, you may send a brief statement to the credit bureau, to be kept in your file, explaining why you think the record is inaccurate. The credit bureau must include a summary of your statement about disputed information with any report it issues about you. “ ‘The Federal Trade Commission regulates credit bureaus and credit repair organizations. For more information contact: “ ‘The Public Reference Branch “ ‘Federal Trade Commission “ ‘Washington, D. C. 20580’. “(b) Separate statement requirement “The written statement required under this section shall be provided as a document which is separate from any written contract or other agreement between the credit repair organization and the consumer or any other written material provided to the consumer. “(c) Retention of compliance records “(1) In general “The credit repair organization shall maintain a copy of the statement signed by the consumer acknowledging receipt of the statement. “(2) Maintenance for 2 years “The copy of any consumer’s statement shall be maintained in the organization’s files for 2 years after the date on which the statement is signed by the consumer.” * * * Section 1679g provides: “(a) Liability established “Any person who fails to comply with any provision of this subchapter with respect to any other person shall be liable to such person in an amount equal to the sum of the amounts determined under each of the following paragraphs: “(1) Actual damages “The greater of— “(A) the amount of any actual damage sustained by such person as a result of such failure; or “(B) any amount paid by the person to the credit repair organization. “(2) Punitive damages “(A) Individual actions “In the case of any action by an individual, such additional amount as the court may allow. “(B) Class actions “In the case of a class action, the sum of— “(i) the aggregate of the amount which the court may allow for each named plaintiff; and “(ii) the aggregate of the amount which the court may allow for each other class member, without regard to any minimum individual recovery. “(3) Attorneys’ fees “In the case of any successful action to enforce any liability under paragraph (1) or (2), the costs of the action, together with reasonable attorneys’ fees. “(b) Factors to be considered in awarding punitive damages “In determining the amount of any liability of any credit repair organization under subsection (a)(2) of this section, the court shall consider, among other relevant factors— “(1) the frequency and persistence of noncompliance by the credit repair organization; “(2) the nature of the noncompliance; “(3) the extent to which such noncompliance was intentional; and “(4) in the case of any class action, the number of consumers adversely affected.” Notes 1 The District Court said that petitioners do not dispute that they come within this definition. See 617 F. Supp. 980, 984, n. 2 (ND Cal. 2009). The Ninth Circuit did not address that issue, see 615 F.3d 1204, 1207, n. 3 (2010), nor do we. 2 Accordingly, when a consumer sues to enforce liability under the CROA, he does so under §1679g(a), not “in light of §1679c,” post, at 4 (Ginsburg, J., dissenting). An action under the CROA need not referto §1679c at all, unless it is based on the company’s failure to provide the statement required under that section. Section 1679g(a) creates the “right” at issue and describes it in detail not contained in §1679c’s summary. When determining the scope of that right, it is therefore §1679g(a)—and not §1679c—that must govern. 3 Gilmer noted that the ADEA had been amended after conclusion of the arbitration agreement in that case to preclude waiver of “rights or claims that may arise after the date the waiver is executed.” 29 U. S. C. §626(f)(1)(C). The Court said in dictum that this provision “did not explicitly preclude arbitration or other nonjudicial resolution of claims,” 500 U. S., at 29. 4 The dissent questions the relevance of these statutes, since they postdated the CROA and since this Court’s intervening decisions compelling arbitration “increasingly alerted Congress to the utility of drafting antiwaiver prescriptions with meticulous care.” Post, at 8. But as the dissent implicitly recognizes, Congress had been “alerted” much before these post-CROA statutes were passed. The CROA itself followed a series of this Court’s seminal decisions compelling arbitration, decisions which held that the FAA had established a “federal policy favoring arbitration,” Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 26 (1991) , and that “[t]he burden is on the party opposing arbitration . . . to show that Congress intended to preclude a waiver of judicial remedies,” Shearson/American Express Inc. v. McMahon, 482 U.S. 220, 227 (1987) . To the extent Congress is ever “stimulated” by this Court’s decisions, post, at 8, there is no reason to think the Congress that enacted the CROA was any less stimulated than subsequent Congresses.

566.US.2011_10-1261

Under §16(b) of the Securities Exchange Act of 1934, a corporation or security holder of that corporation may sue corporate insiders who realize profits from the purchase and sale, or sale and purchase, of the corporation’s securities within any 6-month period. The Act provides that such suits must be brought within “two years after the date such profit was realized.” 15 U. S. C. §78p(b). In 2007, respondent Simmonds filed numerous §16(b) actions, claiming that, in underwriting various initial public offerings in the late 1990’s and 2000, petitioners and others inflated the stocks’ aftermarket prices, allowing them to profit from the aftermarket sales. She also claimed that petitioners had failed to comply with §16(a)’s requirement that insiders disclose any changes to their ownership interests. That failure, according to Simmonds, tolled §16(b)’s 2-year time period. The District Court dismissed the complaints as untimely. The Ninth Circuit reversed. Citing its decision in Whittaker v. Whittaker Corp., 639 F.2d 516, it held that the limitations period is tolled until an insider files the §16(a) disclosure statement “regardless of whether the plaintiff knew or should have known of the conduct at issue.” Held: Even assuming that the 2-year period can be extended (a question on which the Court is equally divided), the Ninth Circuit erred in determining that it is tolled until a §16(a) statement is filed. The text of §16(b)—which starts the clock from “the date such profit was realized,” §78p(b)—simply does not support the Whittaker rule. The rule is also not supported by the background rule of equitable tolling for fraudulent concealment. Under long-settled equitable-tolling principles, a litigant must establish “(1) that he has been pursuing his rights diligently, and (2) that some extraordinary circumstances stood in his way.” Pace v. DiGuglielmo, 544 U.S. 408, 418. Tolling therefore ceases when fraudulently concealed facts are, or should have been, discovered by the plaintiff. Allowing tolling to continue beyond that point would be inequitable and inconsistent with the general purpose of statutes of limitations: “to protect defendants against stale or unduly delayed claims.” John R. Sand & Gravel Co. v. United States, 552 U.S. 130, 133. The Whittaker rule’s inequity is especially apparent here, where the theory of §16(b) liability is so novel that petitioners can plausibly claim that they were not aware they had to file a §16(a) statement. Under the Whittaker rule, alleged insiders who disclaim the necessity of filing are compelled either to file or to face the prospect of §16(b) litigation in perpetuity. Had Congress intended the possibility of such endless tolling, it would have said so. Simmonds’ arguments to the contrary are unpersuasive. The lower courts should consider in the first instance how usual equitable tolling rules apply in this case. Pp. 4–8. 638 F.3d 1072, vacated and remanded. Scalia, J., delivered the opinion of the Court, in which all other Members joined, except Roberts, C. J., who took no part in the consideration or decision of the case.

We consider whether the 2-year period to file suit against a corporate insider under §16(b) of the Securities Exchange Act of 1934, 15 U. S. C. §78p(b), begins to run only upon the insider’s filing of the disclosure statement required by §16(a) of the Act, §78p(a). I Under §16(b) of the Exchange Act, 48Stat. 896, as amended, a corporation or security holder of that corporation may bring suit against the officers, directors, and certain beneficial owners[1] of the corporation who realize any profits from the purchase and sale, or sale and purchase, of the corporation’s securities within any 6-month period. “The statute imposes a form of strict liability” and requires insiders to disgorge these “short-swing” profits “even if they did not trade on inside information or in- tend to profit on the basis of such information.” Gollust v. Mendell, 501 U.S. 115, 122 (1991). Section 16(b) provides that suits must be brought within “two years after the date such profit was realized.”[2] 15 U. S. C. §78p(b). In 2007, respondent Vanessa Simmonds filed 55 nearly identical actions under §16(b) against financial institutions that had underwritten various initial public offerings (IPOs) in the late 1990’s and 2000, including these petitioners.[3] In a representative complaint, she alleged that the underwriters and the issuers’ insiders employed various mechanisms to inflate the aftermarket price of the stock to a level above the IPO price, allowing them to profit from the aftermarket sale. App. 59. She further alleged that, as a group, the underwriters and the insiders owned in excess of 10% of the outstanding stock during the relevant time period, which subjected them to both disgorgement of profits under §16(b) and the reporting requirements of §16(a). Id., at 61. See 15 U. S. C. §78m(d)(3); 17 CFR §§240.13d–5(b)(1) and 240.16a–1(a)(1) (2011). The latter requires insiders to disclose any changes to their ownership interests on a document known as a Form 4, specified in the Securities and Exchange Commission regulations. 15 U. S. C. §78p(a)(2)(C); 17 CFR §240.16a–3(a). Simmonds alleged that the underwriters failed to comply with that requirement, thereby tolling §16(b)’s 2-year time period.[4] App. 62. Simmonds’ lawsuits were consolidated for pretrial purposes, and the United States District Court for the Western District of Washington dismissed all of her complaints.[5] In re: Section 16(b) Litigation, 602 F. Supp. 2d 1202 (2009). As relevant here, the court granted petitioners’ motion to dismiss 24 complaints on the ground that §16(b)’s 2-year time period had expired long before Simmonds filed the suits. The United States Court of Appeals for the Ninth Circuit reversed in relevant part. 638 F.3d 1072 (2011). Citing its decision in Whittaker v. Whittaker Corp., 639 F.2d 516 (1981), the court held that §16(b)’s limitations period is “tolled until the insider discloses his transactions in a Section 16(a) filing, regardless of whether the plaintiff knew or should have known of the conduct at issue.” 638 F. 3d, at 1095. Judge Milan Smith, Jr., the author of the panel opinion, also specially concurred, expressing his disagreement with the Whittaker rule, but noting that the court was compelled to follow Circuit precedent. Id., at 1099–1101. We granted certiorari, 564 U. S. ___ (2011). II Petitioners maintain that these suits were properly dismissed because they were filed more than two years af-ter the alleged profits were realized. Pointing to dictum in Lampf, Pleva, Lipkind, Prupis & Petigrow v. Gilbertson, 501 U.S. 350 (1991), petitioners argue that §16(b)’s limitations period is a period of repose, which is not to be “extended to account for a plaintiff’s discovery of the facts underlying a claim.” Brief for Petitioners 17. See Lampf, supra, at 360, n. 5 (“Section 16(b) . . . sets a 2-year . . . period of repose”). We do not reach that contention, because we conclude that, even assuming that the 2-year period can be extended, the Ninth Circuit erred in de-termining that it is tolled until the filing of a §16(a) statement. In adopting its rule in Whittaker, the Ninth Circuit ex-pressed its concern that “[i]t would be a simple matter for the unscrupulous to avoid the salutary effect of Section 16(b) . . . simply by failing to file . . . reports in violation of subdivision (a) and thereby concealing from prospective plaintiffs the information they would need” to bring a §16(b) action. 639 F. 2d, at 528 (internal quotation marks omitted). Assuming that is correct, it does not follow that the limitations period is tolled until the §16(a) statement is filed. Section 16 itself quite clearly does not extend the period in that manner. The 2-year clock starts from “the date such profit was realized.” §78p(b). Congress could have very easily provided that “no such suit shall be brought more than two years after the filing of a statement under subsection (a)(2)(C).” But it did not. The text of §16 simply does not support the Whittaker rule. The Whittaker court suggested that the background rule of equitable tolling for fraudulent concealment[6] operates to toll the limitations period until the §16(a) statement is filed. See 639 F. 2d, at 527, and n. 9. Even accepting that equitable tolling for fraudulent concealment is triggered by the failure to file a §16(a) statement, the Whittaker rule is completely divorced from long-settled equitable-tolling principles. “Generally, a litigant seeking equitable tolling bears the burden of establishing two elements: (1) that he has been pursuing his rights diligently, and (2) that some extraordinary circumstances stood in his way.” Pace v. DiGuglielmo, 544 U.S. 408, 418 (2005) (emphasis added). It is well established, moreover, that when a limitations period is tolled because of fraudulent concealment of facts, the tolling ceases when those facts are, or should have been, discovered by the plaintiff. 2 C. Corman, Limitation of Actions §9.7.1, pp. 55–57 (1991). Thus, we have explained that the statute does not begin to run until discovery of the fraud “ ‘where the party injured by the fraud remains in ignorance of it without any fault or want of diligence or care on his part.’ ” Lampf, supra, at 363 (quoting Bailey v. Glover, 21 Wall. 342, 348 (1875); emphasis added). Allowing tolling to continue beyond the point at which a §16(b) plaintiff is aware, or should have been aware, of the facts underlying the claim would quite certainly be inequitable and inconsistent with the general purpose of statutes of limitations: “to protect defendants against stale or unduly delayed claims.” John R. Sand & Gravel Co. v. United States, 552 U.S. 130, 133 (2008). The inequity of the Whittaker rule is especially apparent in a case such as this, where the theory of §16(b) liability of underwriters is so novel that petitioners can plausibly claim that they were not aware they were required to file a §16(a) statement. And where they disclaim the necessity of filing, the Whittaker rule compels them either to file or to face the prospect of §16(b) litigation in perpetuity. Simmonds has acknowledged that “under her theory she could buy stocks in companies who had IPOs 20 years ago and bring claims for short-swing transactions if the underwriters had undervalued a stock.” 602 F. Supp. 2d, at 1218. The potential for such endless tolling in cases in which a reasonably diligent plaintiff would know of the facts underlying the action is out of step with the purpose of limitations periods in general. And it is especially at odds with a provision that imposes strict liability on putative insiders, see Gollust, 501 U. S., at 122. Had Congress intended this result, it most certainly would have said so. Simmonds maintains that failing to apply the Whittaker rule would obstruct Congress’s objective of curbing short-swing speculation by corporate insiders. This objective, according to Simmonds, is served by §16(a) statements, which “provide the information necessary to trigger §16(b) enforcement.” Brief for Respondent 24. Simmonds—like the Ninth Circuit in Whittaker—disregards the most glaring indication that Congress did not intend that the limitations period be categorically tolled until the statement is filed: The limitations provision does not say so. This fact alone is reason enough to reject a departure from settled equitable-tolling principles. Moreover, §16’s purpose is fully served by the rules outlined above, under which the limitations period would not expire until two years after a reasonably diligent plaintiff would have learned the facts underlying a §16(b) action. The usual equitable-tolling inquiry will thus take account of the unavailability of sources of information other than the §16(a) filing. Cf., e.g., Ruth v. Unifund CCR Partners, 604 F.3d 908, 911–913 (CA6 2010); Santos ex rel. Beato v. United States, 559 F.3d 189, 202–203 (CA3 2009). The oddity of Simmonds’ position is well demonstrated by the circumstances of this case. Under the Whittaker rule, because petitioners have yet to file §16(a) statements (as noted earlier they do not think themselves subject to that requirement), Simmonds still has two years to bring suit, even though she is so well aware of her alleged cause of action that she has already sued. If §16(a) statements were, as Simmonds suggests, indispensable to a party’s ability to sue, Simmonds would not be here. Simmonds also asserts that application of established equitable-tolling doctrine in this context would be in-consistent with Congress’s intention to establish in §16 a clear rule that is capable of “mechanical application.” Brief for Respondent 57 (internal quotation marks omitted). Equitable tolling, after all, involves fact-intensive disputes “about what the notice was, where it was disseminated, who received it, when it was received, and whether it provides sufficient notice of relevant Section 16(a) facts.” Id, at 56–57. Of course this argument counsels just as much in favor of the “statute of repose” rule that petitioners urge (that is, no tolling whatever) as it does in favor of the Whittaker rule. No tolling is certainly an easily administrable bright-line rule. And assuming some form of tolling does apply, it is preferable to apply that form which Congress was certainly aware of, as opposed to the rule the Ninth Circuit has fashioned.[7] See Meyer v. Holley, 537 U.S. 280, 286 (2003) (“Congress’ silence, while permitting an inference that Congress intended to apply ordinary background tort principles, cannot show that it intended to apply an unusual modification of those rules”). * * * Having determined that §16(b)’s limitations period is not tolled until the filing of a §16(a) statement, we remand for the lower courts to consider how the usual rules of equitable tolling apply to the facts of this case.[8] We are divided 4 to 4 concerning, and thus affirm without precedential effect, the Court of Appeals’ rejection of petitioners’ contention that §16(b) establishes a period of repose that is not subject to tolling. The judgment of the Court of Appeals is vacated, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. The Chief Justice took no part in the consideration or decision of this case. Notes 1 Section 16(b) regulates beneficial owners of more than 10% of any class of equity securities. 15 U. S. C. §78p(a)(1). 2 Section 16(b) provides in full: “For the purpose of preventing the unfair use of information which may have been obtained by such beneficial owner, director, or officer by reason of his relationship to the issuer, any profit realized by him from any purchase and sale, or any sale and purchase, of any equity security of such issuer (other than an exempted security) or a security-based swap agreement (as defined in section 206B of the Gramm-Leach-Bliley Act) involving any such equity security within any period of less than six months, unless such security or security-based swap agreement was acquired in good faith in connection with a debt previously contracted, shall inure to and be recoverable by the issuer, irrespective of any in-tention on the part of such beneficial owner, director, or officer in en-tering into such transaction of holding the security or security-based swap agreement purchased or of not repurchasing the security or security-based swap agreement sold for a period exceeding six months. Suit to recover such profit may be instituted at law or in equity in any court of competent jurisdiction by the issuer, or by the owner of any security of the issuer in the name and in behalf of the issuer if the issuer shall fail or refuse to bring such suit within sixty days after request or shall fail diligently to prosecute the same thereafter; but no such suit shall be brought more than two years after the date such profit was realized. This subsection shall not be construed to cover any transaction where such beneficial owner was not such both at the time of the purchase and sale, or the sale and purchase, of the security or security-based swap agreement (as defined in section 206B of the Gramm-Leach-Bliley Act) involved, or any transaction or transactions which the [Securities and Exchange] Commission by rules and regu-lations may exempt as not comprehended within the purpose of this subsection.” 15 U. S. C. §78p(b). 3 Simmonds also named the issuing companies as nominal defendants. In re: Section 16(b) Litigation, 602 F. Supp. 2d 1202, 1204 (WD Wash. 2009). 4 Petitioners have consistently disputed §16’s application to them, arguing that they, as underwriters, are generally exempt from the statute’s coverage. See 17 CFR §§240.16a–7(a) and 240.16a–10. Simmonds contends that this exemption does not apply where the underwriters do not act in good faith. Brief for Respondent 49. See §240.16a–7(a). We express no view on this issue. 5 Simmonds voluntarily dismissed one of the complaints. 602 F. Supp. 2d, at 1206, n. 4. 6 Relying on our decision in American Pipe & Constr. Co. v. Utah, 414 U.S. 538 (1974), Simmonds argues that the Whittaker rule isbest understood as applying legal—rather than equitable—tolling. In American Pipe, we held that “commencement of a class action suspends the applicable statute of limitations as to all asserted members of the class who would have been parties had the suit been permitted to continue as a class action.” 414 U. S., at 554. We based our conclusion on “the efficiency and economy of litigation which is a principal purpose of [Fed. Rule Civ. Proc. 23 class actions].” Id., at 553. Although we did not employ the term “legal tolling,” some federal courts have used that term to describe our holding on the ground that the rule “is derived from a statutory source,” whereas equitable tolling is “judicially created.” Arivella v. Lucent Technologies, Inc., 623 F. Supp. 2d 164, 176 (Mass. 2009). The label attached to the Whittaker rule does not matter. As we proceed to explain, neither general equitable-tolling principles nor the “statutory source” of §16 supports the conclusion that the limitations period is tolled until the filing of a §16(a) statement. 7 It is for this reason that we also reject the Second Circuit’s rule that the 2-year period is tolled until the plaintiff “gets actual notice that a person subject to Section 16(a) has realized specific short-swing profits that are worth pursuing,” Litzler v. CC Investments, L. D. C., 362 F.3d 203, 208 (2004). As that court itself recognized, this actual-notice rule departs from usual equitable-tolling principles. See id., at 207. 8 The District Court said that “there is no dispute that all of the facts giving rise to Ms. Simmonds’ complaints against [petitioners] were known to the shareholders of the Issuer Defendants for at least five years before these cases were filed,” metricconverter602 F. Supp. 2d, at 1217. The Court of Appeals did not consider the accuracy of that statement, which Simmonds disputes, Brief for Respondent 12, since it concluded the period is tolled until a §16(a) statement is filed.

567.US.2011_11-5683

Under the Anti-Drug Abuse Act (1986 Drug Act), the 5- and 10-year mandatory minimum prison terms for federal drug crimes reflected a 100-to-1 disparity between the amounts of crack cocaine and powder cocaine needed to trigger the minimums. Thus, the 5-year minimum was triggered by a conviction for possessing with intent to distribute 5 grams of crack cocaine but 500 grams of powder, and the 10-year minimum was triggered by a conviction for possessing with intent to distribute 50 grams of crack but 5,000 grams of powder. The United States Sentencing Commission—which is charged under the Sentencing Reform Act of 1984 with writing the Federal Sentencing Guidelines—incorporated the 1986 Drug Act’s 100-to-1 disparity into the Guidelines because it believed that doing so was the best way to keep similar drug-trafficking sentences proportional, thereby satisfying the Sentencing Reform Act’s basic proportionality objective. The Fair Sentencing Act, which took effect on August 3, 2010, reduced the disparity to 18-to-1, lowering the mandatory minimums applicable to many crack offenders, by increasing the amount of crack needed to trigger the 5-year minimum from 5 to 28 grams and the amount for the 10-year minimum from 50 to 280 grams, while leaving the powder cocaine amounts intact. It also directed the Sentencing Commission to make conforming amendments to the Guidelines “as soon as practicable” (but no later than 90 days after the Fair Sentencing Act’s effective date). The new amendments became effective on November 1, 2010. In No. 11−5721, petitioner Hill unlawfully sold 53 grams of crack in 2007, but was not sentenced until December 2010. Sentencing him to the 10-year minimum mandated by the 1986 Drug Act, the District Judge ruled that the Fair Sentencing Act’s 5-year minimum for selling that amount of crack did not apply to those whose offenses were committed before the Act’s effective date. In No. 11−5683, petitioner Dorsey unlawfully sold 5.5 grams of crack in 2008. In September 2010, the District Judge sentenced him to the 1986 Drug Act’s 10-year minimum, finding that it applied because Dorsey had a prior drug conviction and declining to apply the Fair Sentencing Act, under which there would be no mandated minimum term for an amount less than 28 grams, because Dorsey’s offense predated that Act’s effective date. The Seventh Circuit affirmed in both cases. Held: The Fair Sentencing Act’s new, lower mandatory minimums apply to the post-Act sentencing of pre-Act offenders. Pp. 10−20. (a) Language in different statutes argues in opposite directions. The general federal saving statute (1871 Act) provides that a new criminal statute that “repeal[s]” an older criminal statute shall not change the penalties “incurred” under that older statute “unless the repealing Act shall so expressly provide.” 1 U. S. C. §109. The word “repeal” applies when a new statute simply diminishes the penalties that the older statute set forth, see Warden v. Marrero, 417 U.S. 653, 659−664, and penalties are “incurred” under the older statute when an offender becomes subject to them, i.e., commits the underlying conduct that makes the offender liable, see United States v. Reisinger, 128 U.S. 398, 401. In contrast, the Sentencing Reform Act says that, regardless of when the offender’s conduct occurs, the applicable sentencing guidelines are the ones “in effect on the date the defendant is sentenced.” 18 U. S. C. §3553(a)(4)(A)(ii). Six considerations, taken together, show that Congress intended the Fair Sentencing Act’s more lenient penalties to apply to offenders who committed crimes before August 3, 2010, but were sentenced after that date. First, the 1871 saving statute permits Congress to apply a new Act’s more lenient penalties to pre-Act offenders without expressly saying so in the new Act. The 1871 Act creates what is in effect a less demanding interpretive requirement because the statute “cannot justify a disregard of the will of Congress as manifested, either expressly or by necessary implication, in a subsequent enactment.” Great Northern R. Co. v. United States, 208 U.S. 452, 465. Hence, this Court has treated the 1871 Act as setting forth an important background principle of interpretation that requires courts, before interpreting a new criminal statute to apply its new penalties to a set of pre-Act offenders, to assure themselves by the “plain import” or “fair implication” of the new statute that ordinary interpretive considerations point clearly in that direction. Second, the Sentencing Reform Act sets forth a special and different background principle in §3553(a)(4)(A)(ii), which applies unless ex post facto concerns are present. Thus, new, lower Guidelines amendments apply to offenders who committed an offense before the adoption of the amendments but are sentenced thereafter. Third, language in the Fair Sentencing Act implies that Congress intended to follow the Sentencing Reform Act’s special background principle here. Section 8 of the Fair Sentencing Act requires the Commission to promulgate conforming amendments to the Guidelines that “achieve consistency with other guideline provisions and applicable law.” Read most naturally, “applicable law” refers to the law as changed by the Fair Sentencing Act, including the provision reducing the crack mandatory minimums. And consistency with “other guideline provisions” and with prior Commission practice would require application of the new Guidelines amendments to offenders who committed their offense before the new amendments’ effective date but were sentenced thereafter. Fourth, applying the 1986 Drug Act’s old mandatory minimums to the post-August 3 sentencing of pre-August 3 offenders would create sentencing disparities of a kind that Congress enacted the Sentencing Reform Act and the Fair Sentencing Act to prevent. Fifth, not to apply the Fair Sentencing Act would do more than preserve a disproportionate status quo; it would make matters worse by creating new anomalies―new sets of disproportionate sentences―not previously present. That is because sentencing courts must apply the new Guidelines (consistent with the Fair Sentencing Act’s new minimums) to pre-Act offenders, and the 1986 Drug Act’s old minimums would trump those new Guidelines for some pre-Act offenders but not for all of them. Application of the 1986 Drug Act minimums to pre-Act offenders sentenced after the new Guidelines take effect would therefore produce a set of sentences at odds with Congress’ basic efforts to create more uniform, more proportionate sentences. Sixth, this Court has found no strong countervailing considerations that would make a critical difference. Pp. 10−19. (b) The new Act’s lower minimums also apply to those who committed an offense prior to August 3 and were sentenced between that date and November 1, 2010, the effective date of the new Guidelines. The Act simply instructs the Commission to promulgate new Guidelines “as soon as practicable” (but no later than 90 days after the Act took effect), and thus as far as Congress was concerned, the Commission might have promulgated those Guidelines to be effective as early as August 3. In any event, courts, treating the Guidelines as advi-sory, possess authority to sentence in accordance with the new minimums. Finally, applying the new minimums to all who are sentenced after August 3 makes it possible to foresee a reasonably smooth transition, and this Court has no reason to believe Congress would have wanted to impose an unforeseeable, potentially complex application date. Pp. 19−20. No. 11−5683, 635 F.3d 336, and No. 11−5721, 417 Fed. Appx. 560, vacated and remanded. Breyer, J., delivered the opinion of the Court, in which Kennedy, Ginsburg, Sotomayor, and Kagan, JJ., joined. Scalia, J., filed a dissenting opinion, in which Roberts, C. J., and Thomas and Alito, JJ., joined. Notes 1 Together with No. 11–5721, Hill v. United States, also on certiorari to the same court.

Federal statutes impose mandatory minimum prison sentences upon those convicted of federal drug crimes. These statutes typically base the length of a minimum prison term upon the kind and amount of the drug involved. Until 2010, the relevant statute imposed upon an offender who dealt in powder cocaine the same sentence it imposed upon an offender who dealt in one one-hundredth that amount of crack cocaine. It imposed, for example, the same 5-year minimum term upon (1) an offender convicted of possessing with intent to distribute 500 grams of powder cocaine as upon (2) an offender convicted of possessing with intent to distribute 5 grams of crack. In 2010, Congress enacted a new statute reducing the crack-to-powder cocaine disparity from 100-to-1 to 18-to-1. Fair Sentencing Act, 124Stat. 2372. The new statute took effect on August 3, 2010. The question here is whether the Act’s more lenient penalty provisions apply to offenders who committed a crack cocaine crime before August 3, 2010, but were not sentenced until after August 3. We hold that the new, more lenient mandatory minimum provisions do apply to those pre-Act offenders. I The underlying question before us is one of congres-sional intent as revealed in the Fair Sentencing Act’s lan-guage, structure, and basic objectives. Did Congress intend the Act’s more lenient penalties to apply to pre-Act offenders sentenced after the Act took effect? We recognize that, because of important background principles of interpretation, we must assume that Congress did not intend those penalties to apply unless it clearly indicated to the contrary. See infra, at 10–13. But we find that clear indication here. We rest our conclu- sion primarily upon the fact that a contrary determination would seriously undermine basic Federal Sentencing Guidelines objectives such as uniformity and proportionality in sentencing. Indeed, seen from that perspective, a contrary determination would (in respect to relevant groups of drug offenders) produce sentences less uniform and more disproportionate than if Congress had not enacted the Fair Sentencing Act at all. See infra, at 14–18. Because our conclusion rests upon an analysis of the Guidelines-based sentencing system Congress has established, we describe that system at the outset and include an explanation of how the Guidelines interact with federal statutes setting forth specific terms of imprisonment. A The Guidelines originate in the Sentencing Reform Act of 1984, 98Stat. 1987. That statute created a federal Sentencing Commission instructed to write guidelines that judges would use to determine sentences imposed upon offenders convicted of committing federal crimes. 28 U. S. C. §§991, 994. Congress thereby sought to increase transparency, uniformity, and proportionality in sentencing. United States Sentencing Commission (USSC or Commission), Guidelines Manual §1A1.3, p. 2 (Nov. 2011) (USSG); see 28 U. S. C. §§991(b)(1), 994(f). The Sentencing Reform Act directed the Commission to create in the Guidelines categories of offense behavior (e.g., “ ‘bank robbery/committed with a gun/$2500 taken’ ”) and offender characteristics (e.g., “one prior conviction”). USSG §1A1.2, at 1; see 28 U. S. C. §§994(a)–(e). A sentencing judge determines a Guidelines range by (1) finding the applicable offense level and offender category and then (2) consulting a table that lists proportionate sentenc- ing ranges (e.g., 18 to 24 months of imprisonment) at the intersections of rows (marking offense levels) and columns (marking offender categories). USSG ch. 5, pt. A, Sen-tencing Table, §§5E1.2, 7B1.4; see also §1A1.4(h), at 11. The Guidelines, after telling the judge how to determine the applicable offense level and offender category, instruct the judge to apply the intersection’s range in an ordinary case, but they leave the judge free to depart from that range in an unusual case. See 18 U. S. C. §3553(b); USSG §§1A1.2, at 1–2, 1A1.4(b), at 6–7. This Court has held that the Guidelines are now advisory. United States v. Booker, 543 U.S. 220, 245, 264 (2005); see Kimbrough v. United States, 552 U.S. 85, 91 (2007). The Guidelines determine most drug-crime offense lev-els in a special way. They set forth a Drug Quantity Table (or Table) that lists amounts of various drugs and associates different amounts with different “Base Offense Levels” (to which a judge may add or subtract levels depending upon the “specific” characteristics of the offender’s behavior). See USSG §2D1.1. The Table, for example, associates 400 to 499 grams of powder cocaine with a base offense level of 24, a level that would mean for a first-time offender a prison term of 51 to 63 months. §2D1.1(c). In 1986, Congress enacted a more specific, drug-related sentencing statute, the Anti-Drug Abuse Act (1986 Drug Act), 100Stat. 3207. That statute sets forth mandatory minimum penalties of 5 and 10 years applicable to a drug offender depending primarily upon the kind and amount of drugs involved in the offense. See 21 U. S. C. §§841(b)(1) (A)–(C) (2006 ed. and Supp. IV). The minimum applicable to an offender convicted of possessing with intent to distribute 500 grams or more of powder cocaine is 5 years, and for 5,000 grams or more of powder the minimum is 10 years. §§841(b)(1)(A)(ii), (B)(ii). The 1986 Drug Act, however, treated crack cocaine crimes as far more serious. It applied its 5-year minimum to an offender convicted of possessing with intent to distribute only 5 grams of crack (as compared to 500 grams of powder) and its 10-year minimum to one convicted of possessing with intent to distribute only 50 grams of crack (as compared to 5,000 grams of powder), thus producing a 100-to-1 crack-to-powder ratio. §§841(b)(1)(A)(iii), (B)(iii) (2006 ed.). The 1986 Drug Act, like other federal sentencing statutes, interacts with the Guidelines in an important way. Like other sentencing statutes, it trumps the Guidelines. Thus, ordinarily no matter what the Guidelines provide, a judge cannot sentence an offender to a sentence beyond the maximum contained in the federal statute setting forth the crime of conviction. Similarly, ordinarily no matter what range the Guidelines set forth, a sentencing judge must sentence an offender to at least the minimum prison term set forth in a statutory mandatory minimum. See 28 U. S. C. §§994(a), (b)(1); USSG §5G1.1; Neal v. United States, 516 U.S. 284, 289–290, 295 (1996). Not surprisingly, the Sentencing Commission incorporated the 1986 Drug Act’s mandatory minimums into the first version of the Guidelines themselves. Kimbrough, supra, at 96–97. It did so by setting a base offense level for a first-time drug offender that corresponded to the lowest Guidelines range above the applicable mandatory minimum. USSC, Report to the Congress: Mandatory Minimum Penalties in the Federal Criminal Justice System 53–54 (Oct. 2011) (2011 Report). Thus, the first Guidelines Drug Quantity Table associated 500 grams of powder cocaine with an offense level of 26, which for a first-time offender meant a sentencing range of 63 to 78 months (just above the 5-year minimum), and it associated 5,000 grams of powder cocaine with an offense level of 32, which for a first-time offender meant a sentencing range of 121 to 151 months (just above the 10-year minimum). USSG §2D1.1 (Oct. 1987). Further reflecting the 1986 Drug Act’s 100-to-1 crack-to-powder ratio, the Table associated an offense level of 26 with 5 grams of crack and an offense level of 32 with 50 grams of crack. Ibid. In addition, the Drug Quantity Table set offense levels for small drug amounts that did not trigger the 1986 Drug Act’s mandatory minimums so that the resulting Guidelines sentences would remain proportionate to the sentences for amounts that did trigger these minimums. 2011 Report 54. Thus, the Table associated 400 grams of powder cocaine (an amount that fell just below the amount triggering the 1986 Drug Act’s 5-year minimum) with an offense level of 24, which for a first-time offender meant a sentencing range of 51 to 63 months (the range just below the 5-year minimum). USSG §2D1.1 (Oct. 1987). Following the 100-to-1 crack-to-powder ratio, the Table associated four grams of crack (an amount that also fell just below the amount triggering the 1986 Drug Act’s 5-year minimum) with an offense level of 24. Ibid. The Commission did this not because it necessarily thought that those levels were most in keeping with past sentencing practice or would independently have reflected a fair set of sentences, but rather because the Commission believed that doing so was the best way to keep similar drug-trafficking sentences proportional, thereby satisfying the Sentencing Reform Act’s basic “proportionality” objective. See Kimbrough, 552 U. S., at 97; USSG §1A1.3 (Nov. 2011); 2011 Report 53–54, 349, and n. 845. For this reason, the Commission derived the Drug Quantity Table’s entire set of crack and powder cocaine offense levels by using the 1986 Drug Act’s two (5- and 10-year) minimum amounts as reference points and then extrapolating from those two amounts upward and downward to set proportional offense levels for other drug amounts. Ibid. B During the next two decades, the Commission and others in the law enforcement community strongly criticized Congress’ decision to set the crack-to-powder mandatory minimum ratio at 100-to-1. The Commission issued four separate reports telling Congress that the ratio was too high and unjustified because, for example, research showed the relative harm between crack and powder cocaine less severe than 100-to-1, because sentences embodying that ratio could not achieve the Sentencing Reform Act’s “uniformity” goal of treating like offenders alike, because they could not achieve the “proportionality” goal of treating different offenders (e.g., major drug traffickers and low-level dealers) differently, and because the public had come to understand sentences embodying the 100-to-1 ratio as reflecting unjustified race-based differences. Kimbrough, supra, at 97–98; see, e.g., USSC, Special Report to the Congress: Cocaine and Federal Sen-tencing Policy 197–198 (Feb. 1995) (1995 Report); USSC, Special Report to Congress: Cocaine and Federal Sentencing Policy 8 (Apr. 1997) (1997 Report); USSC, Report to Congress: Cocaine and Federal Sentencing Policy 91, 103 (May 2002) (2002 Report); USSC, Report to Congress: Cocaine and Federal Sentencing Policy 8 (May 2007) (2007 Report). The Commission also asked Congress for new legislation embodying a lower crack-to-powder ratio. 1995 Report 198–200; 1997 Report 9–10; 2002 Report 103– 107; 2007 Report 6–9. And the Commission recommended that the legislation “include” an “emergency amendment” allowing “the Commission to incorporate the statutory changes” in the Guidelines while “minimiz[ing] the lag between any statutory and guideline modifications for cocaine offenders.” Id., at 9. In 2010, Congress accepted the Commission’s recommendations, see 2002 Report 104; 2007 Report 8–9, and n. 26, and enacted the Fair Sentencing Act into law. The Act increased the drug amounts triggering mandatory minimums for crack trafficking offenses from 5 grams to 28 grams in respect to the 5-year minimum and from 50 grams to 280 grams in respect to the 10-year minimum (while leaving powder at 500 grams and 5,000 grams respectively). §2(a), 124Stat. 2372. The change had the effect of lowering the 100-to-1 crack-to-powder ratio to 18-to-1. (The Act also eliminated the 5-year mandatory minimum for simple possession of crack. §3, 124Stat. 2372.) Further, the Fair Sentencing Act instructed the Commission to “make such conforming amendments to the Federal sentencing guidelines as the Commission determines necessary to achieve consistency with other guideline provisions and applicable law.” §8(2), id., at 2374. And it directed the Commission to “promulgate the guidelines, policy statements, or amendments provided for in this Act as soon as practicable, and in any event not later than 90 days” after the new Act took effect. §8(1), ibid. The Fair Sentencing Act took effect on August 3, 2010. The Commission promulgated conforming emergency Guidelines amendments that became effective on November 1, 2010. 75 Fed. Reg. 66188 (2010). A permanent version of those Guidelines amendments took effect on November 1, 2011. See 76 id., at 24960 (2011). C With this background in mind, we turn to the relevant facts of the cases before us. Corey Hill, one of the petitioners, unlawfully sold 53 grams of crack in March 2007, before the Fair Sentencing Act became law. App. in No. 11–5721, pp. 6, 83 (hereinafter Hill App.). Under the 1986 Drug Act, an offender who sold 53 grams of crack was subject to a 10-year mandatory minimum. 21 U. S. C. §841(b)(1)(A)(iii) (2006 ed.). Hill was not sentenced, however, until December 2010, after the Fair Sentencing Act became law and after the new Guidelines amendments had become effective. Hill App. 83–94. Under the Fair Sentencing Act, an offender who sold 53 grams of crack was subject to a 5-year, not a 10-year, minimum. §841(b)(1)(B)(iii) (2006 ed., Supp. IV). The sentencing judge stated that, if he thought that the Fair Sentencing Act applied, he would have sentenced Hill to that Act’s 5-year minimum. Id., at 69. But he concluded that the Fair Sentencing Act’s lower minimums apply only to those who committed a drug crime after August 3, 2010—the Act’s effective date. Id., at 65, 68. That is to say, he concluded that the new Act’s more lenient sentences did not apply to those who committed a crime before August 3, even if they were sentenced after that date. Hence, the judge sen-tenced Hill to 10 years of imprisonment. Id., at 78. The Court of Appeals for the Seventh Circuit affirmed. 417 Fed. Appx. 560 (2011). The second petitioner, Edward Dorsey (who had previously been convicted of a drug felony), unlawfully sold 5.5 grams of crack in August 2008, before the Fair Sentencing Act took effect. App. in No. 5683, pp. 9, 48–49, 57–58 (hereinafter Dorsey App.). Under the 1986 Drug Act, an offender such as Dorsey with a prior drug felony who sold 5.5 grams of crack was subject to a 10-year minimum. §841(b)(1)(B)(iii) (2006 ed.). Dorsey was not sentenced, however, until September 2010, after the new Fair Sentencing Act took effect. Id., at 84–95. Under the Fair Sentencing Act, such an offender who sold 5.5 grams of crack was not subject to a mandatory minimum at all, for 5.5 grams is less than the 28 grams that triggers the new Act’s mandatory minimum provisions. §841(b)(1)(B)(iii) (2006 ed., Supp. IV). Dorsey asked the judge to apply the Fair Sentencing Act’s more lenient statutory penalties. Id., at 54–55. Moreover, as of Dorsey’s sentencing in September 2010, the unrevised Guidelines (reflecting the 1986 Drug Act’s old minimums) were still in effect. The Commission had not yet finished revising the Guidelines to reflect the new, lower statutory minimums. And the basic sentencing statute, the Sentencing Reform Act, provides that a judge shall apply the Guidelines that “are in effect on the date the defendant is sentenced.” 18 U. S. C. §3553(a)(4)(A)(ii). The sentencing judge, however, had the legal authority not to apply the Guidelines at all (for they are advisory). But he also knew that he could not ignore a minimum sentence contained in the applicable statute. Dorsey App. 67–68. The judge noted that, even though he was sentencing Dorsey after the effective date of the Fair Sentencing Act, Dorsey had committed the underlying crime prior to that date. Id., at 69–70. And he concluded that the 1986 Drug Act’s old minimums, not the new Fair Sentencing Act, applied in those circumstances. Ibid. He consequently sentenced Dorsey to the 1986 Drug Act’s 10-year man-datory minimum term. Id., at 80. The Court of Appeals for the Seventh Circuit affirmed, United States v. Fisher, 635 F.3d 336 (2011), and denied rehearing en banc, 646 F.3d 429 (2011) (per curiam); see also United States v. Holcomb, 657 F.3d 445 (CA7 2011). The Courts of Appeals have come to different conclusions as to whether the Fair Sentencing Act’s more lenient mandatory minimums apply to offenders whose unlawful conduct took place before, but whose sentencing took place after, the date that Act took effect, namely, August 3, 2010. Compare United States v. Douglas, 644 F.3d 39, 42–44 (CA1 2011) (Act applies), and United States v. Dixon, 648 F.3d 195, 203 (CA3 2011) (same), with 635 F. 3d, at 339–340 (Act does not apply), United States v. Sidney, 648 F.3d 904, 910 (CA8 2011) (same), and United States v. Tickles, 661 F.3d 212, 215 (CA5 2011) (per curiam) (same). In light of that disagreement, we granted Hill’s and Dorsey’s petitions for certiorari. Since petitioners and the Government both take the position that the Fair Sentencing Act’s new minimums do apply in these circumstances, we appointed as amicus curiae Miguel Estrada to argue the contrary position. He has ably discharged his responsibilities. II A The timing issue before us is difficult in part because relevant language in different statutes argues in opposite directions. See Appendix A, infra. On the one hand, a federal saving statute, Act of Feb. 25, 1871 (1871 Act), §4, 16Stat. 432, phrased in general terms, provides that a new criminal statute that “repeal[s]” an older criminal statute shall not change the penalties “incurred” under that older statute “unless the repealing Act shall so expressly provide.” 1 U. S. C. §109. Case law makes clear that the word “repeal” applies when a new statute simply diminishes the penalties that the older statute set forth. See Warden v. Marrero, 417 U.S. 653, 659–664 (1974); see also United States v. Tynen, 11 Wall. 88, 92 (1871). Case law also makes clear that penalties are “incurred” under the older statute when an offender becomes subject to them, i.e., commits the underlying conduct that makes the offender liable. See United States v. Reisinger, 128 U.S. 398, 401 (1888); Great Northern R. Co. v. United States, 208 U.S. 452, 464–470 (1908). On the other hand, the Sentencing Reform Act says that, regardless of when the offender’s conduct occurs, the applicable Guidelines are the ones “in effect on the date the defendant is sentenced.” 18 U. S. C. §3553(a)(4)(A)(ii). And the Fair Sentencing Act requires the Commission to change the Guidelines in the wake of the Act’s new minimums, making them consistent with “other guideline provisions and applicable law.” §8(2), 124Stat. 2374. Courts that have held that they must apply the old, higher 1986 Drug Act minimums to all pre-Act offenders, including those sentenced after the Fair Sentencing Act took effect, have emphasized that the 1871 Act requires that result unless the Fair Sentencing Act either expressly says or at least by fair implication implies the contrary. See 635 F. 3d, at 339–340; Sidney, supra, at 906–908; Tickles, supra, at 214–215; see also Holcomb, supra, at 446–448 (opinion of Easterbrook, J.). Courts that have concluded that the Fair Sentencing Act’s more lenient penalties apply have found in that Act, together with the Sentencing Reform Act and other related circumstances, indicia of a clear congressional intent to apply the new Act’s minimums. See Douglas, supra, at 42–44; Dixon, supra, at 199–203; see also Holcomb, 657 F. 3d, at 454–457 (Williams, J., dissenting from denial of rehearing en banc); id., at 461–463 (Posner, J., dissenting from denial of rehearing en banc). We too take the latter view. Six considerations, taken together, convince us that Congress intended the Fair Sentencing Act’s more lenient penalties to apply to those offenders whose crimes preceded August 3, 2010, but who are sentenced after that date. First, the 1871 saving statute permits Congress to apply a new Act’s more lenient penalties to pre-Act offenders without expressly saying so in the new Act. It is true that the 1871 Act uses the words “expressly provide.” 1 U. S. C. §109. But the Court has long recognized that this saving statute creates what is in effect a less demanding interpretive requirement. That is because statutes en- acted by one Congress cannot bind a later Congress, which remains free to repeal the earlier statute, to exempt the current statute from the earlier statute, to modify the earlier statute, or to apply the earlier statute but as modified. See, e.g., Fletcher v. Peck, 6 Cranch 87, 135 (1810); Reichelderfer v. Quinn, 287 U.S. 315, 318 (1932). And Congress remains free to express any such intention either expressly or by implication as it chooses. Thus, the Court has said that the 1871 Act “cannot justify a disregard of the will of Congress as manifested either expressly or by necessary implication in a subsequent enactment.” Great Northern R. Co., supra, at 465 (emphasis added). And in a comparable context the Court has emphasized that the Administrative Procedure Act’s use of the word “expressly” does not require Congress to use any “magical passwords” to exempt a later statute from the provision. Marcello v. Bonds, 349 U.S. 302, 310 (1955). Without requiring an “express” statement, the Court has described the necessary indicia of congressional intent by the terms “necessary implication,” “clear implication,” and “fair implication,” phrases it has used interchangeably. Great Northern R. Co., supra, at 465, 466; Hertz v. Woodman, 218 U.S. 205, 218 (1910); Marrero, supra, at 660, n. 10. One Member of the Court has said we should determine whether “the plain import of a later statute directly conflicts with an earlier statute,” and, if so, “the later enactment governs, regardless of its compliance with any earlier-enacted requirement of an express reference or other ‘magical password.’ ” Lockhart v. United States, 546 U.S. 142, 149 (2005) (Scalia, J., concurring). Hence, the Court has treated the 1871 Act as setting forth an important background principle of interpretation. The Court has also assumed Congress is well aware of the background principle when it enacts new criminal statutes. E.g., Great Northern R. Co., supra, at 465; Hertz, supra, at 217; cf. Marcello, supra, at 310. And the prin-ciple requires courts, before interpreting a new criminal statute to apply its new penalties to a set of pre-Act offenders, to assure themselves that ordinary interpretive considerations point clearly in that direction. Words such as “plain import,” “fair implication,” or the like reflect the need for that assurance. And it is that assurance, which we shall assume is conveyed by the phrases “plain import” or “fair implication,” that we must look for here. Second, the Sentencing Reform Act sets forth a special and different background principle. That statute says that when “determining the particular sentence to be imposed” in an initial sentencing, the sentencing court “shall consider,” among other things, the “sentencing range” established by the Guidelines that are “in effect on the date the defendant is sentenced.” 18 U. S. C. §3553(a)(4)(A)(ii) (emphasis added). Although the Constitution’s Ex Post Facto Clause, Art. I, §9, cl. 3, prohibits applying a new Act’s higher penalties to pre-Act conduct, it does not prohibit applying lower penalties. See Calder v. Bull, 3 Dall. 386, 390–391 (1798); Collins v. Youngblood, 497 U.S. 37, 41–44 (1990). The Sentencing Commission has consequently instructed sentencing judges to “use the Guidelines Manual in effect on the date that the defendant is sentenced,” regardless of when the defendant committed the offense, unless doing so “would violate the ex post facto clause.” USSG §1B1.11. And therefore when the Commission adopts new, lower Guidelines amendments, those amendments become effective to offenders who committed an offense prior to the adoption of the new amendments but are sentenced thereafter. Just as we assume Congress was aware of the 1871 Act’s background norm, so we assume that Congress was aware of this different background sentencing principle. Third, language in the Fair Sentencing Act implies that Congress intended to follow the Sentencing Reform Act background principle here. A section of the Fair Sentencing Act entitled “Emergency Authority for United States Sentencing Commission” requires the Commission to prom-ulgate “as soon as practicable” (and not later than 90 days after August 3, 2010) “conforming amendments” to the Guidelines that “achieve consistency with other guideline provisions and applicable law.” §8, 124Stat. 2374. Read most naturally, “applicable law” refers to the law as changed by the Fair Sentencing Act, including the provision reducing the crack mandatory minimums. §2(a), id., at 2372. As the Commission understood this provision, achieving consistency with “other guideline provisions” means reducing the base offense levels for all crack amounts proportionally (using the new 18-to-1 ratio), in-cluding the offense levels governing small amounts of crack that did not fall within the scope of the mandatory minimum provisions. 75 Fed. Reg. 66191. And consis-tency with “other guideline provisions” and with prior Com-mission practice would require application of the new Guidelines amendments to offenders who committed their offense prior to the new amendments’ effective date but were sentenced thereafter. See USSG §1B1.11(a); e.g., USSG App. C, amdts. 706, 711 (Supp. Nov. 2004–Nov. 2007); see also Memorandum from G. Schmitt, L. Reed, & K. Cohen, USSC, to Chair Hinojosa et al., Subject: Analysis of the Impact of the Crack Cocaine Amendment if Made Retroactive 23 (Oct. 3, 2007). Cf. USSG App. C, amdt. 571 (amendment increasing restitution, which may present ex post facto and one-book-rule concerns, would apply only to defendants sentenced for post-amendment offenses), discussed post, at 5 (Scalia, J., dissenting). Fourth, applying the 1986 Drug Act’s old mandatory minimums to the post-August 3 sentencing of pre-August 3 offenders would create disparities of a kind that Congress enacted the Sentencing Reform Act and the Fair Sentencing Act to prevent. Two individuals with the same number of prior offenses who each engaged in the same criminal conduct involving the same amount of crack and were sentenced at the same time would receive radically different sentences. For example, a first-time post-Act offender with five grams of crack, subject to a Guidelines range of 21 to 27 months, could receive two years of imprisonment, while an otherwise identical pre-Act offender would have to receive the 5-year mandatory minimum. Compare USSG §2D1.1(c) (Nov. 2011) with 21 U. S. C. §841(b)(1)(B) (2006 ed.). A first-time post-Act 50-gram offender would be subject to a Guidelines range of less than six years of imprisonment, while his otherwise identical pre-Act counterpart would have to receive the 10-year mandatory minimum. Compare USSG §2D1.1(c) (Nov. 2011) with 21 U. S. C. §841(b)(1)(A) (2006 ed.). Moreover, unlike many prechange/postchange discrepancies, the imposition of these disparate sentences involves roughly contemporaneous sentencing, i.e., the same time, the same place, and even the same judge, thereby highlighting a kind of unfairness that modern sentenc- ing statutes typically seek to combat. See, e.g., 28 U. S. C. §991(b)(1)(B) (purposes of Guidelines-based sentencing include “avoiding unwarranted sentencing disparities among defendants with similar records who have been found guilty of similar criminal conduct”); S. Rep. No. 98–223, p. 74 (1983) (explaining rationale for using same, current Guidelines for all roughly contemporaneous sentencings). Further, it would involve imposing upon the pre-Act offender a pre-Act sentence at a time after Congress had specifically found in the Fair Sentencing Act that such a sentence was unfairly long. Finally, one cannot treat such problems as if they were minor ones. Given the 5-year statute of limitations for federal drug offenses, the 11-month median time between indictment and sentencing for those offenses, and the approximately 5,000 federal crack offenders convicted each year, many pre-Act offenders were not (and will not be) sentenced until after August 3, 2010, when the new, more lenient mandatory minimums took effect. See 18 U. S. C. §3282(a); Administrative Office of United States Courts, Judicial Business of the United States Courts, p. 272 (2010) (Table D–10); 2011 Report 191. Fifth, not to apply the Fair Sentencing Act would do more than preserve a disproportionate status quo; it would make matters worse. It would create new anomalies—new sets of disproportionate sentences—not previously present. That is because sentencing courts must apply new Guidelines (consistent with the Fair Sentencing Act’s new minimums) to pre-Act offenders, see supra, at 13–14, and the 1986 Drug Act’s old minimums would trump those new Guidelines for some pre-Act offenders but not for all of them—say, pre-Act offenders who possessed crack in small amounts not directly the subject of mandatory minimums. Consider, for example, a first-time offender convicted of possessing with intent to distribute four grams of crack. No mandatory sentence, under the 1986 Drug Act or the Fair Sentencing Act, applies to an offender possessing so small an amount. Yet under the old law, the Commission, charged with creating proportionate sentences, had created a Guidelines range of 41 to 51 months for such an of-fender, a sentence proportional to the 60 months that the 1986 Drug Act required for one who trafficked five grams of crack. See supra, at 5–6; USSG §2D1.1(c) (Nov. 2009). The Fair Sentencing Act, however, requires the Commission to write new Guidelines consistent with the new law. The Commission therefore wrote new Guidelines that provide a sentencing range of 21 to 27 months—about two years—for the first-time, 4-gram offender. See USSG §2D1.1(c) (Nov. 2011). And the Sentencing Reform Act requires application of those new Guidelines to all of- fenders (including pre-Act offenders) who are sentenced once those new Guidelines take effect. See 18 U. S. C. §3553(a)(4)(A)(ii). Those new Guidelines must take effect and apply to a pre-Act 4-gram offender, for such an offender was never subject to a trumping statutory 1986 Drug Act mandatory minimum. However, unless the Fair Sentencing Act’s new, more lenient mandatory mini- mums apply to pre-Act offenders, an otherwise identical of-fender who possessed five grams would have to receive a 5-year sentence. See 21 U. S. C. §841(b)(1)(B) (2006 ed., Supp. IV). For example, imagine that on July 1, 2010, both Smith and Jones commit a crack crime identical but for the fact that Smith possesses with intent to distribute four grams of crack and Jones five grams. Both are sentenced on December 1, 2010, after the Fair Sentencing Act and the new Guidelines take effect. Smith’s Guidelines sentence would be two years, but unless the Fair Sentencing Act applies, Jones’s sentence would have to be five years. The difference of one gram would make a difference, not of only one year as it did before enactment of the Fair Sentencing Act, but instead of three years. Passage of the new Act, designed to have brought about fairer sentences, would here have created a new disparate sentencing “cliff.” Nor can one say that the new Act would produce disproportionalities like this in only a few cases. In fiscal year 2010, 17.8 percent of all crack offenders were convicted of- offenses not subject to the 1986 Drug Act’s minimums. 2011 Report 191. And since those minimums apply only to some drug offenders and they apply in different ways, one can find many similar examples of disproportionalities. See Appendix B, infra. Thus, application of the 1986 Drug Act minimums to pre-Act offenders sentenced after the new Guidelines take effect would produce a crazy quilt of sentences, at odds with Congress’ basic efforts to achieve more uniform, more proportionate sentences. Congress, when enacting the Fair Sentencing Act, could not have intended any such result. Sixth, we have found no strong countervailing consideration. Amicus and the dissent argue that one might read much of the statutory language we have discussed as embodying exceptions, permitting the old 1986 Drug Act minimums to apply to pre-Act offenders sentenced after August 3, 2010, when the Fair Sentencing Act took effect. The words “applicable law” in the new Act, for example, could, linguistically speaking, encompass the 1986 Drug Act minimums applied to those sentenced after August 3. Post, at 4–6 (Scalia, J., dissenting). Moreover, Congress could have insisted that the Commission write new Guidelines with special speed to assure itself that new, post-August 3 offenders—but not old, pre-August 3 offenders—would receive the benefit of the new Act. Post, at 6–8. Further, amicus and the dissent note that to apply the new Act’s minimums to the old, pre-August 3 offenders will create a new disparity—one between pre-Act offenders sentenced before August 3 and those sentenced after that date. Post, at 9. We do not believe that these arguments make a critical difference. Even if the relevant statutory language can be read as amicus and the dissent suggest and even if Congress might have wanted Guidelines written speedily simply in order to apply them quickly to new offenders, there is scant indication that this is what Congress did mean by the language in question nor that such was in fact Congress’ motivation. The considerations we have set forth, supra, at 13–17 and this page, strongly suggest the contrary. We also recognize that application of the new minimums to pre-Act offenders sentenced after August 3 will create a new set of disparities. But those disparities, reflecting a line-drawing effort, will exist whenever Congress enacts a new law changing sentences (unless Congress intends re-opening sentencing proceedings concluded prior to a new law’s effective date). We have explained how in federal sentencing the ordinary practice is to apply new penalties to defendants not yet sentenced, while withholding that change from defendants already sentenced. Supra, at 13; compare 18 U. S. C. §3553(a)(4)(A)(ii) with §3582(c). And we have explained how, here, continued application of the old 1986 Drug Act minimums to those pre-Act offenders sentenced after August 3 would make matters worse. Supra, at 16–18. We consequently conclude that this particular new disparity (between those pre-Act offenders already sentenced and those not yet sentenced as of August 3) cannot make a critical difference. For these reasons considered as a whole, we conclude that Congress intended the Fair Sentencing Act’s new, lower mandatory minimums to apply to the post-Act sentencing of pre-Act offenders. That is the Act’s “plain import” or “fair implication.” B We add one final point. Several arguments we have discussed involve the language of statutes that determine how new Guidelines take effect. Supra, at 13–14. What about those who committed an offense prior to August 3 and were sentenced after August 3 but before November 1, 2010—a period after the new Act’s effective date but before the new Guidelines first took effect? Do the Fair Sentencing Act’s new mandatory minimums apply to them? In our view, the new Act’s lower minimums apply to them as well. Our reason is that the statute simply instructs the Commission to promulgate new Guidelines “as soon as practicable” (but no later than 90 days after the Act took effect). §8(1), 124Stat. 2374. As far as Congress was concerned, the Commission might have (having prepared new Guidelines in advance) promulgated those Guidelines within a few days—perhaps on August 3 itself. At the same time, the Commission possesses ample authority to permit appropriate adjustments to be made in the Guidelines sentences of those sentenced after August 3 but prior to the new Guidelines promulgation. See 28 U. S. C. §994(u) (power to make Guidelines reductions retroactive); 76 Fed. Reg. 41333–41334 (2011) (amended 18-to-1 Guidelines made retroactive). In any event, courts, treating the Guidelines as advisory, possess authority to sentence in accordance with the new minimums. For these reasons, if the Fair Sentencing Act’s new minimums apply to all of those sentenced after August 3, 2010 (even if the new Guidelines were not yet ready), it is possible to foresee a reasonably smooth transition. On the other hand, it is difficult to foresee such a transition if the new Act’s application is keyed to a later date, thereby leaving the courts unable to take the new Act fully into account, particularly when that circumstance might create additional disparities and uncertainties that courts and the Commission may be helpless to correct. We have no reason to believe Congress would have wanted to impose an unforeseeable, potentially complex application date. * * * We vacate the Court of Appeals’ judgments and remand these cases for further proceedings consistent with this opinion. It is so ordered. APPENDIXES A Act of Feb. 25, 1871, §4, 16Stat. 432, 1 U. S. C. §109 Repeal of statutes as affecting existing liabilities “The repeal of any statute shall not have the effect to release or extinguish any penalty, forfeiture, or liability incurred under such statute, unless the repealing Act shall so expressly provide, and such statute shall be treated as still remaining in force for the purpose of sustaining any proper action or prosecution for the enforcement of such penalty, forfeiture, or liability.” Sentencing Reform Act of 1984, 18 U. S. C. §3553(a)(4) (A)(ii) Imposition of a sentence “Factors To Be Considered in Imposing a Sen-tence. . . . The court, in determining the particu- lar sentence to be imposed, shall consider . . . the kinds of sentence and sentencing range established for . . . the applicable category of offense committed by the applicable category of defendant as set forth in the guidelines . . . that . . . are in effect on the date the defendant is sentenced . . . .” Fair Sentencing Act of 2010, §8, 124Stat. 2374 Emergency Authority for United States Sentencing Commission “The United States Sentencing Commission shall— “(1) promulgate the guidelines, policy statements, or amendments provided for in this Act as soon as practicable, and in any event not later than 90 days after the date of enactment of this Act, in accordance with the procedure set forth in section 21(a) of the Sentencing Act of 1987 (28 U. S. C. [§]994 note), as though the authority under that Act had not expired; and “(2) pursuant to the emergency authority provided under paragraph (1), make such conforming amendments to the Federal sentencing guidelines as the Commission determines necessary to achieve consistency with other guideline provisions and appli-cable law.” B The following chart shows the sentencing scheme that would result for first-time pre-Act crack offenders if the 1986 Drug Act’s old 100-to-1 mandatory minimums remain in effect after the Fair Sentencing Act’s new 18-to-1 Guidelines became effective. 21 U. S. C. §§841(b)(1)(A)–(C) (2006 ed.); USSG §§2D1.1(c), 5G1.1(b) (Nov. 2011). 1986 Drug Act Minimums and Fair Sentencing Act Guidelines for Category I Offenders with No Prior Drug Felonies The chart illustrates the disproportionate sentences that such a scheme would create. See supra, at 16–18. For one thing, it would create sentencing “cliffs” at the 1986 Act’s old triggering amounts of 5 grams and 50 grams (where the old minimums would entirely trump the new Guidelines), resulting in radically different Guidelines sentences for small differences in quantity. For another, because of those “cliffs,” the scheme would create similar Guidelines sentences for offenders who dealt in radically different amounts of crack, e.g., 50 grams versus 500 grams. To be sure, as amicus points out, Congress has provided two mechanisms through which an offender may escape an otherwise applicable mandatory minimum, diminishing this problem for some offenders. First, an offender may escape a minimum by providing substantial assistance in the investigation or prosecution of another person. 18 U. S. C. §3553(e); Fed. Rule Crim. Proc. 35(b); see also 28 U. S. C. §994(n); USSG §5K1.1. Second, under 18 U. S. C. §3553(f), drug offenders who have little or no criminal history and who satisfy other requirements in the provision may obtain “safety valve” relief. See also USSG §5C1.2. And because of these mechanisms a substantial portion of first-time offenders are relieved of application of a manda-tory minimum. However, offenders with a criminal his-tory category of II or higher are ineligible for “safety valve” relief; they escape application of a minimum at a much lower percentage. See 2011 Report 193 (Table 8–8). Crack Offender Categories by Application of 1986 Drug Act Mandatory Min. (FY 2010) Yet similar sentencing anomalies would result for repeat offenders if the 1986 Drug Act’s minimums remain in effect after the Fair Sentencing Act’s Guidelines became effective. Take, for example, Category II offenders. 1986 Drug Act Minimums and Fair Sentencing Act Guidelines for Category II Offenders with No Prior Drug Felonies As the chart illustrates, for Category II offenders accountable for 5 to 22 grams of crack or for 50 to 195 grams, the 100-to-1 minimums would entirely trump the 18-to-1 Guidelines, producing the same anomalies—dissimilar sen-tences for similar quantities and similar sentences for dis-similar quantities—described above. In contrast, a scheme with the Fair Sentencing Act’s 18-to-1 minimums and new Guidelines produces the proportionality in sentencing that Congress intended in enacting the Sentencing Reform Act and the Fair Sentencing Act. Fair Sentencing Act Minimums and Guidelines for Category II Offenders with No Prior Drug Felonies

565.US.2011_09-958

Medicaid is a cooperative federal-state program that provides medical care to needy individuals. To qualify for federal funds, a State must submit its Medicaid plan and any amendments to the federal agency that administers the program, the Centers for Medicare & Medicaid Services (CMS). Before approving a plan or amendments, CMS conducts a review to determine whether they comply with federal requirements. Federal law requires state plans or amendments to “assure that payments are consistent with efficiency, economy, and quality of care and are sufficient to enlist enough providers” to make Medicaid “care and services” available. 42 U. S. C. §1396a(a)(30)(A). After California enacted three statutes reducing the State’s payments to various Medicaid providers, the State submitted plan amendments to CMS. Before the agency finished its review, Medicaid providers and beneficiaries sought, in a series of cases, to enjoin the rate reductions on the ground that they were pre-empted by federal Medicaid law. In seven decisions, the Ninth Circuit ultimately affirmed or ordered preliminary injunctions preventing the State from implementing its statutes. The court (1) held that the providers and beneficiaries could bring a Supremacy Clause action; (2) essentially accepted their claim that the State did not show that its amended plan would provide sufficient services; (3) held that the amendments thus conflicted with §1396a(a)(30)(A); and (4) held that the federal statute pre-empted the new state laws. In the meantime, agency officials disapproved the amendments, and California sought further administrative review. The cases were in this posture when the Court granted certiorari to decide whether respondents could mount a Supremacy Clause challenge. After oral argument, CMS approved several of the State’s amendments, and the State withdrew its requests for approval of the remainder. Held: The judgments are vacated and the cases are remanded, thereby permitting the parties to argue before the Ninth Circuit in the first instance the question whether respondents may maintain Supremacy Clause actions now that CMS has approved the state statutes. Pp. 5–8. (a) CMS’ approval does not make these cases moot, but it does put them in a different posture, since the federal agency charged with administering Medicaid has now found that the rate reductions comply with federal law. That decision does not change the substantive question whether California’s statutes are consistent with federal law, but it may change the answer. It may also require respondents to seek review of CMS’ determination under the Administrative Procedure Act (APA) rather than in a Supremacy Clause action against California. The APA would likely permit respondents to obtain an authoritative judicial determination of the merits of their legal claim. And their basic challenge now presents the kind of legal question ordinarily calling for APA review. The Medicaid Act commits to a federal agency the power to administer a federal program, and the agency has exercised that authority. As CMS is comparatively expert in the statute’s subject matter, its decision carries weight. And §1396a(a)(30)(A)’s broad and general language suggests that CMS’ expertise is relevant in determining the provision’s application. Finally, to allow a Supremacy Clause action to proceed once CMS has reached a decision threatens potential inconsistency or confusion. The Ninth Circuit declined to give weight to the Federal Government’s interpretation of the federal law, but courts are ordinarily required to apply deference standards to agency decisionmaking. The parties suggest no reasons why such standards should not be applied here or why, now that CMS has acted, a court should reach a different result in an APA action than in a Supremacy Clause action. That would make the Supremacy Clause challenge at best redundant. Permitting it to continue would seem inefficient, for the federal agency is not a participant in the action, which will decide whether agency-approved state rates violate federal law. Pp. 5–8. (b) Given the present posture of the cases, the Court does not address whether the Ninth Circuit properly recognized a Supremacy Clause action to enforce the federal law before the agency took final action. To decide whether these cases may proceed under the Supremacy Clause now that the agency has acted, it will be necessary on remand to consider at least the matters addressed by this Court. P. 8. No. 09–958, 572 F.3d 644 (first judgment), 342 Fed. Appx. 306 (second judgment), No. 09–1158, 596 F.3d 1098, 563 F.3d 847, 374 Fed. Appx. 690, 596 F.3d 1087, and No. 10–283, 380 Fed. Appx. 656, vacated and remanded. Breyer, J., delivered the opinion of the Court, in which Kennedy, Ginsburg, Sotomayor, and Kagan, JJ., joined. Roberts, C. J., filed a dissenting opinion, in which Scalia, Thomas, and Alito, JJ., joined. Notes 1 Together with No. 09–1158, Douglas, Director, California Department of Health Care Services v. California Pharmacists Association et al., Douglas, Director, California Department of Health Care Services v. California Hospital Association et al. (see this Court’s Rule 12.4), Douglas, Director, California Department of Health Care Services v. Independent Living Center of Southern California, Inc., et al. (see this Court’s Rule 12.4), Douglas, Director, California Department of Health Care Services v. Dominguez, By and Through her Mother and Next Friend Brown, et al. (see this Court’s Rule 12.4); and No. 10–283, Douglas, Director, California Department of Health Care Services v. Santa Rosa Memorial Hospital et al., also on certiorari to the same court.

We granted certiorari in these cases to decide whether Medicaid providers and recipients may maintain a cause of action under the Supremacy Clause to enforce a federal Medicaid law—a federal law that, in their view, conflicts with (and pre-empts) state Medicaid statutes that reduce payments to providers. Since we granted certiorari, however, the relevant circumstances have changed. The federal agency in charge of administering Medicaid, the Cen- ters for Medicare & Medicaid Services (CMS), has now approved the state statutes as consistent with the federal law. In light of the changed circumstances, we believe that the question before us now is whether, once the agency has approved the state statutes, groups of Medicaid providers and beneficiaries may still maintain a Suprem- acy Clause action asserting that the state statutes are inconsistent with the federal Medicaid law. For the reasons set forth below, we vacate the Ninth Circuit’s judgments and remand these cases for proceedings consistent with this opinion. I A Medicaid is a cooperative federal-state program that provides medical care to needy individuals. To qualify for federal funds, States must submit to a federal agency (CMS, a division of the Department of Health and Human Services) a state Medicaid plan that details the nature and scope of the State’s Medicaid program. It must also submit any amendments to the plan that it may make from time to time. And it must receive the agency’s approval of the plan and any amendments. Before granting approval, the agency reviews the State’s plan and amendments to determine whether they comply with the statutory and regulatory requirements governing the Medicaid program. See 42 U. S. C. §§1316(a)(1), (b), 1396a(a), (b); 42 CFR §430.10 et seq. (2010); Wilder v. Virginia Hospital Assn., 496 U.S. 498, 502 (1990). And the agency’s director has specified that the agency will not provide federal funds for any state plan amendment until the agency approves the amendment. See Letter from Timothy M. Westmoreland, Director, Center for Medicaid & State Operations, Health Care Financing Admin., U. S. Dept. of Health and Human Servs., to State Medicaid Director (Jan. 2, 2001), online at http://www.cms.gov/SMDL/downloads/SMD010201.pdf (as visited Feb. 17, 2012, and available in Clerk of Court’s case file). The federal statutory provision relevant here says that a State’s Medicaid plan and amendments must: “provide such methods and procedures relating to the utilization of, and the payment for, care and services available under the plan . . . as may be necessary to safeguard against unnecessary utilization of such care and services and to assure that payments are consistent with efficiency, economy, and quality of care and are sufficient to enlist enough providers so that care and services are available under the plan at least to the extent that such care and services are available to the general population in the geographic area.” 42 U. S. C. §1396a(a)(30)(A) (emphasis added). B In 2008 and 2009, the California Legislature passed three statutes changing that State’s Medicaid plan. The first statute, enacted in February 2008, reduced by 10% payments that the State makes to various Medicaid providers, such as physicians, pharmacies, and clinics. See 2007–2008 Cal. Sess. Laws, 3d Extraordinary Sess. ch. 3, §§14, 15. The second statute, enacted in September 2008, replaced the 10% rate reductions with a more modest set of cuts. See 2008 Cal. Sess. Laws ch. 758, §§45, 57. And the last statute, enacted in February 2009, placed a cap on the State’s maximum contribution to wages and benefits paid by counties to providers of in-home supportive services. See 2009–2010 Cal. Sess. Laws, 3d Extraordinary Sess. ch. 13, §9. In September and December 2008, the State submitted to the federal agency a series of plan amendments designed to implement most of the reductions contained in these bills. Before the agency finished reviewing the amendments, however, groups of Medicaid providers and beneficiaries filed a series of lawsuits seeking to enjoin the rate reductions on the ground that they conflicted with, and therefore were pre-empted by, federal Medicaid law, in particular the statutory provision that we have just set forth. They argued that California’s Medicaid plan amendments were inconsistent with the federal provision because the State had failed to study whether the rate reductions would be consistent with the statutory factors of efficiency, economy, quality, and access to care. In effect, they argued that California had not shown that its Medicaid plan, as amended, would “enlist enough providers” to make Medicaid “care and services” sufficiently available. 42 U. S. C. §1396a(a)(30)(A). The consolidated cases before us encompass five lawsuits brought by Medicaid providers and beneficiaries against state officials. Those cases produced seven decisions of the Court of Appeals for the Ninth Circuit. See 572 F.3d 644 (2009); 342 Fed. Appx. 306 (2009); 596 F.3d 1098 (2010); 563 F.3d 847 (2009); 374 Fed. Appx. 690 (2010); 596 F.3d 1087 (2010); and 380 Fed. Appx. 656 (2010). The decisions ultimately affirmed or ordered preliminary injunctions that prevented the State from im- plementing its statutes. They (1) held that the Medi- caid providers and beneficiaries could directly bring an action based on the Supremacy Clause; (2) essentially accepted the claim that the State had not demonstrated that its Medicaid plan, as amended, would provide sufficient services; (3) held that the amendments consequently conflicted with the statutory provision we have quoted; and (4) held that, given the Constitution’s Supremacy Clause, the federal statute must prevail. That is to say, the federal statute pre-empted the State’s new laws. In the meantime, the federal agency was also reviewing the same state statutes to determine whether they satisfied the same federal statutory conditions. In November 2010, agency officials concluded that they did not sat- isfy those conditions, and the officials disapproved the amendments. California then exercised its right to further administrative review within the agency. The cases were in this posture when we granted certiorari to decide whether respondents could mount a Supremacy Clause challenge to the state statutes and obtain a court injunction preventing California from implementing its statutes. About a month after we heard oral argument, the federal agency reversed course and approved several of California’s statutory amendments to its plan. See Letter from Donald M. Berwick, Administrator, CMS, to Toby Douglas, Director, Cal. Dept. of Health Care Servs. (Oct. 27, 2011); Letter from Larry Reed, Director, Division of Pharmacy, Disabled and Elderly Health Programs Group, CMS, to Toby Douglas, Director, Cal. Dept. of Health Care Servs. (Oct. 27, 2011). In doing so, the agency also approved a limited retroactive implementation of some of the amendments’ rate reductions. The State, in turn, withdrew its requests for approval of the remaining amendments, in effect agreeing (with one exception) that it would not seek to implement any unapproved reduction. See Letter from Michael E. Kilpatrick, Assistant Chief Counsel, Cal. Dept. of Health Care Servs., to Benjamin R. Cohen, Director, Office of Hearings, CMS (Oct. 27, 2011). (The exception consists of one statute for which California has submitted no amendment and which, by its own terms, cannot take effect unless and until this litigation is complete, see 2010 Cal. Sess. Laws ch. 725, §25.) II All parties agree that the agency’s approval of the enjoined rate reductions does not make these cases moot. For one thing, the providers and beneficiaries continue to believe that the reductions violate the federal provision, the agency’s view to the contrary notwithstanding. For another, federal-court injunctions remain in place, forbidding California to implement the agency-approved rate reductions. And, in light of the agency’s action, California may well ask the lower courts to set those injunctions aside. While the cases are not moot, they are now in a different posture. The federal agency charged with administering the Medicaid program has determined that the challenged rate reductions comply with federal law. That agency decision does not change the underlying substantive question, namely whether California’s statutes are consistent with a specific federal statutory provision (requiring that reimbursement rates be “sufficient to enlist enough providers”). But it may change the answer. And it may require respondents now to proceed by seeking review of the agency determination under the Administrative Procedure Act (APA), 5 U. S. C. §701 et seq., rather than in an action against California under the Supremacy Clause. For one thing, the APA would likely permit respondents to obtain an authoritative judicial determination of the merits of their legal claim. The Act provides for judicial review of final agency action. §704. It permits any person adversely affected or aggrieved by agency action to obtain judicial review of the lawfulness of that action. §702. And it requires a reviewing court to set aside agency action found to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” §706(2)(A). For another thing, respondents’ basic challenge now presents the kind of legal question that ordinarily calls for APA review. The Medicaid Act commits to the federal agency the power to administer a federal program. And here the agency has acted under this grant of authority. That decision carries weight. After all, the agency is comparatively expert in the statute’s subject matter. And the language of the particular provision at issue here is broad and general, suggesting that the agency’s expertise is relevant in determining its application. Finally, to allow a Supremacy Clause action to proceed once the agency has reached a decision threatens potential inconsistency or confusion. In these cases, for example, the Ninth Circuit, in sustaining respondents’ challenges, declined to give weight to the Federal Government’s interpretation of the federal statutory language. (That view was expressed in an amicus curiae brief that the United States submitted in prior litigation.) See Independent Living Center of Southern Cal., Inc. v. Maxwell-Jolly, 572 F.3d 644, 654 (CA9 2009) (referring to the United States’ certiorari-stage invitation brief in Belshe v. Orthopaedic Hospital, 522 U.S. 1044 (1998) (denying writ of certiorari)). And the District Court decisions that underlie injunctions that now forbid California to implement its laws may rest upon similar analysis. But ordinarily review of agency action requires courts to apply certain standards of deference to agency decisionmaking. See National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U.S. 967 (2005) (describing deference reviewing courts must show); Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984) (same). And the parties have not suggested reasons why courts should not now (in the changed posture of these cases) apply those ordinary standards of deference. Nor have the parties suggested reasons why, once the agency has taken final action, a court should reach a different result in a case like this one, depending upon whether the case proceeds in a Supremacy Clause action rather than under the APA for review of an agency decision. Indeed, to permit a difference in result here would subject the States to conflicting interpretations of federal law by several different courts (and the agency), thereby threatening to defeat the uniformity that Congress intended by centralizing administration of the federal program in the agency and to make superfluous or to undermine traditional APA review. Cf. Astra USA, Inc. v. Santa Clara County, 563 U. S. ___, ___ (2011) (slip op., at 2) (noting that the treatment of lawsuits that are “in substance one and the same” “must be the same, ‘[n]o matter the clothing in which [plaintiffs] dress their claims’ ” (quoting Tenet v. Doe, 544 U.S. 1, 8 (2005)). If the two kinds of actions should reach the same result, the Supremacy Clause challenge is at best redundant. And to permit the continuation of the action in that form would seem to be inefficient, for the agency is not a participant in the pending litigation below, litigation that will decide whether the agency-approved state rates violate the federal statute. III In the present posture of these cases, we do not address whether the Ninth Circuit properly recognized a Supremacy Clause action to enforce this federal statute before the agency took final action. To decide whether these cases may proceed directly under the Supremacy Clause now that the agency has acted, it will be necessary to take account, in light of the proceedings that have already taken place, of at least the matters we have set forth above. It must be recognized, furthermore, that the parties have not fully argued this question. Thus, it may be that not all of the considerations that may bear upon the proper resolution of the issue have been presented in the briefs to this Court or in the arguments addressed to and considered by the Court of Appeals. Given the complexity of these cases, rather than ordering reargument, we vacate the Ninth Circuit’s judgments and remand the cases, thereby permitting the parties to argue the matter before that Circuit in the first instance. It is so ordered.

567.US.2011_11-45

The Civil Service Reform Act of 1978 (CSRA) “established a comprehensive system for reviewing personnel action taken against federal employees,” United States v. Fausto, 484 U.S. 439, 455, including removals, 5 U. S. C. §7512. A qualifying employee has the right to a hearing before the Merit Systems Protection Board (MSPB), §§7513(d), 7701(a)(1)–(2), which is authorized to order reinstatement, backpay, and attorney’s fees, §§1204(a)(2), 7701(g). An employee who is dissatisfied with the MSPB’s decision is entitled to judicial review in the Federal Circuit. §§7703(a)(1), (b)(1). Petitioners were federal employees discharged pursuant to 5 U. S. C. §3328, which bars from Executive agency employment anyone who has knowingly and willfully failed to register for the Selective Service as required by the Military Selective Service Act, 50 U. S. C. App. §453. Petitioner Elgin challenged his removal before the MSPB, claiming that §3328 is an unconstitutional bill of attainder and unconstitutionally discriminates based on sex when combined with the Military Selective Service Act’s male-only registration requirement. The MSPB referred the case to an Administrative Law Judge (ALJ), who dismissed the appeal for lack of jurisdiction, concluding that an employee is not entitled to MSPB review of agency action that is based on an absolute statutory bar to employment. The ALJ also concluded that the MSPB lacked authority to determine the constitutionality of a federal statute. Rather than seeking further MSPB review or appealing to the Federal Circuit, Elgin joined other petitioners raising the same constitutional challenges to their removals in a suit in Federal District Court. The District Court found that it had jurisdiction and denied petitioners’ constitutional claims on the merits. The First Circuit vacated and remanded with instructions to dismiss for lack of jurisdiction. The First Circuit held that petitioners were employees entitled to MSPB review despite the statutory bar to their employment. The court further concluded that challenges to a removal are not exempt from the CSRA review scheme simply because an employee challenges the constitutionality of the statute authorizing the removal. Held: The CSRA precludes district court jurisdiction over petitioners’ claims because it is fairly discernible that Congress intended the statute’s review scheme to provide the exclusive avenue to judicial review for covered employees who challenge covered adverse employment actions, even when those employees argue that a federal statute is unconstitutional. Pp. 5–20. (a) Relying on Webster v. Doe, 486 U.S. 592, 603, petitioners claim that 28 U. S. C. §1331’s general grant of federal-question jurisdiction to district courts remains undisturbed unless Congress explicitly directs otherwise. But Webster’s “heightened showing” applies only when a statute purports to “deny any judicial forum for a colorable constitutional claim,” 486 U. S., at 603, not when Congress channels judicial review of a constitutional claim to a particular court, see Thunder Basin Coal Co. v. Reich, 510 U.S. 200. Here, where the claims can be “meaningfully addressed in the” Federal Circuit, id., at 215, the proper inquiry is whether Congress’ intent to preclude district court jurisdiction was “ ‘fairly discernible in the statutory scheme,’ ” id., at 207. Pp. 5–6. (b) It is “fairly discernible” from the CSRA’s text, structure, and purpose that Congress precluded district court jurisdiction over petitioners’ claims. Pp. 6–12. (1) Just as the CSRA’s “elaborate” framework demonstrated Congress’ intent to entirely foreclose judicial review to employees to whom the CSRA denies statutory review in Fausto, 484 U. S., at 443, the CSRA indicates that extrastatutory review is not available to those employees to whom the CSRA grants administrative and judicial review. It “prescribes in great detail the protections and remedies applicable to” adverse personnel actions against federal employees, ibid., specifically enumerating the major adverse actions and employee classifications to which the CSRA’s procedural protections and review provisions apply, §§7511, 7512, setting out the procedures due an employee prior to final agency action, §7513, and exhaustively detailing the system of review before the MSPB and the Federal Circuit, §§7701, 7703. Petitioners and the Government do not dispute that petitioners are removed employees to whom CSRA review is provided, but petitioners claim that there is an exception to the CSRA review scheme for employees who bring constitutional challenges to federal statutes; this claim finds no support in the CSRA’s text and structure. The availability of administrative and judicial review under the CSRA generally turns on the type of civil service employee and adverse employment action at issue. Nothing in the CSRA’s text suggests that its exclusive review scheme is inapplicable simply because a covered employee raises a constitutional challenge. And §7703(b)(2)—which expressly exempts from Federal Circuit review challenges alleging that a covered action was based on discrimination prohibited by enumerated federal employment laws—demonstrates that Congress knew how to provide alternative forums for judicial review based on the nature of an employee’s claim. Pp. 6–10. (2) The CSRA’s purpose also supports the conclusion that the statutory review scheme is exclusive, even for constitutional challenges. The CSRA’s objective of creating an integrated review scheme to replace inconsistent decisionmaking and duplicative judicial review would be seriously undermined if a covered employee could challenge a covered employment action first in a district court, and then again in a court of appeals, simply by challenging the constitutionality of the statutory authorization for the action. Claim-splitting and preclusion doctrines would not necessarily eliminate the possibility of parallel proceedings before the MSPB and the district court, and petitioners point to nothing in the CSRA to support the notion that Congress intended to allow employees to pursue constitutional claims in district court at the expense of forgoing other, potentially meritorious claims before the MSPB. Pp. 10–12. (c) Petitioners invoke the “presum[ption] that Congress does not intend to limit [district court] jurisdiction if ‘a finding of preclusion could foreclose all meaningful judicial review’; if the suit is ‘wholly collateral to a statute’s review provisions’; and if the claims are ‘outside the agency’s expertise.’ ” Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U. S. ___, ___. But none of those characteristics is present here. Pp. 12–20. (1) Petitioners’ constitutional claims can receive meaningful review within the CSRA scheme even if the MSPB, as it claims, is not authorized to decide a federal law’s constitutionality. Their claims can be “meaningfully addressed” in the Federal Circuit, which has held that it can determine the constitutionality of a statute upon which an employee’s removal was based, notwithstanding the MSPB’s professed lack of authority to decide the question. The CSRA review scheme also fully accommodates the potential need for a factual record. Even without factfinding capabilities, the Federal Circuit may take judicial notice of facts relevant to the constitutional question. If further development is necessary, the CSRA empowers the MSPB to take evidence and find facts for Federal Circuit review. See 5 U. S. C. §§1204(b)(1)–(2). Petitioners err in arguing that the MSPB will invariably dismiss a case without ever reaching the factfinding stage in an appeal such as theirs. The MSPB may determine that it lacks authority to decide the issue; but absent another infirmity in the adverse action, it will affirm the employing agency’s decision. The Federal Circuit can then review the decision, including any factual record developed by the MSPB. Petitioners’ argument is not illustrated by Elgin’s case, which was dismissed on the threshold ground that he was not an “employee” with a right to appeal because his employment was absolutely barred by statute. Pp. 12–18. (2) Petitioners’ claims are also not “wholly collateral” to the CSRA scheme. Their constitutional claims are the vehicle by which they seek to reverse the removal decisions, to return to federal employment, and to receive lost compensation. A challenge to removal is precisely the type of personnel action regularly adjudicated by the MSPB and the Federal Circuit within the CSRA scheme, and reinstatement, backpay, and attorney’s fees are precisely the kinds of relief that the CSRA empowers the MSPB and the Federal Circuit to provide. Pp. 18–19. (3) Finally, in arguing that their constitutional claims are not the sort that Congress intended to channel through the MSPB because they are beyond the MSPB’s expertise, petitioners overlook the many threshold questions that may accompany a constitutional claim and to which the MSPB can apply its expertise, e.g., whether a resignation, as in petitioner Tucker’s case, amounts to a constructive discharge. Pp. 19–20. 641 F.3d 6, affirmed. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Breyer, and Sotomayor, JJ., joined. Alito, J., filed a dissenting opinion, in which Ginsburg and Kagan, JJ., joined.

Under the Civil Service Reform Act of 1978 (CSRA), 5 U. S. C. §1101 et seq., certain federal employees may obtain administrative and judicial review of specified ad- verse employment actions. The question before us is whether the CSRA provides the exclusive avenue to judicial review when a qualifying employee challenges an adverse employment action by arguing that a federal sta- tute is unconstitutional. We hold that it does. I The CSRA “established a comprehensive system for reviewing personnel action taken against federal employees.” United States v. Fausto, 484 U.S. 439, 455 (1988). As relevant here, Subchapter II of Chapter 75 governs review of major adverse actions taken against employees “for such cause as will promote the efficiency of the service.” 5 U. S. C. §§7503(a), 7513(a). Employees entitled to review are those in the “competitive service” and “excepted service” who meet certain requirements regarding probationary periods and years of service.[1] §7511(a)(1). The re- viewable agency actions are removal, suspension for more than 14 days, reduction in grade or pay, or furlough for 30 days or less. §7512. When an employing agency proposes a covered action against a covered employee, the CSRA gives the employee the right to notice, representation by counsel, an opportunity to respond, and a written, reasoned decision from the agency. §7513(b). If the agency takes final adverse action against the employee, the CSRA gives the employee the right to a hearing and to be represented by an attorney or other representative before the Merit Systems Pro- tection Board (MSPB). §§7513(d), 7701(a)(1)–(2). The MSPB is authorized to order relief to prevailing employees, including reinstatement, backpay, and attorney’s fees. §§1204(a)(2), 7701(g). An employee who is dissatisfied with the MSPB’s decision is entitled to judicial review in the United States Court of Appeals for the Federal Circuit. That court “shall review the record and hold unlawful and set aside any agency action, findings, or conclusions” that are “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law,” “obtained without procedures required by law, rule, or regulation having been followed,” or “unsupported by substantial evidence.” §§7703(a)(1), (c). The Federal Circuit has “exclusive jurisdiction” over appeals from a final decision of the MSPB. 28 U. S. C. §1295(a)(9); see also 5 U. S. C. §7703(b)(1) (judicial review of an MSPB decision “shall be” in the Federal Circuit). II Petitioners are former federal competitive service employees who failed to comply with the Military Selective Service Act, 50 U. S. C. App. §453. That Act requires male citizens and permanent-resident aliens of the United States between the ages of 18 and 26 to register for the Selective Service. Another federal statute, 5 U. S. C. §3328 (hereinafter Section 3328), bars from employment by an Executive agency anyone who has knowingly and willfully failed to register. Pursuant to Section 3328, petitioners were discharged (or allegedly constructively discharged) by respondents, their employing agencies. Among petitioners, only Michael Elgin appealed his removal to the MSPB. Elgin argued that Section 3328 is an unconstitutional bill of attainder and unconstitution-ally discriminates on the basis of sex when combined with the registration requirement of the Military Selective Service Act. The MSPB referred Elgin’s appeal to an ad- ministrative law judge (ALJ) for an initial decision.[2] The ALJ dismissed the appeal for lack of jurisdiction, concluding that an employee is not entitled to MSPB review of agency action that is based on an absolute statutory bar to employment. App. to Pet. for Cert. 100a–101a. The ALJ also held that Elgin’s constitutional claims could not “confer jurisdiction” on the MSPB because it “lacks authority to determine the constitutionality of a statute.” Id., at 101a. Elgin neither petitioned for review by the full MSPB nor appealed to the Federal Circuit. Instead, he joined the other petitioners in filing suit in the United States District Court for the District of Massachusetts, raising the same constitutional challenges to Section 3328 and the Military Selective Service Act. App. 4, 26–28, 29. Petitioners sought equitable relief in the form of a declaratory judgment that the challenged statutes are unconstitutional, an injunction prohibiting enforcement of Section 3328, reinstatement to their former positions, backpay, benefits, and attorney’s fees. Id., at 29–30. The District Court rejected respondents’ argument that it lacked jurisdiction and denied petitioners’ constitutional claims on the merits. See Elgin v. United States, 697 F. Supp. 2d 187 (Mass. 2010). The District Court held that the CSRA did not preclude it from hearing petitioners’ claims, because the MSPB had no authority to determine the constitutionality of a federal statute. Id., at 193. Hence, the District Court concluded that it retained jurisdiction under the general grant of federal-question jurisdiction in 28 U. S. C. §1331. 697 F. Supp. 2d, at 194. The United States Court of Appeals for the First Circuit vacated the judgment and remanded with instructions to dismiss for lack of jurisdiction. See 641 F.3d 6 (2011). The Court of Appeals held that challenges to a removal are not exempted from the CSRA review scheme simply because the employee argues that the statute authorizing the removal is unconstitutional. Id., at 11–12. According to the Court of Appeals, the CSRA provides a forum—the Federal Circuit—that may adjudicate the constitutionality of a federal statute, and petitioners “were obliged to use it.” Id., at 12–13. We granted certiorari to decide whether the CSRA pre- cludes district court jurisdiction over petitioners’ claims even though they are constitutional claims for equitable relief. See 565 U. S. ___ (2011). We conclude that it does, and we therefore affirm. III We begin with the appropriate standard for determining whether a statutory scheme of administrative and judicial review provides the exclusive means of review for constitutional claims. Petitioners argue that even if they may obtain judicial review of their constitutional claims before the Federal Circuit, they are not precluded from pursuing their claims in federal district court. According to petitioners, the general grant of federal-question jurisdiction in 28 U. S. C. §1331, which gives district courts authority over constitutional claims, remains undisturbed unless Congress explicitly directs otherwise. In support of this argument, petitioners rely on Webster v. Doe, 486 U.S. 592, 603 (1988), which held that “where Congress intends to preclude judicial review of constitutional claims[,] its intent to do so must be clear.” The Webster Court noted that this “heightened showing” was required “to avoid the ‘serious constitutional question’ that would arise if a federal statute were construed to deny any judicial forum for a colorable constitutional claim.” Ibid. (quoting Bowen v. Michigan Academy of Family Physicians, 476 U.S. 667, 681, n. 12 (1986)). Petitioners contend that the CSRA does not meet this standard because it does not expressly bar suits in district court. Petitioners’ argument overlooks a necessary predicate to the application of Webster’s heightened standard: a statute that purports to “deny any judicial forum for a colorable constitutional claim.” 486 U. S., at 603. Webster’s standard does not apply where Congress simply channels judicial review of a constitutional claim to a particular court. We held as much in Thunder Basin Coal Co. v. Reich, 510 U.S. 200 (1994). In that case, we considered whether a statutory scheme of administrative review followed by judicial review in a federal appellate court precluded district court jurisdiction over a plaintiff’s statutory and constitutional claims. Id., at 206. We noted that the plaintiff’s claims could be “meaningfully addressed in the Court of Appeals” and that the case therefore did “not present the ‘serious constitutional question’ that would arise if an agency statute were construed to preclude all judicial review of a constitutional claim.” Id., at 215, and n. 20 (quoting Bowen, supra, at 681, n. 12). Accordingly, we did not require Webster’s “heightened showing,” but instead asked only whether Congress’ intent to preclude district court jurisdiction was “ ‘fairly discernible in the statutory scheme.’ ” 510 U. S., at 207 (quoting Block v. Community Nutrition Institute, 467 U.S. 340, 351 (1984)). Like the statute in Thunder Basin, the CSRA does not foreclose all judicial review of petitioners’ constitutional claims, but merely directs that judicial review shall occur in the Federal Circuit. Moreover, as we explain below, the Federal Circuit is fully capable of providing meaningful review of petitioners’ claims. See infra, at 12–17. Accordingly, the appropriate inquiry is whether it is “fairly discernible” from the CSRA that Congress intended covered employees appealing covered agency actions to proceed exclusively through the statutory review scheme, even in cases in which the employees raise constitutional challenges to federal statutes. IV To determine whether it is “fairly discernible” that Congress precluded district court jurisdiction over petitioners’ claims, we examine the CSRA’s text, structure, and purpose. See Thunder Basin, supra, at 207; Fausto, 484 U. S., at 443. A This is not the first time we have addressed the impact of the CSRA’s text and structure on the availability of judicial review of a federal employee’s challenge to an employment decision. In Fausto, we considered whether a so-called “nonpreference excepted service employe[e]” could challenge his suspension in the United States Claims Court, even though the CSRA did not then afford him a right to review in the MSPB or the Federal Circuit.[3] Id., at 440–441, 448. Citing “[t]he comprehensive nature of the CSRA, the attention that it gives throughout to the rights of nonpreference excepted service employees, and the fact that it does not include them in provisions for administrative and judicial review contained in Chapter 75,” the Court concluded that “the absence of provision for these employees to obtain judicial review” was a “considered congressional judgment.” Id., at 448. The Court thus found it “fairly discernible” that Congress intended to preclude all judicial review of Fausto’s statutory claims.[4] Id., at 452 (citing Block, supra, at 349). Just as the CSRA’s “elaborate” framework, 484 U. S., at 443, demonstrates Congress’ intent to entirely foreclose judicial review to employees to whom the CSRA denies statutory review, it similarly indicates that extrastatutory review is not available to those employees to whom the CSRA grants administrative and judicial review. Indeed, in Fausto we expressly assumed that “competitive service employees, who are given review rights by Chapter 75, cannot expand these rights by resort to” judicial review outside of the CSRA scheme. See id., at 450, n. 3. As Fausto explained, the CSRA “prescribes in great detail the protections and remedies applicable to” adverse personnel actions against federal employees. Id., at 443. For example, Subchapter II of Chapter 75, the portion of the CSRA relevant to petitioners, specifically enumerates the major adverse actions and employee classifications to which the CSRA’s procedural protections and review provisions apply. 5 U. S. C. §§7511, 7512. The subchapter then sets out the procedures due an employee prior to final agency action. §7513. And, Chapter 77 of the CSRA exhaustively details the system of review before the MSPB and the Federal Circuit. §§7701, 7703; see also Fausto, supra, at 449 (emphasizing that the CSRA’s structure evinces “the primacy” of review by the MSPB and the Federal Circuit). Given the painstaking detail with which the CSRA sets out the method for covered employees to obtain review of adverse employment actions, it is fairly discernible that Congress intended to deny such employees an additional avenue of review in district court. Petitioners do not dispute that they are employees who suffered adverse actions covered by the foregoing provisions of the CSRA. Nor do they contest that the CSRA’s text and structure support implied preclusion of district court jurisdiction, at least as a general matter. Petitioners even acknowledge that the MSPB routinely adjudicates some constitutional claims, such as claims that an agency took adverse employment action in violation of an em- ployee’s First or Fourth Amendment rights, and that these claims must be brought within the CSRA scheme. See Brief for Petitioners 33; Tr. of Oral Arg. 7–11, 15, 21; see also, e.g., Smith v. Department of Transp., 106 MSPR 59, 78–79 (2007) (applying Pickering v. Board of Ed. of Township High School Dist. 205, Will Cty., 391 U.S. 563 (1968), to an employee’s claim that he was suspended in retal- iation for the exercise of his First Amendment rights); Garrison v. Department of Justice, 67 MSPR 154 (1995) (considering whether an order directing an employee to submit to a drug test was reasonable under the Fourth Amendment). Nevertheless, petitioners seek to carve out an exception to CSRA exclusivity for facial or as-applied constitutional challenges to federal statutes. The text and structure of the CSRA, however, provide no support for such an exception. The availability of ad- ministrative and judicial review under the CSRA gen- erally turns on the type of civil service employee and adverse employment action at issue. See, e.g., 5 U. S. C. §§7511(a)(1) (defining “employee”), 7512 (defining “[a]c- tions covered”), 7513(d) (providing that “[a]n employee against whom an action is taken under this section is entitled to appeal to the Merit Systems Protection Board”), 7703(a)(1) (providing that “[a]ny employee . . . adversely affected or aggrieved by a final order or decision of the Merit Systems Protection Board may obtain judicial review of the order or decision” in the Federal Circuit). Nothing in the CSRA’s text suggests that its exclusive review scheme is inapplicable simply because a covered employee challenges a covered action on the ground that the statute authorizing that action is unconstitutional. As the Government correctly notes, “[t]he plain language of [the CSRA’s] provisions applies to an employee who challenges his removal on the ground that the statute requiring it is unconstitutional no less than it applies to an employee who challenges his removal on any other ground.” Brief for Respondents 33–34. In only one situation does the CSRA expressly exempt a covered employee’s appeal of a covered action from Federal Circuit review based on the type of claim at issue. When a covered employee “alleges that a basis for the action was discrimination” prohibited by enumerated federal employment laws, 5 U. S. C. §7702(a)(1)(B), the CSRA allows the employee to obtain judicial review of an unfavorable MSPB decision by filing a civil action as provided by the applicable employment law. See §7703(b)(2). Each of the cross-referenced employment laws authorizes an action in federal district court. See 42 U. S. C. §2000e–5(f); 29 U. S. C. §633a(c); §216(b). Title 5 U. S. C. §7703(b)(2) demonstrates that Congress knew how to provide alternative forums for judicial review based on the nature of an employee’s claim. That Congress declined to include an exemption from Federal Circuit review for challenges to a statute’s constitutionality indicates that Congress intended no such exception. B The purpose of the CSRA also supports our conclusion that the statutory review scheme is exclusive, even for employees who bring constitutional challenges to federal statutes. As we have previously explained, the CSRA’s “integrated scheme of administrative and judicial review” for aggrieved federal employees was designed to replace an “ ‘outdated patchwork of statutes and rules’ ” that afforded employees the right to challenge employing agency actions in district courts across the country. Fausto, 484 U. S., at 444–445. Such widespread judicial review, which included appeals in all of the Federal Courts of Appeals produced “wide variations in the kinds of decisions . . . issued on the same or similar matters” and a double layer of judicial review that was “wasteful and irrational.” Id., at 445 (internal quotation marks omitted). The CSRA’s objective of creating an integrated scheme of review would be seriously undermined if, as petitioners would have it, a covered employee could challenge a covered employment action first in a district court, and then again in one of the courts of appeals, simply by alleging that the statutory authorization for such action is unconstitutional. Such suits would reintroduce the very po- tential for inconsistent decisionmaking and duplicative judicial review that the CSRA was designed to avoid. Moreover, petitioners’ position would create the possibility of parallel litigation regarding the same agency action before the MSPB and a district court. An employee could challenge the constitutionality of the statute authorizing an agency’s action in district court, but the MSPB would remain the exclusive forum for other types of challenges to the agency’s decision. See Tr. of Oral Arg. 4–7, 9, 15–16. Petitioners counter that doctrines regarding claim splitting and preclusion would bar parallel suits before the MSPB and the district court. But such doctrines would not invariably eliminate the possibility of simultaneous proceedings, for a tribunal generally has discretion to decide whether to dismiss a suit when a similar suit is pending elsewhere. See 18 C. Wright et al., Federal Practice and Procedure §4406 (2d ed. 2002 and Supp. 2011). In any event, petitioners point to nothing in the CSRA to support the odd notion that Congress intended to allow employees to pursue constitutional claims in district court at the cost of forgoing other, potentially meritorious claims before the MSPB. Finally, we note that a jurisdictional rule based on the nature of an employee’s constitutional claim would deprive the aggrieved employee, the MSPB, and the district court of clear guidance about the proper forum for the employee’s claims at the outset of the case. For example, petitioners contend that facial and as-applied constitutional challenges to statutes may be brought in district court, while other constitutional challenges must be heard by the MSPB. See supra, at 8–9; infra, at 13, n. 5. But, as we explain below, that line is hazy at best and incoherent at worst. See ibid. The dissent’s approach fares no better. The dissent carves out for district court adjudication only facial constitutional challenges to statutes, but we have previously stated that “the distinction between facial and as-applied challenges is not so well defined that it has some automatic effect or that it must always control the pleadings and disposition in every case involving a constitutional challenge.” Citizens United v. Federal Election Comm’n, 558 U. S. ___, ___ (2010) (slip op., at 14). By contrast, a jurisdictional rule based on the type of em- ployee and adverse agency action at issue does not involve such amorphous distinctions. Accordingly, we conclude that the better interpretation of the CSRA is that its exclusivity does not turn on the constitutional nature of an employee’s claim, but rather on the type of the employee and the challenged employment action. V Petitioners raise three additional factors in arguing that their claims are not the type that Congress intended to be reviewed within the CSRA scheme. Specifically, petitioners invoke our “presum[ption] that Congress does not intend to limit [district court] jurisdiction if ‘a finding of preclusion could foreclose all meaningful judicial review’; if the suit is ‘wholly collateral to a statute’s review provisions’; and if the claims are ‘outside the agency’s expertise.’ ” Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U. S. ___, ___ (2010) (slip op., at 8) (quoting Thunder Basin, 510 U. S., at 212–213). Contrary to petitioners’ suggestion, none of those characteristics are present here. A First, petitioners argue that the CSRA review scheme provides no meaningful review of their claims because the MSPB lacks authority to declare a federal statute un- constitutional. Petitioners are correct that the MSPB has repeatedly refused to pass upon the constitutionality of legislation. See, e.g., Malone v. Department of Justice, 13 M. S. P. B. 81, 83 (1983) (“[I]t is well settled that administrative agencies are without authority to determine the constitutionality of statutes”). This Court has also stated that “adjudication of the constitutionality of congressional enactments has generally been thought beyond the jurisdiction of administrative agencies.” Thunder Basin, 510 U. S., at 215 (internal quotation marks and brackets omitted).[5] We need not, and do not, decide whether the MSPB’s view of its power is correct, or whether the oft-stated principle that agencies cannot declare a statute unconstitutional is truly a matter of jurisdiction. See ibid. (describing this rule as “not mandatory”). In Thunder Basin, we held that Congress’ intent to preclude district court jurisdiction was fairly discernible in the statutory scheme “[e]ven if” the administrative body could not decide the constitutionality of a federal law. Ibid. That issue, we reasoned, could be “meaningfully addressed in the Court of Appeals” that Congress had authorized to conduct judicial review. Ibid.[6] Likewise, the CSRA provides review in the Federal Circuit, an Article III court fully competent to adjudicate petitioners’ claims that Section 3328 and the Military Selective Service Act’s registration requirement are unconstitutional. Petitioners insist, however, that the Federal Circuit cannot decide their constitutional claims either. Emphasizing the Federal Circuit’s holdings that its jurisdiction over employee appeals is coextensive with the MSPB’s jurisdiction, petitioners argue that the Federal Circuit likewise lacks jurisdiction to decide their challenge to the constitutionality of a federal statute. Petitioners are incorrect. As we have explained, the CSRA makes MSPB jurisdiction over an appeal dependent only on the nature of the employee and the employment action at issue. See supra, at 1–2, 8–9; see also 5 CFR §1201.3(a) (stating that “[t]he Board has jurisdiction over appeals from agency actions” and enumerating covered actions); Todd v. Merit Systems Protection Bd., 55 F.3d 1574, 1576 (CA Fed. 1995) (explaining that the employee “has the burden of establishing that she and the action she seeks to appeal [are] within the [MSPB’s] jurisdiction”). Accordingly, as the cases cited by petitioners demonstrate, the Federal Circuit has questioned its jurisdiction when an employee appeals from a type of adverse action over which the MSPB lacked jurisdiction.[7] But the Federal Circuit has never held, in an appeal from agency action within the MSPB’s jurisdiction, that its authority to decide particular legal questions is derivative of the MSPB’s authority. To the contrary, in Briggs v. MSPB, 331 F.3d 1307, 1312–1313 (2003), the Federal Circuit concluded that it could determine the constitutionality of a statute upon which an employee’s removal was based, notwithstanding the MSPB’s professed lack of authority to decide the question.[8] Petitioners next contend that even if the Federal Circuit could consider their claims in the first instance, resolution of the claims requires a factual record that neither the MSPB (because it lacks authority to decide the legal question) nor the Federal Circuit (because it is an appellate court) can create. To the contrary, we think the CSRA review scheme fully accommodates an employee’s potential need to establish facts relevant to his constitutional challenge to a federal statute. Even without factfinding capabilities, the Federal Circuit may take judicial notice of facts relevant to the constitutional question. See, e.g., Rothe Development Corp. v. Department of Defense, 545 F.3d 1023, 1045–1046 (CA Fed. 2008) (judicially noticing facts relevant to equal protection challenge). And, if resolution of a constitutional claim requires the development of facts beyond those that the Federal Circuit may judicially notice, the CSRA empowers the MSPB to take evidence and find facts for Federal Circuit review. See 5 U. S. C. §§1204(b)(1)–(2) (providing that the MSPB may administer oaths, examine witnesses, take depositions, issue interrogatories, subpoena testimony and documents, and otherwise receive evidence when a covered employee appeals a covered adverse employment action). Unlike petitioners, we see nothing extraordinary in a statutory scheme that vests reviewable factfinding authority in a non-Article III entity that has jurisdiction over an action but cannot finally decide the legal question to which the facts pertain. Congress has authorized magistrate judges, for example, to conduct evidentiary hearings and make findings of fact relevant to dispositive pretrial motions, although they are powerless to issue a final ruling on such motions. See 28 U. S. C. §§636(b)(1)(A)–(B); United States v. Raddatz, 447 U.S. 667, 673 (1980).[9] Petitioners nonetheless insist that the MSPB will never reach the factfinding stage in an appeal challenging the constitutionality of a federal statute, pointing to the ALJ’s dismissal for lack of jurisdiction in petitioner Elgin’s case. Again, petitioners are incorrect. When a covered employee appeals a covered adverse action, the CSRA grants the MSPB jurisdiction over the appeal. See supra, at 14. If the employee attacks the adverse action on the ground that a statute is unconstitutional, the MSPB may determine that it lacks authority to decide that particular issue; but absent another infirmity in the adverse action, the MSPB will affirm the employing agency’s decision rather than dismiss the appeal. See, e.g., Briggs, supra, at 1311. The Federal Circuit can then review the MSPB decision, including any factual record developed by the MSPB in the course of its decision on the merits. Contrary to petitioners’ suggestion, Elgin’s case does not illustrate that the MSPB will invariably dismiss an appeal challenging the constitutionality of a federal statute before reaching the factfinding stage. The ALJ dismissed Elgin’s case on the threshold jurisdictional ground that he was not an “employee” with a right to appeal to the MSPB because his employment was absolutely barred by statute. See App. to Pet. for Cert. 100a–101a. The Government conceded before the First Circuit that this jurisdictional argument was incorrect, see Brief for United States 10, and the Court of Appeals agreed, see 641 F. 3d, at 10–11. The parties do not raise that issue here, and we do not address it. What matters for present purposes is that the particular circumstances of Elgin’s case do not demonstrate that the MSPB will dismiss an appeal that is otherwise within its jurisdiction merely because it lacks the authority to decide a particular claim.[10] In sum, the CSRA grants the MSPB and the Federal Circuit jurisdiction over petitioners’ appeal because they are covered employees challenging a covered adverse employment action. Within the CSRA review scheme, the Federal Circuit has authority to consider and decide petitioners’ constitutional claims. To the extent such challenges require factual development, the CSRA equips the MSPB with tools to create the necessary record. Thus, petitioners’ constitutional claims can receive meaningful review within the CSRA scheme.[11] B Petitioners next contend that the CSRA does not preclude district court jurisdiction over their claims because they are “wholly collateral” to the CSRA scheme. According to petitioners, their bill-of-attainder and sex discrimination claims “have nothing to do with the types of day-to-day personnel actions adjudicated by the MSPB,” Brief for Petitioners 29, and petitioners “are not seeking the CSRA’s ‘protections and remedies.’ ” Reply Brief for Petitioners 3. We disagree. As evidenced by their district court complaint, petitioners’ constitutional claims are the vehicle by which they seek to reverse the removal decisions, to return to federal employment, and to receive the compensation they would have earned but for the adverse employment action. See App. 29–30. A challenge to removal is precisely the type of personnel action regularly adjudicated by the MSPB and the Federal Circuit within the CSRA scheme. Likewise, reinstatement, backpay, and attorney’s fees are precisely the kinds of relief that the CSRA empowers the MSPB and the Federal Circuit to provide. See supra, at 2; see also Heckler v. Ringer, 466 U.S. 602, 614 (1984) (holding that plaintiffs’ claims were not wholly collateral to a statutory scheme of administrative and judicial review of Medicare payment decisions, where plaintiffs’ constitutional and statutory challenge to an agency’s procedure for reaching payment decisions was “at bottom” an attempt to reverse the agency’s decision to deny payment). Far from a suit wholly collateral to the CSRA scheme, the case before us is a challenge to CSRA-covered employment action brought by CSRA-covered employees requesting relief that the CSRA routinely affords. C Relatedly, petitioners argue that their constitutional claims are not the sort that Congress intended to channel through the MSPB because they are outside the MSPB’s expertise. But petitioners overlook the many threshold questions that may accompany a constitutional claim and to which the MSPB can apply its expertise. Of particular relevance here, preliminary questions unique to the employment context may obviate the need to address the constitutional challenge. For example, petitioner Henry Tucker asserts that his resignation amounted to a constructive discharge. That issue falls squarely within the MSPB’s expertise, and its resolution against Tucker would avoid the need to reach his constitutional claims. In addition, the challenged statute may be one that the MSPB regularly construes, and its statutory interpretation could alleviate constitutional concerns. Or, an employee’s appeal may involve other statutory or constitutional claims that the MSPB routinely considers, in addition to a constitutional challenge to a federal statute. The MSPB’s resolution of those claims in the employee’s favor might fully dispose of the case. Thus, because the MSPB’s expertise can otherwise be “brought to bear” on employee appeals that challenge the constitutionality of a statute, we see no reason to conclude that Congress intended to exempt such claims from exclusive review before the MSPB and the Federal Circuit. See Thunder Basin, 510 U. S., at 214–215 (concluding that, where administrative Commission’s expertise “could be brought to bear” on appeal, Commission’s exclusive review of alleged statutory violation was appropriate despite its lack of expertise in interpreting a particular statute (internal quotation marks and brackets omitted)). * * * For the foregoing reasons, we conclude that it is fairly discernible that the CSRA review scheme was intended to preclude district court jurisdiction over petitioners’ claims. The judgment of the Court of Appeals is affirmed. It is so ordered. Notes 1 The CSRA divides civil service employees into three main categories. Fausto, 484 U. S., at 441, n. 1. “Senior Executive Service” employees occupy high-level positions in the Executive Branch but are not required to be appointed by the President and confirmed by the Senate. 5 U. S. C. §3131(2). “[C]ompetitive service” employees—the relevant category for purposes of this case—are all other Executive Branch employees whose nomination by the President and confirmation by the Senate are not required and who are not specifically excepted from the competitive service by statute. §2102(a)(1). The competitive service also includes employees in other branches of the Federal Govern-ment and in the District of Columbia government who are specifically included by statute. §§2102(a)(2)–(3). Finally, “excepted service” employees are employees who are not in the Senior Executive Service or in the competitive service. §2103. 2 See §7701(b)(1) (authorizing referral of MSPB appeals to an ALJ); 5 CFR §§1201.111–1201.114 (2011) (detailing procedures for an initial decision by an ALJ and review by the MSPB). 3 Certain veterans and their close relatives are considered “preference eligible” civil service employees. Fausto, 484 U. S., at 441, n. 1. 4 Although Fausto interpreted the CSRA to entirely foreclose judicial review, the Court had no need to apply a heightened standard like that applied in Webster v. Doe, 486 U.S. 592 (1988), because Fausto did not press any constitutional claims. 5 According to petitioners, the MSPB can decide claims that anagency violated an employee’s First or Fourth Amendment rights (and those claims consequently must be brought within the CSRA scheme), supra, at 8–9, because such claims allege only that an agency “acted in an unconstitutional manner” and do not challenge the constitutionality of a federal statute either facially or as applied. See Tr. of Oral Arg. 10, 21. That distinction is dubious at best. Agencies are created by and act pursuant to statutes. Thus, unless an action is beyond the scope ofthe agency’s statutory authority, an employee’s claim that the agency “acted in an unconstitutional manner” will generally be a claim that the statute authorizing the agency action was unconstitutionally applied to him. See, e.g., Pickering v. Board of Ed. of Township High SchoolDist. 205, Will Cty., 391 U.S. 563, 565 (1968) (holding that the statute authorizing a government employee’s termination was unconstitutional as applied under the First and Fourteenth Amendments where the employee was fired because of his speech). In any event, the curious line that petitioners draw only highlights the weakness of their position, for it certainly is not “fairly discernible” from the CSRA’s text, structure, or purpose that the statutory review scheme is exclusive for so-called “unconstitutional manner” claims but not for facial or as-applied constitutional challenges to statutes. See supra, at 7–11. 6 The dissent misreads Thunder Basin. The dissent contends that the “heart of the preclusion analysis” in Thunder Basin involved statutory claims reviewable by the administrative body and that the “only constitutional issue” was decided by this Court “ ‘not on preclusion grounds but on the merits.’ ” Post, at 10 (opinion of Alito, J.) (quoting 510 U. S., at 219 (Scalia, J., concurring in part and concurring in judgment)). To be sure, the Thunder Basin Court did decide the merits of the petitioner’s “second constitutional challenge,” namely whether the Court’s finding of preclusion was itself unconstitutional. See i.d., at 219–221, and n. (same); see also id., at 216 (describing this “alternative” argument). But the petitioner’s suit also included another constitutional claim: a due process challenge to a statute that permitted a regulatory agency, before a hearing, to immediately fine the petitioner for noncompliance with the statute. See Brief for Petitioner in Thunder Basin Coal Co. v. Reich, O. T. 1993, No. 92–896, p. 13. The Court expressly found that the statutory review scheme precluded district court jurisdiction over that constitutional claim. See 510 U. S., at 214‒216. 7 See Schmittling v. Department of Army, 219 F.3d 1332, 1336 (CA Fed. 2000) (remanding for MSPB to determine if employee suffered a prohibited personnel action within the scope of its jurisdiction); Perez v. MSPB, 931 F.2d 853, 855 (CA Fed. 1991) (action against employee was not suspension within MSPB’s jurisdiction); Manning v. MSPB, 742 F.2d 1424, 1425–1427 (CA Fed. 1984) (reassignment of employee was not an adverse action within MSPB’s jurisdiction); Rosano v. Department of Navy, 699 F.2d 1315 (CA Fed. 1983) (refusal to prorate employee’s health insurance premiums was not an adverse action within MSPB’s jurisdiction). 8 It is not unusual for an appellate court reviewing the decision ofan administrative agency to consider a constitutional challenge to a federal statute that the agency concluded it lacked authority to decide. See, e.g., Preseault v. ICC, 853 F.2d 145, 148–149 (CA2 1988) (provision of the National Trails System Act Amendments of 1983), aff’d on other grounds, 494 U.S. 1 (1990); Reid v. Engen, 765 F.2d 1457, 1460–1461 (CA9 1985) (provision of the Federal Aviation Act of 1958); Chadha v. INS, 634 F.2d 408, 411, 413 (CA9 1980) (provision of the Immigration and Nationality Act), aff’d, 462 U.S. 919 (1983). 9 The dissent argues that the MSPB may struggle to determine what facts are relevant to the constitutional question, given that it will not decide the claim. See post, at 11. But the MSPB’s professed lack of authority to declare a statute unconstitutional does not mean that the MSPB cannot identify the legal principles that govern the constitu-tional analysis and thus the scope of necessary development of the fac-tual record. The MSPB routinely identifies the relevant constitutional framework from federal court decisions when deciding other constitutional claims. See supra, at 8–9 (citing First and Fourth Amendment cases); see also, e.g., Fitzgerald v. Department of Defense, 80 MSPR 1, 14–15 (1998) (analyzing a claim under the Due Process Clauses of the Fifth and Fourteenth Amendments). We therefore see little reason to credit the dissent’s prediction that our holding will result in a complicated back and forth between a befuddled MSPB and the Federal Circuit. 10 Before this Court, the Government again conceded the error of its argument that Elgin is not an “employee” within the MSPB’s jurisdiction and indicated that it would support a motion by Elgin to reopen his case before the MSPB. See Tr. of Oral Arg. 32. 11 The dissent cites McNary v. Haitian Refugee Center, Inc., 498 U.S. 479 (1991), for the “basic principle,” post, at 8, that preclusion cannot be inferred when “ ‘the administrative appeals process does not address the kind of . . . constitutional claims’ at issue.” See post, at 7–8 (quoting McNary, 498 U. S., at 493). But that statement from McNary was not a reference to an administrative body’s inability to decide a constitutional claim. Rather, McNary was addressing a statutory review scheme that provided no opportunity for the plaintiffs to develop a factual record relevant to their constitutional claims before the administrative body and then restricted judicial review to the administrative record created in the first instance. Ibid. As we have explained, the CSRA review process is not similarly limited. See supra, at 15.

566.US.2011_10-1024

Respondent Cooper, a licensed pilot, failed to disclose his human immunodeficiency virus (HIV) diagnosis to the Federal Aviation Administration (FAA) at a time when the agency did not issue medical certificates, which are required to operate an aircraft, to persons with HIV. Subsequently, respondent applied to the Social Security Administration (SSA) and received long-term disability benefits on the basis of his HIV status. Thereafter, he renewed his certificate with the FAA on several occasions, each time intentionally withholding information about his condition. The Department of Transportation (DOT), the FAA’s parent agency, launched a joint criminal investigation with the SSA to identify medically unfit individuals who had obtained FAA certifications. The DOT provided the SSA with the names of licensed pilots, and the SSA, in turn, provided the DOT with a spreadsheet containing information on those pilots who had also received disability benefits. Respondent’s name appeared on the spreadsheet, and an investigation led to his admission that he had intentionally withheld information about his HIV status from the FAA. His pilot certificate was revoked, and he was indicted for making false statements to a Government agency. He pleaded guilty and was fined and sentenced to probation. He then filed suit, alleging that the FAA, DOT, and SSA violated the Privacy Act of 1974, which contains a detailed set of requirements for the management of records held by Executive Branch agencies. The Act allows an aggrieved individual to sue for “actual damages,” 5 U. S. C. §552a(g)(4)(A), if the Government intentionally or willfully violates the Act’s requirements in such a way as to adversely affect the individual. Specifically, respondent claimed that the unlawful disclosure to the DOT of his confidential medical information had caused him mental and emotional distress. The District Court concluded that the Government had violated the Act. But, finding the term “actual damages” ambiguous, the court relied on the sovereign immunity canon, which provides that sovereign immunity waivers must be strictly construed in the Government’s favor, to hold that the Act does not authorize the recovery of nonpecuniary damages. Reversing the District Court, the Ninth Circuit concluded that “actual damages” in the Act is not ambiguous and includes damages for mental and emotional distress. Held: The Privacy Act does not unequivocally authorize damages for mental or emotional distress and therefore does not waive the Government’s sovereign immunity from liability for such harms. Pp. 4–19. (a) A waiver of sovereign immunity must be unequivocally expressed in statutory text, see e.g., Lane v. Peña, 518 U.S. 187, 192, and any ambiguities are to be construed in favor of immunity, United States v. Williams, 514 U.S. 527, 531. Ambiguity exists if there is a plausible interpretation of the statute that would not allow money damages against the Government. United States v. Nordic Village, Inc., 503 U.S. 30, 37. Pp. 5–6. (b) The term “actual damages” in the Privacy Act is a legal term of art, and Congress, when it employs a term of art, “ ‘presumably knows and adopts the cluster of ideas that were attached to each borrowed word in the body of learning from which it was taken,’ ” Molzof v. United States, 502 U.S. 301, 307. Even as a legal term, the precise meaning of “actual damages” is far from clear. Although the term is sometimes understood to include nonpecuniary harm, it has also been used or construed more narrowly to cover damages for only pecuniary harm. Because of the term’s chameleon-like quality, it must be considered in the particular context in which it appears. Pp. 6–9. (c) The Privacy Act serves interests similar to those protected by defamation and privacy torts. Its remedial provision, under which plaintiffs can recover a minimum award of $1,000 if they first prove at least some “actual damages,” “parallels” the common-law torts of libel per quod and slander, under which plaintiffs can recover “general damages” if they first prove “special damages.” Doe v. Chao, 540 U.S. 614, 625. “Special damages” are limited to actual pecuniary loss, which must be specially pleaded and proved. “General damages” cover nonpecuniary loss and need not be pleaded or proved. This parallel suggests the possibility that Congress intended the term “actual damages” to mean “special damages,” thus barring Privacy Act victims from any recovery unless they can first show some actual pecuniary harm. That Congress would choose “actual damages” instead of “special damages” is not without precedent, as the terms have occasionally been used interchangeably. Furthermore, any doubt about the plausibility of construing “actual damages” as special damages in the Privacy Act is put to rest by Congress’ deliberate refusal to allow recovery for “general damages.” In common-law defamation and privacy cases, special damages is the only category of compensatory damages other than general damages. Because Congress declined to authorize general damages, it is reasonable to infer that Congress intended the term “actual damages” in the Act to mean special dam-ages for proven pecuniary loss. Pp. 9–14. (d) Although the contrary reading of the Privacy Act accepted by the Ninth Circuit and advanced by respondent is not inconceivable, it is plausible to read the Act as authorizing only damages for economic loss. Because Congress did not speak unequivocally, the Court adopts an interpretation of “actual damages” limited to proven pecuniary harm. To do otherwise would expand the scope of Congress’ sovereign immunity waiver beyond what the statutory text clearly requires. P. 14. (e) Respondent raises several counterarguments: (1) common-law cases often define “actual damages” to mean all compensatory damages; (2) the elimination of “general damages” from the Privacy Act means that there can be no recovery for presumed damages, but plaintiffs can still recover for proven mental and emotional distress; (3) because some courts have construed “actual damages” in similar statutes to include mental and emotional distress, Congress must have intended “actual damages” in the Act to include mental and emotional distress as well; and (4) precluding nonpecuniary damages would lead to absurd results, thereby frustrating the Act’s remedial purpose. None of these arguments overcomes the sovereign immu-nity canon. Pp. 14–19. 622 F.3d 1016, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, and Thomas, JJ., joined. Sotomayor, J., filed a dissenting opinion, in which Ginsburg and Breyer, JJ., joined. Kagan, J., took no part in the consideration or decision of the case.

The Privacy Act of 1974, codified in part at 5 U. S. C. §552a, contains a comprehensive and detailed set of requirements for the management of confidential records held by Executive Branch agencies. If an agency fails to comply with those requirements “in such a way as to have an adverse effect on an individual,” the Act authorizes the individual to bring a civil action against the agency. §552a(g)(1)(D). For violations found to be “intentional or willful,” the United States is liable for “actual damages.” §552a(g)(4)(A). In this case, we must decide whether the term “actual damages,” as used in the Privacy Act, includes damages for mental or emotional distress. We hold that it does not. I The Federal Aviation Administration (FAA) requires pilots to obtain a pilot certificate and medical certificate as a precondition for operating an aircraft. 14 CFR §§61.3(a), (c) (2011). Pilots must periodically renew their medical certificates to ensure compliance with FAA medical standards. See §61.23(d). When applying for renewal, pilots must disclose any illnesses, disabilities, or surgeries they have had, and they must identify any medications they are taking. See 14 CFR pt. 67. Respondent Stanmore Cooper has been a private pilot since 1964. In 1985, he was diagnosed with a human im- munodeficiency virus (HIV) infection and began taking antiretroviral medication. At that time, the FAA did not issue medical certificates to persons with respondent’s condition. Knowing that he would not qualify for renewal of his medical certificate, respondent initially grounded himself and chose not to apply. In 1994, however, he ap- plied for and received a medical certificate, but he did so without disclosing his HIV status or his medication. He renewed his certificate in 1998, 2000, 2002, and 2004, each time intentionally withholding information about his condition. When respondent’s health deteriorated in 1995, he applied for long-term disability benefits under Title II of the Social Security Act, 42 U. S. C. §401 et seq. To substantiate his claim, he disclosed his HIV status to the Social Security Administration (SSA), which awarded him benefits for the year from August 1995 to August 1996. In 2002, the Department of Transportation (DOT), the FAA’s parent agency, launched a joint criminal investigation with the SSA, known as “Operation Safe Pilot,” to identify medically unfit individuals who had obtained FAA certifications to fly. The DOT gave the SSA a list of names and other identifying information of 45,000 licensed pilots in northern California. The SSA then compared the list with its own records of benefit recipients and compiled a spreadsheet, which it gave to the DOT. The spreadsheet revealed that respondent had a current medical certificate but had also received disability benefits. After reviewing respondent’s FAA medical file and his SSA disability file, FAA flight surgeons determined in 2005 that the FAA would not have issued a medical cer- tificate to respondent had it known his true medical condition. When investigators confronted respondent with what had been discovered, he admitted that he had intention- ally withheld from the FAA information about his HIV status and other relevant medical information. Because of these fraudulent omissions, the FAA revoked respondent’s pilot certificate, and he was indicted on three counts of making false statements to a Government agency, in violation of 18 placecountry-regionU. S. C. §1001. Respondent ultimately pleaded guilty to one count of making and delivering a false official writing, in violation of §1018. He was sentenced to two years of probation and fined $1,000.[1] Claiming that the FAA, DOT, and SSA (hereinafter Government) violated the Privacy Act by sharing his records with one another, respondent filed suit in the United States District Court for the Northern District of California. He alleged that the unlawful disclosure to the DOT of his confidential medical information, including his HIV status, had caused him “humiliation, embarrassment, mental anguish, fear of social ostracism, and other severe emotional distress.” App. to Pet. for Cert. 120a. Notably, he did not allege any pecuniary or economic loss. The District Court granted summary judgment against respondent. 816 F. Supp. 2d 778, 781 (2008). The court concluded that the Government had violated the Privacy Act and that there was a triable issue of fact as to whether the violation was intentional or willful.[2] But the court held that respondent could not recover damages because he alleged only mental and emotional harm, not economic loss. Finding that the term “actual damages” is “facially ambiguous,” id., at 791, and relying on the sovereign immunity canon, which provides that waivers of sovereign immunity must be strictly construed in favor of the Government, the court concluded that the Act does not authorize the recovery of damages from the Government for nonpecuniary mental or emotional harm. The United States Court of Appeals for the Ninth Circuit reversed and remanded. 622 F.3d 1016, 1024 (2010). The court acknowledged that the term “actual damages” is a “ ‘chameleon’ ” in that “its meaning changes with the specific statute in which it is found.” Id., at 1029. But the court nevertheless held that, as used in the Privacy Act, the term includes damages for mental and emotional distress. Looking to what it described as “[i]ntrinsic” and “[e]xtrinsic” sources, id., at 1028, 1031, the court concluded that the meaning of “actual damages” in the Privacy Act is not ambiguous and that “a construction that limits recovery to pecuniary loss” is not “plausible,” id., at 1034. The Government petitioned for rehearing or rehearing en banc, but a divided court denied the petition. Id., at 1019. The Government then petitioned for certiorari, and we granted review. 564 U. S. ___ (2011). II Because respondent seeks to recover monetary compensation from the Government for mental and emotional harm, we must decide whether the civil remedies provision of the Privacy Act waives the Government’s sovereign immunity with respect to such a recovery. A We have said on many occasions that a waiver of sovereign immunity must be “unequivocally expressed” in statutory text. See, e.g., Lane v. Peña, 518 U.S. 187, 192 (1996); United States v. Nordic Village, Inc., 503 U.S. 30, 33 (1992); Irwin v. Department of Veterans Affairs, 498 U.S. 89, 95 (1990). Legislative history cannot supply a waiver that is not clearly evident from the language of the statute. Lane, supra, at 192. Any ambiguities in the statutory language are to be construed in favor of immu- nity, United States v. Williams, 514 U.S. 527, 531 (1995), so that the Government’s consent to be sued is never en- larged beyond what a fair reading of the text requires, Ruckelshaus v. Sierra Club, 463 U.S. 680, 685–686 (1983) (citing Eastern Transp. Co. v. United States, 272 U.S. 675, 686 (1927)). Ambiguity exists if there is a plausible interpretation of the statute that would not authorize money damages against the Government. Nordic Village, supra, at 34, 37. The question that confronts us here is not whether Congress has consented to be sued for damages under the Privacy Act. That much is clear from the statute, which expressly authorizes recovery from the Government for “actual damages.” Rather, the question at issue concerns the scope of that waiver. For the same reason that we refuse to enforce a waiver that is not unambiguously expressed in the statute, we also construe any ambiguities in the scope of a waiver in favor of the sovereign. Lane, supra, at 192. Although this canon of interpretation requires an unmistakable statutory expression of congressional intent to waive the Government’s immunity, Congress need not state its intent in any particular way. We have never required that Congress use magic words. To the contrary, we have observed that the sovereign immunity canon “is a tool for interpreting the law” and that it does not “displac[e] the other traditional tools of statutory construction.” Richlin Security Service Co. v. Chertoff, 553 U.S. 571, 589 (2008). What we thus require is that the scope of Congress’ waiver be clearly discernable from the statutory text in light of traditional interpretive tools. If it is not, then we take the interpretation most favorable to the Government. B The civil remedies provision of the Privacy Act provides that, for any “intentional or willful” refusal or failure to comply with the Act, the United States shall be liable for “actual damages sustained by the individual as a result of the refusal or failure, but in no case shall a person entitled to recovery receive less than the sum of $1,000.” 5 U. S. C. §552a(g)(4)(A). Because Congress did not define “actual damages,” respondent urges us to rely on the ordinary meaning of the word “actual” as it is defined in standard general-purpose dictionaries. But as the Court of Appeals explained, “actual damages” is a legal term of art, 622 F. 3d, at 1028, and it is a “cardinal rule of statutory construction” that, when Congress employs a term of art, “ ‘it presumably knows and adopts the cluster of ideas that were attached to each borrowed word in the body of learning from which it was taken,’ ” Molzof v. United States, 502 U.S. 301, 307 (1992) (quoting Morissette v. United States, 342 U.S. 246, 263 (1952)). Even as a legal term, however, the meaning of “actual damages” is far from clear. The latest edition of Black’s Law Dictionary available when Congress enacted the Privacy Act defined “actual damages” as “[r]eal, substantial and just damages, or the amount awarded to a complainant in compensation for his actual and real loss or injury, as opposed on the one hand to ‘nominal’ damages, and on the other to ‘exemplary’ or ‘punitive’ damages.” Black’s Law Dictionary 467 (rev. 4th ed. 1968). But this general (and notably circular) definition is of little value here because, as the Court of Appeals accurately observed, the precise meaning of the term “changes with the specific statute in which it is found.” 622 F. 3d, at 1029. The term is sometimes understood to include nonpecuniary harm. Take, for instance, some courts’ interpretations of the Fair Housing Act (FHA), 42 U. S. C. §3613(c), and the Fair Credit Reporting Act (FCRA), 15 U. S. C. §§1681n, 1681o. A number of courts have construed “actual” damages in the remedial provisions of both statutes to include compensation for mental and emotional distress. See, e.g., Seaton v. Sky Realty Co., 491 F.2d 634, 636–638 (CA7 1974) (authorizing compensatory damages under the FHA, 42 U. S. C. §3612, the predecessor to §3613, for humiliation); Steele v. Title Realty Co., 478 F.2d 380, 384 (CA10 1973) (stating that damages under the FHA “are not limited to out-of-pocket losses but may include an award for emotional distress and humiliation”); Thompson v. San Antonio Retail Merchants Assn., 682 F.2d 509, 513–514 (CA5 1982) (per curiam) (explaining that, “[e]ven when there are no out-of-pocket expenses, humiliation and mental distress do constitute recoverable elements of damage” under the FCRA); Millstone v. O’Hanlon Reports, Inc., 528 F.2d 829, 834–835 (CA8 1976) (approving an award of damages under the FCRA for “loss of sleep, nervousness, frustration and mental anguish”). In other contexts, however, the term has been used or construed more narrowly to authorize damages for only pecuniary harm. In the wrongful-death provision of the Federal Tort Claims Act (FTCA), for example, Congress authorized “actual or compensatory damages, measured by the pecuniary injuries resulting from such death.” 28 U. S. C. §2674, ¶2. At least one court has defined “actual damages” in the Copyright Act of 1909, 17 U. S. C. §101(b) (1970 ed.), as “the extent to which the market value of a copyrighted work has been injured or destroyed by an infringement.” Frank Music Corp. v. Metro-Goldwyn-Mayer, Inc., 772 F.2d 505, 512 (CA9 1985); see also Mackie v. Rieser, 296 F.3d 909, 917 (CA9 2002) (holding that “ ‘hurt feelings’ over the nature of the infringement” have no place in the actual damages calculus). And some courts have construed “actual damages” in the Securities Exchange Act of 1934, 15 U. S. C. §78bb(a), to mean “some form of economic loss.” Ryan v. Foster & Marshall, Inc., 556 F.2d 460, 464 (CA9 1977); see also Osofsky v. Zipf, 645 F.2d 107, 111 (CA2 1981) (stating that the purpose of §78bb(a) “is to compensate civil plaintiffs for economic loss suffered as a result of wrongs committed in violation of the 1934 Act”); Herpich v. Wallace, 430 F.2d 792, 810 (CA5 1970) (noting that the “gist” of an action for damages under the Act is “economic injury”).[3] Because the term “actual damages” has this chameleon-like quality, we cannot rely on any all-purpose definition but must consider the particular context in which the term appears.[4] C The Privacy Act directs agencies to establish safeguards to protect individuals against the disclosure of confiden- tial records “which could result in substantial harm, embarrassment, inconvenience, or unfairness to any indi- vidual on whom information is maintained.” 5 U. S. C. §552a(e)(10); see also §2(b), 88Stat. 1896 (stating that the “purpose of this Act is to provide certain safeguards for an individual against an invasion of personal privacy”). Because the Act serves interests similar to those protected by defamation and privacy torts, there is good reason to infer that Congress relied upon those torts in drafting the Act. In Doe v. Chao, 540 U.S. 614 (2004), we held that the Privacy Act’s remedial provision authorizes plaintiffs to recover a guaranteed minimum award of $1,000 for violations of the Act, but only if they prove at least some “actual damages.” Id., at 620, 627; see §552a(g)(4)(A). Although we did not address the meaning of “actual damages,” id., at 622, n. 5, 627, n. 12, we observed that the provision “parallels” the remedial scheme for the common-law torts of libel per quod and slander, under which plaintiffs can recover “general damages,” but only if they prove “special harm” (also known as “special damages”), id., at 625; see also 3 Restatement of Torts §575, Comments a and b (1938) (hereinafter Restatement); D. Dobbs, Law of Remedies §7.2, pp. 511–513 (1973) (hereinafter Dobbs).[5] “Special damages” are limited to actual pecuniary loss, which must be specially pleaded and proved. 1 D. Haggard, Cooley on Torts §164, p. 580 (4th ed. 1932) (hereinafter Cooley).[6] “General damages,” on the other hand, cover “loss of reputation, shame, mortification, injury to the feelings and the like and need not be alleged in detail and require no proof.” Id., §164, at 579.[7] This parallel between the Privacy Act and the common-law torts of libel per quod and slander suggests the possibility that Congress intended the term “actual damages” in the Act to mean special damages. The basic idea is that Privacy Act victims, like victims of libel per quod or slander, are barred from any recovery unless they can first show actual—that is, pecuniary or material—harm. Upon showing some pecuniary harm, no matter how slight, they can recover the statutory minimum of $1,000, presumably for any unproven harm. That Congress would choose to use the term “actual damages” instead of “special damages” was not without precedent. The terms had occasionally been used interchangeably. See, e.g., Wetzel v. Gulf Oil Corp., 455 F.2d 857, 862 (CA9 1972) (holding that plaintiff could not establish libel per quod because he “did not introduce any valid and sufficient evidence of actual damage”); Electric Furnace Corp. v. Deering Milliken Research Corp., 325 F.2d 761, 765 (CA6 1963) (stating that “libel per quod standing alone without proof of actual damages . . . will not support a verdict for the plaintiff”); M & S Furniture Sales Co. v. Edward J. De Bartolo Corp., 249 Md. 540, 544, 241 A.2d 126, 128 (1968) (“In the case of words or conduct actionable only per quod, the injurious effect must be established by allegations and proof of special damage and in such cases it is not only necessary to plead and show that the words or actions were defamatory, but it must also appear that such words or conduct caused actual damage”); Clementson v. Minnesota Tribune Co., 45 Minn. 303, 47 N.W. 781 (1891) (distinguishing “actual, or, as they are sometimes termed, ‘special,’ damages” from “general damages—that is, damages not pecuniary in their nature”).[8] Any doubt about the plausibility of construing “actual damages” in the Privacy Act synonymously with “special damages” is put to rest by Congress’ refusal to authorize “general damages.” In an uncodified section of the Act, Congress established the Privacy Protection Study Commission to consider, among other things, “whether the Federal Government should be liable for general dam- ages.” §5(c)(2)(B)(iii), 88Stat. 1907, note following 5 U. S. C. §552a, p. 712. As we explained in Doe, “Congress left the question of general damages . . . for another day.” 540 U. S., at 622. Although the Commission later recom- mended that general damages be allowed, ibid., n. 4, Congress never amended the Act to include them. For that reason, we held that it was “beyond serious doubt” that general damages are not available for violations of the Privacy Act. Id., at 622. By authorizing recovery for “actual” but not for “general” damages, Congress made clear that it viewed those terms as mutually exclusive. In actions for defamation and related dignitary torts, two categories of compensa- tory damages are recoverable: general damages and special damages. Cooley §164, at 579; see also 4 Restatement §867, Comment d (1939) (noting that damages for interference with privacy “can be awarded in the same way in which general damages are given for defamation”).[9] Because Congress declined to authorize “general damages,” we think it likely that Congress intended “actual dam- ages” in the Privacy Act to mean special damages for proven pecuniary loss. Not surprisingly, this interpretation was accepted by the Privacy Protection Study Commission, an expert body authorized by Congress and highly sensitive to the Act’s goals. The Commission understood “actual damages” in the Act to be “a synonym for special damages as that term is used in defamation cases.” Personal Privacy in an Information Society: The Report of the Privacy Protection Study Commission 530 (July 1977); see also ibid. (“The legislative history and language of the Act suggest that Congress meant to restrict recovery to specific pecuniary losses until the Commission could weigh the propriety of extending the standard of recovery”). Although we are not bound in any way by the Commission’s report, we think it confirms the reasonableness of interpreting “actual damages” in the unique context of the Privacy Act as the equivalent of special damages. D We do not claim that the contrary reading of the statute accepted by the Court of Appeals and advanced now by respondent is inconceivable. But because the Privacy Act waives the Federal Government’s sovereign immunity, the question we must answer is whether it is plausible to read the statute, as the Government does, to authorize only damages for economic loss. Nordic Village, 503 U. S., at 34, 37. When waiving the Government’s sovereign immunity, Congress must speak unequivocally. Lane, 518 U. S., at 192. Here, we conclude that it did not. As a consequence, we adopt an interpretation of “actual damages” limited to proven pecuniary or economic harm. To do otherwise would expand the scope of Congress’ sovereign immunity waiver beyond what the statutory text clearly requires. III None of respondent’s contrary arguments suffices to overcome the sovereign immunity canon. A Respondent notes that the term “actual damages” has often been defined broadly in common-law cases, and in our own, to include all compensatory damages. See Brief for Respondent 18–25. For example, in Birdsall v. Coolidge, 93 U.S. 64 (1876), a patent infringement case, we observed that “[c]ompensatory damages and actual damages mean the same thing.” Ibid. And in Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974), we wrote that actual injury in the defamation context “is not limited to out-of-pocket loss” and that it customarily includes “impairment of reputation and standing in the community, personal humiliation, and mental anguish and suffering.” Id., at 350. These cases and others cited by respondent stand for the unremarkable point that the term “actual damages” can include nonpecuniary loss. But this generic meaning does not establish with the requisite clarity that the Privacy Act, with its distinctive features, authorizes damages for mental and emotional distress. As we already explained, the term “actual damages” takes on different meanings in different contexts. B Respondent’s stronger argument is that the exclusion of “general damages” from the statute simply means that there can be no recovery for presumed damages. Privacy Act victims can still recover for mental and emotional distress, says respondent, so long as it is proved. See Brief for Respondent 54–56.[10] This argument is flawed because it suggests that proven mental and emotional distress does not count as general damages. The term “general damages” is not limited to compensation for unproven injuries; it includes compensation for proven injuries as well. See 3 Restatement §621, Comment a (noting that general damages compensate for “harm which . . . is proved, or, in the absence of proof, is assumed to have caused to [the plaintiff’s] reputation”). To be sure, specific proof of emotional harm is not required to recover general damages for dignitary torts. Dobbs §7.3, at 529. But it does not follow that general damages cannot be recovered for emotional harm that is actually proved. Aside from the fact that general damages need not be proved, what distinguishes those damages, whether proved or not, from the only other category of compensa- tory damages available in the relevant common-law suits is the type of harm. In defamation and privacy cases, “the affront to the plaintiff’s dignity and the emotional harm done” are “called general damages, to distinguish them from proof of actual economic harm,” which is called “special damages.” Id., §3.2, at 139; see also supra, at 10, 12–13, and nn. 6, 7, 9. Therefore, the converse of general damages is special damages, not all proven damages, as respondent would have it. Because Congress removed “general damages” from the Act’s remedial provision, it is reasonable to infer that Congress foreclosed recovery for nonpecuniary harm, even if such harm can be proved, and instead waived the Government’s sovereign immunity only with respect to harm compensable as special damages. C Looking beyond the Privacy Act’s text, respondent points to the use of the term “actual” damages in the remedial provisions of the FHA, 42 U. S. C. §3613(c), and the FCRA, 15 U. S. C. §§1681n, 1681o. As previously mentioned, courts have held that “actual” damages within the meaning of these statutes include compensation for mental and emotional distress. Supra, at 7. Citing the rule of construction that Congress intends the same language in similar statutes to have the same meaning, see Northcross v. Board of Ed. of Memphis City Schools, 412 U.S. 427, 428 (1973) (per curiam), respondent argues that the Privacy Act should also be interpreted as authorizing damages for mental and emotional distress. See Brief for Respondent 25–32. Assuming for the sake of argument that these lower court decisions are correct, they provide only weak support for respondent’s argument here. Since the term “actual damages” can mean different things in different contexts, statutes other than the Privacy Act provide only limited interpretive aid, and that is especially true here. Neither the FHA nor the FCRA contains text that precisely mirrors the Privacy Act.[11] In neither of those statutes did Congress specifically decline to authorize recovery for general damages as it did in the Privacy Act. Supra, at 12–13. And most importantly, none of the lower court cases interpreting the statutes, which respondent has cited, see Brief for Respondent 29–31, involves the sovereign immunity canon. Respondent also points to the FTCA, but the FTCA’s general liability provision does not even use the term “actual damages.” It instead provides that the “United States shall be liable” for certain tort claims “in the same manner and to the same extent as a private individual” under relevant state law. 28 U. S. C. §2674, ¶1. For that reason alone, the FTCA’s general liability provision is not a reliable source for interpreting the term “actual damages” in the Privacy Act. Nor does the FTCA’s wrongful- death provision—which authorizes “actual or compensa- tory damages, measured by the pecuniary injuries resulting from such death,” §2674, ¶2—prove that Congress understood the term “actual damages” in the Privacy Act to include nonpecuniary mental and emotional harm. To the contrary, it proves that actual damages can be understood to entail only pecuniary harm depending on the context. Because the FTCA, like the FHA and FCRA, does not share the same text or design as the Privacy Act, it is not a fitting analog for construing the Act. D Finally, respondent argues that excluding damages for mental and emotional harm would lead to absurd results. Persons suffering relatively minor pecuniary loss would be entitled to recover $1,000, while others suffering only severe and debilitating mental or emotional distress would get nothing. See Brief for Respondent 33–35. Contrary to respondent’s suggestion, however, there is nothing absurd about a scheme that limits the Government’s Privacy Act liability to harm that can be substantiated by proof of tangible economic loss. Respondent insists that such a scheme would frustrate the Privacy Act’s remedial purpose, but that ignores the fact that, by deliberately refusing to authorize general damages, Congress intended to cabin relief, not to maximize it.[12] * * * In sum, applying traditional rules of construction, we hold that the Privacy Act does not unequivocally authorize an award of damages for mental or emotional distress. Accordingly, the Act does not waive the Federal Government’s sovereign immunity from liability for such harms. We therefore reverse the judgment of the United States Court of Appeals for the Ninth Circuit and remand the case for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Notes 1 Respondent eventually applied for recertification as a pilot. After reviewing respondent’s medical records, including information about his HIV diagnosis and treatment, the FAA reissued his pilot certificate and medical certificate. Brief for Respondent 5, n. 1. 2 With certain exceptions, it is unlawful for an agency to disclose a record to another agency without the written consent of the person to whom the record pertains. 5 U. S. C. §552a(b). One exception to this nondisclosure requirement applies when the head of an agency makes a written request for law enforcement purposes to the agency that maintains the record. See §552a(b)(7). The agencies in this case could easily have shared respondent’s medical records pursuant to the procedures prescribed by the Privacy Act, but the District Court concluded that they failed to do so. 3 This narrow usage is reflected in contemporaneous state-court decisions as well. See, e.g., Reist v. Manwiller, 231 Pa. Super. 444, 449, n. 4, 332 A.2d 518, 520, n. 4 (1974) (explaining that recovery for intentional infliction of emotional distress is allowed “despite the total absence of physical injury and actual damages”); Nalder v. Crest Corp., 93 Idaho 744, 749, 472 P.2d 310, 315 (1970) (noting that damages for “mental anguish” due to the wrongful execution of a judgment “are allowable only as an element of punitive but not of actual damages”). It is also reflected in post-Privacy Act statutes and judicial decisions. See, e.g., 17 U. S. C. §1009(d)(1)(A)(ii) (defining “actual damages” in the Audio Home Recording Act of 1992 as “the royalty payments that should have been paid”); 18 U. S. C. §2318(e)(3) (2006 ed., Supp. IV) (calculating “actual damages” for purposes of a counterfeit labeling statute in terms of financial loss); Guzman v. Western State Bank of Devils Lake, 540 F.2d 948, 953 (CA8 1976) (stating that compensatory damages in a civil rights suit “can be awarded for emotional and mental distress even though no actual damages are proven”). 4 The dissent criticizes us for noting that the dictionary definition contains an element of circularity. The dissent says that the definition—“ ‘[a]ctual damages’ compensate for actual injury”—is “plain enough.” Post, at 3 (opinion of Sotomayor, J.). But defining “actual” damages by reference to “actual” injury is hardly helpful when our task is to determine what Congress meant by “actual.” The dissent’s reference to the current version of Black’s Law Dictionary, which provides that “actual damages” can mean “tangible damages,” only highlights the term’s ambiguity. See Black’s Law Dictionary 445 (9th ed. 2009). If “actual damages” can mean “tangible damages,” then it can be construed not to include intangible harm, like mental and emotional distress. Similarly unhelpful is the dissent’s citation to a general-purpose dictionary that defines “actual” as “existing in fact or reality” and “damages” as “compensation or satisfaction imposed by law for a wrong or injury.” Webster’s Third New International Dictionary 22, 571 (2002) (emphasis added). Combining these two lay definitions says nothing about whether compensation for mental and emotional distress is in fact imposed by law. The definitions merely beg the question we are trying to answer. It comes as little surprise, therefore, that “actual damages” has taken on different meanings in different statutes, as our examples amply illustrate. 5 Libel per quod and slander (as opposed to libel and slander per se) apply to a communication that is not defamatory on its face but that is defamatory when coupled with some other extrinsic fact. Dobbs §7.2, at 512–513. 6 See also 3 Restatement §575, Comment b (“Special harm . . . is harm of a material and generally of a pecuniary nature”); Dobbs §7.2, at 520 (“Special damages in defamation cases mean pecuniary damages, or at least ‘material loss’ ” (footnote omitted)). Special damages do not include mental or emotional distress. See 3 Restatement §575, Comment c (“The emotional distress caused to the person slandered by his knowledge that he has been defamed is not special harm and this is so although the distress results in a serious illness”); Dobbs §7.2, at 520 (“Even under the more modern approach, special damages in defamation cases must be economic in nature, and it is not enough that the plaintiff has suffered harm to reputation, mental anguish or other dignitary harm, unless he has also suffered the loss of something having economic value”). 7 See also id., §3.2, at 139 (explaining that noneconomic harms “are called general damages”); W. Prosser, Law of Torts §112, p. 761 (4th ed. 1971) (noting that “ ‘general’ damages may be recovered for the injury to the plaintiff’s reputation, his wounded feelings and humiliation, and resulting physical illness and pain, as well as estimated future dam-ages of the same kind” (footnotes omitted)); 3 Restatement §621, Com-ment a (stating that, in actions for defamation, a plaintiff may recover general damages for “impairment of his reputation or, through loss of reputation, to his other interests”). 8 The dissent disregards these precedents as the product of careless imprecision. Post, at 8, n. 6. But just as we assume that Congress did not act carelessly, we should not be so quick to assume that the courts did. The better explanation for these precedents is not that the courts were careless, but that the term “actual damages” has a varied meaning that, depending on the context, can be limited to compensation for only pecuniary harm. 9 See also Moriarty v. Lippe, 162 Conn. 371, 382–383, metricconverter294 A.2d 326, 332–333 (1972) (“Having admittedly alleged or proven no special damages, the plaintiff here is limited to a recovery of general damages . . .”); Meyerle v. Pioneer Publishing Co., 45 N. D. 568, 574, 178 N.W. 792, 794 (1920) (per curiam) (“Generally speaking, there are recognized two classes of damages in libel cases, general damages and special damages”); Winans v. Chapman, 104 Kan. 664, 666, 180 P. 266, 267 (1919) (“Actual damages include both general and special damages”); Childers v. San Jose Mercury Printing & Publishing Co., 105 Cal. 284, 288–289, 38 P. 903, 904 (1894) (explaining that special damages, “asa branch of actual damages[,] may be recovered when actual pecuniary loss has been sustained” and that the “remaining branch of actual damages embraces recovery for loss of reputation, shame, mortification, injury to feelings, etc.”); see generally Dobbs §7.3, at 531 (“Though the dignitary torts often involve only general damages . . . , they sometimes produce actual pecuniary loss. When this happens, the plaintiff is usually entitled to recover any special damage he can prove . . . ”); metricconverter1 F. Harper & F. James, Law of Torts §5.30, p. 470 (1956) (“When liability for defamation is established, the defendant, in addition to such ‘general’ damages as may be assessed by the jury, is also liable for any special damage which he has sustained”). 10 The dissent advances the same argument. See post, at 9–11. 11 Compare 42 U. S. C. §3613(c)(1) (stating that “the court may award to the plaintiff actual and punitive damages”); 15 U. S. C. §1681n(a)(1) (authorizing “(A) any actual damages sustained by the consumer as a result of the failure or damages of not less than $100 and not more than $1,000; or (B) . . . actual damages sustained by the consumer as a result of the failure or $1,000, whichever is greater”); §1681o(a)(1) (authorizing “any actual damages sustained by the consumer as a result of the failure”) with 5 U. S. C. §552a(g)(4)(A) (authorizing “actual damages sustained by the individual as a result of the refusal or failure, but in no case shall a person entitled to recovery receive less than the sum of $1,000”). 12 Despite its rhetoric, the dissent does not dispute most of the steps in our analysis. For example, although the dissent belittles the sovereign immunity canon, the dissent does not call for its abandonment. See post, at 2–3. Nor does the dissent point out any error in our understanding of the canon’s meaning. See ibid. The dissent acknowledges that statutes and judicial opinions sometimes use the term “actual damages” to mean pecuniary harm, see post, at 5, and that determining its meaning in a particular statute requires consideration of context, see ibid. In addition, the dissent concedes—as it must in light of our reasoning in Doe v. Chao, 540 U.S. 614 (2004)—that the common law of defamation has relevance in construing the term “actual damages” in the Privacy Act. See post, at 7–9. The dissent’s argument thus boils down to this: The text and purpose of the Privacy Act make it clear beyond any reasonable dispute that the term “actual damages,” as used in the Act, means compensatory damages for all proven harm and not just damages for pecuniary harm. The dissent reasons that, because the Act seeks to prevent pecuniary and nonpecuniary harm, Congress must have intended to authorize the recovery of money damages from the Federal Government for both types of harm. This inference is plausible, but it surely is not unavoidable. The Act deters violations of its substantive provisions in other ways—for instance, by permitting recovery for economic injury; by imposing criminal sanctions for some violations, see 5 U. S. C. §552a(i); and possibly by allowing for injunctive relief under the Administrative Procedure Act (APA), 5 U. S. C. §§702, 706; see Doe, supra, at 619, n. 1 (noting that the absence of equitable relief in suits under §§552a(g)(1)(C) or (D) may be explained by the availability of such relief under the APA).

567.US.2011_10-1293

Title 18 U. S. C. §1464 bans the broadcast of “any obscene, indecent, or profane language.” The Federal Communications Commission (Commission) began enforcing §1464 in the 1970’s. In FCC v. Pacif-ica Foundation, 438 U.S. 726, this Court found that the Commission’s order banning George Carlin’s “Filthy Words” monologue passed First Amendment scrutiny, but did not decide whether “an occasional expletive . . . would justify any sanction,” id., at 750. In the ensuing years, the Commission went from strictly observing the narrow circumstances of Pacifica to indicating that it would assess the full context of allegedly indecent broadcasts rather than limit its regulation to an index of indecent words or pictures. However, it continued to note the important difference between isolated and repeated broadcasts of indecent material. And in a 2001 policy statement, it even included, as one of the factors significant to the determination of what was patently offensive, “whether the material dwells on or repeats at length” the offending description or depiction. It was against this regulatory background that the three incidents at issue took place. Two concern isolated utterances of obscene words during two live broadcasts aired by respondent Fox Television Stations, Inc. The third occurred during an episode of a television show broadcast by respondent ABC Television Network, when the nude buttocks of an adult female character were shown for approximately seven seconds and the side of her breast for a moment. After these incidents, but before the Commission issued Notices of Apparent Liability to Fox and ABC, the Commission issued its Golden Globes Order, declaring for the first time that fleeting expletives could be actionable. It then concluded that the Fox and ABC broadcasts violated this new standard. It found the Fox broadcasts indecent, but declined to propose forfeitures. The Second Circuit reversed, finding the Commission’s decision to modify its indecency enforcement regime to regulate fleeting expletives arbitrary and capricious. This Court reversed and remanded for the Second Circuit to address respondents’ First Amendment challenges. FCC v. Fox Television Stations, Inc., 556 U.S. 502. On remand, the Second Circuit found the policy unconstitutionally vague and invalidated it in its entirety. In the ABC case, the Commission found the display actionably indecent, and imposed a $27,500 forfeiture on each of the 45 ABC-affiliated stations that aired the episode. The Second Circuit vacated the order in light of its Fox decision. Held: Because the Commission failed to give Fox or ABC fair notice prior to the broadcasts in question that fleeting expletives and momentary nudity could be found actionably indecent, the Commission’s standards as applied to these broadcasts were vague. Pp. 11–18. (a) The fundamental principle that laws regulating persons or entities must give fair notice of what conduct is required or proscribed, see, e.g., Connally v. General Constr. Co., 269 U.S. 385, 391, is essential to the protections provided by the Fifth Amendment’s Due Process Clause, see United States v. Williams, 553 U.S. 285, 304, which requires the invalidation of impermissibly vague laws. A conviction or punishment fails to comply with due process if the statute or regulation under which it is obtained “fails to provide a person of ordinary intelligence fair notice of what is prohibited, or is so standardless that it authorizes or encourages seriously discriminatory enforcement.” Ibid. The void for vagueness doctrine addresses at least two connected but discrete due process concerns: Regulated parties should know what is required of them so they may act accordingly; and precision and guidance are necessary so that those enforcing the law do not act in an arbitrary or discriminatory way. When speech is involved, rigorous adherence to those requirements is necessary to ensure that ambiguity does not chill protected speech. Pp. 11–12. (b) These concerns are implicated here, where the broadcasters claim that the lengthy procedural history of their cases shows that they did not have fair notice of what was forbidden. Under the 2001 Guidelines in force when the broadcasts occurred, a key consideration was “whether the material dwell[ed] on or repeat[ed] at length” the offending description or depiction, but in the 2004 Golden Globes Order, issued after the broadcasts, the Commission changed course and held that fleeting expletives could be a statutory violation. It then applied this new principle to these cases. Its lack of notice to Fox and ABC of its changed interpretation failed to give them “fair notice of what is prohibited.” Williams, supra, at 304. Pp. 12–13. (c) Neither of the Government’s contrary arguments is persuasive. It claims that Fox cannot establish unconstitutional vagueness because the Commission declined to impose a forfeiture on Fox and said that it would not consider the indecent broadcast in renewing station licenses or in other contexts. But the Commission has the statutory power to take into account “any history of prior offenses” when setting a forfeiture penalty, 47 U. S. C. §503(b)(2)(E), and the due process protection against vague regulations “does not leave [regulated parties] . . . at the mercy of noblesse oblige.” United States v. Stevens, 559 U. S. ___, ___. The challenged orders could also have an adverse impact on Fox’s reputation with audiences and advertisers alike. The Government argues that ABC had notice that its broadcast would be considered indecent. But an isolated statement in a 1960 Commission decision declaring that televising nudes might be contrary to §1464 does not suffice for the fair notice required when the Government intends to impose over a $1 million fine for allegedly impermissible speech. Moreover, previous Commission decisions had declined to find isolated and brief moments of nudity actionably indecent. In light of these agency decisions, and the absence of any notice in the 2001 Guidance that seven seconds of nude buttocks would be found indecent, ABC lacked constitutionally sufficient notice prior to being sanctioned. Pp. 13–17. (d) It is necessary to make three observations about this decision’s scope. First, because the Court resolves these cases on fair notice grounds under the Due Process Clause, it need not address the First Amendment implications of the Commission’s indecency policy or reconsider Pacifica at this time. Second, because the Court rules that Fox and ABC lacked notice at the time of their broadcasts that their material could be found actionably indecent under then-existing policies, the Court need not address the constitutionality of the current indecency policy as expressed in the Golden Globes Order and subsequent adjudications. Third, this opinion leaves the Commission free to modify its current indecency policy in light of its determination of the public interest and applicable legal requirements and leaves courts free to review the current, or any modified, policy in light of its content and application. Pp. 17–18. 613 F.3d 317 (first case) and 404 Fed. Appx. 530 (second case), vacated and remanded. Kennedy, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Thomas, Breyer, Alito, and Kagan, JJ., joined. Ginsburg, J., filed an opinion concurring in the judgment. Sotomayor, J., took no part in the consideration or decision of the cases. Notes 1 Together with Federal Communications Commission v. ABC, Inc., et al. (see this Court’s Rule 12.4), also on certiorari to the same court.

In FCC v. Fox Television Stations, Inc., 556 U.S. 502, 529 (2009) (Fox I), the Court held that the Federal Communication Commission’s decision to modify its indecency enforcement regime to regulate so-called fleeting expletives was neither arbitrary nor capricious. The Court then declined to address the constitutionality of the policy, however, because the United States Court of Appeals for the Second Circuit had yet to do so. On remand, the Court of Appeals found the policy was vague and, as a result, unconstitutional. metricconverter613 F.3d 317 (2010). The case now returns to this Court for decision upon the constitutional question. I In Fox I, the Court described both the regulatory framework through which the Commission regulates broadcast indecency and the long procedural history of this case. The Court need not repeat all that history, but some preliminary discussion is necessary to understand the constitutional issue the case now presents. A Title 18 U. S. C. §1464 provides that “[w]hoever utters any obscene, indecent, or profane language by means of radio communication shall be fined . . . or imprisoned not more than two years, or both.” The Federal Communi-cations Commission (Commission) has been instructed by Congress to enforce §1464 between the hours of 6 a.m. and 10 p.m., see Public Telecommunications Act of 1992, §15(a), 106Stat. 954, note following 47 U. S. C. §303, p. 113 (Broadcasting of Indecent Programming). And the Commission has applied its regulations to radio and television broadcasters alike, see Fox I, supra, at 505–506; see also 47 CFR §73.3999 (2010) (Commission regulation prohibiting the broadcast of any obscene material or any indecent material between 6 a.m. and 10 p.m.). Although the Commission has had the authority to regulate indecent broadcasts under §1464 since 1948 (and its prede-cessor commission, the Federal Radio Commission, since 1927), it did not begin to enforce §1464 until the 1970’s. See Campbell, Pacifica Reconsidered: Implications for the Current Controversy over Broadcast Indecency, 63 Fed. Com. L. J. 195, 198 (2010). This Court first reviewed the Commission’s indecency policy in FCC v. Pacifica Foundation, 438 U.S. 726 (1978). In Pacifica, the Commission determined that George Carlin’s “Filthy Words” monologue was indecent. It contained “ ‘language that describes, in terms patently offensive as measured by contemporary community standards for the broadcast medium, sexual or excretory activities and organs, at times of the day when there is a reasonable risk that children may be in the audience.’ ” Id., at 732 (quoting 56 F. C. C. 2d 94, 98 (1975)). This Court upheld the Commission’s ruling. The broadcaster’s statutory challenge was rejected. The Court held the Commission was not engaged in impermissible censorship within the meaning of 47 U. S. C. §326 (1976 ed.), see 438 U. S., at 735–739, and that §1464’s definition of indecency was not confined to speech with an appeal to the prurient interest, see id., at 738–741. Finding no First Amendment violation, the decision explained the constitutional standard under which regulations of broadcasters are assessed. It observed that “broadcast media have established a uniquely pervasive presence in the lives of all Americans,” id., at 748, and that “broadcasting is uniquely accessible to children, even those too young to read,” id., at 749. In light of these considerations, “broadcasting . . . has received the most limited First Amendment protection.” Id., at 748. Under this standard the Commission’s order passed constitutional scrutiny. The Court did note the narrowness of its holding, explaining that it was not deciding whether “an occasional expletive . . . would justify any sanction.” Id., at 750; see also id., at 760–761 (Powell, J., concur- ring in part and concurring in judgment) (“[C]ertainly the Court’s holding . . . does not speak to cases involving the isolated use of a potentially offensive word in the course of a radio broadcast, as distinguished from the verbal shock treatment administered by respondent here”). From 1978 to 1987, the Commission did not go beyond the narrow circumstances of Pacifica and brought no indecency enforcement actions. See In re Infinity Broadcasting Corp., 3 FCC Rcd. 930 (1987); see also In re Application of WGBH Educ. Foundation, 69 F. C. C. 2d 1250, 1254 (1978) (Commission declaring it “intend[s] strictly to observe the narrowness of the Pacifica holding”). Recognizing that Pacifica provided “no general prerogative to intervene in any case where words similar or identical to those in Pacifica are broadcast over a licensed radio or television station,” the Commission distinguished between the “repetitive occurrence of the ‘indecent’ words” (such as in the Carlin monologue) and an “isolated” or “occasional” expletive, that would not necessarily be actionable. 69 F. C. C. 2d, at 1254. In 1987, the Commission determined it was applying the Pacifica standard in too narrow a way. It stated that in later cases its definition of indecent language would “appropriately includ[e] a broader range of material than the seven specific words at issue in [the Carlin monologue].” In re Pacifica Foundation Inc., 2 FCC Rcd. 2698, 2699. Thus, the Commission indicated it would use the “generic definition of indecency” articulated in its 1975 Pacifica order, Infinity Order, 3 FCC Rcd., at 930, and assess the full context of allegedly indecent broadcasts rather than limiting its regulation to a “comprehensive index . . . of indecent words or pictorial depictions,” id., at 932. Even under this context based approach, the Commission continued to note the important difference between isolated and repeated broadcasts of indecent material. See ibid. (considering variables in determining whether material is patently offensive including “whether allegedly offensive material is isolated or fleeting”). In the context of expletives, the Commission determined “deliberate and repetitive use in a patently offensive manner is a requisite to a finding of indecency.” Pacifica Order, 2 FCC Rcd., at 2699. For speech “involving the description or depiction of sexual or excretory functions . . . [t]he mere fact that specific words or phrases are not repeated does not mandate a finding that material that is otherwise patently offensive . . . is not indecent.” Ibid. In 2001, the Commission issued a policy statement intended “to provide guidance to the broadcast industry regarding [its] caselaw interpreting 18 U. S. C. §1464 and [its] enforcement policies with respect to broadcast indecency.” In re Industry Guidance on Commission’s Case Law Interpreting 18 U. S. C. §1464 and Enforcement Policies Regarding Broadcast Indecency, 16 FCC Rcd. 7999. In that document the Commission restated that for material to be indecent it must depict sexual or excretory organs or activities and be patently offensive as measured by contemporary community standards for the broadcast medium. Id., at 8002. Describing the framework of what it considered patently offensive, the Commission explained that three factors had proved significant: “(1) [T]he explicitness or graphic nature of the description or depiction of sexual or excretory organs or activities; (2) whether the material dwells on or repeats at length descriptions of sexual or excretory organs or activities; (3) whether the material appears to pander or is used to titillate, or whether the material appears to have been presented for its shock value.” Id., at 8003 (emphasis deleted). As regards the second of these factors, the Commission explained that “[r]epetition of and persistent focus on sexual or excretory material have been cited consistently as factors that exacerbate the potential offensiveness of broadcasts. In contrast, where sexual or excretory references have been made once or have been passing or fleeting in nature, this characteristic has tended to weigh against a finding of indecency.” Id., at 8008. The Commission then gave examples of material that was not found indecent because it was fleeting and isolated, id., at 8008–8009 (citing, e.g., L. M. Communications of South Carolina, Inc. (WYBB(FM)), 7 FCC Rcd. 1595 (MMB 1992) (finding “a fleeting and isolated utterance” in the context of live and spontaneous programming not actionable)), and contrasted it with fleeting references that were found patently offensive in light of other factors, 16 FCC Rcd., at 8009 (citing, e.g., Tempe Radio, Inc. (KUPD–FM), 12 FCC Rcd. 21828 (MMB 1997) (finding fleeting language that clearly refers to sexual activity with a child to be patently offensive)). B It was against this regulatory background that the three incidents of alleged indecency at issue here took place. First, in the 2002 Billboard Music Awards, broadcast by respondent Fox Television Stations, Inc., the singer Cher exclaimed during an unscripted acceptance speech: “I’ve also had my critics for the last 40 years saying that I was on my way out every year. Right. So f *** ‘em.” 613 F. 3d, at 323. Second, Fox broadcast the Billboard Music Awards again in 2003. There, a person named Nicole Richie made the following unscripted remark while presenting an award: “Have you ever tried to get cow s*** out of a Prada purse? It’s not so f ***ing simple.” Ibid. The third in-cident involved an episode of NYPD Blue, a regular tele-vision show broadcast by respondent ABC Television Network. The episode broadcast on February 25, 2003, showed the nude buttocks of an adult female character for approximately seven seconds and for a moment the side of her breast. During the scene, in which the character was preparing to take a shower, a child portraying her boy-friend’s son entered the bathroom. A moment of awkwardness followed. 404 Fed. Appx. 530, 533–534 (CA2 2011). The Commission received indecency complaints about all three broadcasts. See Fox I, 556 U. S., at 510; 404 Fed. Appx., at 534. After these incidents, but before the Commission issued Notices of Apparent Liability to Fox and ABC, the Commission issued a decision sanctioning NBC for a comment made by the singer Bono during the 2003 Golden Globe Awards. Upon winning the award for Best Original Song, Bono exclaimed: “ ‘This is really, really, f ***ing brilliant. Really, really great.’ ” In re Complaints Against Various Broadcast Licensees Regarding Their Airing of the “Golden Globe Awards” Program, 19 FCC Rcd. 4975, 4976, n. 4 (2004) (Golden Globes Order). Reversing a decision by its enforcement bureau, the Commission found the use of the F-word actionably indecent. Id., at 4975–4976. The Commission held that the word was “one of the most vul-gar, graphic and explicit descriptions of sexual activity in the English language,” and thus found “any use of that word or a variation, in any context, inherently has a sex-ual connotation.” Id., at 4978–4979. Turning to the isolated nature of the expletive, the Commission reversed prior rulings that had found fleeting expletives not indecent. The Commission held “the mere fact that specific words or phrases are not sustained or repeated does not mandate a finding that material that is otherwise patently offensive to the broadcast medium is not indecent.” Id., at 4980; see also id., at 4982 (“Just as the Court [in Pacifica] held that . . . the George Carlin routine ‘could have enlarged a child’s vocabulary in an instant,’ we believe that even isolated broadcasts of the ‘F-Word’ in situations such as that here could do so as well”). C Even though the incidents at issue in these cases took place before the Golden Globes Order, the Commission applied its new policy regarding fleeting expletives and fleeting nudity. It found the broadcasts by respondents Fox and ABC to be in violation of this standard. 1 As to Fox, the Commission found the two Billboard Awards broadcasts indecent in In re Complaints Regarding Various Television Broadcasts Between February 2, 2002, and March 8, 2005, 21 FCC Rcd. 2664 (2006). Numerous parties petitioned for a review of the order in the United States Court of Appeals for the Second Circuit. The Court of Appeals granted the Commission’s request for a voluntary remand so that it could respond to the parties’ objections. Fox Television Stations, Inc. v. FCC, 489 F.3d 444, 453 (2007). In its remand order, the Commission applied its tripartite definition of patently offensive material from its 2001 Order and found that both broadcasts fell well within its scope. See In re Complaints Regarding Various Television Broadcasts Between February 2, 2002, and March 8, 2005, 21 FCC Rcd. 13299 (2006) (Remand Order); see also Fox I, supra, at 511–513 (discussing in detail the Commission’s findings). As pertains to the constitutional issue in these cases, the Commission noted that under the policy clarified in the Golden Globes Order, “categorically requiring repeated use of expletives in order to find material indecent is inconsistent with our general approach to indecency enforcement.” Remand Order, 21 FCC Rcd., at 13308; see also id., at 13325 (“[U]nder our Golden Globe precedent, the fact that Cher used the ‘F-word’ once does not remove her comment from the realm of actionable indecency”). Though the Commission deemed Fox should have known Nicole Richie’s comments were actionably indecent even prior to the Golden Globes Order, 21 FCC Rcd., at 13307, it declined to propose a forfeiture in light of the limited nature of the Sec-ond Circuit’s remand. Id., at 13321. The Commission acknowledged that “it was not apparent that Fox could be penalized for Cher’s comment at the time it was broadcast.” And so, as in the Golden Globes case it imposed no penalty for that broadcast. Id., at 13324, 13326. Fox and various intervenors returned to the United States Court of Appeals for the Second Circuit, raising ad-ministrative, statutory, and constitutional challenges to the Commission’s indecency regulations. See Fox Television Stations, Inc. v. FCC, 489 F.3d 444. In a 2-to-1 decision, with Judge Leval dissenting, the Court of Appeals found the Remand Order arbitrary and capricious because “the FCC has made a 180-degree turn regarding its treatment of ‘fleeting expletives’ without providing a reasoned explanation justifying the about-face.” 489 F. 3d, at 455. While noting its skepticism as to whether the Commission’s fleeting expletive regime “would pass constitutional muster,” the Court of Appeals found it unnecessary to ad-dress the issue. Id., at 462. The case came here on certiorari. Citing the Administrative Procedure Act, 5 U. S. C. §551 et seq., this Court noted that the Judiciary may set aside agency action that is arbitrary or capricious. In the context of a change in policy (such as the Commission’s determination that fleeting expletives could be indecent), the decision held an agency, in the ordinary course, should acknowledge that it is in fact changing its position and “show that there are good reasons for the new policy.” Fox I, 553 U. S., at 515. There is no need, however, for an agency to provide detailed justifications for every change or to show that the reasons for the new policy are better than the reasons for the old one. Ibid. Judged under this standard, the Court in Fox I found the Commission’s new indecency enforcement policy neither arbitrary nor capricious. Id., at 517. The Court noted the Commission had acknowledged breaking new ground in ruling that fleeting and nonliteral expletives could be indecent under the controlling standards; the Court concluded the agency’s reasons for expanding the scope of its enforcement activity were rational. Ibid. Not only was it “certainly reasonable to determine that it made no sense to distinguish between literal and nonliteral uses of offensive words,” ibid., but the Court agreed that the Commission’s decision to “look at the patent offensiveness of even isolated uses of sexual and excretory words fits with the context-based approach [approved] . . . in Pacifica.” Ibid. Given that “[e]ven isolated utterances can . . . constitute harmful ‘first blow[s]’ to children,” the Court held that the Commission could “decide it needed to step away from its old regime where nonrepetitive use of an expletive was per se nonactionable.” Id., at 518. Having found the agency’s action to be neither arbitrary nor capricious, the Court remanded for the Court of Appeals to address respondents’ First Amendment challenges. Id., at 529–530. On remand from Fox I, the Court of Appeals held the Commission’s indecency policy unconstitutionally vague and invalidated it in its entirety. 613 F. 3d, at 327. The Court of Appeals found the policy, as expressed in the 2001 Guidance and subsequent Commission decisions, failed to give broadcasters sufficient notice of what would be considered indecent. Surveying a number of Commission adjudications, the court found the Commission was inconsistent as to which words it deemed patently offensive. See id., at 330. It also determined that the Com-mission’s presumptive prohibition on the F-word and the S-word was plagued by vagueness because the Commission had on occasion found the fleeting use of those words not indecent provided they occurred during a bona fide news interview or were “demonstrably essential to the nature of an artistic or educational work.” Id., at 331 (internal quotation marks omitted). The Commission’s application of these exceptions, according to the Court of Appeals, left broadcasters guessing whether an expletive would be deemed artistically integral to a program or whether a particular broadcast would be considered a bona fide news interview. The Court of Appeals found the vagueness in-herent in the policy had forced broadcasters to “choose between not airing . . . controversial programs [or] risking massive fines or possibly even loss of their licenses.” Id., at 334. And the court found that there was “ample evidence in the record” that this harsh choice had led to a chill of protected speech. Ibid. 2 The procedural history regarding ABC is more brief. On February 19, 2008, the Commission issued a forfeiture order finding the display of the woman’s nude buttocks in NYPD Blue was actionably indecent. See In re Complaints Against Various Television Licensees Concerning Their February 24, 2003 Broadcast of the Program “NYPD Blue”, 23 FCC Rcd. 3147 (2008). The Commission determined that, regardless of medical definitions, displays of buttocks fell within the category of displays of sexual or excretory organs because the depiction was “widely associated with sexual arousal and closely associated by most people with excretory activities.” Id., at 3150. The scene was deemed patently offensive as measured by contemporary community standards, ibid.; and the Commission determined that “[t]he female actor’s nudity is presented in a manner that clearly panders to and titillates the audience,” id., at 3153. Unlike in the Fox case, the Commission imposed a forfeiture of $27,500 on each of the 45 ABC-affiliated stations that aired the indecent episode. In a summary order the United States Court of Appeals for the Second Circuit vacated the forfeiture order, determining that it was bound by its Fox decision striking down the entirety of the Commission’s indecency policy. See 404 Fed. Appx., at 533. The Government sought review of both judgments, see Brief for Petitioners 1, and this Court granted certiorari, 564 U. S. ____ (2011). These are the cases before us. II A fundamental principle in our legal system is that laws which regulate persons or entities must give fair notice of conduct that is forbidden or required. See Connally v. General Constr. Co., 269 U.S. 385, 391 (1926) (“[A] statute which either forbids or requires the doing of an act in terms so vague that men of common intelligence must necessarily guess at its meaning and differ as to its application, violates the first essential of due process of law”); Papachristou v. Jacksonville, 405 U.S. 156, 162 (1972) (“Living under a rule of law entails various suppositions, one of which is that ‘[all persons] are entitled to be informed as to what the State commands or forbids’ ” (quoting Lanzetta v. New Jersey, 306 U.S. 451, 453 (1939) (alteration in original))). This requirement of clarity in regulation is essential to the protections provided by the Due Process Clause of the Fifth Amendment. See United States v. Williams, 553 U.S. 285, 304 (2008). It requires the invalidation of laws that are impermissibly vague. A conviction or punishment fails to comply with due process if the statute or regulation under which it is obtained “fails to provide a person of ordinary intelligence fair notice of what is prohibited, or is so standardless that it authorizes or encourages seriously discriminatory enforcement.” Ibid. As this Court has explained, a regulation is not vague because it may at times be difficult to prove an incriminating fact but rather because it is unclear as to what fact must be proved. See id., at 306. Even when speech is not at issue, the void for vagueness doctrine addresses at least two connected but discrete due process concerns: first, that regulated parties should know what is required of them so they may act accordingly; second, precision and guidance are necessary so that those enforcing the law do not act in an arbitrary or discriminatory way. See Grayned v. City of Rockford, 408 U.S. 104, 108–109 (1972). When speech is involved, rigorous adherence to those requirements is necessary to ensure that ambiguity does not chill protected speech. These concerns are implicated here because, at the out-set, the broadcasters claim they did not have, and do not have, sufficient notice of what is proscribed. And leaving aside any concerns about facial invalidity, they contend that the lengthy procedural history set forth above shows that the broadcasters did not have fair notice of what was forbidden. Under the 2001 Guidelines in force when the broadcasts occurred, a key consideration was “ ‘whether the material dwell[ed] on or repeat[ed] at length’ ” the offending description or depiction. 613 F. 3d, at 322. In the 2004 Golden Globes Order, issued after the broadcasts, the Commission changed course and held that fleeting expletives could be a statutory violation. Fox I, 556 U. S., at 512. In the challenged orders now under review the Commission applied the new principle promulgated in the Golden Globes Order and determined fleeting expletives and a brief moment of indecency were action-ably indecent. This regulatory history, however, makes it apparent that the Commission policy in place at the time of the broadcasts gave no notice to Fox or ABC that a fleeting expletive or a brief shot of nudity could be actionably indecent; yet Fox and ABC were found to be in violation. The Commission’s lack of notice to Fox and ABC that its interpretation had changed so the fleeting moments of indecency contained in their broadcasts were a violation of §1464 as interpreted and enforced by the agency “fail[ed] to provide a person of ordinary intelligence fair notice of what is prohibited.” Williams, supra, at 304. This would be true with respect to a regulatory change this abrupt on any subject, but it is surely the case when applied to the regulations in question, regulations that touch upon “sensitive areas of basic First Amendment freedoms,” Baggett v. Bullitt, 377 U.S. 360, 372 (1964); see also Reno v. American Civil Liberties Union, 521 U.S. 844, 870–871 (1997) (“The vagueness of [a content-based regulation of speech] raises special First Amendment concerns because of its ob-vious chilling effect”). The Government raises two arguments in response, but neither is persuasive. As for the two fleeting expletives, the Government concedes that “Fox did not have reason-able notice at the time of the broadcasts that the Com-mission would consider non-repeated expletives indecent.” Brief for Petitioners 28, n. 3. The Government argues, nonetheless, that Fox “cannot establish unconstitutional vagueness on that basis . . . because the Commission did not impose a sanction where Fox lacked such notice.” Ibid. As the Court observed when the case was here three Terms ago, it is true that the Commission declined to impose any forfeiture on Fox, see 556 U. S., at 513, and in its order the Commission claimed that it would not con-sider the indecent broadcasts either when considering whether to renew stations’ licenses or “in any other context,” 21 FCC Rcd., at 13321, 13326. This “policy of forbearance,” as the Government calls it, does not suffice to make the issue moot. Brief for Petitioners 31. Though the Commission claims it will not consider the prior indecent broadcasts “in any context,” it has the statutory power to take into account “any history of prior offenses” when setting the level of a forfeiture penalty. See 47 U. S. C. §503(b)(2)(E). Just as in the First Amendment context, the due process protection against vague regulations “does not leave [regulated parties] . . . at the mercy of noblesse oblige.” United States v. Stevens, 559 U. S. ___, ___ (2010) (slip op., at 18). Given that the Commission found it was “not inequitable to hold Fox responsible for [the 2003 broadcast],” 21 FCC Rcd., at 13314, and that it has the statutory authority to use its finding to increase any future penalties, the Government’s assurance it will elect not to do so is insufficient to remedy the constitutional violation. In addition, when combined with the legal consequence described above, reputational injury provides further rea-son for granting relief to Fox. Cf. Paul v. Davis, 424 U.S. 693, 708–709 (1976) (explaining that an “alteration of legal status . . . combined with the injury resulting from the defamation” justifies the invocation of procedural safeguards). As respondent CBS points out, findings of wrongdoing can result in harm to a broadcaster’s “reputation with viewers and advertisers.” Brief for Respondent CBS Television Network Affiliates Assn. et al. 17. This observation is hardly surprising given that the challenged orders, which are contained in the permanent Commission record, describe in strongly disapproving terms the indecent material broadcast by Fox, see, e.g., 21 FCC Rcd., at 13310–13311, ¶30 (noting the “explicit, graphic, vulgar, and shocking nature of Ms. Richie’s comments”), and Fox’s efforts to protect children from being exposed to it, see id., at 13311, ¶33 (finding Fox had failed to exercise “ ‘rea-sonable judgment, responsibility, and sensitivity to the public’s needs and tastes to avoid [a] patently offensive broadcas[t]’ ”). Commission sanctions on broadcasters for indecent material are widely publicized. See, e.g., F. C. C. Fines Fox, N. Y. Times, Feb. 26, 2008, p. E2; F. C. C. Plans Record Fine for CBS, Washington Post, Sept. 24, 2004, p. E1. The challenged orders could have an adverse impact on Fox’s reputation that audiences and advertisers alike are entitled to take into account. With respect to ABC, the Government with good reason does not argue no sanction was imposed. The fine against ABC and its network affiliates for the seven seconds of nudity was nearly $1.24 million. See Brief for Respondent ABC, Inc., et al. 7 (hereinafter ABC Brief). The Government argues instead that ABC had notice that the scene in NYPD Blue would be considered indecent in light of a 1960 decision where the Commission declared that the “televising of nudes might well raise a serious question of programming contrary to 18 U. S. C. §1464.” Brief for Petitioners 32 (quoting Enbanc Programming Inquiry, 44 FCC 2303, 2307 (internal quotation marks omitted)). This argument does not prevail. An isolated and ambiguous statement from a 1960 Commission decision does not suffice for the fair notice required when the Government intends to impose over a $1 million fine for allegedly impermissible speech. The Commission, furthermore, had released decisions before sanctioning ABC that declined to find isolated and brief moments of nudity actionably indecent. See, e.g., In re Application of WGBH, 69 F. C. C. 2d, at 1251, 1255 (declining to find broadcasts contain- ing nudity to be indecent and emphasizing the difference between repeated and isolated expletives); In re WPBN/ WTOM License Subsidiary, Inc., 15 FCC Rcd. 1838, 1840 (2000) (finding full frontal nudity in Schind- ler’s List not indecent). This is not to say, of course, that a graphic scene from Schindler’s List involving nude concentration camp prisoners is the same as the shower scene from NYPD Blue. It does show, however, that the Government can point to nothing that would have given ABC affirmative notice that its broadcast would be considered actionably indecent. It is likewise not sufficient for the Commission to assert, as it did in its order, that though “the depiction [of nudity] here is not as lengthy or repeated” as in some cases, the shower scene nonetheless “does contain more shots or lengthier depictions of nudity” than in other broadcasts found not indecent. 23 FCC Rcd., at 3153. This broad language fails to demonstrate that ABC had fair notice that its broadcast could be found indecent. In fact, a Commission ruling prior to the airing of the NYPD Blue episode had deemed 30 seconds of nude buttocks “very brief” and not actionably indecent in the context of the broadcast. See Letter from Norman Goldstein to David Molina, FCC File No. 97110028 (May 26, 1999), in App. to Brief for Respondent ABC Television Affiliates Assn. et al. 1a; see also Letter from Edythe Wise to Susan Cavin, FCC File No. 91100738 (Aug. 13, 1992), id., at 18a, 19a. In light of this record of agency decisions, and the absence of any notice in the 2001 Guidance that seven seconds of nude buttocks would be found indecent, ABC lacked constitutionally sufficient notice prior to being sanctioned. The Commission failed to give Fox or ABC fair notice prior to the broadcasts in question that fleeting expletives and momentary nudity could be found actionably indecent. Therefore, the Commission’s standards as applied to these broadcasts were vague, and the Commission’s orders must be set aside. III It is necessary to make three observations about the scope of this decision. First, because the Court resolves these cases on fair notice grounds under the Due Process Clause, it need not address the First Amendment implications of the Commission’s indecency policy. It is argued that this Court’s ruling in Pacifica (and the less rigorous standard of scrutiny it provided for the regulation of broadcasters, see 438 U. S. 726) should be overruled because the rationale of that case has been overtaken by technological change and the wide availability of multiple other choices for listeners and viewers. See, e.g., ABC Brief 48–57; Brief for Respondent Fox Television Stations, Inc., et al. 15–26. The Government for its part maintains that when it licenses a conventional broadcast spectrum, the public may assume that the Government has its own interest in setting certain standards. See Brief for Petitioners 40–53. These arguments need not be addressed here. In light of the Court’s holding that the Commission’s policy failed to provide fair notice it is unnecessary to reconsider Pacifica at this time. This leads to a second observation. Here, the Court rules that Fox and ABC lacked notice at the time of their broadcasts that the material they were broadcasting could be found actionably indecent under then-existing policies. Given this disposition, it is unnecessary for the Court to address the constitutionality of the current indecency policy as expressed in the Golden Globes Order and sub-sequent adjudications. The Court adheres to its normal practice of declining to decide cases not before it. See, e.g., Sweatt v. Painter, 339 U.S. 629, 631 (1950) (“Broader issues have been urged for our consideration, but we adhere to the principle of deciding constitutional ques- tions only in the context of the particular case before the Court”). Third, this opinion leaves the Commission free to modify its current indecency policy in light of its determination of the public interest and applicable legal requirements. And it leaves the courts free to review the current policy or any modified policy in light of its content and application. * * * The judgments of the United States Court of Appeals for the Second Circuit are vacated, and the cases are re-manded for further proceedings consistent with the principles set forth in this opinion. It is so ordered. Justice Sotomayor took no part in the consideration or decision of these cases.

566.US.2011_10-1018

Respondent Delia, a firefighter employed by the City of Rialto, California, missed work after becoming ill on the job. Suspicious of Delia’s extended absence, the City hired a private investigation firm to conduct surveillance on him. When Delia was seen buying fiberglass insulation and other building supplies, the City initiated an internal affairs investigation. It hired petitioner Filarsky, a private attorney, to interview Delia. At the interview, which Delia’s attorney and two fire department officials also attended, Delia acknowledged buying the supplies, but denied having done any work on his home. To verify Delia’s claim, Filarsky asked Delia to allow a fire department official to enter his home and view the unused materials. When Delia refused, Filarsky ordered him to bring the materials out of his home for the official to see. This prompted Delia’s attorney to threaten a civil rights action against the City and Filarsky. Nonetheless, after the interview concluded, officials followed Delia to his home, where he produced the materials. Delia brought an action under 42 U. S. C. §1983 against the City, the Fire Department, Filarsky, and other individuals, alleging that the order to produce the building materials violated his Fourth and Fourteenth Amendment rights. The District Court granted summary judgment to the individual defendants on the basis of qualified immunity. The Court of Appeals for the Ninth Circuit affirmed with respect to all individual defendants except Filarsky, concluding that he was not entitled to seek qualified immunity because he was a private attorney, not a City employee. Held: A private individual temporarily retained by the government to carry out its work is entitled to seek qualified immunity from suit under §1983. Pp. 4−16. (a) In determining whether the Court of Appeals made a valid distinction between City employees and Filarsky for qualified immunity purposes, this Court looks to the general principles of tort immunities and defenses applicable at common law, and the reasons the Court has afforded protection from suit under §1983. See Imbler v. Pachtman, 424 U.S. 409, 418. The common law as it existed in 1871, when Congress enacted §1983, did not draw a distinction between full-time public servants and private individuals engaged in public service in according protection to those carrying out government responsibilities. Government at that time was smaller in both size and reach, had fewer responsibilities, and operated primarily at the local level. Government work was carried out to a significant extent by individ- uals who did not devote all their time to public duties, but instead pur- sued private callings as well. In according protection from suit to individuals doing the government’s work, the common law did not draw distinctions based on the nature of a worker’s engagement with the government. Indeed, examples of individuals receiving immunity for actions taken while engaged in public service on a temporary or occasional basis are as varied as the reach of government itself. Common law principles of immunity were incorporated into §1983 and should not be abrogated absent clear legislative intent. See Pulliam v. Allen, 466 U.S. 522, 529. Immunity under §1983 therefore should not vary depending on whether an individual working for the government does so as a permanent or full-time employee, or on some other basis. Pp. 4–11. (b) Nothing about the reasons this Court has given for recognizing immunity under §1983 counsels against carrying forward the common law rule. First, the government interest in avoiding “unwarranted timidity” on the part of those engaged in the public’s business— which has been called “the most important special government immunity-producing concern,” Richardson v. McKnight, 521 U.S. 399, 409—is equally implicated regardless of whether the individual sued as a state actor works for the government full-time or on some other basis. Second, affording immunity to those acting on the government’s behalf serves to “ ‘ensure that talented candidates [are] not deterred by the threat of damages suits from entering public service.’ ” Id., at 408. The government, in need of specialized knowledge or expertise, may look outside its permanent workforce to secure the services of private individuals. But because those individuals are free to choose other work that would not expose them to liability for government actions, the most talented candidates might decline public engagements if they did not receive the same immunity enjoyed by their public employee counterparts. Third, the public interest in ensuring performance of government duties free from the distractions that can accompany lawsuits is implicated whether those duties are discharged by private individuals or permanent government employees. Finally, distinguishing among those who carry out the public’s business based on their particular relationship with the government creates significant line-drawing problems and can deprive state actors of the ability to “ ‘reasonably anticipate when their conduct may give rise to liability for damages,’ ” Anderson v. Creighton, 483 U.S. 635, 646. Pp. 11−13. (c) This conclusion is not contrary to Wyatt v. Cole, 504 U.S. 158, or Richardson v. McKnight, 521 U.S. 399. Wyatt did not implicate the reasons underlying recognition of qualified immunity because the defendant in that case had no connection to government and pursued purely private ends. Richardson involved the unusual circumstances of prison guards employed by a private company who worked in a privately run prison facility. Nothing of the sort is involved here, or in the typical case of an individual hired by the government to assist in carrying out its work. Pp. 13−15. 621 F.3d 1069, reversed. Roberts, C. J., delivered the opinion for a unanimous Court. Ginsburg, J., and Sotomayor, J., filed concurring opinions.

Section 1983 provides a cause of action against state actors who violate an individual’s rights under federal law. 42 U. S. C. §1983. At common law, those who carried out the work of government enjoyed various protections from liability when doing so, in order to allow them to serve the government without undue fear of personal exposure. Our decisions have looked to these common law protections in affording either absolute or qualified immunity to individuals sued under §1983. The question in this case is whether an individual hired by the government to do its work is prohibited from seeking such immunity, solely be- cause he works for the government on something other than a permanent or full-time basis. I A Nicholas Delia, a firefighter employed by the City of Rialto, California, became ill while responding to a toxic spill in August 2006. Under a doctor’s orders, Delia missed three weeks of work. The City became suspicious of Delia’s extended absence, and hired a private investi- gation firm to conduct surveillance on him. The private investigators observed Delia purchasing building supplies— including several rolls of fiberglass insulation—from a home improvement store. The City surmised that Delia was missing work to do construction on his home rather than because of illness, and it initiated a formal internal affairs investigation of him. Delia was ordered to appear for an administrative in- vestigation interview. The City hired Steve Filarsky to conduct the interview. Filarsky was an experienced employ- ment lawyer who had previously represented the City in several investigations. Delia and his attorney attended the interview, along with Filarsky and two fire department officials, Mike Peel and Frank Bekker. During the interview, Filarsky questioned Delia about the building sup- plies. Delia acknowledged that he had purchased the supplies, but claimed that he had not yet done the work on his home. During a break, Filarsky met with Peel, Bekker, and Fire Chief Stephen Wells. Filarsky proposed resolving the investigation by verifying Delia’s claim that he had not done any work on his home. To do so, Filarsky recommended asking Delia to produce the building materials. Chief Wells approved the plan. When the meeting resumed, Filarsky requested permission for Peel to enter Delia’s home to view the materials. On the advice of counsel, Delia refused. Filarsky then asked Delia if he would be willing to bring the materials out onto his lawn, so that Peel could observe them without entering his home. Delia again refused to consent. Un- able to obtain Delia’s cooperation, Filarsky ordered him to produce the materials for inspection. Delia’s counsel objected to the order, asserting that it would violate the Fourth Amendment. When that objection proved unavailing, Delia’s counsel threatened to sue the City. He went on to tell Filarsky that “[w]e might quite possibly find a way to figure if we can name you Mr. Filarsky. . . . If you want to take that chance, you go right ahead.” App. 131–132. The threat was repeated over and over: “[E]verybody is going to get named, and they are going to sweat it out as to whether or not they have individual liability . . . .” “[Y]ou order him and you will be named and that is not an idle threat.” “Whoever issues that order is going to be named in the lawsuit.” “[W]e will seek any and all damages including individual liability . . . . [W]e are coming if you order this.” “[M]ake sure the spell- ing is clear [in the order] so we know who to sue.” Id., at 134–136, 148–149. Despite these threats, Filarsky prepared an order directing Delia to produce the materials, which Chief Wells signed. As soon as the interview concluded, Peel and Bekker followed Delia to his home. Once there, Delia, his attorney, and a union representative went into Delia’s house, brought out the four rolls of insulation, and placed them on Delia’s lawn. Peel and Bekker, who remained in their car during this process, thanked Delia for showing them the insulation and drove off. B Delia brought an action under 42 U. S. C. §1983 against the City, its Fire Department, Chief Wells, Peel, Bekker, Filarsky, and ten unidentified individuals, alleging that the order to produce the building materials violated his rights under the Fourth and Fourteenth Amendments. The District Court granted summary judgment to all the individual defendants, concluding that they were pro- tected by qualified immunity. The court held that Delia had “not demonstrated a violation of a clearly established constitutional right,” because “Delia was not threatened with insubordination or termination if he did not comply with any order given and none of these defendants entered [his] house.” Delia v. Rialto, No. CV 08–03359 (CD Cal., Mar. 9, 2009), App. to Pet. for Cert. 42, 48. The Court of Appeals for the Ninth Circuit affirmed with respect to all defendants except Filarsky. The Court of Appeals concluded that the order violated the Fourth Amendment, but agreed with the District Court that Delia “ha[d] not demonstrated that a constitutional right was clearly established as of the date of Chief Wells’s order, such that defendants would have known that their actions were unlawful.” Delia v. Rialto, 621 F.3d 1069, 1079 (2010). As to Filarsky, however, the court concluded that because he was a private attorney and not a City employee, he was not entitled to seek the protection of qualified immunity. Id., at 1080–1081. The court noted that its decision conflicted with a decision of the Court of Appeals for the Sixth Circuit, see Cullinan v. Abramson, 128 F.3d 301, 310 (1997), but considered itself bound by Circuit precedent and therefore “not free to follow the Cullinan decision.” 621 F. 3d, at 1080 (citing Gonzalez v. Spencer, 336 F.3d 832 (CA9 2003)). Filarsky filed a petition for certiorari, which we granted. 564 U. S. ___ (2011). II Section 1983 provides a cause of action against any person who deprives an individual of federally guaranteed rights “under color” of state law. 42 U. S. C. §1983. Anyone whose conduct is “fairly attributable to the state” can be sued as a state actor under §1983. See Lugar v. Edmondson Oil Co., 457 U.S. 922, 937 (1982). At common law, government actors were afforded certain protections from liability, based on the reasoning that “the public good can best be secured by allowing officers charged with the duty of deciding upon the rights of others, to act upon their own free, unbiased convictions, uninfluenced by any apprehensions.” Wasson v. Mitchell, 18 Iowa 153, 155–156 (1864) (internal quotation marks omitted); see also W. Prosser, Law of Torts §25, p. 150 (1941) (common law protections derived from the need to avoid the “impossible burden [that] would fall upon all our agencies of government” if those acting on behalf of the government were “unduly hampered and intimidated in the discharge of their duties” by a fear of personal liability). Our decisions have recognized similar immunities under §1983, reasoning that common law protections “ ‘well grounded in his- tory and reason’ had not been abrogated ‘by covert inclusion in the general language’ of §1983.” Imbler v. Pachtman, 424 U.S. 409, 418 (1976) (quoting Tenney v. Brandhove, 341 U.S. 367, 376 (1951)). In this case, there is no dispute that qualified immunity is available for the sort of investigative activities at issue. See Pearson v. Callahan, 555 U.S. 223, 243–244 (2009). The Court of Appeals granted this protection to Chief Wells, Peel, and Bekker, but denied it to Filarsky, because he was not a public employee but was instead a private individual “retained by the City to participate in internal affairs investigations.” 621 F. 3d, at 1079–1080. In determining whether this distinction is valid, we look to the “general principles of tort immunities and defenses” applicable at common law, and the reasons we have afforded protection from suit under §1983. Imbler, supra, at 418. A Under our precedent, the inquiry begins with the common law as it existed when Congress passed §1983 in 1871. Tower v. Glover, 467 U.S. 914, 920 (1984). Understanding the protections the common law afforded to those exercising government power in 1871 requires an appreciation of the nature of government at that time. In the mid-nineteenth century, government was smaller in both size and reach. It had fewer responsibilities, and operated primarily at the local level. Local governments faced tight budget constraints, and generally had neither the need nor the ability to maintain an established bureaucracy staffed by professionals. See B. Campbell, The Growth of American Government: Governance From the Cleveland Era to the Present 14–16, 20–21 (1995); id., at 20 (noting that in the 1880s “[t]he governor’s office staff in Wisconsin . . . totaled five workers if we count the lieutenant governor and the janitor”). As one commentator has observed, there was at that time “no very clear conception of a professional office, that is, an office the incumbent of which devotes his entire time to the discharge of public functions, who has no other occupation, and who receives a sufficiently large compensation to enable him to live without resorting to other means.” F. Goodnow, Principles of the Administrative Law of the United States 227 (1905). Instead, to a significant extent, government was “administered by members of society who temporarily or occasionally discharge[d] public functions.” Id., at 228. Whether government relied primarily upon professionals or occasional workers obviously varied across the country and across different government functions. But even at the turn of the twentieth century, a public servant was often one who “does not devote his entire time to his public duties, but is, at the same time that he is holding public office, permitted to carry on some other regular business, and as a matter of fact finds his main means of support in such business or in his private means since he receives from his office a compensation insufficient to support him.” Id., at 227. Private citizens were actively involved in government work, especially where the work most directly touched the lives of the people. It was not unusual, for example, to see the owner of the local general store step behind a window in his shop to don his postman’s hat. See, e.g., Stole Stamps, Maysville, KY, The Evening Bulletin, p. 1, Sept. 25, 1895 (reporting that “[t]he post office and general store at Mount Hope was broken into,” resulting in the loss of $400 worth of cutlery and stamps). Nor would it have been a surprise to find, on a trip to the docks, the local ferryman collecting harbor fees as public wharfmaster. See 3 E. Johnson, A History of Kentucky and Kentuckians 1346 (1912). Even such a core government activity as criminal prosecution was often carried out by a mixture of public employees and private individuals temporarily serving the public. At the time §1983 was enacted, private lawyers were regularly engaged to conduct criminal prosecutions on behalf of the State. See, e.g., Commonwealth v. Gibbs, 70 Mass. 146 (1855); White v. Polk County, 17 Iowa 413 (1864). Abraham Lincoln himself accepted several such appointments. See, e.g., An Awful Crime and Speedy Punishment, Springfield Daily Register, May 14, 1853 (reporting that “A. Lincoln, esq. was appointed prosecutor” in a rape case). In addition, private lawyers often assisted public prosecutors in significant cases. See, e.g., Commonwealth v. Knapp, 10 Mass. 477, 490–491 (1830); Chambers v. State, 22 Tenn. 237 (1842). And public prosecutors themselves continued to represent private clients while in office—sometimes creating odd conflicts of interest. See People v. Bussey, 82 Mich. 49, 46 N.W. 97, 98 (1890) (public prosecutor employed as private counsel by the defendant’s wife in several civil suits against the defendant); Phillip v. Waller, 5 Haw. 609, 617 (1886) (public prosecutor represented plaintiff in a suit for malicious prosecution); Oliver v. Pate, 43 Ind. 132, 139 (1873) (public prosecutor who conducted a state prosecution against a defendant later served as counsel for the defendant in a malicious prosecution suit against the complaining witness). This mixture of public responsibility and private pursuits extended even to the highest levels of government. Until the position became full-time in 1853, for example, the Attorney General of the United States was expected to and did maintain an active private law practice. To cite a notable illustration, in Hayburn’s Case, 2 Dall. 409 (1792), the first Attorney General, Edmund Randolph, sought a writ of mandamus from this Court to compel a lower court to hear William Hayburn’s petition to be put on the pension list. When this Court did not allow the Attorney General to seek the writ in his official capacity, Randolph readily solved the problem by arguing the case as Hayburn’s private lawyer. Ibid.; see also Letter from Edmund Randolph to James Madison (Aug. 12, 1792), reprinted in 14 The Papers of James Madison 348, 349 (R. Rutland, T. Mason, R. Brugger, J. Sisson, & F. Teute eds. 1983); Bloch, The Early Role of the Attorney General in Our Constitutional Scheme: In the Beginning There Was Pragmatism, 1989 Duke L. J. 561, 598–599, n. 121, 619. Given all this, it should come as no surprise that the common law did not draw a distinction between public servants and private individuals engaged in public service in according protection to those carrying out government responsibilities. Government actors involved in adjudicative activities, for example, were protected by an absolute immunity from suit. See Bradley v. Fisher, 13 Wall. 335, 347–348 (1872); J. Bishop, Commentaries on the Non-Contract Law §781 (1889). This immunity applied equally to “the highest judge in the State or nation” and “the lowest officer who sits as a court and tries petty causes,” T. Cooley, Law of Torts 409 (1879), including those who served as judges on a part-time or episodic basis. Justices of the peace, for example, often maintained active private law practices (or even had nonlegal livelihoods), and generally served in a judicial capacity only part-time. See Hubbell v. Harbeck, 54 Hun. 147, 7 N.Y.S. 243 (1889); Ingraham v. Leland, 19 Vt. 304 (1847). In fact, justices of the peace were not even paid a salary by the government, but instead received compensation through fees payable by the parties that came before them. See W. Murfee, The Justice of the Peace §1145 (1886). Yet the common law extended the same immunity “to a justice of the peace as to any other judicial officer.” Pratt v. Gardner, 56 Mass. 63, 70 (1848); see also Mangold v. Thorpe, 33 N. J. L. 134, 137–138 (1868). The common law also extended certain protections to individuals engaged in law enforcement activities, such as sheriffs and constables. At the time §1983 was enacted, however, “[t]he line between public and private policing was frequently hazy. Private detectives and privately em- ployed patrol personnel often were publicly appointed as special policemen, and the means and objects of detective work, in particular, made it difficult to distinguish between those on the public payroll and private detectives.” Sklansky, The Private Police, 46 UCLA L. Rev. 1165, 1210 (1999) (footnotes and internal quotation marks omitted). The protections provided by the common law did not turn on whether someone we today would call a police officer worked for the government full-time or instead for both public and private employers. Rather, at common law, “[a] special constable, duly appointed according to law, ha[d] all the powers of a regular constable so far as may be necessary for the proper discharge of the special duties intrusted to him, and in the lawful discharge of those duties, [was] as fully protected as any other officer.” W. Murfee, A Treatise on the Law of Sheriffs and Other Ministerial Officers §1121, p. 609 (1884). Sheriffs executing a warrant were empowered by the common law to enlist the aid of the able-bodied men of the community in doing so. See 1 W. Blackstone, Commentaries on the Laws of England 332 (1765); In re Quarles, 158 U.S. 532, 535 (1895). While serving as part of this “posse comitatus,” a private individual had the same authority as the sheriff, and was protected to the same extent. See, e.g., Robinson v. State, 93 Ga. 77, 18 S.E. 1018, 1019 (1893) (“A member of a posse comitatus summoned by the sheriff to aid in the execution of a warrant for a felony in the sheriff’s hands is entitled to the same protection in the discharge of his duties as the sheriff himself”); State v. Mooring, 115 N. C. 709, 20 S.E. 182 (1894) (considering it “well settled by the courts” that a sheriff may break open the doors of a house to execute a search warrant and that “if he act in good faith in doing so, both he and his posse comitatus will be protected”); North Carolina v. Gosnell, 74 F. 734, 738–739 (CC WDNC 1896) (“Both judicial and ministerial officers, in the execution of the duties of their office, are under the strong protection of the law; and their legally summoned assistants, for such time as in service, are officers of the law”); Reed v. Rice, 25 Ky. 44, 46–47 (App. 1829) (private individuals summoned by a constable to execute a search warrant were protected from a suit based on the invalidity of the warrant). Indeed, examples of individuals receiving immunity for actions taken while engaged in public service on a temporary or occasional basis are as varied as the reach of government itself. See, e.g., Gregory v. Brooks, 37 Conn. 365, 372 (1870) (public wharfmaster not liable for ordering re- moval of a vessel unless the order was issued maliciously); Henderson v. Smith, 26 W. Va. 829, 836–838 (1885) (notaries public given immunity for discretionary acts taken in good faith); Chamberlain v. Clayton, 56 Iowa 331, 9 N.W. 237 (1881) (trustees of a public institution for the disabled not liable absent a showing of malice); McCormick v. Burt, 95 Ill. 263, 265–266 (1880) (school board members not liable for suspending a student in good faith); Donohue v. Richards, 38 Me. 379, 392 (1854) (same); Downer v. Lent, 6 Cal. 94, 95 (1856) (members of a Board of Pilot Commissioners given immunity for official acts); Rail v. Potts & Baker, 27 Tenn. 225, 228–230 (1847) (private individuals appointed by the sheriff to serve as judges of an election were not liable for refusing a voter absent a showing of malice); Jenkins v. Waldron, 11 Johns. 114, 120–121 (NY Sup. Ct. 1814) (same). We read §1983 “in harmony with general principles of tort immunities and defenses.” Imbler, 424 U. S., at 418. And we “proceed[ ] on the assumption that common-law principles of . . . immunity were incorporated into our judicial system and that they should not be abrogated absent clear legislative intent to do so.” Pulliam v. Allen, 466 U.S. 522, 529 (1984). Under this assumption, immu- nity under §1983 should not vary depending on whether an individual working for the government does so as a full-time employee, or on some other basis. B Nothing about the reasons we have given for recognizing immunity under §1983 counsels against carrying forward the common law rule. As we have explained, such immu- nity “protect[s] government’s ability to perform its traditional functions.” Wyatt v. Cole, 504 U.S. 158, 167 (1992). It does so by helping to avoid “unwarranted timidity” in performance of public duties, ensuring that talented candidates are not deterred from public service, and preventing the harmful distractions from carrying out the work of government that can often accompany damages suits. Richardson v. McKnight, 521 U.S. 399, 409–411 (1997). We have called the government interest in avoiding “unwarranted timidity” on the part of those engaged in the public’s business “the most important special government immunity-producing concern.” Id., at 409. Ensuring that those who serve the government do so “with the decisiveness and the judgment required by the public good,” Scheuer v. Rhodes, 416 U.S. 232, 240 (1974), is of vital importance regardless whether the individual sued as a state actor works full-time or on some other basis. Affording immunity not only to public employees but also to others acting on behalf of the government similarly serves to “ ‘ensure that talented candidates [are] not deterred by the threat of damages suits from entering public service.’ ” Richardson, supra, at 408 (quoting Wyatt, supra, at 167). The government’s need to attract talented individuals is not limited to full-time public employees. Indeed, it is often when there is a particular need for specialized knowledge or expertise that the government must look outside its permanent work force to secure the services of private individuals. This case is a good example: Filarsky had 29 years of specialized experience as an attorney in labor, employment, and personnel matters, with particular expertise in conducting internal affairs investigations. App. to Pet. for Cert. 59, 89; App. 156. The City of Rialto certainly had no permanent employee with anything approaching those qualifications. To the extent such private individuals do not depend on the government for their livelihood, they have freedom to select other work—work that will not expose them to liability for government actions. This makes it more likely that the most talented candidates will decline public en- gagements if they do not receive the same immunity enjoyed by their public employee counterparts. Sometimes, as in this case, private individuals will work in close coordination with public employees, and face threatened legal action for the same conduct. See App. 134 (Delia’s lawyer: “everybody is going to get named” in threatened suit). Because government employees will often be protected from suit by some form of immunity, those working alongside them could be left holding the bag—facing full liability for actions taken in conjunction with government employees who enjoy immunity for the same activity. Under such circumstances, any private individual with a choice might think twice before accepting a government assignment. The public interest in ensuring performance of government duties free from the distractions that can accompany even routine lawsuits is also implicated when individuals other than permanent government employees discharge these duties. See Richardson, supra, at 411. Not only will such individuals’ performance of any ongoing government responsibilities suffer from the distraction of lawsuits, but such distractions will also often affect any public employees with whom they work by embroiling those employees in litigation. This case is again a good example: If the suit against Filarsky moves forward, it is highly likely that Chief Wells, Bekker, and Peel will all be required to tes- tify, given their roles in the dispute. Allowing suit under §1983 against private individuals assisting the government will substantially undermine an important reason immunity is accorded public employees in the first place. Distinguishing among those who carry out the public’s business based on the nature of their particular relationship with the government also creates significant line-drawing problems. It is unclear, for example, how Fil- arsky would be categorized if he regularly spent half his time working for the City, or worked exclusively on one City project for an entire year. See Tr. of Oral Arg. 34–36. Such questions deprive state actors of the ability to “reasonably anticipate when their conduct may give rise to liability for damages,” Anderson v. Creighton, 483 U.S. 635, 646 (1987) (alteration and internal quotation marks omitted), frustrating the purposes immunity is meant to serve. An uncertain immunity is little better than no immunity at all. III Our decisions in Wyatt v. Cole, 504 U.S. 158 (1992), and Richardson v. McKnight, 521 U.S. 399 (1997), are not to the contrary. In Wyatt, we held that individuals who used a state replevin law to compel the local sheriff to seize disputed property from a former business partner were not entitled to seek qualified immunity. Cf. Lugar, 457 U.S. 922 (holding that an individual who uses a state replevin, garnishment, or attachment statute later declared to be unconstitutional acts under color of state law for purposes of §1983). We explained that the reasons underlying recognition of qualified immunity did not sup- port its extension to individuals who had no connection to government and pursued purely private ends. Because such individuals “hold no office requiring them to exercise discretion; nor are they principally concerned with enhancing the public good,” we concluded that extending immunity to them would “have no bearing on whether public officials are able to act forcefully and decisively in their jobs or on whether qualified applicants enter public service.” 504 U. S., at 168. Wyatt is plainly not implicated by the circumstances of this case. Unlike the defendants in Wyatt, who were us- ing the mechanisms of government to achieve their own ends, individuals working for the government in pursuit of government objectives are “principally concerned with en- hancing the public good.” Ibid. Whether such individ- uals have assurance that they will be able to seek protection if sued under §1983 directly affects the government’s ability to achieve its objectives through their public service. Put simply, Wyatt involved no government agents, no government interests, and no government need for immunity. In Richardson, we considered whether guards employed by a privately run prison facility could seek the protection of qualified immunity. Although the Court had previously determined that public-employee prison guards were entitled to qualified immunity, see Procunier v. Navarette, 434 U.S. 555 (1978), it determined that prison guards employed by a private company and working in a privately run prison facility did not enjoy the same protection. We explained that the various incentives characteristic of the private market in that case ensured that the guards would not perform their public duties with unwarranted timidity or be deterred from entering that line of work. 521 U. S., at 410–411. Richardson was a self-consciously “narrow[ ]” decision. Id., at 413 (“[W]e have answered the immunity question narrowly, in the context in which it arose”). The Court made clear that its holding was not meant to foreclose all claims of immunity by private individuals. Ibid. Instead, the Court emphasized that the particular circumstances of that case—“a private firm, systematically organized to assume a major lengthy administrative task (managing an institution) with limited direct supervision by the government, undertak[ing] that task for profit and potentially in competition with other firms”—combined sufficiently to mitigate the concerns underlying recognition of governmental immunity under §1983. Ibid. Nothing of the sort is involved here, or in the typical case of an individual hired by the government to assist in carrying out its work. * * * A straightforward application of the rule set out above is sufficient to resolve this case. Though not a public em- ployee, Filarsky was retained by the City to assist in conducting an official investigation into potential wrongdoing. There is no dispute that government employees performing such work are entitled to seek the protection of qualified immunity. The Court of Appeals rejected Filarsky’s claim to the protection accorded Wells, Bekker, and Peel solely because he was not a permanent, full-time employee of the City. The common law, however, did not draw such distinctions, and we see no justification for doing so under §1983. New York City has a Department of Investigation staffed by full-time public employees who investigate city personnel, and the resources to pay for it. The City of Rialto has neither, and so must rely on the occasional services of private individuals such as Mr. Filarsky. There is no reason Rialto’s internal affairs investigator should be denied the qualified immunity enjoyed by the ones who work for New York. In light of the foregoing, the judgment of the Court of Appeals denying qualified immunity to Filarsky is reversed. It is so ordered.

566.US.2011_10-945

Petitioner was arrested during a traffic stop by a New Jersey state trooper who checked a statewide computer database and found a bench warrant issued for petitioner’s arrest after he failed to appear at a hearing to enforce a fine. He was initially detained in the Burlington County Detention Center and later in the Essex County Correctional Facility, but was released once it was determined that the fine had been paid. At the first jail, petitioner, like every incoming detainee, had to shower with a delousing agent and was checked for scars, marks, gang tattoos, and contraband as he disrobed. Petitioner claims that he also had to open his mouth, lift his tongue, hold out his arms, turn around, and lift his genitals. At the second jail, petitioner, like other arriving detainees, had to remove his clothing while an officer looked for body markings, wounds, and contraband; had an officer look at his ears, nose, mouth, hair, scalp, fingers, hands, armpits, and other body openings; had a mandatory shower; and had his clothes examined. Petitioner claims that he was also required to lift his genitals, turn around, and cough while squatting. He filed a 42 U. S. C. §1983 action in the Federal District Court against the government entities that ran the jails and other defendants, alleging Fourth and Fourteenth Amendment violations, and arguing that persons arrested for minor offenses cannot be subjected to invasive searches unless prison officials have reason to suspect concealment of weapons, drugs, or other contraband. The court granted him summary judgment, ruling that “strip-searching” nonindictable offenders without reasonable suspicion violates the Fourth Amendment. The Third Circuit reversed. Held: The judgment is affirmed. 621 F.3d 296, affirmed. Justice Kennedy delivered the opinion of the Court, except as to Part IV, concluding that the search procedures at the county jails struck a reasonable balance between inmate privacy and the needs of the institutions, and thus the Fourth and Fourteenth Amendments do not require adoption of the framework and rules petitioner pro- poses. Pp. 5−18, 19. (a) Maintaining safety and order at detention centers requires the expertise of correctional officials, who must have substantial discretion to devise reasonable solutions to problems. A regulation impinging on an inmate’s constitutional rights must be upheld “if it is reasonably related to legitimate penological interests.” Turner v. Safley, 482 U.S. 78, 89. This Court, in Bell v. Wolfish, 441 U.S. 520, 558, upheld a rule requiring pretrial detainees in federal correctional facilities “to expose their body cavities for visual inspection as a part of a strip search conducted after every contact visit with a person from outside the institution[s],” deferring to the judgment of correctional officials that the inspections served not only to discover but also to deter the smuggling of weapons, drugs, and other prohibited items. In Block v. Rutherford, 468 U.S. 576, 586−587, the Court upheld a general ban on contact visits in a county jail, noting the smuggling threat posed by such visits and the difficulty of carving out exceptions for certain detainees. The Court, in Hudson v. Palmer, 468 U.S. 517, 522−523, also recognized that deterring the possession of contraband depends in part on the ability to conduct searches without predictable exceptions when it upheld the constitutionality of random searches of inmate lockers and cells even without suspicion that an inmate is concealing a prohibited item. These cases establish that correctional officials must be permitted to devise reasonable search policies to detect and deter the possession of contraband in their facilities, and that “in the absence of substantial evidence in the record to indicate that the officials have exaggerated their response to these considerations courts should ordinarily defer to their expert judgment in such matters,” Block, supra, at 584–585. Persons arrested for minor offenses may be among the detainees to be processed at jails. See Atwater v. Lago Vista, 532 U.S. 318, 354. Pp. 5−9. (b) The question here is whether undoubted security imperatives involved in jail supervision override the assertion that some detainees must be exempt from the invasive search procedures at issue absent reasonable suspicion of a concealed weapon or other contraband. Correctional officials have a significant interest in conducting a thorough search as a standard part of the intake process. The admission of new inmates creates risks for staff, the existing detainee population, and the new detainees themselves. Officials therefore must screen for contagious infections and for wounds or injuries requiring immediate medical attention. It may be difficult to identify and treat medical problems until detainees remove their clothes for a visual inspection. Jails and prisons also face potential gang violence, giving them reasonable justification for a visual inspection of detainees for signs of gang affiliation as part of the intake process. Additionally, correctional officials have to detect weapons, drugs, alcohol, and other prohibited items new detainees may possess. Drugs can make inmates aggressive toward officers or each other, and drug trading can lead to violent confrontations. Contraband has value in a jail’s culture and underground economy, and competition for scarce goods can lead to violence, extortion, and disorder. Pp. 9−13. (c) Petitioner’s proposal―that new detainees not arrested for serious crimes or for offenses involving weapons or drugs be exempt from invasive searches unless they give officers a particular reason to suspect them of hiding contraband―is unworkable. The seriousness of an offense is a poor predictor of who has contraband, and it would be difficult to determine whether individual detainees fall within the proposed exemption. Even persons arrested for a minor offense may be coerced by others into concealing contraband. Exempting people arrested for minor offenses from a standard search protocol thus may put them at greater risk and result in more contraband being brought into the detention facility. It also may be difficult to classify inmates by their current and prior offenses before the intake search. Jail officials know little at the outset about an arrestee, who may be carrying a false ID or lie about his identity. The officers conducting an initial search often do not have access to criminal history records. And those records can be inaccurate or incomplete. Even with accurate information, officers would encounter serious implementation difficulties. They would be required to determine quickly whether any underlying offenses were serious enough to authorize the more invasive search protocol. Other possible classifications based on characteristics of individual detainees also might prove to be unworkable or even give rise to charges of discriminatory application. To avoid liability, officers might be inclined not to conduct a thorough search in any close case, thus creating unnecessary risk for the entire jail population. While the restrictions petitioner suggests would limit the intrusion on the privacy of some detainees, it would be at the risk of increased danger to everyone in the facility, including the less serious offenders. The Fourth and Fourteenth Amendments do not require adoption of the proposed framework. Pp. 13−18, 19. Kennedy, J., delivered the opinion of the Court, except as to Part IV. Roberts, C. J., and Scalia and Alito, JJ., joined that opinion in full, and Thomas, J., joined as to all but Part IV. Roberts, C. J., and Alito, J., filed concurring opinions. Breyer, J., filed a dissenting opinion, in which Ginsburg, Sotomayor, and Kagan, JJ., joined.

except as to Part IV.[1] Correctional officials have a legitimate interest, indeed a responsibility, to ensure that jails are not made less secure by reason of what new detainees may carry in on their bodies. Facility personnel, other inmates, and the new detainee himself or herself may be in danger if these threats are introduced into the jail population. This case presents the question of what rules, or limitations, the Constitution imposes on searches of arrested persons who are to be held in jail while their cases are being processed. The term “jail” is used here in a broad sense to include prisons and other detention facilities. The specific measures being challenged will be described in more detail; but, in broad terms, the controversy concerns whether every detainee who will be admitted to the general population may be required to undergo a close visual inspection while undressed. The case turns in part on the extent to which this Court has sufficient expertise and information in the record to mandate, under the Constitution, the specific restrictions and limitations sought by those who challenge the visual search procedures at issue. In addressing this type of constitutional claim courts must defer to the judgment of correctional officials unless the record contains substantial evidence showing their policies are an unnecessary or un- justified response to problems of jail security. That necessary showing has not been made in this case. I In 1998, seven years before the incidents at issue, petitioner Albert Florence was arrested after fleeing from police officers in Essex County, New Jersey. He was charged with obstruction of justice and use of a deadly weapon. Petitioner entered a plea of guilty to two lesser offenses and was sentenced to pay a fine in monthly installments. In 2003, after he fell behind on his payments and failed to appear at an enforcement hearing, a bench warrant was issued for his arrest. He paid the outstanding balance less than a week later; but, for some unexplained reason, the warrant remained in a statewide computer database. Two years later, in Burlington County, New Jersey, petitioner and his wife were stopped in their automobile by a state trooper. Based on the outstanding warrant in the computer system, the officer arrested petitioner and took him to the Burlington County Detention Center. He was held there for six days and then was transferred to the Essex County Correctional Facility. It is not the arrest or confinement but the search process at each jail that gives rise to the claims before the Court. Burlington County jail procedures required every arrestee to shower with a delousing agent. Officers would check arrestees for scars, marks, gang tattoos, and contraband as they disrobed. App. to Pet. for Cert. 53a–56a. Petitioner claims he was also instructed to open his mouth, lift his tongue, hold out his arms, turn around, and lift his genitals. (It is not clear whether this last step was part of the normal practice. See ibid.) Petitioner shared a cell with at least one other person and interacted with other inmates following his admission to the jail. Tr. of Oral Arg. 17. The Essex County Correctional Facility, where peti- tioner was taken after six days, is the largest county jail in New Jersey. App. 70a. It admits more than 25,000 in- mates each year and houses about 1,000 gang members at any given time. When petitioner was transferred there, all arriving detainees passed through a metal detector and waited in a group holding cell for a more thorough search. When they left the holding cell, they were instructed to remove their clothing while an officer looked for body markings, wounds, and contraband. Apparently without touching the detainees, an officer looked at their ears, nose, mouth, hair, scalp, fingers, hands, arms, armpits, and other body openings. Id., at 57a–59a; App. to Pet. for Cert. 137a–144a. This policy applied regardless of the circumstances of the arrest, the suspected offense, or the detainee’s behavior, demeanor, or criminal history. Petitioner alleges he was required to lift his genitals, turn around, and cough in a squatting position as part of the process. After a mandatory shower, during which his clothes were inspected, petitioner was admitted to the facility. App. 3a–4a, 52a, 258a. He was released the next day, when the charges against him were dismissed. Petitioner sued the governmental entities that operated the jails, one of the wardens, and certain other defendants. The suit was commenced in the United States District Court for the District of New Jersey. Seeking relief under 42 U. S. C. §1983 for violations of his Fourth and Fourteenth Amendment rights, petitioner maintained that per- sons arrested for a minor offense could not be required to remove their clothing and expose the most private areas of their bodies to close visual inspection as a routine part of the intake process. Rather, he contended, officials could conduct this kind of search only if they had reason to suspect a particular inmate of concealing a weapon, drugs, or other contraband. The District Court certified a class of individuals who were charged with a nonindictable offense under New Jersey law, processed at either the Burlington County or Essex County jail, and directed to strip naked even though an officer had not articulated any reasonable suspicion they were concealing contraband. After discovery, the court granted petitioner’s motion for summary judgment on the unlawful search claim. It concluded that any policy of “strip searching” nonindict- able offenders without reasonable suspicion violated the Fourth Amendment. A divided panel of the United States Court of Appeals for the Third Circuit reversed, holding that the procedures described by the District Court struck a reasonable balance between inmate privacy and the security needs of the two jails. 621 F.3d 296 (2010). The case proceeds on the understanding that the officers searched detainees prior to their admission to the general population, as the Court of Appeals seems to have assumed. See id., at 298, 311. Petitioner has not argued this factual premise is incorrect. The opinions in earlier proceedings, the briefs on file, and some cases of this Court refer to a “strip search.” The term is imprecise. It may refer simply to the instruction to remove clothing while an officer observes from a distance of, say, five feet or more; it may mean a visual inspection from a closer, more uncomfortable distance; it may include directing detainees to shake their heads or to run their hands through their hair to dislodge what might be hidden there; or it may involve instructions to raise arms, to display foot insteps, to expose the back of the ears, to move or spread the buttocks or genital areas, or to cough in a squatting position. In the instant case, the term does not include any touching of unclothed areas by the inspecting officer. There are no allegations that the detainees here were touched in any way as part of the searches. The Federal Courts of Appeals have come to differing conclusions as to whether the Fourth Amendment requires correctional officials to exempt some detainees who will be admitted to a jail’s general population from the searches here at issue. This Court granted certiorari to address the question. 563 U. S. ___ (2011). II The difficulties of operating a detention center must not be underestimated by the courts. Turner v. Safley, 482 U.S. 78, 84–85 (1987). Jails (in the stricter sense of the term, excluding prison facilities) admit more than 13 million inmates a year. See, e.g., Dept. of Justice, Bureau of Justice Statistics, T. Minton, Jail Inmates at Midyear 2010—Statistical Tables 2 (2011). The largest facilities process hundreds of people every day; smaller jails may be crowded on weekend nights, after a large police operation, or because of detainees arriving from other jurisdictions. Maintaining safety and order at these institutions requires the expertise of correctional officials, who must have substantial discretion to devise reasonable solutions to the problems they face. The Court has confirmed the importance of deference to correctional officials and explained that a regulation impinging on an inmate’s constitutional rights must be upheld “if it is reasonably related to legitimate penological interests.” Turner, supra, at 89; see Overton v. Bazzetta, 539 U.S. 126, 131–132 (2003). But see Johnson v. California, 543 U.S. 499, 510–511 (2005) (applying strict scrutiny to racial classifications). The Court’s opinion in Bell v. Wolfish, 441 U.S. 520 (1979), is the starting point for understanding how this framework applies to Fourth Amendment challenges. That case addressed a rule requiring pretrial detainees in any correctional facility run by the Federal Bureau of Prisons “to expose their body cavities for visual inspection as a part of a strip search conducted after every contact visit with a person from outside the institution.” Id., at 558. Inmates at the federal Metropolitan Correctional Center in New York City argued there was no security justification for these searches. Officers searched guests before they entered the visiting room, and the inmates were under constant surveillance during the visit. Id., at 577–578 (Marshall, J., dissenting). There had been but one instance in which an inmate attempted to sneak contraband back into the facility. See id., at 559 (majority opinion). The Court nonetheless upheld the search policy. It deferred to the judgment of correctional officials that the inspections served not only to discover but also to deter the smuggling of weapons, drugs, and other prohibited items inside. Id., at 558. The Court explained that there is no mechanical way to determine whether intrusions on an inmate’s privacy are reasonable. Id., at 559. The need for a particular search must be balanced against the resulting invasion of personal rights. Ibid. Policies designed to keep contraband out of jails and prisons have been upheld in cases decided since Bell. In Block v. Rutherford, 468 U.S. 576 (1984), for example, the Court concluded that the Los Angeles County Jail could ban all contact visits because of the threat they posed: “They open the institution to the introduction of drugs, weapons, and other contraband. Visitors can easily conceal guns, knives, drugs, or other contraband in countless ways and pass them to an inmate unnoticed by even the most vigilant observers. And these items can readily be slipped from the clothing of an innocent child, or transferred by other visitors permitted close contact with inmates.” Id., at 586. There were “many justifications” for imposing a general ban rather than trying to carve out exceptions for certain detainees. Id., at 587. Among other problems, it would be “a difficult if not impossible task” to identify “inmates who have propensities for violence, escape, or drug smuggling.” Ibid. This was made “even more difficult by the brevity of detention and the constantly changing nature of the inmate population.” Ibid. The Court has also recognized that deterring the possession of contraband depends in part on the ability to conduct searches without predictable exceptions. In Hudson v. Palmer, 468 U.S. 517 (1984), it addressed the question of whether prison officials could perform random searches of inmate lockers and cells even without reason to suspect a particular individual of concealing a prohibited item. Id., at 522–523. The Court upheld the constitutionality of the practice, recognizing that “ ‘[f]or one to advocate that prison searches must be conducted only pursuant to an enunciated general policy or when suspicion is directed at a particular inmate is to ignore the realities of prison operation.’ ” Id., at 529 (quoting Marrero v. Commonwealth, 222 Va. 754, 757, 284 S.E.2d 809, 811 (1981)). Inmates would adapt to any pattern or loopholes they discovered in the search protocol and then undermine the security of the institution. 468 U. S., at 529. These cases establish that correctional officials must be permitted to devise reasonable search policies to detect and deter the possession of contraband in their facilities. See Bell, 441 U. S., at 546 (“[M]aintaining institutional security and preserving internal order and discipline are essential goals that may require limitation or retraction of retained constitutional rights of both convicted prisoners and pretrial detainees”). The task of determining whether a policy is reasonably related to legitimate security interests is “peculiarly within the province and professional expertise of corrections officials.” Id., at 548. This Court has repeated the admonition that, “ ‘in the absence of substantial evidence in the record to indicate that the officials have exaggerated their response to these considerations courts should ordinarily defer to their expert judgment in such matters.’ ” Block, supra, at 584–585; Bell, supra, at 548. In many jails officials seek to improve security by requiring some kind of strip search of everyone who is to be detained. These procedures have been used in different places throughout the country, from Cranston, Rhode Island, to Sapulpa, Oklahoma, to Idaho Falls, Idaho. See Roberts v. Rhode Island, 239 F.3d 107, 108–109 (CA1 2001); Chapman v. Nichols, 989 F.2d 393, 394 (CA10 1993); Giles v. Ackerman, 746 F.2d 614, 615 (CA9 1984) (per curiam); see also, e.g., Bull v. City and Cty. of San Francisco, 595 F.3d 964 (CA9 2010) (en banc) (San Francisco, California); Powell v. Barrett, 541 F.3d 1298 (CA11 2008) (en banc) (Fulton Cty., Ga.); Masters v. Crouch, 872 F.2d 1248, 1251 (CA6 1989) (Jefferson Cty., Ky.); Weber v. Dell, 804 F.2d 796, 797–798 (CA2 1986) (Monroe Cty., N. Y.); Stewart v. Lubbock Cty., 767 F.2d 153, 154 (CA5 1985) (Lubbock Cty., Tex.). Persons arrested for minor offenses may be among the detainees processed at these facilities. This is, in part, a consequence of the exercise of state authority that was the subject of Atwater v. Lago Vista, 532 U.S. 318 (2001). Atwater addressed the perhaps more fundamental question of who may be deprived of liberty and taken to jail in the first place. The case involved a woman who was arrested after a police officer noticed neither she nor her children were wearing their seatbelts. The arrestee argued the Fourth Amendment prohibited her custodial arrest without a warrant when an offense could not result in jail time and there was no compelling need for immediate detention. Id., at 346. The Court held that a Fourth Amendment restriction on this power would put officers in an “almost impossible spot.” Id., at 350. Their ability to arrest a suspect would depend in some cases on the precise weight of drugs in his pocket, whether he was a repeat offender, and the scope of what counted as a compelling need to detain someone. Id., at 348–349. The Court rejected the proposition that the Fourth Amendment barred custodial arrests in a set of these cases as a matter of constitutional law. It ruled, based on established principles, that officers may make an arrest based upon probable cause to believe the person has committed a criminal offense in their presence. See id., at 354. The Court stated that “a responsible Fourth Amendment balance is not well served by standards requiring sensitive, case-by-case determinations of government need, lest every discretionary judgment in the field be converted into an occasion for constitutional review.” Id., at 347. Atwater did not address whether the Constitution imposes special restrictions on the searches of offenders suspected of committing minor offenses once they are taken to jail. Some Federal Courts of Appeals have held that corrections officials may not conduct a strip search of these detainees, even if no touching is involved, absent reasonable suspicion of concealed contraband. 621 F. 3d, at 303–304, and n. 4. The Courts of Appeals to address this issue in the last decade, however, have come to the opposite conclusion. See 621 F.3d 296 (case below); Bame v. Dillard, 637 F.3d 380 (CADC 2011); Powell, supra; Bull, supra. The current case is set against this precedent and governed by the principles announced in Turner and Bell. III The question here is whether undoubted security imperatives involved in jail supervision override the assertion that some detainees must be exempt from the more invasive search procedures at issue absent reasonable suspicion of a concealed weapon or other contraband. The Court has held that deference must be given to the officials in charge of the jail unless there is “substantial evidence” demonstrating their response to the situation is exaggerated. Block, 468 U. S., at 584–585 (internal quotation marks omitted). Petitioner has not met this standard, and the record provides full justifications for the procedures used. A Correctional officials have a significant interest in conducting a thorough search as a standard part of the intake process. The admission of inmates creates numerous risks for facility staff, for the existing detainee population, and for a new detainee himself or herself. The danger of introducing lice or contagious infections, for example, is well documented. See, e.g., Deger & Quick, The Enduring Menace of MRSA: Incidence, Treatment, and Prevention in a County Jail, 15 J. Correctional Health Care 174, 174–175, 177–178 (2009); Bick, Infection Control in Jails and Prisons, 45 Healthcare Epidemiology 1047, 1049 (2007). The Federal Bureau of Prisons recommends that staff screen new detainees for these conditions. See Clinical Practice Guidelines, Management of Methicillin-Resistant Staphylococcus aureus (MRSA) Infections 2 (2011); Clinical Practice Guidelines, Lice and Scabies Protocol 1 (2011). Persons just arrested may have wounds or other injuries requiring immediate medical attention. It may be difficult to identify and treat these problems until detainees remove their clothes for a visual inspection. See Prison and Jail Administration: Practice and Theory 142 (P. Carlson & G. Garrett eds., 2d ed. 2008) (hereinafter Carlson & Garrett). Jails and prisons also face grave threats posed by the increasing number of gang members who go through the intake process. See Brief for Policemen’s Benevolent As- sociation, Local 249, et al. as Amici Curiae 14 (hereinaf- ter PBA Brief); New Jersey Comm’n of Investigation, Gangland Behind Bars: How and Why Organized Criminal Street Gangs Thrive in New Jersey’s Prisons . . . And What Can Be Done About It 10–11 (2009). “Gang rivalries spawn a climate of tension, violence, and coercion.” Carlson & Garrett 462. The groups recruit new members by force, engage in assaults against staff, and give other inmates a reason to arm themselves. Ibid. Fights among feuding gangs can be deadly, and the officers who must maintain order are put in harm’s way. PBA Brief 17. These considerations provide a reasonable basis to justify a visual inspection for certain tattoos and other signs of gang affiliation as part of the intake process. The identi- fication and isolation of gang members before they are admitted protects everyone in the facility. Cf. Fraise v. Terhune, 283 F.3d 506, 509–510 (CA3 2002) (Alito, J.) (describing a statewide policy authorizing the identification and isolation of gang members in prison). Detecting contraband concealed by new detainees, furthermore, is a most serious responsibility. Weapons, drugs, and alcohol all disrupt the safe operation of a jail. Cf. Hudson, 468 U. S., at 528 (recognizing “the constant fight against the proliferation of knives and guns, illicit drugs, and other contraband”). Correctional officers have had to confront arrestees concealing knives, scissors, razor blades, glass shards, and other prohibited items on their person, including in their body cavities. See Bull, 595 F. 3d, at 967, 969; Brief for New Jersey County Jail Wardens Association as Amicus Curiae 17–18 (hereinafter New Jersey Wardens Brief). They have also found crack, heroin, and marijuana. Brief for City and County of San Francisco et al. as Amici Curiae 9–11 (hereinafter San Francisco Brief). The use of drugs can embolden inmates in aggression toward officers or each other; and, even apart from their use, the trade in these substances can lead to violent confrontations. See PBA Brief 11. There are many other kinds of contraband. The textbook definition of the term covers any unauthorized item. See Prisons: Today and Tomorrow 237 (J. Pollock ed. 1997) (“Contraband is any item that is possessed in violation of prison rules. Contraband obviously includes drugs or weapons, but it can also be money, cigarettes, or even some types of clothing”). Everyday items can undermine security if introduced into a detention facility: “Lighters and matches are fire and arson risks or potential weapons. Cell phones are used to orchestrate violence and criminality both within and without jailhouse walls. Pills and medications enhance suicide risks. Chewing gum can block locking devices; hairpins can open handcuffs; wigs can conceal drugs and weapons.” New Jersey Wardens Brief 8–9. Something as simple as an overlooked pen can pose a significant danger. Inmates commit more than 10,000 assaults on correctional staff every year and many more among themselves. See Dept. of Justice, Bureau of Justice Statistics, J. Stephan & J. Karberg, Census of State and Federal Correctional Facilities, 2000, p. v (2003). Contraband creates additional problems because scarce items, including currency, have value in a jail’s culture and underground economy. Correctional officials inform us “[t]he competition . . . for such goods begets violence, extortion, and disorder.” New Jersey Wardens Brief 2. Gangs exacerbate the problem. They “orchestrate thefts, commit assaults, and approach inmates in packs to take the contraband from the weak.” Id., at 9–10. This puts the entire facility, including detainees being held for a brief term for a minor offense, at risk. Gangs do coerce inmates who have access to the outside world, such as people serving their time on the weekends, to sneak things into the jail. Id., at 10; see, e.g., Pugmire, Vegas Suspect Has Term to Serve, Los Angeles Times, Sept. 23, 2005, p. B1 (“Weekend-only jail sentences are a common punishment for people convicted of nonviolent drug crimes . . .”). These inmates, who might be thought to pose the least risk, have been caught smuggling prohibited items into jail. See New Jersey Wardens Brief 10. Concealing contraband often takes little time and effort. It might be done as an officer approaches a suspect’s car or during a brief commotion in a group holding cell. Something small might be tucked or taped under an armpit, behind an ear, between the buttocks, in the instep of a foot, or inside the mouth or some other body cavity. It is not surprising that correctional officials have sought to perform thorough searches at intake for disease, gang affiliation, and contraband. Jails are often crowded, unsanitary, and dangerous places. There is a substantial interest in preventing any new inmate, either of his own will or as a result of coercion, from putting all who live or work at these institutions at even greater risk when he is admitted to the general population. B Petitioner acknowledges that correctional officials must be allowed to conduct an effective search during the intake process and that this will require at least some detainees to lift their genitals or cough in a squatting position. These procedures, similar to the ones upheld in Bell, are designed to uncover contraband that can go undetected by a patdown, metal detector, and other less invasive searches. See Brief for United States as Amicus Curiae 23 (hereinafter United States Brief); New Jersey Wardens Brief 19, n. 6. Petitioner maintains there is little benefit to conducting these more invasive steps on a new detainee who has not been arrested for a serious crime or for any offense involving a weapon or drugs. In his view these de- tainees should be exempt from this process unless they give officers a particular reason to suspect them of hiding contraband. It is reasonable, however, for correctional officials to conclude this standard would be unworkable. The record provides evidence that the seriousness of an offense is a poor predictor of who has contraband and that it would be difficult in practice to determine whether individual detainees fall within the proposed exemption. 1 People detained for minor offenses can turn out to be the most devious and dangerous criminals. Cf. Clements v. Logan, 454 U.S. 1304, 1305 (1981) (Rehnquist, J., in chambers) (deputy at a detention center shot by misdemeanant who had not been strip searched). Hours after the Oklahoma City bombing, Timothy McVeigh was stopped by a state trooper who noticed he was driving without a license plate. Johnston, Suspect Won’t Answer Any Questions, N. Y. Times, Apr. 25, 1995, p. A1. Police stopped serial killer Joel Rifkin for the same reason. McQuiston, Confession Used to Portray Rifkin as Methodical Killer, N. Y. Times, Apr. 26, 1994, p. B6. One of the terrorists involved in the September 11 attacks was stopped and ticketed for speeding just two days before hijacking Flight 93. The Terrorists: Hijacker Got a Speeding Ticket, N. Y. Times, Jan. 8, 2002, p. A12. Reasonable correctional officials could conclude these uncertainties mean they must conduct the same thorough search of everyone who will be admitted to their facilities. Experience shows that people arrested for minor of- fenses have tried to smuggle prohibited items into jail, sometimes by using their rectal cavities or genitals for the concealment. They may have some of the same incentives as a serious criminal to hide contraband. A detainee might risk carrying cash, cigarettes, or a penknife to survive in jail. Others may make a quick decision to hide unlawful substances to avoid getting in more trouble at the time of their arrest. This record has concrete examples. Officers at the Atlantic County Correctional Facility, for example, discovered that a man arrested for driving under the influence had “2 dime bags of weed, 1 pack of rolling papers, 20 matches, and 5 sleeping pills” taped under his scrotum. Brief for Atlantic County et al. as Amici Curiae 36 (internal quotation marks omitted). A person booked on a misdemeanor charge of disorderly conduct in Washington State managed to hide a lighter, tobacco, tattoo needles, and other prohibited items in his rectal cavity. See United States Brief 25, n. 15. San Francisco officials have discovered contraband hidden in body cavities of people arrested for trespassing, public nuisance, and shoplifting. San Francisco Brief 3. There have been similar incidents at jails throughout the country. See United States Brief 25, n. 15. Even if people arrested for a minor offense do not themselves wish to introduce contraband into a jail, they may be coerced into doing so by others. See New Jersey Wardens Brief 16; cf. Block, 468 U. S., at 587 (“It is not unreasonable to assume, for instance, that low security risk detainees would be enlisted to help obtain contraband or weapons by their fellow inmates who are denied contact visits”). This could happen any time detainees are held in the same area, including in a van on the way to the station or in the holding cell of the jail. If, for example, a person arrested and detained for unpaid traffic citations is not subject to the same search as others, this will be well known to other detainees with jail experience. A hardened criminal or gang member can, in just a few minutes, approach the person and coerce him into hiding the fruits of a crime, a weapon, or some other contraband. As an expert in this case explained, “the interaction and mingling between misdemeanants and felons will only increase the amount of contraband in the facility if the jail can only conduct admission searches on felons.” App. 381a. Exempting people arrested for minor offenses from a standard search protocol thus may put them at greater risk and result in more contraband being brought into the detention facility. This is a substantial reason not to mandate the exception petitioner seeks as a matter of constitutional law. 2 It also may be difficult, as a practical matter, to classify inmates by their current and prior offenses before the intake search. Jails can be even more dangerous than prisons because officials there know so little about the people they admit at the outset. See New Jersey Wardens Brief 11–14. An arrestee may be carrying a false ID or lie about his identity. The officers who conduct an initial search often do not have access to criminal history records. See, e.g., App. 235a; New Jersey Wardens Brief 13. And those records can be inaccurate or incomplete. See Department of Justice v. Reporters Comm. for Freedom of Press, 489 U.S. 749, 752 (1989). Petitioner’s rap sheet is an example. It did not reflect his previous arrest for possession of a deadly weapon. Tr. of Oral Arg. 18–19. In the absence of reliable information it would be illogical to require officers to assume the arrestees in front of them do not pose a risk of smuggling something into the facility. The laborious administration of prisons would become less effective, and likely less fair and evenhanded, were the practical problems inevitable from the rules suggested by petitioner to be imposed as a constitutional mandate. Even if they had accurate information about a detainee’s current and prior arrests, officers, under petitioner’s proposed regime, would encounter serious implementation difficulties. They would be required, in a few minutes, to determine whether any of the underlying offenses were serious enough to authorize the more invasive search protocol. Other possible classifications based on characteristics of individual detainees also might prove to be unworkable or even give rise to charges of discriminatory application. Most officers would not be well equipped to make any of these legal determinations during the pressures of the intake process. Bull, 595 F. 3d, at 985–987 (Kozinski, C. J., concurring); see also Welsh v. Wisconsin, 466 U.S. 740, 761–762 (1984) (White, J., dissenting) (“[T]he Court’s approach will necessitate a case-by-case evaluation of the seriousness of particular crimes, a dif- ficult task for which officers and courts are poorly equipped”). To avoid liability, officers might be inclined not to conduct a thorough search in any close case, thus creating unnecessary risk for the entire jail population. Cf. Atwater, 532 U. S., at 351, and n. 22. The Court addressed an analogous problem in Atwater. The petitioner in that case argued the Fourth Amendment prohibited a warrantless arrest when being convicted of the suspected crime “could not ultimately carry any jail time” and there was “no compelling need for immediate detention.” Id., at 346. That rule “promise[d] very little in the way of administrability.” Id., at 350. Officers could not be expected to draw the proposed lines on a moment’s notice, and the risk of violating the Constitution would have discouraged them from arresting criminals in any questionable circumstances. Id., at 350–351 (“An officer not quite sure the drugs weighed enough to warrant jail time or not quite certain about a suspect’s risk of flight would not arrest, even though it could perfectly well turn out that, in fact, the offense called for incarceration and the defendant was long gone on the day of trial”). The Fourth Amendment did not compel this result in Atwater. The Court held that officers who have probable cause to believe even a minor criminal offense has been committed in their presence may arrest the offender. See id., at 354. Individual jurisdictions can of course choose “to impose more restrictive safeguards through statutes limiting warrantless arrests for minor offenders.” Id., at 352. One of the central principles in Atwater applies with equal force here. Officers who interact with those suspected of violating the law have an “essential interest in readily administrable rules.” Id., at 347; accord, New York v. Belton, 453 U.S. 454, 458 (1981). The officials in charge of the jails in this case urge the Court to reject any complicated constitutional scheme requiring them to conduct less thorough inspections of some detainees based on their behavior, suspected offense, criminal history, and other factors. They offer significant reasons why the Constitution must not prevent them from conducting the same search on any suspected offender who will be admitted to the general population in their facilities. The restrictions suggested by petitioner would limit the intrusion on the privacy of some detainees but at the risk of increased danger to everyone in the facility, including the less serious offenders themselves. IV This case does not require the Court to rule on the types of searches that would be reasonable in instances where, for example, a detainee will be held without assignment to the general jail population and without substantial contact with other detainees. This describes the circumstances in Atwater. See 532 U. S., at 324 (“Officers took Atwater’s ‘mug shot’ and placed her, alone, in a jail cell for about one hour, after which she was taken before a magistrate and released on $310 bond”). The accommodations provided in these situations may diminish the need to conduct some aspects of the searches at issue. Cf. United States Brief 30 (discussing the segregation, and less invasive searches, of individuals held by the Federal Bureau of Prisons for misdemeanors or civil contempt). The circumstances before the Court, however, do not present the opportunity to consider a narrow exception of the sort Justice Alito describes, post, at 2–3 (concurring opinion), which might restrict whether an arrestee whose detention has not yet been reviewed by a magistrate or other judicial officer, and who can be held in available facilities removed from the general population, may be subjected to the types of searches at issue here. Petitioner’s amici raise concerns about instances of officers engaging in intentional humiliation and other abusive practices. See Brief for Sister Bernie Galvin et al. as Amici Curiae; see also Hudson, 468 U. S., at 528 (“[I]ntentional harassment of even the most hardened criminals cannot be tolerated by a civilized society”); Bell, 441 U. S., at 560. There also may be legitimate concerns about the invasiveness of searches that involve the touching of detainees. These issues are not implicated on the facts of this case, however, and it is unnecessary to con- sider them here. V Even assuming all the facts in favor of petitioner, the search procedures at the Burlington County Detention Center and the Essex County Correctional Facility struck a reasonable balance between inmate privacy and the needs of the institutions. The Fourth and Fourteenth Amendments do not require adoption of the framework of rules petitioner proposes. The judgment of the Court of Appeals for the Third Circuit is affirmed. It is so ordered. Notes 1 Justice Thomas joins all but Part IV of this opinion.

566.US.2011_10-1042

The Real Estate Settlement Procedures Act (RESPA), provides, as relevant here, that “[n]o person shall give and no person shall accept any portion, split, or percentage of any charge made or received for the rendering of a real estate settlement service . . . other than for services actually performed.” 12 U. S. C. §2607(b). Petitioners, three couples who obtained mortgage loans from respondent, filed separate state-court actions, alleging that respondent had violated §2607(b) by charging them fees for which no services were provided in return. After the cases were removed to federal court and consolidated, respondent sought summary judgment, arguing that petitioners’ claims were not cognizable under §2607(b) because the allegedly unearned fees were not split with another party. The District Court agreed; and because petitioners had not alleged any splitting of fees, it granted respondent summary judgment. The Fifth Circuit affirmed. Held: In order to establish a violation of §2607(b), a plaintiff must demonstrate that a charge for settlement services was divided between two or more persons. Pp. 3–13. (a) Section 2607(b) unambiguously covers only a settlement-service provider’s splitting of a fee with one or more other persons; it cannot be understood to reach a single provider’s retention of an unearned fee. Pp. 3–11. (1) Section 2607(b) clearly describes two distinct exchanges. First, a “charge” is “made” to or “received” from a consumer by a settlement-service provider. That provider then “give[s],” and another person “accept[s],” a “portion, split, or percentage” of the charge. Congress’s use of different sets of verbs, with distinct tenses, to distinguish between the consumer-provider transaction and the fee-sharing one would be pointless if, as petitioners contend, the two transactions could be collapsed into one. Their reading—that a settlement-service provider can “make” a charge and then “accept” the portion of the charge consisting of 100 percent—does not avoid collapsing the sequential relationship of the two stages and would destroy the tandem character of activities that the text envisions at stage two (i.e., a giving and accepting). And if the consumer were the person who “give[s]” a “portion, split, or percentage” of the charge to the provider who “accepts” it, consumers would become lawbreakers themselves. Pp. 3–8. (2) The normal usage of the terms “portion,” “split,” and “percentage”—which, when referring to a portion or percentage of a whole, usually mean less than 100 percent—reinforces the conclusion that §2607(b) does not apply where a settlement-service provider retains the entirety of a fee received from a consumer. The meaning is also confirmed by the “commonsense canon of noscitur a sociis, which counsels that a word is given more precise content by the neighboring words with which it is associated.” United States v. Williams, 553 U.S. 285, 294. This connation is not undermined by the canon against surplusage. “Portion,” “split,” and “percentage” may all mean the same thing, but the canon merely favors that interpretation which avoids surplusage, see Microsoft Corp. v. i4i Ltd. Partnership, 564 U. S. ___, ___, and petitioners’ interpretation no more achieves that end than the Court’s does. Pp. 8–11. (b) Petitioners’ arguments in favor of their contrary interpretation are unpersuasive. Section 2607(b), as interpreted here, is not rendered surplusage by §2607(a)’s express prohibition of kickbacks, for each subsection reaches conduct that the other does not. RESPA’s gen- eral purpose—to protect consumers from “certain abusive practices,” §2601(a)—also provides no warrant for expanding §2607(b)’s pro- hibition beyond the field to which it is unambiguously limited: the splitting of fees paid for settlement services. And giving §2607(b) its natural meaning would not lead to absurd results. Pp. 11–13. 626 F.3d 799, affirmed. Scalia, J., delivered the opinion for a unanimous Court.

A provision of the Real Estate Settlement Procedures Act (RESPA), codified at 12 U. S. C. §2607(b), prohibits giving and accepting “any portion, split, or percentage of any charge made or received for the rendering of a real estate settlement service . . . other than for services actually performed.” We consider whether, to establish a vio-lation of §2607(b),[1] a plaintiff must demonstrate that a charge was divided between two or more persons. I Enacted in 1974, RESPA regulates the market for real estate “settlement services,” a term defined by statute to include “any service provided in connection with a real estate settlement,” such as “title searches, . . . title insurance, services rendered by an attorney, the preparation of documents, property surveys, the rendering of credit reports or appraisals, . . . services rendered by a real estate agent or broker, the origination of a federally re-lated mortgage loan[[2]] . . . , and the handling of the processing, and closing or settlement.” §2602(3). Among RESPA’s consumer-protection provisions is §2607, which directly furthers Congress’s stated goal of “eliminat[ing] . . . kickbacks or referral fees that tend to increase un-necessarily the costs of certain settlement services,” §2601(b)(2). Section 2607(a) provides: “No person shall give and no person shall accept any fee, kickback, or thing of value pursuant to any agreement or understanding, oral or otherwise, that business incident to or a part of a real estate settlement service involving a federally related mortgage loan shall be referred to any person.” The neighboring provision, subsection (b), adds the following: “No person shall give and no person shall accept any portion, split, or percentage of any charge made or received for the rendering of a real estate settlement service in connection with a transaction involving a federally related mortgage loan other than for services actually performed.” These substantive provisions are enforceable through, in-ter alia, actions for damages brought by consumers of settlement services against “[a]ny person or persons who violate the prohibitions or limitations” of §2607, with recovery set at an amount equal to three times the charge paid by the plaintiff for the settlement service at issue. §2607(d)(2). Petitioners in this case are three married couples who obtained mortgage loans from respondent Quicken Loans, Inc. In 2008, they filed separate actions in Louisiana state court, alleging, as pertinent here, that respondent had violated §2607(b) by charging them fees for which no services were provided. In particular, the Freemans and the Bennetts allege that they were charged loan discount fees of $980 and $1,100, respectively, but that respondent did not give them lower interest rates in return. The Smiths’ allegations focus on a $575 loan “processing fee” and a “loan origination” fee of more than $5,100.[3] Respondent removed petitioners’ lawsuits to federal court, where the cases were consolidated. Respondent thereafter moved for summary judgment on the ground that petitioners’ claims are not cognizable under §2607(b) because the allegedly unearned fees were not split with another party. The District Court agreed; and because petitioners did not allege any splitting of fees it granted summary judgment in favor of respondent. A divided panel of the United States Court of Appeals for the Fifth Circuit affirmed. 626 F.3d 799 (2010). We granted certiorari. 565 U. S. ___ (2011). II The question in this case pertains to the scope of §2607(b), which as we have said provides that “[n]o person shall give and no person shall accept any portion, split, or percentage of any charge made or received for the rendering of a real estate settlement service . . . other than for services actually performed.” The dispute between the parties boils down to whether this provision prohibits the collection of an unearned charge by a single settlement-service provider—what we might call an undivided unearned fee—or whether it covers only transactions in which a provider shares a part of a settlement-service charge with one or more other persons who did nothing to earn that part. Petitioners’ argument that the former interpretation should prevail finds support in a 2001 policy statement issued by the Department of Housing and Urban Development (HUD), the agency that was until recently authorized by Congress to “prescribe such rules and regulations” and “to make such interpretations” as “may be necessary to achieve the purposes of [RESPA],” §2617(a).[4] That policy statement says that §2607(b) “prohibit[s] any person from giving or accepting any unearned fees, i.e., charges or payments for real estate settlement services other than for goods or facilities provided or services performed.” 66 Fed. Reg. 53057 (2001). It “specifically interprets [§2607(b)] as not being limited to situations where at least two persons split or share an unearned fee.” Ibid. More broadly, the policy statement construes §2607(b) as authority for regulation of the charges paid by consumers for the provision of settlements. It says that “a settlement service provider may not mark-up the cost of another provider’s services without providing additional settlement services; such payment must be for services that are actual, necessary and distinct.” Id., at 53059. Moreover, in addition to facing liability when it collects a fee that is entirely unearned, a provider may also “be liable under [§2607(b)] when it charges a fee that exceeds the reason-able value of goods, facilities, or services provided,” ibid., on the theory that the excess over reasonable value constitutes a “portion” of the charge “other than for services actually performed,” §2607(b). The last mentioned point, however, is manifestly in-consistent with the statute HUD purported to construe. When Congress enacted RESPA in 1974, it included a directive that HUD make a report to Congress within five years regarding the need for further legislation in the area. See §2612(a) (1976 ed.). Among the topics required to be included in the report were “recommendations on whether Federal regulation of the charges for real estate settlement services in federally related mortgage transactions is necessary and desirable,” and, if so, recommendations with regard to what reforms should be adopted. §2612(b)(2). The directive for recommendations regarding the desirability of price regulation would make no sense if Congress had already resolved the issue—if §2607(b) already carried with it authority for HUD to proscribe the collection of unreasonably high fees for settlement services, i.e., to engage in price regulation. No doubt recognizing as much, petitioners do not fully adopt HUD’s construction of §2607(b). Noting that even those Courts of Appeals which have found §2607(b) not to be limited to fee-splitting situations have held that the statute does not reach unreasonably high fees, see Kruse v. Wells Fargo Home Mortgage, Inc., 383 F.3d 49, 56 (CA2 2004); Santiago v. GMAC Mortgage Group, Inc., 417 F.3d 384, 387 (CA3 2005); Friedman v. Market Street Mortgage Corp., 520 F.3d 1289, 1297 (CA11 2008), petitioners ac-knowledge that the statute does not cover overcharges. They nonetheless embrace HUD’s construction of §2607(b) insofar as it holds that a provider violates the statute by retaining a fee after providing no services at all in return. In short, petitioners contend that, by allegedly charg- ing each of them an unearned fee, respondent “accept[ed]” a “portion, split, or percentage” of a settlement, service charge (i.e., 100 percent of the charge) “other than for services actually performed.” §2607(b) (2006 ed.). The parties vigorously dispute whether the position set forth in HUD’s 2001 policy statement should be accorded deference under the framework announced by this Court in Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). We need not resolve that dispute—or address whether, if Chevron deference would otherwise apply, it is eliminated by the policy statement’s palpable overreach with regard to price controls. For we conclude that even the more limited position espoused by the policy statement and urged by petitioners “goes beyond the meaning that the statute can bear,” MCI Telecommunications Corp. v. American Telephone & Telegraph Co., 512 U.S. 218, 229 (1994). In our view, §2607(b) unambiguously covers only a settlement-service provider’s splitting of a fee with one or more other persons; it cannot be understood to reach a single provider’s retention of an unearned fee.[5] By providing that no person “shall give” or “shall accept” a “portion, split, or percentage” of a “charge” that has been “made or received,” “other than for services actually performed,” §2607(b) clearly describes two distinct exchanges. First, a “charge” is “made” to or “received” from a consumer by a settlement-service provider. That provider then “give[s],” and another person “accept[s],” a “portion, split, or percentage” of the charge. Congress’s use of different sets of verbs, with distinct tenses, to distinguish between the consumer-provider transaction (the “charge” that is “made or received”) and the fee-sharing transaction (the “portion, split, or percentage” that is “give[n]” or “accept[ed]”) would be pointless if, as petitioners contend, the two transactions could be collapsed into one. Petitioners try to merge the two stages by arguing that a settlement-service provider can “make” a charge (stage one) and then “accept” (stage two) the portion of the charge consisting of 100 percent. See Reply Brief for Petitioners 6. But then is not the provider also “receiv[ing]” the charge at the same time he is “accept[ing]” the portion of it? And who “give[s]” the portion of the charge consisting of 100 percent? The same provider who “accept[s]” it? This reading does not avoid collapsing the sequential relationship of the two stages, and it would simply destroy the tandem character of activities that the text envisions at stage two (i.e., a giving and accepting). Petitioners seek to avoid this consequence, at stage two at least, by saying that the consumer is the person who “give[s]” a “portion, split, or percentage” of the charge to the provider who “accept[s]” it. See Brief for Petitioners 21; Reply Brief for Petitioners 5. But since under this statute it is (so to speak) as accursed to give as to receive, this would make lawbreakers of consumers—the very class for whose benefit §2607(b) was enacted, see §2601. It is no answer to say that a consumer would not face damages liability because a violator is liable only “to the person or persons charged for the settlement service,” §2607(d)(2), and it would not make sense to render a consumer liable to himself. It is the logical consequence that a consumer would be liable to himself, not the specter of actual damages liability, which provides strong indication that something in petitioners’ interpretation is amiss. At any rate, §2607(b) is also enforceable through criminal prosecutions, §2607(d)(1), and actions for injunctive relief brought by federal and state regulators, §2607(d)(4). HUD’s 2001 policy statement asserts that “HUD is, of course, unlikely to direct any enforcement actions against consumers for the payment of unearned fees,” 66 Fed. Reg. 53059, n. 6, but that assurance is cold comfort. Moreover, even assuming (as seems realistic) that the Justice Department would be similarly reluctant to prosecute consumers for criminal violations of §2607(b), “prosecutorial discretion is not a reason for courts to give improbable breadth to criminal statutes.” Abuelhawa v. United States, 556 U.S. 816, 823, n. 3 (2009). Nor is the problem of consumer criminal liability solved by petitioners’ suggestion that an unstated mens rea requirement be read into the criminal enforcement provision, §2607(d)(1), see, e.g., Staples v. United States, 511 U.S. 600, 605 (1994). If that would excuse only those consumers who are unaware that they are paying for unearned services, some consumers would remain criminally liable—those who know that the fee is unearned but decide to pay it anyway, perhaps because the provider’s proposal is still the best deal. And if it would immunize all consumers, the statute’s criminalization of the entire “giving” portion of consumer-provider transactions would make little sense. We find it virtually unthinkable that Congress would leave it to imputed mens rea to preserve from criminal liability some or all of the class RESPA was designed to protect—and entirely unthinkable that Congress would have created that strange disposition through language as obscure as that relied upon here. The phrase “portion, split, or percentage” reinforces the conclusion that §2607(b) does not cover a situation in which a settlement-service provider retains the entirety of a fee received from a consumer. It is certainly true that “portion” or “percentage” can be used to include the entirety, or 100 percent. See, e.g., 18 U. S. C. §648 (“portion”); 5 U. S. C. §8348(g) (2006 ed., Supp. IV) (“percentag[e]”); 5 U. S. C. §8351(b)(2)(B) (2006 ed.) (same); 12 U. S. C. §1467a(m)(7)(B)(ii)(II) (same). But that is not the normal meaning of “portion” when one speaks of “giv[ing]” or “accept[ing]” a portion of the whole, as dictionary definitions uniformly show.[6] Aesop’s fable would be just as wryly humorous if the lion’s claim to the entirety of the kill he hunted in partnership with less ferocious animals had been translated into English as the “lion’s portion” instead of the lion’s share. As for “percentage,” that word can include 100 percent—or even 300 percent—when it refers to merely a ratable measure (“unemployment claims were up 300 percent”).[7] But, like “portion,” it normally means less than all when referring to a “percentage” of a specific whole (“he demanded a percentage of the profits”).[8] And it is normal usage that, in the absence of contrary indication, governs our interpretation of texts. Crawford v. Metropolitan Government of Nashville and Davidson Cty., 555 U.S. 271, 276 (2009); Asgrow Seed Co. v. Winterboer, 513 U.S. 179, 187 (1995). In the present statute, that meaning is confirmed by the “commonsense canon of noscitur a sociis—which counsels that a word is given more precise content by the neighboring words with which it is associated.” United States v. Williams, 553 U.S. 285, 294 (2008). For “portion” and “percentage” do not stand in isolation, but are part of a phrase in which they are joined together by the intervening word “split”—which, as petitioners acknowledge, Brief for Petitioners 19, cannot possibly mean the entirety. We think it clear that, in employing the phrase “portion, split, or percentage,” Congress sought to invoke the words’ common “core of meaning,” Graham County Soil and Water Conservation Dist. v. United States ex rel. Wilson, 559 U. S. ___, ___, n. 7 (2010) (slip op., at 7, n. 7), which is to say, a part of a whole. That is so even though the phrase is preceded by “any”—a word that, we have observed, has an “ ‘expansive meaning,’ ” Department of Housing and Urban Development v. Rucker, 535 U.S. 125, 131 (2002). Expansive, yes; transformative, no. It can broaden to the maximum, but never change in the least, the clear meaning of the phrase selected by Congress here. Contrary to petitioners’ contention, the natural connotation of “portion, split, or percentage” is not undermined in this context by our “general ‘reluctan[ce] to treat statutory terms as surplusage.’ ” Board of Trustees of Leland Stanford Junior Univ. v. Roche Molecular Systems, Inc., 563 U. S. ___, ___ (2011) (slip op., at 9) (quoting Duncan v. Walker, 533 U.S. 167, 174 (2001)). Petitioners rightly point out that under our interpretation “portion,” “split,” and “percentage” all mean the same thing—a perhaps regrettable but not uncommon sort of lawyerly iteration (“give, grant, bargain, sell, and convey”). But the canon against surplusage merely favors that interpretation which avoids surplusage, see Microsoft Corp. v. i4i Ltd. Partnership, 564 U. S. ___, ___–___ (2011) (slip op., at 12–13)—and petitioners’ interpretation no more achieves that end than ours does. It is impossible to imagine a “portion” (even a portion consisting of the entirety) or a “split” that is not also a “percentage.” Petitioners invoke the presumption against surplusage a second time, urging that if §2607(b) is not construed to reach undivided unearned fees, it would be rendered “largely surplusage” in light of §2607(a)’s express prohibition of kickbacks. Brief for Petitioners 24. Not so. Section 2607(a) prohibits giving or accepting “any fee, kickback, or thing of value pursuant to any agreement or understanding . . . that business incident to or a part of a real estate settlement service . . . shall be referred to any person.” §2607(a). That prohibition is at once broader than §2607(b)’s (because it applies to the transfer of any “thing of value,” rather than to the dividing of a “charge” paid by a consumer) and narrower (because it requires an “agreement or understanding” to refer business). Thus, a settlement-service provider who agrees to exchange valuable tickets to a sporting event in return for a referral of business would violate §2607(a), but not §2607(b). So too a provider who agrees to pay a monetary referral fee that is not tied in any respect to a charge paid by a particular consumer—for instance, a “retainer” agreement pursuant to which the provider pays a monthly lump sum in exchange for the recipient’s agreement to refer any business that comes his way. By contrast, a settlement-service provider who gives a portion of a charge to another person who has not rendered any services in return would violate §2607(b), even if an express referral arrangement does not exist or cannot be shown. In short, each subsection reaches conduct that the other does not; there is no need to adopt petitioners’ improbable reading of §2607(b) to avoid rendering any portion of §2607 superfluous. It follows that petitioners can derive no support from §2607’s caption: “Prohibition against kickbacks and unearned fees.” Subsection (a) prohibits certain kickbacks (those agreed to in exchange for referrals) and subsection (b) prohibits certain unearned fees (those paid from a part of the charge to the customer).[9] Petitioners also appeal to statutory purpose, arguing that a prohibition against the charging of undivided unearned fees would fit comfortably with RESPA’s stated goal of “insur[ing] that consumers . . . are protected from unnecessarily high settlement charges caused by certain abusive practices,” §2601(a). It bears noting that RESPA’s declaration of purpose is by its terms limited to “certain abusive practices”—making the statute an even worse candidate than most for the expansion of limited text by the positing of an unlimited purpose. RESPA’s particular language ultimately serves to drive home a broader point: “[N]o legislation pursues its purposes at all costs,” Rodriguez v. United States, 480 U.S. 522, 525–526 (1987) (per curiam), and “[e]very statute purposes, not only to achieve certain ends, but also to achieve them by particular means,” Director, Office of Workers’ Compensation Programs v. Newport News Shipbuilding & Dry Dock Co., 514 U.S. 122, 136 (1995). Vague notions of statutory purpose provide no warrant for expanding §2607(b)’s prohibition beyond the field to which it is unambiguously limited: the splitting of fees paid for settlement services. Nor is there any merit to petitioners’ related contention that §2607(b) should not be given its natural meaning because doing so leads to the allegedly absurd result of permitting a provider to charge and keep the entirety of a $1,000 unearned fee, while imposing liability if the provider shares even a nickel of a $10 charge with someone else. That result does not strike us as particularly anomalous. Congress may well have concluded that existing remedies, such as state-law fraud actions, were sufficient to deal with the problem of entirely fictitious fees, whereas legislative action was required to deal with the problems posed by kickbacks and fee splitting. In any event, petitioners’ reading of the statute leads to an “absurdity” of its own: Because §2607(b) manifestly cannot be understood to prohibit unreasonably high fees, see supra, at 5, a service provider could avoid liability by providing just a dollar’s worth of services in exchange for the $1,000 fee. Acknowledging that §2607(b)’s coverage is limited to fee-splitting transactions at least has the virtue of making it a coherent response to that particular problem, rather than an incoherent response to the broader problem of unreasonably high fees. * * * In order to establish a violation of §2607(b), a plaintiff must demonstrate that a charge for settlement services was divided between two or more persons. Because petitioners do not contend that respondent split the challenged charges with anyone else, summary judgment was properly granted in favor of respondent. We therefore affirm the judgment of the Court of Appeals. It is so ordered. Notes 1 This and all subsequent section references pertain to Title 12 unless otherwise specified. 2 The statutory definition of “federally related mortgage loan” is set forth in §2602(1). 3 Respondent maintains that at least the “loan origination” fee charged to the Smiths was in fact a mislabeled loan discount fee, like the allegedly unearned fees charged to the Freemans and the Bennetts. Respondent contends that loan discount fees fall outside the scope of §2607(b) because they are not fees for settlement services, but rather, as the Eleventh Circuit has held, are part of the pricing of a loan. See Wooten v. Quicken Loans, Inc., 626 F.3d 1187 (2010). Petitioners dispute this point on the merits and further argue that respondent forfeited the contention in the lower courts. We express no view on this issue. 4 On July 21, 2011, HUD’s consumer-protection functions under the RESPA were transferred to the Bureau of Consumer Financial Protection. See Dodd-Frank Wall Street Reform and Consumer Protection Act, §§1061(b)(7) and (d), 1062, 1098, 1100H, 124Stat. 2038, 2039–2040, 2103–2104, 2113. That day, the Bureau issued a notice stating that it would enforce HUD’s RESPA regulations and that, pending further Bureau action, it would apply HUD’s previously issued official policy statements regarding RESPA. 76 Fed. Reg. 43570–43571. 5 Petitioners also contend that the position set forth in the 2001 policy statement is consistent with a HUD regulation, 24 CFR §3500.14(c) (2011), and with prior administrative guidance. In light of our conclusion that §2607(b) unambiguously forecloses petitioners’ position, we have no need to address this issue. 6 See, e.g., Webster’s New International Dictionary 1924 (2d ed. 1954) (hereinafter Webster’s) (defs. 1, 4: defining “portion” as “[a]n allotted part; a share; a parcel; a division in a distribution[;] . . . [a] part ofa whole”); 12 Oxford English Dictionary 154–155 (2d ed. 1989) (here-inafter OED) (defs. 1a, 5: defining “portion” as “[t]he part (of any-thing) allotted or belonging to one person; a share[;] . . . [a] part ofany whole”); American Heritage Dictionary 1373 (5th ed. 2011) (def. 1: defining “portion” as “[a] section or quantity within a larger thing; a part of a whole”). 7 See, e.g., Webster’s 1815 (def. 1: defining “percentage” as “[a] certain rate per cent”); 11 OED 521 (def. a: defining “percentage” as “[a] rate or proportion per cent”). 8 See, e.g., Webster’s 1815 (def. 1: defining “percentage” as “a part or proportion of a whole expressed as so much or many per hundred”); 11 OED 521 (def. a: defining “percentage” as “a quantity or amount reckoned as so much in the hundred, i.e. as so many hundredth parts of another, esp. of the whole of which it is a part; hence loosely, a part or portion considered in its quantitative relation to the whole”); American Heritage Dictionary, supra, at 1307 (def. 2: defining “percentage” as“[a] proportion or share in relation to a whole; a part”). 9 The United States, as amicus curiae, raises an additional argument from the statutory context: that coverage of undivided unearned feesin §2607(b) can be inferred from the text of §2607(d), which sets out penalties for the “person or persons” who violate §2607(a) or §2607(b). §2607(d)(1), (2), and (3) (emphasis added). But Congress’s use ofthe singular “person” does not remotely establish that §2607(b) can be violated by a single culpable actor who accepts an unearned charge from a consumer. In fact, any such inference is negated by the history of §2607. When RESPA was first enacted, §2607(d) separately provided for damages liability of “any person or persons who violate the provisions of subsection (a)” and of “any person or persons who violate the provisions of subsection (b).” §2607(d)(2) (1976 ed.). Because §2607(a), with its reference to an “agreement or understanding,” has always required two culpable parties for a violation, Congress’s use of the phrase “any person or persons” in connection with that subsection demonstrates that the phrase does not have the significance attributed to it by the United States.

565.US.2011_10-545

The Berne Convention for the Protection of Literary and Artistic Works (Berne), which took effect in 1886, is the principal accord governing international copyright relations. Berne’s 164 member states agree to provide a minimum level of copyright protection and to treat authors from other member countries as well as they treat their own. Of central importance in this case, Article 18 of Berne requires countries to protect the works of other member states unless the works’ copyright term has expired in either the country where protection is claimed or the country of origin. A different system of transnational copyright protection long prevailed in this country. Throughout most of the 20th century, the only foreign authors eligible for Copyright Act protection were those whose countries granted reciprocal rights to American authors and whose works were printed in the United States. Despite Article 18, when the United States joined Berne in 1989, it did not protect any foreign works lodged in the U. S. public domain, many of them works never protected here. In 1994, however, the Agreement on Trade-Related Aspects of Intellectual Property Rights mandated implementation of Berne’s first 21 articles, on pain of enforcement by the World Trade Organization. In response, Congress applied the term of protection available to U. S. works to preexisting works from Berne member countries. Section 514 of the Uruguay Round Agreements Act (URAA) grants copyright protection to works protected in their country of origin, but lacking protection in the United States for any of three reasons: The United States did not protect works from the country of origin at the time of publication; the United States did not protect sound recordings fixed before 1972; or the author had not complied with certain U. S. statutory formalities. Works encompassed by §514 are granted the protection they would have enjoyed had the United States maintained copyright relations with the author’s country or removed formalities incompatible with Berne. As a consequence of the barriers to U. S. copyright protection prior to §514’s enactment, foreign works “restored” to protection by the measure had entered the public domain in this country. To cushion the impact of their placement in protected status, §514 provides ameliorating accommodations for parties who had exploited affected works before the URAA was enacted. Petitioners are orchestra conductors, musicians, publishers, and others who formerly enjoyed free access to works §514 removed from the public domain. They maintain that Congress, in passing §514, exceeded its authority under the Copyright Clause and transgressed First Amendment limitations. The District Court granted the Attorney General’s motion for summary judgment. Affirming in part, the Tenth Circuit agreed that Congress had not offended the Copyright Clause, but concluded that §514 required further First Amendment inspection in light of Eldred v. Ashcroft, 537 U.S. 186. On remand, the District Court granted summary judgment to petitioners on the First Amendment claim, holding that §514’s constriction of the public domain was not justified by any of the asserted federal interests. The Tenth Circuit reversed, ruling that §514 was narrowly tailored to fit the important government aim of protecting U. S. copyright holders’ interests abroad. Held: 1. Section 514 does not exceed Congress’ authority under the Copyright Clause. Pp. 13–23. (a) The text of the Copyright Clause does not exclude application of copyright protection to works in the public domain. Eldred is largely dispositive of petitioners’ claim that the Clause’s confinement of a copyright’s lifespan to a “limited Tim[e]” prevents the removal of works from the public domain. In Eldred, the Court upheld the Copyright Term Extension Act (CTEA), which extended, by 20 years, the terms of existing copyrights. The text of the Copyright Clause, the Court observed, contains no “command that a time prescription, once set, becomes forever ‘fixed’ or ‘inalterable,’ ” and the Court declined to infer any such command. 537 U. S., at 199. The construction petitioners tender here is similarly infirm. The terms afforded works restored by §514 are no less “limited” than those the CTEA lengthened. Nor had the “limited Tim[e]” already passed for the works at issue here—many of them works formerly denied any U. S. copyright protection—for a period of exclusivity must begin before it may end. Petitioners also urge that the Government’s position would allow Congress to legislate perpetual copyright terms by instituting successive “limited” terms as prior terms expire. But as in Eldred, such hypothetical misbehavior is far afield from this case. In aligning the United States with other nations bound by Berne, Congress can hardly be charged with a design to move stealthily toward a perpetual copyright regime. Pp. 13–15. (b) Historical practice corroborates the Court’s reading of the Copyright Clause to permit the protection of previously unprotected works. In the Copyright Act of 1790, the First Congress protected works that had been freely reproducible under State copyright laws. Subsequent actions confirm that Congress has not understood the Copyright Clause to preclude protection for existing works. Several private bills restored the copyrights and patents of works and inventions previously in the public domain. Congress has also passed generally applicable legislation granting copyrights and patents to works and inventions that had lost protection. Pp. 15–19. (c) Petitioners also argue that §514 fails to “promote the Progress of Science” as contemplated by the initial words of the Copyright Clause. Specifically, they claim that because §514 affects only works already created, it cannot meet the Clause’s objective. The creation of new works, however, is not the sole way Congress may promote “Science,” i.e., knowledge and learning. In Eldred, this Court rejected a nearly identical argument, concluding that the Clause does not demand that each copyright provision, examined discretely, operate to induce new works. Rather the Clause “empowers Congress to determine the intellectual property regimes that, overall, in that body’s judgment, will serve the ends of the Clause.” 537 U. S., at 222. Nothing in the text or history of the Copyright Clause, moreover, confines the “Progress of Science” exclusively to “incentives for creation.” Historical evidence, congressional practice, and this Court’s decisions, in fact, suggest that inducing the dissemination of existing works is an appropriate means to promote science. Pp. 20–22. (d) Considered against this backdrop, §514 falls comfortably within Congress’ Copyright Clause authority. Congress had reason to believe that a well-functioning international copyright system would encourage the dissemination of existing and future works. And testimony informed Congress that full compliance with Berne would expand the foreign markets available to U. S. authors and invigorate protection against piracy of U. S. works abroad, thus benefitting copyright-intensive industries stateside and inducing greater investment in the creative process. This Court has no warrant to reject Congress’ rational judgment that exemplary adherence to Berne would serve the objectives of the Copyright Clause. Pp. 22–23. 2. The First Amendment does not inhibit the restoration authorized by §514. Pp. 23–32. (a) The pathmarking Eldred decision is again instructive. There, the Court held that the CTEA’s enlargement of a copyright’s duration did not offend the First Amendment’s freedom of expression guarantee. Recognizing that some restriction on expression is the inherent and intended effect of every grant of copyright, the Court observed that the Framers regarded copyright protection not simply as a limit on the manner in which expressive works may be used, but also as an “engine of free expression.” 537 U. S., at 219. The “traditional contours” of copyright protection, i.e., the “idea/expression dichotomy” and the “fair use” defense, moreover, serve as “built-in First Amendment accommodations.” Ibid. Given the speech-protective purposes and safeguards embraced by copyright law, there was no call for the heightened review sought in Eldred. The Court reaches the same conclusion here. Section 514 leaves undisturbed the idea/expression distinction and the fair use defense. Moreover, Congress adopted measures to ease the transition from a national scheme to an international copyright regime. Pp. 23–26. (b) Petitioners claim that First Amendment interests of a higher order are at stake because they—unlike their Eldred counterparts—enjoyed “vested rights” in works that had already entered the public domain. Their contentions depend on an argument already considered and rejected, namely, that the Constitution renders the public domain largely untouchable by Congress. Nothing in the historical record, subsequent congressional practice, or this Court’s jurisprudence warrants exceptional First Amendment solicitude for copyrighted works that were once in the public domain. Congress has several times adjusted copyright law to protect new categories of works as well as works previously in the public domain. Section 514, moreover, does not impose a blanket prohibition on public access. The question is whether would-be users of certain foreign works must pay for their desired use of the author’s expression, or else limit their exploitation to “fair use” of those works. By fully implementing Berne, Congress ensured that these works, like domestic and most other foreign works, would be governed by the same legal regime. Section 514 simply placed foreign works in the position they would have occupied if the current copyright regime had been in effect when those works were created and first published. Pp. 26–30. 609 F.3d 1076, affirmed. Ginsburg, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Thomas, and Sotomayor, JJ., joined. Breyer, J., filed a dissenting opinion, in which Alito, J., joined. Kagan, J., took no part in the consideration or decision of the case.

The Berne Convention for the Protection of Literary and Artistic Works (Berne Convention or Berne), which took effect in 1886, is the principal accord governing international copyright relations. Latecomer to the international copyright regime launched by Berne, the United States joined the Convention in 1989. To perfect U. S. implementation of Berne, and as part of our response to the Uruguay Round of multilateral trade negotiations, Congress, in 1994, gave works enjoying copyright protection abroad the same full term of protection available to U. S. works. Congress did so in §514 of the Uruguay Round Agreements Act (URAA), which grants copyright protection to preexisting works of Berne member countries, protected in their country of origin, but lacking protection in the United States for any of three reasons: The United States did not protect works from the country of origin at the time of publication; the United States did not protect sound recordings fixed before 1972; or the author had failed to comply with U. S. statutory formalities (formalities Congress no longer requires as prerequisites to copyright protection). The URAA accords no protection to a foreign work after its full copyright term has expired, causing it to fall into the public domain, whether under the laws of the country of origin or of this country. Works encompassed by §514 are granted the protection they would have enjoyed had the United States maintained copyright relations with the author’s country or removed formalities incompatible with Berne. Foreign authors, however, gain no credit for the protection they lacked in years prior to §514’s enactment. They therefore enjoy fewer total years of exclusivity than do their U. S. counterparts. As a consequence of the barriers to U. S. copyright protection prior to the enactment of §514, foreign works “restored” to protection by the measure had entered the public domain in this country. To cushion the impact of their placement in protected status, Congress included in §514 ameliorating accommodations for parties who had exploited affected works before the URAA was enacted. Petitioners include orchestra conductors, musicians, pub-lishers, and others who formerly enjoyed free access to works §514 removed from the public domain. They maintain that the Constitution’s Copyright and Patent Clause, Art. I, §8, cl. 8, and First Amendment both decree the invalidity of §514. Under those prescriptions of our highest law, petitioners assert, a work that has entered the public domain, for whatever reason, must forever remain there. In accord with the judgment of the Tenth Circuit, we conclude that §514 does not transgress constitutional limitations on Congress’ authority. Neither the Copyright and Patent Clause nor the First Amendment, we hold, makes the public domain, in any and all cases, a territory that works may never exit. I A Members of the Berne Union agree to treat authors from other member countries as well as they treat their own. Berne Convention, Sept. 9, 1886, as revised at Stockholm on July 14, 1967, Art. 1, 5(1), 828 U. N. T. S. 221, 225, 231–233. Nationals of a member country, as well as any author who publishes in one of Berne’s 164 member states, thus enjoy copyright protection in nations across the globe. Art. 2(6), 3. Each country, moreover, must afford at least the minimum level of protection specified by Berne. The copyright term must span the author’s lifetime, plus at least 50 additional years, whether or not the author has complied with a member state’s legal formalities. Art. 5(2), 7(1). And, as relevant here, a work must be protected abroad unless its copyright term has expired in either the country where protection is claimed or the country of origin. Art. 18(1)–(2). [ 1 ] A different system of transnational copyright protection long prevailed in this country. Until 1891, foreign works were categorically excluded from Copyright Act protection. Throughout most of the 20th century, the only eligible foreign authors were those whose countries granted reciprocal rights to U. S. authors and whose works were print ed in the United States. See Act of Mar. 3, 1891, §3, 13, 26Stat. 1107, 1110; Patry, The United States and Inter-national Copyright Law, 40 Houston L. Rev. 749, 750 (2003). [ 2 ] For domestic and foreign authors alike, protection hinged on compliance with notice, registration, and renewal formalities. The United States became party to Berne’s multilateral, formality-free copyright regime in 1989. Initially, Congress adopted a “minimalist approach” to compliance with the Convention. H. R. Rep. No. 100–609, p. 7 (1988) (hereinafter BCIA House Report). The Berne Convention Implementation Act of 1988 (BCIA), 102Stat. 2853, made “only those changes to American copyright law that [were] clearly required under the treaty’s provisions,” BCIA House Report, at 7. Despite Berne’s instruction that member countries—including “new accessions to the Union”—protect foreign works under copyright in the country of origin, Art. 18(1) and (4), 828 U. N. T. S., at 251, the BCIA accorded no protection for “any work that is in the public domain in the United States,” §12, 102Stat. 2860. Protection of future foreign works, the BCIA indicated, satisfied Article 18. See §2(3), 102Stat. 2853 (“The amendments made by this Act, together with the law as it exists on the date of the enactment of this Act, satisfy the obligations of the United States in adhering to the Berne Convention . . . .”). Congress indicated, however, that it had not definitively rejected “retroactive” protection for preexisting foreign works; instead it had punted on this issue of Berne’s implementation, deferring consideration until “a more thorough examination of Constitutional, commercial, and consumer considerations is possible.” BCIA House Report, at 51, 52. [ 3 ] The minimalist approach essayed by the United States did not sit well with other Berne members. [ 4 ] While negotiations were ongoing over the North American Free Trade Agreement (NAFTA), Mexican authorities complained about the United States’ refusal to grant protection, in accord with Article 18, to Mexican works that remained under copyright domestically. See Intellectual Property and International Issues, Hearings before the Subcommittee on Intellectual Property and Judicial Administration, House Committee on the Judiciary, 102d Cong., 1st Sess., 168 (1991) (statement of Ralph Oman, U. S. Register of Copyrights). [ 5 ] The Register of Copyrights also reported “questions” from Turkey, Egypt, and Austria. Ibid. Thailand and Russia balked at protecting U. S. works, copyrighted here but in those countries’ public domains, until the United States reciprocated with respect to their authors’ works. URAA Joint Hearing 137 (statement of Ira S. Shapiro, General Counsel, Office of the U. S. Trade Representative (USTR)); id., at 208 (statement of Professor Shira Perlmutter); id., at 291 (statement of Jason S. Berman, Recording Industry Association of America (RIAA)). [ 6 ] Berne, however, did not provide a potent enforcement mechanism. The Convention contemplates dispute resolution before the International Court of Justice. Art. 33(1). But it specifies no sanctions for noncompliance and allows parties, at any time, to declare themselves “not . . . bound” by the Convention’s dispute resolution provision. Art. 33(2)–(3) 828 U. N. T. S., at 277. Unsurprisingly, no enforcement actions were launched before 1994. D. Gervais, The TRIPS Agreement 213, and n. 134 (3d ed. 2008). Although “several Berne Union Members disagreed with [our] interpretation of Article 18,” the USTR told Congress, the Berne Convention did “not provide a meaningful dispute resolution process.” URAA Joint Hearing 137 (statement of Shapiro). This shortcoming left Congress “free to adopt a minimalist approach and evade Article 18.” Karp, Final Report, Berne Article 18 Study on Retroactive United States Copyright Protection for Berne and other Works, 20 Colum.-VLA J. L. & Arts 157, 172 (1996). The landscape changed in 1994. The Uruguay round of multilateral trade negotiations produced the World Trade Organization (WTO) and the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS). [ 7 ] The United States joined both. TRIPS mandates, on pain of WTO enforcement, implementation of Berne’s first 21 articles. TRIPS, Art. 9.1, 33 I. L. M. 1197, 1201 (requiring adherence to all but the “moral rights” provisions of Article 6bis). The WTO gave teeth to the Convention’s requirements: Noncompliance with a WTO ruling could subject member countries to tariffs or cross-sector retaliation. See Gervais, supra, at 213; 7 W. Patry, Copyright §24:1, pp. 24–8 to 24–9 (2011). The specter of WTO enforcement proceedings bolstered the credibility of our trading partners’ threats to challenge the United States for inadequate compliance with Article 18. See URAA Joint Hearing 137 (statement of Shapiro, USTR) (“It is likely that other WTO members would challenge the current U. S. implementation of Berne Article 18 under [WTO] procedures.”). [ 8 ] Congress’ response to the Uruguay agreements put to rest any questions concerning U. S. compliance with Article 18. Section 514 of the URAA, 108Stat. 4976 (codified at 17 U. S. C. §104A, 109(a)), [ 9 ] extended copyright to works that garnered protection in their countries of origin, [ 10 ] but had no right to exclusivity in the United States for any of three reasons: lack of copyright relations between the country of origin and the United States at the time of publication; lack of subject-matter protection for sound recordings fixed before 1972; and failure to comply with U. S. statutory formalities (e.g., failure to provide notice of copyright status, or to register and renew a copyright). See §104A(h)(6)(B)–(C). [ 11 ] Works that have fallen into the public domain after the expiration of a full copyright term—either in the United States or the country of origin—receive no further protection under §514. Ibid. [ 12 ] Copyrights “restored” [ 13 ] under URAA §514 “subsist for the remainder of the term of copyright that the work would have otherwise been granted . . . if the work never entered the public domain.” §104A(a)(1)(B). Prospectively, restoration places foreign works on an equal footing with their U. S. counterparts; assuming a foreign and domestic author died the same day, their works will enter the public domain simultaneously. See §302(a) (copyrights generally expire 70 years after the author’s death). Restored works, however, receive no compensatory time for the period of exclusivity they would have enjoyed before §514’s enactment, had they been protected at the outset in the United States. Their total term, therefore, falls short of that available to similarly situated U. S. works. The URAA’s disturbance of the public domain hardly escaped Congress’ attention. Section 514 imposed no liability for any use of foreign works occurring before restoration. In addition, anyone remained free to copy and use restored works for one year following §514’s enactment. See 17 U. S. C. §104A(h)(2)(A). Concerns about §514’s compatibility with the Fifth Amendment’s Takings Clause led Congress to include additional protections for “reliance parties”—those who had, before the URAA’s enactment, used or acquired a foreign work then in the public domain. See §104A(h)(3)–(4). [ 14 ] Reliance parties may continue to exploit a restored work until the owner of the restored copyright gives notice of intent to enforce—either by filing with the U. S. Copyright Office within two years of restoration, or by actually notifying the reliance party. §104A(c), (d)(2)(A)(i), and (B)(i). After that, reliance parties may continue to exploit existing copies for a grace period of one year. §104A(d)(2)(A)(ii), and (B)(ii). Finally, anyone who, before the URAA’s enactment, created a “derivative work” based on a restored work may indefinitely exploit the derivation upon payment to the copyright holder of “reasonable compensation,” to be set by a district judge if the parties cannot agree. §104A(d)(3). B In 2001, petitioners filed this lawsuit challenging §514. They maintain that Congress, when it passed the URAA, exceeded its authority under the Copyright Clause and transgressed First Amendment limitations. [ 15 ] The District Court granted the Attorney General’s motion for summary judgment. Golan v. Gonzales, No. Civ. 01–B–1854, 2005 WL 914754 (D. Colo., Apr. 20, 2005). In rejecting petitioners’ Copyright Clause argument, the court stated that Congress “has historically demonstrated little compunction about removing copyrightable materials from the public domain.” Id., at *14. The court next declined to part from “the settled rule that private censorship via copyright enforcement does not implicate First Amendment concerns.” Id., at *17. The Court of Appeals for the Tenth Circuit affirmed in part. Golan v. Gonzales, 501 F.3d 1179 (2007). The public domain, it agreed, was not a “threshold that Congress” was powerless to “traverse in both directions.” Id., at 1187 (internal quotations marks omitted). But §514, as the Court of Appeals read our decision in Eldred v. Ashcroft, 537 U.S. 186 (2003) , required further First Amendment inspection, 501 F. 3d, at 1187. The measure “ ‘altered the traditional contours of copyright protection,’ ” the court said—specifically, the “bedrock principle” that once works enter the public domain, they do not leave. Ibid. (quoting Eldred, 537 U. S., at 221). The case was remanded with an instruction to the District Court to address the First Amendment claim in light of the Tenth Circuit’s opinion. On remand, the District Court’s starting premise was uncontested: Section 514 does not regulate speech on the basis of its content; therefore the law would be upheld if “narrowly tailored to serve a significant government interest.” 611 F. Supp. 2d 1165, 1170–1171 (Colo. 2009) (quoting Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989) ). Summary judgment was due petitioners, the court concluded, because §514’s constriction of the public domain was not justified by any of the asserted federal interests: compliance with Berne, securing greater protection for U. S. authors abroad, or remediation of the inequitable treatment suffered by foreign authors whose works lacked protection in the United States. 611 F. Supp. 2d, at 1172–1177. The Tenth Circuit reversed. Deferring to Congress’ predictive judgments in matters relating to foreign affairs, the appellate court held that §514 survived First Amendment scrutiny. Specifically, the court determined that the law was narrowly tailored to fit the important government aim of protecting U. S. copyright holders’ interests abroad. 609 F.3d 1076 (2010). We granted certiorari to consider petitioners’ challenge to §514 under both the Copyright Clause and the First Amendment, 562 U. S. ___ (2011), and now affirm. II We first address petitioners’ argument that Congress lacked authority, under the Copyright Clause, to enact §514. The Constitution states that “Congress shall have Power . . . [t]o promote the Progress of Science . . . by securing for limited Times to Authors . . . the exclusive Right to their . . . Writings.” Art. I, §8, cl. 8. Petitioners find in this grant of authority an impenetrable barrier to the extension of copyright protection to authors whose writings, for whatever reason, are in the public domain. We see no such barrier in the text of the Copyright Clause, historical practice, or our precedents. A The text of the Copyright Clause does not exclude application of copyright protection to works in the public domain. Symposium, Congressional Power and Limitations Inherent in the Copyright Clause, 30 Colum. J. L. & Arts 259, 266 (2007). Petitioners’ contrary argument relies primarily on the Constitution’s confinement of a copyright’s lifespan to a “limited Tim[e].” “Removing works from the public domain,” they contend, “violates the ‘limited [t]imes’ restriction by turning a fixed and predictable period into one that can be reset or resurrected at any time, even after it expires.” Brief for Petitioners 22. Our decision in Eldred is largely dispositive of petitioners’ limited-time argument. There we addressed the question whether Congress violated the Copyright Clause when it extended, by 20 years, the terms of existing copyrights. 537 U. S., at 192–193 (upholding Copyright Term Extension Act (CTEA)). Ruling that Congress acted within constitutional bounds, we declined to infer from the text of the Copyright Clause “the command that a time prescription, once set, becomes forever ‘fixed’ or ‘inalterable.’ ” Id., at 199. “The word ‘limited,’ ” we observed, “does not convey a meaning so constricted.” Ibid. Rather, the term is best understood to mean “confine[d] within certain bounds,” “restrain[ed],” or “circumscribed.” Ibid. (internal quotation marks omitted). The construction petitioners tender closely resembles the definition rejected in Eldred and is similarly infirm. The terms afforded works restored by §514 are no less “limited” than those the CTEA lengthened. In light of Eldred, petitioners do not here contend that the term Congress has granted U. S. authors—their lifetimes, plus 70 years—is unlimited. See 17 U. S. C. §302(a). Nor do petitioners explain why terms of the same duration, as applied to foreign works, are not equally “circumscribed” and “confined.” See Eldred, 537 U. S., at 199. Indeed, as earlier noted, see supra, at 2, 10, the copyrights of restored foreign works typically last for fewer years than those of their domestic counterparts. The difference, petitioners say, is that the limited time had already passed for works in the public domain. What was that limited term for foreign works once excluded from U. S. copyright protection? Exactly “zero,” petitioners respond. Brief for Petitioners 22 (works in question “received a specific term of protection . . . sometimes expressly set to zero”; “at the end of that period,” they “entered the public domain”); Tr. of Oral Arg. 52 (by “refusing to provide any protection for a work,” Congress “set[s] the term at zero,” and thereby “tell[s] us when the end has come”). We find scant sense in this argument, for surely a “limited time” of exclusivity must begin before it may end. [ 16 ] Carried to its logical conclusion, petitioners persist, the Government’s position would allow Congress to institute a second “limited” term after the first expires, a third after that, and so on. Thus, as long as Congress legislated in installments, perpetual copyright terms would be achievable. As in Eldred, the hypothetical legislative misbehavior petitioners posit is far afield from the case before us. See 537 U. S., at 198–200, 209–210. In aligning the United States with other nations bound by the Berne Convention, and thereby according equitable treatment to once dis-favored foreign authors, Congress can hardly be charged with a design to move stealthily toward a regime of perpetual copyrights. B Historical practice corroborates our reading of the Copyright Clause to permit full U. S. compliance with Berne. Undoubtedly, federal copyright legislation generally has not affected works in the public domain. Section 514’s disturbance of that domain, petitioners argue, distin- guishes their suit from Eldred’s. In adopting the CTEA, petitioners note, Congress acted in accord with “an unbroken congressional practice” of granting pre-expiration term extensions, 537 U. S., at 200. No comparable practice, they maintain, supports §514. On occasion, however, Congress has seen fit to protect works once freely available. Notably, the Copyright Act of 1790 granted protection to many works previously in the public domain. Act of May 31, 1790 (1790 Act), §1, 1Stat. 124 (covering “any map, chart, book, or books already printed within these United States”). Before the Act launched a uniform national system, three States provided no statutory copyright protection at all. [ 17 ] Of those that did afford some protection, seven failed to protect maps; [ 18 ] eight did not cover previously published books; [ 19 ] and all ten denied protection to works that failed to comply with formalities. [ 20 ] The First Congress, it thus appears, did not view the public domain as inviolate. As we have recognized, the “construction placed upon the Constitution by [the drafters of] the first [copyright] act of 1790 and the act of 1802 . . . men who were contemporary with [the Constitution’s] formation, many of whom were members of the convention which framed it, is of itself entitled to very great weight.” Burrow-Giles Lithographic Co. v. Sarony, 111 U.S. 53, 57 (1884) . [ 21 ] Subsequent actions confirm that Congress has not understood the Copyright Clause to preclude protection for existing works. Several private bills restored the copyrights of works that previously had been in the public domain. See Act of Feb. 19, 1849 (Corson Act), ch. 57, 9Stat. 763; Act of June 23, 1874 (Helmuth Act), ch. 534, 18Stat. 618; Act of Feb. 17, 1898 (Jones Act), ch. 29, 30Stat. 1396. These bills were unchallenged in court. Analogous patent statutes, however, were upheld in litigation. [ 22 ] In 1808, Congress passed a private bill restoring patent protection to Oliver Evans’ flour mill. When Evans sued for infringement, first Chief Justice Marshall in the Circuit Court, Evans v. Jordan, 8 F. Cas. 872 (No. 4,564) (Va. 1813), and then Justice Bushrod Washington for this Court, Evans v. Jordan, 9 Cranch 199 (1815), upheld the restored patent’s validity. After the patent’s expiration, the Court said, “a general right to use [Evans’] discovery was not so vested in the public” as to allow the defendant to continue using the machinery, which he had constructed between the patent’s expiration and the bill’s passage. Id., at 202. See also Blanchard v. Sprague, 3 F. Cas. 648, 650 (No. 1,518) (CC Mass. 1839) (Story, J.) (“I never have entertained any doubt of the constitutional authority of congress” to “give a patent for an invention, which . . . was in public use and enjoyed by the community at the time of the passage of the act.”). This Court again upheld Congress’ restoration of an invention to protected status in McClurg v. Kingsland, 1 How. 202 (1843). There we enforced an 1839 amendment that recognized a patent on an invention despite its prior use by the inventor’s employer. Absent such dispensation, the employer’s use would have rendered the invention unpatentable, and therefore open to exploitation without the inventor’s leave. Id., at 206–209. Congress has also passed generally applicable legislation granting patents and copyrights to inventions and works that had lost protection. An 1832 statute authorized a new patent for any inventor whose failure, “by inadvertence, accident, or mistake,” to comply with statutory formalities rendered the original patent “invalid or inoperative.” Act of July 3, §3, 4Stat. 559. An 1893 measure similarly allowed authors who had not timely deposited their work to receive “all the rights and privileges” the Copyright Act affords, if they made the required deposit by March 1, 1893. Act of Mar. 3, ch. 215, 27Stat. 743. [ 23 ] And in 1919 and 1941, Congress authorized the President to issue proclamations granting protection to foreign works that had fallen into the public domain during World Wars I and II. See Act of Dec. 18, 1919, ch. 11, 41Stat. 368; Act of Sept. 25, 1941, ch. 421, 55Stat. 732. [ 24 ] Pointing to dictum in Graham v. John Deere Co. of Kansas City, 383 U.S. 1 (1966) , petitioners would have us look past this history. In Graham, we stated that “Congress may not authorize the issuance of patents whose effects are to remove existent knowledge from the public domain, or to restrict free access to materials already available.” Id., at 6; post, at 15. But as we explained in Eldred, this passage did not speak to the constitutional limits on Congress’ copyright and patent authority. Rather, it “addressed an invention’s very eligibility for patent protection.” 537 U. S., at 202, n. 7. Installing a federal copyright system and ameliorating the interruptions of global war, it is true, presented Congress with extraordinary situations. Yet the TRIPS accord, leading the United States to comply in full measure with Berne, was also a signal event. See supra, at 7–8; cf. Eldred, 537 U. S., at 259, 264–265 (Breyer, J., dissenting) (acknowledging importance of international uniformity advanced by U. S. efforts to conform to the Berne Convention). Given the authority we hold Congress has, we will not second-guess the political choice Congress made between leaving the public domain untouched and embracing Berne unstintingly. Cf. id., at 212–213. C Petitioners’ ultimate argument as to the Copyright and Patent Clause concerns its initial words. Congress is empowered to “promote the Progress of Science and useful Arts” by enacting systems of copyright and patent protection. U. S. Const., Art. I, §8, cl. 8. Perhaps counterintuitively for the contemporary reader, Congress’ copyright authority is tied to the progress of science; its patent authority, to the progress of the useful arts. See Graham, 383 U. S., at 5, and n. 1; Evans, 8 F. Cas., at 873 (Marshall, J.). The “Progress of Science,” petitioners acknowledge, refers broadly to “the creation and spread of knowledge and learning.” Brief for Petitioners 21; accord post, at 1. They nevertheless argue that federal legislation cannot serve the Clause’s aim unless the legislation “spur[s] the creation of . . . new works.” Brief for Petitioners 24; accord post, at 1–2, 8, 17. Because §514 deals solely with works already created, petitioners urge, it “provides no plausible incentive to create new works” and is therefore invalid. Reply Brief 4. [ 25 ] The creation of at least one new work, however, is not the sole way Congress may promote knowledge and learning. In Eldred, we rejected an argument nearly identical to the one petitioners rehearse. The Eldred petitioners urged that the “CTEA’s extension of existing copyrights categorically fails to ‘promote the Progress of Science,’ . . . because it does not stimulate the creation of new works.” 537 U. S., at 211–212. In response to this argument, we held that the Copyright Clause does not demand that each copyright provision, examined discretely, operate to induce new works. Rather, we explained, the Clause “empowers Congress to determine the intellectual property regimes that, overall, in that body’s judgment, will serve the ends of the Clause.” Id., at 222. And those permissible ends, we held, extended beyond the creation of new works. See id., at 205–206 (rejecting the notion that “ ‘the only way to promote the progress of science [is] to provide incentives to create new works’ ” (quoting Perlmutter, Participation in the International Copyright System as a Means to Promote the Progress of Science and Useful Arts, 36 Loyola (LA) L. Rev. 323, 332 (2002))). [ 26 ] Even were we writing on a clean slate, petitioners’ argument would be unavailing. Nothing in the text of the Copyright Clause confines the “Progress of Science” exclusively to “incentives for creation.” Id., at 324, n. 5 (internal quotation marks omitted). Evidence from the founding, moreover, suggests that inducing dissemination—as opposed to creation—was viewed as an appropriate means to promote science. See Nachbar, Constructing Copyright’s Mythology, 6 Green Bag 2d 37, 44 (2002) (“The scope of copyright protection existing at the time of the framing,” trained as it was on “publication, not creation,” “is inconsistent with claims that copyright must promote creative activity in order to be valid.” (internal quotation marks omitted)). Until 1976, in fact, Congress made “federal copyright contingent on publication[,] [thereby] providing incentives not primarily for creation,” but for dissemination. Perlmutter, supra, at 324, n. 5. Our decisions correspondingly recognize that “copyright supplies the economic incentive to create and disseminate ideas.” Harper & Row, Publishers, Inc. v. Nation Enterprises, 471 U.S. 539, 558 (1985) (emphasis added). See also Eldred, 537 U. S., at 206. [ 27 ] Considered against this backdrop, §514 falls comfortably within Congress’ authority under the Copyright Clause. Congress rationally could have concluded that adherence to Berne “promotes the diffusion of knowledge,” Brief for Petitioners 4. A well-functioning international copyright system would likely encourage the dissemination of existing and future works. See URAA Joint Hearing 189 (statement of Professor Perlmutter). Full compliance with Berne, Congress had reason to believe, would expand the foreign markets available to U. S. authors and invigorate protection against piracy of U. S. works abroad, S. Rep. No. 103–412, pp. 224, 225 (1994); URAA Joint Hearing 291 (statement of Berman, RIAA); id., at 244, 247 (statement of Smith, IIPA), thereby benefitting copyright-intensive industries stateside and inducing greater investment in the creative process. The provision of incentives for the creation of new works is surely an essential means to advance the spread of knowledge and learning. We hold, however, that it is not the sole means Congress may use “[t]o promote the Progress of Science.” See Perlmutter, supra, at 332 (United States would “lose all flexibility” were the provision of incentives to create the exclusive way to promote the progress of science). [ 28 ] Congress determined that exem-plary adherence to Berne would serve the objectives of the Copyright Clause. We have no warrant to reject the rational judgment Congress made. III A We next explain why the First Amendment does not inhibit the restoration authorized by §514. To do so, we first recapitulate the relevant part of our pathmarking decision in Eldred. The petitioners in Eldred, like those here, argued that Congress had violated not only the “limited Times” prescription of the Copyright Clause. In addition, and independently, the Eldred petitioners charged, Congress had offended the First Amendment’s freedom of expression guarantee. The CTEA’s 20-year enlargement of a copyright’s duration, we held in Eldred, offended neither provision. Concerning the First Amendment, we recognized that some restriction on expression is the inherent and in-tended effect of every grant of copyright. Noting that the “Copyright Clause and the First Amendment were adopted close in time,” 537 U. S., at 219, we observed that the Framers regarded copyright protection not simply as a limit on the manner in which expressive works may be used. They also saw copyright as an “engine of free expression[:] By establishing a marketable right to the use of one’s expression, copyright supplies the economic incentive to create and disseminate ideas.” Ibid. (quoting Harper & Row, 471 U. S., at 558 (internal quotation marks omitted)); see id., at 546 (“rights conferred by copyright are designed to assure contributors to the store of knowledge a fair return for their labors”). We then described the “traditional contours” of copyright protection, i.e., the “idea/expression dichotomy” and the “fair use” defense. [ 29 ] Both are recognized in our jurisprudence as “built-in First Amendment accommodations.” Eldred, 537 U. S., at 219; see Harper & Row, 471 U. S., at 560 ( First Amendment protections are “embodied in the Copyright Act’s distinction between copyrightable expression and uncopyrightable facts and ideas,” and in the “latitude for scholarship and comment” safeguarded by the fair use defense). The idea/expression dichotomy is codified at 17 U. S. C. §102(b): “In no case does copyright protec[t] . . . any idea, procedure, process, system, method of operation, concept, principle, or discovery . . . described, explained, illustrated, or embodied in [the copyrighted] work.” “Due to this [idea/expression] distinction, every idea, theory, and fact in a copyrighted work becomes instantly available for public exploitation at the moment of publication”; the author’s expression alone gains copyright protection. Eldred, 537 U. S., at 219; see Harper & Row, 471 U. S., at 556 (“idea/expression dichotomy strike[s] a definitional balance between the First Amendment and the Copyright Act by permitting free communication of facts while still protecting an author’s expression” (internal quotation marks omitted)). The second “traditional contour,” the fair use defense, is codified at 17 U. S. C. §107: “[T]he fair use of a copyrighted work, including such use by reproduction in copies . . . , for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright.” This limitation on exclusivity “allows the public to use not only facts and ideas contained in a copyrighted work, but also [the author’s] expression itself in certain circumstances.” Eldred, 537 U. S., at 219; see id., at 220 (“fair use defense affords considerable latitude for scholarship and comment, . . . even for parody” (internal quotation marks omitted)). Given the “speech-protective purposes and safeguards” embraced by copyright law, see id., at 219, we concluded in Eldred that there was no call for the heightened review petitioners sought in that case. [ 30 ] We reach the same conclusion here. [ 31 ] Section 514 leaves undisturbed the “idea/expression” distinction and the “fair use” defense. Moreover, Congress adopted measures to ease the transition from a national scheme to an international copyright regime: It deferred the date from which enforcement runs, and it cushioned the impact of restoration on “reliance parties” who exploited foreign works denied protection before §514 took effect. See supra, at 10–11 (describing 17 U. S. C. §104A(c), (d), and (h)). See also Eldred, 537 U. S., at 220 (describing supplemental allowances and exemp- tions available to certain users to mitigate the CTEA’s impact). B Petitioners attempt to distinguish their challenge from the one turned away in Eldred. First Amendment interests of a higher order are at stake here, petitioners say, because they—unlike their counterparts in Eldred—enjoyed “vested rights” in works that had already entered the public domain. The limited rights they retain under copyright law’s “built-in safeguards” are, in their view, no substitute for the unlimited use they enjoyed before §514’s enactment. Nor, petitioners urge, does §514’s “unprecedented” foray into the public domain possess the historical pedigree that supported the term extension at issue in Eldred. Brief for Petitioners 42–43. However spun, these contentions depend on an argument we considered and rejected above, namely, that the Constitution renders the public domain largely untouch-able by Congress. Petitioners here attempt to achieve under the banner of the First Amendment what they could not win under the Copyright Clause: On their view of the Copyright Clause, the public domain is inviolable; as they read the First Amendment, the public domain is policed through heightened judicial scrutiny of Congress’ means and ends. As we have already shown, see supra, at 13–19, the text of the Copyright Clause and the historical record scarcely establish that “once a work enters the public domain,” Congress cannot permit anyone—“not even the creator—[to] copyright it,” 501 F. 3d, at 1184. And nothing in the historical record, congressional practice, or our own jurisprudence warrants exceptional First Amendment solicitude for copyrighted works that were once in the public domain. [ 32 ] Neither this challenge nor that raised in Eldred, we stress, allege Congress transgressed a gener-ally applicable First Amendment prohibition; we are not faced, for example, with copyright protection that hinges on the author’s viewpoint. The Tenth Circuit’s initial opinion determined that petitioners marshaled a stronger First Amendment challenge than did their predecessors in Eldred, who never “possessed unfettered access to any of the works at issue.” 501 F. 3d, at 1193. See also id., at 1194 (“[O]nce the works at issue became free for anyone to copy, [petitioners] had vested First Amendment interests in the expressions, [thus] §514’s interference with [petitioners’] rights is subject to First Amendment scrutiny.”). As petitioners put it in this Court, Congress impermissibly revoked their right to exploit foreign works that “belonged to them” once the works were in the public domain. Brief for Petitioners 44–45. To copyright lawyers, the “vested rights” formulation might sound exactly backwards: Rights typically vest at the outset of copyright protection, in an author or rightholder. See, e.g., 17 U. S. C. §201(a) (“Copyright in a work protected . . . vests initially in the author . . . .”). Once the term of protection ends, the works do not revest in any rightholder. Instead, the works simply lapse into the public domain. See, e.g., Berne, Art. 18(1), 828 U. N. T. S., at 251 (“This Convention shall apply to all works which . . . have not yet fallen into the public domain . . . .”). Anyone has free access to the public domain, but no one, after the copyright term has expired, acquires ownership rights in the once-protected works. Congress recurrently adjusts copyright law to protect categories of works once outside the law’s compass. For example, Congress broke new ground when it extended copyright protection to foreign works in 1891, Act of Mar. 3, §13, 26Stat. 1110; to dramatic works in 1856, Act of Aug. 18, 11Stat. 138; to photographs and photographic negatives in 1865, Act of Mar. 3, §1, 13Stat. 540; to motion pictures in 1912, Act of Aug. 24, 37Stat. 488; to fixed sound recordings in 1972, Act of Oct. 15, 1971, 85Stat. 391; and to architectural works in 1990, Architectural Works Copyright Protection Act, 104Stat. 5133. And on several occasions, as recounted above, Congress protected works previously in the public domain, hence freely usable by the public. See supra, at 15–19. If Congress could grant protection to these works without hazarding heightened First Amendment scrutiny, then what free speech principle disarms it from protecting works prematurely cast into the public domain for reasons antithetical to the Berne Convention? [ 33 ] Section 514, we add, does not impose a blanket prohibition on public access. Petitioners protest that fair use and the idea/expression dichotomy “are plainly inadequate to protect the speech and expression rights that Section 514 took from petitioners, or . . . the public”—that is, “the unrestricted right to perform, copy, teach and distribute the entire work, for any reason.” Brief for Petitioners 46–47. “Playing a few bars of a Shostakovich symphony,” petitioners observe, “is no substitute for performing the entire work.” Id., at 47. [ 34 ] But Congress has not put petitioners in this bind. The question here, as in Eldred, is whether would-be users must pay for their desired use of the author’s expression, or else limit their exploitation to “fair use” of that work. Prokofiev’s Peter and the Wolf could once be performed free of charge; after §514 the right to perform it must be obtained in the marketplace. This is the same marketplace, of course, that exists for the music of Prokofiev’s U. S. contemporaries: works of Copland and Bernstein, for example, that enjoy copyright protection, but nevertheless appear regularly in the programs of U. S. concertgoers. Before we joined Berne, domestic works and some foreign works were protected under U. S. statutes and bilateral international agreements, while other foreign works were available at an artificially low (because royalty-free) cost. By fully implementing Berne, Congress ensured that most works, whether foreign or domestic, would be governed by the same legal regime. The phenomenon to which Congress responded is not new: Distortions of the same order occurred with greater frequency—and to the detriment of both foreign and domestic authors—when, before 1891, foreign works were excluded entirely from U. S. copyright protection. See Kampelman, The United States and International Copyright, 41 Am. J. Int’l L. 406, 413 (1947) (“American readers were less inclined to read the novels of Cooper or Hawthorne for a dollar when they could buy a novel of Scott or Dickens for a quarter.”). Section 514 continued the trend toward a harmonized copyright regime by placing foreign works in the position they would have occupied if the current regime had been in effect when those works were created and first published. Authors once deprived of protection are spared the continuing effects of that initial deprivation; §514 gives them nothing more than the benefit of their labors during whatever time remains before the normal copyright term expires. [ 35 ] Unlike petitioners, the dissent makes much of the so-called “orphan works” problem. See post, at 11–14, 23–24. We readily acknowledge the difficulties would-be users of copyrightable materials may face in identifying or locating copyright owners. See generally U. S. Copyright Office, Report on Orphan Works 21–40 (2006). But as the dissent concedes, see post, at 13, this difficulty is hardly peculiar to works restored under §514. It similarly afflicts, for instance, U. S. libraries that attempt to catalogue U. S. books. See post, at 12. See also Brief for American Library Association et al. as Amici Curiae 22 (Section 514 “exacerbated,” but did not create, the problem of orphan works); U. S. Copyright Office, supra, at 41–44 (tracing orphan-works problem to Congress’ elimination of formalities, commencing with the 1976 Copyright Act). [ 36 ] Nor is this a matter appropriate for judicial, as opposed to legislative, resolution. Cf. Authors Guild v. Google, Inc., 770 F. Supp. 2d 666, 677–678 (SDNY 2011) (rejecting proposed “Google Books” class settlement because, inter alia, “the establishment of a mechanism for exploiting unclaimed books is a matter more suited for Congress than this Court” (citing Eldred, 537 U. S., at 212)). Indeed, the host of policy and logistical questions identified by the dissent speak for themselves. Post, at 12. Despite “longstanding efforts,” see Authors Guild, 770 F. Supp. 2d, at 678 (quoting statement of Marybeth Peters), Congress has not yet passed ameliorative orphan-works legislation of the sort enacted by other Berne members, see, e.g., Canada Copyright Act, R. S. C., 1985, c. C–42, §77 (authorizing Copyright Board to license use of orphan works by persons unable, after making reasonable efforts, to locate the copyright owner). Heretofore, no one has suggested that the orphan-works issue should be addressed through our implementation of Berne, rather than through overarching legislation of the sort proposed in Congress and cited by the dissent. See post, at 23–24; U. S. Copyright Office, Legal Issues in Mass Digitization 25–29 (2011) (discussing recent legislative efforts). Our unstinting adherence to Berne may add impetus to calls for the enactment of such legislation. But resistance to Berne’s prescriptions surely is not a necessary or proper response to the pervasive question, what should Congress do about orphan works. IV Congress determined that U. S. interests were best served by our full participation in the dominant system of international copyright protection. Those interests include ensuring exemplary compliance with our international obligations, securing greater protection for U. S. authors abroad, and remedying unequal treatment of foreign authors. The judgment §514 expresses lies well within the ken of the political branches. It is our obligation, of course, to determine whether the action Congress took, wise or not, encounters any constitutional shoal. For the reasons stated, we are satisfied it does not. The judgment of the Court of Appeals for the Tenth Circuit is therefore Affirmed. Justice Kagan took no part in the consideration or decision of this case. APPENDIX Title 17 U. S. C. §104A provides: “(a) Automatic Protection and Term.— “(1) Term.— “(A) Copyright subsists, in accordance with this section, in restored works, and vests automatically on the date of restoration. “(B) Any work in which copyright is restored under this section shall subsist for the remainder of the term of copyright that the work would have otherwise been granted in the United States if the work never entered the public domain in the United States. “(2) Exception.—Any work in which the copyright was ever owned or administered by the Alien Property Custodian and in which the restored copyright would be owned by a government or instrumentality thereof, is not a restored work. “(b) Ownership of Restored Copyright.—A restored work vests initially in the author or initial rightholder of the work as determined by the law of the source country of the work. “(c) Filing of Notice of Intent to Enforce Restored Copyright Against Reliance Parties.—On or after the date of restoration, any person who owns a copyright in a restored work or an exclusive right therein may file with the Copyright Office a notice of intent to enforce that person’s copyright or exclusive right or may serve such a notice directly on a reliance party. Acceptance of a notice by the Copyright Office is effective as to any reliance parties but shall not create a presumption of the validity of any of the facts stated therein. Service on a reliance party is effective as to that reliance party and any other reliance parties with actual knowledge of such service and of the contents of that notice. “(d) Remedies for Infringement of Restored Copyrights.— “(1) Enforcement of copyright in restored works in the absence of a reliance party.—As against any party who is not a reliance party, the remedies provided in chapter 5 of this title shall be available on or after the date of restoration of a restored copyright with respect to an act of infringement of the restored copyright that is commenced on or after the date of restoration. “(2) Enforcement of copyright in restored works as against reliance parties.—As against a reliance party, except to the extent provided in paragraphs (3) and (4), the remedies provided in chapter 5 of this title shall be available, with respect to an act of infringement of a restored copyright, on or after the date of restoration of the restored copyright if the requirements of either of the following subparagraphs are met: “(A)(i) The owner of the restored copyright (or such owner’s agent) or the owner of an exclusive right therein (or such owner’s agent) files with the Copyright Office, during the 24-month period beginning on the date of res-toration, a notice of intent to enforce the restored copyright; and “(ii)(I) the act of infringement commenced after the end of the 12-month period beginning on the date of publication of the notice in the Federal Register; “(II) the act of infringement commenced before the end of the 12-month period described in subclause (I) and continued after the end of that 12-month period, in which case remedies shall be available only for infringement occurring after the end of that 12-month period; or “(III) copies or phonorecords of a work in which copyright has been restored under this section are made after publication of the notice of intent in the Federal Register. “(B)(i) The owner of the restored copyright (or such owner’s agent) or the owner of an exclusive right therein (or such owner’s agent) serves upon a reliance party a notice of intent to enforce a restored copyright; and “(ii)(I) the act of infringement commenced after the end of the 12-month period beginning on the date the notice of intent is received; “(II) the act of infringement commenced before the end of the 12-month period described in subclause (I) and continued after the end of that 12-month period, in which case remedies shall be available only for the infringement occurring after the end of that 12-month period; or “(III) copies or phonorecords of a work in which copyright has been restored under this section are made after receipt of the notice of intent. “In the event that notice is provided under both subparagraphs (A) and (B), the 12-month period referred to in such subparagraphs shall run from the earlier of publication or service of notice. “(3) Existing derivative works.—(A) In the case of a derivative work that is based upon a restored work and is created— “(i) before the date of the enactment of the Uruguay Round Agreements Act, if the source country of the restored work is an eligible country on such date, or “(ii) before the date on which the source country of the restored work becomes an eligible country, if that country is not an eligible country on such date of enactment, “a reliance party may continue to exploit that derivative work for the duration of the restored copyright if the reliance party pays to the owner of the restored copyright reasonable compensation for conduct which would be subject to a remedy for infringement but for the provisions of this paragraph. “(B) In the absence of an agreement between the parties, the amount of such compensation shall be determined by an action in United States district court, and shall reflect any harm to the actual or potential market for or value of the restored work from the reliance party’s continued exploitation of the work, as well as compensation for the relative contributions of expression of the author of the restored work and the reliance party to the derivative work. “(4) Commencement of infringement for reliance parties.—For purposes of section 412, in the case of reliance parties, infringement shall be deemed to have commenced before registration when acts which would have constituted infringement had the restored work been subject to copyright were commenced before the date of restoration. “(e) Notices of Intent to Enforce a Restored Copyright.— “(1) Notices of intent filed with the copyright office.—(A)(i) A notice of intent filed with the Copyright Office to enforce a restored copyright shall be signed by the owner of the restored copyright or the owner of an exclusive right therein, who files the notice under subsection (d)(2)(A)(i) (hereafter in this paragraph referred to as the “owner”), or by the owner’s agent, shall identify the title of the restored work, and shall include an English translation of the title and any other alternative titles known to the owner by which the restored work may be identified, and an address and telephone number at which the owner may be contacted. If the notice is signed by an agent, the agency relationship must have been constituted in a writing signed by the owner before the filing of the notice. The Copyright Office may specifically require in regulations other information to be included in the notice, but failure to provide such other information shall not invalidate the notice or be a basis for refusal to list the restored work in the Federal Register. “(ii) If a work in which copyright is restored has no formal title, it shall be described in the notice of intent in detail sufficient to identify it. “(iii) Minor errors or omissions may be corrected by further notice at any time after the notice of intent is filed. Notices of corrections for such minor errors or omissions shall be accepted after the period established in subsection (d)(2)(A)(i). Notices shall be published in the Federal Register pursuant to subparagraph (B). “(B)(i) The Register of Copyrights shall publish in the Federal Register, commencing not later than 4 months after the date of restoration for a particular nation and every 4 months thereafter for a period of 2 years, lists identifying restored works and the ownership thereof if a notice of intent to enforce a restored copyright has been filed. “(ii) Not less than 1 list containing all notices of intent to enforce shall be maintained in the Public Information Office of the Copyright Office and shall be available for public inspection and copying during regular business hours pursuant to sections 705 and 708. “(C) The Register of Copyrights is authorized to fix reasonable fees based on the costs of receipt, processing, recording, and publication of notices of intent to enforce a restored copyright and corrections thereto. “(D)(i) Not later than 90 days before the date the Agreement on Trade-Related Aspects of Intellectual Property referred to in section 101(d)(15) of the Uruguay Round Agreements Act enters into force with respect to the United States, the Copyright Office shall issue and publish in the Federal Register regulations governing the filing under this subsection of notices of intent to enforce a restored copyright. “(ii) Such regulations shall permit owners of restored copyrights to file simultaneously for registration of the restored copyright. “(2) Notices of intent served on a reliance party.—(A) Notices of intent to enforce a restored copyright may be served on a reliance party at any time after the date of restoration of the restored copyright. “(B) Notices of intent to enforce a restored copyright served on a reliance party shall be signed by the owner or the owner’s agent, shall identify the restored work and the work in which the restored work is used, if any, in detail sufficient to identify them, and shall include an English translation of the title, any other alternative titles known to the owner by which the work may be identified, the use or uses to which the owner objects, and an address and telephone number at which the reliance party may contact the owner. If the notice is signed by an agent, the agency relationship must have been constituted in writing and signed by the owner before service of the notice. “(3) Effect of material false statements.—Any material false statement knowingly made with respect to any restored copyright identified in any notice of intent shall make void all claims and assertions made with respect to such restored copyright. “(f) Immunity From Warranty and Related Liability.— “(1) In general.—Any person who warrants, promises, or guarantees that a work does not violate an exclusive right granted in section 106 shall not be liable for legal, equitable, arbitral, or administrative relief if the war-ranty, promise, or guarantee is breached by virtue of the restoration of copyright under this section, if such warranty, promise, or guarantee is made before January 1, 1995. “(2) Performances.—No person shall be required to perform any act if such performance is made infringing by virtue of the restoration of copyright under the provisions of this section, if the obligation to perform was undertaken before January 1, 1995. “(g) Proclamation of Copyright Restoration.—Whenever the President finds that a particular foreign nation extends, to works by authors who are nationals or domiciliaries of the United States, restored copyright protection on substantially the same basis as provided under this section, the President may by proclamation extend restored protection provided under this section to any work— “(1) of which one or more of the authors is, on the date of first publication, a national, domiciliary, or sovereign authority of that nation; or “(2) which was first published in that nation. “The President may revise, suspend, or revoke any such proclamation or impose any conditions or limitations on protection under such a proclamation. “(h) Definitions.—For purposes of this section and section 109(a): “(1) The term “date of adherence or proclamation” means the earlier of the date on which a foreign nation which, as of the date the WTO Agreement enters into force with respect to the United States, is not a nation adhering to the Berne Convention or a WTO member country, becomes— “(A) a nation adhering to the Berne Convention; “(B) a WTO member country; “(C) a nation adhering to the WIPO Copyright Treaty; “(D) a nation adhering to the WIPO Performances and Phonograms Treaty; or “(E) subject to a Presidential proclamation under subsection (g). “(2) The “date of restoration” of a restored copyright is— “(A) January 1, 1996, if the source country of the restored work is a nation adhering to the Berne Convention or a WTO member country on such date, or “(B) the date of adherence or proclamation, in the case of any other source country of the restored work. “(3) The term “eligible country” means a nation, other than the United States, that— “(A) becomes a WTO member country after the date of the enactment of the Uruguay Round Agreements Act; “(B) on such date of enactment is, or after such date of enactment becomes, a nation adhering to the Berne Convention; “(C) adheres to the WIPO Copyright Treaty; “(D) adheres to the WIPO Performances and Phonograms Treaty; or “(E) after such date of enactment becomes subject to a proclamation under subsection (g). “(4) The term “reliance party” means any person who— “(A) with respect to a particular work, engages in acts, before the source country of that work becomes an eligible country, which would have violated section 106 if the restored work had been subject to copyright protection, and who, after the source country becomes an eligible country, continues to engage in such acts; “(B) before the source country of a particular work becomes an eligible country, makes or acquires 1 or more copies or phonorecords of that work; or “(C) as the result of the sale or other disposition of a derivative work covered under subsection (d)(3), or significant assets of a person described in subparagraph (A) or (B), is a successor, assignee, or licensee of that person. “(5) The term “restored copyright” means copyright in a restored work under this section. “(6) The term “restored work” means an original work of authorship that— “(A) is protected under subsection (a); “(B) is not in the public domain in its source country through expiration of term of protection; “(C) is in the public domain in the United States due to— “(i) noncompliance with formalities imposed at any time by United States copyright law, including failure of renewal, lack of proper notice, or failure to comply with any manufacturing requirements; “(ii) lack of subject matter protection in the case of sound recordings fixed before February 15, 1972; or “(iii) lack of national eligibility; “(D) has at least one author or rightholder who was, at the time the work was created, a national or domiciliary of an eligible country, and if published, was first published in an eligible country and not published in the United States during the 30-day period following publication in such eligible country; and “(E) if the source country for the work is an eligible country solely by virtue of its adherence to the WIPO Performances and Phonograms Treaty, is a sound recording. “(7) The term “rightholder” means the person— “(A) who, with respect to a sound recording, first fixes a sound recording with authorization, or “(B) who has acquired rights from the person described in subparagraph (A) by means of any conveyance or by operation of law. “(8) The “source country” of a restored work is— “(A) a nation other than the United States “(B) in the case of an unpublished work— “(i) the eligible country in which the author or rightholder is a national or domiciliary, or, if a restored work has more than 1 author or rightholder, of which the majority of foreign authors or rightholders are nationals or domiciliaries; or “(ii) if the majority of authors or rightholders are not foreign, the nation other than the United States which has the most significant contacts with the work; and “(C) in the case of a published work— “(i) the eligible country in which the work is first published, or “(ii) if the restored work is published on the same day in 2 or more eligible countries, the eligible country which has the most significant contacts with the work.” Notes 1 Article 18 of the Berne Convention provides: “(1) This Convention shall apply to all works which, at the moment of its coming into force, have not yet fallen into the public domain in the country of origin through the expiry of the term of protection. “(2) If, however, through the expiry of the term of protection which was previously granted, a work has fallen into the public domain of the country where protection is claimed, that work shall not be protected anew. “(3) The application of this principle shall be subject to any provisions contained in special conventions to that effect existing or to be concluded between countries of the Union. In the absence of such provisions, the respective countries shall determine, each in so far as it is concerned, the conditions of application of this principle. “(4) The preceding provisions shall also apply in the case of new accessions to the Union and to cases in which protection is extended by the application of Article 7 or by the abandonment of reservations.” 828 U. N. T. S. 251. 2 As noted by the Government’s amici, the United States excluded foreign works from copyright not to swell the number of unprotected works available to the consuming public, but to favor domestic publishing interests that escaped paying royalties to foreign authors. See Brief for International Publishers Association et al. as Amici Curiae 8–15. This free-riding, according to Senator Jonathan Chace, champion of the 1891 Act, made the United States “the Barbary coast of literature” and its people “the buccaneers of books.” S. Rep. No. 622, 50th Cong., 1st Sess., p. 2 (1888). 3 See also S. Rep. No. 103–412, p. 225 (1994) (“While the United States declared its compliance with the Berne Convention in 1989, it never addressed or enacted legislation to implement Article 18 of the Convention.”); Memorandum from Chris Schroeder, Counselor to the Assistant Attorney General, Office of Legal Counsel, Dept. of Justice (DOJ), to Ira S. Shapiro, General Counsel, Office of the U. S. Trade Representative (July 29, 1994), in W. Patry, Copyright and the GATT, p. C–15 (1995) (“At the time Congress was debating the BCIA, it reserved the issue of removing works from the public domain.”); General Agreement on Tariffs and Trade (GATT): Intellectual Property Provisions, Joint Hearing before the Subcommittee on Intellectual Property and Judicial Administration of the House Committee on the Judiciary and the Subcommittee on Patents, Copyrights and Trademarks of the Senate Committee on the Judiciary, 103d Cong., 2d Sess., p. 120 (1994) (URAA Joint Hearing) (app. to statement of Bruce A. Lehman, Assistant Secretary of Commerce and Commissioner of Patents and Trademarks (Commerce Dept.)) (“When the United States adhered to the Berne Convention, Congress . . . acknowledged that the possibility of restoring copyright protection for foreign works that had fallen into the public domain in the United States for failure to comply with formalities was an issue that merited further discussion.”). 4 The dissent implicitly agrees that, whatever tentative conclusion Congress reached in 1988, Article 18 requires the United States to “protect the foreign works at issue,” at least absent a special conven-tion the United States did not here negotiate. Post, at 22. Seealso post, at 23 (citing Gervais, Golan v. Holder: A Look at the Constraints Imposed by the Berne Convention, 64 Vand. L. Rev. En Banc 147, 151–152 (2011)); id., at 152 (“[T]he Convention clearly requires that some level of protection be given to foreign authors whose works have entered the public domain (other than by expiration of previous copyright).”). Accord S. Ricketson, The Berne Convention for the Protection of Literary and Artistic Works 1886–1986, p. 675 (1987) (“There is no basis on which [protection of existing works under Article 18] can be completely denied. The conditions and reservations,” authorized by Article 18(3) [and stressed by the dissent, post, at 23–24] are of “limited” and “transitional” duration and “would not be permitted to deny [protection] altogether in relation to a particular class . . . of works.”). 5 NAFTA ultimately included a limited retroactivity provision—a precursor to §514 of the URAA—granting U. S. copyright protection to certain Mexican and Canadian films. These films had fallen into the public domain, between 1978 and 1988, for failure to meet U. S. notice requirements. See North American Free Trade Agreement Implementation Act, §334, 107Stat. 2115; Brief for Franklin Pierce Center for Intellectual Property as Amicus Curiae 14–16. One year later, Congress replaced this provision with the version of 17 U. S. C. §104A at issue here. See 3 M. Nimmer & D. Nimmer, Copyright §9 A. 03, 9 A. 04, pp. 9A–17, 9A–22 (2011) (hereinafter Nimmer). 6 This tension between the United States and its new Berne counterparties calls into question the dissent’s assertion that, despite the 1988 Act’s minimalist approach, “[t]he United States obtained the benefits of Berne for many years.” Post, at 22–23. During this six-year period, Congress had reason to doubt that U. S. authors enjoyed the full benefits of Berne membership. 7 Marrakesh Agreement Establishing the World Trade Organization, Apr. 15, 1994, 1867 U. N. T. S. 154. 8 Proponents of prompt congressional action urged that avoiding a trade enforcement proceeding—potentially the WTO’s first—would be instrumental in preserving the United States’ “reputation as a world leader in the copyright field.” URAA Joint Hearing 241 (statement of Eric Smith, International Intellectual Property Alliance (IIPA)). In this regard, U. S. negotiators reported that widespread perception of U. S. noncompliance was undermining our leverage in copyright negotiations. Unimpeachable adherence to Berne, Congress was told, would help ensure enhanced foreign protection, and hence profitable dissemination, for existing and future U. S. works. See id., at 120 (app. to statement of Lehman, Commerce Dept.) (“Clearly, providing for [retroactive] protection for existing works in our own law will improve our position in future negotiations.”); id., at 268 (statement of Berman, RIAA). 9 Title 17 U. S. C. §104A is reproduced in full in an appendix to this opinion. 10 Works from most, but not all, foreign countries are eligible for protection under §514. The provision covers only works that have “at least one author or rightholder who was, at the time the work was created, a national or domiciliary of an eligible country.” 17 U. S. C. §104A(h)(6)(D). An “eligible country” includes any “nation, other than the United States, that—(A) becomes a WTO member country after the date of the enactment of the [URAA]; [or] (B) on such date of enactment is, or after such date of enactment becomes, a nation adhering tothe Berne Convention.” §104A(h)(3). As noted above, see supra, at 3, 164 countries adhere to the Berne Convention. World Intellec-tual Property Organization, Contracting Parties: Berne Convention, www.wipo.int/treaties (as visited Jan. 13, 2012, and in Clerk of Court’s case file). 11 From the first Copyright Act until late in the 20th century, Congress conditioned copyright protection on compliance with certain statutory formalities. The most notable required an author to register her work, renew that registration, and affix to published copies notice of copyrighted status. The formalities drew criticism as a trap for the unwary. See, e.g., 2 Nimmer §7.01[A], p. 7–8; Doyle, Cary, McCannon, & Ringer, Notice of Copyright, Study No. 7, p. 46 (1957), reprinted in 1 Studies on Copyright 229, 272 (1963). In 1976, Congress eliminated the registration renewal requirement for future works. Copyright Act of 1976, §302, 408, 90Stat. 2572, 2580. In 1988, it repealed the mandatory notice prerequisite. BCIA §7, 102Stat. 2857. And in 1992, Congress made renewal automatic for works still in their first term of protection. Copyright Amendments Act of 1992, 106Stat. 264–266. The Copyright Act retains, however, incentives for authors to register their works and provide notice of the works’ copyrighted status. See, e.g., 17 U. S. C. §405(b) (precluding actual and statutory damages against “innocent infringers” of a work that lacked notice of copyrighted status); §411(a) (requiring registration of U. S. “work[s],” but not foreign works, before an owner may sue for infringement). The revisions successively made accord with Berne Convention Article 5(2), which proscribes application of copyright formalities to foreign authors. Berne, however, affords domestic authors no escape from domestic formalities. See Art. 5(3) (protection within country of origin is a matter of domestic law). 12 Title 17 U. S. C. §104A(h)(6)(B) defines a “restored work” to exclude “an original work of authorship” that is “in the public domain in its source country through expiration of [its] term of protection.” This provision tracks Berne’s denial of protection for any work that has “fallen into the public domain in the country of origin through the expiry of the term of protection.” Art. 18(1), 828 U. N. T. S., at 251. 13 Restoration is a misnomer insofar as it implies that all works protected under §104A previously enjoyed protection. Each work in the public domain because of lack of national eligibility or subject-matter protection, and many that failed to comply with formalities, never enjoyed U. S. copyright protection. See, e.g., 3 Nimmer §9A.04[A][1][b][iii], at 9A–26, and n. 29.4. 14 A reliance party must have used the work in a manner that would constitute infringement had a valid copyright been in effect. See §104A(h)(4)(A). After restoration, the reliance party is limited to her previous uses. A performer of a restored work, for example, cannot, post-restoration, venture to sell copies of the script. See 3 Nimmer §9A.04[C][1][a], at 9A–45 to 9A–46. 15 Petitioners’ complaint also challenged the constitutionality of the Copyright Term Extension Act, 112Stat. 2827, which added 20 years to the duration of existing and future copyrights. After this Court rejected a similar challenge in Eldred v. Ashcroft, 537 U.S. 186 (2003) , the District Court dismissed this portion of petitioners’ suit on the pleadings, Golan v. Ashcroft, 310 F. Supp. 2d 1215 (D. Colo. 2004). The Tenth Circuit affirmed, Golan v. Gonzales, 501 F.3d 1179 (2007), and petitioners do not attempt to revive that claim in this Court, Pet. for Cert. 7, n. 2. Neither have petitioners challenged the District Court’s entry of summary judgment for the Government on the claim that §514 violates the substantive component of the Due Process Clause. 16 Cf. 3 Nimmer §9A.02[A][2], at 9A–11, n. 28 (“[I]t stretches the language of the Berne Convention past the breaking point to posit that following ‘expiry of the zero term’ the . . . work need not be resurrected.”). 17 See B. Bugbee, Genesis of American Patent and Copyright Law 123–124 (1967) (hereinafter Bugbee) (Delaware, Maryland, and Pennsylvania). 18 See 1783 Mass. Acts p. 236; 1783 N. J. Laws p. 47; 1783 N. H. Laws p. 521; 1783 R. I. Laws pp. 6–7; 1784 S. C. Acts p. 49; 1785 Va. Acts ch. VI; 1786 N. Y. Laws p. 298. 19 1783 Conn. Pub. Acts no. 617; 1783 N. J. Laws p. 47; 1785 N. C. Laws p. 563; 1786 Ga. Laws p. 323. In four States, copyright enforcement was restricted to works “not yet printed” or “hereinafter published.” 1783 Mass. Acts p. 236; 1783 N. H. Laws p. 521; 1783 R. I. Laws pp. 6–7; 1784 S. C. Acts p. 49. 20 See Bugbee 109–123. 21 The parties debate the extent to which the First Congress removed works from the public domain. We have held, however, that at least some works protected by the 1790 Act previously lacked protection. In Wheaton v. Peters, 8 Pet. 591 (1834), the Court ruled that before enactment of the 1790 Act, common-law copyright protection expired upon first publication. Id., at 657, 663. Thus published works covered by the 1790 Act previously would have been in the public domain unless protected by state statute. Had the founding generation perceived the constitutional boundary petitioners advance today, the First Congress could have designed a prospective scheme that left the public domain undisturbed. Accord Luck’s Music Library, Inc. v. Gonzales, 407 F.3d 1262, 1265 (CADC 2005) (Section 514 does not offend the Copyright Clause because, inter alia, “evidence from the First Congress,” as confirmed by Wheaton, “points toward constitutionality.”). 22 Here, as in Eldred, “[b]ecause the Clause empowering Congress to confer copyrights also authorizes patents, congressional practice with respect to patents informs our inquiry.” 537 U. S., at 201. 23 Section 514 is in line with these measures; like them, it accords protection to works that had lapsed into the public domain because of failure to comply with U. S. statutory formalities. See supra, at 9, and n. 11. 24 Legislation of this order, petitioners argue, is best understood as an exercise of Congress’ power to remedy excusable neglect. Even so, the remedy sheltered creations that, absent congressional action, would have been open to free exploitation. Such action, according to petitioners’ dominant argument, see supra, at 13–14, is ever and always impermissible. Accord Luck’s Music Library, 407 F. 3d, at 1265–1266 (“Plaintiffs urge that [the 1790 Act and the wartime legislation] simply extended the time limits for filing and [did] not purport to modify the prohibition on removing works from the public domain. But to the extent that potential copyright holders failed to satisfy procedural requirements, such works”—like those protected by §514—“would necessarily have already entered the public domain . . . .”). 25 But see Brief for Motion Picture Association of America as Amicus Curiae 27 (observing that income from existing works can finance the creation and publication of new works); Eldred, 537 U. S., at 208, n. 15 (noting that Noah Webster “supported his entire family from the earnings on his speller and grammar during the twenty years he took to complete his dictionary” (internal quotation marks omitted)). 26 The dissent also suggests, more tentatively, that at least where copyright legislation extends protection to works previously in the public domain, Congress must counterbalance that restriction with new incentives to create. Post, at 8. Even assuming the public domain were a category of constitutional significance, contra supra, at 13–19, we would not understand “the Progress of Science” to have this contingent meaning. 27 That the same economic incentives might also induce the dissemination of futons, fruit, or Bibles, see post, at 20, is no answer to this evidence that legislation furthering the dissemination of literary property has long been thought a legitimate way to “promote the Progress of Science.” 28 The dissent suggests that the “utilitarian view of copyrigh[t]” embraced by Jefferson, Madison, and our case law sets us apart from continental Europe and inhibits us from harmonizing our copyright laws with those of countries in the civil-law tradition. See post, at 5–6, 22. For persuasive refutation of that suggestion, see Austin, Does the Copyright Clause Mandate Isolationism? 26 Colum. J. L. & Arts 17, 59 (2002) (cautioning against “an isolationist reading of the Copyright Clause that is in tension with . . . America’s international copyright relations over the last hundred or so years”). 29 On the initial appeal in this case, the Tenth Circuit gave an unconfined reading to our reference in Eldred to “traditional contours of copyright.” 501 F. 3d, at 1187–1196. That reading was incorrect, as we here clarify. 30 See Eldred, 537 U. S., at 221 (“Protection of [an author’s original expression from unrestricted exploitation] does not raise the free speech concerns present when the government compels or burdens the communication of particular facts or ideas.”). 31 Focusing narrowly on the specific problem of orphan works, the dissent overlooks these principal protections against “the dissemination-restricting harms of copyright.” Post, at 14. 32 “[R]equir[ing] works that have already fallen into the public domain to stay there” might, as the dissent asserts, supply an “easily administrable standard.” Post, at 14. However attractive this bright-line rule might be, it is not a rule rooted in the constitutional text or history. Nor can it fairly be gleaned from our case law. The dissent cites three decisions to document its assertion that “this Court has assumed the particular importance of public domain material in roughly analogous circumstances.” Post, at 15. The dictum in Graham v. John Deere Co. of Kansas City, 383 U.S. 1, 6 (1966) , noted earlier, did not treat the public domain as a constitutional limit—certainly not under the rubric of the First Amendment. See supra, at 19. The other two decisions the dissent cites considered whether the federal Patent Act preempted a state trade-secret law, Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470 –484 (1974), and whether the freedom of the press shielded reporters from liability for publishing material drawn from public court documents, Cox Broadcasting Corp. v. Cohn, 420 U.S. 469 –497 (1975). Neither decision remotely ascribed constitutional significance to a work’s public domain status. 33 It was the Fifth Amendment’s Takings Clause—not the First Amendment—that Congress apparently perceived to be a potential check on its authority to protect works then freely available to the public. See URAA Joint Hearing 3 (statement of Rep. Hughes); id., at 121 (app. to statement of Lehman, Commerce Dept.); id., at 141 (statement of Shapiro, USTR); id., at 145 (statement of Christopher Schroe-der, DOJ). The reliance-party protections supplied by §514, see supra, at 10–11, were meant to address such concerns. See URAA Joint Hearing 148–149 (prepared statement of Schroeder). 34 Because Shostakovich was a pre-1973 Russian composer, his works were not protected in the United States. See U. S. Copyright Office, Circular No. 38A: The International Copyright Relations of the United States 9, 11, n. 2 (2010) (copyright relations between the Soviet Union and the United States date to 1973). 35 Persistently deploring “ ‘restored copyright’ protection [because it] removes material from the public domain,” post, at 14, the dissent does not pause to consider when and why the material came to be lodged in that domain. Most of the works affected by §514 got there after a term of zero or a term cut short by failure to observe U. S. formalities. See supra, at 9. 36 The pervasive problem of copyright piracy, noted post, at 13, likewise is scarcely limited to protected foreign works formerly in the public domain.

565.US.2011_10-895

After the intermediate state appellate court affirmed his state-court conviction, petitioner Gonzalez allowed his time for seeking discretionary review with the State’s highest court for criminal appeals to expire. Roughly six weeks later, the intermediate state appellate court issued its mandate. When Gonzalez subsequently sought federal habeas relief, the District Court dismissed Gonzalez’s petition as time barred by the 1-year statute of limitations in the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA). Under 28 U. S. C. §2244(d)(1)(A), state prisoners have one year to file federal habeas petitions running from “the date on which the judgment became final by the conclusion of direct review or the expiration of the time for seeking such review.” The District Court held that Gonzalez’s judgment had become “final” when his time for seeking discretionary review in the State’s highest court expired, and that running the limitations period from that date, his petition was untimely. Under AEDPA, a habeas petitioner must obtain a certificate of appealability (COA) to appeal a district court’s final order in a habeas proceeding. 28 U. S. C. §2253(c)(1). The COA may issue only if the petitioner has made a “substantial showing of the denial of a constitutional right,” §2253(c)(2), and “shall indicate which specific issue” satisfies that showing, §2253(c)(3). A Fifth Circuit judge granted Gonzalez a COA on the question whether his petition was timely. The issued COA, however, failed to “indicate” a constitutional issue. The Fifth Circuit affirmed, holding that Gonzalez’s petition was untimely because the limitations period begins to run for petitioners who fail to appeal to a State’s highest court when the time for seeking further direct review in the state court expires. The Fifth Circuit did not mention, and the State did not raise, the §2253(c)(3) defect. When Gonzalez petitioned this Court for review, the State argued for the first time that the Fifth Circuit lacked jurisdiction to adjudicate Gonzalez’s appeal based on the §2253(c)(3) defect. Held: 1. Section 2253(c)(3) is a mandatory but nonjurisdictional rule. A COA’s failure to “indicate” a constitutional issue does not deprive a Court of Appeals of jurisdiction to adjudicate the appeal. Pp. 4−13. (a) A rule is jurisdictional “[i]f the Legislature clearly states that a threshold limitation on a statute’s scope shall count as jurisdictional,” Arbaugh v. Y & H Corp., 546 U.S. 500, 515. Here, the only clear jurisdictional language in §2253(c) appears in §2253(c)(1). The parties agree that §2253(c)(1)’s plain terms make the issuance of a COA a jurisdictional prerequisite. The parties also agree that §2253(c)(2), which speaks only to when a COA may issue and does not contain §2253(c)(1)’s jurisdictional terms, is nonjurisdictional. It follows that §2253(c)(3) is also nonjurisdictional. Like §2253(c)(2), it reflects a threshold condition for issuing a COA, and “does not speak in jurisdictional terms or refer . . . to the jurisdiction of the [appeals] courts.” Arbaugh, 546 U. S., at 515. Jurisdictional treatment also would thwart Congress’s intent in AEDPA “to eliminate delays in the federal habeas review process.” Holland v. Florida, 560 U. S. ___, ___. Once a judge has determined that a COA is warranted and resources are deployed in briefing and argument, the COA has fulfilled its gatekeeping function. Pp. 4−9. (b) The State’s contrary arguments are unpersuasive. Section 2253(c)(3)’s cross-reference to §2253(c)(1) does not mean §2253(c)(3) can be read as part of §2253(c)(1), as Congress set off the requirements in distinct paragraphs with distinct terms. The word “shall” in §2253(c)(3), meanwhile, underscores the rule’s mandatory nature, but not all mandatory rules are jurisdictional. Nor does §2253(c)(3)’s mere proximity to other jurisdictional provisions turn a rule that speaks in nonjurisdictional terms into a jurisdictional hurdle. Finally, the Court rejects the State’s attempt to analogize a COA to a notice of appeal. Pp. 10−13. 2. For a state prisoner who does not seek review in a State’s highest court, the judgment becomes “final” for purposes of §2244(d)(1)(A) on the date that the time for seeking such review expires. Pp. 13−19. (a) In Clay v. United States, 537 U.S. 522, the Court held that a federal conviction becomes final “when this Court affirms a conviction on the merits on direct review or denies a petition for a writ of certiorari,” or, if a petitioner does not seek certiorari, “when the time for filing a certiorari petition expires.” Id., at 527. In Jimenez v. Quarterman, 555 U.S. 113, the Court adopted Clay’s “most natural reading of the statutory text” in construing “the similar language of §2244(d)(1)(A).” Id., at 119. The Court made no mention of when Jimenez’s appeal concluded and held that his judgment became final when his time for seeking certiorari expired. Section 2244(d)(1)(A) thus consists of two prongs corresponding to two categories of petitioners. For petitioners pursuing direct review all the way to this Court, the judgment becomes final at the “conclusion of direct review,” when this Court affirms a conviction on the merits or denies certiorari. For all other petitioners, the judgment becomes final at the “expiration of the time for seeking such review,” when the time for pursuing direct review in this Court, or in state court, expires. Because Gonzalez did not appeal to the State’s highest court, his judgment became final when his time for seeking review with that court expired. Pp. 13–15. (b) Gonzalez argues that courts should determine both prongs for every petitioner who does not seek certiorari, then start the 1-year clock from the latest of the two dates. Gonzalez further contends that when a petitioner does not seek certiorari, state law should define the “conclusion of direct review.” The words “latest of,” however, appear in §2244(d)(1), not §2244(d)(1)(A). Nothing in §2244(d)(1)(A) contemplates any conflict between the two prongs or instructs that the later of the two shall prevail. Gonzalez’s approach of scouring each State’s laws and cases to determine how the State defines finality, moreover, would contradict the uniform meaning of “conclusion of direct review” that Clay and Jimenez accepted. It will be a rare situation in which a delay in the mandate’s issuance is so severe as to prevent a petitioner from filing a federal habeas petition within a year or requesting a stay and abeyance. Finally, the Court rejects Gonzalez’s alternative argument that his petition is timely because it was filed within a year of when his time for seeking certiorari review expired. Pp. 15−19. 623 F.3d 222, affirmed. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, Ginsburg, Breyer, Alito, and Kagan, JJ., joined. Scalia, J., filed a dissenting opinion.

This case interprets two provisions of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA). The first, 28 U. S. C. §2253(c), provides that a habeas peti- tioner must obtain a certificate of appealability (COA) to appeal a federal district court’s final order in a habeas proceeding. §2253(c)(1). The COA may issue only if the petitioner has made a “substantial showing of the denial of a constitutional right,” §2253(c)(2), and “shall indicate which specific issue” satisfies that showing. §2253(c)(3). We hold that §2253(c)(3) is not a jurisdictional requirement. Accordingly, a judge’s failure to “indicate” the requisite constitutional issue in a COA does not deprive a court of appeals of subject-matter jurisdiction to adjudicate the habeas petitioner’s appeal. The second provision, 28 U. S. C. §2244(d)(1)(A), establishes a 1-year limitations period for state prisoners to file federal habeas petitions, running from “the date on which the judgment became final by the conclusion of direct review or the expiration of the time for seeking such review.” We hold that, for a state prisoner who does not seek review in a State’s highest court, the judgment becomes “final” on the date that the time for seeking such review expires. I Petitioner Rafael Gonzalez was convicted of murder in Texas state court. The intermediate state appellate court, the Texas Court of Appeals, affirmed Gonzalez’s conviction on July 12, 2006. Gonzalez then allowed his time for seeking discretionary review with the Texas Court of Criminal Appeals (Texas CCA)—the State’s highest court for criminal appeals—to expire on August 11, 2006. Tex. Rule App. Proc. 68.2(a) (2011). The Texas Court of Appeals issued its mandate on September 26, 2006. After Gonzalez, proceeding pro se, petitioned unsuccessfully for state habeas relief, he filed a federal habeas petition under 28 U. S. C. §2254 on January 24, 2008, in the U. S. District Court for the Northern District of Texas. His petition alleged, inter alia, that the nearly 10-year delay between his indictment and trial violated his Sixth Amendment right to a speedy trial. The District Court, without discussing Gonzalez’s constitutional claims, dismissed Gonzalez’s petition as time barred by the 1-year statute of limitations in §2244(d)(1)(A). Although Gonzalez argued that his judgment had not become final until the Texas Court of Appeals issued its mandate, the District Court held that Gonzalez’s judgment had become final when his time for seeking discretionary review in the Texas CCA expired on August 11, 2006. Counting from that date, and tolling the limitations period for the time during which Gonzalez’s state habeas petition was pending, Gonzalez’s limitations period elapsed on December 17, 2007—over a month before he filed his federal habeas petition. The District Court denied a COA. Gonzalez applied to the U. S. Court of Appeals for the Fifth Circuit for a COA on two grounds: (1) his habeas petition was timely, and (2) his Sixth Amendment speedy-trial right was violated. A Court of Appeals judge granted a COA on the question “whether the habeas application was timely filed because Gonzalez’s conviction became final, and thus the limitations period commenced, on the date the intermediate state appellate court issued its mandate.” App. 347. The COA did not mention the Sixth Amendment question. The Court of Appeals affirmed. 623 F.3d 222 (2010). Acknowledging that a sister Circuit had run the limitations period from the date of a state court’s issuance of a mandate, the Court of Appeals deemed the mandate’s issuance “irrelevant” to determining finality under §2244(d)(1)(A). Id., at 224, 226 (disagreeing with Riddle v. Kemna, 523 F.3d 850 (CA8 2008) (en banc)). The Court of Appeals held that because a judgment becomes final at “the conclusion of direct review or the expiration of the time for seeking such review,” §2244(d)(1)(A), the limitations period begins to run for petitioners who fail to appeal to a State’s highest court when the time for seeking further direct review in the state court expires. The Court of Appeals therefore concluded that Gonzalez’s conviction became final on August 11, 2006, and his habeas petition was time barred. The Court of Appeals did not address Gonzalez’s Sixth Amendment claim or discuss whether the COA had been improperly issued. Nor did the State allege any defect in the COA or move to dismiss for lack of jurisdiction. Gonzalez petitioned this Court for a writ of certiorari. In its brief in opposition, the State argued for the first time that the Court of Appeals lacked jurisdiction to adjudicate Gonzalez’s appeal because the COA identified only a procedural issue, without also “indicat[ing]” a constitutional issue as required by §2253(c)(3). We granted certiorari to decide two questions, both of which implicate splits in authority: (1) whether the Court of Appeals had jurisdiction to adjudicate Gonzalez’s appeal, notwithstanding the §2253(c)(3) defect; [ 1 ] and (2) whether Gonzalez’s habeas petition was time barred under §2244(d)(1) due to the date on which his judgment became final. [ 2 ] 564 U. S. ___ (2011). II We first consider whether the Court of Appeals had juris- diction to adjudicate Gonzalez’s appeal. A Section 2253, as amended by AEDPA, governs appeals in habeas corpus proceedings. The first subsection, §2253(a), is a general grant of jurisdiction, providing that district courts’ final orders in habeas proceedings “shall be subject to review, on appeal, by the court of appeals.” 28 U. S. C. §2253(a). The second, §2253(b), limits jurisdiction over a particular type of final order. See §2253(b) (“There shall be no right of appeal from a final order in a proceeding to test the validity of a warrant [of] remov[al] . . .”). This case concerns the third, §2253(c), which provides: “(1) Unless a circuit justice or judge issues a certificate of appealability, an appeal may not be taken to the court of appeals . . . . . . . . “(2) A certificate of appealability may issue under paragraph (1) only if the applicant has made a substantial showing of the denial of a constitutional right. “(3) The certificate of appealability under paragraph (1) shall indicate which specific issue or issues satisfy the showing required by paragraph (2).” When, as here, the district court denies relief on procedural grounds, the petitioner seeking a COA must show both “that jurists of reason would find it debatable whether the petition states a valid claim of the denial of a constitutional right and that jurists of reason would find it debatable whether the district court was correct in its procedural ruling.” Slack v. McDaniel, 529 U.S. 473, 484 (2000) . In this case, the Court of Appeals judge granted a COA that identified a debatable procedural ruling, but did not “indicate” the issue on which Gonzalez had made a substantial showing of the denial of a constitutional right, as required by §2253(c)(3). The question before us is whether that defect deprived the Court of Appeals of the power to adjudicate Gonzalez’s appeal. We hold that it did not. This Court has endeavored in recent years to “bring some discipline” to the use of the term “jurisdictional.” Henderson v. Shinseki, 562 U. S. ___, ___ (2011) (slip op., at 5). Recognizing our “less than meticulous” use of the term in the past, we have pressed a stricter distinction between truly jurisdictional rules, which govern “a court’s adjudicatory authority,” and nonjurisdictional “claim-processing rules,” which do not. Kontrick v. Ryan, 540 U.S. 443 –455 (2004). When a requirement goes to subject-matter jurisdiction, courts are obligated to consider sua sponte issues that the parties have disclaimed or have not presented. See United States v. Cotton, 535 U.S. 625, 630 (2002) . Subject-matter jurisdiction can never be waived or forfeited. The objections may be resurrected at any point in the litigation, and a valid objection may lead a court midway through briefing to dismiss a complaint in its entirety. “[M]any months of work on the part of the attorneys and the court may be wasted.” Henderson, 562 U. S., at ___ (slip op., at 5). Courts, we have said, should not lightly attach those “drastic” consequences to limits Congress has enacted. Ibid. We accordingly have applied the following principle: A rule is jurisdictional “[i]f the Legislature clearly states that a threshold limitation on a statute’s scope shall count as jurisdictional.” Arbaugh v. Y & H Corp., 546 U.S. 500, 515 (2006) . But if “Congress does not rank a statutory limitation on coverage as jurisdictional, courts should treat the restriction as nonjurisdictional.” Id., at 516. [ 3 ] That clear-statement principle makes particular sense in this statute, as we consider—against the backdrop of §2253(a)’s clear jurisdictional grant to the courts of appeals and §2253(b)’s clear limit on that grant—the extent to which Congress intended the COA process outlined in §2253(c) to further limit the courts of appeals’ jurisdiction over habeas appeals. Here, the only “clear” jurisdictional language in §2253(c) appears in §2253(c)(1). As we explained in Miller-El v. Cockrell, 537 U.S. 322 (2003) , §2253(c)(1)’s plain terms—“Unless a circuit justice or judge issues a certificate of appealability, an appeal may not be taken to the court of appeals”—establish that “until a COA has been issued federal courts of appeals lack jurisdiction to rule on the merits of appeals from habeas petitioners.” Id., at 336. The parties thus agree that §2253(c)(1) is jurisdictional. The parties also agree that §2253(c)(2) is nonjurisdictional. [ 4 ] That is for good reason. Section 2253(c)(2) speaks only to when a COA may issue—upon “a substantial showing of the denial of a constitutional right.” It does not contain §2253(c)(1)’s jurisdictional terms. See Russello v. United States, 464 U.S. 16, 23 (1983) (“[W]here Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally . . .”). And it would be passing strange if, after a COA has issued, each court of appeals adjudicating an appeal were dutybound to revisit the threshold showing and gauge its “substantial[ity]” to verify its jurisdiction. That inquiry would be largely duplicative of the merits question before the court. It follows that §2253(c)(3) is nonjurisdictional as well. Like §2253(c)(2), it too reflects a threshold condition for the issuance of a COA—the COA’s indication of “which specific issue or issues satisfy the showing required by paragraph (2).” It too “does not speak in jurisdictional terms or refer in any way to the jurisdiction of the [appeals] courts.” Arbaugh, 546 U. S., at 515 (internal quo- tation marks omitted). The unambiguous jurisdictional terms of §§2253(a), (b), and (c)(1) show that Congress would have spoken in clearer terms if it intended §2253(c)(3) to have similar jurisdictional force. Instead, the contrast underscores that the failure to obtain a COA is jurisdictional, while a COA’s failure to indicate an issue is not. A defective COA is not equivalent to the lack of any COA. It is telling, moreover, that Congress placed the power to issue COAs in the hands of a “circuit justice or judge.” [ 5 ] It would seem somewhat counterintuitive to render a panel of court of appeals judges powerless to act on appeals based on COAs that Congress specifically empowered one court of appeals judge to grant. Indeed, whereas §2253(c)(2)’s substantial-showing requirement at least de- scribes a burden that “the applicant” seeking a COA bears, §2253(c)(3)’s indication requirement binds only the judge issuing the COA. Notably, Gonzalez advanced both the timeliness and Sixth Amendment issues in his application for a COA. A petitioner, having successfully obtained a COA, has no control over how the judge drafts the COA and, as in Gonzalez’s case, may have done everything required of him by law. That fact would only compound the “unfai[r] prejudice” resulting from the sua sponte dismissals and remands that jurisdictional treatment would entail. Henderson, 562 U. S., at ___ (slip op., at 5). [ 6 ] Treating §2253(c)(3) as jurisdictional also would thwart Congress’ intent in AEDPA “to eliminate delays in the federal habeas review process.” Holland v. Florida, 560 U. S. ___, ___ (2010) (slip op., at 16). The COA process screens out issues unworthy of judicial time and attention and ensures that frivolous claims are not assigned to merits panels. Once a judge has made the determination that a COA is warranted and resources are deployed in briefing and argument, however, the COA has fulfilled that gatekeeping function. Even if additional screening of already-issued COAs for §2253(c)(3) defects could further winnow the cases before the courts of appeals, that would not outweigh the costs of further delay from the extra layer of review. This case, in which the alleged defect would be dispositive, exemplifies those inefficiencies; the State requests that we vacate and remand with instructions to dismiss the appeal based on a §2253(c)(3) defect that it raised for the first time in response to a petition for certiorari. And delay would be particularly fruitless in the numerous cases where, as here, the district court dismissed the petition on procedural grounds and the court of appeals affirms, without having to address the omitted constitutional issue at all. B The State, aided by the United States as amicus curiae, makes several arguments in support of jurisdictional treatment of §2253(c)(3). None is persuasive. First, the State notes that although §2253(c)(3) does not speak in jurisdictional terms, it refers back to §2253(c)(1), which does. The State argues that it is as if §2253(c)(1) provided: “Unless a circuit justice or judge issues a certificate of appealability that shall indicate the specific issue or issues that satisfy the showing required by paragraph (2), an appeal may not be taken to the court of appeals.” The problem is that the statute provides no such thing. Instead, Congress set off the requirements in distinct paragraphs and, rather than mirroring their terms, excluded the jurisdictional terms in one from the other. Notably, the State concedes that §2253(c)(2) is nonjurisdictional, even though it too cross-references §2253(c)(1) and is cross-referenced by §2253(c)(3). Second, the State seizes on the word “shall” in §2253(c)(3), arguing that an omitted indication renders the COA no COA at all. But calling a rule nonjurisdictional does not mean that it is not mandatory or that a timely objection can be ignored. If a party timely raises the COA’s failure to indicate a constitutional issue, the court of appeals panel must address the defect by considering an amendment to the COA or remanding to the district judge for specification of the issues. [ 7 ] This Court, moreover, has long “rejected the notion that ‘all mandatory prescriptions, however emphatic, are . . . properly typed jurisdictional.’ ” Henderson, 562 U. S., at ___ (slip op., at 9); see also Dolan v. United States, 560 U. S. ___, ___ (2010) (slip op., at 5) (statute’s reference to “shall” alone does not render statutory deadline jurisdictional). Nothing in §2253(c)(3)’s prescription establishes that an omitted indication should remain an open issue throughout the case. Third, the United States argues that the placement of §2253(c)(3) in a section containing jurisdictional provisions signals that it too is jurisdictional. In characterizing certain requirements as nonjurisdictional, we have on occasion observed their “ ‘separat[ion]’ ” from jurisdictional provisions. E.g., Reed Elsevier, Inc. v. Muchnick, 559 U. S. ___, ___ (2010) (slip op., at 7); Arbaugh, 546 U. S., at 515. The converse, however, is not necessarily true: Mere proximity will not turn a rule that speaks in nonjurisdictional terms into a jurisdictional hurdle. In fact, §2253(c)(3)’s proximity to §§2253(a), (b), and (c)(1) highlights the absence of clear jurisdictional terms in §2253(c)(3). Finally, the State analogizes a COA to a notice of appeal, pointing out that both a notice and its contents are jurisdictional prerequisites. Federal Rule of Appellate Procedure 3(c)(1) provides that a notice of appeal must: “(A) specify the party or parties taking the appeal”; “(B) designate the judgment, order, or part thereof being appealed”; and “(C) name the court to which the appeal is taken.” We have held that “Rule 3’s dictates are jurisdictional in nature.” Smith v. Barry, 502 U.S. 244, 248 (1992) . We reject this analogy. We construed the content requirements for notices of appeal as jurisdictional because we were “convinced that the harshness of our construction [wa]s ‘imposed by the legislature.’ ” Torres v. Oakland Scavenger Co., 487 U.S. 312, 318 (1988) . Rule 4, we noted, establishes mandatory time limits for filing a notice of appeal. Excusing a failure to name a party in a notice of appeal, in violation of Rule 3, would be “equivalent to permitting courts to extend the time for filing a notice of appeal,” in violation of Rule 4. Id., at 315. And “time limits for filing a notice of appeal have been treated as jurisdictional in American law for well over a century.” Bowles v. Russell, 551 U.S. 205 , n. 2 (2007). Accordingly, the Advisory Committee Note “makes no distinction among the various requirements of Rule 3 and Rule 4,” treating them “as a single jurisdictional threshold.” Torres, 487 U. S., at 315; see also id., at 316 (“the Advisory Committee viewed the requirements of Rule 3 as jurisdictional in nature”). Here, we find no similar basis for treating the paragraphs of §2253(c) as a single jurisdictional threshold. Moreover, in explaining why the naming requirement was jurisdictional in Torres, we reasoned that an unnamed party leaves the notice’s “intended recipient[s]”—the appellee and court—“unable to determine with certitude whether [that party] should be bound by an adverse judgment or held liable for costs or sanctions.” Id., at 318. The party could sit on the fence, await the outcome, and opt to participate only if it was favorable. That possibility of gamesmanship is not present here. Unlike the party who fails to submit a compliant notice of appeal, the habeas petitioner who obtains a COA cannot control how that COA is drafted. [ 8 ] And whereas a party’s failure to be named in a notice of appeal gives absolutely no “notice of [his or her] appeal,” a judge’s issuance of a COA reflects his or her judgment that the appeal should proceed and supplies the State with notice that the habeas litigation will continue. Because we conclude that §2253(c)(3) is a nonjurisdictional rule, the Court of Appeals had jurisdiction to adjudicate Gonzalez’s appeal. III We next consider whether Gonzalez’s habeas petition was time barred. AEDPA establishes a 1-year limitations period for state prisoners to file for federal habeas relief, which “run[s] from the latest of” four specified dates. [ 9 ] §2244(d)(1). This case concerns the first of those dates: “the date on which the judgment became final by the conclusion of direct review or the expiration of the time for seeking such review.” §2244(d)(1)(A). The question before us is when the judgment becomes “final” if a petitioner does not appeal to a State’s highest court. A In construing the language of §2244(d)(1)(A), we do not write on a blank slate. In Clay v. United States, 537 U.S. 522 (2003) , we addressed AEDPA’s statute of limitations for federal prisoners seeking postconviction relief. See §2255(f)(1) (2006 ed., Supp. III) (beginning 1-year period of limitations from “the date on which the judgment of conviction becomes final”). We held that the federal judgment becomes final “when this Court affirms a conviction on the merits on direct review or denies a petition for a writ of certiorari,” or, if a petitioner does not seek certiorari, “when the time for filing a certiorari petition expires.” Id., at 527. In so holding, we rejected the argument that, if a petitioner declines to seek certiorari, the limitations period “starts to run on the date the court of appeals issues its mandate.” Id., at 529. In Jimenez v. Quarterman, 555 U.S. 113 (2009) , we described Clay’s interpretation as comporting “with the most natural reading of the statutory text” and saw “no reason to depart” from it in “construing the similar language of §2244(d)(1)(A).” 555 U. S., at 119. The state court had permitted Jimenez to file an out-of-time direct appeal. We held that this “reset” the limitations period; Jimenez’s judgment would now become final at “the conclusion of the out-of-time direct appeal, or the expiration of the time for seeking review of that [out-of-time] appeal.” Id., at 120–121. Because Jimenez did not seek certiorari, we made no mention of when the out-of-time appeal “conclu[ded].” Rather, we held that his judgment became final when his “time for seeking certiorari review in this Court expired.” Id., at 120. Nor did we mention the date on which the state court issued its mandate. Both Clay and Jimenez thus suggested that the direct review process either “concludes” or “expires,” depending on whether the petitioner pursues or forgoes direct appeal to this Court. We now make clear what we suggested in those cases: The text of §2244(d)(1)(A), which marks finality as of “the conclusion of direct review or the expiration of the time for seeking such review,” consists of two prongs. Each prong—the “conclusion of direct review” and the “expiration of the time for seeking such review”—relates to a distinct category of petitioners. For petitioners who pursue direct review all the way to this Court, the judgment becomes final at the “conclusion of direct review”—when this Court affirms a conviction on the merits or denies a petition for certiorari. For all other petitioners, the judgment becomes final at the “expiration of the time for seeking such review”—when the time for pursuing direct review in this Court, or in state court, expires. We thus agree with the Court of Appeals that because Gonzalez did not appeal to the State’s highest court, his judgment became final when his time for seeking review with the State’s highest court expired. B Gonzalez offers an alternative reading of §2244(d)(1)(A): Courts should determine both the “conclusion of direct review” and the “expiration of the time for seeking such review” for every petitioner who does not seek certiorari, then start the 1-year clock from the “latest of” the two dates. Gonzalez rejects our uniform definition of the “conclusion of direct review” as the date on which this Court affirms a conviction on the merits or denies a petition for certiorari. In his view, whenever a petitioner does not seek certiorari, the “conclusion of direct review” is the date on which state law marks finality—in Texas, the date on which the mandate issues. Ex parte Johnson, 12 S.W.3d 472, 473 (Crim. App. 2000) (per curiam). Applying this approach, Gonzalez contends that his habeas petition was timely because his direct review “concluded” when the mandate issued (on September 26, 2006), later than the date on which his time for seeking Texas CCA review “expired” (August 11, 2006). We find his construction of the statute unpersuasive. First, Gonzalez lacks a textual anchor for his later- in-time approach. The words “latest of” do not appear anywhere in §2244(d)(1)(A). Rather, they appear in §2244(d)(1) and refer to the “latest of” the dates in subparagraphs (A), (B), (C), and (D)—the latter three of which are inapplicable here. Nothing in §2244(d)(1)(A) contemplates any conflict between the “conclusion of direct review” and the “expiration of the time for seeking such review,” much less instructs that the later of the two shall prevail. Nor is Gonzalez’s later-in-time reading necessary to give both prongs of §2244(d)(1)(A) full effect. Our reading does so by applying one “or” the other, depending on whether the direct review process concludes or expires. Treating the judgment as final on one date “or” the other is consistent with the disjunctive language of the provision. Second, Gonzalez misreads our precedents. Gonzalez asserts that in Jimenez, we made a later-in-time choice between the two prongs. That is mistaken. Rather, we chose between two “expiration” dates corresponding to different appeals: Jimenez initially failed to appeal to the Texas Court of Appeals and that appeal became final when his “time for seeking discretionary review . . . expired.” 555 U. S., at 117, 119. When Jimenez was later allowed to file an out-of-time appeal, he pursued appeals with both the Texas Court of Appeals and Texas CCA; the out-of-time appeal thus became final when his “[t]ime for seeking certiorari review . . . with this Court expired.” Id., at 116, 120. We adopted the out-of-time appeal’s date of finality over the initial appeal’s date of finality. Id., at 119–121. Critically, by deeming the initial appeal final at the expiration of time for seeking review in state court, and the out-of-time appeal final at the expiration of time for seeking certiorari in this Court, we reinforced Clay’s suggestion that the “expiration” prong governs all petitioners who do not pursue direct review all the way to this Court. [ 10 ] Third, Gonzalez argues that AEDPA’s federalism concerns and respect for state-law procedures mean that we should not read §2244(d)(1)(A) to disregard state law. We agree. That is why a state court’s reopening of direct review will reset the limitations period. 555 U. S., at 121. That is also why, just as we determine the “expiration of the time for seeking [direct] review” from this Court’s filing deadlines when petitioners forgo certiorari, we look to state-court filing deadlines when petitioners forgo state-court appeals. Referring to state-law procedures in that context makes sense because such deadlines are inherently court specific. There is no risk of relying on “state-law rules that may differ from the general federal rule.” Clay, 537 U. S., at 531. By contrast, Gonzalez urges us to scour each State’s laws and cases to determine how it defines finality for every petitioner who forgoes a state-court appeal. That ap- proach would usher in state-by-state definitions of the con- clusion of direct review. It would be at odds with the uniform definition we adopted in Clay and accepted in the §2244(d)(1)(A) context in Jimenez. And it would pose serious administrability concerns. Even if roughly “half of the States define the conclusion of direct review as the issuance of the mandate or similar process,” Brief for Petitioner 40, that still leaves half with either different rules or no settled rules at all. [ 11 ] Fourth, Gonzalez speculates that our reading will rob some habeas petitioners of the full 1-year limitations pe- riod. Gonzalez asserts that our reading starts the clock running from the date that his time for seeking Texas CCA review expired, even though, under Texas law, he could not file for state habeas relief until six weeks later, on the date the Texas Court of Appeals issued its mandate. Tex. Code Crim. Proc. Ann., Art. 11.07, §3(a) (Vernon Supp. 2011). His inability to initiate state habeas proceedings during those six weeks, he argues, reduced his 1-year federal habeas filing period by six weeks. We expect, however, that it will be a rare situation where a petitioner confronting similar state laws faces a delay in the mandate’s issuance so excessive that it prevents him or her from filing a federal habeas petition within a year. [ 12 ] A petitioner who has exhausted his or her claims in state court need not await state habeas proceedings to seek federal habeas relief on those claims. To the extent a petitioner has had his or her federal filing period severely truncated by a delay in the mandate’s issuance and has unexhausted claims that must be raised on state habeas review, such a petitioner could file a request for a stay and abeyance from the federal district court. See Rhines v. Weber, 544 U.S. 269, 277 (2005) . Finally, Gonzalez argues, as an alternative to his later-in-time construction, that his petition should be considered timely because it was filed within a year of when his time for seeking this Court’s review—as opposed to the Texas CCA’s review—expired. We can review, however, only judgments of a “state court of last resort” or of a lower state court if the “state court of last resort” has denied discretionary review. This Court’s Rule 13.1; see also 28 U. S. C. §1257(a) (2006 ed.). Because Gonzalez did not appeal to the Texas CCA, this Court would have lacked jurisdiction over a petition for certiorari from the Texas Court of Appeals’ decision affirming Gonzalez’s conviction. We therefore decline to incorporate the 90-day period for seeking certiorari in determining when Gonzalez’s judgment became final. * * * In sum, we hold that §2253(c)(3) is a mandatory but nonjurisdictional rule. Here, the COA’s failure to “indicate” a constitutional issue did not deprive the Court of Appeals of jurisdiction to adjudicate Gonzalez’s appeal. We further hold that, with respect to a state prisoner who does not seek review in a State’s highest court, the judgment becomes “final” under §2244(d)(1)(A) when the time for seeking such review expires—here, August 11, 2006. We thus agree with the Court of Appeals that Gonzalez’s federal habeas petition was time barred. For the reasons stated, the judgment of the Court of Appeals for the Fifth Circuit is Affirmed. Notes 1 The Circuits have divided over whether a defect in a COA is a jurisdictional bar. Compare, e.g., Phelps v. Alameda, 366 F.3d 722, 726 (CA9 2004) (no); Porterfield v. Bell, 258 F.3d 484, 485 (CA6 2001) (no); Young v. United States, 124 F.3d 794, 798–799 (CA7 1997) (no), with United States v. Cepero, 224 F.3d 256, 259–262 (CA3 2000) (en banc) (yes). 2 The Circuits have divided over when a judgment becomes final if a petitioner forgoes review in a State’s highest court. Compare, e.g., 623 F.3d 222, 226 (CA5 2010) (case below) (date when time for seeking such review expires); Hemmerle v. Schriro, 495 F.3d 1069, 1073–1074 (CA9 2007) (same), with Riddle v. Kemna, 523 F.3d 850, 855–856 (CA8 2008) (en banc) (date when state court issues its mandate). 3 We have also held that “context, including this Court’s interpretation of similar provisions in many years past, is relevant to whether a statute ranks a requirement as jurisdictional.” Reed Elsevier, Inc. v. Muchnick, 559 U. S. ___, ___ (2010) (slip op., at 13). Here, however, even though the requirement of a COA (or its predecessor, the certificate of probable cause (CPC)) dates back to 1908, Congress did not enact the indication requirement until 1996. There is thus no “long line of this Court’s decisions left undisturbed by Congress” on which to rely. Union Pacific R. Co. v. Locomotive Engineers and Trainmen Gen. Comm. of Adjustment, Central Region, 558 U. S. ___, ___ (2009) (slip op., at 13). The issuance of a CPC, like the issuance of a COA, was jurisdictional. Contrary to the dissent’s assertions, post, at 8–10 (opinion of Scalia, J.), that fact does not suggest that the indication requirement is jurisdictional as well. If anything, the inference runs the other way. For nearly a century, a judge’s granting or withholding of a CPC, absent any indication of issues, was the fully effective “expression of opinion,” post, at 8, required for an appeal to proceed. AEDPA’s new requirement that judges indicate the specific issues to be raised on appeal has no predecessor provision—indeed, it is the primary difference between a CPC and COA. 4 The United States as amicus curiae contends that §2253(c)(2) is jurisdictional, but the State concedes that it is not. Tr. of Oral Arg. 31. 5 The courts of appeals uniformly interpret “circuit justice or judge” to encompass district judges. See United States v. Mitchell, 216 F.3d 1126, 1129 (CADC 2000) (collecting cases); Fed. Rule App. Proc. 22(b). Habeas Corpus Rule 11(a) requires district judges to decide whether to grant or deny a COA in the first instance. 6 That fact also distinguishes the indication requirement from every “ ‘similar provisio[n]’ ” that the dissent claims we have deemed jurisdictional. Post, at 5–6. None of our cases addressing those provisions, moreover, recognized or relied on the sweeping “rule” that the dissent now invokes, whereby this Court should enforce as jurisdictional all “procedural conditions for appealing a case from one Article III court to another.” Ibid.; but see, e.g., post, at 6–7, n. 2 (conceding that the “rule” does not apply to criminal appeals); Becker v. Montgomery, 532 U.S. 757, 763 (2001) (failure to sign notice of appeal is a nonjurisdictional omission). All the cases, meanwhile, involved time limits (save one involving Federal Rule of Appellate Procedure 3(c)(1), which we address infra). In Bowles v. Russell, 551 U.S. 205 (2007) , we emphasized our “century’s worth of precedent” for treating statutory time limits on appeals as jurisdictional, id., at 209, n. 2, but even “Bowles did not hold . . . that all statutory conditions imposing a time limit should be con-sidered jurisdictional,” Reed Elsevier, 559 U. S., at ___ (slip op.,at 12). This case, in any event, involves a different type of procedural condition. 7 The dissent’s insistence that there is “no practical, real-world effect” to treating this rule as mandatory, post, at 4, ignores the real world. Courts of appeals regularly amend COAs or remand for specification of issues, notwithstanding the supposed potential to “embarras[s] a colleague.” Post, at 5; see, e.g., Saunders v. Senkowski, 587 F.3d 543, 545 (CA2 2009) (per curiam) (amending COA to add issue); United States v. Weaver, 195 F.3d 52, 53 (CADC 1999) (remanding for specification of issues). The government frequently alleges COA defects as grounds for dismissal (as the State did here, at this late stage), apparently not sharing the dissent’s concern that such efforts “yield nothing but additional litigation expenses.” Post, at 5; see, e.g., Porterfield, 258 F. 3d, at 485; Cepero, 224 F. 3d, at 257. Habeas petitioners, too, have every incentive to request that defects be resolved, not only to defuse potential problems later in the litigation, but also to ensure that the issue on which they sought appeal is certified and will receive full briefing and consideration. 8 The dissent claims that we fail to give stare decisis effect to Torres. Post, at 10. Setting aside the fact that Torres involved an unrelated Federal Rule featuring a different textual, contextual, and historical backdrop, the dissent notably fails to grapple with—indeed, its opinion is bereft of quotation to—any supporting reasoning in that opinion. That reasoning is simply not applicable here. 9 Title 28 U. S. C. §2244(d)(1) provides: “A 1-year period of limitation shall apply to an application for a writ of habeas corpus by a person in custody pursuant to the judgment of a State court. The limitation period shall run from the latest of— “(A) the date on which the judgment became final by the conclusion of direct review or the expiration of the time for seeking such review; “(B) the date on which the impediment to filing an application created by State action in violation of the Constitution or laws of the United States is removed, if the applicant was prevented from filing by such State action; “(C) the date on which the constitutional right asserted was initially recognized by the Supreme Court, if the right has been newly recognized by the Supreme Court and made retroactively applicable to cases on collateral review; or “(D) the date on which the factual predicate of the claim or claims presented could have been discovered through the exercise of due diligence.” 10 Gonzalez also argues that Lawrence v. Florida, 549 U.S. 327 (2007) , supports his focus on the state court’s issuance of the mandate because it referred to a mandate in determining when state postconviction proceedings were no longer pending. Lawrence, however, is inapposite. The case involved a different provision, 28 U. S. C. §2244(d)(2), which by its terms refers to “State” procedures. 11 Compare, e.g., PSL Realty Co. v. Granite Inv. Co., 86 Ill. 2d 291, 304, 427 N.E.2d 563, 569 (1981) (judgment is final “when entered”); Gillis v. F & A Enterprises, 934 P.2d 1253, 1256 (Wyo. 1997) (judgment is final when “opinion is filed with the clerk”), with Ex parte Johnson, 12 S.W.3d 472, 473 (Texas CCA 2000) (per curiam) (judgment is final at “issuance of the mandate”). 12 We note that Gonzalez waited four months from the date of the mandate’s issuance before filing a state habeas petition. See 623 F. 3d, at 223. When that petition was dismissed as improperly filed, Gonzalez waited another three months before refiling. Ibid. Even then, his state habeas proceedings concluded several weeks before his 1-year federal deadline elapsed. Id., at 225.

565.US.2011_10-637

During petitioner Greene’s trial for murder, robbery, and conspiracy, the prosecution introduced the redacted confessions of two of Greene’s nontestifying codefendants. A jury convicted Greene. The Pennsylvania Superior Court upheld the conviction, reasoning that the rule announced in Bruton v. United States, 391 U.S. 123, did not apply because the confessions were redacted to remove any specific reference to Greene. While Greene’s petition to the Pennsylvania Supreme Court was pending, this Court announced in Gray v. Maryland, 523 U.S. 185, that Bruton does apply to some redacted confessions. The Pennsylvania Supreme Court declined to hear Greene’s appeal, and he then sought federal habeas relief. Under the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), a federal court may not grant such relief to a state prisoner on any claim that has been “adjudicated on the merits in State court proceedings” unless that adjudication “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.” 28 U. S. C. §2254(d)(1). Here, the District Court concluded that, because the United States Supreme Court’s opinion in Gray had not yet been issued when the Pennsylvania Superior Court adjudicated Greene’s claim, the condition for granting habeas relief had not been met. The Third Circuit affirmed. Held: 1. Under §2254(d)(1), “clearly established Federal law, as determined by the Supreme Court of the United States” includes only this Court’s decisions as of the time of the relevant state-court adjudication on the merits. The Court’s decision last Term in Cullen v. Pinholster, 563 U. S. ___, established that §2254(d)(1)’s “backward-looking language requires an examination of the state-court decision at the time it was made.” Id., at ___. As the Court explained in Cullen, §2254(d)(1) requires federal courts to measure state-court decisions “against this Court’s precedents as of ‘the time the state court renders its decision.’ ” Id., at ___. That reasoning determines the result here. Pp. 3–6. 2. Because the Pennsylvania Superior Court’s decision—the last state-court adjudication on the merits of Greene’s claim—predated Gray by nearly three months, the Third Circuit correctly held that Gray was not “clearly established Federal law” against which it could measure the state-court decision. It therefore correctly concluded that the state court’s decision neither was “contrary to,” nor “involved an unreasonable application of,” any “clearly established Federal law.” Pp. 6–7. 606 F.3d 85, affirmed. Scalia, J., delivered the opinion for a unanimous Court.

Under the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), a federal court may not grant habeas relief to a state prisoner with respect to any claim that has been “adjudicated on the merits in State court proceedings” unless the state-court adjudication “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.” 28 U. S. C. §2254(d)(1). We consider whether “clearly established Federal law” includes decisions of this Court that are announced after the last adjudication of the merits in state court but before the defendant’s conviction becomes final. I In December 1993, petitioner Eric Greene and four co-conspirators robbed a grocery store in North Philadelphia, Pennsylvania. During the robbery, one of the men shot and killed the store’s owner. The five were apprehended, and two of them confessed to taking part in the robbery. Greene did not confess, but he was implicated by the others’ statements. When the Commonwealth sought to try all of the co-conspirators jointly, Greene sought severance, arguing, inter alia, that the confessions of his nontestifying codefendants should not be introduced at his trial. The trial court denied the motion to sever, but agreed to require redaction of the confessions to eliminate proper names. As redacted, the confessions replaced names with words like “this guy,” “someone,” and “other guys,” or with the word “blank,” or simply omitted the names without substitution. A jury convicted Greene of second-degree murder, robbery, and conspiracy. He appealed to the Pennsylvania Superior Court, arguing that severance of his trial was demanded by the rule announced in Bruton v. United States, 391 U.S. 123 (1968), that the Confrontation Clause forbids the prosecution to introduce a nontestifying co- defendant’s confession implicating the defendant in the crime. The Pennsylvania Superior Court affirmed the conviction, holding that the redaction had cured any problem under Bruton. Greene filed a petition for allowance of appeal to the Pennsylvania Supreme Court, raising the same Bruton claim. While that petition was pending, we held in Gray v. Maryland, 523 U.S. 185, 195 (1998), that “considered as a class, redactions that replace a proper name with an obvious blank, the word ‘delete,’ a symbol, or similarly notify the jury that a name has been deleted are similar enough to Bruton’s unredacted confessions as to warrant the same legal results.” The Pennsylvania Supreme Court granted the petition for allowance of appeal, limited to the question whether admission of the redacted confessions violated Greene’s Sixth Amendment rights. After the parties filed merits briefs, however, the Pennsylvania Supreme Court dismissed the appeal as improvidently granted. Greene then filed a federal habeas corpus petition in the United States District Court for the Eastern District of Pennsylvania, alleging, inter alia, that the introduction of his nontestifying codefendants’ statements violated the Confrontation Clause. Adopting the report and recommendation of a Magistrate Judge, the District Court denied the petition. It concluded that since our decision in Gray was not “clearly established Federal law” when the Pennsylvania Superior Court adjudicated Greene’s Confrontation Clause claim, that court’s decision was not “contrary to,” or “an unreasonable application of, clearly established Federal law.” 28 U. S. C. §2254(d)(1). A divided panel of the United States Court of Appeals for the Third Circuit affirmed. Greene v. Palakovich, 606 F.3d 85 (2010). The majority held that the “clearly established Federal law” referred to in §2254(d)(1) is the law at the time of the state-court adjudication on the merits. Id., at 99. The dissenting judge contended that it is the law at the time the conviction becomes final. Id., at 108. We granted certiorari. 563 U. S. ___ (2011). II Section 2254(d) of Title 28, U. S. C., as amended by AEDPA, provides: “An application for a writ of habeas corpus on behalf of a person in custody pursuant to the judgment of a State court shall not be granted with respect to any claim that was adjudicated on the merits in State court proceedings unless the adjudication of the claim— “(1) resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States; or “(2) resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” The issue here pertains to the first exception. We have said that its standard of “contrary to, or involv[ing] an unreasonable application of, clearly established Federal law” is “difficult to meet,” because the purpose of AEDPA is to ensure that federal habeas relief functions as a “ ‘guard against extreme malfunctions in the state criminal justice systems,’ ” and not as a means of error correction. Harrington v. Richter, 562 U. S. ___, ___ (2011) (slip op., at 12–13) (quoting Jackson v. Virginia, 443 U.S. 307, 332, n. 5 (1979) (Stevens, J., concurring in judgment)). In light of that objective, and relying upon the text of the provision, we held last Term, in Cullen v. Pinholster, 563 U. S. ___ (2011), that review under §2254(d)(1) is limited to the record that was before the state court that adjudicated the prisoner’s claim on the merits. We said that the provision’s “backward-looking language requires an examination of the state-court decision at the time it was made.” Id., at ___ (slip op., at 9). The reasoning of Cullen determines the result here. As we explained, §2254(d)(1) requires federal courts to “focu[s] on what a state court knew and did,” and to measure state-court decisions “against this Court’s precedents as of ‘the time the state court renders its decision.’ ” Id., at __ (slip op., at 10) (quoting Lockyer v. Andrade, 538 U.S. 63, 71–72 (2003); emphasis added). Greene resists that conclusion by appealing to our decision in Teague v. Lane, 489 U.S. 288 (1989). Teague held that, with two exceptions not pertinent here, a prisoner seeking federal habeas relief may rely on new constitutional rules of criminal procedure announced before the prisoner’s conviction became final. Id., at 310 (plurality opinion); see also Penry v. Lynaugh, 492 U.S. 302, 313 (1989) (affirming and applying Teague rule). Finality occurs when direct state appeals have been exhausted and a petition for writ of certiorari from this Court has become time barred or has been disposed of. Griffith v. Kentucky, 479 U.S. 314, 321, n. 6 (1987). Greene contends that, because finality marks the temporal cutoff for Teague purposes, it must mark the temporal cutoff for “clearly established Federal law” under AEDPA. The analogy has been rejected by our cases. We have explained that AEDPA did not codify Teague, and that “the AEDPA and Teague inquiries are distinct.” Horn v. Banks, 536 U.S. 266, 272 (2002) (per curiam). The retroactivity rules that govern federal habeas review on the merits—which include Teague—are quite separate from the relitigation bar imposed by AEDPA; neither abrogates or qualifies the other. If §2254(d)(1) was, indeed, pegged to Teague, it would authorize relief when a state-court merits adjudication “resulted in a decision that became contrary to, or an unreasonable application of, clearly established Federal law, before the conviction became final.” The statute says no such thing, and we see no reason why Teague should alter AEDPA’s plain meaning.[1] Greene alternatively contends that the relevant “decision” to which the “clearly established Federal law” criterion must be applied is the decision of the state supreme court that disposes of a direct appeal from a defendant’s conviction or sentence, even when (as here) that decision does not adjudicate the relevant claim on the merits. This is an implausible reading of §2254(d)(1). The text, we repeat, provides that habeas relief “shall not be granted with respect to any claim that was adjudicated on the merits in State court proceedings unless the adjudication of the claim . . . resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law . . . .” (Emphasis added.) The words “the adjudication” in the “unless” clause obviously refer back to the “adjudicat[ion] on the merits,” and the phrase “resulted in a decision” in the “unless” clause obviously refers to the decision produced by that same adjudication on the merits. A later affirmance of that decision on alternative procedural grounds, for example, would not be a decision resulting from the merits adjudication. And much less would be (what is at issue here) a decision by the state supreme court not to hear the appeal—that is, not to decide at all. III The Third Circuit held, and the parties do not dispute, that the last state-court adjudication on the merits of Greene’s Confrontation Clause claim occurred on direct appeal to the Pennsylvania Superior Court. 606 F. 3d, at 92, and n. 1. The Pennsylvania Superior Court’s decision predated our decision in Gray by nearly three months. The Third Circuit thus correctly held that Gray was not “clearly established Federal law” against which it could measure the Pennsylvania Superior Court’s decision. 606 F. 3d, at 99. The panel then concluded (and the parties do not dispute) that the Pennsylvania Superior Court’s decision neither was “contrary to,” nor “involved an unreasonable application of,” any “clearly established Federal law” that existed at the time. Id., at 106. Consequently, §2254(d)(1) bars the federal courts from granting Greene’s application for a writ of habeas corpus. We must observe that Greene’s predicament is an unusual one of his own creation. Before applying for federal habeas, he missed two opportunities to obtain relief under Gray: After the Pennsylvania Supreme Court dismissed his appeal, he did not file a petition for writ of certiorari from this Court, which would almost certainly have produced a remand in light of the intervening Gray decision. “Where intervening developments . . . reveal a reasonable probability that the decision below rests upon a premise that the lower court would reject if given the opportunity for further consideration, and where it appears that such a redetermination may determine the ultimate outcome of the litigation, [an order granting the petition, vacating the judgment below, and remanding the case (GVR)] is, we believe, potentially appropriate.” Lawrence v. Chater, 516 U.S. 163, 167 (1996) (per curiam). See, e.g., Stanbridge v. New York, 395 U.S. 709 (1969) (per curiam) (GVR in light of Bruton). Nor did Greene assert his Gray claim in a petition for state postconviction relief. Having forgone two obvious means of asserting his claim, Greene asks us to provide him relief by interpreting AEDPA in a manner contrary to both its text and our precedents. We decline to do so, and affirm the judgment of the Court of Appeals. It is so ordered. Notes 1 Whether §2254(d)(1) would bar a federal habeas petitioner from relying on a decision that came after the last state-court adjudication on the merits, but fell within one of the exceptions recognized in Teague, 489 U. S., at 311, is a question we need not address to resolve this case.

566.US.2011_10-875

Chapter 12 of the Bankruptcy Code allows farmer debtors with regular annual income to adjust their debts subject to a reorganization plan. The plan must provide for full payment of priority claims. 11 U. S. C. §1222(a)(2). Under §1222(a)(2)(A), however, certain governmental claims arising from the disposition of farm assets are stripped of priority status and downgraded to general, unsecured claims that are dischargeable after less than full payment. That exception applies only to claims “entitled to priority under [ 11 U. S. C. §507]” in the first place. As relevant here, §507(a)(2) covers “administrative expenses allowed under §503(b),” which includes “any tax . . . incurred by the estate.” §503(b)(B)(i). Petitioners filed for Chapter 12 bankruptcy and then sold their farm. They proposed a plan under which they would pay off outstanding liabilities with proceeds from the sale. The Internal Revenue Service (IRS) objected, asserting a tax on the capital gains from the sale. Petitioners then proposed treating the tax as an unsecured claim to be paid to the extent funds were available, with the unpaid balance being discharged. The Bankruptcy Court sustained an IRS objection, the District Court reversed, and the Ninth Circuit reversed the District Court. The Ninth Circuit held that because a Chapter 12 estate is not a separate taxable entity under the Internal Revenue Code (IRC), 26 U. S. C. §§1398, 1399, it does not “incur” postpetition federal income taxes. The Ninth Circuit concluded that because the tax was not “incurred by the estate” under §503(b), it was not a priority claim eligible for the §1222(a)(2)(A) exception. Held: The federal income tax liability resulting from petitioners’ postpetition farm sale is not “incurred by the estate” under §503(b) of the Bankruptcy Code and thus is neither collectible nor dischargeable in the Chapter 12 plan. Pp. 4−17. (a) The phrase “incurred by the estate” bears a plain and natural reading. A tax “incurred by the estate” is a tax for which the estate itself is liable. Only certain estates are liable for federal income taxes. IRC §§1398 and 1399 define the division of responsibilities for the payment of taxes between the estate and the debtor on a chapter-by-chapter basis. Under those provisions, a Chapter 12 estate is not a separately taxable entity. The debtor—not the trustee—is generally liable for taxes and files the only tax return. The postpetition income taxes are thus not “incurred by the estate.” Pp. 4−5. (b) Section 346 of the Bankruptcy Code and its longstanding interplay with IRC §§1398 and 1399 reinforce that whether an estate “incurs” taxes turns on Congress’ chapter-specific guidance on which estates are separately taxable. The original §346 established that state or local income taxes could be imposed only on the estate in an individual-debtor Chapter 7 or 11 bankruptcy, and only on the debtor in a Chapter 13 bankruptcy. Congress applied the framework of §346 to federal taxes two years later: IRC §1398 and 1399 established that the estate is separately taxable in individual-debtor Chapter 7 or 11 cases, and not separately taxable in Chapter 13 (and now Chapter 12) cases. The Bankruptcy Abuse Prevention and Consumer Protection Act of 2005 subsequently amended §346, expressly aligning its assignment of state or local taxes with the IRC separate taxable entity rules for federal taxes. This Court assumes that Congress is aware of existing law when it passes legislation, and the existing law at the enactment of §1222(a)(2)(A) indicated that an estate’s liability for taxes turned on separate taxable entity rules. Pp. 6−9. (c) Chapter 13, on which Chapter 12 was modeled, further bolsters this Court’s holding. Established understandings hold that postpetition income taxes are not “incurred by the [Chapter 13] estate” under §503(b) because they are the liability of the Chapter 13 debtor alone. The Government has also long hewed to this position. Section 1305(a)(1), which gives holders of postpetition claims the option of collecting postpetition taxes within the bankruptcy case, would be superfluous if postpetition tax liabilities were automatically collectible inside the bankruptcy. It is thus clear that postpetition income taxes are not automatically collectible in a Chapter 13 plan and are not administrative expenses under §503(b). To hold otherwise in Chapter 12 would disrupt settled practices in Chapter 13 cases. Pp. 9−12. (d) None of the contrary arguments by petitioners and the dissent overcomes the statute’s plain language, context, and structure. There is no textual basis for giving “incurred by the estate” a temporal meaning, such that it refers to all taxes “incurred postpetition.” Nor does the text support deeming a tax “incurred by the estate” whenever it is paid by the debtor out of property of the estate. Section 503’s legislative history is not inconsistent with this Court’s holding, and the Court has cautioned against allowing ambiguous legislative history to muddy clear statutory language. See Milner v. Department of Navy, 562 U. S. ___, ___. Meanwhile, any cases suggesting that postpetition taxes were treated as administrative expenses are inapposite because they involve corporate debtors, which Congress has singled out for responsibilities paralleling those borne by a separate taxable entity’s trustee. Finally, petitioners contend that the purpose of §1222(a)(2)(A) was to provide debtors with robust relief from tax debts. There may be compelling policy reasons for treating postpetition income tax liabilities as dischargeable. But if Congress intended petitioners’ result, it did not so provide in the statute. Pp. 12−17. 617 F.3d 1161, affirmed. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Thomas, and Alito, JJ., joined. Breyer, J., filed a dissenting opinion, in which Kennedy, Ginsburg, and Kagan, JJ., joined.

Under Chapter 12 of the Bankruptcy Code, farmer debtors may treat certain claims owed to a governmental unit resulting from the disposition of farm assets as dischargeable, unsecured liabilities. 11 U. S. C. §1222(a) (2)(A). One such claim is for “any tax . . . incurred by the estate.” §503(b)(B)(i). The question presented is whether a federal income tax liability resulting from individual debtors’ sale of a farm during the pendency of a Chapter 12 bankruptcy is “incurred by the estate” and thus dischargeable. We hold that it is not. I A In 1986, Congress enacted Chapter 12 of the Bankruptcy Code, §1201 et seq., to allow farmer debtors with regu- lar annual income to adjust their debts. Chapter 12 was modeled on Chapter 13, §1301 et seq., which permits individual debtors with regular annual income to preserve existing assets subject to a “court-approved plan under which they pay creditors out of their future income.” Hamilton v. Lanning, 560 U. S. ___, ___ (2010) (slip op., at 1). Chapter 12 debtors similarly file a plan of reorganization. §1221. To be confirmed, the plan must provide for the full payment of priority claims. §1222(a)(2). In the Bankruptcy Abuse Prevention and Consumer Protection Act of 2005 (BAPCPA), §1003, 119Stat. 186, Congress created an exception to that requirement: “Contents of plan “(a) The plan shall— . . . . . “(2) provide for the full payment, in deferred cash payments, of all claims entitled to priority under section 507, unless— “(A) the claim is a claim owed to a governmental unit that arises as a result of the sale, transfer, exchange, or other disposition of any farm asset used in the debtor’s farming operation, in which case the claim shall be treated as an unsecured claim that is not entitled to priority under section 507, but the debt shall be treated in such manner only if the debtor receives a discharge.” 11 U. S. C. §1222. Under §1222(a)(2)(A), certain governmental claims resulting from the disposition of farm assets are downgraded to general, unsecured claims that are dischargeable after less than full payment. See §1228(a). The claims are stripped of their priority status. That exception, however, applies only to claims in the plan that are “entitled to priority under section 507” in the first place. Section 507 lists 10 categories of such claims. Two pertain to taxes: One category, §507(a)(8), covers prepetition taxes, and is inapplicable in this case. The other, §507(a)(2), covers “administrative expenses allowed under section 503(b),” which in turn includes “any tax . . . incurred by the estate.” §503(b)(B)(i). Thus, for postpetition taxes to be entitled to priority under §507 and eligible for the §1222(a)(2)(A) exception, the taxes must be “incurred by the estate.” B Petitioners Lynwood and Brenda Hall petitioned for bankruptcy under Chapter 12 and sold their farm shortly thereafter. Petitioners initially proposed a plan of reorganization under which they would pay off outstanding liabilities with proceeds from the sale. The Internal Revenue Service (IRS) objected, asserting a federal income tax of $29,000 on the capital gains from the farm sale. Petitioners amended their proposal to treat the income tax as a general, unsecured claim to be paid to the extent funds were available, with the unpaid balance discharged. Again the IRS objected. Taxes on income from a postpetition farm sale, the IRS argued, remain the debtors’ independent responsibility because they are neither collectible nor dischargeable in bankruptcy. The Bankruptcy Court sustained the objection. The court reasoned that because a Chapter 12 estate is not a separate taxable entity under the Internal Revenue Code (IRC), see 26 U. S. C. §§1398, 1399, it cannot “incur” taxes for purposes of 11 U. S. C. §503(b). The District Court reversed, expressing doubt that IRC provisions are relevant to interpreting §503(b). Based on its reading of legislative history, the District Court determined that Congress intended §1222(a)(2)(A) to extend to petitioners’ postpetition taxes. The Court of Appeals for the Ninth Circuit reversed. 617 F.3d 1161 (2010). The Court of Appeals held that the Chapter 12 estate does not “incur” the postpetition federal income taxes for purposes of §503(b) because it is not a separate taxable entity under the IRC, and noted that Congress repeatedly has indicated the relevance of the IRC’s taxable entity provisions to the Bankruptcy Code. Although “sympathetic” to the view that the postpetition tax liabilities should be dischargeable, the Court of Appeals held that “the operative language simply failed to make its way into the statute.” Id., at 1167. The Court of Appeals concluded that because the taxes do not qualify under §503(b), they are not priority claims in the plan eligible for the §1222(a)(2)(A) exception. Judge Paez dissented, siding with a sister Circuit that had concluded that Congress intended §1222(a)(2)(A) to extend to such postpetition federal income taxes. We granted certiorari to resolve the split of authority.[1] 564 U. S. ___ (2011). II A Our resolution of this case turns on the meaning of a phrase in §503(b) of the Bankruptcy Code: “incurred by the estate.” The parties agree that §1222(a)(2)(A) applies only to priority claims collectible in the bankruptcy plan and that postpetition federal income taxes so qualify only if they constitute a “tax . . . incurred by the estate.” §503(b)(B)(i). The phrase “incurred by the estate” bears a plain and natural reading. See FCC v. AT&T Inc., 562 U. S. ___, ___ (2011) (slip op., at 5) (“When a statute does not define a term, we typically ‘give the phrase its ordinary meaning’ ”). To “incur,” one must “suffer or bring on oneself (a liability or expense).” Black’s Law Dictionary 836 (9th ed. 2009); see also Webster’s Third New International Dictionary 1146 (1976) (“to . . . become liable or subject to: bring down upon oneself”); Random House Dictionary 722 (1966) (“to become liable or subject to through one’s own action; bring upon oneself”). A tax “incurred by the estate” is a tax for which the estate itself is liable. As the IRC makes clear, only certain estates are liable for federal income taxes. Title 26 U. S. C. §§1398 and 1399 address taxation in bankruptcy and define the division of responsibilities for the payment of taxes between the estate and the debtor on a chapter-by-chapter basis. Section 1398 provides that when an individual debtor files for Chapter 7 or 11 bankruptcy, the estate shall be liable for taxes. In such cases, the trustee files a separate re- turn on the estate’s behalf and “[t]he tax” on “the taxable income of the estate . . . shall be paid by the trustee.” §1398(c)(1); see also §6012(b)(4) (“Returns of . . . an estate of an individual under chapter 7 or 11 . . . shall be made by the fiduciary thereof”). Section 1399 provides that “[e]xcept in any case to which section 1398 applies, no separate taxable entity shall result from the commencement of a [bankruptcy] case.” In Chapter 12 and 13 cases, then, there is no separately taxable estate. The debtor—not the trustee—is generally liable for taxes and files the only tax return. See In re Lindsey, 142 B.R. 447, 448 (Bkrtcy. Ct. WD Okla. 1992) (“It is clear that, pursuant to 26 U. S. C. §1398 and 1399, the standing Chapter 12 trustee neither files a return nor pays federal income tax”); cf. infra, at 15 (discussing special trustee duties in corporate-debtor cases). These provisions suffice to resolve this case: Chapter 12 estates are not taxable entities. Petitioners, not the estate itself, are required to file the tax return and are liable for the taxes resulting from their postpetition farm sale. The postpetition federal income tax liability is not “incurred by the estate” and thus is neither collectible nor discharge- able in the Chapter 12 plan.[2] B Our reading of “incurred by the estate” as informed by the IRC’s separate taxable entity rules draws support from a related provision of the Bankruptcy Code, 11 U. S. C. §346, and its longstanding interplay with 26 U. S. C. §§1398 and 1399. That relationship illustrates that from the inception of the current Bankruptcy Code, Congress has specified on a chapter-by-chapter basis which estates are separately taxable and therefore liable for taxes. That relationship also refutes the dissent’s suggestion that applying such rules is an incongruous importation of “tax law” unconnected to “bankruptcy principles (as Congress understood them).” Post, at 8–9 (opinion of Breyer, J.). And it reinforces the reason- ableness of our view that whether an estate “incurs” taxes under §503(b) turns on such chapter-by-chapter distinctions. In the original Bankruptcy Code, Congress included a provision, §346, that set out a chapter-specific division of tax liabilities between the estate and the debtor. Bankruptcy Reform Act of 1978, 92Stat. 2565. Section 346(b)(1) provided that in an individual-debtor Chapter 7 or 11 bankruptcy, “any income of the estate may be taxed under a State or local law imposing a tax . . . only to the estate, and may not be taxed to such individual.” 92Stat. 2565 (emphasis added); see also 11 Collier on Bankruptcy ¶TX12.03[5][b][i], p. TX12–21 (16th ed. 2011) (hereinafter Collier) (§346(b) “provided that in a case under chapter 7 [or] 11 . . . the estate of an individual is a taxable entity”). Section 346(d) provided, meanwhile, that in a Chapter 13 bankruptcy, “any income of the estate or the debtor may be taxed under a State or local law imposing a tax . . . only to the debtor, and may not be taxed to the estate.” 92Stat. 2566 (emphasis added). Congress thus established that the estate in an individual-debtor Chapter 7 or 11 bankruptcy is a separate taxable entity; the estate in a Chapter 13 bankruptcy is not.[3] Although §346 concerned state or local taxes,[4] Congress applied its framework to federal taxes two years later. In the Bankruptcy Tax Act of 1980, 94Stat. 3397, Congress enacted 26 U. S. C. §§1398 and 1399. Section 1398 of the IRC, much like §346(b) in the Bankruptcy Code, established that the estate is separately taxable in individual-debtor Chapter 7 or 11 cases. Section 1399 of the IRC, much like §346(d) in the Bankruptcy Code, clarified that the estate is not separately taxable in Chapter 13 (and now Chapter 12) cases. In 2005, Congress in BAPCPA amended §346 and crystallized the connection between the Bankruptcy Code and the IRC. Section 346 now expressly aligns its assignment of state or local taxes with the rules for federal taxes, providing in relevant part: “(a) Whenever the Internal Revenue Code of 1986 provides that a separate taxable estate or entity is created in a case concerning a debtor under this title, and the income . . . of such estate shall be taxed to or claimed by the estate, a separate taxable estate is also created for purposes of any State and local law imposing a tax on or measured by income and such income . . . shall be taxed to or claimed by the estate and may not be taxed to or claimed by the debtor. “(b) Whenever the Internal Revenue Code of 1986 provides that no separate taxable estate shall be created in a case concerning a debtor under this title, and the income . . . of an estate shall be taxed to or claimed by the debtor, such income . . . shall be taxed to or claimed by the debtor under a State or local law imposing a tax on or measured by income and may not be taxed to or claimed by the estate.” (Emphasis added.) Thus, whenever the estate is separately taxable under federal income tax law, that “is also” the case under state or local income tax law, §346(a), and vice versa, §346(b). And given that the Bankruptcy Code instructs that the as- signment of state or local tax liabilities shall turn on the IRC’s separate taxable entity rules, there is parity in turning to such rules in assigning federal tax liabilities. In the same Act, Congress added §1222(a)(2)(A). Section 1222(a)(2)(A) carves out an exception to the ordinary priority classification scheme. But §1222(a)(2)(A) did not purport to redefine which claims are otherwise entitled to priority, much less alter the underlying division of tax liability between the estate and the debtor in Chapter 12 cases. “We assume that Congress is aware of existing law when it passes legislation,” Miles v. Apex Marine Corp., 498 U.S. 19, 32 (1990), and the existing law at the enactment of §1222(a)(2)(A) indicated that an estate’s liability for taxes turned on chapter-by-chapter separate taxable entity rules. C The statutory structure further reinforces our holding that petitioners’ postpetition income taxes are not “incurred by the estate.” As a leading bankruptcy treatise and lower courts recognize, “[b]ecause chapter 12 was modeled on chapter 13, and because so many of the provisions are identical, chapter 13 cases construing provisions corresponding to chapter 12 provisions may be relied on as authority in chapter 12 cases.” 8 Collier ¶1200.01[5], at 1200–10; In re Lopez, 372 B.R. 40, 45, n. 13 (Bkrtcy. App. Panel CA9 2007); Justice v. Valley Nat. Bank, 849 F.2d 1078, 1083 (CA8 1988). We agree. Section 1322(a)(2), like §1222(a)(2), requires full payment of “all claims entitled to priority under section 507” under the plan. Both provisions cross-reference the same section of the Code, §507, and in turn, the same subsection, §503(b). Both are treated alike by IRC §§1398 and 1399. Whether postpetition taxes qualify under §503(b) in Chapter 13 thus sheds light on whether they so qualify in petitioners’ Chapter 12 case. Bankruptcy courts and commentators have reasoned that postpetition income taxes are not “incurred by the estate” under §503(b) because “a tax on postpetition income of the debtor or of the chapter 13 estate is not a liability of the chapter 13 estate; it is a liability of the debtor alone.” 8 Collier ¶1305.02[1], at 1305–5 and 1305–6.[5] For over a decade, the Government has likewise hewed to the position that “since post-petition tax liabilities are, in Chapter 13 cases, incurred by the debtor, rather than the bankruptcy estate, characterizing such liabilities as administrative expenses is inconsistent with section 503.” IRS Chief Counsel Advice No. 200113027, p. 6 (Mar. 30, 2001), 2001 WL 307746, *4; see also Internal Revenue Manual §5.9.10.9.2(3) (2006) (hereinafter IRM); IRS Litigation Guideline Memorandum GL–26, p. 9 (Dec. 16, 1996), 1996 WL 33107107, *6. We see no reason to depart from those established understandings. To “ ‘hold the Chapter 13 estate liable for [a] tax when it does not exist as a taxable entity defies common sense as well as Congress’ intent.’ ” In re Whall, 391 B.R. 1, 4 (Bkrtcy. Ct. Mass. 2008). The same holds true for a Chapter 12 estate. A provision in Chapter 13 confirms that postpetition income taxes fall outside §503(b). Section 1305(a)(1) pro- vides that “[a] proof of claim may be filed by any entity that holds a claim against the debtor . . . for taxes that become payable to a governmental unit while the case is pending.” (Emphasis added.) That provision gives holders of postpetition claims the option of collecting postpetition taxes within the bankruptcy case—an option that the Government would never need to invoke if postpetition tax liabilities were already collectible inside the bankruptcy. Accordingly, lest we render §1305 “ ‘inoperative or superfluous,’ ” Hibbs v. Winn, 542 U.S. 88, 101 (2004), it is clear that postpetition income taxes are not automatically collectible in a Chapter 13 plan and, a fortiori, are not administrative expenses under §503(b). It follows that postpetition income taxes are not automatically collectible in petitioners’ Chapter 12 plan.[6] Because both chapters cross-reference §503(b) in an identical manner, see §§1222(a)(2), 1322(a)(2), we are cognizant that any conflicting reading of §503(b) here could disrupt settled Chapter 13 practices. See Cohen v. de la Cruz, 523 U.S. 213, 221 (1998) (the Court “ ‘will not read the Bankruptcy Code to erode past bankruptcy practice absent a clear indication that Congress intended such a departure’ ”). Chapter 13 filings outnumber Chapter 12 filings six-hundred-fold. See U. S. Bankruptcy Courts—Cases Commenced During the 12-Month Period Ending September 30, 2011 (Table F–2), http://www.uscourts.gov/ Statistics/BankruptcyStatistics.aspx (estimating 676 and 417,503 annual Chapter 12 and 13 filings, respectively) (as visited May 14, 2012, and available in Clerk of Court’s case file). Yet adopting petitioners’ reading of §503(b) would mean that, in every Chapter 13 case, the Government could ignore §1305 and expect priority payment of postpetition income taxes in every plan. At bottom, “identical words and phrases within the same statute should normally be given the same meaning.” Powerex Corp. v. Reliant Energy Services, Inc., 551 U.S. 224, 232 (2007). Absent any indication that Congress intended a conflict between two closely related chapters, we decline to create one.[7] III Petitioners and the dissent advance several arguments for why the postpetition income taxes at issue should be considered “incurred by the estate,” notwithstanding the IRC’s separate taxable entity rules. But none provides sufficient reason to overcome the statute’s plain language, context, and structure. Petitioners primarily argue that “incurred by the estate” has a temporal meaning. Petitioners emphasize that the estate only comes into existence after a bankruptcy petition is filed. Thus, they reason, taxes “incurred by the estate” refers to all taxes “incurred postpetition,” regardless of whether the estate is liable for the tax and regardless of the chapter under which a case is filed. Although all taxes “incurred by the estate” are necessarily incurred postpetition, not all taxes incurred postpetition are “incurred by the estate.” That an estate cannot incur liability until it exists does not mean that every liability that arises after that point automatically becomes the estate’s liability. And there is no textual basis to focus on when the liability is incurred, as opposed to whether the liability is incurred “by the estate.” Alternately, petitioners contend that a tax should be considered “incurred by the estate” so long as it is payable out of estate assets. Income from postpetition sales of farm assets is considered property of the estate. See §1207(a). Petitioners argue that even if the debtor—and not the estate—is liable for a tax, the tax is still “incurred by the estate” because the funds the debtor uses to pay the tax are property of the estate. But that too strains the text beyond what it can bear. To concede that someone other than the estate is liable for filing the return and paying the tax, and yet maintain that the estate is the one that has “incurred” the tax, defies the ordinary meaning of “incur” as bringing a liability upon oneself. The dissent, echoing both of these points, urges that we “simply . . . consider the debtor and estate as merged.” Post, at 11. “The English language,” the dissent reasons, “permits this reading” and “do[es] not require” our reading. Post, at 8–9. But any reading of “tax . . . incurred by the estate” that is contingent on merging the debtor and estate—despite Congress’ longstanding efforts to distinguish between when tax liabilities are borne by the debtor or borne by the estate—is not a natural construction of the statute as written. Moreover, these alternative readings create a conflict between §503(b) and §346(b). Petitioners consider postpetition state or local income taxes, like federal income taxes, to be “incurred by the estate” under §503(b). See Tr. of Oral Arg. 4–5. But §346(b) requires that such taxes be borne by the Chapter 12 debtor, not the estate. It is implausible to maintain that taxes are “incurred by the estate” when §346(b) specifically prohibits such taxes from being “taxed to or claimed by the estate.” To buttress their counterintuitive readings of the text, petitioners and the dissent suggest that there is a long history of treating postpetition taxes as administrative expenses entitled to priority. Both point to two legislative Reports accompanying the 1978 enactment of §503. But neither snippet from which they quote is inconsistent with today’s holding,[8] and we have cautioned against “allowing ambiguous legislative history to muddy clear statutory language.” Milner v. Department of Navy, 562 U. S. ___, ___ (2011) (slip op., at 9). Petitioners also point to cases suggesting that postpetition taxes were treated as administrative expenses. E.g., United States v. Noland, 517 U.S. 535, 543 (1996) (cor- porate Chapter 11 debtor); Nicholas v. United States, 384 U.S. 678, 687–688 (1966) (corporate Chapter XI case under predecessor Bankruptcy Act). But those cases involve corporate debtors and are therefore inapposite. Among estates that are not separately taxable, those involving corporate debtors have long been singled out by Congress for special responsibilities.[9] See H. R. Rep., at 277 (even “[i]f the estate is not a separate taxable entity,” administrative responsibility can “var[y] according to the nature of the debtor”). Although estates of corporate debtors are not separate taxable entities under 26 U. S. C. §§1398 and 1399, the IRC requires a trustee that “has possession of or holds title to all or substantially all the property or business of a corporation” to “make the return of income for such corporation.” §6012(b)(3). In effect, Congress provided that the trustee in a corporate-debtor case may shoulder responsibility that parallels that borne by the trustee of a separate taxable entity. In any event, petitioners do not deny that neither the separate taxable entity provisions nor the special provisions for corporate debtors apply to them. Finally, petitioners and the dissent contend that the purpose of 11 U. S. C. §1222(a)(2)(A) was to provide debtors with robust relief from tax debts, relying on statements by a single Senator on unenacted bills introduced in years preceding the enactment. See Brief for Petitioners 23–36. They argue that deeming §1222(a)(2)(A) inapplicable to their postpetition income taxes would undermine that purpose and confine the exception to prepetition taxes. But we need not resolve here what other claims, if any, are covered by §1222(a)(2)(A).[10] Whatever the 2005 Congress’ intent with respect to §1222(a)(2)(A), that provision merely carved out an exception to the pre-existing priority classification scheme. The exception could only apply to claims “entitled to priority under section 507” in the first place. That pre-existing scheme was in turn premised on antecedent, decades-old understandings about the scope of §503(b) and the division of tax liabilities between estates and debtors. See Dewsnup v. Timm, 502 U.S. 410, 419 (1992) (“When Congress amends the bankruptcy laws, it does not write ‘on a clean slate’ ”). If Congress wished to alter these background norms, it needed to enact a provision to enable postpetition income taxes to be collected in the Chapter 12 plan in the first place. The dissent concludes otherwise by an inverted analysis. Rather than demonstrate that such claims were treated as §507 priority claims in the first place, the dissent begins with the single Senator’s stated purpose for the exception to that priority scheme. Post, at 7. It then reasons backwards from there, and in the process upsets background norms in both Chapters 12 and 13. Certainly, there may be compelling policy reasons for treating postpetition income tax liabilities as discharge- able. But if Congress intended that result, it did not so provide in the statute. Given the statute’s plain language, context, and structure, it is not for us to rewrite the statute, particularly in this complex terrain of interconnected provisions and exceptions enacted over nearly three decades. Petitioners’ position threatens ripple effects beyond this individual case for debtors in Chapter 13 and the broader bankruptcy scheme that we need not invite. As the Court of Appeals noted, “Congress is entirely free to change the law by amending the text.” 617 F. 3d, at 1167. * * * We hold that the federal income tax liability resulting from petitioners’ postpetition farm sale is not “incurred by the estate” under §503(b) and thus is neither collectible nor dischargeable in the Chapter 12 plan. We therefore affirm the judgment of the Court of Appeals for the Ninth Circuit. It is so ordered. Notes 1 Compare In re Dawes, 652 F.3d 1236 (CA10 2011), and 617 F.3d 1161 (CA9 2010) (case below), with Knudsen v. IRS, 581 F.3d 696 (CA8 2009) (postpetition federal taxes are eligible for the §1222(a)(2)(A) exception and thus dischargeable). 2 Because we hold that the postpetition federal income taxes at issue are not collectible in the plan because they are not “incurred by the estate,” we need not address the Government’s broader alternative argument that Chapter 12 plans are exclusively limited to prepetition claims. 3 For those of us for whom it is relevant, the legislative historyconfirms that Congress viewed §346 as defining which estates were separate taxable entities. See H. R. Rep. No. 95–595, p. 275 (1977) (here-inafter H. R. Rep.) (“A threshold issue to be considered when a debtor files a petition under title 11 is whether the estate created . . . should be treated as a separate taxable entity”); id., at 334 (“Subsection (d) indicates that the estate in a chapter 13 case is not a separate taxable entity”); accord, S. Rep. No. 95–989, p. 45 (1978) (hereinafter S. Rep.); H. R. Rep., at 335 (noting “the creation of the estate of an individual under chapters 7 or 11 of title 11 as a separate taxable entity”); accord, S. Rep., at 46. The Reports also tie separate taxable entity status to the responsibility to file returns and pay taxes. See H. R. Rep., at 277 (“If the estateis a separate taxable entity, then the representative of the estate is responsible for filing any income tax returns and paying any taxes due by the estate”); id., at 278 (“When the estate is not a separate taxable entity, then taxation of the debtor should be conducted on the same basis as if no petition were filed”). 4 A dispute over Committee jurisdiction led to the insertion of “State or local” before each mention of “law imposing a tax.” Compare H. R. 8200, 95th Cong., 1st Sess., §346 (1977), with §346, 92Stat. 2565. Nonetheless, the House Report underscored that the policy behind §346 applied equally to federal taxes: “[T]here is a strong bankruptcy policy that these provisions apply equally to Federal, State, and local taxes. However, in order to avoid any possible jurisdictional conflict with the Ways and Means Committee over the applicability of these provisions to Federal taxes, H. R. 8200 has been amended to make the sections inapplicable to Federal taxes. The amendment . . . will obviate the need for a sequential referral of the bill to Ways and Means, which will be considering these provisions and other bankruptcy-related tax law later in this Congress.” H. R. Rep., at 275. 5 See, e.g., In re Maxfield, No. 04–60355, 2009 WL 2105953, *5–*6 (Bkrtcy. Ct. ND Ind. 2009); In re Jagours, 236 B.R. 616, 620 (Bkrtcy. Ct. ED Tex. 1999); In re Whall, 391 B.R. 1, 5–6 (Bkrtcy. Ct. Mass. 2008); In re Brown, No. 05–41071, 2006 WL 3370867, *3 (Bkrtcy. Ct. Mass. 2006); In re Gyulafia, 65 B.R. 913, 916 (Bkrtcy. Ct. Kan. 1986). 6 The dissent suggests that Chapter 12 can be distinguished from Chapter 13 because Chapter 12 bankruptcies tend to be longer, such that the treatment of taxes is more “important.” Post, at 13. Asa practical matter, it is not clear that Chapter 12 bankruptcies are substantially longer. Compare Brief for Neil E. Harl. et al. as Amici Curiae 33 (median Chapter 12 case duration is under 8 months) with Tr. of Oral Arg. 49 (“on average we’re talking about 4 months in a chapter 13 case”). In any event, there is no indication that Congress intended any difference in duration—if it anticipated a difference at all—to flip the characterization of postpetition income taxes from one chapter to the other. Nor does the absence of a §1305 equivalent in Chapter 12 justify shoehorning postpetition taxes into §503(b), as the dissent argues. That Chapter 12 lacks a provision allowing such taxes to be brought inside the plan only clarifies that such taxes fall outside of the plan. The dissent alternatively suggests that it “do[es] not see the serious harm in treating the relevant taxes as ‘administrative expenses’ in both Chapter 12 and Chapter 13 cases.” Post, at 13–14. The “harm” is to settled understandings in Chapter 13 to the contrary. The “harm” is also to §1305; to avoid rendering §1305 a nullity, the dissent recasts the provision as applicable not to all “taxes that become payable . . . while the case is pending,” but only those payable “after the Chapter 13 Plan is confirmed.” Post, at 14. The dissent does not claim, however, that this was Congress’ intent for §1305, as Congress’ choice of words would be exceedingly overbroad if it were. And the dissent’s novel reading contravenes ample Chapter 13 authority recognizing no such limitation on §1305’s scope. E.g., 8 Collier ¶1305.02 (citing cases). 7 IRS manuals dating back to 1998 indicate that the Government did not view postpetition federal income taxes as collectible in an individ-ual debtor’s Chapter 12 plan, even when that view was adverse to its interests. See IRM §25.17.12.9.3 (2004); id., §25.17.12.9.3(1) (2002);id., §5.9, ch. 10.8(4) (1999); id., §5.9, ch. 10.8(4) (1998). Until the en-actment of 11 U. S. C. §1222(a)(2)(A), treating such taxes as priority claims in the plan would have assured the Government of full payment before or at the time of the plan. 8 The House Report stated—after noting that, in addition to prepetition taxes, “certain other taxes are entitled to priority”—that “[t]axes arising from the operation of the estate after bankruptcy are entitled to priority as administrative expenses.” H. R. Rep., at 193. That is still true. Many taxes arising after bankruptcy, as in individual-debtor Chapter 7 or 11 cases, remain entitled to priority as administrative expenses. The Senate Report, meanwhile, stated: “In general, administrative expenses include taxes which the trustee incurs in administering the debtor’s estate, including taxes on capital gains from sales of property by the trustee and taxes on income earned by the estate during the case.” S. Rep., at 66 (emphasis added). That likewise remains true. Administrative expenses still include income taxes that “the trustee,” as opposed to the debtor, has incurred—again, as in individual-debtor Chapter 7 or 11 cases. 9 The original §346 established that the estate of a corporate debtoris not a separate taxable entity, but nonetheless provided that “the trustee shall make any [State or local] tax return otherwise required . . . to be filed by or on behalf of such . . . corporation.” §§346(c)(1)–(2), 92Stat. 2565, 2566. The current §346 similarly states, in the same provision deeming the debtor taxable when there is no separate taxable estate, that “[t]he trustee shall make such tax returns of income of corporations . . . . The estate shall be liable for any [State or local] tax imposed on such corporation.” §346(b). 10 The dissent opines that employment taxes must be administrative expenses “incurred by the estate” because, in its view, they “do notfit easily” within the category of administrative expenses under §503(b)(1)(A)(i), notwithstanding the Government’s contrary representations on both points. Post, at 12. Because employment taxes are not at issue in this case, we offer no opinion on either question.

566.US.2011_10-1542

Title 8 U. S. C. §1229b(a) authorizes the Attorney General to cancel the removal of an alien from the United States who, among other things, has held the status of a lawful permanent resident (LPR) for at least five years, §1229b(a)(1), and has lived in the United States for at least seven continuous years after a lawful admission, §1229b(a)(2). These cases concern whether the Board of Immigration Appeals (BIA or Board) should impute a parent’s years of continuous residence or LPR status to his or her child. That issue arises because a child may enter the country lawfully, or may gain LPR status, after one of his parents does—meaning that a parent may satisfy §1229b(a)(1) or §1229b(a)(2), while his child, considered independently, does not. In In re Escobar, 24 I. & N. Dec. 231, the BIA concluded that an alien must meet §1229b(a)’s requirements on his own. But the Ninth Circuit found the Board’s position unreasonable, holding that §1229b(a)(1) and §1229b(a)(2) require imputation. See Mercado-Zazueta v. Holder, 580 F.3d 1102; Cuevas-Gaspar v. Gonzales, 430 F.3d 1013. Respondent Martinez Gutierrez illegally entered the country with his family in 1989, when he was 5 years old. Martinez Gutierrez’s father was lawfully admitted to the country two years later as an LPR. But Martinez Gutierrez was neither lawfully admitted nor given LPR status until 2003. Two years after that, he was apprehended for smuggling undocumented aliens across the border. Admitting the offense, he sought cancellation of removal. The Immigration Judge concluded that Martinez Gutierrez qualified for relief because of his father’s immigration history, even though Martinez Gutierrez could not satisfy §1229b(a)(1) or §1229b(a)(2) on his own. Relying on Escobar, the BIA reversed. The Ninth Circuit then granted Martinez Gutierrez’s petition for review and remanded the case to the Board for reconsideration in light of its contrary decisions. Respondent Sawyers was lawfully admitted as an LPR in October 1995, when he was 15 years old. At that time, his mother had already resided in the country for six consecutive years following a lawful entry. After Sawyers was convicted of a drug offense in August 2002, the Government began removal proceedings. The Immigration Judge found Sawyers ineligible for cancellation of removal because he could not satisfy §1229b(a)(2). The BIA affirmed, and Sawyers petitioned the Ninth Circuit for review. There, he argued that the Board should have counted his mother’s years of residency while he was a minor toward §1229b(a)(2)’s 7-year continuous-residency requirement. The Court of Appeals granted the petition and remanded the case to the BIA. Held: The BIA’s rejection of imputation is based on a permissible construction of §1229b(a). Pp. 6–13. (a) The Board has required each alien seeking cancellation of removal to satisfy §1229b(a)’s requirements on his own, without relying on a parent’s years of continuous residence or immigration status. That position prevails if it is a reasonable construction of the statute, whether or not it is the only possible interpretation or even the one a court might think best. See e.g., Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843–844, and n. 11. The BIA’s approach satisfies this standard. The Board’s position is consistent with the statute’s text. Section 1229b(a) does not mention—much less require—imputation. Instead, it simply calls for “the alien” to meet the prerequisites for cancellation of removal. See §§1101(a)(13)(A) and (a)(33). Respondents contend that this language does not foreclose imputation, but even if so, that is not enough to require the Board to adopt that policy. Pp. 6–7. (b) Neither does the statute’s history and context mandate imputation. Section 1229b(a) replaced former §212(c) of the Immigration and Nationality Act (INA), which allowed the Attorney General to prevent the removal of an alien with LPR status who had maintained a “lawful unrelinquished domicile of seven consecutive years” in this country. Like §1229b(a), §212(c) was silent on imputation. But every Court of Appeals that confronted the question concluded that, in determining eligibility for §212(c) relief, the Board should impute a parent’s years of domicile to his or her child. Based on this history, Sawyers contends that Congress would have understood §1229b(a)’s language to provide for imputation. But in enacting §1229b(a), Congress eliminated the very term—“domicile”—on which the appeals courts had founded their imputation decisions. And the doctrine of congressional ratification applies only when Congress reenacts a statute without relevant change. See Jama v. Immigration and Customs Enforcement, 543 U.S. 335, 349. Nor do the INA’s purposes demand imputation. As respondents correctly observe, many provisions of immigration law advance the goals of promoting family unity and providing relief to aliens with strong ties to this country. But these are not the INA’s only goals, and Congress did not pursue them at all costs. For example, aliens convicted of aggravated felonies are ineligible for cancellation of removal, regardless of the strength of their family ties, see §1229b(a)(3). In addition, as these cases show, not every alien with LPR status can immediately get the same for a spouse or minor child. A silent statute cannot be read as requiring imputation just because that rule would be family-friendly. Pp. 7–10. (c) Respondents advance two additional arguments for why the Board’s position is not entitled to Chevron deference. First, they claim that the Board’s approach to §1229b(a) is arbitrary because it is inconsistent with the Board’s acceptance of imputation under other, similar provisions that are silent on the matter. See §1182(k) and §1181(b). But the Board’s decision in Escobar provided a reasoned explanation for these divergent results: The Board imputes matters involving an alien’s state of mind, while declining to impute objective conditions or characteristics. See 24 I. & N. Dec., at 233–234, and n. 4. Section 1229b(a) hinges on the objective facts of immigration status and place of residence. See id., at 233. So the Board’s approach to §1229b(a) largely follows from one straightforward distinction. Second, respondents claim that the BIA adopted its no-imputation rule only because it thought Congress had left it no other choice. But Escobar belies this contention. The Board did explain how §1229b(a)’s text supports its no-imputation policy. But the Board also brought its experience and expertise to bear on the matter: It noted that there was no precedent in its decisions for imputing status or residence, and it argued that allowing imputation under §1229b(a) would create anomalies in the statutory scheme. Escobar thus expressed the BIA’s view that statutory text, administrative practice, and regulatory policy all pointed toward disallowing imputation. In making that case, the opinion reads like a multitude of agency interpretations to which this and other courts have routinely deferred. Pp. 10–13. No. 10–1542, 411 Fed. Appx. 121; No. 10–1543, 399 Fed. Appx. 313, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court. Notes 1 Together with No. 10–1543, Holder, Attorney General v. Sawyers, also on certiorari to the same court.

An immigration statute, 8 U. S. C. §1229b(a), authorizes the Attorney General to cancel the removal of an alien from the United States so long as the alien satisfies certain criteria. One of those criteria relates to the length of time an alien has lawfully resided in the United States, and another to the length of time he has held permanent resident status here. We consider whether the Board of Immigration Appeals (BIA or Board) could reasonably conclude that an alien living in this country as a child must meet those requirements on his own, without counting a parent’s years of residence or immigration status. We hold that the BIA’s approach is based on a permissible construction of the statute. I A The immigration laws have long given the Attorney General discretion to permit certain otherwise-removable aliens to remain in the United States. See Judulang v. Holder, 565 U. S. ___, ___ (2011) (slip op., at 2–4). The Attorney General formerly exercised this authority by virtue of §212(c) of the Immigration and Nationality Act (INA), 66Stat. 187, 8 U. S. C. §1182(c) (1994 ed.), a provision with some lingering relevance here, see infra, at 7–9. But in 1996, Congress replaced §212(c) with §1229b(a) (2006 ed.). That new section, applicable to the cases before us, provides as follows: “(a) Cancellation of removal for certain permanent residents “The Attorney General may cancel removal in the case of an alien who is inadmissible or deportable from the United States if the alien— “(1) has been an alien lawfully admitted for permanent residence for not less than 5 years, “(2) has resided in the United States continuously for 7 years after having been admitted in any status, and “(3) has not been convicted of any aggravated felony.” Ibid. Section 1229b(a) thus specifies the criteria that make an alien eligible to obtain relief from the Attorney General. The first paragraph requires that the alien have held the status of a lawful permanent resident (LPR) for at least five years. And the second adds that the alien must have lived in the United States for at least seven continuous years after a lawful admission, whether as an LPR or in some other immigration status.[1] (The third paragraph is not at issue in these cases.) The question we consider here is whether, in applying this statutory provision, the BIA should impute a parent’s years of continuous residence or LPR status to his or her child. That question arises because a child may enter the country lawfully, or may gain LPR status, after one of his parents does. A parent may therefore satisfy the re- quirements of §§1229b(a)(1) and (2), while his or her child, considered independently, does not. In these circum- stances, is the child eligible for cancellation of removal? The Ninth Circuit, the first court of appeals to confront this issue, held that such an alien could obtain relief. See Cuevas-Gaspar v. Gonzales, 430 F.3d 1013 (2005). Enrique Cuevas-Gaspar and his parents came to the United States illegally in 1985, when he was one year old. Cuevas-Gaspar’s mother was lawfully admitted to the country in 1990, as an LPR. But Cuevas-Gaspar was lawfully admitted only in 1997, when he too received LPR status. That meant that when Cuevas-Gaspar committed a removable offense in 2002, he could not independently sat- isfy §1229b(a)(2)’s requirement of seven consecutive years of residence after a lawful entry.[2] (The parties agreed that he just met §1229b(a)(1)’s 5-year status requirement.) The Board deemed Cuevas-Gaspar ineligible for relief on that account, but the Ninth Circuit found that position unreasonable. According to the Court of Appeals, the Board should have “imputed” to Cuevas-Gaspar his mother’s years of continuous residence during the time he lived with her as an “unemancipated minor.” Id., at 1029. That approach, the Ninth Circuit reasoned, followed from both the INA’s “priorit[ization]” of familial relations and the Board’s “consistent willingness” to make imputations from a parent to a child in many areas of immigration law. Id., at 1026. The Board responded by reiterating its opposition to imputation under both relevant paragraphs of §1229b(a). In In re Escobar, 24 I. & N. Dec. 231 (2007), the Board considered whether a child could rely on a parent’s period of LPR status to satisfy §1229b(a)(1)’s 5-year clock. The Board expressly “disagree[d] with the reasoning” of Cuevas-Gaspar, rejecting the Ninth Circuit’s understanding of both the statute and the Board’s prior policies. 24 I. & N. Dec., at 233–234, and n. 4. Accordingly, the Board announced that it would “decline to extend” Cuevas-Gaspar to any case involving §1229b(a)(1), and that it would ignore the decision even as to §1229b(a)(2) outside the Ninth Circuit. 24 I. & N. Dec., at 235. A year later, in Matter of Ramirez-Vargas, 24 I. & N. Dec. 599 (2008), the BIA took the final step: It rejected imputation under §1229b(a)(2) in a case arising in the Ninth Circuit, maintaining that the court should abandon Cuevas-Gaspar and defer to the Board’s intervening reasoned decision in Escobar. See Ramirez-Vargas, 24 I. & N. Dec., at 600–601 (citing National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U.S. 967 (2005)). The BIA’s position on imputation touched off a split in the courts of appeals. The Third and Fifth Circuits both deferred to the BIA’s approach as a reasonable construction of §1229b(a). See Augustin v. Attorney Gen., 520 F.3d 264 (CA3 2008); Deus v. Holder, 591 F.3d 807 (CA5 2009). But in Mercado-Zazueta v. Holder, 580 F.3d 1102 (2009), the Ninth Circuit doubled down on its contrary view, declaring the BIA’s position unreasonable and requiring imputation under both §§1229b(a)(1) and (a)(2). See id., at 1103 (“[T]he rationale and holding of Cuevas-Gaspar apply equally to the five-year permanent residence and the seven-year continuance residence requirements” of §1229b(a)). B Two cases are before us. In 1989, at the age of five, respondent Carlos Martinez Gutierrez illegally entered the United States with his family. Martinez Gutierrez’s father was lawfully admitted to the country two years later as an LPR. But Martinez Gutierrez himself was neither lawfully admitted nor given LPR status until 2003. Two years after that, Martinez Gutierrez was apprehended for smuggling undocumented aliens across the border. He admitted the offense, and sought cancellation of removal. The Immigration Judge concluded that Martinez Gutierrez qualified for relief because of his father’s immigration history, even though Martinez Gutierrez could not satisfy either §1229b(a)(1) or §1229b(a)(2) on his own. See App. to Pet. for Cert. in No. 10–1542, pp. 20a–22a (citing Cuevas-Gaspar, 430 F. 3d 1013). The BIA reversed, and after entry of a removal order on remand, reaffirmed its disposition in an order relying on Escobar, see App. to Pet. for Cert. in No. 10–1542, at 5a–6a. The Ninth Circuit then granted Martinez Gutierrez’s petition for review and remanded the case to the Board for reconsideration in light of the court’s contrary decisions. See 411 Fed. Appx. 121 (2011). Respondent Damien Sawyers was lawfully admitted as an LPR in October 1995, when he was 15 years old. At that time, his mother had already resided in the country for six consecutive years following a lawful entry. After Sawyers’s conviction of a drug offense in August 2002, the Government initiated removal proceedings. The Immigration Judge found Sawyers ineligible for cancellation of removal because he was a few months shy of the seven years of continuous residence required under §1229b(a)(2). See App. to Pet. for Cert. in No. 10–1543, p. 13a. (No one doubted that Sawyers had by that time held LPR status for five years, as required under §1229b(a)(1).) The Board affirmed, relying on its reasoning in Escobar. See In re Sawyers, No. A44 852 478, 2007 WL 4711443 (Dec. 26, 2007). Sawyers petitioned the Ninth Circuit for review, arguing that the Board should have counted his mother’s years of residency while he was a minor toward §1229b(a)(2)’s 7-year requirement. As in Gutierrez, the Court of Appeals granted the petition and remanded the case to the BIA. See 399 Fed. Appx. 313 (2010). We granted the Government’s petitions for certiorari, 564 U. S. ___ (2011), consolidated the cases, and now reverse the Ninth Circuit’s judgments. II The Board has required each alien seeking cancellation of removal to satisfy §1229b(a)’s requirements on his own, without counting a parent’s years of continuous residence or LPR status. That position prevails if it is a reasonable construction of the statute, whether or not it is the only possible interpretation or even the one a court might think best. See Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843–844, and n. 11 (1984); see also INS v. Aguirre-Aguirre, 526 U.S. 415, 424–425 (1999) (according Chevron deference to the Board’s interpretations of the INA). We think the BIA’s view on imputation meets that standard, and so need not decide if the statute permits any other construction. The Board’s approach is consistent with the statute’s text, as even respondents tacitly concede. Section 1229b(a) does not mention imputation, much less require it. The provision calls for “the alien”—not, say, “the alien or one of his parents”—to meet the three prerequisites for cancellation of removal. Similarly, several of §1229b(a)’s other terms have statutory definitions referring to only a single individual. See, e.g., §1101(a)(13)(A) (“The terms ‘admission’ and ‘admitted’ mean, with respect to an alien, the lawful entry of the alien into the United States” (emphasis added)); §1101(a)(33) (“The term ‘residence’ means the place of general abode; the place of general abode of a person means his principal, actual dwelling” (emphasis added)). Respondents contend that none of this language “forecloses” imputation: They argue that if the Board allowed imputation, “[t]he alien” seeking cancellation would “still have to satisfy the provision’s durational requirements”—just pursuant to a different computational rule. Brief for Respondent Martinez Gutierrez in No. 10–1542, p. 16 (hereinafter Martinez Gutierrez Brief); see Brief for Respondent Sawyers in No. 10–1543, pp. 11, 15 (hereinafter Sawyers Brief). And they claim that the Board’s history of permitting imputation under similarly “silent” statutes supports this construction. Martinez Gutierrez Brief 16; see Sawyers Brief 15–16; infra, at 10–11. But even if so—even if the Board could adopt an imputation rule consistent with the statute’s text—that would not avail respondents. Taken alone, the language of §1229b(a) at least permits the Board to go the other way—to say that “the alien” must meet the statutory conditions independently, without relying on a parent’s history. For this reason, respondents focus on §1229b(a)’s history and context—particularly, the provision’s relationship to the INA’s former §212(c) and its associated imputation rule. Section 212(c)—§1229b(a)’s predecessor—generally allowed the Attorney General to prevent the removal of an alien with LPR status who had maintained a “lawful unrelinquished domicile of seven consecutive years” in this country. 8 U. S. C. §1182(c) (1994 ed.). Like §1229b(a), §212(c) was silent on imputation. Yet the Second, Third, and Ninth Circuits (the only appellate courts to consider the question) concluded that, in determining eligibility for relief under §212(c), the Board should impute a parent’s years of domicile to his or her child. See Rosario v. INS, 962 F.2d 220 (CA2 1992); Lepe-Guitron v. INS, 16 F.3d 1021, 1024–1026 (CA9 1994); Morel v. INS, 90 F.3d 833, 840–842 (CA3 1996). Those courts reasoned that at common law, a minor’s domicile was “the same as that of its parents, since most children are presumed not legally capable of forming the requisite intent to establish their own domicile.” Rosario, 962 F. 2d, at 224; see Mississippi Band of Choctaw Indians v. Holyfield, 490 U.S. 30, 48 (1989) (defining “domicile” as “physical presence in a place in connection with a certain state of mind concerning one’s intent to remain there”). So by the time Congress replaced §212(c) with §1229b(a), the BIA often imputed a parent’s years of domicile to a child in determining eligibility for cancellation of removal. Sawyers argues that against this backdrop, Congress “would have understood the language it chose [in §1229b(a)] to provide for imputation.” Sawyers Brief 10. But we cannot conclude that Congress ratified an imputation requirement when it passed §1229b(a). As all parties agree, Congress enacted §§1229b(a)(1) and (a)(2) to resolve an unrelated question about §212(c)’s meaning. See id., at 17; Martinez Gutierrez Brief 28; Brief for Petitioner 25. Courts had differed on whether an alien’s “seven consecutive years” of domicile under §212(c) all had to post-date the alien’s obtaining LPR status. See Cuevas-Gaspar, 430 F. 3d, at 1027–1028 (canvassing split). Congress addressed that split by creating two distinct durational conditions: the 5-year status requirement of subsection (a)(1), which runs from the time an alien becomes an LPR, and the 7-year continuous-residency requirement of subsection (a)(2), which can include years preceding the acquisition of LPR status. In doing so, Congress elimi- nated the very term—“domicile”—on which the appeals courts had founded their imputation decisions. See supra, at 8. That alteration dooms respondents’ position, because the doctrine of congressional ratification applies only when Congress reenacts a statute without relevant change. See Jama v. Immigration and Customs Enforcement, 543 U.S. 335, 349 (2005).[3] So the statutory history here provides no basis for holding that the BIA flouted a congressional command in adopting its no-imputation policy. Nor do the INA’s purposes demand imputation here, as both respondents claim. According to Martinez Gutierrez, the BIA’s approach contradicts that statute’s objectives of “providing relief to aliens with strong ties to the United States” and “promoting family unity.” Martinez Gutierrez Brief 40, 44; see Sawyers Brief 37. We agree—indeed, we have stated—that the goals respondents identify underlie or inform many provisions of immigration law. See Fiallo v. Bell, 430 U.S. 787, 795, n. 6 (1977); INS v. Errico, 385 U.S. 214, 220 (1966). But they are not the INA’s only goals, and Congress did not pursue them to the nth degree. To take one example, §1229b(a)’s third paragraph makes aliens convicted of aggravated felonies ineligible for cancellation of removal, regardless of the strength of their family ties. See §1229b(a)(3). And more generally—as these very cases show—not every alien who obtains LPR status can immediately get the same for her spouse or minor children. See Brief for Petitioner 31–32, and n. 9 (providing program-specific examples). We cannot read a silent statute as requiring (not merely allowing) imputation just because that rule would be family-friendly. Respondents’ stronger arguments take a different tack—that we should refuse to defer to the Board’s decision even assuming Congress placed the question of imputation in its hands. Respondents offer two main reasons. First, they contend that the Board’s approach to §1229b(a) cannot be squared with its acceptance of imputation under other, similar statutory provisions. This “wil[d]” and “ ‘[u]n- explained inconsistency,’ ” Sawyers asserts, is the very “paradigm of arbitrary agency action.” Sawyers Brief 13, 41 (emphasis deleted); see Martinez Gutierrez Brief 52–54. Second, they argue that the Board did not appreciate its own discretion over whether to allow imputation. The Board, they say, thought Congress had forbidden imputation, and so did not bring its “ ‘experience and expertise to bear’ ” on the issue. Id., at 31 (quoting PDK Labs. Inc. v. DEA, 362 F.3d 786, 797 (CADC 2004)); see Sawyers Brief 38–39. These arguments are not insubstantial, but in the end neither persuades us to deny the Board the usual deference we accord to agency interpretations. Start with the claim of inconsistency. The BIA has indeed imputed parental attributes to children under other INA provisions that do not mention the matter. Section 1182(k), for example, enables the Attorney General to let certain inadmissible aliens into the country if he finds “that inadmissibility was not known to, and could not have been ascertained by the exercise of reasonable diligence by, the immigrant before the time of departure.” Like §1229b(a), that provision refers to a single person (“the immigrant”) and says nothing about imputation. But the BIA has consistently imputed a parent’s knowledge of inadmissibility (or lack thereof) to a child. See, e.g., Senica v. INS, 16 F.3d 1013, 1015 (CA9 1994) (“Therefore, the BIA reasoned, the children were not entitled to relief under [§1182(k)] because [their mother’s] knowledge was imputed to them”); In re Mushtaq, No. A43 968 082, 2007 WL 4707539 (BIA, Dec. 10, 2007) (per curiam); In re Ahmed, No. A41 982 631, 2006 WL 448156 (BIA, Jan. 17, 2006) (per curiam). Similarly, the Board imputes a parent’s abandonment (or non-abandonment) of LPR status to her child when determining whether that child can reenter the country as a “returning resident immigran[t]” under §1181(b). See Matter of Zamora, 17 I. & N. Dec. 395, 396 (1980) (hold- ing that a “voluntary and intended abandonment by the mother is imputed” to an unemancipated minor child for purposes of applying §1181(b)); Matter of Huang, 19 I. & N. Dec. 749, 755–756 (1988) (concluding that a mother and her children abandoned their LPR status based solely on the mother’s intent); In re Ali, No. A44 143 723, 2006 WL 3088820 (BIA, Sept. 11, 2006) (holding that a child could not have abandoned his LPR status if his mother had not abandoned hers). And once again, that is so even though neither §1181(b) nor any other statutory provision says that the BIA should look to the parent in assessing the child’s eligibility for reentry. But Escobar provided a reasoned explanation for these divergent results: The Board imputes matters involving an alien’s state of mind, while declining to impute objective conditions or characteristics. See 24 I. & N. Dec., at 233–234, and n. 4. On one side of the line, knowledge of inadmissibility is all and only about a mental state. See, e.g., Senica, 16 F. 3d, at 1015; In re Ahmed, 2006 WL 448156. Likewise, abandonment of status turns on an alien’s “intention of . . . returning to the United States” to live as a permanent resident, Zamora, 17 I. & N. Dec., at 396; the Board thus explained that imputing abandonment is “consistent with the . . . longstanding policy that a child cannot form the intent necessary to establish his or her own domicile,” Escobar, 24 I. & N. Dec., at 234, n. 4. And as that analogy recalls, the 7-year domicile requirement of the former §212(c) also involved intent and so lent itself to imputation. See Rosario, 962 F. 2d, at 224; supra, at 8. But the 5- and 7-year clocks of §1229b(a) fall on the other side of the line, because they hinge not on any state of mind but on the objective facts of immigration status and place of residence. See Escobar, 24 I. & N. Dec., at 233 (“[W]e find that residence is different from domicile because it ‘contains no element of subjective intent’ ” (quoting Cuevas-Gaspar, 430 F. 3d, at 1031 (Fernandez, J., dissenting))). The BIA’s varied rulings on imputation thus largely follow from one straightforward distinction.[4] Similarly, Escobar belies respondents’ claim that the BIA adopted its no-imputation rule only because it thought Congress had left it no other choice. The Board, to be sure, did not highlight the statute’s gaps or ambiguity; rather, it read §1229b(a)’s text to support its conclusion that each alien must personally meet that section’s durational requirements. See 24 I. & N. Dec., at 235. But the Board also explained that “there [was] no precedent” in its decisions for imputing status or residence, and distinguished those statutory terms, on the ground just explained, from domicile or abandonment of LPR status. Id., at 234; see id., at 233–234, and n. 4. And the Board argued that allowing imputation under §1229b(a) would create anomalies in administration of the statutory scheme by permitting even those who had not obtained LPR status—or could not do so because of a criminal history—to become eligible for cancellation of removal. See id., at 234–235, and n. 5. The Board therefore saw neither a “logical” nor a “legal” basis for adopting a policy of imputation. Id., at 233. We see nothing in this decision to suggest that the Board thought its hands tied, or that it might have reached a different result if assured it could do so. To the contrary, the decision expressed the BIA’s view, based on its experience implementing the INA, that statutory text, administrative practice, and regulatory policy all pointed in one direction: toward disallowing imputation. In making that case, the decision reads like a multitude of agency interpretations—not the best example, but far from the worst—to which we and other courts have routinely deferred. We see no reason not to do so here. Because the Board’s rejection of imputation under §1229b(a) is “based on a permissible construction of the statute,” Chevron, 467 U. S., at 843, we reverse the Ninth Circuit’s judgments and remand the cases for further proceedings consistent with this opinion. It is so ordered. Notes 1 The INA defines “admitted” as referring to “the lawful entry ofthe alien into the United States after inspection and authorization by an immigration officer.” 8 U. S. C. §1101 (a)(13)(A). The 7-year clock of §1229b(a)(2) thus begins with an alien’s lawful entry. 2 The 7-year clock stopped running on the date of Cuevas-Gaspar’s offense under a statutory provision known as the “stop-time” rule. See §1229b(d)(1) (“For purposes of this section, any period of continuous residence . . . in the United States shall be deemed to end . . . when the alien is served a notice to appear . . . or . . . when the alien has committed an offense . . . that renders the alien . . . removable from the United States . . . , whichever is earliest”). 3 Sawyers contends that §1229b(a)(2)’s replacement term—“resided continuously”—is a “term of art” in the immigration context which incorporates “an intent component” and so means the same thing as “domiciled.” Sawyers Brief 25–26 (emphasis deleted). Thus, Sawyers argues, we should read §1229b(a) as reenacting §212(c) without meaningful change. See id., at 25. But even assuming that Congress could ratify judicial decisions based on the term “domicile” through a new statute using a synonym for that term, we do not think “resided continuously” qualifies. The INA defines “residence” as a person’s “princi-pal, actual dwelling place in fact, without regard to intent,” 8 U. S. C. §1101(a)(33) (emphasis added), and we find nothing to suggest that Congress added an intent element, inconsistent with that definition, by requiring that the residence have been maintained “continuously for 7 years.” 4 Respondents aver that the BIA deviates from this principle in imputing to a child his parent’s “ ‘firm resettlement’ ” in another country, which renders an alien ineligible for asylum without regard to intent. See Sawyers Brief 39; Martinez Gutierrez Brief 52. But the Government denies that it has a “settled imputation rule” in that context. Reply Brief for Petitioner 13. And the sources on which respondents rely are slender reeds: a 40-year old ruling by a regional commissioner (not the Board itself) that considered the conduct of both the parents and the child, see Matter of Ng, 12 I. & N. Dec. 411 (1967), and a Ninth Circuit decision imputing a parent’s resettlement even though the Board had focused only on the child’s actions, see Vang v. INS, 146 F.3d 1114, 1117 (1998). Based on these scant decisions, we cannot conclude that the Board has any policy on imputing resettlement, let alone one inconsistent with Escobar.

565.US.2011_10-553

Petitioner Hosanna-Tabor Evangelical Lutheran Church and School is a member congregation of the Lutheran Church–Missouri Synod. The Synod classifies its school teachers into two categories: “called” and “lay.” “Called” teachers are regarded as having been called to their vocation by God. To be eligible to be considered “called,” a teacher must complete certain academic requirements, including a course of theological study. Once called, a teacher receives the formal title “Minister of Religion, Commissioned.” “Lay” teachers, by contrast, are not required to be trained by the Synod or even to be Lutheran. Although lay and called teachers at Hosanna-Tabor generally performed the same duties, lay teachers were hired only when called teachers were unavailable. After respondent Cheryl Perich completed the required training, Hosanna-Tabor asked her to become a called teacher. Perich accepted the call and was designated a commissioned minister. In addition to teaching secular subjects, Perich taught a religion class, led her students in daily prayer and devotional exercises, and took her students to a weekly school-wide chapel service. Perich led the chapel service herself about twice a year. Perich developed narcolepsy and began the 2004–2005 school year on disability leave. In January 2005, she notified the school principal that she would be able to report to work in February. The principal responded that the school had already contracted with a lay teacher to fill Perich’s position for the remainder of the school year. The principal also expressed concern that Perich was not yet ready to return to the classroom. The congregation subsequently offered to pay a portion of Perich’s health insurance premiums in exchange for her resignation as a called teacher. Perich refused to resign. In February, Perich presented herself at the school and refused to leave until she received written documentation that she had reported to work. The principal later called Perich and told her that she would likely be fired. Perich responded that she had spoken with an attorney and intended to assert her legal rights. In a subsequent letter, the chairman of the school board advised Perich that the congregation would consider whether to rescind her call at its next meeting. As grounds for termination, the letter cited Perich’s “insubordination and disruptive behavior,” as well as the damage she had done to her “working relationship” with the school by “threatening to take legal action.” The congregation voted to rescind Perich’s call, and Hosanna-Tabor sent her a letter of termination. Perich filed a charge with the Equal Employment Opportunity Commission, claiming that her employment had been terminated in violation of the Americans with Disabilities Act. The EEOC brought suit against Hosanna-Tabor, alleging that Perich had been fired in retaliation for threatening to file an ADA lawsuit. Perich intervened in the litigation. Invoking what is known as the “ministerial exception,” Hosanna-Tabor argued that the suit was barred by the First Amendment because the claims concerned the employment relationship between a religious institution and one of its ministers. The District Court agreed and granted summary judgment in Hosanna-Tabor’s favor. The Sixth Circuit vacated and remanded. It recognized the existence of a ministerial exception rooted in the First Amendment, but concluded that Perich did not qualify as a “minister” under the exception. Held: 1. The Establishment and Free Exercise Clauses of the First Amendment bar suits brought on behalf of ministers against their churches, claiming termination in violation of employment discrimination laws. Pp. 6–15. (a) The First Amendment provides, in part, that “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof.” Familiar with life under the established Church of England, the founding generation sought to foreclose the possibility of a national church. By forbidding the “establishment of religion” and guaranteeing the “free exercise thereof,” the Religion Clauses ensured that the new Federal Government—unlike the English Crown—would have no role in filling ecclesiastical offices. Pp. 6–10. (b) This Court first considered the issue of government interference with a church’s ability to select its own ministers in the context of disputes over church property. This Court’s decisions in that area confirm that it is impermissible for the government to contradict a church’s determination of who can act as its ministers. See Watson v. Jones, 13 Wall. 679; Kedroff v. Saint Nicholas Cathedral of Russian Orthodox Church in North America, 344 U.S. 94; Serbian Eastern Orthodox Diocese for United States and Canada v. Milivojevich, 426 U.S. 696. Pp. 10–12. (c) Since the passage of Title VII of the Civil Rights Act of 1964 and other employment discrimination laws, the Courts of Appeals have uniformly recognized the existence of a “ministerial exception,” grounded in the First Amendment, that precludes application of such legislation to claims concerning the employment relationship between a religious institution and its ministers. The Court agrees that there is such a ministerial exception. Requiring a church to accept or retain an unwanted minister, or punishing a church for failing to do so, intrudes upon more than a mere employment decision. Such action interferes with the internal governance of the church, depriving the church of control over the selection of those who will personify its beliefs. By imposing an unwanted minister, the state infringes the Free Exercise Clause, which protects a religious group’s right to shape its own faith and mission through its appointments. According the state the power to determine which individuals will minister to the faithful also violates the Establishment Clause, which prohibits government involvement in such ecclesiastical decisions. The EEOC and Perich contend that religious organizations can defend against employment discrimination claims by invoking their First Amendment right to freedom of association. They thus see no need—and no basis—for a special rule for ministers grounded in the Religion Clauses themselves. Their position, however, is hard to square with the text of the First Amendment itself, which gives special solicitude to the rights of religious organizations. The Court cannot accept the remarkable view that the Religion Clauses have nothing to say about a religious organization’s freedom to select its own ministers. The EEOC and Perich also contend that Employment Div., Dept. of Human Resources of Ore. v. Smith, 494 U.S. 872, precludes recognition of a ministerial exception. But Smith involved government regulation of only outward physical acts. The present case, in contrast, concerns government interference with an internal church decision that affects the faith and mission of the church itself. Pp. 13–15. 2. Because Perich was a minister within the meaning of the ministerial exception, the First Amendment requires dismissal of this employment discrimination suit against her religious employer. Pp. 15–21. (a) The ministerial exception is not limited to the head of a religious congregation. The Court, however, does not adopt a rigid formula for deciding when an employee qualifies as a minister. Here, it is enough to conclude that the exception covers Perich, given all the circumstances of her employment. Hosanna-Tabor held her out as a minister, with a role distinct from that of most of its members. That title represented a significant degree of religious training followed by a formal process of commissioning. Perich also held herself out as a minister by, for example, accepting the formal call to religious service. And her job duties reflected a role in conveying the Church’s message and carrying out its mission: As a source of religious instruction, Perich played an important part in transmitting the Lutheran faith. In concluding that Perich was not a minister under the exception, the Sixth Circuit committed three errors. First, it failed to see any relevance in the fact that Perich was a commissioned minister. Although such a title, by itself, does not automatically ensure coverage, the fact that an employee has been ordained or commissioned as a minister is surely relevant, as is the fact that significant religious training and a recognized religious mission underlie the description of the employee’s position. Second, the Sixth Circuit gave too much weight to the fact that lay teachers at the school performed the same religious duties as Perich. Though relevant, it cannot be dispositive that others not formally recognized as ministers by the church perform the same functions—particularly when, as here, they did so only because commissioned ministers were unavailable. Third, the Sixth Circuit placed too much emphasis on Perich’s performance of secular duties. Although the amount of time an employee spends on particular activities is relevant in assessing that employee’s status, that factor cannot be considered in isolation, without regard to the other considerations discussed above. Pp. 15–19. (b) Because Perich was a minister for purposes of the exception, this suit must be dismissed. An order reinstating Perich as a called teacher would have plainly violated the Church’s freedom under the Religion Clauses to select its own ministers. Though Perich no longer seeks reinstatement, she continues to seek frontpay, backpay, compensatory and punitive damages, and attorney’s fees. An award of such relief would operate as a penalty on the Church for terminating an unwanted minister, and would be no less prohibited by the First Amendment than an order overturning the termination. Such relief would depend on a determination that Hosanna-Tabor was wrong to have relieved Perich of her position, and it is precisely such a ruling that is barred by the ministerial exception. Any suggestion that Hosanna-Tabor’s asserted religious reason for firing Perich was pretextual misses the point of the ministerial exception. The purpose of the exception is not to safeguard a church’s decision to fire a minister only when it is made for a religious reason. The exception instead ensures that the authority to select and control who will minister to the faithful is the church’s alone. Pp. 19–20. (c) Today the Court holds only that the ministerial exception bars an employment discrimination suit brought on behalf of a minister, challenging her church’s decision to fire her. The Court expresses no view on whether the exception bars other types of suits. Pp. 20–21. 597 F.3d 769, reversed. Roberts, C. J., delivered the opinion for a unanimous Court. Thomas, J., filed a concurring opinion. Alito, J., filed a concurring opinion, in which Kagan, J., joined.

Certain employment discrimination laws authorize employees who have been wrongfully terminated to sue their employers for reinstatement and damages. The question presented is whether the Establishment and Free Exercise Clauses of the First Amendment bar such an action when the employer is a religious group and the employee is one of the group’s ministers. I A Petitioner Hosanna-Tabor Evangelical Lutheran Church and School is a member congregation of the Lutheran Church–Missouri Synod, the second largest Lutheran denomination in America. Hosanna-Tabor operated a small school in Redford, Michigan, offering a “Christ-centered education” to students in kindergarten through eighth grade. 582 F. Supp. 2d 881, 884 (ED Mich. 2008) (internal quotation marks omitted). The Synod classifies teachers into two categories: “called” and “lay.” “Called” teachers are regarded as having been called to their vocation by God through a congregation. To be eligible to receive a call from a congregation, a teacher must satisfy certain academic requirements. One way of doing so is by completing a “colloquy” program at a Lutheran college or university. The program requires candidates to take eight courses of theological study, obtain the endorsement of their local Synod district, and pass an oral examination by a faculty committee. A teacher who meets these requirements may be called by a congregation. Once called, a teacher receives the formal title “Minister of Religion, Commissioned.” App. 42, 48. A commissioned minister serves for an open-ended term; at Hosanna-Tabor, a call could be rescinded only for cause and by a supermajority vote of the congregation. “Lay” or “contract” teachers, by contrast, are not required to be trained by the Synod or even to be Lutheran. At Hosanna-Tabor, they were appointed by the school board, without a vote of the congregation, to one-year renewable terms. Although teachers at the school generally performed the same duties regardless of whether they were lay or called, lay teachers were hired only when called teachers were unavailable. Respondent Cheryl Perich was first employed by Hosanna-Tabor as a lay teacher in 1999. After Perich com-pleted her colloquy later that school year, Hosanna-Tabor asked her to become a called teacher. Perich accepted the call and received a “diploma of vocation” designating her a commissioned minister. Id., at 42. Perich taught kindergarten during her first four years at Hosanna-Tabor and fourth grade during the 2003–2004 school year. She taught math, language arts, social studies, science, gym, art, and music. She also taught a religion class four days a week, led the students in prayer and devotional exercises each day, and attended a weekly school-wide chapel service. Perich led the chapel service herself about twice a year. Perich became ill in June 2004 with what was eventually diagnosed as narcolepsy. Symptoms included sudden and deep sleeps from which she could not be roused. Because of her illness, Perich began the 2004–2005 school year on disability leave. On January 27, 2005, however, Perich notified the school principal, Stacey Hoeft, that she would be able to report to work the following month. Hoeft responded that the school had already contracted with a lay teacher to fill Perich’s position for the remainder of the school year. Hoeft also expressed concern that Perich was not yet ready to return to the classroom. On January 30, Hosanna-Tabor held a meeting of its congregation at which school administrators stated that Perich was unlikely to be physically capable of returning to work that school year or the next. The congregation voted to offer Perich a “peaceful release” from her call, whereby the congregation would pay a portion of her health insurance premiums in exchange for her resignation as a called teacher. Id., at 178, 186. Perich refused to resign and produced a note from her doctor stating that she would be able to return to work on February 22. The school board urged Perich to reconsider, informing her that the school no longer had a position for her, but Perich stood by her decision not to resign. On the morning of February 22—the first day she was medically cleared to return to work—Perich presented herself at the school. Hoeft asked her to leave but she would not do so until she obtained written documentation that she had reported to work. Later that afternoon, Hoeft called Perich at home and told her that she would likely be fired. Perich responded that she had spoken with an attorney and intended to assert her legal rights. Following a school board meeting that evening, board chairman Scott Salo sent Perich a letter stating that Hosanna-Tabor was reviewing the process for rescinding her call in light of her “regrettable” actions. Id., at 229. Salo subsequently followed up with a letter advising Perich that the congregation would consider whether to rescind her call at its next meeting. As grounds for ter-mination, the letter cited Perich’s “insubordination and disruptive behavior” on February 22, as well as the damage she had done to her “working relationship” with the school by “threatening to take legal action.” Id., at 55. The congregation voted to rescind Perich’s call on April 10, and Hosanna-Tabor sent her a letter of termination the next day. B Perich filed a charge with the Equal Employment Opportunity Commission, alleging that her employment had been terminated in violation of the Americans with Dis-abilities Act, 104Stat. 327, 42 U. S. C. §12101 et seq. (1990). The ADA prohibits an employer from discriminating against a qualified individual on the basis of disability. §12112(a). It also prohibits an employer from retaliating “against any individual because such individual has opposed any act or practice made unlawful by [the ADA] or because such individual made a charge, testified, assisted, or participated in any manner in an investigation, proceeding, or hearing under [the ADA].” §12203(a). [ 1 ] The EEOC brought suit against Hosanna-Tabor, alleging that Perich had been fired in retaliation for threat-ening to file an ADA lawsuit. Perich intervened in the litigation, claiming unlawful retaliation under both the ADA and the Michigan Persons with Disabilities Civil Rights Act, Mich. Comp. Laws §37.1602(a) (1979). The EEOC and Perich sought Perich’s reinstatement to her former position (or frontpay in lieu thereof), along with backpay, compensatory and punitive damages, attorney’s fees, and other injunctive relief. Hosanna-Tabor moved for summary judgment. Invoking what is known as the “ministerial exception,” the Church argued that the suit was barred by the First Amendment because the claims at issue concerned the employment relationship between a religious institution and one of its ministers. According to the Church, Perich was a minister, and she had been fired for a religious reason—namely, that her threat to sue the Church vio-lated the Synod’s belief that Christians should resolve their disputes internally. The District Court agreed that the suit was barred by the ministerial exception and granted summary judgment in Hosanna-Tabor’s favor. The court explained that “Hosanna-Tabor treated Perich like a minister and held her out to the world as such long before this litigation began,” and that the “facts surrounding Perich’s employment in a religious school with a sectarian mission” supported the Church’s characterization. 582 F. Supp. 2d, at 891–892. In light of that determination, the court concluded that it could “inquire no further into her claims of retaliation.” Id., at 892. The Court of Appeals for the Sixth Circuit vacated and remanded, directing the District Court to proceed to the merits of Perich’s retaliation claims. The Court of Appeals recognized the existence of a ministerial exception barring certain employment discrimination claims against religious institutions—an exception “rooted in the First Amendment’s guarantees of religious freedom.” 597 F.3d 769, 777 (2010). The court concluded, however, that Perich did not qualify as a “minister” under the exception, noting in particular that her duties as a called teacher were identical to her duties as a lay teacher. Id., at 778–781. Judge White concurred. She viewed the question whether Perich qualified as a minister to be closer than did the majority, but agreed that the “fact that the duties of the contract teachers are the same as the duties of the called teachers is telling.” Id., at 782, 784. We granted certiorari. 563 U. S. ___ (2011). II The First Amendment provides, in part, that “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof.” We have said that these two Clauses “often exert conflicting pressures,” Cutter v. Wilkinson, 544 U.S. 709, 719 (2005) , and that there can be “internal tension . . . between the Establishment Clause and the Free Exercise Clause,” Tilton v. Richardson, 403 U.S. 672, 677 (1971) (plurality opinion). Not so here. Both Religion Clauses bar the government from interfering with the decision of a religious group to fire one of its ministers. A Controversy between church and state over religious offices is hardly new. In 1215, the issue was addressed in the very first clause of Magna Carta. There, King John agreed that “the English church shall be free, and shall have its rights undiminished and its liberties unimpaired.” The King in particular accepted the “freedom of elections,” a right “thought to be of the greatest necessity and importance to the English church.” J. Holt, Magna Carta App. IV, p. 317, cl. 1 (1965). That freedom in many cases may have been more the-oretical than real. See, e.g., W. Warren, Henry II 312 (1973) (recounting the writ sent by Henry II to the electors of a bishopric in Winchester, stating: “I order you to hold a free election, but forbid you to elect anyone but Richard my clerk”). In any event, it did not survive the reign of Henry VIII, even in theory. The Act of Supremacy of 1534, 26 Hen. 8, ch. 1, made the English monarch the supreme head of the Church, and the Act in Restraint of Annates, 25 Hen. 8, ch. 20, passed that same year, gave him the authority to appoint the Church’s high officials. See G. Elton, The Tudor Constitution: Documents and Commentary 331–332 (1960). Various Acts of Uniformity, enacted subsequently, tightened further the government’s grip on the exercise of religion. See, e.g., Act of Uniformity, 1559, 1 Eliz., ch. 2; Act of Uniformity, 1549, 2 & 3 Edw. 6, ch. 1. The Uniformity Act of 1662, for instance, limited service as a minister to those who formally assented to prescribed tenets and pledged to follow the mode of worship set forth in the Book of Common Prayer. Any minister who refused to make that pledge was “deprived of all his Spiritual Promotions.” Act of Uniformity, 1662, 14 Car. 2, ch. 4. Seeking to escape the control of the national church, the Puritans fled to New England, where they hoped to elect their own ministers and establish their own modes of worship. See T. Curry, The First Freedoms: Church and State in America to the Passage of the First Amendment 3 (1986); McConnell, The Origins and Historical Understanding of Free Exercise of Religion, 103 Harv. L. Rev. 1409, 1422 (1990). William Penn, the Quaker proprietor of what would eventually become Pennsylvania and Delaware, also sought independence from the Church of England. The charter creating the province of Pennsylvania contained no clause establishing a religion. See S. Cobb, The Rise of Religious Liberty in America 440–441 (1970). Colonists in the South, in contrast, brought the Church of England with them. But even they sometimes chafed at the control exercised by the Crown and its representatives over religious offices. In Virginia, for example, the law vested the governor with the power to induct ministers presented to him by parish vestries, 2 Hening’s Statutes at Large 46 (1642), but the vestries often refused to make such presentations and instead chose ministers on their own. See H. Eckenrode, Separation of Church and State in Virginia 13–19 (1910). Controversies over the selection of ministers also arose in other Colonies with Anglican establishments, including North Carolina. See C. Antieau, A. Downey, & E. Roberts, Freedom from Federal Establishment: Formation and Early History of the First Amendment Religion Clauses 10–11 (1964). There, the royal governor insisted that the right of presentation lay with the Bishop of London, but the colonial assembly enacted laws placing that right in the vestries. Authorities in England intervened, repealing those laws as inconsistent with the rights of the Crown. See id., at 11; Weeks, Church and State in North Carolina, Johns Hopkins U. Studies in Hist. & Pol. Sci., 11th Ser., Nos. 5–6, pp. 29–36 (1893). It was against this background that the First Amendment was adopted. Familiar with life under the established Church of England, the founding generation sought to foreclose the possibility of a national church. See 1 Annals of Cong. 730–731 (1789) (noting that the Establishment Clause addressed the fear that “one sect might obtain a pre-eminence, or two combine together, and establish a religion to which they would compel others to conform” (remarks of J. Madison)). By forbidding the “establishment of religion” and guaranteeing the “free exercise thereof,” the Religion Clauses ensured that the new Federal Government—unlike the English Crown—would have no role in filling ecclesiastical offices. The Establishment Clause prevents the Government from appointing ministers, and the Free Exercise Clause prevents it from interfering with the freedom of religious groups to select their own. This understanding of the Religion Clauses was reflected in two events involving James Madison, “ ‘the leading architect of the religion clauses of the First Amendment.’ ” Arizona Christian School Tuition Organization v. Winn, 563 U. S. ___, ___ (2011) (slip op., at 13) (quoting Flast v. Cohen, 392 U.S. 83, 103 (1968) ). The first occurred in 1806, when John Carroll, the first Catholic bishop in the United States, solicited the Executive’s opinion on who should be appointed to direct the affairs of the Catholic Church in the territory newly acquired by the Louisiana Purchase. After consulting with President Jefferson, then-Secretary of State Madison responded that the selection of church “functionaries” was an “entirely ecclesiastical” matter left to the Church’s own judgment. Letter from James Madison to Bishop Carroll (Nov. 20, 1806), reprinted in 20 Records of the American Catholic Historical Society 63 (1909). The “scrupulous policy of the Constitution in guarding against a political interference with religious affairs,” Madison explained, prevented the Government from rendering an opinion on the “selection of ecclesiastical individuals.” Id., at 63–64. The second episode occurred in 1811, when Madison was President. Congress had passed a bill incorporating the Protestant Episcopal Church in the town of Alexandria in what was then the District of Columbia. Madison vetoed the bill, on the ground that it “exceeds the rightful authority to which Governments are limited, by the essential distinction between civil and religious functions, and violates, in particular, the article of the Constitution of the United States, which declares, that ‘Congress shall make no law respecting a religious establishment.’ ” 22 Annals of Cong. 982–983 (1811). Madison explained: “The bill enacts into, and establishes by law, sundry rules and proceedings relative purely to the organi-zation and polity of the church incorporated, and comprehending even the election and removal of the Minister of the same; so that no change could be made therein by the particular society, or by the general church of which it is a member, and whose authority it recognises.” Id., at 983 (emphasis added). B Given this understanding of the Religion Clauses—and the absence of government employment regulation generally—it was some time before questions about government interference with a church’s ability to select its own ministers came before the courts. This Court touched upon the issue indirectly, however, in the context of disputes over church property. Our decisions in that area confirm that it is impermissible for the government to contradict a church’s determination of who can act as its ministers. In Watson v. Jones, 13 Wall. 679 (1872), the Court considered a dispute between antislavery and proslavery factions over who controlled the property of the Walnut Street Presbyterian Church in Louisville, Kentucky. The General Assembly of the Presbyterian Church had recognized the antislavery faction, and this Court—applying not the Constitution but a “broad and sound view of the relations of church and state under our system of laws”—declined to question that determination. Id., at 727. We explained that “whenever the questions of discipline, or of faith, or ecclesiastical rule, custom, or law have been decided by the highest of [the] church judicatories to which the matter has been carried, the legal tribunals must accept such decisions as final, and as binding on them.” Ibid. As we would put it later, our opinion in Watson “radiates . . . a spirit of freedom for religious organizations, an independence from secular control or manipulation—in short, power to decide for themselves, free from state interference, matters of church government as well as those of faith and doctrine.” Kedroff v. Saint Nicholas Cathedral of Russian Orthodox Church in North America, 344 U.S. 94, 116 (1952) . Confronting the issue under the Constitution for the first time in Kedroff, the Court recognized that the “[f ]reedom to select the clergy, where no improper methods of choice are proven,” is “part of the free exercise of religion” protected by the First Amendment against government interference. Ibid. At issue in Kedroff was the right to use a Russian Orthodox cathedral in New York City. The Russian Orthodox churches in North America had split from the Supreme Church Authority in Moscow, out of concern that the Authority had become a tool of the Soviet Government. The North American churches claimed that the right to use the cathedral belonged to an archbishop elected by them; the Supreme Church Authority claimed that it belonged instead to an archbishop appointed by the patriarch in Moscow. New York’s highest court ruled in favor of the North American churches, based on a state law requiring every Russian Orthodox church in New York to recognize the determination of the governing body of the North American churches as authoritative. Id., at 96–97, 99, n. 3, 107, n. 10. This Court reversed, concluding that the New York law violated the First Amendment. Id., at 107. We explained that the controversy over the right to use the cathedral was “strictly a matter of ecclesiastical government, the power of the Supreme Church Authority of the Russian Orthodox Church to appoint the ruling hierarch of the archdiocese of North America.” Id., at 115. By “pass[ing] the control of matters strictly ecclesiastical from one church authority to another,” the New York law intruded the “power of the state into the forbidden area of reli- gious freedom contrary to the principles of the First Amend-ment.” Id., at 119. Accordingly, we declared the law unconstitutional because it “directly prohibit[ed] the free exercise of an ecclesiastical right, the Church’s choice of its hierarchy.” Ibid. This Court reaffirmed these First Amendment principles in Serbian Eastern Orthodox Diocese for United States and Canada v. Milivojevich, 426 U.S. 696 (1976) , a case involving a dispute over control of the American-Canadian Diocese of the Serbian Orthodox Church, including its property and assets. The Church had removed Dionisije Milivojevich as bishop of the American-Canadian Diocese because of his defiance of the church hierarchy. Following his removal, Dionisije brought a civil action in state court challenging the Church’s decision, and the Illinois Supreme Court “purported in effect to reinstate Dionisije as Diocesan Bishop,” on the ground that the proceedings resulting in his removal failed to comply with church laws and regulations. Id., at 708. Reversing that judgment, this Court explained that the First Amendment “permit[s] hierarchical religious organizations to establish their own rules and regulations for internal discipline and government, and to create tribunals for adjudicating disputes over these matters.” Id., at 724. When ecclesiastical tribunals decide such disputes, we further explained, “the Constitution requires that civil courts accept their decisions as binding upon them.” Id., at 725. We thus held that by inquiring into whether the Church had followed its own procedures, the State Supreme Court had “unconstitutionally undertaken the resolution of quintessentially religious controversies whose resolution the First Amendment commits exclusively to the highest ecclesiastical tribunals” of the Church. Id., at 720. C Until today, we have not had occasion to consider whether this freedom of a religious organization to select its ministers is implicated by a suit alleging discrimination in employment. The Courts of Appeals, in contrast, have had extensive experience with this issue. Since the passage of Title VII of the Civil Rights Act of 1964, 42 U. S. C. §2000e et seq., and other employment discrimination laws, the Courts of Appeals have uniformly recognized the existence of a “ministerial exception,” grounded in the First Amendment, that precludes application of such legislation to claims concerning the employment relationship between a religious institution and its ministers. [ 2 ] We agree that there is such a ministerial exception. The members of a religious group put their faith in the hands of their ministers. Requiring a church to accept or retain an unwanted minister, or punishing a church for failing to do so, intrudes upon more than a mere employment decision. Such action interferes with the internal governance of the church, depriving the church of control over the selection of those who will personify its beliefs. By imposing an unwanted minister, the state infringes the Free Exercise Clause, which protects a religious group’s right to shape its own faith and mission through its appointments. According the state the power to determine which individuals will minister to the faithful also violates the Establishment Clause, which prohibits government involvement in such ecclesiastical decisions. The EEOC and Perich acknowledge that employment discrimination laws would be unconstitutional as applied to religious groups in certain circumstances. They grant, for example, that it would violate the First Amendment for courts to apply such laws to compel the ordination of women by the Catholic Church or by an Orthodox Jewish seminary. Brief for Federal Respondent 31; Brief for Respondent Perich 35–36. According to the EEOC and Perich, religious organizations could successfully defend against employment discrimination claims in those circum-stances by invoking the constitutional right to freedom of association—a right “implicit” in the First Amendment. Roberts v. United States Jaycees, 468 U.S. 609, 622 (1984) . The EEOC and Perich thus see no need—and no basis—for a special rule for ministers grounded in the Religion Clauses themselves. We find this position untenable. The right to freedom of association is a right enjoyed by religious and secular groups alike. It follows under the EEOC’s and Perich’s view that the First Amendment analysis should be the same, whether the association in question is the Lutheran Church, a labor union, or a social club. See Perich Brief 31; Tr. of Oral Arg. 28. That result is hard to square with the text of the First Amendment itself, which gives spe-cial solicitude to the rights of religious organizations. We cannot accept the remarkable view that the Religion Clauses have nothing to say about a religious organization’s freedom to select its own ministers. The EEOC and Perich also contend that our decision in Employment Div., Dept. of Human Resources of Ore. v. Smith, 494 U.S. 872 (1990) , precludes recognition of a ministerial exception. In Smith, two members of the Native American Church were denied state unemployment benefits after it was determined that they had been fired from their jobs for ingesting peyote, a crime under Oregon law. We held that this did not violate the Free Exercise Clause, even though the peyote had been ingested for sacramental purposes, because the “right of free exercise does not relieve an individual of the obligation to comply with a valid and neutral law of general applicability on the ground that the law proscribes (or prescribes) conduct that his religion prescribes (or proscribes).” Id., at 879 (internal quotation marks omitted). It is true that the ADA’s prohibition on retaliation, like Oregon’s prohibition on peyote use, is a valid and neutral law of general applicability. But a church’s selection of its ministers is unlike an individual’s ingestion of peyote. Smith involved government regulation of only outward physical acts. The present case, in contrast, concerns government interference with an internal church decision that affects the faith and mission of the church itself. See id., at 877 (distinguishing the government’s regulation of “physical acts” from its “lend[ing] its power to one or the other side in controversies over religious authority or dogma”). The contention that Smith forecloses recognition of a ministerial exception rooted in the Religion Clauses has no merit. III Having concluded that there is a ministerial exception grounded in the Religion Clauses of the First Amendment, we consider whether the exception applies in this case. We hold that it does. Every Court of Appeals to have considered the question has concluded that the ministerial exception is not limited to the head of a religious congregation, and we agree. We are reluctant, however, to adopt a rigid formula for deciding when an employee qualifies as a minister. It is enough for us to conclude, in this our first case involving the ministerial exception, that the exception covers Perich, given all the circumstances of her employment. To begin with, Hosanna-Tabor held Perich out as a minister, with a role distinct from that of most of its members. When Hosanna-Tabor extended her a call, it issued her a “diploma of vocation” according her the title “Minister of Religion, Commissioned.” App. 42. She was tasked with performing that office “according to the Word of God and the confessional standards of the Evangelical Lutheran Church as drawn from the Sacred Scriptures.” Ibid. The congregation prayed that God “bless [her] ministrations to the glory of His holy name, [and] the building of His church.” Id., at 43. In a supplement to the diploma, the congregation undertook to periodically review Perich’s “skills of ministry” and “ministerial responsibilities,” and to provide for her “continuing education as a professional person in the ministry of the Gospel.” Id., at 49. Perich’s title as a minister reflected a significant degree of religious training followed by a formal process of commissioning. To be eligible to become a commissioned minister, Perich had to complete eight college-level courses in subjects including biblical interpretation, church doctrine, and the ministry of the Lutheran teacher. She also had to obtain the endorsement of her local Synod district by submitting a petition that contained her academic transcripts, letters of recommendation, personal statement, and written answers to various ministry-related questions. Finally, she had to pass an oral examination by a faculty committee at a Lutheran college. It took Perich six years to fulfill these requirements. And when she eventually did, she was commissioned as a minister only upon election by the congregation, which recognized God’s call to her to teach. At that point, her call could be rescinded only upon a supermajority vote of the congregation—a protection designed to allow her to “preach the Word of God boldly.” Brief for Lutheran Church-Missouri Synod as Amicus Curiae 15. Perich held herself out as a minister of the Church by accepting the formal call to religious service, according to its terms. She did so in other ways as well. For example, she claimed a special housing allowance on her taxes that was available only to employees earning their compensation “ ‘in the exercise of the ministry.’ ” App. 220 (“If you are not conducting activities ‘in the exercise of the ministry,’ you cannot take advantage of the parsonage or housing allowance exclusion” (quoting Lutheran Church-Missouri Synod Brochure on Whether the IRS Considers Employees as a Minister (2007)). In a form she submitted to the Synod following her termination, Perich again indicated that she regarded herself as a minister at Hosanna-Tabor, stating: “I feel that God is leading me to serve in the teaching ministry . . . . I am anxious to be in the teaching ministry again soon.” App. 53. Perich’s job duties reflected a role in conveying the Church’s message and carrying out its mission. Hosanna-Tabor expressly charged her with “lead[ing] others toward Christian maturity” and “teach[ing] faithfully the Word of God, the Sacred Scriptures, in its truth and purity and as set forth in all the symbolical books of the Evangelical Lutheran Church.” Id., at 48. In fulfilling these responsibilities, Perich taught her students religion four days a week, and led them in prayer three times a day. Once a week, she took her students to a school-wide chapel service, and—about twice a year—she took her turn leading it, choosing the liturgy, selecting the hymns, and delivering a short message based on verses from the Bible. During her last year of teaching, Perich also led her fourth graders in a brief devotional exercise each morning. As a source of religious instruction, Perich performed an important role in transmitting the Lutheran faith to the next generation. In light of these considerations—the formal title given Perich by the Church, the substance reflected in that title, her own use of that title, and the important religious functions she performed for the Church—we conclude that Perich was a minister covered by the ministerial exception. In reaching a contrary conclusion, the Court of Appeals committed three errors. First, the Sixth Circuit failed to see any relevance in the fact that Perich was a commissioned minister. Although such a title, by itself, does not automatically ensure coverage, the fact that an employee has been ordained or commissioned as a minister is surely relevant, as is the fact that significant religious training and a recognized religious mission underlie the description of the employee’s position. It was wrong for the Court of Appeals—and Perich, who has adopted the court’s view, see Perich Brief 45—to say that an employee’s title does not matter. Second, the Sixth Circuit gave too much weight to the fact that lay teachers at the school performed the same religious duties as Perich. We express no view on whether someone with Perich’s duties would be covered by the ministerial exception in the absence of the other considerations we have discussed. But though relevant, it cannot be dispositive that others not formally recognized as ministers by the church perform the same functions—particularly when, as here, they did so only because commissioned ministers were unavailable. Third, the Sixth Circuit placed too much emphasis on Perich’s performance of secular duties. It is true that her religious duties consumed only 45 minutes of each workday, and that the rest of her day was devoted to teaching secular subjects. The EEOC regards that as conclusive, contending that any ministerial exception “should be limited to those employees who perform exclusively religious functions.” Brief for Federal Respondent 51. We cannot accept that view. Indeed, we are unsure whether any such employees exist. The heads of congregations themselves often have a mix of duties, including secular ones such as helping to manage the congregation’s finances, supervising purely secular personnel, and overseeing the upkeep of facilities. Although the Sixth Circuit did not adopt the extreme position pressed here by the EEOC, it did regard the relative amount of time Perich spent performing religious functions as largely determinative. The issue before us, however, is not one that can be resolved by a stopwatch. The amount of time an employee spends on particular activities is relevant in assessing that employee’s status, but that factor cannot be considered in isolation, without regard to the nature of the religious functions performed and the other considerations discussed above. Because Perich was a minister within the meaning of the exception, the First Amendment requires dismissal of this employment discrimination suit against her religious employer. The EEOC and Perich originally sought an order reinstating Perich to her former position as a called teacher. By requiring the Church to accept a minister it did not want, such an order would have plainly violated the Church’s freedom under the Religion Clauses to select its own ministers. Perich no longer seeks reinstatement, having abandoned that relief before this Court. See Perich Brief 58. But that is immaterial. Perich continues to seek frontpay in lieu of reinstatement, backpay, compensatory and punitive damages, and attorney’s fees. An award of such relief would operate as a penalty on the Church for terminating an unwanted minister, and would be no less prohibited by the First Amendment than an order overturning the termination. Such relief would depend on a determination that Hosanna-Tabor was wrong to have relieved Perich of her position, and it is precisely such a ruling that is barred by the ministerial exception. [ 3 ] The EEOC and Perich suggest that Hosanna-Tabor’s asserted religious reason for firing Perich—that she violated the Synod’s commitment to internal dispute resolution—was pretextual. That suggestion misses the point of the ministerial exception. The purpose of the exception is not to safeguard a church’s decision to fire a minister only when it is made for a religious reason. The exception instead ensures that the authority to select and control who will minister to the faithful—a matter “strictly ecclesiastical,” Kedroff, 344 U. S., at 119—is the church’s alone. [ 4 ] IV The EEOC and Perich foresee a parade of horribles that will follow our recognition of a ministerial exception to employment discrimination suits. According to the EEOC and Perich, such an exception could protect religious organizations from liability for retaliating against employees for reporting criminal misconduct or for testifying before a grand jury or in a criminal trial. What is more, the EEOC contends, the logic of the exception would confer on religious employers “unfettered discretion” to violate employment laws by, for example, hiring children or aliens not authorized to work in the United States. Brief for Federal Respondent 29. Hosanna-Tabor responds that the ministerial exception would not in any way bar criminal prosecutions for in-terfering with law enforcement investigations or other proceedings. Nor, according to the Church, would the exception bar government enforcement of general laws restricting eligibility for employment, because the exception applies only to suits by or on behalf of ministers themselves. Hosanna-Tabor also notes that the ministe-rial exception has been around in the lower courts for 40 years, see McClure v. Salvation Army, 460 F.2d 553, 558 (CA5 1972), and has not given rise to the dire consequences predicted by the EEOC and Perich. The case before us is an employment discrimination suit brought on behalf of a minister, challenging her church’s decision to fire her. Today we hold only that the ministerial exception bars such a suit. We express no view on whether the exception bars other types of suits, including actions by employees alleging breach of contract or tortious conduct by their religious employers. There will be time enough to address the applicability of the exception to other circumstances if and when they arise. * * * The interest of society in the enforcement of employment discrimination statutes is undoubtedly important. But so too is the interest of religious groups in choosing who will preach their beliefs, teach their faith, and carry out their mission. When a minister who has been fired sues her church alleging that her termination was discriminatory, the First Amendment has struck the balance for us. The church must be free to choose those who will guide it on its way. The judgment of the Court of Appeals for the Sixth Circuit is reversed. It is so ordered. Notes 1 The ADA itself provides religious entities with two defenses to claims of discrimination that arise under subchapter I of the Act. The first provides that “[t]his subchapter shall not prohibit a religious corporation, association, educational institution, or society from giving preference in employment to individuals of a particular religion to perform work connected with the carrying on by such [entity] of its activities.” §12113(d)(1) (2006 ed., Supp. III). The second provides that “[u]nder this subchapter, a religious organization may require that all applicants and employees conform to the religious tenets of such organization.” §12113(d)(2). The ADA’s prohibition against retaliation, §12203(a), appears in a different subchapter—subchapter IV. The EEOC and Perich contend, and Hosanna-Tabor does not dispute, that these defenses therefore do not apply to retaliation claims. 2 See Natal v. Christian and Missionary Alliance, 878 F.2d 1575, 1578 (CA1 1989); Rweyemamu v. Cote, 520 F.3d 198, 204–209 (CA2 2008); Petruska v. Gannon Univ., 462 F.3d 294, 303–307 (CA3 2006); EEOC v. Roman Catholic Diocese, 213 F.3d 795, 800–801 (CA4 2000); Combs v. Central Tex. Annual Conference, 173 F.3d 343, 345–350 (CA5 1999); Hollins v. Methodist Healthcare, Inc., 474 F.3d 223, 225–227 (CA6 2007); Schleicher v. Salvation Army, 518 F.3d 472, 475 (CA7 2008); Scharon v. St. Luke’s Episcopal Presbyterian Hospitals, 929 F.2d 360, 362–363 (CA8 1991); Werft v. Desert Southwest Annual Conference, 377 F.3d 1099, 1100–1104 (CA9 2004); Bryce v. Episcopal Church, 289 F.3d 648, 655–657 (CA10 2002); Gellington v. Christian Methodist Episcopal Church, Inc., 203 F.3d 1299, 1301–1304 (CA11 2000); EEOC v. Catholic Univ., 83 F.3d 455, 460–463 (CADC 1996). 3 Perich does not dispute that if the ministerial exception bars her retaliation claim under the ADA, it also bars her retaliation claim under Michigan law. 4 A conflict has arisen in the Courts of Appeals over whether the ministerial exception is a jurisdictional bar or a defense on the merits. Compare Hollins, 474 F. 3d, at 225 (treating the exception as jurisdictional); and Tomic v. Catholic Diocese of Peoria, 442 F.3d 1036, 1038–1039 (CA7 2006) (same), with Petruska, 462 F. 3d, at 302 (treating the exception as an affirmative defense); Bryce, 289 F. 3d, at 654 (same); Bollard v. California Province of Soc. of Jesus, 196 F.3d 940, 951 (CA9 1999) (same); and Natal, 878 F. 2d, at 1576 (same). We conclude that the exception operates as an affirmative defense to an otherwise cognizable claim, not a jurisdictional bar. That is because the issue presented by the exception is “whether the allegations the plaintiff makes entitle him to relief,” not whether the court has “power to hear [the] case.” Morrison v. National Australia Bank Ltd., 561 U. S. ___, ___ (2010) (slip op., at 4–5) (internal quotation marks omitted). District courts have power to consider ADA claims in cases of this sort, and to decide whether the claim can proceed or is instead barred by the ministerial exception.

565.US.2011_10-680

Respondent Fields, a Michigan state prisoner, was escorted from his prison cell by a corrections officer to a conference room where he was questioned by two sheriff’s deputies about criminal activity he had allegedly engaged in before coming to prison. At no time was Fields given Miranda warnings or advised that he did not have to speak with the deputies. As relevant here: Fields was questioned for between five and seven hours; Fields was told more than once that he was free to leave and return to his cell; the deputies were armed, but Fields remained free of restraints; the conference room door was sometimes open and sometimes shut; several times during the interview Fields stated that he no longer wanted to talk to the deputies, but he did not ask to go back to his cell; after Fields confessed and the interview concluded, he had to wait an additional 20 minutes for an escort and returned to his cell well after the hour when he generally retired. The trial court denied Fields’ motion to suppress his confession under Miranda v. Arizona, 384 U.S. 436, and he was convicted. The Michigan Court of Appeals affirmed, rejecting Fields’ contention that his statements should have been suppressed because he was subjected to custodial interrogation without a Miranda warning. The United States District Court for the Eastern District of Michigan subsequently granted Fields habeas relief under 28 U. S. C. §2254(d)(1). Affirming, the Sixth Circuit held that the interview was a custodial interrogation within the meaning of Miranda, reasoning that Mathis v. United States, 391 U.S. 1, “clearly established,” §2254(d)(1), that isolation from the general prison population, combined with questioning about conduct occurring outside the prison, makes any such interrogation custodial per se. Held: 1. This Court’s precedents do not clearly establish the categorical rule on which the Sixth Circuit relied. The Court has repeatedly declined to adopt any such rule. See, e.g., Illinois v. Perkins, 496 U.S. 292. The Sixth Circuit misread Mathis, which simply held, as relevant here, that a prisoner who otherwise meets the requirements for Miranda custody is not taken outside the scope of Miranda because he was incarcerated for an unconnected offense. It did not hold that imprisonment alone constitutes Miranda custody. Nor does the statement in Maryland v. Shatzer, 559 U. S. ___, ___, that “[n]o one questions that [inmate] Shatzer was in custody for Miranda purposes” support a per se rule. It means only that the issue of custody was not contested in that case. Finally, contrary to respondent’s suggestion, Miranda itself did not hold that the inherently compelling pressures of custodial interrogation are always present when a prisoner is taken aside and questioned about events outside the prison walls. Pp. 4–7. 2. The Sixth Circuit’s categorical rule—that imprisonment, questioning in private, and questioning about events in the outside world create a custodial situation for Miranda purposes—is simply wrong. Pp. 8–13. (a) The initial step in determining whether a person is in Miranda custody is to ascertain, given “all of the circumstances surrounding the interrogation,” how a suspect would have gauged his freedom of movement. Stansbury v. California, 511 U.S. 318, 322, 325. However, not all restraints on freedom of movement amount to Miranda custody. See, e.g., Berkemer v. McCarty, 468 U.S. 420, 423. Shatzer, distinguishing between restraints on freedom of movement and Miranda custody, held that a break in Miranda custody between a suspect’s invocation of the right to counsel and the initiation of subsequent questioning may occur while a suspect is serving an uninterrupted term of imprisonment. If a break in custody can occur, it must follow that imprisonment alone is not enough to create a custodial situation within the meaning of Miranda. At least three strong grounds support this conclusion: Questioning a person who is already in prison does not generally involve the shock that very often accompanies arrest; a prisoner is unlikely to be lured into speaking by a longing for prompt release; and a prisoner knows that his questioners probably lack authority to affect the duration of his sentence. Thus, service of a prison term, without more, is not enough to constitute Miranda custody. Pp. 8–12. (b) The other two elements in the Sixth Circuit’s rule are likewise insufficient. Taking a prisoner aside for questioning may necessitate some additional limitations on the prisoner’s freedom of movement, but it does not necessarily convert a noncustodial situation into Miranda custody. Isolation may contribute to a coercive atmosphere when a nonprisoner is questioned, but questioning a prisoner in private does not generally remove him from a supportive atmosphere and may be in his best interest. Neither does questioning a prisoner about criminal activity outside the prison have a significantly greater potential for coercion than questioning under otherwise identical circumstances about criminal activity within the prison walls. The coercive pressure that Miranda guards against is neither mitigated nor magnified by the location of the conduct about which questions are asked. Pp. 12–13. 3. When a prisoner is questioned, the determination of custody should focus on all of the features of the interrogation. The record in this case reveals that respondent was not taken into custody for Miranda purposes. While some of the facts lend support to his argument that Miranda’s custody requirement was met, they are offset by others. Most important, he was told at the outset of the interrogation, and reminded thereafter, that he was free to leave and could go back to his cell whenever he wanted. Moreover, he was not physically restrained or threatened, was interviewed in a well-lit, average-sized conference room where the door was sometimes left open, and was offered food and water. These facts are consistent with an environment in which a reasonable person would have felt free to terminate the interview and leave, subject to the ordinary restraints of life behind bars. Pp. 13–16. 617 F.3d 813, reversed. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Thomas, and Kagan, JJ., joined. Ginsburg, J., filed an opinion concurring in part and dissenting in part, in which Breyer and Sotomayor, JJ., joined.

The United States Court of Appeals for the Sixth Circuit held that our precedents clearly establish that a prisoner is in custody within the meaning of Miranda v. Arizona, 384 U.S. 436 (1966), if the prisoner is taken aside and questioned about events that occurred outside the prison walls. Our decisions, however, do not clearly establish such a rule, and therefore the Court of Appeals erred in holding that this rule provides a permissible basis for federal habeas relief under the relevant provision of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), 28 U. S. C. §2254(d)(1). Indeed, the rule applied by the court below does not represent a correct interpretation of our Miranda case law. We therefore reverse. I While serving a sentence in a Michigan jail, Randall Fields was escorted by a corrections officer to a conference room where two sheriff’s deputies questioned him about allegations that, before he came to prison, he had engaged in sexual conduct with a 12-year-old boy. In order to get to the conference room, Fields had to go down one floor and pass through a locked door that separated two sections of the facility. See App. to Pet. for Cert. 66a, 69a. Fields arrived at the conference room between 7 p.m. and 9 p.m.[1] and was questioned for between five and seven hours.[2] At the beginning of the interview, Fields was told that he was free to leave and return to his cell. See id., at 70a. Later, he was again told that he could leave whenever he wanted. See id., at 90a. The two interviewing deputies were armed during the interview, but Fields remained free of handcuffs and other restraints. The door to the conference room was sometimes open and sometimes shut. See id., at 70a–75a. About halfway through the interview, after Fields had been confronted with the allegations of abuse, he became agitated and began to yell. See id., at 80a, 125a. Fields testified that one of the deputies, using an expletive, told him to sit down and said that “if [he] didn’t want to cooperate, [he] could leave.” Id., at 89a; see also id., at 70a–71a. Fields eventually confessed to engaging in sex acts with the boy. According to Fields’ testimony at a suppression hearing, he said several times during the interview that he no longer wanted to talk to the deputies, but he did not ask to go back to his cell prior to the end of the interview. See id., at 92a–93a. When he was eventually ready to leave, he had to wait an additional 20 minutes or so because a corrections officer had to be summoned to escort him back to his cell, and he did not return to his cell until well after the hour when he generally retired.[3] At no time was Fields given Miranda warnings or advised that he did not have to speak with the deputies. The State of Michigan charged Fields with criminal sexual conduct. Relying on Miranda, Fields moved to suppress his confession, but the trial court denied his motion. Over the renewed objection of defense counsel, one of the interviewing deputies testified at trial about Fields’ admissions. The jury convicted Fields of two counts of third-degree criminal sexual conduct, and the judge sentenced him to a term of 10 to 15 years of imprisonment. On direct appeal, the Michigan Court of Appeals affirmed, rejecting Fields’ contention that his statements should have been suppressed because he was subjected to custodial interrogation without a Miranda warning. The court ruled that Fields had not been in custody for pur- poses of Miranda during the interview, so no Miranda warnings were required. The court emphasized that Fields was told that he was free to leave and return to his cell but that he never asked to do so. The Michigan Supreme Court denied discretionary review. Fields then filed a petition for a writ of habeas corpus in Federal District Court, and the court granted relief. The Sixth Circuit affirmed, holding that the interview in the conference room was a “custodial interrogation” within the meaning of Miranda because isolation from the general prison population combined with questioning about conduct occurring outside the prison makes any such interrogation custodial per se. The Court of Appeals reasoned that this Court clearly established in Mathis v. United States, 391 U.S. 1 (1968), that “Miranda warnings must be administered when law enforcement officers remove an inmate from the general prison population and interrogate him regarding criminal conduct that took place outside the jail or prison.” 617 F.3d 813, 820 (CA6 2010); see also id., at 818 (“The central holding of Mathis is that a Miranda warning is required whenever an incarcerated individual is isolated from the general prison population and interrogated, i.e.[,] questioned in a manner likely to lead to self-incrimination, about conduct occurring outside of the prison”). Because Fields was isolated from the general prison population and interrogated about conduct occurring in the outside world, the Court of Appeals found that the state court’s decision was contrary to clearly established federal law as determined by this Court in Mathis. 617 F. 3d, at 823. We granted certiorari. 562 U. S. ___ (2011). II Under AEDPA, a federal court may grant a state prisoner’s application for a writ of habeas corpus if the state-court adjudication pursuant to which the prisoner is held “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.” 28 U. S. C. §2254(d)(1). In this context, “clearly established law” signifies “the holdings, as opposed to the dicta, of this Court’s decisions.” Williams v. Taylor, 529 U.S. 362, 412 (2000). In this case, it is abundantly clear that our precedents do not clearly establish the categorical rule on which the Court of Appeals relied, i.e., that the questioning of a prisoner is always custodial when the prisoner is removed from the general prison population and questioned about events that occurred outside the prison. On the contrary, we have repeatedly declined to adopt any categorical rule with respect to whether the questioning of a prison inmate is custodial. In Illinois v. Perkins, 496 U.S. 292 (1990), where we upheld the admission of un-Mirandized statements elicited from an inmate by an undercover officer masquerading as another inmate, we noted that “[t]he bare fact of cus- tody may not in every instance require a warning even when the suspect is aware that he is speaking to an official, but we do not have occasion to explore that issue here.” Id., at 299 (emphasis added). Instead, we simply “reject[ed] the argument that Miranda warnings are required whenever a suspect is in custody in a technical sense and converses with someone who happens to be a government agent.” Id., at 297. Most recently, in Maryland v. Shatzer, 559 U. S. ___ (2010), we expressly declined to adopt a bright-line rule for determining the applicability of Miranda in prisons. Shatzer considered whether a break in custody ends the presumption of involuntariness established in Edwards v. Arizona, 451 U.S. 477 (1981), and, if so, whether a prisoner’s return to the general prison population after a custodial interrogation constitutes a break in Miranda custody. See 559 U. S., at ___ (slip op., at 3–4). In considering the latter question, we noted first that “[w]e have never decided whether incarceration constitutes custody for Miranda purposes, and have indeed explicitly declined to address the issue.” Id., at ___ (slip op., at 13) (citing Perkins, supra, at 299; emphasis added). The answer to this question, we noted, would “depen[d] upon whether [incar- ceration] exerts the coercive pressure that Miranda was designed to guard against—the ‘danger of coercion [that] results from the interaction of custody and official inter- rogation.’ ” 559 U. S., at ___ (slip op., at 13) (quoting Perkins, supra, at 297). In concluding that our precedents establish a categorical rule, the Court of Appeals placed great weight on the decision in Mathis, but the Court of Appeals misread the holding in that case. In Mathis, an inmate in a state prison was questioned by an Internal Revenue agent and was subsequently convicted for federal offenses. The Court of Appeals held that Miranda did not apply to this interview for two reasons: A criminal investigation had not been commenced at the time of the interview, and the prisoner was incarcerated for an “unconnected offense.” Mathis v. United States, 376 F.2d 595, 597 (CA5 1967). This Court rejected both of those grounds for distinguishing Miranda, 391 U. S., at 4, and thus the holding in Mathis is simply that a prisoner who otherwise meets the requirements for Miranda custody is not taken outside the scope of Miranda by either of the two factors on which the Court of Appeals had relied. Mathis did not hold that imprisonment, in and of itself, is enough to constitute Miranda custody.[4] Nor, contrary to respondent’s submission, see Brief for Respondent 14, did Oregon v. Mathiason, 429 U.S. 492, 494 (1977) (per curiam), which simply restated in dictum the holding in Mathis. The Court of Appeals purported to find support for its per se rule in Shatzer, relying on our statement that “[n]o one questions that Shatzer was in custody for Miranda purposes” when he was interviewed. 559 U. S., at ___ (slip op., at 13). But this statement means only that the issue of custody was not contested before us. It strains credulity to read the statement as constituting an “unambiguous conclusion” or “finding” by this Court that Shatzer was in custody. 617 F. 3d, at 822. Finally, contrary to respondent’s suggestion, see Brief for Respondent 12–15, Miranda itself did not clearly establish the rule applied by the Court of Appeals. Miranda adopted a “set of prophylactic measures” designed to ward off the “ ‘inherently compelling pressures’ of custodial interrogation,” Shatzer, supra, at ___ (slip op., at 4) (quoting Miranda, 384 U. S., at 467), but Miranda did not hold that such pressures are always present when a prisoner is taken aside and questioned about events outside the prison walls. Indeed, Miranda did not even establish that police questioning of a suspect at the station house is always custodial. See Mathiason, supra, at 495 (declining to find that Miranda warnings are required “simply because the questioning takes place in the station house, or because the questioned person is one whom the police suspect”). In sum, our decisions do not clearly establish that a prisoner is always in custody for purposes of Miranda whenever a prisoner is isolated from the general prison population and questioned about conduct outside the prison.[5] III Not only does the categorical rule applied below go well beyond anything that is clearly established in our prior decisions, it is simply wrong. The three elements of that rule—(1) imprisonment, (2) questioning in private, and (3) questioning about events in the outside world—are not necessarily enough to create a custodial situation for Miranda purposes. A As used in our Miranda case law, “custody” is a term of art that specifies circumstances that are thought generally to present a serious danger of coercion. In determining whether a person is in custody in this sense, the initial step is to ascertain whether, in light of “the objective cir- cumstances of the interrogation,” Stansbury v. Califor- nia, 511 U.S. 318, 322–323, 325 (1994) (per curiam), a “reasonable person [would] have felt he or she was not at liberty to terminate the interrogation and leave.” Thompson v. Keohane, 516 U.S. 99, 112 (1995). And in order to determine how a suspect would have “gauge[d]” his “freedom of movement,” courts must examine “all of the circumstances surrounding the interrogation.” Stansbury, supra, at 322, 325 (internal quotation marks omitted). Relevant factors include the location of the questioning, see Shatzer, supra, at ___–___ (slip op., at 13–16), its duration, see Berkemer v. McCarty, 468 U.S. 420, 437–438 (1984), statements made during the interview, see Mathiason, supra, at 495; Yarborough v. Alvarado, 541 U.S. 652, 665 (2004); Stansbury, supra, at 325, the presence or absence of physical restraints during the questioning, see New York v. Quarles, 467 U.S. 649, 655 (1984), and the release of the interviewee at the end of the questioning, see California v. Beheler, 463 U.S. 1121, 1122–1123 (1983) (per curiam). Determining whether an individual’s freedom of movement was curtailed, however, is simply the first step in the analysis, not the last. Not all restraints on freedom of movement amount to custody for purposes of Miranda. We have “decline[d] to accord talismanic power” to the freedom-of-movement inquiry, Berkemer, supra, at 437, and have instead asked the additional question whether the relevant environment presents the same inherently coercive pressures as the type of station house questioning at issue in Miranda. “Our cases make clear . . . that the freedom-of-movement test identifies only a necessary and not a sufficient condition for Miranda custody.” Shatzer, 559 U. S., at ___ (slip op., at 14). This important point is illustrated by our decision in Berkemer v. McCarty, supra. In that case, we held that the roadside questioning of a motorist who was pulled over in a routine traffic stop did not constitute custodial interrogation. Id., at 423, 441–442. We acknowledged that “a traffic stop significantly curtails the ‘freedom of action’ of the driver and the passengers,” and that it is generally “a crime either to ignore a policeman’s signal to stop one’s car or, once having stopped, to drive away without permission.” Id., at 436. “[F]ew motorists,” we noted, “would feel free either to disobey a directive to pull over or to leave the scene of a traffic stop without being told they might do so.” Ibid. Nevertheless, we held that a person detained as a result of a traffic stop is not in Miranda custody because such detention does not “sufficiently impair [the detained person’s] free exercise of his privilege against self-incrimination to require that he be warned of his consti- tutional rights.” 468 U. S., at 437. As we later put it, the “temporary and relatively nonthreatening detention in- volved in a traffic stop or Terry stop does not constitute Miranda custody,” Shatzer, supra, at ___ (slip op., at 14) (citation omitted). See Terry v. Ohio, 392 U.S. 1 (1968). It may be thought that the situation in Berkemer—the questioning of a motorist subjected to a brief traffic stop—is worlds away from those present when an inmate is questioned in a prison, but the same cannot be said of Shatzer, where we again distinguished between restraints on freedom of movement and Miranda custody. Shatzer, as noted, concerned the Edwards prophylactic rule, which limits the ability of the police to initiate further questioning of a suspect in Miranda custody once the suspect invokes the right to counsel. We held in Shatzer that this rule does not apply when there is a sufficient break in custody between the suspect’s invocation of the right to counsel and the initiation of subsequent questioning. See 559 U. S., at ___ (slip op., at 13-16). And, what is significant for present purposes, we further held that a break in custody may occur while a suspect is serving a term in prison. If a break in custody can occur while a prisoner is serving an uninterrupted term of imprisonment, it must follow that imprisonment alone is not enough to create a custodial situation within the meaning of Miranda. There are at least three strong grounds for this conclusion. First, questioning a person who is already serving a prison term does not generally involve the shock that very often accompanies arrest. In the paradigmatic Miranda situation—a person is arrested in his home or on the street and whisked to a police station for questioning—detention represents a sharp and ominous change, and the shock may give rise to coercive pressures. A person who is “cut off from his normal life and companions,” Shatzer, supra, at ___ (slip op., at 7), and abruptly transported from the street into a “police-dominated atmosphere,” Miranda, 384 U. S., at 456, may feel coerced into answering questions. By contrast, when a person who is already serving a term of imprisonment is questioned, there is usually no such change. “Interrogated suspects who have previously been convicted of crime live in prison.” Shatzer, 559 U. S., at ___ (slip op., at 14). For a person serving a term of incarceration, we reasoned in Shatzer, the ordinary restrictions of prison life, while no doubt unpleasant, are expected and familiar and thus do not involve the same “inherently compelling pressures” that are often present when a suspect is yanked from familiar surroundings in the outside world and subjected to interrogation in a police station. Id., at ___ (slip op., at 4). Second, a prisoner, unlike a person who has not been sentenced to a term of incarceration, is unlikely to be lured into speaking by a longing for prompt release. When a person is arrested and taken to a station house for interrogation, the person who is questioned may be pressured to speak by the hope that, after doing so, he will be allowed to leave and go home. On the other hand, when a prisoner is questioned, he knows that when the questioning ceases, he will remain under confinement. Id., at ___– ___, n. 8 (slip op., at 14–15, n. 8). Third, a prisoner, unlike a person who has not been convicted and sentenced, knows that the law enforcement officers who question him probably lack the authority to affect the duration of his sentence. Id., at ___–___ (slip op., at 14–15). And “where the possibility of parole exists,” the interrogating officers probably also lack the power to bring about an early release. Ibid. “When the suspect has no reason to think that the listeners have official power over him, it should not be assumed that his words are motivated by the reaction he expects from his listeners.” Perkins, 496 U. S., at 297. Under such circumstances, there is little “basis for the assumption that a suspect . . . will feel compelled to speak by the fear of reprisal for remaining silent or in the hope of [a] more lenient treatment should he confess.” Id., at 296–297. In short, standard conditions of confinement and associated restrictions on freedom will not necessarily implicate the same interests that the Court sought to protect when it afforded special safeguards to persons subjected to custodial interrogation. Thus, service of a term of imprisonment, without more, is not enough to constitute Miranda custody. B The two other elements included in the Court of Appeals’ rule—questioning in private and questioning about events that took place outside the prison—are likewise insufficient. Taking a prisoner aside for questioning—as opposed to questioning the prisoner in the presence of fellow inmates—does not necessarily convert a “noncustodial situation . . . to one in which Miranda applies.” Mathiason, 429 U. S., at 495. When a person who is not serving a prison term is questioned, isolation may contribute to a coercive atmosphere by preventing family members, friends, and others who may be sympathetic from providing either advice or emotional support. And without any such assistance, the person who is questioned may feel overwhelming pressure to speak and to refrain from asking that the interview be terminated. By contrast, questioning a prisoner in private does not generally remove the prisoner from a supportive atmosphere. Fellow inmates are by no means necessarily friends. On the contrary, they may be hostile and, for a variety of reasons, may react negatively to what the questioning reveals. In the present case, for example, would respondent have felt more at ease if he had been questioned in the presence of other inmates about the sexual abuse of an adolescent boy? Isolation from the general prison population is often in the best interest of the interviewee and, in any event, does not suggest on its own the atmosphere of coercion that concerned the Court in Miranda. It is true that taking a prisoner aside for questioning may necessitate some additional limitations on his freedom of movement. A prisoner may, for example, be removed from an exercise yard and taken, under close guard, to the room where the interview is to be held. But such procedures are an ordinary and familiar attribute of life behind bars. Escorts and special security precautions may be standard procedures regardless of the purpose for which an inmate is removed from his regular routine and taken to a special location. For example, ordinary prison procedure may require such measures when a prisoner is led to a meeting with an attorney. Finally, we fail to see why questioning about criminal activity outside the prison should be regarded as having a significantly greater potential for coercion than questioning under otherwise identical circumstances about criminal activity within the prison walls. In both instances, there is the potential for additional criminal liability and punishment. If anything, the distinction would seem to cut the other way, as an inmate who confesses to misconduct that occurred within the prison may also incur administrative penalties, but even this is not enough to tip the scale in the direction of custody. “The threat to a citizen’s Fifth Amendment rights that Miranda was designed to neutralize” is neither mitigated nor magnified by the location of the conduct about which questions are asked. Berkemer, 468 U. S., at 435, n. 22. For these reasons, the Court of Appeals’ categorical rule is unsound. IV A When a prisoner is questioned, the determination of custody should focus on all of the features of the interrogation. These include the language that is used in summoning the prisoner to the interview and the manner in which the interrogation is conducted. See Yarborough, 541 U. S., at 665. An inmate who is removed from the general prison population for questioning and is “thereafter . . . subjected to treatment” in connection with the interrogation “that renders him ‘in custody’ for practical purposes . . . will be entitled to the full panoply of protections prescribed by Miranda.” Berkemer, 468 U. S., at 440. “Fidelity to the doctrine announced in Miranda requires that it be enforced strictly, but only in those types of situations in which the concerns that powered the decision are implicated.” Id., at 437; see Shatzer, 559 U. S., at ___ (slip op., at 9); Mathiason, supra, at 495. Confessions voluntarily made by prisoners in other situations should not be suppressed. “Voluntary confessions are not merely a proper element in law enforcement, they are an unmitigated good, essential to society’s compelling interest in finding, convicting, and punishing those who violate the law.” Shatzer, supra, at ___ (slip op., at 9) (internal quotation marks and citations omitted). B The record in this case reveals that respondent was not taken into custody for purposes of Miranda. To be sure, respondent did not invite the interview or consent to it in advance, and he was not advised that he was free to decline to speak with the deputies. The following facts also lend some support to respondent’s argument that Miranda’s custody requirement was met: The interview lasted for between five and seven hours in the evening and continued well past the hour when respondent generally went to bed; the deputies who questioned respondent were armed; and one of the deputies, according to respondent, “[u]sed a very sharp tone,” App. to Pet. for Cert. 76a, and, on one occasion, profanity, see id., at 77a. These circumstances, however, were offset by others. Most important, respondent was told at the outset of the interrogation, and was reminded again thereafter, that he could leave and go back to his cell whenever he wanted. See id., at 89a–90a (“I was told I could get up and leave whenever I wanted”); id., at 70a–71a. Moreover, respondent was not physically restrained or threatened and was interviewed in a well-lit, average-sized conference room, where he was “not uncomfortable.” Id., at 90a; see id., at 71a, 88a–89a. He was offered food and water, and the door to the conference room was sometimes left open. See id., at 70a, 74a. “All of these objective facts are consistent with an interrogation environment in which a reasonable person would have felt free to terminate the interview and leave.” Yarborough, supra, at 664–665. Because he was in prison, respondent was not free to leave the conference room by himself and to make his own way through the facility to his cell. Instead, he was escorted to the conference room and, when he ultimately decided to end the interview, he had to wait about 20 minutes for a corrections officer to arrive and escort him to his cell. But he would have been subject to this same restraint even if he had been taken to the conference room for some reason other than police questioning; under no circumstances could he have reasonably expected to be able to roam free.[6] And while respondent testified that he “was told . . . if I did not want to cooperate, I needed to go back to my cell,” these words did not coerce cooperation by threatening harsher conditions. App. to Pet. for Cert. 71a; see id., at 89a (“I was told, if I didn’t want to cooperate, I could leave”). Returning to his cell would merely have returned him to his usual environment. See Shatzer, supra, at ___ (slip op., at 14) (“Interrogated suspects who have previously been convicted of crime live in prison. When they are released back into the general prison population, they return to their accustomed surroundings and daily routine—they regain the degree of control they had over their lives prior to the interrogation”). Taking into account all of the circumstances of the questioning—including especially the undisputed fact that respondent was told that he was free to end the questioning and to return to his cell—we hold that respondent was not in custody within the meaning of Miranda. * * * The judgment of the Court of Appeals is Reversed. Notes 1 Fields testified that he left his cell around 8 p.m. and that the in-terview began around 8:30 p.m. App. to Pet. for Cert. 77a. Both the Michigan Court of Appeals and the Sixth Circuit stated that the interview began between 7 p.m. and 9 p.m. See id., at 4a, 54a. 2 The Court of Appeals stated that the interview lasted for approximately seven hours, see id., at 4a, a figure that appears to be basedon the testimony of one of the interviewing deputies, see id., at 123a. Fields put the number of hours between five and five and a half, saying the interview began around 8:30 p.m. and continued until 1:30 a.m. or 2 a.m. See id., at 77a. The Michigan Court of Appeals stated that the interview ended around midnight, which would put the length of the interview at between three and five hours. 3 Fields testified that his normal bedtime was 10:30 p.m. or 11 p.m. See id., at 78a. 4 Indeed, it is impossible to tell from either the opinion of this Court or that of the court below whether the prisoner’s interview was routine or whether there were special features that may have created an especially coercive atmosphere. 5 The state-court decision applied the traditional context-specific analysis to determine whether the circumstances of respondent’s interrogation gave rise to “the coercive pressure that Miranda was designed to guard against.” Shatzer, 559 U. S., at ___ (slip op., at 13). The court first observed: “That a defendant is in prison for an unrelated offense when being questioned does not, without more, mean that he was in custody for the purpose of determining whether Miranda warnings were required.” App. to Pet. for Cert. 56a (internal quotation marks omitted and emphasis added). In this case, the court noted, the “defendant was unquestionably in custody, but on a matter unrelated to the interrogation.” Ibid. The Sixth Circuit concluded that the state court thereby limited Miranda in a way rejected by Mathis v. United States, 391 U.S. 1 (1968), and “curtail[ed] the warnings to be given persons under interrogation by officers based on the reason why the person is in custody.” Id., at 4–5. We think the better reading is that the state court merely meant to draw a distinction between incarceration and Miranda custody. This reading is supported by the state court’s subsequent consideration of whether the facts of the case were likely to create an atmosphere of coercion. App. to Pet. for Cert. 56a. 6 Respondent did not testify to the contrary. The following colloquy occurred at his Miranda hearing: “Q. You’re not generally allowed to just roam around Lenawee County Jail on your own, are you? “A. No, I never have. “Q. So wouldn’t it make sense to you, since you had that experience, that in fact you would have been escorted just like you were escorted . . . into this conference room? “A. That makes common sense. “Q. So when they said that you were free to leave and you get up—could get up and go and all you had to do was tell them you wanted to go, in your mind, did you understand that to mean that somebody would come get you and take you back to your cell? “A. But that doesn’t give me freedom to just get up and walk away. “Q. I understand it doesn’t— “A. So, no. “Q. The question is this, sir, not whether you had freedom to get up and walk away, but did you understand that what that meant was that a jailer would come get you and— “A. No— “Q. —take you back to your cell? “A. I did not understand that. “Q. You didn’t? “A. No. “Q. Why not? That’s how you got there. “A. Because I did not know if a jailer would take me back or if one of those gentlemen would take me back. “Q. But you understood that, if you asked, one of them or a jailer would take you back to your cell? “A. I assumed that. “Q. And you believed that to be true? “A. I assumed that.” App. to Pet. for Cert. 91a–92a.

565.US.2011_10-694

Federal immigration law governs both the exclusion of aliens from admission to this country and the deportation of aliens previously admitted. Before 1996, these two kinds of action occurred in different procedural settings; since then, the Government has employed a unified “removal proceeding” for exclusions and deportations alike. But the immigration laws have always provided separate lists of substantive grounds for the two actions. One list specifies what crimes render an alien excludable, see 8 U. S. C. §1182(a), while another—sometimes overlapping and sometimes divergent—list specifies what crimes render an alien deportable, see §1227(a). Until repealed in 1996, §212(c) of the Immigration and Nationality Act permitted the Attorney General to grant discretionary relief to an excludable alien, if the alien had lawfully resided in the United States for at least seven years before temporarily leaving the country and if the alien was not excludable on one of two specified grounds. By its terms, §212(c) applied only in exclusion proceedings, but the Board of Immigration Appeals (BIA) extended it decades ago to deportation proceedings as well. Although Congress substituted a narrower discretionary remedy for §212(c) in 1996, see §1229b, §212(c)’s broader relief remains available to an alien whose removal is based on a guilty plea entered before §212(c)’s repeal, INS v. St. Cyr, 533 U.S. 289, 326. In deciding whether to exclude such an alien, the BIA first checks the statutory ground identified by the Department of Homeland Security (DHS) as the basis for exclusion. Unless that ground is one of the two falling outside §212(c)’s scope, the alien is eligible for discretionary relief. The BIA then determines whether to grant relief based on such factors as the seriousness of the offense. This case concerns the BIA’s method for applying §212(c) in the deportation context. The BIA’s approach, known as the “comparable-grounds” rule, evaluates whether the charged deportation ground has a close analogue in the statute’s list of exclusion grounds. If the deportation ground consists of a set of crimes “substantially equivalent” to the set making up an exclusion ground, the alien can seek §212(c) relief. But if the deportation ground covers different or more or fewer offenses than any exclusion ground, the alien is ineligible for relief, even if the alien’s particular offense falls within an exclusion ground. Petitioner Judulang, who has lived continuously in the United States as a lawful permanent resident since 1974, pleaded guilty to voluntary manslaughter in 1988. After he pleaded guilty to another crime in 2005, DHS commenced a deportation action, charging him with having committed an “aggravated felony” involving “a crime of violence” based on his manslaughter conviction. The Immigration Judge ordered Judulang’s deportation, and the BIA affirmed, finding Judulang ineligible for §212(c) relief because the “crime of violence” deportation ground is not comparable to any exclusion ground. The Ninth Circuit, having previously upheld the BIA’s comparable-grounds rule, denied Judulang’s petition for review. Held: The BIA’s policy for applying §212(c) in deportation cases is “arbitrary and capricious” under the Administrative Procedure Act, 5 U. S. C. §706(2)(A). Pp. 9–21. (a) While agencies have expertise and experience in administering their statutes that no court may properly ignore, courts retain a narrow but important role in ensuring that agencies have engaged in reasoned decisionmaking. Thus, in reviewing the BIA’s action, this Court must assess, among other matters, “whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.” Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U.S. 29, 43. That task involves examining the reasons for agency decisions, or the absence of such reasons. The comparable-grounds approach cannot survive scrutiny under this standard. By hinging a deportable alien’s eligibility for discretionary relief on the chance correspondence between statutory categories—a matter irrelevant to the alien’s fitness to reside in this country—the BIA has failed to exercise its discretion in a reasoned manner. Pp. 9–10. (b) Even if the BIA has legitimate reasons for limiting §212(c)’s scope in deportation cases, it must do so in some rational way. In other words, the BIA must use an approach that is tied to the purposes of the immigration laws or the appropriate operation of the immigration system. The comparable-grounds rule has no connection to these factors. Instead, it makes §212(c) eligibility turn on an irrelevant comparison between statutory provisions. Whether the set of offenses in a particular deportation ground lines up with the set in an exclusion ground has nothing to do with whether a deportable alien whose prior conviction falls within both grounds merits the ability to stay in this country. Here, Judulang was found ineligible for §212(c) relief because the “crime of violence” deportation ground includes a few offenses—simple assault, minor burglary, and unauthorized use of a vehicle—not found in the similar moral turpitude exclusion ground. But the inclusion of simple assaults and minor burglaries in the deportation ground is irrelevant to the merits of Judulang’s case. The BIA’s approach has other odd features. In applying the comparable-grounds rule, the BIA has denied relief to aliens whose deportation ground fits entirely within a much broader exclusion ground. Yet providing relief in exclusion cases to a broad class of aliens hardly justifies denying relief in deportation cases to a subset of that group. In addition, the outcome of the comparable-grounds analysis may itself rest on an arbitrary decision. An alien’s prior conviction could fall within a number of deportation grounds, only one of which corresponds to an exclusion ground. In such cases, an alien’s eligibility for relief would hinge on an individual official’s decision as to which deportation ground to charge. An alien appearing before one official may suffer deportation, while an identically situated alien appearing before another may gain the right to stay in this country. In short, the comparable-grounds approach does not rest on any factors relevant to whether an alien should be deported. Instead, it turns deportation decisions into a “sport of chance.” Rosenberg v. Fleuti, 374 U.S. 449, 455. That is what the APA’s “arbitrary and capricious” standard is designed to prevent. Pp. 10–15. (c) The Government’s arguments in defense of the comparable-grounds rule are not persuasive. First, §212(c)’s text does not support the rule. That section cannot provide a textual anchor for any method of providing discretionary relief in deportation cases because it addresses only exclusion. Second, the history of the comparable-grounds rule does not work in the Government’s favor. The BIA repeatedly vacillated in its method for applying §212(c) to deportable aliens, settling on the current rule only in 2005. Third, the Government’s claim that the comparable-grounds rule saves time and money falls short. Cost may be an important factor for agencies to consider in many contexts, but cheapness alone cannot save an arbitrary agency policy. In any event, it is unclear that the comparable-grounds rule saves money when compared with alternative approaches. Pp. 16–21. 249 Fed. Appx. 499, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court.

This case concerns the Board of Immigration Appeals’ (BIA or Board) policy for deciding when resident aliens may apply to the Attorney General for relief from deportation under a now-repealed provision of the immigration laws. We hold that the BIA’s approach is arbitrary and capricious. The legal background of this case is complex, but the principle guiding our decision is anything but. When an administrative agency sets policy, it must provide a reasoned explanation for its action. That is not a high bar, but it is an unwavering one. Here, the BIA has failed to meet it. I A Federal immigration law governs both the exclusion of aliens from admission to this country and the deportation of aliens previously admitted. Before 1996, these two kinds of action occurred in different procedural settings, with an alien seeking entry (whether for the first time or upon return from a trip abroad) placed in an “exclusion proceeding” and an alien already here channeled to a “deportation proceeding.” See Landon v. Plasencia, 459 U. S. 21 –26 (1982) (comparing the two). Since that time, the Government has used a unified procedure, known as a “removal proceeding,” for exclusions and deportations alike. See 8 U. S. C. §§1229, 1229a. But the statutory bases for excluding and deporting aliens have always varied. Now, as before, the immigration laws provide two separate lists of substantive grounds, principally involving criminal offenses, for these two actions. One list specifies what kinds of crime render an alien excludable (or in the term the statute now uses, “inadmissible”), see §1182(a) (2006 ed., Supp. IV), while another—sometimes overlapping and sometimes divergent—list specifies what kinds of crime render an alien deportable from the country, see §1227(a). An additional, historic difference between exclusion and deportation cases involved the ability of the Attorney General to grant an alien discretionary relief. Until repealed in 1996, §212(c) of the Immigration and Nationality Act, 66Stat. 187, 8 U. S. C. §1182(c) (1994 ed.), authorized the Attorney General to admit certain excludable aliens. See also §136(p) (1926 ed.) (predecessor provision to §212(c)). The Attorney General could order this relief when the alien had lawfully resided in the United States for at least seven years before temporarily leaving the country, unless the alien was excludable on one of two specified grounds. See §1182(c) (1994 ed.). [ 1 ] But by its terms, §212(c) did not apply when an alien was being deported. This discrepancy threatened to produce an odd result in a case called Matter of L-, 1 I. & N. Dec. 1 (1940), leading to the first-ever grant of discretionary relief in a deportation case. L- was a permanent resident of the United States who had been convicted of larceny. Although L-’s crime made him inadmissible, he traveled abroad and then returned to the United States without any immigration official’s preventing his entry. A few months later, the Government caught up with L- and initiated a deportation action based on his larceny conviction. Had the Government apprehended L- at the border a short while earlier, he would have been placed in an exclusion proceeding where he could have applied for discretionary relief. But because L- was instead in a deportation proceeding, no such relief was available. Responding to this apparent anomaly, Attorney General Robert Jackson (on referral of the case from the BIA) determined that L- could receive a waiver: L-, Jackson said, “should be permitted to make the same appeal to discretion that he could have made if denied admission” when returning from his recent trip. Id., at 6. In accord with this decision, the BIA adopted a policy of allowing aliens in deportation proceedings to apply for discretionary relief under §212(c) whenever they had left and reentered the country after be- coming deportable. See Matter of S-, 6 I. & N. Dec. 392, 394–396 (1954). But this approach created another peculiar asymmetry: Deportable aliens who had traveled abroad and returned could receive §212(c) relief, while those who had never left could not. In Francis v. INS, 532 F. 2d 268 (1976), the Court of Appeals for the Second Circuit concluded that this disparity violated equal protection. Id., at 273 (“[A]n alien whose ties with this country are so strong that he has never departed after his initial entry should receive at least as much consideration as an individual who may leave and return from time to time”). The BIA acquiesced in the Second Circuit’s decision, see Matter of Silva, 16 I. & N. Dec. 26 (1976), thus applying §212(c) in deportation proceedings regardless of an alien’s travel history. All this might have become academic when Congress repealed §212(c) in 1996 and substituted a new discretionary remedy, known as “cancellation of removal,” which is available in a narrow range of circumstances to excludable and deportable aliens alike. See 8 U. S. C. §1229b. But in INS v. St. Cyr, 533 U. S. 289, 326 (2001) , this Court concluded that the broader relief afforded by §212(c) must remain available, on the same terms as before, to an alien whose removal is based on a guilty plea entered before §212(c)’s repeal. We reasoned that aliens had agreed to those pleas with the possibility of discretionary relief in mind and that eliminating this prospect would ill comport with “ ‘familiar considerations of fair notice, reasonable reliance, and settled expectations.’ ” Id., at 323 (quoting Landgraf v. USI Film Products, 511 U. S. 244, 270 (1994) ). Accordingly, §212(c) has had an afterlife for resident aliens with old criminal convictions. When the BIA is deciding whether to exclude such an alien, applying §212(c) is an easy matter. The Board first checks the statutory ground that the Department of Homeland Security (DHS) has identified as the basis for exclusion; the Board may note, for example, that DHS has charged the alien with previously committing a “crime involving moral turpitude,” see 8 U. S. C. §1182(a)(2)(A)(i)(I). Unless the charged ground is one of the pair falling outside §212(c)’s scope, see n. 1, supra, the alien is eligible for discretionary relief. The Board then determines whether to grant that relief based on such factors as “the seriousness of the offense, evidence of either rehabilitation or recidivism, the duration of the alien’s residence, the impact of deportation on the family, the number of citizens in the family, and the character of any service in the Armed Forces.” St. Cyr, 533 U. S., at 296, n. 5. By contrast, when the BIA is deciding whether to deport an alien, applying §212(c) becomes a tricky business. Recall that §212(c) applies on its face only to exclusion decisions. So the question arises: How is the BIA to determine when an alien should receive §212(c) relief in the deportation context? One approach that the BIA formerly used considered how the alien would fare in an exclusion proceeding. To perform this analysis, the Board would first determine whether the criminal conviction making the alien deportable fell within a statutory ground for exclusion. Almost all convictions did so, largely because the “crime involving moral turpitude” ground encompasses so many offenses. [ 2 ] Assuming that threshold inquiry were met, the Board would mimic its approach in exclusion cases—first making sure the statutory ground at issue was not excepted from §212(c) and then conducting the multi-factor analysis. See Matter of Tanori, 15 I. & N. Dec. 566, 567–568 (1976); In re Manzueta, No. A93 022 672, 2003 WL 23269892 (BIA, Dec. 1, 2003). A second approach is the one challenged here; definitively adopted in 2005 (after decades of occasional use), it often is called the “comparable-grounds” rule. See, e.g., De la Rosa v. U. S. Attorney General, 579 F. 3d 1327, 1332 (CA11 2009). That approach evaluates whether the ground for deportation charged in a case has a close analogue in the statute’s list of exclusion grounds. See In re Blake, 23 I. & N. Dec. 722, 728 (2005); In re Brieva-Perez, 23 I. & N. Dec. 766, 772–773 (2005). [ 3 ] If the deportation ground consists of a set of crimes “substantially equivalent” to the set of offenses making up an exclusion ground, then the alien can seek §212(c) relief. Blake, 23 I. & N. Dec., at 728. But if the deportation ground charged covers significantly different or more or fewer offenses than any exclusion ground, the alien is not eligible for a waiver. Such a divergence makes §212(c) inapplicable even if the particular offense committed by the alien falls within an exclusion ground. Two contrasting examples from the BIA’s cases may help to illustrate this approach. Take first an alien convicted of conspiring to distribute cocaine, whom DHS seeks to deport on the ground that he has commit- ted an “aggravated felony” involving “illicit trafficking in a controlled substance.” 8 U. S. C. §§1101(a)(43)(B), 1227(a)(2)(A)(iii). Under the comparable-grounds rule, the immigration judge would look to see if that deportation ground covers substantially the same offenses as an exclusion ground. And according to the BIA in Matter of Meza, 20 I. & N. Dec. 257 (1991), the judge would find an adequate match—the exclusion ground applicable to aliens who have committed offenses “relating to a controlled substance,” 8 U. S. C. §§1182(a)(2)(A)(i)(II) and (a)(2)(C). Now consider an alien convicted of first-degree sexual abuse of a child, whom DHS wishes to deport on the ground that he has committed an “aggravated felony” involving “sexual abuse of a minor.” §§1101(a)(43)(A), 1227(a)(2)(A)(iii). May this alien seek §212(c) relief ? According to the BIA, he may not do so—not because his crime is too serious (that is irrelevant to the analysis), but instead because no statutory ground of exclusion covers substantially the same offenses. To be sure, the alien’s own offense is a “crime involving moral turpitude,” 8 U. S. C. §1182(a)(2)(A)(i)(I), and so fits within an exclusion ground. Indeed, that will be true of most or all offenses included in this deportation category. See supra, at 5. But on the BIA’s view, the “moral turpitude” exclusion ground “addresses a distinctly different and much broader category of offenses than the aggravated felony sexual abuse of a minor charge.” Blake, 23 I. & N. Dec., at 728. And the much greater sweep of the exclusion ground prevents the alien from seeking discretionary relief from deportation. [ 4 ] Those mathematically inclined might think of the comparable-grounds approach as employing Venn dia- grams. Within one circle are all the criminal offenses com- posing the particular ground of deportation charged. Within other circles are the offenses composing the various exclusion grounds. When, but only when, the “deportation circle” sufficiently corresponds to one of the “exclusion circles” may an alien apply for §212(c) relief. B Petitioner Joel Judulang is a native of the Philippines who entered the United States in 1974 at the age of eight. Since that time, he has lived continuously in this country as a lawful permanent resident. In 1988, Judulang took part in a fight in which another person shot and killed someone. Judulang was charged as an accessory and eventually pleaded guilty to voluntary manslaughter. He received a 6-year suspended sentence and was released on probation immediately after his plea. In 2005, after Judulang pleaded guilty to another criminal offense (this one involving theft), DHS commenced an action to deport him. DHS charged Judulang with having committed an “aggravated felony” involving “a crime of violence,” based on his old manslaughter conviction. 8 U. S. C. §§1101(a)(43)(F), 1227(a)(2)(A)(iii). [ 5 ] The Immigration Judge ordered Judulang’s deportation, and the BIA affirmed. As part of its decision, the BIA considered whether Judulang could apply for §212(c) relief. It held that he could not do so because the “crime of violence” deportation ground is not comparable to any exclusion ground, including the one for crimes involving moral turpitude. App. to Pet. for Cert. 8a. The Court of Appeals for the Ninth Circuit denied Judulang’s petition for review in reliance on circuit precedent upholding the BIA’s comparable-grounds approach. Judulang v. Gonzales, 249 Fed. Appx. 499, 502 (2007) (citing Abebe v. Gonzales, 493 F. 3d 1092 (2007)). We granted certiorari, 563 U. S. ___ (2011), to resolve a circuit split on the approach’s validity. [ 6 ] We now reverse. II This case requires us to decide whether the BIA’s policy for applying §212(c) in deportation cases is “arbitrary [or] capricious” under the Administrative Procedure Act (APA), 5 U. S. C. §706(2)(A). [ 7 ] The scope of our review under this standard is “narrow”; as we have often recognized, “a court is not to substitute its judgment for that of the agency.” Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U. S. 29, 43 (1983) ; see Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U. S. 402, 416 (1971) . Agencies, the BIA among them, have expertise and experience in administering their statutes that no court can properly ignore. But courts retain a role, and an important one, in ensuring that agencies have engaged in reasoned decisionmaking. When reviewing an agency action, we must assess, among other matters, “ ‘whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.’ ” State Farm, 463 U. S., at 43 (quoting Bowman Transp., Inc. v. Arkansas-Best Freight System, Inc., 419 U. S. 281, 285 (1974) ). That task involves examining the reasons for agency decisions—or, as the case may be, the absence of such reasons. See FCC v. Fox Television Stations, Inc., 556 U. S. 502, 515 (2009) (noting “the requirement that an agency provide reasoned explanation for its action”). The BIA has flunked that test here. By hinging a deportable alien’s eligibility for discretionary relief on the chance correspondence between statutory categories—a matter irrelevant to the alien’s fitness to reside in this country—the BIA has failed to exercise its discretion in a reasoned manner. A The parties here spend much time disputing whether the BIA must make discretionary relief available to deportable and excludable aliens on identical terms. As this case illustrates, the comparable-grounds approach does not do so. If Judulang were seeking entry to this country, he would be eligible for §212(c) relief; voluntary manslaughter is “a crime involving moral turpitude,” and so his conviction falls within an exclusion ground. But Judulang cannot apply for relief from deportation because the “crime of violence” ground charged in his case does not match any exclusion ground (including the one for “turpitudinous” crimes). See infra, at 13. Judulang argues that this disparity is impermissible because any disparity between excludable and deportable aliens is impermissible: If an alien may seek §212(c) relief in an exclusion case, he also must be able to seek such relief in a deportation case. See Brief for Petitioner 47–51. [ 8 ] But the Government notes that the immigration laws have always drawn distinctions between exclusion and deportation. See Brief for Respondent 51. And the Government presses a policy reason for making §212(c) relief more readily available in exclusion cases. Doing so, it argues, will provide an incentive for some resident aliens (i.e., those eligible for a waiver from exclusion, but not deportation) to report themselves to immigration officials, by applying for advance permission to exit and reenter the country. In contrast, applying §212(c) uniformly might lead all aliens to “try to evade immigration officials for as long as possible,” because they could in any event “seek [discretionary] relief if caught.” Id., at 52. In the end, we think this dispute beside the point, and we do not resolve it. The BIA may well have legitimate reasons for limiting §212(c)’s scope in deportation cases. But still, it must do so in some rational way. If the BIA proposed to narrow the class of deportable aliens eligible to seek §212(c) relief by flipping a coin—heads an alien may apply for relief, tails he may not—we would reverse the policy in an instant. That is because agency action must be based on non-arbitrary, “ ‘relevant factors,’ ” State Farm, 463 U. S., at 43 (quoting Bowman Transp., 419 U. S., at 285), which here means that the BIA’s approach must be tied, even if loosely, to the purposes of the immigration laws or the appropriate operation of the immigration system. A method for disfavoring deportable aliens that bears no relation to these matters—that neither focuses on nor relates to an alien’s fitness to remain in the country—is arbitrary and capricious. And that is true regardless whether the BIA might have acted to limit the class of deportable aliens eligible for §212(c) relief on other, more rational bases. The problem with the comparable-grounds policy is that it does not impose such a reasonable limitation. Rather than considering factors that might be thought germane to the deportation decision, that policy hinges §212(c) eligibility on an irrelevant comparison between statutory provisions. Recall that the BIA asks whether the set of offenses in a particular deportation ground lines up with the set in an exclusion ground. But so what if it does? Does an alien charged with a particular deportation ground become more worthy of relief because that ground happens to match up with another? Or less worthy of relief because the ground does not? The comparison in no way changes the alien’s prior offense or his other attributes and circumstances. So it is difficult to see why that comparison should matter. Each of these statutory grounds contains a slew of offenses. Whether each contains the same slew has nothing to do with whether a deportable alien whose prior conviction falls within both grounds merits the ability to seek a waiver. [ 9 ] This case well illustrates the point. In commencing Judulang’s deportation proceeding, the Government charged him with an “aggravated felony” involving a “crime of violence” based on his prior manslaughter conviction. See App. to Pet. for Cert. 11a–12a. That made him ineligible for §212(c) relief because the “crime of violence” deportation ground does not sufficiently overlap with the most similar exclusion ground, for “crime[s] involving moral turpitude.” The problem, according to the BIA, is that the “crime of violence” ground includes a few offenses—simple assault, minor burglary, and unauthorized use of a vehicle—that the “moral turpitude” ground does not. See Brieva-Perez, 23 I. & N. Dec., at 772–773; Tr. of Oral Arg. 28–29, 40–41. But this statutory difference in no way relates to Judulang—or to most other aliens charged with committing a “crime of violence.” Perhaps aliens like Judulang should be eligible for §212(c) relief, or perhaps they should not. But that determination is not sensibly made by establishing that simple assaults and minor burglaries fall outside a ground for exclusion. That fact is as extraneous to the merits of the case as a coin flip would be. It makes Judulang no less deserving of the opportunity to seek discretionary relief—just as its converse (the inclusion of simple assaults and burglaries in the “moral turpitude” exclusion ground) would make him no more so. Or consider a different headscratching oddity of the comparable-grounds approach—that it may deny §212(c) eligibility to aliens whose deportation ground fits entirely inside an exclusion ground. The BIA’s Blake decision, noted earlier, provides an example. See supra, at 6–7. The deportation ground charged was “aggravated felony” involving “sexual abuse of a minor”; the closest exclusion ground was, once again, a “crime [of] moral turpitude.” 23 I. & N. Dec., at 727. Here, the BIA’s problem was not that the deportation ground covered too many offenses; all or virtually all the crimes within that ground also are crimes of moral turpitude. Rather, the BIA objected that the deportation ground covered too few crimes—or put oppositely, that “the moral turpitude ground of exclusion addresses a . . . much broader category of offenses.” Id., at 728. But providing relief in exclusion cases to a broad class of aliens hardly justifies denying relief in deportation cases to a subset of that group. [ 10 ] (The better argument would surely be the reverse—that giving relief in the one context supports doing so in the other.) Again, we do not say today that the BIA must give all deportable aliens meeting §212(c)’s requirements the chance to apply for a waiver. See supra, at 11–12. The point is instead that the BIA cannot make that opportunity turn on the meaningless matching of statutory grounds. And underneath this layer of arbitrariness lies yet another, because the outcome of the Board’s comparable-grounds analysis itself may rest on the happenstance of an immigration official’s charging decision. This problem arises because an alien’s prior conviction may fall within a number of deportation grounds, only one of which corresponds to an exclusion ground. Consider, for example, an alien who entered the country in 1984 and commit- ted voluntary manslaughter in 1988. That person could be charged (as Judulang was) with an “aggravated fel- ony” involving a “crime of violence,” see 8 U. S. C. §§1101(a)(43)(F), 1227(a)(2)(A)(iii). If so, the alien could not seek a waiver because of the absence of a comparable exclusion ground. But the alien also could be charged with “a crime involving moral turpitude committed within five years . . . after the date of admission,” see §1227(a)(2)(A)(i)(I). And if that were the deportation charge, the alien could apply for relief, because the ground corresponds to the “moral turpitude” ground used in exclusion cases. See In re Salmon, 16 I. & N. Dec. 734 (1978). So everything hangs on the charge. And the Government has provided no reason to think that immigration officials must adhere to any set scheme in deciding what charges to bring, or that those officials are exercising their charging discretion with §212(c) in mind. See Tr. of Oral Arg. 34–36. So at base everything hangs on the fortuity of an individual official’s decision. An alien appearing before one official may suffer deportation; an identically situated alien appearing before another may gain the right to stay in this country. In a foundational deportation case, this Court recognized the high stakes for an alien who has long resided in this country, and reversed an agency decision that would “make his right to remain here dependent on circumstan- ces so fortuitous and capricious.” Delgadillo v. Carmichael, 332 U. S. 388, 391 (1947) . We think the policy before us similarly flawed. The comparable-grounds approach does not rest on any factors relevant to whether an alien (or any group of aliens) should be deported. It instead distinguishes among aliens—decides who should be eligible for discretionary relief and who should not—solely by comparing the metes and bounds of diverse statutory categories into which an alien falls. The resulting Venn diagrams have no connection to the goals of the deportation process or the rational operation of the immigration laws. Judge Learned Hand wrote in another early immigration case that deportation decisions cannot be made a “sport of chance.” See Di Pasquale v. Karnuth, 158 F. 2d 878, 879 (CA2 1947) (quoted in Rosenberg v. Fleuti, 374 U. S. 449, 455 (1963) ). That is what the comparable-grounds rule brings about, and that is what the APA’s “arbitrary and capricious” standard is designed to thwart. B The Government makes three arguments in defense of the comparable-grounds rule—the first based on statutory text, the next on history, the last on cost. We find none of them persuasive. 1 The Government initially contends that the comparable-grounds approach is more faithful to “the statute’s language,” Brief for Respondent 21—or otherwise said, that “lifting that limit ‘would take immigration practice even further from the statutory text,’ ” id., at 22 (quoting Matter of Hernandez-Casillas, 20 I. & N. Dec. 262, 287 (1990)). In the Government’s view, §212(c) is “phrased in terms of waiving statutorily specified grounds of exclusion”; that phrasing, says the Government, counsels a comparative analysis of grounds when applying §212(c) in the deportation context. Brief for Respondent 21; see Tr. of Oral Arg. 34 (“[T]he reason [the comparable-grounds approach] makes sense is because the statute only provides for relief from grounds of . . . exclusion”). The first difficulty with this argument is that it is based on an inaccurate description of the statute. Section 212(c) instructs that certain resident aliens “may be admitted in the discretion of the Attorney General” notwithstanding any of “the provisions of subsection (a) . . . (other than paragraphs (3) and (9)(C)).” 8 U. S. C. §1182(c) (1994 ed.). Subsection (a) contains the full list of exclusion grounds; paragraphs (3) and (9)(C) (which deal with national security and international child abduction) are two among these. What §212(c) actually says, then, is that the Attorney General may admit any excludable alien, except if the alien is charged with two specified grounds. And that means that once the Attorney General determines that the alien is not being excluded for those two reasons, the ground of exclusion no longer matters. At that point, the alien is eligible for relief, and the thing the Attorney General waives is not a particular exclusion ground, but the simple denial of entry. So the premise of the Government’s argument is wrong. And if the premise, so too the conclusion—that is, because §212(c)’s text is not “phrased in terms of waiving statutorily specified grounds of exclusion,” Brief for Respondent 21, it cannot counsel a search for corresponding grounds of deportation. More fundamentally, the comparable-grounds approach would not follow from §212(c) even were the Government right about the section’s phrasing. That is because §212(c) simply has nothing to do with deportation: The provision was not meant to interact with the statutory grounds for deportation, any more than those grounds were designed to interact with the provision. Rather, §212(c) refers solely to exclusion decisions; its extension to deportation cases arose from the agency’s extra-textual view that some similar relief should be available in that context to avoid unreasonable distinctions. Cf., e.g., Matter of L-, 1 I. & N. Dec., at 5; see also supra, at 3–4. [ 11 ] Accordingly, the text of §212(c), whether or not phrased in terms of “waiving grounds of exclusion,” cannot support the BIA’s use of the comparable-grounds rule—or, for that matter, any other method for extending discretionary relief to deportation cases. We well understand the difficulties of operating in such a text-free zone; indeed, we appreciate the Government’s yearning for a textual anchor. But §212(c), no matter how many times read or parsed, does not provide one. 2 In disputing Judulang’s contentions, the Government also emphasizes the comparable-grounds rule’s vintage. See Brief for Respondent 22–23, 30–43. As an initial matter, we think this a slender reed to support a significant government policy. Arbitrary agency action becomes no less so by simple dint of repetition. (To use a prior analogy, flipping coins to determine §212(c) eligibility would remain as arbitrary on the thousandth try as on the first.) And longstanding capriciousness receives no special exemption from the APA. In any event, we cannot detect the consistency that the BIA claims has marked its approach to this issue. To the contrary, the BIA has repeatedly vacillated in its method for applying §212(c) to deportable aliens. Prior to 1984, the BIA endorsed a variety of approaches. In Matter of T-, 5 I. & N. Dec. 389, 390 (1953), for example, the BIA held that an alien was not eligible for §212(c) relief because her “ground of deportation” did not appear in the exclusion statute. That decision anticipated the comparable-grounds approach that the BIA today uses. But in Tanori, the BIA pronounced that a deportable alien could apply for a waiver because “the same facts”—in that case, a marijuana conviction—would have allowed him to seek §212(c) relief in an exclusion proceeding. 15 I. & N. Dec., at 568. That approach is more nearly similar to the one Judulang urges here. And then, in Matter of Granados, 16 I. & N. Dec. 726, 728 (1979), the BIA tried to have it both ways: It denied §212(c) eligibility both because the deportation ground charged did not correspond to, and because the alien’s prior conviction did not fall within, a waivable ground of exclusion. In short, the BIA’s cases were all over the map. The Government insists that the BIA imposed order in Matter of Wadud, 19 I. & N. Dec. 182, 185–186 (1984), when it held that a deportable alien could not seek §212(c) relief unless the deportation ground charged had an “analogous ground of inadmissibility.” See Brief for Respondent 40–41. But the BIA’s settlement, if any, was fleeting. Just seven years later, the BIA adopted a new policy entirely, extending §212(c) eligibility to “aliens deportable under any ground of deportability except those where there is a comparable ground of exclusion which has been specifically excepted from section 212(c).” Hernandez-Casillas, 20 I. & N. Dec., at 266. That new rule turned the comparable-grounds approach inside-out, allowing aliens to seek §212(c) relief in deportation cases except when the ground charged corresponded to an exclusion ground that could not be waived. To be sure, the Attorney General (on referral of the case from the BIA), disavowed this position in favor of the more standard version of the comparable-grounds rule. Id., at 287. But even while doing so, the Attorney General stated that “an alien subject to deportation must have the same opportunity to seek discretionary relief as an alien . . . subject to exclusion.” Ibid. That assertion is exactly the one Judulang makes in this case; it is consonant not with the comparable-grounds rule the BIA here defends, but instead with an inquiry into whether an alien’s prior conviction falls within an exclusion ground. Given these mixed signals, it is perhaps not surprising that the BIA continued to alternate between approaches in the years that followed. Immediately after the Attorney General’s opinion in Hernandez-Casillas, the BIA endorsed the comparable-grounds approach on several occasions. See Meza, 20 I. & N. Dec., at 259; Matter of Montenegro, 20 I. & N. Dec. 603, 604–605 (1992); Matter of Gabryelsky, 20 I. & N. Dec. 750, 753–754 (1993); In re Esposito, 21 I. & N. Dec. 1, 6–7 (1995); In re Jimenez-Santillano, 21 I. & N. Dec. 567, 571–572 (1996). But just a few years later, the BIA issued a series of unpublished opinions that asked only whether a deportable alien’s prior conviction fell within an exclusion ground. See, e.g., In re Manzueta, No. A93 022 672, 2003 WL 23269892 (Dec. 1, 2003). Not until the BIA’s decisions in Blake and Brieva-Perez did the pendulum stop swinging. That history hardly supports the Government’s view of a consistent agency practice. [ 12 ] 3 The Government finally argues that the comparable-grounds rule saves time and money. The Government claims that comparing deportation grounds to exclusion grounds can be accomplished in just a few “precedential decision[s],” which then can govern broad swaths of cases. See Brief for Respondent 46. By contrast, the Government argues, Judulang’s approach would force it to determine whether each and every crime of conviction falls within an exclusion ground. Further, the Government contends that Judulang’s approach would grant eligibility to a greater number of deportable aliens, which in turn would force the Government to make additional individualized assessments of whether to actually grant relief. Id., at 47. Once again, the Government’s rationale comes up short. Cost is an important factor for agencies to consider in many contexts. But cheapness alone cannot save an arbitrary agency policy. (If it could, flipping coins would be a valid way to determine an alien’s eligibility for a waiver.) And in any event, we suspect the Government exaggerates the cost savings associated with the comparable-grounds rule. Judulang’s proposed approach asks immigration officials only to do what they have done for years in exclusion cases; that means, for one thing, that officials can make use of substantial existing precedent governing whether a crime falls within a ground of exclusion. And Judulang’s proposal may not be the only alternative to the comparable-grounds rule. See supra, at 11–12. In rejecting that rule, we do not preclude the BIA from trying to devise another, equally economical policy respecting eligibility for §212(c) relief, so long as it comports with everything held in both this decision and St. Cyr. III We must reverse an agency policy when we cannot discern a reason for it. That is the trouble in this case. The BIA’s comparable-grounds rule is unmoored from the purposes and concerns of the immigration laws. It allows an irrelevant comparison between statutory provisions to govern a matter of the utmost importance—whether lawful resident aliens with longstanding ties to this country may stay here. And contrary to the Government’s protestations, it is not supported by text or practice or cost considerations. The BIA’s approach therefore cannot pass muster under ordinary principles of administrative law. The judgment of the Ninth Circuit is hereby reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 The relevant part of §212(c), in the version of the exclusion statute all parties use, read as follows: “Aliens lawfully admitted for permanent residence who temporarily proceeded abroad voluntarily and not under an order of deportation, and who are returning to a lawful unrelinquished domicile of seven consecutive years, may be admitted in the discretion of the Attorney General without regard to the provisions of subsection (a) of this section (other than paragraphs (3) and (9)(C)).” (1994 ed.). The parenthetical clause of this section prevented the Attorney General from waiving exclusion for aliens who posed a threat to national security, §1182(a)(3), and aliens who engaged in international child abduction, §1182(a)(9)(C). 2 Firearms offenses are the most significant crimes falling outside the statutory grounds for exclusion. See Matter of Hernandez-Casillas, 20 I. & N. Dec. 262, 282, n. 4 (1990). 3 Blake and Brieva-Perez clarified a 2004 regulation issued by the BIA stating that an alien is ineligible for §212(c) relief when deportable “on a ground which does not have a statutory counterpart in section 212.” 8 CFR §1212.3(f)(5) (2010). 4 Careful readers may note that the example involving controlled substances offered in the last paragraph also involves an exclusion ground that sweeps more broadly than the deportation ground charged. The deportation ground requires “trafficking” in a controlled substance, whereas the exclusion ground includes all possession offenses as well. The BIA nonetheless held in Meza that the degree of overlap between the two grounds was sufficient to make the alien eligible for §212(c) relief. That holding reveals the broad discretion that the BIA currently exercises in deciding when two statutory grounds are comparable enough. 5 DHS also charged two other grounds for deportation, but the BIA did not rule on those grounds and they are not before us. 6 Compare Blake v. Carbone, 489 F. 3d 88, 103 (CA2 2007) (rejecting the BIA’s approach and holding instead that “[i]f the offense that renders [an alien] deportable would render a similarly situated [alien] excludable, the deportable [alien] is eligible for a waiver of deportation”), with Koussan v. Holder, 556 F. 3d 403, 412–414 (CA6 2009) (upholding the comparable-grounds policy); Caroleo v. Gonzales, 476 F. 3d 158, 162–163, 168 (CA3 2007) (same); Kim v. Gonzales, 468 F. 3d 58, 62–63 (CA1 2006) (same). 7 The Government urges us instead to analyze this case under the second step of the test we announced in Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., , to govern judicial review of an agency’s statutory interpretations. See Brief for Respondent 19. Were we to do so, our analysis would be the same, because under Chevron step two, we ask whether an agency interpretation is “ ‘arbitrary or capricious in substance.’ ” Mayo Foundation for Medical Ed. and Research v. United States, 562 U. S. ___ , ___ (2011) (slip op., at 7) (quoting Household Credit Services, Inc. v. Pfennig, ). But we think the more apt analytic framework in this case is standard “arbitrary [or] capricious” review under the APA. The BIA’s comparable-grounds policy, as articulated in In re Blake, 23 I. & N. Dec. 722 (2005) and In re Brieva-Perez, 23 I. & N. Dec. 766 (2005), is not an interpretation of any statutory language—nor could it be, given that §212(c) does not mention deportation cases, see infra, at 16–17, and n. 11. 8 Judulang also argues that the BIA is making an impermissible distinction between two groups of deportable aliens—those who have recently left and returned to the country and those who have not. According to Judulang, the BIA is treating the former as if they were seeking admission, while applying the “comparable grounds” approach only to the latter. See Reply Brief for Petitioner 16–18. That is the kind of distinction the Second Circuit held in Francis v. INS, 532 F. 2d 268 (1976), violated equal protection. See supra, at 3–4. But the Government contends that it is drawing no such line—that it is applying the comparable-grounds policy to all deportable aliens. Brief for Respondent 29. We think the available evidence tends to support the Government’s representation. See In re Meza-Castillo, No. A091 366 529, 2009 WL 455596 (BIA, Feb. 9, 2009) (applying comparable-grounds analysis to a deportable alien who had left and returned to the country); In re Valenzuela-Morales, No. A40 443 512, 2008 WL 2079382 (BIA, Apr. 23, 2008) (same). But in light of our holding that thecomparable-grounds approach is arbitrary and capricious, we need notresolve this dispute about the BIA’s practice. 9 The case would be different if Congress had intended for §212(c) relief to depend on the interaction of exclusion grounds and deportation grounds. But the Government has presented us with no evidence to this effect, nor have we found any. See Blake, 489 F. 3d, at 102 (Congress never contemplated, in drafting the immigration laws, “that its grounds of deportation would have any connection with the grounds of exclusion” in the application of §212(c)); see also infra, at 16–17. 10 Perhaps that is why the BIA declined to apply similar reasoning in Meza—a case also involving an exclusion ground that sweeps more broadly than a deportation ground (although not to the same extent as in Blake). See supra, at 6. 11 Congress amended §212(c), just five months before repealing it, to include a first-time reference to deportation cases. That amendment prohibited the Attorney General from granting discretionary relief to aliens deportable on several specified grounds. See Antiterrorism and Effective Death Penalty Act of 1996, (effective Apr. 24, 1996). The change does not affect our analysis, nor does the Government argue it should. As the Government notes, the amendment “did not speak to the viability of the Board’s” comparable-grounds rule, but instead made categorically ineligible for §212(c) relief “those deportable by reason of certain crimes.” Brief for Respondent 20. Presumably, Congress thought those crimes particularly incompatible with an alien’s continued residence in this country. 12 Because we find the BIA’s prior practice so unsettled, we likewise reject Judulang’s argument that Blake and Brieva-Perez were impermissibly retroactive. To succeed on that theory, Judulang would have to show, at a minimum, that in entering his guilty plea, he had reasonably relied on a legal rule from which Blake and Brieva-Perez departed. See Landgraf v. USI Film Products, (stating that retroactivity analysis focuses on “considerations of fair notice, reasonable reliance, and settled expectations”). The instability of the BIA’s prior practice prevents Judulang from making this showing:The BIA sometimes recognized aliens in Judulang’s position as eligible for §212(c) relief, but sometimes did not.

566.US.2011_10-1219

Under the Patent Act of 1952, if a Patent and Trade Office (PTO) examiner denies a patent application, 35 U. S. C. §131, the applicant may file an administrative appeal with the PTO’s Board of Patent Appeals and Interferences, §134. If the Board also denies the application, the applicant may appeal directly to the Court of Appeals for the Federal Circuit under §141. Alternatively, the applicant may file a civil action against the PTO Director under §145, which permits the applicant to present evidence that was not presented to the PTO. Respondent Hyatt filed a patent application covering multiple claims. The patent examiner denied all of the claims for lack of an adequate written description. Hyatt appealed to the Board, which approved some claims but denied others. Pursuant to §145, Hyatt filed a civil action against the Director, but the District Court declined to consider Hyatt’s newly proffered written declaration in support of the adequacy of his description, thus limiting its review to the administrative record. Applying the deferential “substantial evidence” standard of the Administrative Procedure Act (APA) to the PTO’s factual findings, the court granted summary judgment to the Director. On appeal, the Federal Circuit vacated the judgment, holding that patent applicants can introduce new evidence in §145 proceedings, subject only to the limitations in the Federal Rules of Evidence and the Federal Rules of Civil Procedure. It also reaffirmed its precedent that when new, conflicting evidence is introduced, the district court must make de novo findings to take such evidence into account. Held: There are no limitations on a patent applicant’s ability to introduce new evidence in a §145 proceeding beyond those already present in the Federal Rules of Evidence and the Federal Rules of Civil Procedure. If new evidence is presented on a disputed question of fact, the district court must make de novo factual findings that take account of both the new evidence and the administrative record before the PTO. Pp. 5−14. (a) Section 145, by its express terms, neither imposes unique evidentiary limits in district court proceedings nor establishes a heightened standard of review for PTO factual findings. Nonetheless, the Director contends that background principles of administrative law govern the admissibility of new evidence and impose a deferential standard of review in §145 proceedings. As the Director concedes, however, judicial review in §145 proceedings is not limited to the administrative record because the district court may consider new evidence. If it does so, the district court must act as a factfinder and cannot apply the APA’s deferential standard to PTO factual findings when those findings are contradicted by new evidence. Moreover, the doctrine of administrative exhaustion―the primary purpose of which is “the avoidance of premature interruption of the administrative process,” McKart v. United States, 395 U.S. 185, 193―does not apply because the PTO process is complete by the time a §145 proceeding occurs. Pp. 5−7. (b) The core language of the 1870 Patent Act, codified as Revised Statute §4915 (R. S. 4915), remains largely unchanged in §145. Decisions interpreting R. S. 4915 thus inform this Court’s understanding of §145. Both Butterworth v. United States ex rel. Hoe, 112 U.S. 50, and Morgan v. Daniels, 153 U.S. 120, describe the nature of R. S. 4915 proceedings, but the two opinions can be perceived as being in some tension. Butterworth described the proceeding as an original civil action seeking de novo adjudication of the merits of a patent application, while Morgan described it as a suit for judicial review of agency action under a deferential standard. The cases are distinguishable, however, because they addressed different circumstances. Butterworth discussed a patent applicant’s challenge to the denial of his application, whereas Morgan involved an interference proceeding that would now be governed by §146, not §145, and in which no new evidence was presented. Here, this Court is concerned only with a §145 proceeding in which new evidence was presented to the District Court, so Butterworth guides this Court’s decision. Thus, a district court conducting a §145 proceeding may consider all competent evidence adduced and is not limited to considering only new evidence that could not have been presented to the PTO. The introduction of new evidence in §145 proceedings is subject only to the Federal Rules of Evidence and the Federal Rules of Civil Procedure, and if new evidence is presented to the district court on a disputed factual question, de novo findings by the district court will be necessary for that new evidence to be taken into account along with the evidence before the Board. Pp. 7−13. (c) The district court may, however, consider whether the applicant had an opportunity to present the newly proffered evidence before the PTO in deciding what weight to afford that evidence. Pp. 13−14. 625 F.3d 1320, affirmed and remanded. Thomas, J., delivered the opinion for a unanimous Court. Soto- mayor, J., filed a concurring opinion, in which Breyer, J., joined.

The Patent Act of 1952, 35 U. S. C. §100 et seq., grants a patent applicant whose claims are denied by the Patent and Trademark Office (PTO) the opportunity to challenge the PTO’s decision by filing a civil action against the Director of the PTO in federal district court. In such a proceeding, the applicant may present evidence to the district court that he did not present to the PTO. This case requires us to consider two questions. First, we must decide whether there are any limitations on the applicant’s ability to introduce new evidence before the district court. For the reasons set forth below, we conclude that there are no evidentiary restrictions beyond those already imposed by the Federal Rules of Evidence and the Federal Rules of Civil Procedure. Second, we must determine what standard of review the district court should apply when considering new evidence. On this question, we hold that the district court must make a de novo finding when new evidence is presented on a disputed question of fact. In deciding what weight to afford that evidence, the district court may, however, consider whether the applicant had an opportunity to present the evidence to the PTO. I The Patent Act of 1952 establishes the process by which the PTO examines patent applications. A patent exam- iner first determines whether the application satisfies the statutory prerequisites for granting a patent. 35 U. S. C. §131. If the examiner denies the application, the applicant may file an administrative appeal with the PTO’s Board of Patent Appeals and Interferences (Board). §134. If the Board also denies the application, the Patent Act gives the disappointed applicant two options for judicial review of the Board’s decision. The applicant may either: (1) appeal the decision directly to the United States Court of Appeals for the Federal Circuit, pursuant to §141; or (2) file a civil action against the Director of the PTO in the United States District Court for the District of Columbia pursuant to §145.[1] In a §141 proceeding, the Federal Circuit must review the PTO’s decision on the same administrative record that was before the PTO. §144. Thus, there is no opportunity for the applicant to offer new evidence in such a proceeding. In Dickinson v. Zurko, 527 U.S. 150 (1999), we addressed the standard that governs the Federal Circuit’s review of the PTO’s factual findings. We held that the Administrative Procedure Act (APA), 5 U. S. C. §701 et seq., applies to §141 proceedings and that the Federal Circuit therefore should set aside the PTO’s factual findings only if they are “ ‘unsupported by substantial evidence.’ ” 527 U. S., at 152 (quoting 5 U. S. C. §706). In Zurko, we also noted that, unlike §141, §145 permits the applicant to present new evidence to the district court that was not presented to the PTO. 527 U. S., at 164. This opportunity to present new evidence is significant, not the least because the PTO generally does not accept oral testimony. See Brief for Petitioner 40, n. 11. We have not yet addressed, however, whether there are any limitations on the applicant’s ability to introduce new evidence in such a proceeding or the appropriate standard of review that a district court should apply when considering such evidence. II In 1995, respondent Gilbert Hyatt filed a patent application that, as amended, included 117 claims. The PTO’s patent examiner denied each claim for lack of an adequate written description. See 35 U. S. C. §112 (requiring pat- ent applications to include a “specification” that provides, among other information, a written description of the invention and of the manner and process of making and using it). Hyatt appealed the examiner’s decision to the Board, which eventually approved 38 claims, but denied the rest. Hyatt then filed a §145 action in Federal Dis- trict Court against the Director of the PTO (Director), peti- tioner here. To refute the Board’s conclusion that his patent application lacked an adequate written description, Hyatt submitted a written declaration to the District Court. In the declaration, Hyatt identified portions of the patent specification that, in his view, supported the claims that the Board held were not patentable. The District Court determined that it could not consider Hyatt’s declaration because applicants are “ ‘precluded from presenting new is- sues, at least in the absence of some reason of justice put forward for failure to present the issue to the Patent Office.’ ” Hyatt v. Dudas, Civ. Action No. 03–0901 (D DC, Sept. 30, 2005), p. 9, App. to Pet. for Cert. 182a (quoting DeSeversky v. Brenner, 424 F.2d 857, 858 (CADC 1970)). Because the excluded declaration was the only additional evidence submitted by Hyatt in the §145 proceeding, the evidence remaining before the District Court consisted entirely of the PTO’s administrative record. Therefore, the District Court reviewed all of the PTO’s factual findings under the APA’s deferential “substantial evidence” standard. See supra, at 2; see also Mazzari v. Rogan, 323 F.3d 1000, 1004–1005 (CA Fed. 2003). Applying that standard, the District Court granted summary judgment to the Director. Hyatt appealed to the Federal Circuit. A divided panel affirmed, holding that the APA imposed restrictions on the admission of new evidence in a §145 proceeding and that the district court’s review is not “wholly de novo.” Hyatt v. Doll, 576 F.3d 1246, 1269–1270 (2009). The Federal Circuit granted rehearing en banc and vacated the District Court’s grant of summary judgment. The en banc court first held “that Congress intended that applicants would be free to introduce new evidence in §145 proceedings subject only to the rules applicable to all civil actions, the Federal Rules of Evidence and the Federal Rules of Civil Procedure,” even if the applicant had no justification for failing to present the evidence to the PTO. 625 F.3d 1320, 1331 (2010). Reaffirming its precedent, the court also held that when new, conflicting evidence is introduced in a §145 proceeding, the district court must make de novo findings to take such evidence into account. Id., at 1336. We granted certiorari, 564 U. S. ___ (2011), and now affirm. III The Director challenges both aspects of the Federal Circuit’s decision. First, the Director argues that a district court should admit new evidence in a §145 action only if the proponent of the evidence had no reasonable opportunity to present it to the PTO in the first instance. Second, the Director contends that, when new evidence is introduced, the district court should overturn the PTO’s fac- tual findings only if the new evidence clearly establishes that the agency erred. Both of these arguments share the premise that §145 creates a special proceeding that is distinct from a typical civil suit filed in federal district court and that is thus governed by a different set of procedural rules. To support this interpretation of §145, the Director relies on background principles of administrative law and pre-existing practice under a patent statute that predated §145. For the reasons discussed below, we find that neither of these factors justifies a new evidentiary rule or a heightened standard of review for factual findings in §145 proceedings. A To address the Director’s challenges, we begin with the text of §145. See, e.g., Magwood v. Patterson, 561 U. S. ___, ___ (2010) (slip op., at 10). Section 145 grants a disappointed patent applicant a “remedy by civil action against the Director.” The section further explains that the district court “may adjudge that such applicant is entitled to receive a patent for his invention, as specified in any of his claims involved in the decision of the [PTO], as the facts in the case may appear and such adjudication shall authorize the Director to issue such patent on compliance with the requirements of law.” By its terms, §145 neither imposes unique evidentiary limits in district court proceedings nor establishes a heightened standard of re- view for factual findings by the PTO. B In the absence of express support for his position in the text of §145, the Director argues that the statute should be read in light of traditional principles of administrative law, which Congress codified in the APA. The Director notes that §145 requires a district court to review the reasoned decisionmaking of the PTO, an executive agency with specific authority and expertise. Accordingly, the Director contends that a district court should defer to the PTO’s factual findings. The Director further contends that, given the traditional rule that a party must exhaust his administrative remedies, a district court should consider new evidence only if the party did not have an opportunity to present it to the agency. We reject the Director’s contention that background principles of administrative law govern the admissibility of new evidence and require a deferential standard of review in a §145 proceeding. Under the APA, judicial review of an agency decision is typically limited to the administrative record. See 5 U. S. C. §706. But, as the Director concedes, §145 proceedings are not so limited, for the district court may consider new evidence. When the district court does so, it must act as a factfinder. Zurko, 527 U. S., at 164. In that role, it makes little sense for the district court to apply a deferential standard of review to PTO factual findings that are contradicted by the new evidence. The PTO, no matter how great its authority or expertise, cannot account for evidence that it has never seen. Consequently, the district court must make its own findings de novo and does not act as the “reviewing court” envisioned by the APA. See 5 U. S. C. §706. We also conclude that the principles of administrative exhaustion do not apply in a §145 proceeding. The Director argues that applicants must present all available evidence to the PTO to permit the PTO to develop the necessary facts and to give the PTO the opportunity to properly apply the Patent Act in the first instance. Brief for Petitioner 21–22 (citing McKart v. United States, 395 U.S. 185, 193–194 (1969)). But as this Court held in McKart, a primary purpose of administrative exhaustion “is, of course, the avoidance of premature interruption of the administrative process.” Id., at 193. That rationale does not apply here because, by the time a §145 proceeding occurs, the PTO’s process is complete. Section 145, moreover, does not provide for remand to the PTO to consider new evidence, and there is no pressing need for such a procedure because a district court, unlike a court of appeals, has the ability and the competence to receive new evidence and to act as a factfinder. In light of these aspects of §145 proceedings—at least in those cases in which new evidence is presented to the district court on a dis- puted question of fact—we are not persuaded by the Director’s suggestion that §145 proceedings are governed by the deferential principles of agency review. C Having concluded that neither the statutory text nor background principles of administrative law support an evidentiary limit or a heightened standard of review for factual findings in §145 proceedings, we turn to the evidentiary and procedural rules that were in effect when Congress enacted §145 in 1952. Although §145 is a relatively modern statute, the language in that provision originated in the Act of July 8, 1870 (1870 Act), ch. 230, 16Stat. 198, and the history of §145 proceedings can be traced back to the Act of July 4, 1836 (1836 Act), ch. 357, 5Stat. 117. Thus, we begin our inquiry with the 1836 Act, which established the Patent Office, the PTO’s predecessor, and first authorized judicial review of its decisions. 1 The 1836 Act provided that a patent applicant could bring a bill in equity in federal district court if his application was denied on the ground that it would interfere with another patent. Id., at 123–124; see also B. Shipman, Handbook of the Law of Equity Pleading §§101–103, pp. 168–171 (1897). Three years later, Congress expanded that provision, making judicial review available whenever a patent was refused on any ground. Act of Mar. 3, 1839 (1839 Act), 5Stat. 354. Pursuant to these statutes, any disappointed patent applicant could file a bill in equity to have the district court “adjudge” whether the applicant was “entitled, according to the principles and provisions of [the Patent Act], to have and receive a patent for his invention.” 1836 Act, 5Stat. 124. In 1870, Congress amended the Patent Act again, adding intermediate layers of administrative review and in- troducing language describing the proceeding in the district court. 16Stat. 198. Under the 1870 Act, an applicant denied a patent by the primary examiner could appeal first to a three-member board of examiners-in-chief, then to the Commissioner for Patents, and finally to an en banc sitting of the Supreme Court of the District of Columbia.[2] Id., at 205. Notably, Congress described that court’s review as an “appeal” based “on the evidence produced before the commissioner.” Ibid. The 1870 Act preserved the prior remedy of a bill in equity in district court for the applicant whose appeal was denied either by the Commissioner or by the Supreme Court of the District of Columbia. Ibid. The district court, in a proceeding that was distinct from the appeal considered on the administrative record by the Supreme Court of the District of Columbia, would “adjudge” whether the applicant was “entitled, according to law, to receive a patent for his invention . . . as the facts in the case may appear.” Ibid. In 1878, Congress codified this provision of the 1870 Act as Revised Statute §4915 (R. S. 4915). That statute was the immediate predecessor to §145, and its core language remains largely unchanged in §145. Accordingly, both parties agree that R. S. 4915 and the judicial decisions interpreting that statute should inform our understanding of §145. 2 This Court described the nature of R. S. 4915 proceedings in two different cases: Butterworth v. United States ex rel. Hoe, 112 U.S. 50 (1884), and Morgan v. Daniels, 153 U.S. 120 (1894). In Butterworth, the Court held that the Secretary of the Interior, the head of the federal department in which the Patent Office was a bureau, had no authority to review a decision made by the Commissioner of Patents in an interference proceeding. In its discussion, the Court described the remedy provided by R. S. 4915 as “a proceeding in a court of the United States having original equity jurisdiction under the patent laws, according to the ordinary course of equity practice and procedure. It is not a technical appeal from the Patent-Office, like that authorized [before the Supreme Court of the District of Columbia], confined to the case as made in the record of that office, but is prepared and heard upon all competent evidence adduced and upon the whole merits.” 112 U. S., at 61. The Butterworth Court also cited several lower court cases, which similarly described R. S. 4915 proceedings as “altogether independent” from the hearings before the Patent Office and made clear that the parties were “at liberty to introduce additional evidence” under “the rules and practice of a court of equity.” In re Squire, 22 F. Cas. 1015, 1016 (No. 13,269) (CC ED Mo. 1877); see also Whipple v. Miner, 15 F. 117, 118 (CC Mass. 1883) (describing the federal court’s jurisdiction in an R. S. 4915 proceeding as “an independent, original jurisdiction”); Butler v. Shaw, 21 F. 321, 327 (CC Mass. 1884) (holding that “the court may receive new evidence, and has the same powers as in other cases in equity”). Ten years later, in Morgan, this Court again confronted a case involving proceedings under R. S. 4915. 153 U.S. 120. There, a party challenged a factual finding by the Patent Office, but neither side presented additional evidence in the District Court. Id., at 122–123. This Court described the parties’ dispute as one over a question of fact that had already “been settled by a special tribunal [e]ntrusted with full power in the premises” and characterized the resulting District Court proceeding not as an independent civil action, but as “something in the nature of a suit to set aside a judgment.” Id., at 124. Consistent with that view, the Court held that the agency’s findings should not be overturned by “a mere preponderance of evidence.” Ibid. Viewing Butterworth and Morgan together, one might perceive some tension between the two cases. Butterworth appears to describe an R. S. 4915 proceeding as an original civil action, seeking de novo adjudication of the merits of a patent application. Morgan, on the other hand, appears to describe an R. S. 4915 proceeding as a suit for judicial review of agency action, governed by a deferential standard of review. To resolve that apparent tension, the Director urges us to disregard the language in Butterworth as mere dicta and to follow Morgan. He argues that Butterworth “shed[s] no light on the extent to which new evidence was admissible in R. S. 4915 proceedings or on the standard of review that applied in such suits.” Brief for Petitioner 33. The Director maintains that Morgan, in contrast, firmly established that a district court in such a proceeding performs a deferential form of review, governed by traditional principles of administrative law. We reject the Director’s position.[3] We think that the differences between Butterworth and Morgan are best explained by the fact that the two cases addressed different circumstances. Butterworth discussed the character of an R. S. 4915 proceeding in which a disappointed patent applicant challenged the Board’s denial of his application. Although that discussion was not strictly necessary to Butterworth’s holding it was also not the kind of ill-considered dicta that we are inclined to ignore. The Butterworth Court carefully examined the various pro- visions providing relief from the final denial of a patent application by the Commissioner of Patents to determine that the Secretary of the Interior had no role to play in that process. 112 U. S., at 59–64. The Court further surveyed the decisions of the lower courts with regard to the nature of an R. S. 4915 proceeding and concluded that its view was “the uniform and correct practice in the Circuit Courts.” Id., at 61. We note that this Court reiter- ated Butterworth’s well-reasoned interpretation of R. S. 4915 in three later cases.[4] Morgan, on the other hand, concerned a different situation from the one presented in this case. First, Morgan addressed an interference proceeding. See 153 U. S., at 125 (emphasizing that “the question decided in the Pat- ent Office is one between contesting parties as to priority of invention”). Although interference proceedings were pre- viously governed by R. S. 4915, they are now governed by a separate section of the Patent Act, 35 U. S. C. §146, and therefore do not implicate §145. In addition, Morgan did not involve a proceeding in which new evidence was presented to the District Court. See 153 U. S., at 122 (stating that the case “was submitted, without any additional testimony, to the Circuit Court”). 3 Because in this case we are concerned only with §145 proceedings in which new evidence has been presented to the District Court, Butterworth rather than Morgan guides our decision. In Butterworth, this Court observed that an R. S. 4915 proceeding should be conducted “according to the ordinary course of equity practice and procedure” and that it should be “prepared and heard upon all competent evidence adduced and upon the whole merits.” 112 U. S., at 61. Likewise, we conclude that a district court conducting a §145 proceeding may consider “all com- petent evidence adduced,” id., at 61, and is not limited to considering only new evidence that could not have been presented to the PTO. Thus, we agree with the Federal Circuit that “Congress intended that applicants would be free to introduce new evidence in §145 proceedings subject only to the rules applicable to all civil actions, the Federal Rules of Evidence and the Federal Rules of Civil Procedure.” 625 F. 3d, at 1331. We also agree with the Federal Circuit’s longstanding view that, “where new evidence is presented to the district court on a disputed fact question, a de novo finding will be necessary to take such evidence into account together with the evidence before the board.” Fregeau v. Mossinghoff, 776 F.2d 1034, 1038 (1985). As we noted in Zurko, the district court acts as a factfinder when new evidence is introduced in a §145 proceeding. 527 U. S., at 164. The district court must assess the credibility of new witnesses and other evidence, determine how the new evidence comports with the existing administrative record, and decide what weight the new evidence deserves. As a logical matter, the district court can only make these determinations de novo because it is the first tribunal to hear the evidence in question. Furthermore, a de novo standard adheres to this Court’s instruction in Butterworth that an R. S. 4915 proceeding be heard “upon the whole merits” and conducted “according to the ordinary course of equity practice and procedure.” 112 U. S., at 61. D Although we reject the Director’s proposal for a stricter evidentiary rule and an elevated standard of review in §145 proceedings, we agree with the Federal Circuit that the district court may, in its discretion, “consider the proceedings before and findings of the Patent Office in deciding what weight to afford an applicant’s newly-admitted evidence.” 625 F. 3d, at 1335. Though the PTO has special expertise in evaluating patent applications, the district court cannot meaningfully defer to the PTO’s factual findings if the PTO considered a different set of facts. Supra, at 8; cf. Microsoft Corp. v. i4i Ltd. Partnership, 564 U. S. ___, ___ (2011) (slip op., at 19) (noting that “if the PTO did not have all material facts before it, its considered judgment may lose significant force”). For this reason, we conclude that the proper means for the district court to accord respect to decisions of the PTO is through the court’s broad discretion over the weight to be given to evidence newly adduced in the §145 proceedings. The Director warns that allowing the district court to consider all admissible evidence and to make de novo findings will encourage patent applicants to withhold evidence from the PTO intentionally with the goal of pre- senting that evidence for the first time to a nonexpert judge. Brief for Petitioner 23. We find that scenario unlikely. An applicant who pursues such a strategy would be intentionally undermining his claims before the PTO on the speculative chance that he will gain some advantage in the §145 proceeding by presenting new evidence to a district court judge. IV For these reasons, we conclude that there are no limitations on a patent applicant’s ability to introduce new evidence in a §145 proceeding beyond those already present in the Federal Rules of Evidence and the Federal Rules of Civil Procedure. Moreover, if new evidence is presented on a disputed question of fact, the district court must make de novo factual findings that take account of both the new evidence and the administrative record before the PTO. In light of these conclusions, the Federal Circuit was correct to vacate the judgment of the District Court, which excluded newly presented evidence under the view that it “need not consider evidence negligently submitted after the end of administrative proceedings.” Civ. Action No. 03–0901, at 15, App. to Pet. for Cert. 189a. The judgment is affirmed, and the case is remanded to the Court of Appeals for further proceedings consistent with this opinion. It is so ordered. Notes 1 On September 16, 2011, the President signed the Leahy-Smith America Invents Act, 125Stat. 284, into law. That Act made significant changes to Title 35 of the United States Code, some of which are related to the subject matter of this case. For example, the Act changed the venue for §145 actions from the United States District Court for the District of Columbia to the United States District Court for the East-ern District of Virginia, id., at 316, changed the name of the Boardof Patent Appeals and Interferences to the Patent Trial and Appeal Board, id., at 290, and changed the name of interferences to derivation proceedings, ibid. Neither party contends that the Act has any effect on the questions before us, and all references and citations in this opinion are to the law as it existed prior to the Act. 2 The Supreme Court of the District of Columbia was a trial court created by Congress in 1863. Act of Mar. 3, 1863, ch. 91, 12Stat. 762. Although the court was generally one of first instance, it also functioned as an appellate court when it sat en banc. Voorhees, The District of Columbia Courts: A Judicial Anomaly, 29 Cath. U. L. Rev. 917, 923 (1980). 3 Both parties cite additional cases from the lower courts that they claim support their view of the statute, but these cases are too diverse to support any firm inferences about Congress’ likely intent in enacting §145. 4 In Gandy v. Marble, 122 U.S. 432 (1887), the Court described an R. S. 4915 proceeding as “a suit according to the ordinary course of equity practice and procedure” rather than a “technical appeal fromthe Patent Office.” Id., at 439 (citing Butterworth, 112 U. S., at 61). Likewise, in In re Hien, 166 U.S. 432 (1897), the Court distinguished an R. S. 4915 proceeding from the “ ‘technical appeal from the Patent Office’ ” authorized under R. S. 4911, the predecessor to current §141. Id., at 439 (quoting Butterworth, supra, at 61). And, finally, in Hoover Co. v. Coe, 325 U.S. 79 (1945), the Court cited Butterworth to support its description of an R. S. 4915 proceeding as a “formal trial.” 325 U. S., at 83, and n. 4.

565.US.2011_10-577

An Immigration Judge ordered the removal of resident aliens Akio and Fusako Kawashima, determining that Mr. Kawashima’s conviction for willfully making and subscribing a false tax return, 26 U. S. C. §7206(1), and Mrs. Kawashima’s conviction for aiding and assisting in the preparation of a false tax return, §7206(2), qualified as crimes involving fraud or deceit under 8 U. S. C. §1101(a)(43)(M)(i) (Clause (i)) and thus were aggravated felonies for which they could be deported under §1227(a)(2)(A)(iii). The Board of Immigration Appeals affirmed. Holding that convictions under 26 U. S. C. §§7206(1) and (2) in which the Government’s revenue loss exceeds $10,000 constitute aggravated felonies under Clause (i), the Ninth Circuit affirmed, but remanded for the Board to determine whether Mrs. Kawashima’s conviction had caused a Government loss in excess of $10,000. Held: Convictions under 26 U. S. C. §§7206(1) and (2) in which the Government’s revenue loss exceeds $10,000 qualify as aggravated felonies pursuant to Clause (i). Pp. 3−11. (a) The Kawashimas’ argument that they cannot be deported for the commission of an “aggravated felony” because crimes under §§7206(1) and (2) do not involve the fraud or deceit required by Clause (i) is rejected. This Court looks to the statute defining the crime of conviction, rather than the specific facts underlying the crime, see Gonzales v. Duenas-Alvarez, 549 U.S. 183, 186, to determine whether the Kawashimas’ offenses involve fraud or deceit within the meaning of Clause (i). Section 7206(1) provides that any person who “willfully makes and subscribes any return . . . which contains or is verified by a written declaration that it is made under the penalties of perjury, and which he does not believe to be true and correct as to every material matter,” shall be guilty of a felony. Although the words “fraud” and “deceit” are absent from §7206(1) and are not themselves formal elements of the crime, it does not follow that Mr. Kawashima’s offense falls outside Clause (i). Clause (i) is not limited to offenses that include fraud or deceit as formal elements. Rather, it refers more broadly to offenses involving fraud or deceit―meaning offenses with elements that necessarily entail fraudulent or deceitful conduct. Mr. Kawashima’s conviction under §7206(1) involved deceitful conduct in that he knowingly and willfully submitted a tax return that was false as to a material matter. Mrs. Kawashima was convicted of violating §7206(2), which declares that any person who “[w]illfully aids or assists in . . . the preparation or presentation . . . of a return . . . which is fraudulent or is false as to any material matter” has committed a felony. She committed a felony involving deceit by knowingly and willfully assisting her husband’s filing of a materially false tax return. Pp. 3−6. (b) The Kawashimas’ argument that Clause (i), when considered in light of 8 U. S. C. §1101(a)(43)(M)(ii) (Clause (ii)), must be interpreted as being inapplicable to tax crimes is also rejected. Clause (i) defines “aggravated felony” to mean an offense that “involves fraud or deceit in which the loss to the victim or victims exceeds $10,000.” Clause (ii) defines “aggravated felony” as an offense that is “described in section 7201 of title 26 (relating to tax evasion) in which the revenue loss to the Government exceeds $10,000.” Contrary to the Kawashimas’ claim, the difference in the clauses’ language—“revenue loss to the Government” in Clause (ii) compared to “loss to the victim” in Clause (i)—does not establish Congress’ intent to remove tax crimes from the scope of Clause (i). By its plain language, Clause (i) covers a broad class of offenses that involve fraud or deceit, and Congress’ decision to tailor Clause (ii)’s language to match the sole type of offense it covers does not demonstrate that Congress intended to implicitly circumscribe Clause (i)’s broad scope. Furthermore, interpreting Clause (i) to include tax crimes does not violate the presumption against superfluities. The specific inclusion of tax evasion in Clause (ii) does not make it redundant to Clause (i) because the inclusion was intended to ensure that tax evasion pursuant to 26 U. S. C. §7201 was a deportable offense. Pp. 6−10. (c) The United States Sentencing Guidelines’ separate treatment of tax crimes and crimes involving fraud and deceit does not support the Kawashimas’ contention that Congress did not intend to include tax crimes within Clause (i). No evidence suggests that Congress considered the Guidelines when drafting 8 U. S. C. §1101(a)(43)(M). Moreover, the differences between §1101(a)(43)(M) and the Guidelines undercut any inference that Congress was incorporating the distinction drawn by the Guidelines into §1101(a)(43)(M). Pp. 10−11. (d) Construing §1101(a)(43)(M) in the Kawashimas’ favor under the rule of lenity is not warranted in light of the statute’s clear application. P. 11. 615 F.3d 1043, affirmed. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Alito, and Sotomayor, JJ., joined. Ginsburg, J., filed a dissenting opinion, in which Breyer and Kagan, JJ., joined.

This case concerns whether aliens who commit certain federal tax crimes are subject to deportation as aliens who have been convicted of an aggravated felony. We hold that violations of 26 U. S. C. §§7206(1) and (2) are crimes “involv[ing] fraud or deceit” under 8 U. S. C. §1101(a)(43)(M)(i) and are therefore aggravated felonies as that term is defined in the Immigration and Nationality Act, 8 U. S. C. §1101 et seq., when the loss to the Government exceeds $10,000. I Petitioners, Akio and Fusako Kawashima, are natives and citizens of Japan who have been lawful permanent residents of the United States since June 21, 1984. In 1997, Mr. Kawashima pleaded guilty to one count of willfully making and subscribing a false tax return in violation of 26 U. S. C. §7206(1). Mrs. Kawashima pleaded guilty to one count of aiding and assisting in the prep- aration of a false tax return in violation of 26 U. S. C. §7206(2). Following their convictions, the Immigration and Naturalization Service charged the Kawashimas with being deportable from the United States as aliens who had been convicted of an aggravated felony.[1] See 8 U. S. C. §1227(a)(2)(A)(iii) (“Any alien who is convicted of an aggravated felony at any time after admission is deportable”).[2] In the Immigration and Nationality Act, Congress listed categories of offenses that qualify as “aggravated felonies” for the purpose of deportation. See §1101(a)(43). Here, the Government charged the Kawashimas with be- ing deportable for committing offenses under subparagraph (M) of §1101(a)(43). That subparagraph classifies as an aggravated felony an offense that either: “(i) involves fraud or deceit in which the loss to the victim or vic- tims exceeds $10,000; or (ii) is described in section 7201 of title 26 (relating to tax evasion) in which the revenue loss to the Government exceeds $10,000.” Hereinafter, we refer to §1101(a)(43)(M)(i) as “Clause (i)” and to §1101(a)(43)(M)(ii) as “Clause (ii).” At their deportation hearing, the Kawashimas argued that their convictions under 26 U. S. C. §7206 did not qualify as aggravated felonies under subparagraph (M). The Immigration Judge disagreed and ordered removal, concluding that the Kawashimas’ convictions qualified as aggravated felonies under Clause (i). The Kawashimas appealed the removal order to the Board of Immigration Appeals (Board), which affirmed the Immigration Judge’s decision. After unsuccessfully petitioning the Board to reopen its decision, the Kawashimas filed petitions for review of the Board’s decision in the United States Court of Appeals for the Ninth Circuit. The Ninth Circuit held that “convictions for violating §§7206(1) and (2) in which the tax loss to the Government exceeds $10,000 constitute aggravated felonies under subsection (M)(i).” metricconverter615 F.3d 1043, 1053 (2010). The court concluded that Mr. Kawashima’s conviction un- der §7206(1) qualified as an aggravated felony within Clause (i)’s definition “because it involved ‘fraud or deceit’ and because his offense resulted in a loss to the government in excess of $10,000.” Id., at 1055. The Ninth Circuit also determined that Mrs. Kawashima’s conviction under §7206(2) “necessarily ‘involve[d] fraud or deceit.’ ” Id., at 1055. But because Mrs. Kawashima’s plea agreement was not in the administrative record, the Ninth Circuit remanded to the Board to determine whether Mrs. Kawashima’s conviction had caused a loss to the Government in excess of $10,000. Id., at 1056–1057. We granted the Kawashimas’ petition for a writ of certiorari to determine whether their convictions for violations of 26 U. S. C. §§7206(1) and (2) respectively qualify as aggravated felonies under 8 U. S. C. §1101(a)(43)(M)(i). 563 U. S. ___ (2011). We now affirm. II The Kawashimas argue that they cannot be deported for commission of an “aggravated felony” because crimes under §§7206(1) and (2) do not “involv[e] fraud or deceit” as required by Clause (i). The Kawashimas also assert that their convictions under §7206 are not “aggravated felonies” because tax crimes are not included within Clause (i) at all. We address each argument in turn. A The Kawashimas contend that their offenses of conviction do not fall within the scope of Clause (i) because neither “fraud” nor “deceit” is a formal element of a con- viction under §7206(1) or §7206(2). The Government responds that the Kawashimas’ convictions necessarily in- volved deceit because they required a showing that the Kawashimas willfully made materially false statements. To determine whether the Kawashimas’ offenses “involv[e] fraud or deceit” within the meaning of Clause (i), we employ a categorical approach by looking to the statute defining the crime of conviction, rather than to the specific facts underlying the crime. See Gonzales v. Duenas-Alvarez, 549 U.S. 183, 186 (2007) (applying the approach set forth in Taylor v. United States, 495 U.S. 575, 599–600 (1990)). If the elements of the offenses establish that the Kawashimas committed crimes involving fraud or deceit, then the first requirement of Clause (i) is satisfied.[3] Mr. Kawashima was convicted of violating 26 U. S. C. §7206(1), which provides that any person who “[w]illfully makes and subscribes any return, statement, or other document, which contains or is verified by a written declaration that it is made under the penalties of perjury, and which he does not believe to be true and correct as to every material matter,” shall be guilty of a felony. Mr. Ka- washima does not dispute that the elements of a violation of §7206(1) include, inter alia, that the document in question was false as to a material matter, that the defendant did not believe the document to be true and correct as to every material matter, and that he acted willfully with the specific intent to violate the law. See, e.g., United States v. Aramony, 88 F.3d 1369, 1382 (CA4 1996); United States v. Kaiser, 893 F.2d 1300, 1305 (CA11 1990); United States v. Marabelles, 724 F.2d 1374, 1380 (CA9 1984); United States v. Whyte, 699 F.2d 375, 381 (CA7 1983). Although the words “fraud” and “deceit” are absent from the text of §7206(1) and are not themselves formal elements of the crime, it does not follow that his offense falls outside of Clause (i). The scope of that clause is not limited to offenses that include fraud or deceit as formal elements. Rather, Clause (i) refers more broadly to offenses that “involv[e]” fraud or deceit—meaning offenses with elements that necessarily entail fraudulent or deceitful conduct. When subparagraph (M) was enacted, the term “deceit” meant a “the act or process of deceiving (as by falsification, concealment, or cheating).” Webster’s Third New International Dictionary 584 (1993). Mr. Kawashima’s conviction under §7206(1) establishes that he knowingly and willfully submitted a tax return that was false as to a material matter. He therefore committed a felony that involved “deceit.” Turning to Mrs. Kawashima, our analysis follows a similar path. Mrs. Kawashima was convicted of violating 26 U. S. C. §7206(2), which declares that any person who “[w]illfully aids or assists in . . . the preparation or presentation under, or in connection with any matter arising under, the internal revenue laws, of a return, affidavit, claim, or other document, which is fraudulent or is false as to any material matter,” has committed a felony. Mrs. Kawashima does not dispute that the elements of a violation of §7206(2) include, inter alia, that the document in question was false as to a material matter and that the defendant acted willfully. See Aramony, supra, at 1382; United States v. Sassak, 881 F.2d 276, 278 (CA6 1989); United States v. Hooks, 848 F.2d 785, 788–789 (CA7 1988); United States v. Dahlstrom, metricconverter713 F.2d 1423, 1426–1427 (CA9 1983). We conclude that Mrs. Kawashima’s conviction establishes that, by knowingly and willfully assisting her husband’s filing of a materially false tax return, Mrs. Kawashima also committed a felony that involved “deceit.” The language of Clause (i) is clear. Anyone who is convicted of an offense that “involves fraud or deceit in which the loss to the victim or victims exceeds $10,000” has committed an aggravated felony and is subject to depor- tation pursuant to 8 U. S. C. §1227(a)(2)(A)(iii). The elements of willfully making and subscribing a false corporate tax return, in violation of 26 U. S. C. §7206(1), and of aiding and assisting in the preparation of a false tax return, in violation of 26 U. S. C. §7206(2), establish that those crimes are deportable offenses because they necessarily entail deceit. B The Kawashimas’ second argument is based on inferences drawn from the interaction of Clause (i) and Clause (ii). The full text of subparagraph (M) reads as follows: “(43) The term ‘aggravated felony’ means— . . . . . “(M) an offense that— “(i) involves fraud or deceit in which the loss to the victim or victims exceeds $10,000; or “(ii) is described in section 7201 of title 26 (relating to tax evasion) in which the revenue loss to the Government exceeds $10,000.” The Kawashimas argue that when Clause (i) is read together with Clause (ii), Clause (i) must be interpreted as being inapplicable to tax crimes. In their view, subparagraph (M), when considered in its entirety, demonstrates that Congress was addressing two mutually exclusive categories of crimes in subparagraph (M)’s two clauses: general, non-tax crimes involving fraud or deceit that cause actual losses to real victims in Clause (i), and tax crimes involving revenue losses to the Government in Clause (ii). For the reasons discussed below, this argument cannot overcome the plain language of Clause (i), which encompasses the Kawashimas’ offenses of conviction. 1 The Kawashimas contend that textual differences between Clauses (i) and (ii) indicate that Congress intended to exclude tax crimes from Clause (i). Specifically, they note that Clause (i) addresses “loss to the victim,” whereas Clause (ii) addresses “revenue loss to the Government.” This difference in language does not establish Congress’ intent to remove tax crimes from the scope of Clause (i). Clause (i) covers a broad class of offenses that involve fraud or deceit. Clause (i) thus uses correspondingly broad language to refer to the wide range of potential losses and victims. Clause (ii), on the other hand, is limited to the single type of offense “described in section 7201 of title 26 (relating to tax evasion),” which, by definition, can only cause one type of loss (revenue loss) to one type of victim (the Government). Congress’ decision to tailor Clause (ii)’s language to match the sole type of offense covered by Clause (ii) does not demonstrate that Congress also intended to implicitly circumscribe the broad scope of Clause (i)’s plain language. 2 Next, the Kawashimas argue that interpreting Clause (i) to include tax crimes violates the presumption against superfluities by rendering Clause (ii) completely redundant to Clause (i). Clause (ii) explicitly states that convictions for tax evasion pursuant to 26 U. S. C. §7201 that cause a revenue loss of at least $10,000 to the Government are aggravated felonies. The Kawashimas assert that, if Clause (i) applies to tax crimes, then qualifying convictions for tax evasion under Clause (ii) would also qualify as aggravated felonies under Clause (i), because tax evasion is a crime involving fraud or deceit. To buttress this argument, the Kawashimas point to a body of law providing that a conviction for tax evasion under §7201 collaterally estops the convicted taxpayer from contesting a civil penalty under 26 U. S. C. §6663(b) for “underpayment . . . attributable to fraud.” See, e.g., Gray v. Commissioner, 708 F.2d 243, 246 (CA6 1983) (“Numerous federal courts have held that a conviction for federal income tax evasion, either upon a plea of guilty, or upon a jury verdict of guilt, conclusively establishes fraud in a subsequent civil tax fraud proceeding through application of the doctrine of collateral estoppel”). Therefore, according to the Kawashimas, if Clause (i) covers tax offenses, then Clause (ii) is mere surplusage. We disagree with the Kawashimas’ contention that the specific mention of one type of tax crime in Clause (ii) impliedly limits the scope of Clause (i)’s plain language, which extends to any offense that “involves fraud or deceit.” We think it more likely that Congress specifically included tax evasion offenses under 26 U. S. C. §7201 in Clause (ii) to remove any doubt that tax evasion qualifies as an aggravated felony. Several considerations support this conclusion. Like §§7206(1) and (2), §7201 does not, on its face, mention fraud or deceit. Instead, §7201 simply provides that “[a]ny person who willfully attempts in any manner to evade or defeat any tax imposed by [the Internal Revenue Code] or the payment thereof shall, in addition to other penalties provided by law, be guilty of a felony.” Accordingly, neither fraud nor deceit is among the elements of a conviction under §7201, which include: (1) willfulness; (2) the existence of a tax deficiency; and (3) an affirmative act constituting an evasion or an attempted evasion of the tax. Boulware v. United States, 552 U.S. 421, 424, n. 2 (2008). A conviction under §7201, therefore, only qualifies as an aggravated felony under Clause (i) if a willful, affirmative attempt to evade a tax necessarily entails fraud or deceit. This Court’s decision in United States v. Scharton, 285 U.S. 518 (1932), gave Congress good reason to doubt that a conviction under §7201 satisfies that condition. In Scharton, the defendant was indicted for attempting to evade income taxes by falsely understating his taxable income. The question before the Court was whether the crime was subject to the 3-year statute of limitations generally applicable to tax crimes, or whether it was instead subject to the 6-year statute of limitations applicable to “ ‘offenses involving the defrauding or attempting to defraud the United States or any agency thereof, whether by conspiracy or not, and in any manner.’ ” Id., at 520, n. 2 (quoting 18 U. S. C. §585 (1962 ed., Supp. V)). The Government argued that the 6-year statute of limitations applied because “fraud is implicit in the concept of evading or defeating” and because any effort to evade a tax is tantamount to an attempt to defraud the taxing body. 285 U. S., at 520–521. The Court rejected that argument, noting that, in an indictment for evasion, “an averment [of intent to defraud] would be surplusage, for it would be sufficient to plead and prove a willful attempt to evade or defeat.” Id., at 521. Moreover, §7201 includes two offenses: “the offense of willfully attempting to evade or defeat the assessment of a tax as well as the offense of willfully attempting to evade or defeat the payment of a tax.” Sansone v. United States, 380 U.S. 343, 354 (1965) (emphasis in original). As the Government notes, it is possible to willfully evade or de- feat payment of a tax under §7201 without making any misrepresentation. For example, §7201 can be violated by a taxpayer who files a truthful tax return, but who also takes affirmative steps to evade payment by moving his assets beyond the reach of the Internal Revenue Service. Although the Government concedes that evasion-of-payment cases will almost invariably involve some affirmative acts of fraud or deceit, it is still true that the elements of tax evasion pursuant to §7201 do not necessarily involve fraud or deceit. Thus, we conclude that the specific inclusion of tax evasion in Clause (ii) was intended to ensure that tax evasion pursuant to §7201 was a deportable offense. Clause (ii) does not implicitly remove all other tax offenses from the scope of Clause (i)’s plain language. 3 The Kawashimas also assert that the separate treatment of tax crimes and crimes involving fraud and deceit in the United State Sentencing Guidelines supports their contention that Congress did not intend to include tax crimes within Clause (i). They point to the fact that, in 1987, the United States Sentencing Commission included within the Guidelines a category of “offenses involving fraud or deceit.” USSG §§2F1.1 to 2F1.2 (deleted effec- tive Nov. 1, 2001). The Commission simultaneously included “offenses involving taxation” as a separate category. §§2T1.1 et seq. (Nov. 2011). Although the Kawashimas acknowledge that they have found no evidence that Congress actually considered the Guidelines, they contend that “it is likely that the language of [Clause (i)] and [Clause (ii)] was taken from the Sentencing Guidelines” by the sponsors of the bill that expanded the definition of aggravated felony to include subparagraph (M). Brief for Petitioners 29. Therefore, the theory goes, we can infer from the similar language in the Guidelines that Congress did not intend Clause (i) to include tax crimes. We reject the Kawashimas’ reliance on the Guidelines. The Kawashimas’ argument is at odds with the fact that, unlike the Guideline that the Kawashimas cite, Clause (ii) does not refer to all offenses “involving taxation.” Rather, Clause (ii) is expressly limited to tax evasion offenses under §7201. That textual difference undercuts any inference that Congress was considering, much less incorporating, the distinction drawn by the Guidelines. C Finally, the Kawashimas argue that subparagraph (M)’s treatment of tax crimes other than tax evasion is ambiguous, and that we should therefore construe the statute in their favor. It is true that we have, in the past, construed ambiguities in deportation statutes in the alien’s favor. See INS v. St. Cyr, 533 U.S. 289, 320 (2001). We think the application of the present statute clear enough that resort to the rule of lenity is not warranted. * * * For the foregoing reasons, we conclude that convictions under 26 U. S. C. §§7206(1) and (2) in which the revenue loss to the Government exceeds $10,000 qualify as aggravated felonies pursuant to 8 U. S. C. §1101(a)(43)(M)(i). Because the Kawashimas are subject to deportation as aliens who have been convicted of aggravated felonies pursuant to 8 U. S. C. §1227(a)(2)(A)(iii), the judgment of the Court of Appeals is affirmed. It is so ordered. Notes 1 On March 1, 2003, most of the functions of the Immigration and Naturalization Service were transferred to the Bureau of Immigration and Customs Enforcement, and the Immigration and Naturalization Service ceased to exist. 2 Before 1996, there were two procedures for removing aliens from the country: “deportation” of aliens who were already present, and “exclusion” of aliens seeking entry or reentry into the country. Since 1996, the Government has used a unified procedure, known as “removal,” for both exclusion and deportation. See 8 U. S. C. §§1229, 1229a. We use the terms “deportation” and “removal” interchangeably in this opinion. 3 We note that the issue whether the Kawashimas’ offenses satisfy the second requirement of Clause (i)—that the loss to the victim exceeded $10,000—is not before us. We address only whether their offenses of conviction qualify as crimes “involv[ing] fraud or deceit.”

569.US.108

Petitioners, Nigerian nationals residing in the United States, filed suit in federal court under the Alien Tort Statute, alleging that respondents—certain Dutch, British, and Nigerian corporations—aided and abetted the Nigerian Government in committing violations of the law of nations in Nigeria. The ATS provides that “[t]he district courts shall have original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States.” 28 U. S. C. §1350. The District Court dismissed several of petitioners’ claims, but on interlocutory appeal, the Second Circuit dismissed the entire complaint, reasoning that the law of nations does not recognize corporate liability. This Court granted certiorari, and ordered supplemental briefing on whether and under what circumstances courts may recognize a cause of action under the ATS, for violations of the law of nations occurring within the territory of a sovereign other than the United States. Held: The presumption against extraterritoriality applies to claims under the ATS, and nothing in the statute rebuts that presumption. Pp. 3–14. (a) Passed as part of the Judiciary Act of 1789, the ATS is a jurisdictional statute that creates no causes of action. It permits federal courts to “recognize private claims [for a modest number of international law violations] under federal common law.” Sosa v. Alvarez-Machain, 542 U.S. 692, 732. In contending that a claim under the ATS does not reach conduct occurring in a foreign sovereign’s territory, respondents rely on the presumption against extraterritorial application, which provides that “[w]hen a statute gives no clear indication of an extraterritorial application, it has none,” Morrison v. National Australia Bank Ltd., 561 U. S. ___, ___. The presumption “serves to protect against unintended clashes between our laws and those of other nations which could result in international discord.” EEOC v. Arabian American Oil Co., 499 U.S. 244, 248. It is typically applied to discern whether an Act of Congress regulating conduct applies abroad, see, e.g., id., at 246, but its underlying principles similarly constrain courts when considering causes of action that may be brought under the ATS. Indeed, the danger of unwarranted judicial interference in the conduct of foreign policy is magnified in this context, where the question is not what Congress has done but what courts may do. These foreign policy concerns are not diminished by the fact that Sosa limited federal courts to recognizing causes of action only for alleged violations of international law norms that are “ ‘specific, universal, and obligatory,” 542 U. S., at 732. Pp. 3–6. (b) The presumption is not rebutted by the text, history, or purposes of the ATS. Nothing in the ATS’s text evinces a clear indication of extraterritorial reach. Violations of the law of nations affecting aliens can occur either within or outside the United States. And generic terms, like “any” in the phrase “any civil action,” do not rebut the presumption against extraterritoriality. See, e.g., Morrison, supra, at ___. Petitioners also rely on the common-law “transitory torts” doctrine, but that doctrine is inapposite here; as the Court has explained, “the only justification for allowing a party to recover when the cause of action arose in another civilized jurisdiction is a well-founded belief that it was a cause of action in that place,” Cuba R. Co. v. Crosby, 222 U.S. 473, 479. The question under Sosa is not whether a federal court has jurisdiction to entertain a cause of action provided by foreign or even international law. The question is instead whether the court has authority to recognize a cause of action under U. S. law to enforce a norm of international law. That question is not answered by the mere fact that the ATS mentions torts. The historical background against which the ATS was enacted also does not overcome the presumption. When the ATS was passed, “three principal offenses against the law of nations” had been identified by Blackstone: violation of safe conducts, infringement of the rights of ambassadors, and piracy. Sosa, supra, at 723, 724. Prominent contemporary examples of the first two offenses—immediately before and after passage of the ATS—provide no support for the proposition that Congress expected causes of action to be brought under the statute for violations of the law of nations occurring abroad. And although the offense of piracy normally occurs on the high seas, beyond the territorial jurisdiction of the United States or any other country, applying U. S. law to pirates does not typically impose the sovereign will of the United States onto conduct occurring within the territorial jurisdiction of another sovereign, and therefore carries less direct foreign policy consequences. A 1795 opinion of Attorney General William Bradford regarding the conduct of U. S. citizens on both the high seas and a foreign shore is at best ambiguous about the ATS’s extraterritorial application; it does not suffice to counter the weighty concerns underlying the presumption against extraterritoriality. Finally, there is no indication that the ATS was passed to make the United States a uniquely hospitable forum for the enforcement of international norms. Pp. 6–14. 621 F.3d 111, affirmed. Roberts, C. J., delivered the opinion of the Court, in which Scalia, Kennedy, Thomas, and Alito, JJ., joined. Kennedy, J., filed a concurring opinion. Alito, J., filed a concurring opinion, in which Thomas, J., joined. Breyer, J., filed an opinion concurring in the judgment, in which Ginsburg, Sotomayor, and Kagan, JJ., joined.

Petitioners, a group of Nigerian nationals residing in the United States, filed suit in federal court against certain Dutch, British, and Nigerian corporations. Petitioners sued under the Alien Tort Statute, 28 U. S. C. §1350, alleging that the corporations aided and abetted the Nigerian Government in committing violations of the law of nations in Nigeria. The question presented is whether and under what circumstances courts may recognize a cause of action under the Alien Tort Statute, for violations of the law of nations occurring within the territory of a sovereign other than the United States. I Petitioners were residents of Ogoniland, an area of 250 square miles located in the Niger delta area of Nigeria and populated by roughly half a million people. When the complaint was filed, respondents Royal Dutch Petroleum Company and Shell Transport and Trading Company, p.l.c., were holding companies incorporated in the Netherlands and England, respectively. Their joint subsidiary, respondent Shell Petroleum Development Company of Nigeria, Ltd. (SPDC), was incorporated in Nigeria, and engaged in oil exploration and production in Ogoniland. According to the complaint, after concerned residents of Ogoniland began protesting the environmental effects of SPDC’s practices, respondents enlisted the Nigerian Government to violently suppress the burgeoning demonstrations. Throughout the early 1990’s, the complaint alleges, Nigerian military and police forces attacked Ogoni vil- lages, beating, raping, killing, and arresting residents and destroying or looting property. Petitioners further allege that respondents aided and abetted these atrocities by, among other things, providing the Nigerian forces with food, transportation, and compensation, as well as by al- lowing the Nigerian military to use respondents’ property as a staging ground for attacks. Following the alleged atrocities, petitioners moved to the United States where they have been granted political asylum and now reside as legal residents. See Supp. Brief for Petitioners 3, and n. 2. They filed suit in the United States District Court for the Southern District of New York, alleging jurisdiction under the Alien Tort Statute and requesting relief under customary international law. The ATS provides, in full, that “[t]he district courts shall have original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States.” 28 U. S. C. §1350. According to petitioners, respondents violated the law of nations by aiding and abetting the Nigerian Government in committing (1) extrajudicial killings; (2) crimes against humanity; (3) torture and cruel treatment; (4) arbitrary arrest and detention; (5) violations of the rights to life, liberty, security, and association; (6) forced exile; and (7) property destruction. The District Court dismissed the first, fifth, sixth, and seventh claims, reasoning that the facts alleged to support those claims did not give rise to a violation of the law of nations. The court denied respondents’ motion to dismiss with respect to the remaining claims, but certified its order for interlocutory appeal pursuant to §1292(b). The Second Circuit dismissed the entire complaint, rea- soning that the law of nations does not recognize corpo- rate liability. 621 F.3d 111 (2010). We granted certiorari to consider that question. 565 U. S. ___ (2011). After oral argument, we directed the parties to file supplemen- tal briefs addressing an additional question: “Whether and under what circumstances the [ATS] allows courts to recognize a cause of action for violations of the law of nations occurring within the territory of a sovereign other than the United States.” 565 U. S. ___ (2012). We heard oral argument again and now affirm the judgment below, based on our answer to the second question. II Passed as part of the Judiciary Act of 1789, the ATS was invoked twice in the late 18th century, but then only once more over the next 167 years. Act of Sept. 24, 1789, §9, 1 Stat 77; see Moxon v. The Fanny, 17 F. Cas. 942 (No. 9,895) (DC Pa. 1793); Bolchos v. Darrel, 3 F. Cas. 810 (No. 1,607) (DC SC 1795); O’Reilly de Camara v. Brooke, 209 U.S. 45 (1908); Khedivial Line, S.A.E. v. Seafarers’ Int’l Union, 278 F.2d 49, 51–52 (CA2 1960) (per curiam). The statute provides district courts with jurisdiction to hear certain claims, but does not expressly provide any causes of action. We held in Sosa v. Alvarez-Machain, 542 U.S. 692, 714 (2004), however, that the First Congress did not intend the provision to be “stillborn.” The grant of jurisdiction is instead “best read as having been enacted on the understanding that the common law would provide a cause of action for [a] modest number of international law violations.” Id., at 724. We thus held that federal courts may “recognize private claims [for such violations] under federal common law.” Id., at 732. The Court in Sosa rejected the plaintiff’s claim in that case for “arbitrary arrest and detention,” on the ground that it failed to state a violation of the law of nations with the requisite “definite content and acceptance among civilized nations.” Id., at 699, 732. The question here is not whether petitioners have stated a proper claim under the ATS, but whether a claim may reach conduct occurring in the territory of a foreign sovereign. Respondents contend that claims under the ATS do not, relying primarily on a canon of statutory interpretation known as the presumption against extraterritorial application. That canon provides that “[w]hen a statute gives no clear indication of an extraterritorial application, it has none,” Morrison v. National Australia Bank Ltd., 561 U. S. ___, ___ (2010) (slip op., at 6), and reflects the “presumption that United States law governs domestically but does not rule the world,” Microsoft Corp. v. AT&T Corp., 550 U.S. 437, 454 (2007). This presumption “serves to protect against unintended clashes between our laws and those of other nations which could result in international discord.” EEOC v. Arabian American Oil Co., 499 U.S. 244, 248 (1991) (Aramco). As this Court has explained: “For us to run interference in . . . a delicate field of international relations there must be present the affirmative intention of the Congress clearly expressed. It alone has the facilities necessary to make fairly such an important policy decision where the possibilities of international discord are so evident and retaliative action so certain.” Benz v. Compania Naviera Hidalgo, S. A., 353 U.S. 138, 147 (1957). The presumption against extraterritorial application helps ensure that the Judiciary does not erroneously adopt an interpretation of U. S. law that carries foreign pol- icy consequences not clearly intended by the political branches. We typically apply the presumption to discern whether an Act of Congress regulating conduct applies abroad. See, e.g., Aramco, supra, at 246 (“These cases present the issue whether Title VII applies extraterritorially to regulate the employment practices of United States employers who employ United States citizens abroad”); Morrison, supra, at ___ (slip op., at 4) (noting that the question of extraterritorial application was a “merits question,” not a question of jurisdiction). The ATS, on the other hand, is “strictly jurisdictional.” Sosa, 542 U. S., at 713. It does not directly regulate conduct or afford relief. It instead allows federal courts to recognize certain causes of action based on sufficiently definite norms of international law. But we think the principles underlying the canon of interpretation similarly constrain courts considering causes of action that may be brought under the ATS. Indeed, the danger of unwarranted judicial interference in the conduct of foreign policy is magnified in the context of the ATS, because the question is not what Congress has done but instead what courts may do. This Court in Sosa repeatedly stressed the need for judicial caution in considering which claims could be brought under the ATS, in light of foreign policy concerns. As the Court explained, “the potential [foreign policy] implications . . . of recog- nizing . . . . causes [under the ATS] should make courts particularly wary of impinging on the discretion of the Legislative and Executive Branches in managing foreign affairs.” Id., at 727; see also id., at 727–728 (“Since many attempts by federal courts to craft remedies for the violation of new norms of international law would raise risks of adverse foreign policy consequences, they should be undertaken, if at all, with great caution”); id., at 727 (“[T]he possible collateral consequences of making international rules privately actionable argue for judicial caution”). These concerns, which are implicated in any case arising under the ATS, are all the more pressing when the question is whether a cause of action under the ATS reaches conduct within the territory of another sovereign. These concerns are not diminished by the fact that Sosa limited federal courts to recognizing causes of action only for alleged violations of international law norms that are “ ‘specific, universal, and obligatory.’ ” Id., at 732 (quoting In re Estate of Marcos, Human Rights Litigation, 25 F.3d 1467, 1475 (CA9 1994)). As demonstrated by Congress’s enactment of the Torture Victim Protection Act of 1991, 106Stat. 73, note following 28 U. S. C. §1350, identifying such a norm is only the beginning of defining a cause of action. See id., §3 (providing detailed definitions for extrajudicial killing and torture); id., §2 (specifying who may be liable, creating a rule of exhaustion, and establishing a statute of limitations). Each of these decisions carries with it significant foreign policy implications. The principles underlying the presumption against ex- traterritoriality thus constrain courts exercising their power under the ATS. III Petitioners contend that even if the presumption applies, the text, history, and purposes of the ATS rebut it for causes of action brought under that statute. It is true that Congress, even in a jurisdictional provision, can indicate that it intends federal law to apply to conduct occurring abroad. See, e.g., 18 U. S. C. §1091(e) (2006 ed., Supp. V) (providing jurisdiction over the offense of genocide “regardless of where the offense is committed” if the alleged offender is, among other things, “present in the United States”). But to rebut the presumption, the ATS would need to evince a “clear indication of extraterritoriality.” Morrison, 561 U. S., at ___ (slip op., at 16). It does not. To begin, nothing in the text of the statute suggests that Congress intended causes of action recognized under it to have extraterritorial reach. The ATS covers actions by aliens for violations of the law of nations, but that does not imply extraterritorial reach—such violations affect- ing aliens can occur either within or outside the United States. Nor does the fact that the text reaches “any civil action” suggest application to torts committed abroad; it is well established that generic terms like “any” or “every” do not rebut the presumption against extraterritoriality. See, e.g., id., at ___ (slip op., at 13–14); Small v. United States, 544 U.S. 385, 388 (2005); Aramco, 499 U. S., at 248–250; Foley Bros., Inc. v. Filardo, 336 U.S. 281, 287 (1949). Petitioners make much of the fact that the ATS provides jurisdiction over civil actions for “torts” in violation of the law of nations. They claim that in using that word, the First Congress “necessarily meant to provide for jurisdiction over extraterritorial transitory torts that could arise on foreign soil.” Supp. Brief for Petitioners 18. For support, they cite the common-law doctrine that allowed courts to assume jurisdiction over such “transitory torts,” including actions for personal injury, arising abroad. See Mostyn v. Fabrigas, 1 Cowp. 161, 177, 98 Eng. Rep. 1021, 1030 (1774) (Mansfield, L.) (“[A]ll actions of a transitory nature that arise abroad may be laid as happening in an English county”); Dennick v. Railroad Co., 103 U.S. 11, 18 (1881) (“Wherever, by either the common law or the statute law of a State, a right of action has become fixed and a legal liability incurred, that liability may be enforced and the right of action pursued in any court which has jurisdiction of such matters and can obtain jurisdiction of the parties”). Under the transitory torts doctrine, however, “the only justification for allowing a party to recover when the cause of action arose in another civilized jurisdiction is a well founded belief that it was a cause of action in that place.” Cuba R. Co. v. Crosby, 222 U.S. 473, 479 (1912) (majority opinion of Holmes, J.). The question under Sosa is not whether a federal court has jurisdiction to entertain a cause of action provided by foreign or even international law. The question is instead whether the court has authority to recognize a cause of action under U. S. law to enforce a norm of international law. The reference to “tort” does not demonstrate that the First Congress “necessarily meant” for those causes of action to reach conduct in the territory of a foreign sovereign. In the end, nothing in the text of the ATS evinces the requisite clear indication of extraterritoriality. Nor does the historical background against which the ATS was enacted overcome the presumption against ap- plication to conduct in the territory of another sovereign. See Morrison, supra, at ___ (slip op., at 16) (noting that “[a]ssuredly context can be consulted” in determining whether a cause of action applies abroad). We explained in Sosa that when Congress passed the ATS, “three principal offenses against the law of nations” had been identified by Blackstone: violation of safe conducts, infringement of the rights of ambassadors, and piracy. 542 U. S., at 723, 724; see 4 W. Blackstone, Commentaries on the Laws of England 68 (1769). The first two offenses have no necessary extraterritorial application. Indeed, Blackstone—in describing them—did so in terms of conduct occur- ring within the forum nation. See ibid. (describing the right of safe conducts for those “who are here”); 1 id., at 251 (1765) (explaining that safe conducts grant a member of one society “a right to intrude into another”); id., at 245–248 (recognizing the king’s power to “receiv[e] ambassadors at home” and detailing their rights in the state “wherein they are appointed to reside”); see also E. De Vattel, Law of Nations 465 (J. Chitty et al. transl. and ed. 1883) (“[O]n his entering the country to which he is sent, and making himself known, [the ambassador] is under the protection of the law of nations . . .”). Two notorious episodes involving violations of the law of nations occurred in the United States shortly before passage of the ATS. Each concerned the rights of ambas- sadors, and each involved conduct within the Union. In 1784, a French adventurer verbally and physically assaulted Francis Barbe Marbois—the Secretary of the French Legion—in Philadelphia. The assault led the French Minister Plenipotentiary to lodge a formal protest with the Continental Congress and threaten to leave the country unless an adequate remedy were provided. Respublica v. De Longschamps, 1 Dall. 111 (O. T. Phila. 1784); Sosa, supra, at 716–717, and n. 11. And in 1787, a New York constable entered the Dutch Ambassador’s house and arrested one of his domestic servants. See Casto, The Federal Courts’ Protective Jurisdiction over Torts Committed in Violation of the Law of Nations, 18 Conn. L. Rev. 467, 494 (1986). At the request of Secretary of Foreign Affairs John Jay, the Mayor of New York City arrested the constable in turn, but cautioned that because “ ‘neither Congress nor our [State] Legislature have yet passed any act respecting a breach of the privileges of Ambassadors,’ ” the extent of any available relief would depend on the common law. See Bradley, The Alien Tort Statute and Article III, 42 Va. J. Int’l L. 587, 641–642 (2002) (quoting 3 Dept. of State, The Diplomatic Correspondence of the United States of America 447 (1837)). The two cases in which the ATS was invoked shortly after its passage also concerned conduct within the territory of the United States. See Bolchos, 3 F. Cas. 810 (wrongful seizure of slaves from a vessel while in port in the United States); Moxon, 17 F. Cas. 942 (wrongful seizure in United States territorial waters). These prominent contemporary examples—immediately before and after passage of the ATS—provide no support for the proposition that Congress expected causes of action to be brought under the statute for violations of the law of nations occurring abroad. The third example of a violation of the law of nations familiar to the Congress that enacted the ATS was piracy. Piracy typically occurs on the high seas, beyond the territorial jurisdiction of the United States or any other country. See 4 Blackstone, supra, at 72 (“The offence of piracy, by common law, consists of committing those acts of robbery and depredation upon the high seas, which, if committed upon land, would have amounted to felony there”). This Court has generally treated the high seas the same as foreign soil for purposes of the presumption against extraterritorial application. See, e.g., Sale v. Haitian Centers Council, Inc., 509 U.S. 155, 173–174 (1993) (declining to apply a provision of the Immigration and Nationality Act to conduct occurring on the high seas); Argentine Republic v. Amerada Hess Shipping Corp., 488 U.S. 428, 440 (1989) (declining to apply a provision of the Foreign Sovereign Immunities Act of 1976 to the high seas). Petitioners contend that because Congress surely intended the ATS to provide jurisdiction for actions against pirates, it necessarily anticipated the statute would apply to conduct occurring abroad. Applying U. S. law to pirates, however, does not typi- cally impose the sovereign will of the United States onto conduct occurring within the territorial jurisdiction of another sovereign, and therefore carries less direct foreign policy consequences. Pirates were fair game wherever found, by any nation, because they generally did not operate within any jurisdiction. See 4 Blackstone, supra, at 71. We do not think that the existence of a cause of action against them is a sufficient basis for concluding that other causes of action under the ATS reach conduct that does occur within the territory of another sovereign; pirates may well be a category unto themselves. See Morrison, 561 U. S., at ___ (slip op., at 16) (“[W]hen a statute provides for some extraterritorial application, the presumption against extraterritoriality operates to limit that provision to its terms”); see also Microsoft Corp., 550 U. S., at 455–456. Petitioners also point to a 1795 opinion authored by Attorney General William Bradford. See Breach of Neutrality, 1 Op. Atty. Gen. 57. In 1794, in the midst of war between France and Great Britain, and notwithstanding the American official policy of neutrality, several U. S. citizens joined a French privateer fleet and attacked and plundered the British colony of Sierra Leone. In response to a protest from the British Ambassador, Attorney General Bradford responded as follows: So far . . . as the transactions complained of originated or took place in a foreign country, they are not within the cognizance of our courts; nor can the actors be legally prosecuted or punished for them by the United States. But crimes committed on the high seas are within the jurisdiction of the . . . courts of the United States; and, so far as the offence was committed thereon, I am inclined to think that it may be legally prosecuted in . . . those courts . . . . But some doubt rests on this point, in consequence of the terms in which the [applicable criminal law] is expressed. But there can be no doubt that the company or individuals who have been injured by these acts of hostil- ity have a remedy by a civil suit in the courts of the United States; jurisdiction being expressly given to these courts in all cases where an alien sues for a tort only, in violation of the laws of nations, or a treaty of the United States . . . .” Id., at 58–59. Petitioners read the last sentence as confirming that “the Founding generation understood the ATS to apply to law of nations violations committed on the territory of a foreign sovereign.” Supp. Brief for Petitioners 33. Respondents counter that when Attorney General Bradford referred to “these acts of hostility,” he meant the acts only insofar as they took place on the high seas, and even if his conclusion were broader, it was only because the applicable treaty had extraterritorial reach. See Supp. Brief for Respondents 28–30. The Solicitor General, having once read the opinion to stand for the proposition that an “ATS suit could be brought against American citizens for breaching neutrality with Britain only if acts did not take place in a foreign country,” Supp. Brief for United States as Amicus Curiae 8, n. 1 (internal quotation marks and brackets omitted), now suggests the opinion “could have been meant to encompass . . . conduct [occurring within the foreign territory],” id., at 8. Attorney General Bradford’s opinion defies a definitive reading and we need not adopt one here. Whatever its pre- cise meaning, it deals with U. S. citizens who, by partic- ipating in an attack taking place both on the high seas and on a foreign shore, violated a treaty between the United States and Great Britain. The opinion hardly suffices to counter the weighty concerns underlying the presumption against extraterritoriality. Finally, there is no indication that the ATS was passed to make the United States a uniquely hospitable forum for the enforcement of international norms. As Justice Story put it, “No nation has ever yet pretended to be the custos morum of the whole world . . . .” United States v. The La Jeune Eugenie, 26 F. Cas. 832, 847 (No. 15,551) (CC. Mass. 1822). It is implausible to suppose that the First Congress wanted their fledgling Republic—struggling to receive international recognition—to be the first. Indeed, the parties offer no evidence that any nation, meek or mighty, presumed to do such a thing. The United States was, however, embarrassed by its potential inability to provide judicial relief to foreign officials injured in the United States. Bradley, 42 Va. J. Int’l L., at 641. Such offenses against ambassadors vio- lated the law of nations, “and if not adequately redressed could rise to an issue of war.” Sosa, 542 U. S., at 715; cf. The Federalist No. 80, p. 536 (J. Cooke ed. 1961) (A. Hamilton) (“As the denial or perversion of justice . . . is with reason classed among the just causes of war, it will follow that the federal judiciary ought to have cognizance of all causes in which the citizens of other countries are concerned”). The ATS ensured that the United States could provide a forum for adjudicating such incidents. See Sosa, supra, at 715–718, and n. 11. Nothing about this historical context suggests that Congress also intended federal common law under the ATS to provide a cause of action for conduct occurring in the territory of another sovereign. Indeed, far from avoiding diplomatic strife, providing such a cause of action could have generated it. Recent experience bears this out. See Doe v. Exxon Mobil Corp., 654 F.3d 11, 77–78 (CADC 2011) (Kavanaugh, J., dissenting in part) (listing recent objections to extraterritorial applications of the ATS by Canada, Germany, Indonesia, Papua New Guinea, South Africa, Switzerland, and the United Kingdom). Moreover, accepting petitioners’ view would imply that other nations, also applying the law of nations, could hale our citizens into their courts for alleged violations of the law of nations occurring in the United States, or anywhere else in the world. The presumption against extraterritoriality guards against our courts triggering such serious foreign policy consequences, and instead defers such decisions, quite appropriately, to the political branches. We therefore conclude that the presumption against extraterritoriality applies to claims under the ATS, and that nothing in the statute rebuts that presumption. “[T]here is no clear indication of extraterritoriality here,” Morrison, 561 U. S., at ___ (slip op., at 16), and petitioners’ case seeking relief for violations of the law of nations occurring outside the United States is barred. IV On these facts, all the relevant conduct took place outside the United States. And even where the claims touch and concern the territory of the United States, they must do so with sufficient force to displace the presumption against extraterritorial application. See Morrison, 561 U. S. ___ (slip op. at 17–24). Corporations are often present in many countries, and it would reach too far to say that mere corporate presence suffices. If Congress were to determine otherwise, a statute more specific than the ATS would be required. The judgment of the Court of Appeals is affirmed. It is so ordered.

567.US.2011_10-1121

California law permits public-sector employees in a bargaining unit to decide by majority vote to create an “agency shop” arrangement under which all the employees are represented by a union. Even employees who do not join the union must pay an annual fee for “chargeable expenses,” i.e., the cost of nonpolitical union services related to collective bargaining. Under Abood v. Detroit Bd. of Ed., 431 U.S. 209, a public-sector union can bill nonmembers for chargeable expenses but may not require them to fund its political or ideological projects. Teachers v. Hudson, 475 U.S. 292, 302–311, sets out requirements that a union must meet in order to collect regular fees from nonmembers without violating their rights. In June 2005, respondent, a public-sector union (SEIU), sent to California employees its annual Hudson notice, setting and capping monthly dues and estimating that 56.35% of its total expenditures in the coming year would be chargeable expenses. A nonmember had 30 days to object to full payment of dues but would still have to pay the chargeable portion. The notice stated that the fee was subject to increase without further notice. That same month, the Governor called for a special election on, inter alia, two ballot propositions opposed by the SEIU. After the 30-day objection period ended, the SEIU sent a letter to unit employees announcing a temporary 25% increase in dues and a temporary elimination of the monthly dues cap, billing the move as an “Emergency Temporary Assessment to Build a Political Fight-Back Fund.” The purpose of the fund was to help achieve the union’s political objectives in the special election and in the upcoming November 2006 election. The union noted that the fund would be used “for a broad range of political expenses, including television and radio advertising, direct mail, voter registration, voter education, and get out the vote activities in our work sites and in our communities across California.” Nonunion employees were not given any choice as to whether they would pay into the fund. Petitioners, on behalf of nonunion employees who paid into the fund, brought a class action against the SEIU alleging violation of their First Amendment rights. The Federal District Court granted petitioners summary judgment. Ruling that the special assessment was for entirely political purposes, it ordered the SEIU to send a new notice giving class members 45 days to object and to provide those who object a full refund of contributions to the fund. The Ninth Circuit reversed, concluding that Hudson prescribed a balancing test under which the proper inquiry is whether the SEIU’s procedures reasonably accommodated the interests of the union, the employer, and the nonmember employees. Held: 1. This case is not moot. Although the SEIU offered a full refund to all class members after certiorari was granted, a live controversy remains. The voluntary cessation of challenged conduct does not ordinarily render a case moot because that conduct could be resumed as soon as the case is dismissed. See City of Mesquite v. Aladdin’s Castle, Inc., 455 U.S. 283, 289. Since the SEIU continues to defend the fund’s legality, it would not necessarily refrain from collecting similar fees in the future. Even if concerns about voluntary cessation were inapplicable because petitioners did not seek prospective relief, there would still be a live controversy as to the adequacy of the refund notice the SEIU sent pursuant to the District Court’s order. Pp. 6−8. 2. Under the First Amendment, when a union imposes a special assessment or dues increase levied to meet expenses that were not disclosed when the regular assessment was set, it must provide a fresh notice and may not exact any funds from nonmembers without their affirmative consent. Pp. 8−23. (a) A close connection exists between this Nation’s commitment to self-government and the rights protected by the First Amendment, see, e.g., Brown v. Hartlage, 456 U.S. 45, 52−53, which creates “an open marketplace” in which differing ideas about political, economic, and social issues can compete freely for public acceptance without improper government interference, New York State Bd. of Elections v. Lopez Torres, 552 U. S 196, 202. The government may not prohibit the dissemination of ideas it disfavors, nor compel the endorsement of ideas that it approves. See, e.g., R. A. V. v. St. Paul, 505 U.S. 377, 382. And the ability of like-minded individuals to associate for the purpose of expressing commonly held views may not be curtailed. See, e.g., Roberts v. United States Jaycees, 468 U.S. 609, 623. Closely related to compelled speech and compelled association is compelled funding of the speech of private speakers or groups. Compulsory subsidies for private speech are thus subject to exacting First Amendment scrutiny and cannot be sustained unless, first, there is a comprehensive regulatory scheme involving a “mandated association” among those who are required to pay the subsidy, United States v. United Foods, Inc., 533 U.S. 405, and, second, compulsory fees are levied only insofar as they are a “necessary incident” of the “larger regulatory purpose which justified the required association,” ibid. Pp. 8−10. (b) When a State establishes an “agency shop” that exacts compulsory union fees as a condition of public employment, “[t]he dissenting employee is forced to support financially an organization with whose principles and demands he may disagree.” Ellis v. Railway Clerks, 466 U.S. 435, 455. This form of compelled speech and association imposes a “significant impingement on First Amendment rights.” Ibid. The justification for permitting a union to collect fees from nonmembers—to prevent them from free-riding on the union’s efforts—is an anomaly. Similarly, requiring objecting nonmembers to opt out of paying the nonchargeable portion of union dues―rather than exempting them unless they opt in―represents a remarkable boon for unions, creating a risk that the fees nonmembers pay will be used to further political and ideological ends with which they do not agree. Thus, Hudson, far from calling for a balancing of rights or interests, made it clear that any procedure for exacting fees from unwilling contributors must be “carefully tailored to minimize the infringement” of free speech rights, 475 U. S. 302−303, and it cited cases holding that measures burdening the freedom of speech or association must serve a compelling interest and must not be significantly broader than necessary to serve that interest. Pp. 10−13. (c) There is no justification for the SEIU’s failure to provide a fresh Hudson notice. Hudson rests on the principle that nonmembers should not be required to fund a union’s political and ideological projects unless they choose to do so after having “a fair opportunity” to assess the impact of paying for nonchargeable union activities. 475 U. S., at 303. The SEIU’s procedure cannot be considered to have met Hudson’s requirement that fee-collection procedures be carefully tailored to minimize impingement on First Amendment rights. The SEIU argues that nonmembers who objected to the special assessment but were not given the opportunity to opt out would have been given the chance to recover the funds by opting out when the next annual notice was sent, and that the amount of dues payable the following year by objecting nonmembers would decrease if the special assessment were found to be for nonchargeable purposes. But this decrease would not fully recompense nonmembers, who would not have paid to support the special assessment if given the choice. In any event, even a full refund would not undo the First Amendment violations, since the First Amendment does not permit a union to extract a loan from unwilling nonmembers even if the money is later paid back in full. Pp. 14−17. (d) The SEIU’s treatment of nonmembers who opted out when the initial Hudson notice was sent also ran afoul of the First Amendment. They were required to pay 56.35% of the special assessment even though all the money was slated for nonchargeable, electoral uses. And the SEIU’s claim that the assessment was a windfall because chargeable expenses turned out to be 66.26% is unpersuasive. First, the SEIU’s understanding of the breadth of chargeable expenses is so expansive that it is hard to place much reliance on its statistics. “Lobbying the electorate,” which the SEIU claims is chargeable, is nothing more than another term for supporting political causes and candidates. Second, even if the SEIU’s statistics are accurate, it does not follow that it was proper to charge objecting nonmembers any particular percentage of the special assessment. If, as the SEIU argues, it is not possible to accurately determine in advance the percentage of union funds that will be used for an upcoming year’s chargeable purposes, there is a risk that unconsenting nonmembers will have paid too much or too little. That risk should be borne by the side whose constitutional rights are not at stake. If the nonmembers pay too much, their First Amendment rights are infringed. But, if they pay too little, no constitutional right of the union is violated because it has no constitutional right to receive any payment from those employees. Pp. 17−23. 628 F.3d 1115, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, and Thomas, JJ., joined. Sotomayor, J., filed an opinion concurring in the judgment, in which Ginsburg, J., joined. Breyer, J., filed a dissenting opinion, in which Kagan, J., joined.

In this case, we decide whether the First Amendment allows a public-sector union to require objecting nonmembers to pay a special fee for the purpose of financing the union’s political and ideological activities. I A Under California law, public-sector employees in a bargaining unit may decide by majority vote to create an “agency shop” arrangement under which all the employees are represented by a union selected by the majority. Cal. Govt. Code Ann. §3502.5(a) (West 2010). While employees in the unit are not required to join the union, they must nevertheless pay the union an annual fee to cover the cost of union services related to collective bargaining (so-called chargeable expenses). See Lehnert v. Ferris Faculty Assn., 500 U.S. 507, 524 (1991); Machinists v. Street, 367 U.S. 740, 760 (1961). Our prior cases have recognized that such arrangements represent an “impingement” on the First Amendment rights of nonmembers. Teachers v. Hudson, 475 U.S. 292, 307, n. 20 (1986). See also Davenport v. Washington Ed. Assn., 551 U.S. 177, 181 (2007) (“[A]gency-shop arrangements in the public sector raise First Amendment concerns because they force individuals to contribute money to unions as a condition of government employment”); Street, supra, at 749 (union shop presents First Amendment “questions of the utmost gravity”). Thus, in Abood v. Detroit Bd. of Ed., 431 U.S. 209 (1977), we held that a public-sector union, while permitted to bill nonmembers for chargeable expenses, may not require nonmembers to fund its political and ideological projects. And in Hudson, we identified procedural requirements that a union must meet in order to collect fees from nonmembers without violating their rights. 475 U. S., at 302–311. The First Amendment, we held, does not permit a public-sector union to adopt procedures that have the effect of requiring objecting nonmembers to lend the union money to be used for political, ideological, and other purposes not germane to collective bargaining. Id., at 305. In the interest of administrative convenience, however, we concluded that a union “cannot be faulted” for calculating the fee that nonmembers must pay “on the basis of its expenses during the preceding year.” Id., at 307, n. 18. Hudson concerned a union’s regular annual fees. The present case, by contrast, concerns the First Amendment requirements applicable to a special assessment or dues increase that is levied to meet expenses that were not disclosed when the amount of the regular assessment was set. B In June 2005, respondent, the Service Employees International Union, Local 1000 (SEIU), sent out its regular Hudson notice informing employees what the agency fee would be for the year ahead. The notice set monthly dues at 1% of an employee’s gross monthly salary but capped monthly dues at $45. Based on the most recently audited year, the SEIU estimated that 56.35% of its total expenditures in the coming year would be dedicated to chargeable collective-bargaining activities. Thus, if a nonunion employee objected within 30 days to payment of the full amount of union dues, the objecting employee was required to pay only 56.35% of total dues. The SEIU’s notice also included a feature that was not present in Hudson: The notice stated that the agency fee was subject to increase at any time without further notice. During this time, the citizens of the State of California were engaged in a wide-ranging political debate regarding state budget deficits, and in particular the budget consequences of growing compensation for public employees backed by powerful public-sector unions. On June 13, 2005, Governor Arnold Schwarzenegger called for a special election to be held in November 2005, where voters would consider various ballot propositions aimed at state-level structural reforms. Two of the most controversial issues on the ballot were Propositions 75 and 76. Proposition 75 would have required unions to obtain employees’ affirmative consent before charging them fees to be used for political purposes. Proposition 76 would have limited state spending and would have given the Governor the ability under some circumstances to reduce state appropriations for public-employee compensation. The SEIU joined a coalition of public-sector unions in vigorously opposing these measures. Calling itself the “Alliance for a Better California,” the group would eventually raise “more than $10 million, with almost all of it coming from public employee unions, including $2.75 million from state worker unions, $4.7 million from the California Teachers Association, and $700,000 from school workers unions.”[1] On July 30, shortly after the end of the 30-day objection period for the June Hudson notice, the SEIU proposed a temporary 25% increase in employee fees, which it billed as an “Emergency Temporary Assessment to Build a Po- litical Fight-Back Fund.” App. 25. The proposal stated that the money was needed to achieve the union’s political objectives, both in the special November 2005 election and in the November 2006 election. Id., at 26. According to the proposal, money in the Fight-Back Fund would be used “for a broad range of political expenses, including television and radio advertising, direct mail, voter registration, voter education, and get out the vote activities in our work sites and in our communities across California.” Ibid. The proposal specifically stated that “[t]he Fund will not be used for regular costs of the union—such as office rent, staff salaries or routine equipment replacement, etc.” Ibid. It noted that “all other public worker unions are in the process of raising the extraordinary funds needed to defeat the Governor.” Id., at 27. And it concluded: “Each of us must do our part to turn back these initiatives which would allow the Governor to destroy our wages and benefits and even our jobs, and threaten the well-being of all Californians.” Ibid. On August 27, the SEIU’s General Council voted to implement the proposal. On August 31, the SEIU sent out a letter addressed to “Local 1000 Members and Fair Share Fee Payers,” an- nouncing that, for a limited period, their fees would be raised to 1.25% of gross monthly salary and the $45-per-month cap on regular dues would not apply. Id., at 31. The letter explained that the union would use the fund to “defeat Proposition 76 and Proposition 75 on November 8,” and to “defeat another attack on [its] pension plan” in June 2006. Ibid. The letter also informed employees that, in the following year, the money would help “to elect a governor and a legislature who support public employees and the services [they] provide.” Ibid. After receiving this letter, one of the plaintiffs in this case called the SEIU’s offices to complain that the union was levying the special assessment for political purposes without giving employees a fair opportunity to object. An SEIU area manager responded that “even if [the employee] objected to the payment of the full agency fee, there was nothing he could do about the September increase for the Assessment.” Knox v. Westly, No. 2:05–cv–02198, 2008 WL 850128, *3 (ED Cal., Mar. 28, 2008). “She also stated that ‘we are in the fight of our lives,’ that the Assessment was needed, and that there was nothing that could be done to stop the Union’s expenditure of that Assessment for political purposes.” Ibid. As a consolation, however, those employees who had filed timely objections after the regular June Hudson notice were required to pay only 56.35% of the temporary increase. Petitioners filed this class-action suit on behalf of 28,000 nonunion employees who were forced to contribute money to the Political Fight-Back Fund. Some of the class members had filed timely objections after receiving the regular Hudson notice in June, and others had not. Those who had objected argued that it was wrong to require them to pay 56.35% of the temporary assessment, which had been billed as intended for use in making political expenditures that they found objectionable. Those who had not objected after receiving the June Hudson notice contended that they should have received a new opportunity to object when the SEIU levied the special assessment for its Political Fight-Back Fund. The District Court granted summary judgment for the petitioners, finding that the union “fully intended to use the 12 million additional dollars it anticipated to raise for political purposes.” 2008 WL 850128, *7. “Even if every cent of the assessment was not intended to be used for entirely political purposes,” the court stated, “it is clear that the Union’s intent was to depart drastically from its typical spending regime and to focus on activities that were political or ideological in nature.” Id., at *8. In light of this fact, the court held that it would be inappropriate for the union to rely on previous annual expenditures to estimate that 56.35% of the new fee would go toward chargeable expenses. The court ordered the SEIU to send out a new notice giving all class members 45 days to object and to provide those who objected with a full refund of their contributions to the Political Fight-Back Fund. Id., at *12. A divided panel of the Ninth Circuit reversed. Knox v. California State Employees Assn., Local 1000, 628 F.3d 1115 (2010). According to the panel majority, Hudson prescribed the use of a balancing test. 628 F. 3d, at 1119–1120. The majority therefore inquired whether the procedure that the SEIU employed reasonably accommodated the interests of the union, the employer, and nonmember employees. Id., at 1120–1123. Judge Wallace dissented, arguing that the majority had misinterpreted Hudson and sanctioned the abridgment of the First Amendment rights of nonmembers. 628 F. 3d, at 1123–1139. We granted certiorari. 564 U. S. ___ (2011). II The SEIU argues that we should dismiss this case as moot. In opposing the petition for certiorari, the SEIU defended the decision below on the merits. After certiorari was granted, however, the union sent out a notice offering a full refund to all class members, and the union then promptly moved for dismissal of the case on the ground of mootness. Such postcertiorari maneuvers designed to insulate a decision from review by this Court must be viewed with a critical eye. See City News & Novelty, Inc. v. Waukesha, 531 U.S. 278, 283–284 (2001). The vol-untary cessation of challenged conduct does not ordinar- ily render a case moot because a dismissal for mootness would permit a resumption of the challenged conduct as soon as the case is dismissed. See City of Mesquite v. Aladdin’s Castle, Inc., 455 U.S. 283, 289 (1982). And here, since the union continues to defend the legality of the Political Fight-Back fee, it is not clear why the union would necessarily refrain from collecting similar fees in the future. The union argues that concerns about voluntary cessation are inapplicable in this case because petitioners do not seek any prospective relief. See Motion to Dismiss as Moot 11–12. But even if that is so, the union’s mootness argument fails because there is still a live controversy as to the adequacy of the SEIU’s refund notice. A case becomes moot only when it is impossible for a court to grant “ ‘ “any effectual relief whatever” to the prevailing party.’ ” Erie v. Pap’s A. M., 529 U.S. 277, 287 (2000) (quoting Church of Scientology of Cal. v. United States, 506 U.S. 9, 12 (1992), in turn quoting Mills v. Green, 159 U.S. 651, 653 (1895)). “[A]s long as the parties have a concrete interest, however small, in the outcome of the litigation, the case is not moot.” Ellis v. Railway Clerks, 466 U.S. 435, 442 (1984). The District Court ordered the SEIU to send out a “proper” notice giving employees an adequate opportunity to receive a full refund. 2008 WL 850128, *12. Petitioners argue that the notice that the SEIU sent was improper because it includes a host of “conditions, caveats, and confusions as unnecessary complications aimed at reducing the number of class members who claim a refund.” Brief for Petitioners in Opposition to Motion to Dismiss 19. In particular, petitioners allege that the union has refused to accept refund requests by fax or e-mail and has made refunds conditional upon the provision of an original signature and a Social Security number. Id., at 18–19. As this dispute illustrates, the nature of the notice may affect how many employees who object to the union’s special assessment will be able to get their money back. The union is not entitled to dictate unilaterally the manner in which it advertises the availability of the refund. For this reason, we conclude that a live controversy remains, and we proceed to the merits. III A Our cases have often noted the close connection between our Nation’s commitment to self-government and the rights protected by the First Amendment. See, e.g., Brown v. Hartlage, 456 U.S. 45, 52 (1982) (“At the core of the First Amendment are certain basic conceptions about the manner in which political discussion in a representative democracy should proceed”); Buckley v. Valeo, 424 U.S. 1, 93, n. 127 (1976) (per curiam) (“[T]he central purpose of the Speech and Press Clauses was to assure a society in which ‘uninhibited, robust, and wide-open’ public debate concerning matters of public interest would thrive, for only in such a society can a healthy representative democracy flourish”); Cox v. Louisiana, 379 U.S. 536, 552 (1965) (“Maintenance of the opportunity for free political discussion is a basic tenet of our constitutional democracy”); Whitney v. California, 274 U.S. 357, 375 (1927) (Brandeis, J., concurring); Patterson v. Colorado ex rel. Attorney General of Colo., 205 U.S. 454, 465 (1907) (Harlan, J., dissenting). The First Amendment creates “an open marketplace” in which differing ideas about political, economic, and social issues can compete freely for public acceptance without improper government interference. New York State Bd. of Elections v. Lopez Torres, 552 U.S. 196, 208 (2008). See also Hustler Magazine, Inc. v. Falwell, 485 U.S. 46, 51 (1988); Mills v. Alabama, 384 U.S. 214, 218–219 (1966). The government may not prohibit the dissemination of ideas that it disfavors, nor compel the endorsement of ideas that it approves. See R. A. V. v. St. Paul, 505 U.S. 377, 382 (1992); Brandenburg v. Ohio, 395 U.S. 444, 447–448 (1969) (per curiam); West Virginia Bd. of Ed. v. Barnette, 319 U.S. 624 (1943); Wooley v. Maynard, 430 U.S. 705, 713–715 (1977); Riley v. National Federation of Blind of N. C., Inc., 487 U.S. 781, 797 (1988) (The First Amendment protects “the decision of both what to say and what not to say” (emphasis deleted)). And the ability of like-minded individuals to associate for the purpose of expressing commonly held views may not be curtailed. See Roberts v. United States Jaycees, 468 U.S. 609, 623 (1984) (“Freedom of association . . . plainly presupposes a freedom not to associate”); NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 460–461 (1958). Closely related to compelled speech and compelled association is compelled funding of the speech of other private speakers or groups. See Abood, 431 U. S., at 222–223. In United States v. United Foods, Inc., 533 U.S. 405 (2001), we considered the constitutionality of a state scheme that compelled such funding. The subject of the speech at issue—promoting the sale of mushrooms—was not one that is likely to stir the passions of many, but the mundane commercial nature of that speech only highlights the importance of our analysis and our holding. The federal Mushroom Promotion, Research, and Consumer Information Act required that fresh mushroom handlers pay assessments used primarily to fund advertisements promoting mushroom sales. A large producer objected to subsidizing these generic ads, and even though we applied the less demanding standard used in prior cases to judge laws affecting commercial speech, we held that the challenged scheme violated the First Amendment. We made it clear that compulsory subsidies for private speech are subject to exacting First Amendment scrutiny and cannot be sustained unless two criteria are met. First, there must be a comprehensive regulatory scheme involving a “mandated association” among those who are required to pay the subsidy. Id., at 414. Such situations are exceedingly rare because, as we have stated elsewhere, mandatory associations are permissible only when they serve a “compelling state interes[t] . . . that cannot be achieved through means significantly less restrictive of associational freedoms.” Roberts, supra, at 623. Second, even in the rare case where a mandatory association can be justified, compulsory fees can be levied only insofar as they are a “necessary incident” of the “larger regulatory purpose which justified the required association.” United Foods, supra, at 414. B When a State establishes an “agency shop” that ex- acts compulsory union fees as a condition of public employment, “[t]he dissenting employee is forced to support financially an organization with whose principles and demands he may disagree.” Ellis, 466 U. S., at 455. Because a public-sector union takes many positions during collective bargaining that have powerful political and civic consequences, see Tr. of Oral Arg. 48–49, the compulsory fees constitute a form of compelled speech and association that imposes a “significant impingement on First Amendment rights.” Ellis, supra, at 455. Our cases to date have tolerated this “impingement,” and we do not revisit today whether the Court’s former cases have given adequate recognition to the critical First Amendment rights at stake. “The primary purpose” of permitting unions to collect fees from nonmembers, we have said, is “to prevent nonmembers from free-riding on the union’s efforts, sharing the employment benefits obtained by the union’s collective bargaining without sharing the costs incurred.” Davenport, 551 U. S., at 181. Such free-rider arguments, however, are generally insufficient to overcome First Amend- ment objections. Consider the following examples: “If a community association engages in a clean-up campaign or opposes encroachments by industrial development, no one suggests that all residents or property owners who benefit be required to contribute. If a parent-teacher association raises money for the school library, assessments are not levied on all parents. If an association of university professors has as a major function bringing pressure on universities to observe standards of tenure and academic freedom, most professors would consider it an outrage to be required to join. If a medical association lobbies against regulation of fees, not all doctors who share in the benefits share in the costs.”[2] Acceptance of the free-rider argument as a justification for compelling nonmembers to pay a portion of union dues represents something of an anomaly—one that we have found to be justified by the interest in furthering “labor peace.” Hudson, 475 U. S., at 303. But it is an anomaly nevertheless. Similarly, requiring objecting nonmembers to opt out of paying the nonchargeable portion of union dues—as opposed to exempting them from making such payments unless they opt in—represents a remarkable boon for unions. Courts “do not presume acquiescence in the loss of fundamental rights.” College Savings Bank v. Florida Prepaid Postsecondary Ed. Expense Bd., 527 U.S. 666, 682 (1999) (internal quotation marks omitted). Once it is recognized, as our cases have, that a nonmember cannot be forced to fund a union’s political or ideological activities, what is the justification for putting the burden on the nonmember to opt out of making such a payment? Shouldn’t the default rule comport with the probable preferences of most nonmembers? And isn’t it likely that most employees who choose not to join the union that represents their bargaining unit prefer not to pay the full amount of union dues? An opt-out system creates a risk that the fees paid by nonmembers will be used to further political and ideological ends with which they do not agree. But a “[u]nion should not be permitted to exact a service fee from nonmembers without first establishing a procedure which will avoid the risk that their funds will be used, even temporarily, to finance ideological activities unrelated to collective bargaining.” Hudson, supra, at 305 (internal quotation marks omitted). Although the difference between opt-out and opt-in schemes is important, our prior cases have given sur-prisingly little attention to this distinction. Indeed, acceptance of the opt-out approach appears to have come about more as a historical accident than through the careful application of First Amendment principles. The trail begins with dicta in Street, where we considered whether a federal collective-bargaining statute authorized a union to impose compulsory fees for political activities. 367 U. S., at 774. The plaintiffs were employees who had affirmatively objected to the way their fees were being used, and so we took that feature of the case for granted. We held that the statute did not authorize the use of the objecting employees’ fees for ideological purposes, and we stated in passing that “dissent is not to be presumed—it must affirmatively be made known to the union by the dissenting employee.” Ibid. In making that offhand remark, we did not pause to consider the broader constitutional implications of an affirmative opt-out requirement. Nor did we explore the extent of First Amendment protection for employees who might not qual-ify as active “dissenters” but who would nonetheless prefer to keep their own money rather than subsidizing by default the political agenda of a state-favored union. In later cases such as Abood and Hudson, we assumed without any focused analysis that the dicta from Street had authorized the opt-out requirement as a constitutional matter. Thus in Hudson we did not take issue with the union’s practice of giving employees annual notice and an opportunity to object to expected political expenditures. At the same time, however, we made it clear that the procedures used by a union to collect money from nonmembers must satisfy a high standard. Contrary to the view of the Ninth Circuit panel major-ity, we did not call for a balancing of the “right” of the union to collect an agency fee against the First Amendment rights of nonmembers. 628 F. 3d, at 1119–1120. As we noted in Davenport, “unions have no constitutional entitlement to the fees of nonmember-employees.” 551 U. S., at 185. A union’s “collection of fees from nonmembers is authorized by an act of legislative grace,” 628 F. 3d, at 1126 (Wallace, J., dissenting)—one that we have termed “unusual” and “extraordinary,” Davenport, supra, at 184, 187. Far from calling for a balancing of rights or interests, Hudson made it clear that any procedure for exacting fees from unwilling contributors must be “care-fully tailored to minimize the infringement” of free speech rights. 475 U. S., at 303. And to underscore the meaning of this careful tailoring, we followed that statement with a citation to cases holding that measures burdening the freedom of speech or association must serve a “compelling interest” and must not be significantly broader than necessary to serve that interest.[3] IV By authorizing a union to collect fees from nonmembers and permitting the use of an opt-out system for the collection of fees levied to cover nonchargeable expenses, our prior decisions approach, if they do not cross, the limit of what the First Amendment can tolerate. The SEIU, however, asks us to go farther. It asks us to approve a procedure under which (a) a special assessment billed for use in electoral campaigns was assessed without providing a new opportunity for nonmembers to decide whether they wished to contribute to this effort and (b) nonmembers who previously opted out were nevertheless required to pay more than half of the special assessment even though the union had said that the purpose of the fund was to mount a political campaign and that it would not be used for ordinary union expenses. This aggressive use of power by the SEIU to collect fees from nonmembers is indefensible. A First, we see no justification for the union’s failure to provide a fresh Hudson notice. Hudson rests on the principle that nonmembers should not be required to fund a union’s political and ideological projects unless they choose to do so after having “a fair opportunity” to assess the im- pact of paying for nonchargeable union activities. 475 U. S., at 303. Giving employees only one opportunity per year to make this choice is tolerable if employees are able at the time in question to make an informed choice. But a nonmember cannot make an informed choice about a special assessment or dues increase that is unknown when the annual notice is sent. When a union levies a special assessment or raises dues as a result of unexpected developments, the factors influencing a nonmember’s choice may change. In particular, a nonmember may take special exception to the uses for which the additional funds are sought.[4] The present case provides a striking example. The special assessment in this case was billed for use in a broad electoral campaign designed to defeat two important and controversial ballot initiatives and to elect sympa- thetic candidates in the 2006 gubernatorial and legislative elections. There were undoubtedly nonmembers who, for one reason or another, chose not to opt out or neglected to do so when the standard Hudson notice was sent but who took strong exception to the SEIU’s political objectives and did not want to subsidize those efforts. These nonmembers might have favored one or both of the ballot initiatives; they might have wished to support the reelection of the incumbent Governor; or they might not have wanted to delegate to the union the authority to decide which candidates in the 2006 elections would receive a share of their money. The effect on nonmembers was particularly striking with respect to the union’s campaign against Proposition 75 because that initiative would have bolstered nonmember rights. If Proposition 75 had passed, nonmembers would have been exempt from paying for the SEIU’s extensive political projects unless they affirmatively consented. Thus, the effect of the SEIU’s procedure was to force many nonmembers to subsidize a political effort designed to restrict their own rights. As Hudson held, procedures for collecting fees from nonmembers must be carefully tailored to minimize impingement on First Amendment rights, and the procedure used in this case cannot possibly be considered to have met that standard. After the dues increase was adopted, the SEIU wrote to all employees in the relevant bargaining units to inform them of this development. It would have been a relatively simple matter for the union to cast this letter in the form of a new Hudson notice, so that nonmembers could decide whether they wanted to pay for the union’s electoral project. The SEIU argues that we should not be troubled by its failure to provide a new notice because nonmembers who objected to the special assessment but were nonetheless required to pay it would have been given the chance to recover the funds in question by opting out when the next annual notice was sent. If the special assessment was used entirely or in part for nonchargeable purposes, they suggest, the percentage of the union’s annual expenditures for chargeable purposes would decrease, and therefore the amount of the dues payable by objecting nonmembers the following year would also decrease. This decrease, how- ever, would not fully recompense nonmembers who did not opt out after receiving the regular notice but would have opted out if they had been permitted to do so when the special assessment was announced.[5] And in any event, even a full refund would not undo the violation of First Amendment rights. As we have recognized, the First Amendment does not permit a union to extract a loan from unwilling nonmembers even if the money is later paid back in full. See Hudson, supra, at 305; Ellis, 466 U. S., at 444. Here, for nonmembers who disagreed with the SEIU’s electoral objectives, a refund provided after the union’s objectives had already been achieved would be cold comfort.[6] To respect the limits of the First Amendment, the union should have sent out a new notice allowing nonmembers to opt in to the special fee rather than requiring them to opt out. Our cases have tolerated a substantial impingement on First Amendment rights by allowing unions to impose an opt-out requirement at all. Even if this burden can be justified during the collection of regular dues on an annual basis, there is no way to justify the additional burden of imposing yet another opt-out requirement to collect special fees whenever the union desires. B 1 The SEIU’s treatment of nonmembers who opted out when the initial Hudson notice was sent also ran afoul of the First Amendment. The SEIU required these employees to pay 56.35% of the special assessment, just as they had been required to pay 56.35% of the regular annual dues. But the union proclaimed that the special assessment would be used to support an electoral campaign and would not be used for ordinary union expenses. Accordingly, there is no reason to suppose that 56.35% of the new assessment was used for properly chargeable expenses. On the contrary, if the union is to be taken at its word, virtually all of the money was slated for nonchargeable uses. The procedure accepted in Hudson is designed for use when a union sends out its regular annual dues notices. The procedure is predicated on the assumption that a union’s allocation of funds for chargeable and noncharge- able purposes is not likely to vary greatly from one year to the next.[7] No such assumption is reasonable, however, when a union levies a special assessment or raises dues as a result of events that were not anticipated or disclosed at the time when a yearly Hudson notice was sent. Accordingly, use of figures based on an audit of the union’s operations during an entire previous year makes no sense. Nor would it be feasible to devise a new breakdown of chargeable and nonchargeable expenses for the special assessment. Determining that breakdown is problematic enough when it is done on a regular annual basis because auditors typically do not make a legal determination as to whether particular expenditures are chargeable. Instead, the auditors take the union’s characterization for granted and perform the simple accounting function of “ensur[ing] that the expenditures which the union claims it made for certain expenses were actually made for those expenses.” Andrews v. Education Assn. of Cheshire, 829 F.2d 335, 340 (CA2 1987). Thus, if a union takes a very broad view of what is chargeable—if, for example, it believes that supporting sympathetic political candidates is chargeable and bases its classification on that view—the auditors will classify these political expenditures as chargeable. Objecting employees may then contest the union’s chargeability determinations, but the onus is on the employees to come up with the resources to mount the legal challenge in a timely fashion.[8] See, e.g., Lehnert, 500 U. S., at 513; Jibson v. Michigan Ed. Assn., 30 F.3d 723, 730 (CA6 1994). This is already a significant burden for employees to bear simply to avoid having their money taken to subsidize speech with which they disagree, and the burden would become insupportable if unions could impose a new assessment at any time, with a new chargeability determination to be challenged. 2 The SEIU argues that objecting nonmembers who were required to pay 56.35% of the special assessment, far from subsidizing the union’s political campaign, actually received a windfall. According to the union’s statistics, the actual percentage of regular dues and fees spent for chargeable purposes in 2005 turned out to be quite a bit higher (66.26%), and therefore, even if all of the money obtained through the special assessment is classified as nonchargeable, the union’s total expenditures for 2005 were at least 66.26% chargeable. See Brief for Respond- ent 5, n. 6. This argument is unpersuasive for several reasons. First, the SEIU’s understanding of the breadth of charge- able expenses is so expansive that it is hard to place much reliance on its statistics. In its brief, the SEIU argues broadly that all funds spent on “lobbying . . . the electorate” are chargeable. See id., at 51. But “lobbying . . . the electorate” is nothing but another term for supporting political causes and candidates, and we have never held that the First Amendment permits a union to compel nonmembers to support such political activities. On the contrary, as long ago as Street, we noted the important difference between a union’s authority to engage in collective bargaining and related activities on behalf of nonmember employees in a bargaining unit and the union’s use of nonmembers’ money “to support candidates for public office” or “to support political causes which [they] oppos[e].” 367 U. S., at 768. The sweep of the SEIU’s argument is highlighted by its discussion of the use of fees paid by objecting nonmembers to defeat Proposition 76. According to the SEIU, these expenditures were “germane” to the implementation of its contracts because, if Proposition 76 had passed, it would have “effectively permitted the Governor to abrogate the Union’s collective bargaining agreements under certain circumstances, undermining the Union’s ability to perform its representation duty of negotiating effective collective bargaining agreements.” Brief for Respondent 49–50 (internal quotation marks omitted). If we were to accept this broad definition of germaneness, it would effectively eviscerate the limitation on the use of compulsory fees to support unions’ controversial political activities. Public-employee salaries, pensions, and other benefits constitute a substantial percentage of the budgets of many States and their subdivisions. As a result, a broad array of ballot questions and campaigns for public office may be said to have an effect on present and future contracts between public-sector workers and their employers. If the concept of “germaneness” were as broad as the SEIU advocates, public-sector employees who do not endorse the unions’ goals would be essentially unprotected against being compelled to subsidize political and ideological activities to which they object. Second, even if the SEIU’s statistics are accurate, it does not follow that it was proper for the union to charge objecting nonmembers 56.35%—or any other particular per- centage—of the special assessment. Unless it is possible to determine in advance with some degree of accuracy the percentage of union funds that will be used during an upcoming year for chargeable purposes—and the SEIU argues that this is not possible—there is at least a risk that, at the end of the year, unconsenting nonmembers will have paid either too much or too little. Which side should bear this risk? The answer is obvious: the side whose constitutional rights are not at stake. “Given the existence of acceptable alternatives, [a] union cannot be allowed to commit dissenters’ funds to improper uses even temporarily.” Ellis, 466 U. S., at 444. Thus, if unconsenting nonmembers pay too much, their First Amendment rights are infringed. On the other hand, if unconsenting nonmembers pay less than their proportionate share, no constitutional right of the union is violated because the union has no constitutional right to receive any payment from these employees. See Davenport, 551 U. S., at 185. The union has simply lost for a few months the “extraordinary” benefit of being em- powered to compel nonmembers to pay for services that they may not want and in any event have not agreed to fund. As we have noted, by allowing unions to collect any fees from nonmembers and by permitting unions to use opt-out rather than opt-in schemes when annual dues are billed, our cases have substantially impinged upon the First Amendment rights of nonmembers. In the new situation presented here, we see no justification for any further impingement. The general rule—individuals should not be compelled to subsidize private groups or private speech—should prevail. Public-sector unions have the right under the First Amendment to express their views on political and social issues without government interference. See, e.g., Citizens United v. Federal Election Comm’n, 558 U. S. ___ (2010). But employees who choose not to join a union have the same rights. The First Amendment creates a forum in which all may seek, without hindrance or aid from the State, to move public opinion and achieve their political goals. “ First Amendment values [would be] at serious risk if the government [could] compel a particular citizen, or a discrete group of citizens, to pay special subsidies for speech on the side that [the government] favors.” United Foods, 533 U. S., at 411. Therefore, when a public-sector union imposes a special assessment or dues increase, the union must provide a fresh Hudson notice and may not exact any funds from nonmembers without their affirmative consent.[9] * * * The judgment of the Ninth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 Marinucci & Wildermuth, Schwarzenegger Adds Prop. 75 to His Agenda, San Francisco Chronicle, Sept. 18, 2005, p. A–17. 2 Summers, Book Review, Sheldon Leader, Freedom of Association: A Study in Labor Law and Political Theory, 16 Comparative Labor L. J. 262, 268 (1995). 3 The specific citation was as follows: “See Roberts v. United States Jaycees, [ 468 U.S. 609, 623 (1984)] (Infringements on freedom of association ‘may be justified by regu-lations adopted to serve compelling state interests, unrelated to the suppression of ideas, that cannot be achieved through means significantly less restrictive of associational freedoms’); Elrod v. Burns, 427 U.S. 347, 363 (1976) (government means must be ‘least restrictiveof freedom of belief and association’); Kusper v. Pontikes, 414 U.S. 51, 58–59 (1973) (‘[E]ven when pursuing a legitimate interest, a State may not choose means that unnecessarily restrict constitutionally protected liberty’); NAACP v. Button, 371 U.S. 415, 438 (1963) (‘Precision of reg-ulation must be the touchstone’ in the First Amendment context).” Hudson, 475 U. S., at 303, n. 11. 4 The dissent suggests that the union gave fair notice because it announced at the beginning of the year that “ ‘[d]ues are subject to change without further notice to fee payers.’ ” Post, at 12 (opinion of Breyer, J.). But a union cannot define the scope of its own notice obligations simply by promulgating a clause giving itself the power to increase fees at any time for any purpose without further notice. 5 These nonmembers, after paying the full amount of the special assessment, would be required during the subsequent year to pay at least as much as those nonmembers who did opt out when they received the initial Hudson notice. 6 Justice Sotomayor contends that a new Hudson notice should be required only when a special assessment is imposed for political purposes. Post, at 2 (opinion concurring in judgment). But as even the dissent acknowledges, post, at 7, such a rule would be unworkable. First, our cases have recognized that a union’s money is fungible, so even if the new fee were spent entirely for nonpolitical activities, it would free up other funds to be spent for political purposes. See Retail Clerks v. Schermerhorn, 373 U.S. 746, 753 (1963) (noting that particular fee earmarks are “of bookkeeping significance only rather than a matter of real substance”). And second, unless we can rely on unions to advertise the true purpose behind every special fee, it is not clear how a court could make a timely determination of whether each new fee is political in nature. It would be practically impossible to require the parties to litigate the purpose of every fee merely to determine whether notice is required. 7 The SEIU contends that “[s]ignificant fluctuations in the chargeable and nonchargeable proportions of a union’s spending are inevitable,” Brief for Respondent 13, and the dissent appears to agree, post, at 10. But if the Hudson Court had proceeded on this assumption it is doubtful that it would have found it acceptable for a union to rely solely on the breakdown in the most recent year rather than computing the average breakdown over a longer period. 8 The dissent is comforted by the fact that the union “has offered to pay for neutral arbitration of such disputes before the American Arbitration Association,” post, at 9, but the painful burden of initiating and participating in such disputes cannot be so easily relieved. 9 Contrary to Justice Sotomayor’s suggestion, our holding does not venture beyond the scope of the questions on which we granted review or the scope of the parties’ dispute. The second question on which we granted review broadly asks us to determine the circumstances under which a State may deduct from the pay of nonunion employees money that is used by a union for general electioneering. See Pet. for Cert. (i) (“May a State, consistent with the First and Fourteenth Amendments, condition continued public employment on the payment of unionagency fees for purposes of financing political expenditures for ballot measures?”). Our holding—that this may be done only when the em-ployee affirmatively consents—falls within that question. Our holding also addresses the primary remaining dispute between the parties, namely, the particular procedures that must be followed on remand in order to provide adequate assurance that members ofthe class are not compelled to subsidize nonchargeable activities to which they object. See supra, at 7–8. Petitioners argue strenuously that these procedures must be narrowly tailored to minimize intrusion on their free-speech rights. See Brief for Petitioners 11–17. We see no sensible way to address this dispute without confronting the question whether, in the particular context present here, an opt-out regime suffices. Justice Sotomayor would apparently have us proceed on the assumption that an opt-out regime is permitted. She would then have us decide what sort of opt-out procedures would be sufficient if such a regime were allowed at all. But that is a question that simply cannot be answered. It would be like asking what sort of procedural requirements would be required if the government set out to do somethingelse that the First Amendment flatly prohibits—for example, requiring prepublication approval of newspapers. There is also no merit in Justice Sotomayor’s and Justice Breyer’s comments about prior precedent. This case concerns the procedures that must be followed when a public-sector union announces a special assessment or mid-year dues increase. No prior decision of this Court has addressed that question, and Hudson says not one word on the subject.

565.US.2011_10-879

George Corson worked as a welder and machinist for a railroad carrier. After retirement, Corson was diagnosed with mesothelioma. He and his wife, a petitioner here, sued respondents Railroad Friction Products Corporation and Viad Corp in state court, claiming injury from Corson’s exposure to asbestos in locomotives and locomotive parts distributed by respondents. The Corsons alleged state-law claims of defective design and failure to warn of the dangers posed by asbestos. After Corson died, petitioner Kurns, executrix of his estate, was substituted as a party. Respondents removed the case to the Federal District Court, which granted them summary judgment, ruling that the state-law claims were pre-empted by the Locomotive Inspection Act (LIA), 49 U. S. C. §20701 et seq. The Third Circuit affirmed. Held: Petitioners’ state-law design-defect and failure-to-warn claims fall within the field of locomotive equipment regulation pre-empted by the LIA, as that field was defined in Napier v. Atlantic Coast Line R. Co., 272 U.S. 605. Pp. 2−11. (a) The LIA provides that a railroad carrier may use or allow to be used a locomotive or tender on its railroad line only when the locomotive or tender and its parts or appurtenances are in proper condition and safe to operate without unnecessary danger of personal injury, have been inspected as required by the LIA and regulations prescribed thereunder by the Secretary of Transportation, and can withstand every test prescribed under the LIA by the Secretary. See §20701. Pp. 2–3. (b) Congress may expressly pre-empt state law. But even without an express pre-emption provision, state law must yield to a congressional Act to the extent of any conflict with a federal statute, see Crosby v. National Foreign Trade Council, 530 U.S. 363, 372, or when the federal statute’s scope indicates that Congress intended federal law to occupy a field exclusively, see Freightliner Corp. v. Myrick, 514 U.S. 280, 287. This case involves only the latter, so-called “field pre-emption.” Pp. 3–4. (c) In Napier, this Court held two state laws prescribing the use of locomotive equipment pre-empted by the LIA, concluding that the broad power conferred by the LIA on the Interstate Commerce Commission (the agency then vested with authority to carry out the LIA’s requirements) was a “general one” that “extends to the design, the construction and the material of every part of the locomotive and tender and of all appurtenances.” 272 U. S., at 611. The Court rejected the States’ contention that the scope of the pre-empted field was to “be determined by the object sought through legislation, rather than the physical elements affected by it,” id., at 612, and found it dispositive that “[t]he federal and state statutes are directed to the same subject―the equipment of locomotives.” Ibid. Pp. 4−5. (d) The Federal Railroad Safety Act of 1970 (FRSA) did not alter the LIA’s pre-emptive scope. By its terms, the FRSA—which instructs that “[t]he Secretary of Transportation . . . shall prescribe regulations and issue orders for every area of railroad safety supplementing laws and regulations in effect on October 16, 1970,” 49 U. S. C. §20103(a)—does not alter pre-existing federal railroad safety statutes. Rather, it leaves those statutes intact and authorizes the Secretary to fill interstitial areas of railroad safety with supplemental regulation. Because the LIA was already in effect when the FRSA was enacted, the FRSA left the LIA, and its pre-emptive scope as defined by Napier, intact. P. 6. (e) Petitioners do not argue that Napier should be overruled. Pp. 6–7. Instead, petitioners contend that their claims fall outside the LIA’s pre-empted field, as it was defined in Napier. Petitioners’ arguments are unpersuasive. First, the argument that the pre-empted field does not extend to state-law claims arising from the repair or maintenance of locomotives is inconsistent with Napier’s holding that Congress, in enacting the LIA, “manifest[ed] the intention to occupy the entire field of regulating locomotive equipment.” 272 U. S., at 611. Second, the argument that petitioners’ failure-to-warn claims are not pre-empted because they do not base liability on the design or manufacture of a product ignores that a failure-to-warn claim alleges that the product itself is defective unless accompanied by sufficient warnings or instructions. Because petitioners’ failure-to-warn claims are therefore directed at the equipment of locomotives, they fall within the pre-empted field defined by Napier. Third, the argument that petitioners’ claims are not pre-empted because manufacturers were not regulated under the LIA when Corson was exposed to asbestos is inconsistent with Napier, which defined the pre-empted field on the basis of the physical elements regulated, not on the basis of the entity directly subject to regulation. Finally, contrary to petitioners’ argument, the LIA’s pre-emptive scope is not limited to state legislation or regulation but extends to state common-law duties and standards of care directed to the subject of locomotive equipment. Pp. 6−11. 620 F.3d 392, affirmed. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Alito, and Kagan, JJ., joined. Kagan, J., filed a concurring opinion. Sotomayor, J., filed an opinion concurring in part and dissenting in part, in which Ginsburg and Breyer, JJ., joined.

This case requires us to determine whether petitioners’ state-law tort claims for defective design and failure to warn are pre-empted by the Locomotive Inspection Act (LIA), 49 U. S. C. §20701 et seq. The United States Court of Appeals for the Third Circuit determined that petitioners’ claims fall within the field pre-empted by that Act, as that field was defined by this Court’s decision in Napier v. Atlantic Coast Line R. Co., 272 U.S. 605 (1926). We agree. I George Corson was employed as a welder and machinist by the Chicago, Milwaukee, St. Paul & Pacific Railroad from 1947 until 1974. Corson worked in locomotive repair and maintenance facilities, where his duties included installing brakeshoes on locomotives and stripping insulation from locomotive boilers. In 2005, Corson was diagnosed with malignant mesothelioma. In 2007, Corson and his wife filed suit in Pennsylvania state court against 59 defendants, including respondents Railroad Friction Products Corporation (RFPC) and Viad Corp (Viad). According to the complaint, RFPC distributed locomotive brakeshoes containing asbestos, and Viad was the successor-in-interest to a company that manufactured and sold locomotives and locomotive engine valves containing asbestos. Corson alleged that he handled this equipment and that he was injured by exposure to asbestos. The complaint asserted state-law claims that the equipment was defectively designed because it contained asbestos, and that respondents failed to warn of the dangers of asbestos or to provide instructions regarding its safe use. After the complaint was filed, Corson passed away, and the executrix of his estate, Gloria Kurns, was substituted as a party. Corson’s widow and the executrix are petitioners here. Respondents removed the case to the United States District Court for the Eastern District of Pennsylvania and moved for summary judgment. Respondents argued that petitioners’ state-law claims were pre-empted by the LIA. The District Court agreed and granted summary judgment for respondents. See Kurns v. A. W. Chesterton, Civ. Action No. 08–2216 (ED Pa., Feb. 3, 2009), App. to Pet. for Cert. 39a. The Third Circuit affirmed. See Kurns v. A. W. Chesterton, Inc., 620 F.3d 392 (2010). We granted certiorari. 563 U. S. ___ (2011). II Congress enacted the predecessor to the LIA, the Boiler Inspection Act (BIA), in 1911. The BIA made it unlawful to use a steam locomotive “unless the boiler of said locomotive and appurtenances thereof are in proper condition and safe to operate . . . without unnecessary peril to life or limb.” Act of Feb. 17, 1911, ch. 103, §2, 36Stat. 913–914. In 1915, Congress amended the BIA to apply to “the entire locomotive and tender and all parts and appurtenances thereof.”[1] Act of Mar. 4, 1915, ch. 169, §1, 38Stat. 1192. The BIA as amended became commonly known as the Locomotive Inspection Act. As relevant here, the LIA provides: “A railroad carrier may use or allow to be used a locomotive or tender on its railroad line only when the locomotive or tender and its parts and appurtenances— “(1) are in proper condition and safe to operate without unnecessary danger of personal injury; “(2) have been inspected as required under this chapter and regulations prescribed by the Secretary of Transportation under this chapter; and “(3) can withstand every test prescribed by the Secretary under this chapter.” 49 U. S. C. §20701.[2] The issue presented in this case is whether the LIA pre-empts petitioners’ state-law claims that respondents defectively designed locomotive parts and failed to warn Corson of dangers associated with those parts. In light of this Court’s prior decision in Napier, supra, we conclude that petitioners’ claims are pre-empted. III A The Supremacy Clause provides that federal law “shall be the supreme Law of the Land . . . any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.” U. S. Const., Art. VI, cl. 2. Pre-emption of state law thus occurs through the “direct operation of the Supremacy Clause.” Brown v. Hotel Employees, 468 U.S. 491, 501 (1984). Congress may, of course, expressly pre-empt state law, but “[e]ven without an express provision for preemption, we have found that state law must yield to a congressional Act in at least two circumstances.” Crosby v. National Foreign Trade Council, 530 U.S. 363, 372 (2000). First, “state law is naturally preempted to the extent of any conflict with a federal statute.” Ibid. Second, we have deemed state law pre-empted “when the scope of a [federal] statute indicates that Congress intended federal law to occupy a field exclusively.” Freightliner Corp. v. Myrick, 514 U.S. 280, 287 (1995). We deal here only with the latter, so-called field pre-emption. B We do not, however, address the LIA’s pre-emptive ef- fect on a clean slate, because this Court addressed that issue 85 years ago in Napier. In that case, railroads challenged two state laws that “prohibit[ed] use within the State of locomotives not equipped with” certain prescribed devices, on the ground that the Interstate Commerce Commission (ICC), the agency then vested with the authority to carry out the LIA’s requirements, had not required the devices in question.[3] 272 U. S., at 607, 609. In response, the States argued that their requirements were not pre-empted because they were directed at a different objective than the LIA. Id., at 612. According to the States, their regulations were intended to protect railroad workers from sickness and disease, whereas “the federal regulation endeavors solely to prevent accidental injury in the operation of trains.” Ibid. To determine whether the state requirements were pre-empted, this Court asked whether the LIA “manifest[s] the intention to occupy the entire field of regulating locomotive equipment[.]” Id., at 611. The Court answered that question in the affirmative, stating that “[t]he broad scope of the authority conferred upon the [ICC]” by Congress in the LIA led to that conclusion. Id., at 613. The power delegated to the ICC, the Court explained, was a “general one” that “extends to the design, the construction and the material of every part of the locomotive and tender and of all appurtenances.” Id., at 611. The Court rejected the States’ contention that the scope of the pre-empted field was to “be determined by the object sought through the legislation, rather than the physical elements affected by it.” Id., at 612. The Court found it dispositive that “[t]he federal and the state statutes are directed to the same subject—the equipment of locomotives.” Ibid. Because the States’ requirements operated upon the same physical elements as the LIA, the Court held that the state laws, “however commendable or however different their purpose,” id., at 613, fell within the LIA’s pre-empted field. IV Against the backdrop of Napier, petitioners advance two arguments in support of their position that their state-law claims related to the use of asbestos in locomotive equipment do not fall within the LIA’s pre-empted field. Petitioners first contend that Napier no longer defines the scope of the LIA’s pre-empted field because that field has been narrowed by a subsequently enacted federal statute. Alternatively, petitioners argue that their claims do not fall within the LIA’s pre-empted field, even as that field was defined by Napier. We address each of petitioners’ arguments in turn. A First, petitioners suggest that the Federal Railroad Safety Act of 1970 (FRSA), 84Stat. 971 (codified at 49 U. S. C. §20102 et seq.), altered the LIA’s pre-emptive scope. The FRSA grants the Secretary of Transportation broad regulatory authority over railroad safety. See §20103(a). Petitioners point to the FRSA’s pre-emption provision, which provides in part that “[a] State may adopt or continue in force a law, regulation, or order related to railroad safety . . . until the Secretary of Transportation . . . prescribes a regulation or issues an order covering the subject matter of the State requirement.” §20106(a)(2) (2006 ed., Supp. III). According to petitioners, the FRSA’s pre-emption provision supplanted the LIA’s pre-emption of the field, with the result that petitioners’ claims are not pre-empted because the Secretary has not issued a regulation or order addressing the use of asbestos in locomotives or locomotive parts. Petitioners’ reliance on the FRSA is misplaced. The FRSA instructs that “[t]he Secretary of Transportation, as necessary, shall prescribe regulations and issue orders for every area of railroad safety supplementing laws and regulations in effect on October 16, 1970.” §20103(a) (2006 ed.) (emphasis added). By its terms, the FRSA does not alter pre-existing federal statutes on railroad safety. “Rather, it leaves existing statutes intact, . . . and authorizes the Secretary to fill interstitial areas of railroad safety with supplemental regulation.” Marshall v. Burlington Northern, Inc., 720 F.2d 1149, 1152–1153 (CA9 1983) (Kennedy, J.). Because the LIA was already in effect when the FRSA was enacted, we conclude that the FRSA left the LIA, and its pre-emptive scope as defined by Napier, intact. B Since the LIA’s pre-emptive scope remains unaltered, petitioners must contend with Napier. Petitioners do not ask us to overrule Napier and thus do not seek to overcome the presumption of stare decisis that attaches to this 85-year-old precedent. See Global-Tech Appliances, Inc. v. stocktickerSEB S. A., 563 U. S. ___, ___ (2011) (slip op., at 9) (noting the “special force of the doctrine of stare decisis with regard to questions of statutory interpretation” (internal quotation marks omitted)). Instead, petitioners advance several arguments aimed at demonstrating that their claims fall outside of the field pre-empted by the LIA, as it was defined in Napier. Each is unpersuasive. 1 Petitioners, along with the Solicitor General as amicus curiae, first argue that petitioners’ claims do not fall within the LIA’s pre-empted field because the claims arise out of the repair and maintenance of locomotives, rather than the use of locomotives on a railroad line. Specifically, they contend that the scope of the field pre-empted by the LIA is coextensive with the scope of the Federal Government’s regulatory authority under the LIA, which, they argue, does not extend to the regulation of hazards arising from the repair or maintenance of locomotives. Therefore, the argument goes, state-law claims arising from repair or maintenance—as opposed to claims arising from use on the line—do not fall within the pre-empted field. We reject this attempt to redefine the pre-empted field. In Napier, the Court held that Congress, in enacting the LIA, “manifest[ed] the intention to occupy the entire field of regulating locomotive equipment,” and the Court did not distinguish between hazards arising from repair and maintenance as opposed to those arising from use on the line. 272 U. S., at 611. The pre-empted field as defined by Napier plainly encompasses the claims at issue here. Petitioners’ common-law claims for defective design and failure to warn are aimed at the equipment of locomotives. Because those claims “are directed to the same subject” as the LIA, Napier dictates that they fall within the pre-empted field. Id., at 612. 2 Petitioners further argue that, even if their design-defect claims are pre-empted, their failure-to-warn claims do not suffer the same fate. In their complaint, petitioners alleged in closely related claims (1) that respondents negligently failed to warn of the risks associated with asbestos and to provide instructions concerning safe- guards for working with asbestos; and (2) that the asbestos- containing products were defective because respondents failed to give sufficient warnings or instructions con- cerning the “risks, dangers, and harm inherent in said asbestos products.” See App. 20–27 (¶¶7–10, 12), 42 (¶8); see also Brief for Petitioners 11. According to petitioners, these claims do not fall within the LIA’s pre-empted field because “[t]he basis of liability for failure to warn . . . is not the ‘design’ or ‘manufacture’ of a product,” but is instead “the failure to provide adequate warnings regarding the product’s risks.” Reply Brief for Petitioners 16. We disagree. A failure-to-warn claim alleges that the product itself is unlawfully dangerous unless accompanied by sufficient warnings or instructions. Restatement (Third) of Torts: Products Liability §2(c) (1997) (A failure-to-warn claim alleges that a product is defective “when the foreseeable risks of harm posed by the product could have been reduced or avoided by the provision of reasonable instructions or warnings by the seller or other distributor, . . . and the omission of the instructions or warnings renders the product not reasonably safe”); see also id., Comment l, at 33 (“Reasonable designs and instructions or warnings both play important roles in the production and distribution of reasonably safe products”). Thus, the “gravamen” of petitioners’ failure-to-warn claims “is still that [Corson] suffered harmful consequences as a result of his exposure to asbestos contained in locomotive parts and appurtenances.” 620 F. 3d, at 398, n. 8. Because petitioners’ failure-to-warn claims are therefore directed at the equipment of locomotives, they fall within the pre-empted field defined by Napier. 272 U. S., at 612.[4] 3 Petitioners also contend that their state-law claims against manufacturers of locomotives and locomotive parts fall outside of the LIA’s pre-empted field because manufacturers were not regulated under the LIA at the time that Corson was allegedly exposed to asbestos. Petitioners point out that the LIA, as originally enacted in the BIA, subjected only common carriers to civil penalties. Act of dateFeb. 17, 1911, §9, 36Stat. 916. It was not until 1988, well after the events of this case, that the LIA’s penalty provision was revised to apply to “[a]ny person” violating the LIA. Rail Safety Improvement Act of 1988, §14(7)(A), 102Stat. 633; see also §14(7)(B) (amending penalty provision to provide that “an act by an individual that causes a railroad to be in violation . . . shall be deemed a violation”). This argument fails for the same reason as the two preceding arguments: It is inconsistent with Napier. Napier defined the field pre-empted by the LIA on the basis of the physical elements regulated—“the equipment of locomotives”—not on the basis of the entity directly subject to regulation. 272 U. S., at 612. Because petitioners’ claims are directed at the equipment of locomotives, they fall within the pre-empted field. Petitioners’ proposed rule is also contrary to common sense. Under petitioners’ approach, a State could not require railroads to equip their locomotives with parts meeting state-imposed specifications, but could require manufacturers of locomotive parts to produce only parts meeting those state-imposed specifications. We rejected a similar approach in an express pre-emption context in Engine Mfrs. Assn. v. South Coast Air Quality Management Dist., 541 U.S. 246 (2004). There, a state entity argued that its rules prohibiting the purchase or lease of vehicles that failed to meet stringent emissions requirements were not pre-empted by the Clean Air Act, 42 U. S. C. §7543(a), because the rules in question were aimed at the purchase of vehicles, rather than their manufacture or sale. 541 U. S., at 248. We observed, however, that “treating sales restrictions and purchase restrictions differently for pre-emption purposes would make no sense,” because the “manufacturer’s right to sell federally approved vehicles is meaningless in the absence of a purchaser’s right to buy them.” Id., at 255. Similarly, a railroad’s ability to equip its fleet of locomotives in compliance with federal standards is meaningless if manufacturers are not allowed to produce locomotives and locomotive parts that meet those standards. Petitioners’ claims thus do not avoid pre-emption simply because they are aimed at the manufacturers of locomotives and locomotive parts. 4 Finally, petitioners contend that the LIA’s pre-emptive scope does not extend to state common-law claims, as opposed to state legislation or regulation. Petitioners note that “a preempted field does not necessarily include state common law.” Brief for Petitioners 38–39 (citing Silkwood v. Kerr-McGee Corp., 464 U.S. 238 (1984); Sprietsma v. Mercury Marine, 537 U.S. 51 (2002)). Napier, however, held that the LIA “occup[ied] the entire field of regulating locomotive equipment” to the exclusion of state regulation. 272 U. S., at 611–612. That categorical conclusion admits of no exception for state common-law duties and standards of care. As we have recognized, state “regulation can be . . . effectively exerted through an award of damages,” and “[t]he obligation to pay compensation can be, indeed is designed to be, a potent method of governing conduct and controlling policy.” San Diego Building Trades Council v. Garmon, 359 U.S. 236, 247 (1959). Cf. Riegel v. Med- tronic, Inc., 552 U.S. 312, 324 (2008) (“Absent other in- dication, reference to a State’s ‘requirements’ [in a federal express pre-emption provision] includes its common-law duties”). We therefore conclude that state common-law duties and standards of care directed to the subject of locomotive equipment are pre-empted by the LIA. * * * For the foregoing reasons, we hold that petitioners’ state-law design-defect and failure-to-warn claims fall within the field of locomotive equipment regulation pre-empted by the LIA, as that field was defined in Napier. Accordingly, the judgment of the Court of Appeals is affirmed. It is so ordered. Notes 1 A “tender” is a “[a] car attached to a locomotive, for carrying a supply of fuel and water.” Webster’s New International Dictionary of the English Language 2126 (1917). 2 At the time of Corson’s employment, this provision of the LIA was worded somewhat differently. See 45 U. S. C. §23 (1946 ed.). Petitioners do not argue that the change in statutory language makes any difference in this case. 3 Act of Feb. 17, 1911, §6, 36Stat. 915. That authority has sincebeen transferred to the Secretary of Transportation. Department of Transportation Act, §§6(e)(1)(E) and (F), 80Stat. 939; see 49 U. S. C. §§20701–20702. 4 Justice Sotomayor apparently agrees that petitioners’ failure-to-warn claims are directed at the equipment of locomotives. Post, at 5 (opinion concurring in part and dissenting in part). Yet, she argues, those claims affect locomotive equipment only “ ‘tangentially.’ ” Ibid. (quoting English v. General Elec. Co., 496 U.S. 72, 85 (1990)). Notso. A failure-to-warn claim imposes liability on a particular design of locomotive equipment unless warnings deemed sufficient under state law are given. This duty to warn and the accompanying threat of liability will inevitably influence a manufacturer’s choice whether to use that particular design. By influencing design decisions in that manner, failure-to-warn liability has a “ ‘direct and substantial effect’ ” on the “physical elements” of a locomotive. Post, at 5 (quoting English, supra, at 85).

566.US.2011_10-209

Respondent was charged under Michigan law with assault with intent to murder and three other offenses. The prosecution offered to dismiss two of the charges and to recommend a 51-to-85-month sentence on the other two, in exchange for a guilty plea. In a communication with the court, respondent admitted his guilt and expressed a willingness to accept the offer. But he rejected the offer, allegedly after his attorney convinced him that the prosecution would be unable to establish intent to murder because the victim had been shot below the waist. At trial, respondent was convicted on all counts and received a mandatory minimum 185-to-360-month sentence. In a subsequent hearing, the state trial court rejected respondent’s claim that his attorney’s advice to reject the plea constituted ineffective assistance. The Michigan Court of Appeals affirmed, rejecting the ineffective-assistance claim on the ground that respondent knowingly and intelligently turned down the plea offer and chose to go to trial. Respondent renewed his claim in federal habeas. Finding that the state appellate court had unreasonably applied the constitutional effective-assistance standards laid out in Strickland v. Washington, 466 U.S. 668, and Hill v. Lockhart, 474 U.S. 52, the District Court granted a conditional writ and ordered specific performance of the original plea offer. The Sixth Circuit affirmed. Applying Strickland, it found that counsel had provided deficient performance by advising respondent of an incorrect legal rule, and that respondent suffered prejudice because he lost the opportunity to take the more favorable sentence offered in the plea. Held: 1. Where counsel’s ineffective advice led to an offer’s rejection, and where the prejudice alleged is having to stand trial, a defendant must show that but for the ineffective advice, there is a reasonable probability that the plea offer would have been presented to the court, that the court would have accepted its terms, and that the conviction or sentence, or both, under the offer’s terms would have been less severe than under the actual judgment and sentence imposed. Pp. 3–11. (a) Because the parties agree that counsel’s performance was deficient, the only question is how to apply Strickland’s prejudice test where ineffective assistance results in a rejection of the plea offer and the defendant is convicted at the ensuing trial. Pp. 3–4. (b) In that context, the Strickland prejudice test requires a defendant to show a reasonable possibility that the outcome of the plea process would have been different with competent advice. The Sixth Circuit and other federal appellate courts have agreed with the Strickland prejudice test for rejected pleas adopted here by this Court. Petitioner and the Solicitor General propose a narrow view—that Strickland prejudice cannot arise from plea bargaining if the defendant is later convicted at a fair trial—but their reasoning is unpersuasive. First, they claim that the Sixth Amendment’s sole purpose is to protect the right to a fair trial, but the Amendment actually requires effective assistance at critical stages of a criminal proceeding, including pretrial stages. This is consistent with the right to effective assistance on appeal, see, e.g., Halbert v. Michigan, 545 U.S. 605, and the right to counsel during sentencing, see, e.g., Glover v. United States, 531 U.S. 198, 203–204. This Court has not followed a rigid rule that an otherwise fair trial remedies errors not occurring at trial, but has instead inquired whether the trial cured the particular error at issue. See, e.g., Vasquez v. Hillery, 474 U.S. 254, 263. Second, this Court has previously rejected petitioner’s argument that Lockhart v. Fretwell, 506 U.S. 364, modified Strickland and does so again here. Fretwell and Nix v. Whiteside, 475 U.S. 157, demonstrate that “it would be unjust to characterize the likelihood of a different outcome as legitimate ‘prejudice,’ ” Williams v. Taylor, 529 U.S. 362, 391–392, where defendants would receive a windfall as a result of the application of an incorrect legal principle or a defense strategy outside the law. Here, however, respondent seeks relief from counsel’s failure to meet a valid legal standard. Third, petitioner seeks to preserve the conviction by arguing that the Sixth Amendment’s purpose is to ensure a conviction’s reliability, but this argument fails to comprehend the full scope of the Sixth Amendment and is refuted by precedent. Here, the question is the fairness or reliability not of the trial but of the processes that preceded it, which caused respondent to lose benefits he would have received but for counsel’s ineffective assistance. Furthermore, a reliable trial may not foreclose relief when counsel has failed to assert rights that may have altered the outcome. See Kimmelman v. Morrison, 477 U.S. 365, 379. Petitioner’s position that a fair trial wipes clean ineffective assistance during plea bargaining also ignores the reality that criminal justice today is for the most part a system of pleas, not a system of trials. See Missouri v. Frye, ante, at ___. Pp. 4–11. 2. Where a defendant shows ineffective assistance has caused the rejection of a plea leading to a more severe sentence at trial, the remedy must “neutralize the taint” of a constitutional violation, United States v. Morrison, 449 U.S. 361, 365, but must not grant a windfall to the defendant or needlessly squander the resources the State properly invested in the criminal prosecution, see United States v. Mechanik, 475 U.S. 66, 72. If the sole advantage is that the defendant would have received a lesser sentence under the plea, the court should have an evidentiary hearing to determine whether the defendant would have accepted the plea. If so, the court may exercise discretion in determining whether the defendant should receive the term offered in the plea, the sentence received at trial, or something in between. However, resentencing based on the conviction at trial may not suffice, e.g., where the offered guilty plea was for less serious counts than the ones for which a defendant was convicted after trial, or where a mandatory sentence confines a judge’s sentencing discretion. In these circumstances, the proper remedy may be to require the prosecution to reoffer the plea. The judge can then exercise discretion in deciding whether to vacate the conviction from trial and accept the plea, or leave the conviction undisturbed. In either situation, a court must weigh various factors. Here, it suffices to give two relevant considerations. First, a court may take account of a defendant’s earlier expressed willingness, or unwillingness, to accept responsibility for his or her actions. Second, it is not necessary here to decide as a constitutional rule that a judge is required to disregard any information concerning the crime discovered after the plea offer was made. Petitioner argues that implementing a remedy will open the floodgates to litigation by defendants seeking to unsettle their convictions, but in the 30 years that courts have recognized such claims, there has been no indication that the system is overwhelmed or that defendants are receiving windfalls as a result of strategically timed Strickland claims. In addition, the prosecution and trial courts may adopt measures to help ensure against meritless claims. See Frye, ante, at ___. Pp. 11–14. 3. This case arises under the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), but because the Michigan Court of Appeals’ analysis of respondent’s ineffective-assistance-of-counsel claim was contrary to clearly established federal law, AEDPA presents no bar to relief. Respondent has satisfied Strickland’s two-part test. The parties concede the fact of deficient performance. And respondent has shown that but for that performance there is a reasonable probability he and the trial court would have accepted the guilty plea. In addition, as a result of not accepting the plea and being convicted at trial, he received a minimum sentence 3½ times greater than he would have received under the plea. As a remedy, the District Court ordered specific performance of the plea agreement, but the correct remedy is to order the State to reoffer the plea. If respondent accepts the offer, the state trial court can exercise its discretion in determining whether to vacate respondent’s convictions and resentence pursuant to the plea agreement, to vacate only some of the convictions and resentence accordingly, or to leave the conviction and sentence resulting from the trial undisturbed. Pp. 14–16. 376 Fed. Appx. 563, vacated and remanded. Kennedy, J., delivered the opinion of the Court, in which Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined. Scalia, J., filed a dissenting opinion, in which Thomas, J., joined, and in which Roberts, C. J., joined as to all but Part IV. Alito, J., filed a dissenting opinion.

In this case, as in Missouri v. Frye, ante, p. ___, also decided today, a criminal defendant seeks a remedy when inadequate assistance of counsel caused nonacceptance of a plea offer and further proceedings led to a less favorable outcome. In Frye, defense counsel did not inform the defendant of the plea offer; and after the offer lapsed the defendant still pleaded guilty, but on more severe terms. Here, the favorable plea offer was reported to the client but, on advice of counsel, was rejected. In Frye there was a later guilty plea. Here, after the plea offer had been rejected, there was a full and fair trial before a jury. After a guilty verdict, the defendant received a sentence harsher than that offered in the rejected plea bargain. The instant case comes to the Court with the concession that counsel’s advice with respect to the plea offer fell below the standard of adequate assistance of counsel guaranteed by the Sixth Amendment, applicable to the States through the Fourteenth Amendment. I On the evening of March 25, 2003, respondent pointed a gun toward Kali Mundy’s head and fired. From the record, it is unclear why respondent did this, and at trial it was suggested that he might have acted either in self-defense or in defense of another person. In any event the shot missed and Mundy fled. Respondent followed in pur- suit, firing repeatedly. Mundy was shot in her buttock, hip, and abdomen but survived the assault. Respondent was charged under Michigan law with as- sault with intent to murder, possession of a firearm by a felon, possession of a firearm in the commission of a fel- ony, misdemeanor possession of marijuana, and for being a habitual offender. On two occasions, the prosecution offered to dismiss two of the charges and to recommend a sentence of 51 to 85 months for the other two, in exchange for a guilty plea. In a communication with the court respondent admitted guilt and expressed a willingness to accept the offer. Respondent, however, later rejected the offer on both occasions, allegedly after his attorney convinced him that the prosecution would be unable to establish his intent to murder Mundy because she had been shot below the waist. On the first day of trial the prosecution offered a significantly less favorable plea deal, which respondent again rejected. After trial, respondent was convicted on all counts and received a mandatory minimum sentence of 185 to 360 months’ imprisonment. In a so-called Ginther hearing before the state trial court, see People v. Ginther, 390 Mich. 436, 212 N.W.2d 922 (1973), respondent argued his attorney’s advice to reject the plea constituted ineffective assistance. The trial judge rejected the claim, and the Michigan Court of Appeals affirmed. People v. Cooper, No. 250583, 2005 WL 599740 (Mar. 15, 2005) (per curiam), App. to Pet. for Cert. 44a. The Michigan Court of Appeals rejected the claim of ineffective assistance of counsel on the ground that re- spondent knowingly and intelligently rejected two plea offers and chose to go to trial. The Michigan Supreme Court denied respondent’s application for leave to file an appeal. People v. Cooper, 474 Mich. 905, 705 N.W.2d 118 (2005) (table). Respondent then filed a petition for federal habeas relief under 28 U. S. C. §2254, renewing his ineffective-assistance-of-counsel claim. After finding, as required by the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), that the Michigan Court of Appeals had un- reasonably applied the constitutional standards for effective assistance of counsel laid out in Strickland v. Washington, 466 U.S. 668 (1984), and Hill v. Lockhart, 474 U.S. 52 (1985), the District Court granted a conditional writ. Cooper v. Lafler, No. 06–11068, 2009 WL 817712, *10 (ED Mich., Mar. 26, 2009), App. to Pet. for Cert. 41a–42a. To remedy the violation, the District Court ordered “specific performance of [respondent’s] original plea agreement, for a minimum sentence in the range of fifty-one to eighty-five months.” Id., at *9, App. to Pet. for Cert. 41a. The United States Court of Appeals for the Sixth Circuit affirmed, 376 Fed. Appx. 563 (2010), finding “[e]ven full deference under AEDPA cannot salvage the state court’s decision,” id., at 569. Applying Strickland, the Court of Appeals found that respondent’s attorney had provided deficient performance by informing respondent of “an incorrect legal rule,” 376 Fed. Appx., at 570–571, and that respondent suffered prejudice because he “lost out on an opportunity to plead guilty and receive the lower sentence that was offered to him.” Id., at 573. This Court granted certiorari. 562 U. S. ___ (2011). II A Defendants have a Sixth Amendment right to counsel, a right that extends to the plea-bargaining process. Frye, ante, at 8; see also Padilla v. Kentucky, 559 U. S. ___, ___ (2010) (slip op., at 16); Hill, supra, at 57. During plea negotiations defendants are “entitled to the effective assistance of competent counsel.” McMann v. Richardson, 397 U.S. 759, 771 (1970). In Hill, the Court held “the two-part Strickland v. Washington test applies to challenges to guilty pleas based on ineffective assistance of counsel.” 474 U. S., at 58. The performance prong of Strickland requires a defendant to show “ ‘that counsel’s representation fell below an objective standard of reasonableness.’ ” 474 U. S., at 57 (quoting Strickland, 466 U. S., at 688). In this case all parties agree the performance of respondent’s counsel was deficient when he advised respondent to reject the plea offer on the grounds he could not be convicted at trial. In light of this concession, it is unnecessary for this Court to explore the issue. The question for this Court is how to apply Strickland’s prejudice test where ineffective assistance results in a re- jection of the plea offer and the defendant is convicted at the ensuing trial. B To establish Strickland prejudice a defendant must “show that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Id., at 694. In the context of pleas a defendant must show the outcome of the plea process would have been different with competent advice. See Frye, ante, at 12 (noting that Strickland’s inquiry, as applied to advice with respect to plea bargains, turns on “whether ‘the result of the proceeding would have been different’ ” (quoting Strickland, supra, at 694)); see also Hill, 474 U. S., at 59 (“The . . . ‘prejudice,’ requirement . . . focuses on whether counsel’s constitutionally ineffective performance affected the outcome of the plea process”). In Hill, when evaluating the petitioner’s claim that ineffective assistance led to the improvident acceptance of a guilty plea, the Court required the petitioner to show “that there is a reasonable probability that, but for counsel’s errors, [the defendant] would not have pleaded guilty and would have insisted on going to trial.” Ibid. In contrast to Hill, here the ineffective advice led not to an offer’s acceptance but to its rejection. Having to stand trial, not choosing to waive it, is the prejudice alleged. In these circumstances a defendant must show that but for the ineffective advice of counsel there is a reasonable prob- ability that the plea offer would have been presented to the court (i.e., that the defendant would have accepted the plea and the prosecution would not have withdrawn it in light of intervening circumstances), that the court would have accepted its terms, and that the conviction or sentence, or both, under the offer’s terms would have been less severe than under the judgment and sentence that in fact were imposed. Here, the Court of Appeals for the Sixth Circuit agreed with that test for Strickland prejudice in the context of a rejected plea bargain. This is consistent with the test adopted and applied by other appellate courts without demonstrated difficulties or systemic disruptions. See 376 Fed. Appx., at 571–573; see also, e.g., United States v. Rodriguez Rodriguez, 929 F.2d 747, 753, n. 1 (CA1 1991) (per curiam); United States v. Gordon, 156 F.3d 376, 380–381 (CA2 1998) (per curiam); United States v. Day, 969 F.2d 39, 43–45 (CA3 1992); Beckham v. Wainwright, 639 F.2d 262, 267 (CA5 1981); Julian v. Bartley, 495 F.3d 487, 498–500 (CA7 2007); Wanatee v. Ault, 259 F.3d 700, 703–704 (CA8 2001); Nunes v. Mueller, 350 F.3d 1045, 1052–1053 (CA9 2003); Williams v. Jones, 571 F.3d 1086, 1094–1095 (CA10 2009) (per curiam); United States v. Gaviria, 116 F.3d 1498, 1512–1514 (CADC 1997) (per curiam). Petitioner and the Solicitor General propose a different, far more narrow, view of the Sixth Amendment. They contend there can be no finding of Strickland prejudice arising from plea bargaining if the defendant is later convicted at a fair trial. The three reasons petitioner and the Solicitor General offer for their approach are unpersuasive. First, petitioner and the Solicitor General claim that the sole purpose of the Sixth Amendment is to protect the right to a fair trial. Errors before trial, they argue, are not cognizable under the Sixth Amendment unless they affect the fairness of the trial itself. See Brief for Petitioner 12–21; Brief for United States as Amicus Curiae 10–12. The Sixth Amendment, however, is not so narrow in its reach. Cf. Frye, ante, at 11 (holding that a defendant can show prejudice under Strickland even absent a showing that the deficient performance precluded him from going to trial). The Sixth Amendment requires effective assistance of counsel at critical stages of a criminal proceeding. Its protections are not designed simply to protect the trial, even though “counsel’s absence [in these stages] may derogate from the accused’s right to a fair trial.” United States v. Wade, 388 U.S. 218, 226 (1967). The constitutional guarantee applies to pretrial critical stages that are part of the whole course of a criminal proceeding, a proceeding in which defendants cannot be presumed to make critical decisions without counsel’s advice. This is consistent, too, with the rule that defendants have a right to effective assistance of counsel on appeal, even though that cannot in any way be characterized as part of the trial. See, e.g., Halbert v. Michigan, 545 U.S. 605 (2005); Evitts v. Lucey, 469 U.S. 387 (1985). The precedents also establish that there exists a right to counsel during sentencing in both noncapital, see Glover v. United States, 531 U.S. 198, 203–204 (2001); Mempa v. Rhay, 389 U.S. 128 (1967), and capital cases, see Wiggins v. Smith, 539 U.S. 510, 538 (2003). Even though sentencing does not concern the defendant’s guilt or innocence, ineffective assistance of counsel during a sentencing hearing can result in Strickland prejudice because “any amount of [additional] jail time has Sixth Amendment significance.” Glover, supra, at 203. The Court, moreover, has not followed a rigid rule that an otherwise fair trial remedies errors not occurring at the trial itself. It has inquired instead whether the trial cured the particular error at issue. Thus, in Vasquez v. Hillery, 474 U.S. 254 (1986), the deliberate exclusion of all African-Americans from a grand jury was prejudicial be- cause a defendant may have been tried on charges that would not have been brought at all by a properly constituted grand jury. Id., at 263; see Ballard v. United States, 329 U.S. 187, 195 (1946) (dismissing an indictment returned by a grand jury from which women were excluded); see also Stirone v. United States, 361 U.S. 212, 218–219 (1960) (reversing a defendant’s conviction because the jury may have based its verdict on acts not charged in the indictment). By contrast, in United States v. Mechanik, 475 U.S. 66 (1986), the complained-of error was a violation of a grand jury rule meant to ensure probable cause existed to believe a defendant was guilty. A subsequent trial, resulting in a verdict of guilt, cured this error. See id., at 72–73. In the instant case respondent went to trial rather than accept a plea deal, and it is conceded this was the result of ineffective assistance during the plea negotiation process. Respondent received a more severe sentence at trial, one 3½ times more severe than he likely would have received by pleading guilty. Far from curing the error, the trial caused the injury from the error. Even if the trial itself is free from constitutional flaw, the defendant who goes to trial instead of taking a more favorable plea may be prejudiced from either a conviction on more serious counts or the imposition of a more severe sentence. Second, petitioner claims this Court refined Strickland’s prejudice analysis in Fretwell to add an additional requirement that the defendant show that ineffective assistance of counsel led to his being denied a substantive or procedural right. Brief for Petitioner 12–13. The Court has rejected the argument that Fretwell modified Strickland before and does so again now. See Williams v. Taylor, 529 U.S. 362, 391 (2000) (“The Virginia Supreme Court erred in holding that our decision in Lockhart v. Fretwell, 506 U.S. 364 (1993), modified or in some way supplanted the rule set down in Strickland”); see also Glover, supra, at 203 (“The Court explained last Term [in Williams] that our holding in Lockhart does not supplant the Strickland analysis”). Fretwell could not show Strickland prejudice resulting from his attorney’s failure to object to the use of a sentencing factor the Eighth Circuit had erroneously (and temporarily) found to be impermissible. Fretwell, 506 U. S., at 373. Because the objection upon which his ineffective-assistance-of-counsel claim was premised was meritless, Fretwell could not demonstrate an error entitling him to relief. The case presented the “unusual circumstance where the defendant attempts to demonstrate prejudice based on considerations that, as a matter of law, ought not inform the inquiry.” Ibid. (O’Connor, J., concurring). See also ibid. (recognizing “[t]he determinative question—whether there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different—remains unchanged” (internal quotation marks and citation omitted)). It is for this same reason a defendant cannot show prejudice based on counsel’s refusal to present perjured testimony, even if such testimony might have affected the outcome of the case. See Nix v. Whiteside, 475 U.S. 157, 175 (1986) (holding first that counsel’s refusal to present perjured testimony breached no professional duty and second that it cannot establish prejudice under Strickland). Both Fretwell and Nix are instructive in that they demonstrate “there are also situations in which it would be unjust to characterize the likelihood of a different outcome as legitimate ‘prejudice,’ ” Williams, supra, at 391–392, because defendants would receive a windfall as a result of the application of an incorrect legal principle or a defense strategy outside the law. Here, however, the injured client seeks relief from counsel’s failure to meet a valid legal standard, not from counsel’s refusal to violate it. He maintains that, absent ineffective counsel, he would have accepted a plea offer for a sentence the prosecution evidently deemed consistent with the sound administration of criminal justice. The favorable sentence that eluded the defendant in the criminal proceeding appears to be the sentence he or others in his position would have received in the ordinary course, absent the failings of counsel. See Bibas, Regulating the Plea-Bargaining Market: From Caveat Emptor to Consumer Protection, 99 Cal. L. Rev. 1117, 1138 (2011) (“The expected post-trial sentence is imposed in only a few percent of cases. It is like the sticker price for cars: only an ignorant, ill-advised consumer would view full price as the norm and anything less a bargain”); see also Frye, ante, at 7–8. If a plea bargain has been offered, a defendant has the right to effective assistance of counsel in considering whether to accept it. If that right is denied, prejudice can be shown if loss of the plea opportunity led to a trial resulting in a conviction on more serious charges or the imposition of a more severe sentence. It is, of course, true that defendants have “no right to be offered a plea . . . nor a federal right that the judge accept it.” Frye, ante, at 12. In the circumstances here, that is beside the point. If no plea offer is made, or a plea deal is accepted by the defendant but rejected by the judge, the issue raised here simply does not arise. Much the same reasoning guides cases that find criminal defendants have a right to effective assistance of counsel in direct appeals even though the Constitution does not require States to provide a system of appellate review at all. See Evitts, 469 U.S. 387; see also Douglas v. California, 372 U.S. 353 (1963). As in those cases, “[w]hen a State opts to act in a field where its action has significant discretionary elements, it must nonetheless act in accord with the dictates of the Constitution.” Evitts, supra, at 401. Third, petitioner seeks to preserve the conviction obtained by the State by arguing that the purpose of the Sixth Amendment is to ensure “the reliability of [a] conviction following trial.” Brief for Petitioner 13. This argument, too, fails to comprehend the full scope of the Sixth Amendment’s protections; and it is refuted by precedent. Strickland recognized “[t]he benchmark for judging any claim of ineffectiveness must be whether counsel’s conduct so undermined the proper functioning of the adversarial process that the trial cannot be relied on as having produced a just result.” 466 U. S., at 686. The goal of a just result is not divorced from the reliability of a conviction, see United States v. Cronic, 466 U.S. 648, 658 (1984); but here the question is not the fairness or reliability of the trial but the fairness and regularity of the processes that preceded it, which caused the defendant to lose benefits he would have received in the ordinary course but for counsel’s ineffective assistance. There are instances, furthermore, where a reliable trial does not foreclose relief when counsel has failed to assert rights that may have altered the outcome. In Kimmelman v. Morrison, 477 U.S. 365 (1986), the Court held that an attorney’s failure to timely move to suppress evidence during trial could be grounds for federal habeas relief. The Court rejected the suggestion that the “failure to make a timely request for the exclusion of illegally seized evidence” could not be the basis for a Sixth Amendment violation because the evidence “is ‘typically reliable and often the most probative information bearing on the guilt or innocence of the defendant.’ ” Id., at 379 (quoting Stone v. Powell, 428 U.S. 465, 490 (1976)). “The constitutional rights of criminal defendants,” the Court observed, “are granted to the innocent and the guilty alike. Consequently, we decline to hold either that the guarantee of effective assistance of counsel belongs solely to the innocent or that it attaches only to matters affecting the determination of actual guilt.” 477 U. S., at 380. The same logic applies here. The fact that respondent is guilty does not mean he was not entitled by the Sixth Amendment to effective assistance or that he suffered no prejudice from his attorney’s deficient performance during plea bargaining. In the end, petitioner’s three arguments amount to one general contention: A fair trial wipes clean any deficient performance by defense counsel during plea bargaining. That position ignores the reality that criminal justice today is for the most part a system of pleas, not a system of trials. Ninety-seven percent of federal convictions and ninety-four percent of state convictions are the result of guilty pleas. See Frye, ante, at 7. As explained in Frye, the right to adequate assistance of counsel cannot be defined or enforced without taking account of the central role plea bargaining plays in securing convictions and determining sentences. Ibid. (“[I]t is insufficient simply to point to the guarantee of a fair trial as a backstop that inoculates any errors in the pretrial process”). C Even if a defendant shows ineffective assistance of counsel has caused the rejection of a plea leading to a trial and a more severe sentence, there is the question of what constitutes an appropriate remedy. That question must now be addressed. Sixth Amendment remedies should be “tailored to the injury suffered from the constitutional violation and should not unnecessarily infringe on competing interests.” United States v. Morrison, 449 U.S. 361, 364 (1981). Thus, a remedy must “neutralize the taint” of a constitutional violation, id., at 365, while at the same time not grant a windfall to the defendant or needlessly squander the considerable resources the State properly invested in the criminal prosecution. See Mechanik, 475 U. S., at 72 (“The reversal of a conviction entails substantial social costs: it forces jurors, witnesses, courts, the prosecution, and the defendants to expend further time, energy, and other resources to repeat a trial that has already once taken place; victims may be asked to relive their disturbing experiences”). The specific injury suffered by defendants who decline a plea offer as a result of ineffective assistance of counsel and then receive a greater sentence as a result of trial can come in at least one of two forms. In some cases, the sole advantage a defendant would have received under the plea is a lesser sentence. This is typically the case when the charges that would have been admitted as part of the plea bargain are the same as the charges the defendant was convicted of after trial. In this situation the court may conduct an evidentiary hearing to determine whether the defendant has shown a reasonable probability that but for counsel’s errors he would have accepted the plea. If the showing is made, the court may exercise discretion in determining whether the defendant should receive the term of imprisonment the government offered in the plea, the sentence he received at trial, or something in between. In some situations it may be that resentencing alone will not be full redress for the constitutional injury. If, for example, an offer was for a guilty plea to a count or counts less serious than the ones for which a defendant was convicted after trial, or if a mandatory sentence confines a judge’s sentencing discretion after trial, a resentencing based on the conviction at trial may not suffice. See, e.g., Williams, 571 F. 3d, at 1088; Riggs v. Fairman, 399 F.3d 1179, 1181 (CA9 2005). In these circumstances, the proper exercise of discretion to remedy the constitutional injury may be to require the prosecution to reoffer the plea proposal. Once this has occurred, the judge can then exercise discretion in deciding whether to vacate the conviction from trial and accept the plea or leave the conviction undisturbed. In implementing a remedy in both of these situations, the trial court must weigh various factors; and the boundaries of proper discretion need not be defined here. Principles elaborated over time in decisions of state and federal courts, and in statutes and rules, will serve to give more complete guidance as to the factors that should bear upon the exercise of the judge’s discretion. At this point, however, it suffices to note two considerations that are of relevance. First, a court may take account of a defendant’s earlier expressed willingness, or unwillingness, to accept responsibility for his or her actions. Second, it is not necessary here to decide as a constitutional rule that a judge is re- quired to prescind (that is to say disregard) any information concerning the crime that was discovered after the plea offer was made. The time continuum makes it difficult to restore the defendant and the prosecution to the precise positions they occupied prior to the rejection of the plea offer, but that baseline can be consulted in finding a remedy that does not require the prosecution to incur the expense of conducting a new trial. Petitioner argues that implementing a remedy here will open the floodgates to litigation by defendants seeking to unsettle their convictions. See Brief for Petitioner 20. Petitioner’s concern is misplaced. Courts have recognized claims of this sort for over 30 years, see supra, at 5, and yet there is no indication that the system is overwhelmed by these types of suits or that defendants are receiving windfalls as a result of strategically timed Strickland claims. See also Padilla, 559 U. S., at ___ (slip op., at 14) (“We confronted a similar ‘floodgates’ concern in Hill,” but a “flood did not follow in that decision’s wake”). In addition, the “prosecution and the trial courts may adopt some measures to help ensure against late, frivolous, or fabricated claims after a later, less advantageous plea offer has been accepted or after a trial leading to conviction.” Frye, ante, at 10. See also ibid. (listing procedures currently used by various States). This, too, will help ensure against meritless claims. III The standards for ineffective assistance of counsel when a defendant rejects a plea offer and goes to trial must now be applied to this case. Respondent brings a federal collateral challenge to a state-court conviction. Under AEDPA, a federal court may not grant a petition for a writ of habeas corpus unless the state court’s adjudication on the merits was “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.” 28 U. S. C. §2254(d)(1). A decision is contrary to clearly established law if the state court “applies a rule that con- tradicts the governing law set forth in [Supreme Court] cases.” Williams v. Taylor, 529 U.S. 362, 405 (2000) (opinion for the Court by O’Connor, J.). The Court of Appeals for the Sixth Circuit could not determine whether the Michigan Court of Appeals addressed respondent’s ineffective-assistance-of-counsel claim or, if it did, “what the court decided, or even whether the correct legal rule was identified.” 376 Fed. Appx., at 568–569. The state court’s decision may not be quite so opaque as the Court of Appeals for the Sixth Circuit thought, yet the federal court was correct to note that AEDPA does not present a bar to granting respondent relief. That is because the Michigan Court of Appeals identified respondent’s ineffective-assistance-of-counsel claim but failed to apply Strickland to assess it. Rather than applying Strickland, the state court simply found that respondent’s rejection of the plea was knowing and voluntary. Cooper, 2005 WL 599740, *1, App. to Pet. for Cert. 45a. An inquiry into whether the rejection of a plea is knowing and voluntary, however, is not the correct means by which to address a claim of ineffective assistance of counsel. See Hill, 474 U. S., at 370 (applying Strickland to assess a claim of ineffective assistance of counsel arising out of the plea negotiation process). After stating the incorrect standard, moreover, the state court then made an irrelevant observation about counsel’s performance at trial and mischaracterized respondent’s claim as a complaint that his attorney did not obtain a more favorable plea bargain. By failing to apply Strickland to assess the ineffective-assistance-of-counsel claim respondent raised, the state court’s adjudication was contrary to clearly established federal law. And in that circumstance the federal courts in this habeas action can determine the principles necessary to grant relief. See Panetti v. Quarterman, 551 U.S. 930, 948 (2007). Respondent has satisfied Strickland’s two-part test. Regarding performance, perhaps it could be accepted that it is unclear whether respondent’s counsel believed respondent could not be convicted for assault with intent to murder as a matter of law because the shots hit Mundy below the waist, or whether he simply thought this would be a persuasive argument to make to the jury to show lack of specific intent. And, as the Court of Appeals for the Sixth Circuit suggested, an erroneous strategic prediction about the outcome of a trial is not necessarily deficient performance. Here, however, the fact of deficient performance has been conceded by all parties. The case comes to us on that assumption, so there is no need to address this question. As to prejudice, respondent has shown that but for counsel’s deficient performance there is a reasonable probability he and the trial court would have accepted the guilty plea. See 376 Fed. Appx., at 571–572. In addition, as a result of not accepting the plea and being convicted at trial, respondent received a minimum sentence 3½ times greater than he would have received under the plea. The standard for ineffective assistance under Strickland has thus been satisfied. As a remedy, the District Court ordered specific performance of the original plea agreement. The correct remedy in these circumstances, however, is to order the State to reoffer the plea agreement. Presuming respondent accepts the offer, the state trial court can then exercise its discretion in determining whether to vacate the convictions and resentence respondent pursuant to the plea agreement, to vacate only some of the convictions and resentence respondent accordingly, or to leave the convictions and sentence from trial undisturbed. See Mich. Ct. Rule 6.302(C)(3) (2011) (“If there is a plea agreement and its terms provide for the defendant’s plea to be made in exchange for a specific sentence disposition or a prosecuto- rial sentence recommendation, the court may . . . reject the agreement”). Today’s decision leaves open to the trial court how best to exercise that discretion in all the circumstances of the case. The judgment of the Court of Appeals for the Sixth Circuit is vacated, and the case is remanded for further proceedings consistent with this opinion. It is so ordered.

565.US.2011_10-1265

Respondent Clair was charged with capital murder for the 1984 slaying of Linda Rodgers. The main evidence at his trial in California state court came from statements Clair made to his former girlfriend in a conversation that she secretly recorded for the police. He was convicted and sentenced to death, and his verdict was upheld on direct review. In 1994, Clair commenced federal habeas proceedings by filing a request for appointment of counsel, which the District Court granted under 18 U. S. C. §3599. That statute entitles indigent capital habeas petitioners like Clair to appointed counsel. It also contemplates that appointed counsel may be “replaced . . . upon motion of the defendant,” §3599(e), but it does not specify a standard for district courts to use in evaluating those motions. Clair’s counsel filed his initial habeas petition in 1994, and in the late 1990’s, when two associates from the firm representing Clair moved to the Office of the Federal Public Defender (FPD), the FPD was substituted as counsel of record. The District Court held an evidentiary hearing on Clair’s habeas petition in August 2004, and the parties submitted their post-hearing briefs by February 2005. The court subsequently told the parties that it did not wish to receive further material about the petition. In March, Clair moved to substitute counsel, claiming that his attorneys were seeking only to overturn his death sentence, not to prove his innocence. After the court asked the parties to address the motion, Clair’s counsel informed it that they had met with Clair and that he wanted the FPD to continue representing him. The court accordingly decided that it would take no action. Six weeks later, however, Clair filed another substitution motion, adding one more charge to his earlier claims: that his private investigator had discovered that certain physical evidence from the crime scene had never been fully tested, but that Clair’s attorneys had done nothing to analyze this evidence or follow up on its discovery. The court denied the renewed motion without further inquiry. On the same day, it also denied Clair’s habeas petition. Clair sought review of his substitution motion pro se, and the FPD appealed the habeas ruling. The Ninth Circuit asked the FPD to address whether substitution was now warranted, and after the FPD informed the court that the attorney-client relationship had broken down, the court provided Clair with a new lawyer. Clair then asked the District Court to vacate the denial of his habeas petition under Federal Rule of Civil Procedure 60(b), arguing that he should be allowed to explore the significance of the new physical evidence for his case. The District Court rejected his request, and Clair appealed. Consolidating his appeals, the Ninth Circuit vacated the District Court’s denial of both his substitution request and his habeas petition. Holding that the “interests of justice” standard used in non-capital cases, see 18 U. S. C. §3006A, should govern substitution motions like Clair’s, it ruled that the District Court abused its discretion by failing to inquire into the complaints in Clair’s second letter. Because Clair had already received new counsel on appeal, the court decided the best remedy was to treat Clair’s new counsel as though he had been appointed in June 2005 and to allow him to make whatever submissions he would have made then, including a motion to amend Clair’s habeas petition in light of new evidence. Held: 1. When evaluating motions to substitute counsel in capital cases under 18 U. S. C. §3599, courts should employ the same “interests of justice” standard that applies in non-capital cases under §3006A. Pp. 6–12. (a) Although §3599 guarantees that indigent capital defendants and petitioners seeking federal habeas relief in capital cases will receive the assistance of counsel, see, e.g., §§3599(a)(1), (a)(2), (e), and contemplates that an appointed attorney may be “replaced by similarly qualified counsel upon the attorney’s own motion or upon motion of the defendant,” §3599(e), the statute fails to specify how a court should decide such a motion. Clair argues, and the Ninth Circuit agreed, that district courts should use the “in the interests of justice” standard of §3006A, which governs the appointment and substitution of counsel in federal non-capital litigation. By contrast, the State contends that an appointed lawyer can only be replaced when the defendant has suffered an “actual or constructive denial” of counsel—that is, when the lawyer lacks the requisite statutory qualifications, has a conflict of interest, or has completely abandoned the client. The Court adopts Clair’s approach, based on the history of §3599. Before 1988, §3006A governed both capital and non-capital cases, authorizing courts to appoint counsel for federal habeas petitioners and providing that in all cases in which a court had appointed counsel, substitution motions should be decided “in the interests of justice.” §3006A(c). Thus, a court in those days would have used that standard to evaluate a request like Clair’s. In 1988, Congress enacted what is now §3599, thus displacing §3006A for persons facing execution. The new statute grants federal capital defendants and capital habeas petitioners enhanced rights of representation. Habeas petitioners facing execution now receive counsel as a matter of right, see §3599(a)(2), and in multiple ways the statute aims to improve the quality of their representation: it provides them with more experienced counsel than §3006A demands, authorizes higher rates of compensation, and provides more money for investigative and expert services. These measures “reflec[t] a determination that quality legal representation is necessary” in all capital proceedings to foster “fundamental fairness in the imposition of the death penalty.” McFarland v. Scott, 512 U.S. 849, 855, 859. Given this context, the Court cannot conclude that Congress silently prescribed a substitution standard that would make it more difficult for those facing capital punishment to substitute counsel. Adopting a more stringent test than §3006A’s would deprive capital defendants of a tool they formerly had, and non-capital defendants still have, to handle serious rep- resentational problems. By contrast, utilizing §3006A’s standard comports with the myriad ways that §3599 seeks to promote effective representation for persons facing capital punishment. Pp. 6–9. (b) The dearth of support for the State’s alternative standard reinforces this conclusion. The State concedes that Congress has not considered its standard in any context; neither has a federal court used it in any case. The Court prefers to use a familiar standard, already known to work, than to try out a new one. Moreover, the State’s proposed test would gut §3599’s substitution provision, because even absent that provision courts would have an obligation to ensure that the defendant’s statutory right to counsel was satisfied throughout the litigation. The State counters that only its approach comports with the rule that habeas petitioners generally have no Sixth Amendment right to counsel. But Congress declined to track that Amendment in providing statutory rights to counsel in both §3006A and §3599. Thus, the scope of the Amendment cannot answer the statutory question presented here. The State also contends that the “interests of justice” standard will permit substitution motions to become a mechanism to defer enforcement of a death sentence. But the “interests of justice” standard takes into account whether a substitution motion will cause undue delay. Pp. 9–12. 2. The District Court did not abuse its discretion in denying Clair’s second request for new counsel under §3599’s “interests of justice” standard. In reviewing substitution motions, the courts of appeals have pointed to several relevant considerations, including: the timeliness of the motion; the adequacy of the district court’s inquiry into the defendant’s complaint; and the asserted cause for that complaint, including the extent of the conflict or breakdown in communication between lawyer and client. Because a trial court’s decision on substitution is so fact-specific, it deserves deference and may be overturned only for an abuse of discretion. Here, the District Court received Clair’s second substitution motion on the eve of deciding his 10-year-old habeas petition. Just three months earlier, Clair had written the court to complain about his attorneys. After making proper inquiry, the court learned that Clair and his counsel had settled their dispute and turned once more to ruling on Clair’s habeas petition, only to receive a second letter six weeks later. In it Clair maintained his general assertion that his lawyers were not trying to prove his innocence, but he also alleged a new and significant charge of attorney error: that counsel had refused to investigate particular, newly located physical evidence. Such a charge normally would require the court to make further inquiry; a district court cannot usually rule on a substitution motion without exploring why a defendant wants new counsel. But here, the motion’s timing precludes a holding that the District Court abused its discretion. The court received the letter while putting the finishing touches on its denial of Clair’s habeas petition. After years of litigation, an evidentiary hearing, and post-hearing briefing, the court had instructed the parties that it would accept no further submissions. All proceedings had therefore come to a close, and a new attorney could have done nothing further in the District Court. In those circumstances, the District Court acted within its discretion in denying Clair’s substitution motion. Pp. 13–16. 403 Fed. Appx. 276, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court.

A federal statute, §3599 of Title 18, entitles indigent defendants to the appointment of counsel in capital cases, including habeas corpus proceedings. The statute contemplates that appointed counsel may be “replaced . . . upon motion of the defendant,” §3599(e), but it does not specify the standard that district courts should use in evaluating those motions. We hold that courts should employ the same “interests of justice” standard that they apply in non-capital cases under a related statute, §3006A of Title 18. We also hold that the District Court here did not abuse its discretion in denying respondent Kenneth Clair’s motion to change counsel. I This case arises from the murder of Linda Rodgers in 1984. Rodgers resided at the home of Kai Henriksen and Margaret Hessling in Santa Ana, California. Clair was a squatter in a vacant house next door. About a week prior to the murder, police officers arrested Clair for burglarizing the Henriksen-Hessling home, relying on information Henriksen had provided. On the night the police released Clair from custody, Hessling returned from an evening out to find Rodgers’ dead body in the master bedroom, naked from the waist down and beaten, stabbed, and strangled. Some jewelry and household items were missing from the house. See People v. Clair, 2 Cal. 4th 629, 644–647, 828 P.2d 705, 713–714 (1992); App. to Pet. for Cert. 23–24. The district attorney charged Clair with Rodgers’ murder and sought the death penalty. No forensic evidence linked Clair to the crime; instead, the main evidence against Clair came from his former girlfriend, Pauline Flores. Although she later recanted her testimony, see App. 36–42, Flores stated at trial that she and Clair were walking in the neighborhood on the night of the murder and split up near the Henriksen-Hessling house. When they reunited about an hour later, Flores recounted, Clair was carrying jewelry and other items and had blood on his right hand. According to Flores, Clair explained to her that he had “just finished beating up a woman.” Clair, 2 Cal. 4th, at 647, 828 P. 2d, at 714. The prosecution then introduced a tape recording of a talk between Flores and Clair several months after the murder, which Flores had made in cooperation with the police. On that tape, Clair at one point denied committing the murder, but also made several inculpatory statements. For example, when Flores told Clair that she had seen blood on him, he replied “Ain’t on me no more” and “They can’t prove nothing.” App. to Pet. for Cert. 53 (internal quotation marks omitted). And in response to her continued probing, Clair explained “[W]hat you fail to realize, how . . . they gonna prove I was there . . . ? There ain’t no . . . fingerprints, ain’t no . . . body seen me go in there and leave out there.” Id., at 53–54 (internal quotation marks omitted). The jury convicted Clair and sentenced him to death. The California Supreme Court upheld the verdict, and this Court denied review, Clair v. California, 506 U.S. 1063 (1993). Clair commenced federal habeas proceedings by filing a request for appointment of counsel, which the District Court granted under §3599. Clair and his counsel filed an initial petition for habeas relief in 1994 and, after exhausting state remedies, an amended petition the following year. The petition alleged more than 40 claims, involving such matters as jury selection and composition, sufficiency of the evidence, prosecutorial misconduct, nondisclosure of exculpatory materials relating to state witnesses, and ineffectiveness of trial counsel. In the late 1990’s, two associates from the firm representing Clair took jobs at the Office of the Federal Public Defender (FPD), and the court substituted that office as counsel of record. The court held an evidentiary hearing on Clair’s habeas petition in August 2004, and the parties submitted post-hearing briefs by February 2005. The court subsequently informed the parties that it viewed the briefing “to be complete and d[id] not wish to receive any additional material” about the petition. App. 3–4. On March 16, 2005, Clair sent a letter to the court stating that the FPD attorneys “no longer . . . ha[d] [his] best interest at hand” and that he did not want them to continue to represent him. Id., at 24; see id., at 18–25. Clair alleged that the lawyers had repeatedly dismissed his efforts to participate in his own defense. Prior to the evidentiary hearing, Clair wrote, he had become so frustrated with the attorneys that he enlisted a private detective to look into his case. But the lawyers, Clair charged, refused to cooperate with the investigator; they were seeking only to overturn his death sentence, rather than to prove his innocence. As a result, Clair felt that he and his counsel were not “on the same team.” Id., at 23. The District Court responded by asking both parties to address Clair’s motion to substitute counsel. See id., at 18. The State noted that “[w]hat the trial court does with respect to appointing counsel is within its discretion, providing the interests of justice are served.” Id., at 29. The State further advised the court that “nothing in [Clair’s] letter require[d] a change” of counsel because the FPD lawyers had provided appropriate representation and substitution would delay the case. Ibid. Clair replied to the court’s request through his FPD attorneys on April 26, 2005. Their letter stated: “After meeting with Mr. Clair, counsel understands that Mr. Clair wants the [FPD] to continue to serve as his counsel in this case at this time.” Id., at 27. On the basis of that representation, the court determined that it would “take no further action on the matter at this time.” Id., at 33. But the issue resurfaced just six weeks after the court’s decision. On June 16, 2005, Clair wrote a second letter to the court asking for substitution of counsel. That letter again asserted a “total break down of communication” between Clair and the FPD; according to Clair, he was “no longer able to trust anybody within that office.” Id., at 62–63. In explaining the source of the problem, Clair reiterated each of the points made in his prior complaint. And then he added one more. Clair recounted that his private investigator had recently learned that the police and dis-trict attorney’s office were in possession of fingerprints and other physical evidence from the crime scene that had never been fully tested. The FPD lawyers, Clair asserted, were doing nothing to analyze this evidence or otherwise follow up on its discovery. Clair attributed this failure, too, to the FPD’s decision to focus on his sentence, rather than on questions of guilt. Two weeks later, the District Court denied Clair’s renewed request for substitution without further inquiry. The court stated: “It does not appear to the Court that a change of counsel is appropriate. It appears that [Clair’s] counsel is doing a proper job. No conflict of interest or inadequacy of counsel is shown.” Id., at 61. On the same day, the court denied Clair’s habeas petition in a detailed opinion. Clair v. Brown, Case No. CV 93–1133 GLT (CD Cal., June 30, 2005), App. to Pet. for Cert. 20–91. Clair sought review of his substitution motion pro se, while the FPD filed a notice of appeal from the denial of his habeas petition. The Court of Appeals for the Ninth Circuit instructed the FPD to address whether substitution of counsel was now warranted, and in October 2005, the FPD informed the court that “the attorney-client relationship ha[d] broken down to such an extent that sub-stitution of counsel [would be] appropriate.” Attorney for Appellant’s Response to Court’s Sept. 15, 2005 Order, in No. 05–99005 (CA9), Record, Doc. 9, p. 1. The State did not comment or object, and the Court of Appeals provided Clair with a new lawyer going forward. Clair then asked the District Court to vacate the denial of his habeas petition under Federal Rule of Civil Procedure 60(b), arguing that he should be allowed to explore the significance of the new physical evidence for his case. The District Court (with a new judge assigned, because the judge previously handling the case had retired) rejected that request on the ground that the new evidence did not pertain to any of the claims presented in Clair’s habeas petition. See App. to Pet. for Cert. 9–10. Clair appealed that decision as well.[1] After consolidating Clair’s appeals, the Ninth Circuit vacated the trial court’s denial of both Clair’s request for new counsel and his habeas petition. See Clair v. Ayers, 403 Fed. Appx. 276 (2010). The Court of Appeals’ opinion focused on Clair’s substitution motion. Holding that the “interests of justice” standard should apply to that motion, the Ninth Circuit ruled that the District Court abused its discretion by failing to inquire into the complaints in Clair’s second letter. See id., at 278. The Court of Appeals then considered how to remedy that error, given that Clair had received new counsel while on appeal. It decided that “the most reasonable solution” was to “treat Clair’s current counsel as if he were the counsel who might have been appointed” in June 2005, and to allow him to make whatever submissions he would have made then, including a motion to amend Clair’s habeas petition in light of new evidence. Id., at 279. We granted certiorari to review this judgment, 564 U. S. __ (2011), and now reverse. II We first consider the standard that district courts should use to adjudicate federal habeas petitioners’ motions to substitute counsel in capital cases. The question arises because the relevant statute, 18 U. S. C. §3599, contains a notable gap. Section 3599 first guarantees that indigent defendants in federal capital cases will receive the assistance of counsel, from pretrial proceedings through stay applications. See §§3599(a)(1), (a)(2), (e). It next grants a corresponding right to people like Clair who seek federal habeas relief from a state death sentence, for all post-conviction proceedings and related activities. See §§3599(a)(2), (e); McFarland v. Scott, 512 U.S. 849, 854–855 (1994); Harbison v. Bell, 556 U.S. 180, 183–185 (2009). And the statute contemplates that both sets of litigants may sometimes substitute counsel; it notes that an attorney appointed under the section may be “replaced by similarly qualified counsel upon the attorney’s own motion or upon motion of the defendant.” §3599(e).[2] But here lies the rub: The statute fails to specify how a court should decide such a motion. Section 3599 says not a word about the standard a court should apply when addressing a request for a new lawyer. The parties offer us two alternative ways to fill this statutory hole. Clair argues, and the Ninth Circuit agreed, that district courts should decide substitution motions brought under §3599 “in the interests of justice.” That standard derives from 18 U. S. C. §3006A, which governs the appointment and substitution of counsel in federal non-capital litigation. By contrast, the State contends that district courts may replace an appointed lawyer under §3599 only when the defendant has suffered an “actual or constructive denial” of counsel. Brief for Petitioner 33. That denial occurs, the State asserts, in just three situations: when the lawyer lacks the qualifications necessary for appointment under the statute; when he has a “disabling conflict of interest”; or when he has “completely abandoned” the client. Id., at 34. On this matter, we think Clair, not the State, gets it right. A trip back in time begins to show why. Prior to 1988, §3006A governed the appointment of counsel in all federal criminal cases and habeas litigation, regardless whether the matter involved a capital or a non-capital offense. That section provided counsel as a matter of right to most indigent criminal defendants, from pre-trial proceedings through appeal. See §§3006A(a)(1), (c) (1982 ed.). In addition, the statute authorized courts to appoint counsel for federal habeas petitioners when “the interests of justice so require[d],” §3006A(g); and under that provi-sion, courts almost always appointed counsel to represent petitioners convicted of capital offenses, see Ruthenbeck, Dueling with Death in Federal Courts, 4 ABA Criminal Justice, No. 3, pp. 2, 42 (Fall, 1989). In all cases in which a court had appointed counsel, §3006A further provided (as it continues to do) that substitution motions should be decided “in the interests of justice.” §3006A(c). So in those days, a court would have used that standard to evaluate a request like Clair’s. In 1988, Congress enacted the legislation now known as §3599 to govern appointment of counsel in capital cases, thus displacing §3006A for persons facing execution (but retaining that section for all others). See Anti-Drug Abuse Act, 102Stat. 4393–4394, 21 U. S. C. §§848(q)(4)–(10) (1988 ed.) (recodified at 18 U. S. C. §3599 (2006 ed. and Supp. IV)). The new statute grants federal capital defendants and capital habeas petitioners enhanced rights of representation, in light of what it calls “the seriousness of the possible penalty and . . . the unique and complex nature of the litigation.” §3599(d) (2006 ed.). Habeas petitioners facing execution now receive counsel as a matter of right, not an exercise of the court’s discretion. See §3599(a)(2). And the statute aims in multiple ways to improve the quality of representation afforded to capital petitioners and defendants alike. Section 3599 requires lawyers in capital cases to have more legal experience than §3006A demands. Compare §§3599(b)–(d) with §3006A(b). Similarly, §3599 authorizes higher rates of compensation, in part to attract better counsel. Compare §3599(g)(1) with §3006A(d) (2006 ed. and Supp. IV). And §3599 provides more money for investigative and expert services. Compare §§3599(f) (2006 ed.), (g)(2) (2006 ed., Supp. IV), with §3006A(e) (2006 ed. and Supp. IV). As we have previously noted, those measures “reflec[t] a determination that quality legal representation is necessary” in all capital proceedings to foster “fundamental fairness in the imposition of the death penalty.” McFarland, 512 U. S., at 855, 859. That understanding of §3599’s terms and origins goes far toward resolving the parties’ dispute over what standard should apply. We know that before §3599’s passage, courts used an “interests of justice” standard to decide substitution motions in all cases—and that today, they continue to do so in all non-capital proceedings. We know, too, that in spinning off §3599, Congress enacted a set of reforms to improve the quality of lawyering in capital litigation. With all those measures pointing in one direction, we cannot conclude that Congress silently prescribed a substitution standard that would head the opposite way. Adopting a more stringent test than §3006A’s would deprive capital defendants of a tool they formerly had, and defendants facing lesser penalties still have, to handle serious representational problems. That result clashes with everything else §3599 does. By contrast, utilizing §3006A’s standard comports with the myriad ways that §3599 seeks to promote effective representation for persons threatened with capital punishment. The dearth of support for the State’s alternative standard reinforces the case for borrowing from §3006A. Recall that the State thinks substitution proper “only when . . . counsel is completely denied”—which, the State says, occurs when counsel lacks the requisite experience; “actively represents conflicting interests”; or has “total[ly] desert[ed]” the client. Brief for Petitioner 15, 35, 38. As the State acknowledges, this test comes from . . . well, from nowhere. The State conceded during argument that Congress has not considered (much less adopted) the standard in any context; neither has a federal court used it in any case. See Tr. of Oral Arg. 16. Indeed, the standard is new to the State’s own attorneys. As noted earlier, when Clair first requested a change of counsel, the State responded that substitution is a “matter . . . of trial court discretion,” based on “the interests of justice.” App. 29; see supra, at 3–4. Only later did the State devise its present proposal. Inventiveness is often an admirable quality, but here we think the State overdoes it. To be sure, we must infer a substitution standard for §3599; in that sense, we are writing on a blank slate. But in undertaking that task, we prefer to copy something familiar than concoct something novel. That enables courts to rely on experience and precedent, with a standard already known to work effectively. Still worse, the State’s proposed test guts §3599’s pro-vision for substitution motions. See §3599(e) (2006 ed.) (appointed counsel may be “replaced . . . upon motion of the defendant”). According to the State, a court may not change counsel under §3599 even if the attorney-client relationship has broken down, so long as the lawyer has the required qualifications and is “act[ing] as an advocate.” Brief for Petitioner 35. And that is so, continues the State, even when substitution will not cause delay or other prejudice—because again, the defendant retains a functioning lawyer. See id., at 34. That approach, as already noted, undermines Congress’s efforts in §3599 to enhance representation in capital cases. See supra, at 8–9. And beyond that, it renders §3599’s substitution provision superfluous. Even in the absence of that provision, a court would have to ensure that the defendant’s statutory right to counsel was satisfied throughout the litigation; for example, the court would have to appoint new counsel if the first lawyer developed a conflict with or abandoned the client. So by confining substitution to cases in which the defendant has no counsel at all, the State’s proposal effectively deletes §3599’s substitution clause. The State counters that only its approach comports with “this Court’s long-established jurisprudence that habeas prisoners, including capital prisoners,” have no right to counsel under the Sixth Amendment. Brief for Petitioner 18; see Murray v. Giarratano, 492 U.S. 1, 10, 12 (1989) (plurality opinion); id., at 14–15 (Kennedy, J., concurring in judgment); cf. Coleman v. Thompson, 501 U.S. 722, 755 (1991) (reserving question of whether the Sixth Amendment guarantees counsel when a habeas proceeding provides the first opportunity to raise a claim). But we do not understand the State’s basis for linking use of the “interests of justice” standard to cases in which an individual has a Sixth Amendment right. A statute need not draw the same lines as the Constitution, and neither §3006A nor §3599 does so in addressing the substitution of counsel. Section 3006A applies the “interests of justice” standard to substitution motions even when the Sixth Amendment does not require representation; that is presumptively so, for example, when a court provides counsel to a non-capital habeas petitioner. See §§3006A(a)(2)(B), (c). And whatever standard we adopt for §3599 will likewise apply both to litigants who have and to litigants who lack a Sixth Amendment right, because the section offers counsel on the same terms to capital defendants and habeas petitioners. In providing statutory rights to counsel, Congress declined to track the Sixth Amendment; accordingly, the scope of that Amendment cannot answer the statutory question presented here. The State’s stronger argument relates to delay in capital proceedings. Under the “interests of justice” standard, the State contends, substitution motions will become a mechanism to defer enforcement of a death sentence, contrary to historic restrictions on “abuse of the writ” and to the goals of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA). See Brief for Petitioner 19–22. But this argument, like the last, forgets that §3599 reaches not just habeas petitioners but also criminal defendants, who have not been convicted or sentenced and therefore have no incentive to delay. Moreover, the State’s claim misjudges the capacity of the “interests of justice” standard to deal with such issues. Protecting against abusive delay is an interest of justice. Because that is so, courts addressing substitution motions in both capital and non-capital cases routinely consider issues of timeliness. See, e.g., Hunter v. Delo, 62 F.3d 271, 274 (CA8 1995) (citing “the need to thwart abusive delay” in affirming the denial of a habeas petitioner’s substitution motion); United States v. White, 451 F.2d 1225, 1226 (CA6 1971) (per curiam) (approving a District Court’s refusal to change counsel under §3006A(c) “on the morning of the trial”). Indeed, we will do so, just paragraphs from here, in this very case. See infra, at 15–16. The standard we adopt thus takes account of, rather than ignores or opposes, the State’s interest in avoiding undue delay.[3] III The remaining question is whether the District Court abused its discretion in denying Clair’s second request for new counsel under §3599’s “interests of justice” standard. We do not think the court did so, although the court’s failure to make any inquiry into Clair’s allegations makes this decision harder than necessary. As its name betrays, the “interests of justice” standard contemplates a peculiarly context-specific inquiry. So we doubt that any attempt to provide a general definition of the standard would prove helpful. In reviewing substitution motions, the courts of appeals have pointed to several relevant considerations. Those factors may vary a bit from circuit to circuit, but generally include: the timeliness of the motion; the adequacy of the district court’s inquiry into the defendant’s complaint; and the asserted cause for that complaint, including the extent of the conflict or breakdown in communication between lawyer and client (and the client’s own responsibility, if any, for that conflict). See, e.g., United States v. Prime, 431 F.3d 1147, 1154 (CA9 2005); United States v. Doe, 272 F.3d 116, 122–123 (CA2 2001); Hunter, 62 F. 3d, at 274; United States v. Welty, 674 F.2d 185, 188 (CA3 1982). Because a trial court’s decision on substitution is so fact-specific, it deserves deference; a reviewing court may overturn it only for an abuse of discretion. The District Court here received Clair’s second substitution motion on the eve of deciding his 10-year-old habeas petition. Recall that three months earlier, following an evidentiary hearing and post-hearing briefing, Clair had written the court to complain about his attorneys. In that first letter, Clair accused his lawyers of refusing to co-operate with a private detective and, more generally, of forgoing efforts to prove his innocence. After making proper inquiry, the court learned that Clair and his attorneys had worked through their dispute and Clair no longer wanted to substitute counsel. The court thus turned its attention once again to ruling on Clair’s habeas petition—only to receive another letter requesting a change in representation. If that second letter had merely recapitulated the charges in the first, this case would be relatively simple. Even then, the court might have done well to make further inquiry of Clair and his counsel. As all Circuits agree, courts cannot properly resolve substitution motions without probing why a defendant wants a new lawyer. See, e.g., United States v. Iles, 906 F.2d 1122, 1130 (CA6 1990) (“It is hornbook law that ‘[w]hen an indigent defendant makes a timely and good faith motion requesting that appointed counsel be discharged and new counsel appointed, the trial court clearly has a responsibility to determine the reasons for defendant’s dissatisfaction . . .’ ” (quoting 2 W. LaFave & J. Israel, Criminal Procedure §11.4, p. 36 (1984))). Moreover, an on-the-record inquiry into the defendant’s allegations “permit[s] meaningful appellate review” of a trial court’s exercise of discretion. United States v. Taylor, 487 U.S. 326, 336–337 (1988). But here the court had inquired, just a short time earlier, into Clair’s relationship with his lawyers. The court knew that Clair had responded to that inquiry by dropping his initial complaints. And the court had reason to think, based on 10 years of handling the case, that those charges lacked merit: Perhaps most important, the court knew that the lawyers had raised many challenges not just to Clair’s sentence, but to his conviction, including to the sufficiency of the State’s evidence. See, e.g., App. to Pet. for Cert. 27–69. Especially at this stage of the litigation, those factors would have provided ample basis to reject a simple reprise of Clair’s allegations. What complicates this case is that in his second letter, Clair added a new and significant charge of attorney error. Beyond asserting generally that his lawyers were not trying to prove his innocence, Clair now alleged that counsel had refused to investigate particular, newly located physical evidence. That evidence, according to Clair, might have shown that the police had suppressed Brady material, that his trial counsel had been ineffective in investigating the murder, or that he had not committed the offense. See Tr. of Oral Arg. 45–46. Especially in a case lacking physical evidence, built in part on since-recanted witness testimony, those possibilities cannot be blithely dismissed. In the mine run of circumstances, Clair’s new charge would have required the court to make further inquiry before ruling on his motion for a new attorney. But here, the timing of that motion precludes a holding that the District Court abused its discretion. The court received Clair’s second letter while putting the finishing touches on its denial of his habeas petition. (That lengthy decision issued just two weeks later.) After many years of litigation, an evidentiary hearing, and substantial post-hearing briefing, the court had instructed the parties that it would accept no further submissions. See App. 3–4; Tr. of Oral Arg. 4–5. The case was all over but the deciding; counsel, whether old or new, could do nothing more in the trial court proceedings. At that point and in that forum, Clair’s conflict with his lawyers no longer mattered. Clair, to be sure, wanted to press his case further in the District Court. He desired a new lawyer, after examining the physical evidence, to make whatever claims followed from it. But, notably, all of those claims would have been new; as the District Court later found in ruling on Clair’s Rule 60(b) motion, the physical evidence did not relate to any of the claims Clair had previously made in his habeas petition. See App. to Pet. for Cert. 9–10. A substitute lawyer thus would have had to seek an amendment of that petition, as well as an evidentiary hearing or, more likely, a stay to allow exhaustion of remedies in state court. See 403 Fed. Appx., at 279. The District Court could properly have rejected that motion, consistent with its order precluding further submissions (effectively remitting Clair to state court to pursue the matter). See Mayle v. Felix, 545 U.S. 644, 663 (2005). And if that is so, the court also acted within its discretion in denying Clair’s request to substitute counsel, even without the usually appropriate inquiry. The court was not required to appoint a new lawyer just so Clair could file a futile motion. We accordingly find that the Court of Appeals erred in overturning the District Court’s decision.[4] The judgment below is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 While litigating his Rule 60(b) motion in the District Court, Clair also pursued discovery in the California state courts relating to the newly found physical evidence. On the basis of material he obtained, Clair filed another petition for state habeas relief, alleging (among other claims) actual innocence and improper suppression of exculpatory material under Brady v. Maryland, 373 U.S. 83 (1963). The California Supreme Court summarily denied that petition. See In re Clair, No. S169188 (Aug. 24, 2011). 2 Section 3599(e) provides in full: “Unless replaced by similarly qualified counsel upon the attorney’s own motion or upon motion of the defendant, each attorney so appointed shall represent the defendant throughout every subsequent stage of available judicial proceedings, including pretrial proceedings, trial, sentencing, motions for new trial, appeals, applications for writ of certiorari to the Supreme Court of the United States, and all available post-conviction process, together with applications for stays of exe-cution and other appropriate motions and procedures, and shall also represent the defendant in such competency proceedings and pro-ceedings for executive or other clemency as may be available to the defendant.” 3 The State also makes a more specific argument based on AEDPA, see Brief for Petitioner 26–29, but we think it is not well taken. The State notes that the “interests of justice” standard enables a court, when ruling on a substitution motion, to take account of a lawyer’s effectiveness. That consideration, according to the State, conflictswith AEDPA’s injunction that “[t]he ineffectiveness or incompetence of counsel during Federal or State collateral post-conviction proceedings shall not be a ground for relief in a [habeas] proceeding arising under section 2254.” 28 U. S. C. §2254(i); see §2261(e) (using similar language). But most naturally read, §2254(i) prohibits a court from granting substantive habeas relief on the basis of a lawyer’s ineffectiveness in post-conviction proceedings, not from substituting counsel on that ground. Cf. Holland v. Florida, 560 U. S. ___, ___ (2010) (slip op., at 18) (holding that §2254(i) does not preclude equitable tolling of a statute of limitations based on attorney misconduct in habeas proceedings). Indeed, if the State were right, we would also have to find that AEDPA silently repealed §3006A’s instruction to courts to apply the “interests of justice” standard in non-capital habeas cases. We see nothing to suggest that Congress had that result in mind. 4 We note as well that the Court of Appeals ordered the wrong remedy even assuming the District Court had abused its discretion in denying Clair’s substitution motion without inquiry. The way to cure that error would have been to remand to the District Court to decide whether substitution was appropriate at the time of Clair’s letter. Unless that court determined that counsel should have been changed, the Court of Appeals had no basis for vacating the denial of Clair’s habeas petition.

566.US.2011_10-1001

Arizona prisoners may raise claims of ineffective assistance of trial counsel only in state collateral proceedings, not on direct review. In petitioner Martinez’s first state collateral proceeding, his counsel did not raise such a claim. On federal habeas review with new counsel, Martinez argued that he received ineffective assistance both at trial and in his first state collateral proceeding. He also claimed that he had a constitutional right to an effective attorney in the collateral proceeding because it was the first place to raise his claim of ineffective assistance at trial. The District Court denied the petition, finding that Arizona’s preclusion rule was an adequate and independent state-law ground barring federal review, and that under Coleman v. Thompson, 501 U.S. 722, the attorney’s errors in the postconviction proceeding did not qualify as cause to excuse the procedural default. The Court of Appeals for the Ninth Circuit affirmed. Held: 1. Where, under state law, ineffective-assistance-of-trial-counsel claims must be raised in an initial-review collateral proceeding, a procedural default will not bar a federal habeas court from hearing those claims if, in the initial-review collateral proceeding, there was no counsel or counsel in that proceeding was ineffective. Pp. 5–14. (a) Given that the precise question here is whether ineffective assistance in an initial-review collateral proceeding on an ineffective-assistance-at-trial claim may provide cause for a procedural default in a federal habeas proceeding, this is not the case to resolve the question left open in Coleman: whether a prisoner has a constitutional right to effective counsel in initial-review collateral proceedings. However, to protect prisoners with potentially legitimate ineffective-assistance claims, it is necessary to recognize a narrow exception to Coleman’s unqualified statement that an attorney’s ignorance or inadvertence in a postconviction proceeding does not qualify as cause to excuse a procedural default, namely, that inadequate assistance of counsel at initial-review collateral proceedings may establish cause. Pp. 5–6. (b) A federal court can hear Martinez’s ineffective-assistance claim only if he can establish cause to excuse the procedural default and prejudice from a violation of federal law. Coleman held that a postconviction attorney’s negligence “does not qualify as ‘cause,’ ” because “the attorney is the prisoner’s agent,” and “the principal bears the risk of” his agent’s negligent conduct, Maples v. Thomas, ante, at 12. However, in Coleman, counsel’s alleged error was on appeal from an initial-review collateral proceeding. Thus, his claims had been addressed by the state habeas trial court. This marks a key difference between initial-review collateral proceedings and other collateral proceedings. Here, where the initial-review collateral proceeding is the first designated proceeding for a prisoner to raise the ineffective-assistance claim, the collateral proceeding is the equivalent of a prisoner’s direct appeal as to that claim because the state habeas court decides the claim’s merits, no other court has addressed the claim, and defendants “are generally ill equipped to represent themselves” where they have no brief from counsel and no court opinion addressing their claim. Halbert v. Michigan, 545 U.S. 605, 617. An attorney’s errors during an appeal on direct review may provide cause to excuse a procedural default; for if the attorney appointed by the State is ineffective, the prisoner has been denied fair process and the opportunity to comply with the State’s procedures and obtain an adjudication on the merits of his claim. Without adequate representation in an initial-review collateral proceeding, a prisoner will have similar difficulties vindicating a substantial ineffective-assistance-at-trial claim. The same would be true if the State did not appoint an attorney for the initial-review collateral proceeding. A prisoner’s inability to present an ineffective-assistance claim is of particular concern because the right to effective trial counsel is a bedrock principle in this Nation’s justice system. Allowing a federal habeas court to hear a claim of ineffective assistance at trial when an attorney’s errors (or an attorney’s absence) caused a procedural default in an initial-review collateral proceeding acknowledges, as an equitable matter, that a collateral proceeding, if undertaken with no counsel or ineffective counsel, may not have been sufficient to ensure that proper consideration was given to a substantial claim. It thus follows that, when a State requires a prisoner to raise a i claim of ineffective assistance at trial in a collateral proceeding, a prisoner may establish cause for a procedural default of such claim in two circumstances: where the state courts did not appoint counsel in the initial-review collateral proceeding for an ineffective-assistance-at-trial claim; and where appointed counsel in the initial-review collateral proceeding, where that claim should have been raised, was ineffective under Strickland v. Washington, 466 U.S. 668. To overcome the default, a prisoner must also demonstrate that the underlying ineffective-assistance-at-trial claim is substantial. Most jurisdictions have procedures to ensure counsel is appointed for substantial ineffective-assistance claims. It is likely that such attorneys are qualified to perform, and do perform, according to prevailing professional norms. And where that is so, States may enforce a procedural default in federal habeas proceedings. Pp. 6–12. (c) This limited qualification to Coleman does not implicate stare decisis concerns. Coleman’s holding remains true except as to initial-review collateral proceedings for claims of ineffective assistance at trial. The holding in this case should not put a significant strain on state resources. A State facing the question of cause for an apparent default may answer that the ineffective-assistance-of-trial-counsel claim is insubstantial. The limited circumstances recognized here also reflect the importance of the right to effective assistance at trial. Other claims may not implicate the same fundamentals of the adversary system. The Antiterrorism and Effective Death Penalty Act of 1996 does not speak to the question presented here, and thus does not bar Martinez from asserting attorney error as cause for a procedural default. Pp. 12–14. 2. Whether Martinez’s attorney in his first collateral proceeding was ineffective and whether his ineffective-assistance-at-trial claim is substantial, as well as the question of prejudice, are questions that remain open for a decision on remand. P. 15. 623 F.3d 731, reversed and remanded. Kennedy, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Alito, Sotomayor, and Kagan, JJ., joined. Scalia, J., filed a dissenting opinion, in which Thomas, J., joined.

The State of Arizona does not permit a convicted person alleging ineffective assistance of trial counsel to raise that claim on direct review. Instead, the prisoner must bring the claim in state collateral proceedings. In the instant case, however, petitioner’s postconviction counsel did not raise the ineffective-assistance claim in the first collateral proceeding, and, indeed, filed a statement that, after reviewing the case, she found no meritorious claims helpful to petitioner. On federal habeas review, and with new counsel, petitioner sought to argue he had received ineffective assistance of counsel at trial and in the first phase of his state collateral proceeding. Because the state collateral proceeding was the first place to challenge his con- viction on grounds of ineffective assistance, petitioner maintained he had a constitutional right to an effective attorney in the collateral proceeding. While petitioner frames the question in this case as a constitutional one, a more narrow, but still dispositive, formulation is whether a federal habeas court may excuse a procedural default of an ineffective-assistance claim when the claim was not properly presented in state court due to an attorney’s errors in an initial-review collateral proceeding. I A jury convicted petitioner, Luis Mariano Martinez, of two counts of sexual conduct with a minor under the age of 15. The prosecution introduced a videotaped forensic interview with the victim, Martinez’s 11-year-old stepdaughter. It also put in evidence the victim’s nightgown, with traces of Martinez’s DNA. As part of his defense, Martinez introduced evidence of the victim’s recantations, including testimony from the victim’s grandmother and mother and a second videotaped interview in which the victim denied any abuse. The victim also denied any abuse when she testified at trial. App. to Pet. for Cert. 38a–39a. To explain the inconsistencies, a prosecution expert testified that recantations of child-abuse accusations are caused often by reluctance on the part of the victim’s mother to lend support to the child’s claims. Pet. for Cert. 3. After considering the conflicting evidence, the jury convicted Martinez. He was sentenced to two consecutive terms of life imprisonment with no possibility of parole for 35 years. App. to Pet. for Cert. 39a. The State appointed a new attorney to represent Martinez in his direct appeal. Ibid.; Pet. for Cert. 4. She made numerous arguments on Martinez’s behalf, including a claim that the evidence was insufficient and that newly discovered evidence warranted a new trial. App. to Pet. for Cert. 39a. Arizona law, however, did not permit her to argue on direct appeal that trial counsel was ineffective. State v. Spreitz, 202 Ariz. 1, 3, 39 P.3d 525, 527 (2002). Arizona instead requires claims of ineffective assistance at trial to be reserved for state collateral proceedings. While Martinez’s direct appeal was pending, the attorney began a state collateral proceeding by filing a “Notice of Post-Conviction Relief.” Martinez v. Schriro, 623 F.3d 731, 733–734 (CA9 2010); Ariz. Rule Crim. Proc. 32.4(a) (2011). Despite initiating this proceeding, counsel made no claim trial counsel was ineffective and later filed a statement asserting she could find no colorable claims at all. 623 F. 3d, at 734. Cf. State v. Smith, 184 Ariz. 456, 459, 910 P.2d 1, 4 (1996). The state trial court hearing the collateral proceeding gave Martinez 45 days to file a pro se petition in support of postconviction relief and to raise any claims he believed his counsel overlooked. 623 F. 3d, at 734; see Smith, supra, at 459, 910 P. 2d, at 4. Martinez did not respond. He later alleged that he was unaware of the ongoing collateral proceedings and that counsel failed to advise him of the need to file a pro se petition to preserve his rights. The state trial court dismissed the action for postconviction relief, in effect affirming counsel’s determination that Martinez had no meritorious claims. 623 F. 3d, at 734. The Arizona Court of Appeals affirmed Martinez’s conviction, and the Arizona Supreme Court denied review. Id., at 733. About a year and a half later, Martinez, now represented by new counsel, filed a second notice of postconviction relief in the Arizona trial court. Id., at 734. Martinez claimed his trial counsel had been ineffective for failing to challenge the prosecution’s evidence. He argued, for example, that his trial counsel should have objected to the expert testimony explaining the victim’s recantations or should have called an expert witness in rebuttal. Martinez also faulted trial counsel for not pursuing an exculpatory explanation for the DNA on the nightgown. App. to Brief in Opposition B–6 to B–12. Martinez’s petition was dismissed, in part in reliance on an Arizona Rule barring relief on a claim that could have been raised in a previous collateral proceeding. Id., at B–27; see Ariz. Rule Crim. Proc. 32.2(a)(3). Martinez, the theory went, should have asserted the claims of ineffective assistance of trial counsel in his first notice for postconviction relief. The Arizona Court of Appeals agreed. It denied Martinez relief because he failed to raise his claims in the first collateral proceeding. 623 F. 3d, at 734. The Arizona Supreme Court declined to review Martinez’s appeal. Martinez then sought relief in United States District Court for the District of Arizona, where he filed a petition for a writ of habeas corpus, again raising the ineffective-assistance-of-trial-counsel claims. Martinez acknowledged the state courts denied his claims by relying on a well-established state procedural rule, which, under the doctrine of procedural default, would prohibit a federal court from reaching the merits of the claims. See, e.g., Wainwright v. Sykes, 433 U.S. 72, 84–85, 90–91 (1977). He could overcome this hurdle to federal review, Martinez argued, because he had cause for the default: His first postconviction counsel was ineffective in failing to raise any claims in the first notice of postconviction relief and in failing to notify Martinez of her actions. See id., at 84–85. On the Magistrate Judge’s recommendation, the District Court denied the petition, ruling that Arizona’s preclusion rule was an adequate and independent state-law ground to bar federal review. App. to Pet. for Cert. 36a. Martinez had not shown cause to excuse the procedural default, the District Court reasoned, because under Coleman v. Thompson, 501 U.S. 722, 753–754 (1991), an attorney’s errors in a postconviction proceeding do not qualify as cause for a default. See id., at 754–755. The Court of Appeals for the Ninth Circuit affirmed. The Court of Appeals relied on general statements in Coleman that, absent a right to counsel in a collateral proceeding, an attorney’s errors in the proceeding do not establish cause for a procedural default. Expanding on the District Court’s opinion, the Court of Appeals, citing Coleman, noted the general rule that there is no constitutional right to counsel in collateral proceedings. 623 F. 3d, at 736. The Court of Appeals recognized that Coleman reserved ruling on whether there is “an exception” to this rule in those cases “where ‘state collateral review is the first place a prisoner can present a challenge to his conviction.’ ” 623 F. 3d, at 736 (quoting Coleman, supra, at 755). It concluded, nevertheless, that the controlling cases established no basis for the exception. Certiorari was granted. 563 U. S. ___ (2011). II Coleman v. Thompson, supra, left open, and the Court of Appeals in this case addressed, a question of constitutional law: whether a prisoner has a right to effective counsel in collateral proceedings which provide the first occasion to raise a claim of ineffective assistance at trial. These proceedings can be called, for purposes of this opinion, “initial-review collateral proceedings.” Coleman had sug- gested, though without holding, that the Constitution may require States to provide counsel in initial-review collateral proceedings because “in [these] cases . . . state collateral review is the first place a prisoner can present a challenge to his conviction.” Id., at 755. As Coleman noted, this makes the initial-review collateral proceeding a prisoner’s “one and only appeal” as to an ineffective-assistance claim, id., at 756 (emphasis deleted; internal quotation marks omitted), and this may justify an exception to the constitutional rule that there is no right to counsel in collateral proceedings. See id., at 755; Douglas v. California, 372 U.S. 353, 357 (1963) (holding States must appoint counsel on a prisoner’s first appeal). This is not the case, however, to resolve whether that exception exists as a constitutional matter. The pre- cise question here is whether ineffective assistance in an initial-review collateral proceeding on a claim of ineffective assistance at trial may provide cause for a procedural default in a federal habeas proceeding. To protect prisoners with a potentially legitimate claim of ineffective assistance of trial counsel, it is necessary to modify the unqualified statement in Coleman that an attorney’s ignorance or inadvertence in a postconviction proceeding does not qualify as cause to excuse a procedural default. This opinion qualifies Coleman by recognizing a narrow exception: Inadequate assistance of counsel at initial-review collateral proceedings may establish cause for a prisoner’s procedural default of a claim of ineffective assistance at trial. A Federal habeas courts reviewing the constitutionality of a state prisoner’s conviction and sentence are guided by rules designed to ensure that state-court judgments are accorded the finality and respect necessary to preserve the integrity of legal proceedings within our system of federalism. These rules include the doctrine of procedural default, under which a federal court will not review the merits of claims, including constitutional claims, that a state court declined to hear because the prisoner failed to abide by a state procedural rule. See, e.g., Coleman, supra, at 747–748; Sykes, supra, at 84–85. A state court’s invocation of a procedural rule to deny a prisoner’s claims precludes federal review of the claims if, among other requisites, the state procedural rule is a nonfederal ground adequate to support the judgment and the rule is firmly established and consistently followed. See, e.g., Walker v. Martin, 562 U. S. ___, ___ (2011) (slip op., at 7–8); Beard v. Kindler, 558 U. S. ___, ___ (2009) (slip op., at 7). The doctrine barring procedurally defaulted claims from being heard is not without exceptions. A prisoner may obtain federal review of a defaulted claim by showing cause for the default and prejudice from a violation of federal law. See Coleman, 501 U. S., at 750. There is no dispute that Arizona’s procedural bar on successive petitions is an independent and adequate state ground. Thus, a federal court can hear Martinez’s ineffective-assistance claim only if he can establish cause to excuse the procedural default. Coleman held that “[n]egligence on the part of a prisoner’s postconviction attorney does not qualify as ‘cause.’ ” Maples v. Thomas, 565 U. S ___, ___ (2011) (slip op., at 12). Coleman reasoned that “because the attorney is the prisoner’s agent . . . under ‘well-settled principles of agency law,’ the principal bears the risk of negligent conduct on the part of his agent.” Maples, supra, at ___ (slip op., at 12). Coleman, however, did not present the occasion to apply this principle to determine whether attorney errors in initial-review collateral proceedings may qualify as cause for a procedural default. The alleged failure of counsel in Coleman was on appeal from an initial-review collateral proceeding, and in that proceeding the prisoner’s claims had been addressed by the state habeas trial court. See 501 U. S., at 755. As Coleman recognized, this marks a key difference between initial-review collateral proceedings and other kinds of collateral proceedings. When an attorney errs in initial-review collateral proceedings, it is likely that no state court at any level will hear the prisoner’s claim. This Court on direct review of the state proceeding could not consider or adjudicate the claim. See, e.g., Fox Film Corp. v. Muller, 296 U.S. 207 (1935); Murdock v. Memphis, 20 Wall. 590 (1875); cf. Coleman, supra, at 730–731. And if counsel’s errors in an initial-review collateral proceeding do not establish cause to excuse the procedural default in a federal habeas proceeding, no court will review the prisoner’s claims. The same is not true when counsel errs in other kinds of postconviction proceedings. While counsel’s errors in these proceedings preclude any further review of the prisoner’s claim, the claim will have been addressed by one court, whether it be the trial court, the appellate court on direct review, or the trial court in an initial-review collateral proceeding. See, e.g., Coleman, supra, at 756. Where, as here, the initial-review collateral proceeding is the first designated proceeding for a prisoner to raise a claim of ineffective assistance at trial, the collateral proceeding is in many ways the equivalent of a prisoner’s direct appeal as to the ineffective-assistance claim. This is because the state habeas court “looks to the merits of the clai[m]” of ineffective assistance, no other court has addressed the claim, and “defendants pursuing first-tier review . . . are generally ill equipped to represent themselves” because they do not have a brief from counsel or an opinion of the court addressing their claim of error. Halbert v. Michigan, 545 U.S. 605, 617 (2005); see Douglas, 372 U. S., at 357–358. As Coleman recognized, an attorney’s errors during an appeal on direct review may provide cause to excuse a procedural default; for if the attorney appointed by the State to pursue the direct appeal is ineffective, the prisoner has been denied fair process and the opportunity to comply with the State’s procedures and obtain an adjudication on the merits of his claims. See 501 U. S., at 754; Evitts v. Lucey, 469 U.S. 387, 396 (1985); Douglas, supra, at 357–358. Without the help of an adequate attorney, a prisoner will have similar difficulties vindicating a substantial ineffective-assistance-of-trial-counsel claim. Claims of ineffective assistance at trial often require investigative work and an understanding of trial strategy. When the issue cannot be raised on direct review, more- over, a prisoner asserting an ineffective-assistance-of-trial-counsel claim in an initial-review collateral proceeding cannot rely on a court opinion or the prior work of an attorney addressing that claim. Halbert, 545 U. S., at 619. To present a claim of ineffective assistance at trial in accordance with the State’s procedures, then, a prisoner likely needs an effective attorney. The same would be true if the State did not appoint an attorney to assist the prisoner in the initial-review collateral proceeding. The prisoner, unlearned in the law, may not comply with the State’s procedural rules or may misapprehend the substantive details of federal constitutional law. Cf., e.g., id., at 620–621 (describing the educational background of the prison population). While confined to prison, the prisoner is in no position to develop the evidentiary basis for a claim of ineffective assistance, which often turns on evidence outside the trial record. A prisoner’s inability to present a claim of trial error is of particular concern when the claim is one of ineffective assistance of counsel. The right to the effective assistance of counsel at trial is a bedrock principle in our justice system. It is deemed as an “obvious truth” the idea that “any person haled into court, who is too poor to hire a lawyer, cannot be assured a fair trial unless counsel is provided for him.” Gideon v. Wainwright, 372 U.S. 335, 344 (1963). Indeed, the right to counsel is the foundation for our adversary system. Defense counsel tests the pros- ecution’s case to ensure that the proceedings serve the function of adjudicating guilt or innocence, while protecting the rights of the person charged. See, e.g., Powell v. Alabama, 287 U.S. 45, 68–69 (1932) (“[The defendant] requires the guiding hand of counsel at every step in the proceedings against him. Without it, though he be not guilty, he faces the danger of conviction because he does not know how to establish his innocence”). Effective trial counsel preserves claims to be considered on appeal, see, e.g., Fed. Rule Crim. Proc. 52(b), and in federal habeas proceedings, Edwards v. Carpenter, 529 U.S. 446 (2000). This is not to imply the State acted with any impropri- ety by reserving the claim of ineffective assistance for a collateral proceeding. See Massaro v. United States, 538 U.S. 500, 505 (2003). Ineffective-assistance claims often depend on evidence outside the trial record. Direct appeals, without evidentiary hearings, may not be as effective as other proceedings for developing the factual basis for the claim. Ibid. Abbreviated deadlines to expand the record on direct appeal may not allow adequate time for an attorney to investigate the ineffective-assistance claim. See Primus, Structural Reform in Criminal Defense, 92 Cornell L. Rev. 679, 689, and n. 57 (2004) (most rules give between 5 and 30 days from the time of conviction to file a request to expand the record on appeal). Thus, there are sound reasons for deferring consideration of ineffective-assistance-of-trial-counsel claims until the collateral-review stage, but this decision is not without consequences for the State’s ability to assert a procedural default in later proceedings. By deliberately choosing to move trial-ineffectiveness claims outside of the direct-appeal process, where counsel is constitutionally guaranteed, the State significantly diminishes prisoners’ ability to file such claims. It is within the context of this state procedural framework that counsel’s ineffectiveness in an initial-review collateral proceeding qualifies as cause for a procedural default. The rules for when a prisoner may establish cause to excuse a procedural default are elaborated in the exercise of the Court’s discretion. McCleskey v. Zant, 499 U.S. 467, 490 (1991); see also Coleman, supra, at 730–731; Sykes, 433 U. S., at 83; Reed v. Ross, 468 U.S. 1, 9 (1984); Fay v. Noia, 372 U.S. 391, 430 (1963), overruled in part by Sykes, supra. These rules reflect an equitable judgment that only where a prisoner is impeded or obstructed in complying with the State’s established procedures will a federal habeas court excuse the prisoner from the usual sanction of default. See, e.g., Strickler v. Greene, 527 U.S. 263, 289 (1999); Reed, supra, at 16. Allowing a federal habeas court to hear a claim of ineffective assistance of trial counsel when an attorney’s errors (or the absence of an attorney) caused a procedural default in an initial-review collateral proceeding acknowledges, as an equitable matter, that the initial-review collateral proceeding, if undertaken without counsel or with ineffective counsel, may not have been sufficient to ensure that proper consideration was given to a substantial claim. From this it follows that, when a State requires a prisoner to raise an ineffective-assistance-of-trial-counsel claim in a collateral proceeding, a prisoner may establish cause for a default of an ineffective-assistance claim in two circumstances. The first is where the state courts did not appoint counsel in the initial-review collateral proceeding for a claim of ineffective assistance at trial. The second is where appointed counsel in the initial-review collateral proceeding, where the claim should have been raised, was ineffective under the standards of Strickland v. Washington, 466 U.S. 668 (1984). To overcome the default, a prisoner must also demonstrate that the underlying ineffective-assistance-of-trial-counsel claim is a substantial one, which is to say that the prisoner must demonstrate that the claim has some merit. Cf. Miller-El v. Cockrell, 537 U.S. 322 (2003) (describing standards for certificates of appealability to issue). Most jurisdictions have in place procedures to ensure counsel is appointed for substantial ineffective-assistance claims. Some States, including Arizona, appoint counsel in every first collateral proceeding. See, e.g., Alaska Stat. 18.85.100(c) (2010); Ariz. Rule Crim. Proc. 32.4(c)(2) (2011); Conn. Gen. Stat. §51–296(a) (2011); Me. Rules Crim. Proc. 69, 70(c) (2010); N. C. Gen. Stat. Ann. §7A–451(a)(2) (2009); N. J. Ct. Rule 3:22–6(b) (2012); R. I. Gen. Laws §10–9.1–5 (Lexis 1997); Tenn. Code Ann. §8–14–205 (2011). Some States appoint counsel if the claims require an evidentiary hearing, as claims of ineffective assistance often do. See, e.g., Ky. Rule Crim. Proc. 11.42(5) (2011); La. Code Crim. Proc. Ann., Art. 930.7(C) (West 2008); Mich. Rule Crim. Proc. 6.505(A) (2011); S. C. Rule Civ. Proc. 71.1(d) (2011). Other States appoint counsel if the claims have some merit to them or the state habeas trial court deems the record worthy of further development. See, e.g., Ark. Rule Crim. Proc. 37.3(b) (2011); Colo. Rule Crim. Proc. 35(b) (2011); Del. Super. Ct. Rule Crim. Proc. 61(e)(1) (2011); Indiana Rule Post-Conviction Remedies Proc. 1, §9(a) (rev. 2011); Kan. Stat. Ann. §22–4506 (2007); N. M. Dist. Ct. Rule Crim. Proc. 5–802 (2011); Hust v. State, 147 Idaho 682, 683–684, 214 P.3d 668, 669–670 (2009); Hardin v. Arkansas, 350 Ark. 299, 301, 86 S.W.3d 384, 385 (2007) (per curiam); Jensen v. State, 2004 ND 200, ¶13, 688 N.W.2d 374, 378; Wu v. United States, 798 A.2d 1083, 1089 (D. C. 2002); Kostal v. People, 167 Colo. 317, 447 P.2d 536 (1968). It is likely that most of the attorneys appointed by the courts are qualified to per- form, and do perform, according to prevailing professional norms; and, where that is so, the States may enforce a procedural default in federal habeas proceedings. B This limited qualification to Coleman does not implicate the usual concerns with upsetting reliance interests protected by stare decisis principles. Cf., e.g., Montejo v. Louisiana, 556 U.S. 778, 792–793 (2009). Coleman held that an attorney’s negligence in a postconviction proceeding does not establish cause, and this remains true except as to initial-review collateral proceedings for claims of ineffective assistance of counsel at trial. Coleman itself did not involve an occasion when an attorney erred in an initial-review collateral proceeding with respect to a claim of ineffective trial counsel; and in the 20 years since Coleman was decided, we have not held Coleman applies in circumstances like this one. The holding here ought not to put a significant strain on state resources. When faced with the question whether there is cause for an apparent default, a State may answer that the ineffective-assistance-of-trial-counsel claim is in- substantial, i.e., it does not have any merit or that it is wholly without factual support, or that the attorney in the initial-review collateral proceeding did not perform below constitutional standards. This is but one of the differences between a constitutional ruling and the equitable ruling of this case. A constitutional ruling would provide defendants a freestanding constitutional claim to raise; it would require the appointment of counsel in initial-review collateral proceedings; it would impose the same system of appointing counsel in every State; and it would require a reversal in all state collateral cases on direct review from state courts if the States’ system of appointing counsel did not conform to the constitutional rule. An equitable ruling, by contrast, permits States a variety of systems for appointing counsel in initial-review collateral proceedings. And it permits a State to elect between appointing counsel in initial-review collateral proceedings or not asserting a procedural default and raising a defense on the merits in federal habeas proceedings. In addition, state collateral cases on direct review from state courts are unaffected by the ruling in this case. The rule of Coleman governs in all but the limited circumstances recognized here. The holding in this case does not concern attorney errors in other kinds of proceedings, including appeals from initial-review collateral proceedings, second or successive collateral proceedings, and petitions for discretionary review in a State’s appellate courts. See 501 U. S., at 754; Carrier, 477 U. S., at 488. It does not extend to attorney errors in any proceeding beyond the first occasion the State allows a prisoner to raise a claim of ineffective assistance at trial, even though that initial-review collateral proceeding may be deficient for other reasons. In addition, the limited nature of the qualification to Coleman adopted here reflects the importance of the right to the effective assistance of trial counsel and Arizona’s decision to bar defendants from raising ineffective-assistance claims on direct appeal. Our holding here addresses only the constitutional claims presented in this case, where the State barred the defendant from raising the claims on direct appeal. Arizona contends that the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), 28 U. S. C. §2254, bars Martinez from asserting attorney error as cause for a procedural default. AEDPA refers to attorney error in collateral proceedings, but it does not speak to the question presented in this case. Section 2254(i) provides that “the ineffectiveness or incompetence of counsel during Federal or State collateral post-conviction proceedings shall not be a ground for relief.” “Cause,” however, is not synonymous with “a ground for relief.” A finding of cause and prejudice does not entitle the prisoner to habeas relief. It merely allows a federal court to consider the merits of a claim that otherwise would have been procedurally de- faulted. In this case, for example, Martinez’s “ground for relief” is his ineffective-assistance-of-trial-counsel claim, a claim that AEDPA does not bar. Martinez relies on the ineffectiveness of his postconviction attorney to excuse his failure to comply with Arizona’s procedural rules, not as an independent basis for overturning his conviction. In short, while §2254(i) precludes Martinez from relying on the ineffectiveness of his postconviction attorney as a “ground for relief,” it does not stop Martinez from using it to establish “cause.” Holland v. Florida, 560 U. S. ___, ___ (2010) (slip op., at 18). III Where, under state law, claims of ineffective assistance of trial counsel must be raised in an initial-review collateral proceeding, a procedural default will not bar a federal habeas court from hearing a substantial claim of ineffective assistance at trial if, in the initial-review collateral proceeding, there was no counsel or counsel in that proceeding was ineffective. In this case Martinez’s attorney in the initial-review collateral proceeding filed a notice akin to an Anders brief, in effect conceding that Martinez lacked any meritorious claim, including his claim of ineffective assistance at trial. See Anders v. California, 386 U.S. 738 (1967). Martinez argued before the federal habeas court that filing the Anders brief constituted ineffective assistance. The Court of Appeals did not decide whether that was so. Rather, it held that because Martinez did not have a right to an attorney in the initial-review collateral proceeding, the attorney’s errors in the initial-review collateral proceeding could not establish cause for the failure to comply with the State’s rules. Thus, the Court of Appeals did not determine whether Martinez’s attorney in his first collateral proceeding was ineffective or whether his claim of ineffective assistance of trial counsel is substantial. And the court did not address the question of prejudice. These issues remain open for a decision on remand. * * * The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered.

567.US.2011_11-246

The Indian Reorganization Act (IRA) authorizes the Secretary of the Interior to acquire property “for the purpose of providing land to Indians.” 25 U. S. C. §465. Petitioner Match-E-Be-Nash-She-Wish Band of Pottawatomi Indians (Band), an Indian tribe federally recognized in 1999, requested that the Secretary take into trust on its behalf a tract of land known as the Bradley Property, which the Band intended to use “for gaming purposes.” The Secretary took title to the Bradley Property in 2009. Respondent David Patchak, who lives near the Bradley Property, filed suit under the Administrative Procedure Act (APA), asserting that §465 did not authorize the Secretary to acquire the property because the Band was not a federally recognized tribe when the IRA was enacted in 1934. Patchak alleged a variety of economic, environmental, and aesthetic harms as a result of the Band’s proposed use of the property to operate a casino, and requested injunctive and declaratory relief reversing the Secretary’s decision to take title to the land. The Band intervened to defend the Secretary’s decision. The District Court did not reach the merits of Patchak’s suit, but ruled that he lacked prudential standing to challenge the Secretary’s acquisition of the Bradley Property. The D. C. Circuit reversed and also rejected the Secretary’s and the Band’s alternative argument that sovereign immunity barred the suit. Held: 1. The United States has waived its sovereign immunity from Patchak’s action. The APA’s general waiver of the Federal Government’s immunity from suit does not apply “if any other statute that grants consent to suit expressly or impliedly forbids the relief which is sought” by the plaintiff. 5 U. S. C. §702. The Government and Band contend that the Quiet Title Act (QTA) is such a statute. The QTA authorizes (and so waives the Government’s sovereign immunity from) a suit by a plaintiff asserting a “right, title, or interest” in real property that conflicts with a “right, title, or interest” the United States claims. 28 U. S. C. §2409a(d). But it contains an exception for “trust or restricted Indian lands.” §2409a(a). To determine whether the “Indian lands” exception bars Patchak’s suit, the Court considers whether the QTA addresses the kind of grievance Patchak advances. It does not, because Patchak’s action is not a quiet title action. The QTA, from its title to its jurisdictional grant to its venue provision, speaks specifically and repeatedly of “quiet title” actions, a term universally understood to refer to suits in which a plaintiff not only challenges someone else’s claim, but also asserts his own right to disputed property. Although Patchak’s suit contests the Secretary’s title, it does not claim any competing interest in the Bradley Property. Contrary to the argument of the Band and Government, the QTA does not more broadly encompass any “civil action . . . to adjudicate a disputed title to real property in which the United States claims an interest.” §2409(a). Rather, §2409a includes a host of indications that the “civil action” at issue is an ordinary quiet title suit. The Band and Government also contend that the QTA’s specific authorization of adverse claimants’ suits creates the negative implication that non-claimants like Patchak cannot challenge Government ownership of land under any statute. That argument is faulty for the reason already given: Patchak is bringing a different claim, seeking different relief, from the kind the QTA addresses. Finally, the Band and Government argue that Patchak’s suit should be treated the same as an adverse claimant’s because both equally implicate the “Indian lands” exception’s policies. That argument must be addressed to Congress. The “Indian lands” exception reflects Congress’s judgment about how far to allow quiet title suits—not all suits challenging the Government’s ownership of property. Pp. 4−14. 2. Patchak has prudential standing to challenge the Secretary’s acquisition. A person suing under the APA must assert an interest that is “arguably within the zone of interests to be protected or regulated by the statute” that he says was violated. Association of Data Processing Service Organizations, Inc. v. Camp, 397 U.S. 150, 153. The Government and Band claim that Patchak’s economic, environmental, and aesthetic injuries are not within §465’s zone of interests because the statute focuses on land acquisition, while Patchak’s injuries relate to the land’s use as a casino. However, §465 has far more to do with land use than the Government and Band acknowledge. Section 465 is the capstone of the IRA’s land provisions, and functions as a primary mechanism to foster Indian tribes’ economic development. The Secretary thus takes title to properties with an eye toward how tribes will use those lands to support such development. The Department’s regulations make this statutory concern with land use clear, requiring the Secretary to acquire land with its eventual use in mind, after assessing potential conflicts that use might create. And because §465 encompasses land’s use, neighbors to the use (like Patchak) are reasonable―indeed, predictable―challengers of the Secretary’s decisions: Their interests, whether economic, environmental, or aesthetic, come within §465’s regulatory ambit. Pp. 14–18. 632 F.3d 702, affirmed and remanded. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Thomas, Ginsburg, Breyer, and Alito, JJ., joined. Sotomayor, J., filed a dissenting opinion. Notes 1 Together with No. 11–247, Salazar, Secretary of the Interior, et al. v. Patchak et al., also on certiorari to the same court.

A provision of the Indian Reorganization Act (IRA), 25 U. S. C. §465, authorizes the Secretary of the Interior (Secretary) to acquire property “for the purpose of providing land for Indians.” Ch. 576, §5, 48Stat. 985. The Secretary here acquired land in trust for an Indian tribe seeking to open a casino. Respondent David Patchak lives near that land and challenges the Secretary’s decision in a suit brought under the Administrative Procedure Act (APA), 5 U. S. C. §701 et seq. Patchak claims that the Secretary lacked authority under §465 to take title to the land, and alleges economic, environmental, and aesthetic harms from the casino’s operation. We consider two questions arising from Patchak’s action. The first is whether the United States has sovereign immunity from the suit by virtue of the Quiet Title Act (QTA), 86Stat. 1176. We think it does not. The second is whether Patchak has prudential standing to challenge the Secretary’s acquisition. We think he does. We therefore hold that Patchak’s suit may proceed. I The Match-E-Be-Nash-She-Wish Band of Pottawatomi Indians (Band) is an Indian tribe residing in rural Michigan. Although the Band has a long history, the Department of the Interior (DOI) formally recognized it only in 1999. See 63 Fed. Reg. 56936 (1998). Two years later, the Band petitioned the Secretary to exercise her authority under §465 by taking into trust a tract of land in Wayland Township, Michigan, known as the Bradley Property. The Band’s application explained that the Band would use the property “for gaming purposes,” with the goal of generating the “revenue necessary to promote tribal economic development, self-sufficiency and a strong tribal government capable of providing its members with sorely needed social and educational programs.” App. 52, 41.[1] In 2005, after a lengthy administrative review, the Secretary announced her decision to acquire the Bradley Property in trust for the Band. See 70 Fed. Reg. 25596. In accordance with applicable regulations, the Secretary committed to wait 30 days before taking action, so that interested parties could seek judicial review. See ibid.; 25 CFR §151.12(b) (2011). Within that window, an organization called Michigan Gambling Opposition (or MichGO) filed suit alleging that the Secretary’s decision violated environmental and gaming statutes. The Secretary held off taking title to the property while that litigation proceeded. Within the next few years, a District Court and the D. C. Circuit rejected MichGO’s claims. See Michigan Gambling Opposition v. Kempthorne, 525 F.3d 23, 27–28 (CADC 2008); Michigan Gambling Opposition v. Norton, 477 F. Supp. 2d 1 (DC 2007). Shortly after the D. C. Circuit ruled against MichGO (but still before the Secretary took title), Patchak filed this suit under the APA advancing a different legal theory. He asserted that §465 did not authorize the Secretary to acquire property for the Band because it was not a feder- ally recognized tribe when the IRA was enacted in 1934. See App. 37. To establish his standing to bring suit, Patchak contended that he lived “in close proximity to” the Bradley Property and that a casino there would “destroy the lifestyle he has enjoyed” by causing “increased traffic,” “increased crime,” “decreased property values,” “an irreversible change in the rural character of the area,” and “other aesthetic, socioeconomic, and environmental problems.” Id., at 30–31. Notably, Patchak did not assert any claim of his own to the Bradley Property. He requested only a declaration that the decision to acquire the land violated the IRA and an injunction to stop the Secretary from accepting title. See id., at 38–39. The Band intervened in the suit to defend the Secretary’s decision. In January 2009, about five months after Patchak filed suit, this Court denied certiorari in MichGO’s case, 555 U.S. 1137, and the Secretary took the Bradley Property into trust. That action mooted Patchak’s request for an injunction to prevent the acquisition, and all parties agree that the suit now effectively seeks to divest the Federal Government of title to the land. See Brief for Match-E-Be-Nash-She-Wish Band of Pottawatomi Indians 17 (hereinafter Tribal Petitioner); Brief for Federal Petitioners 11; Brief for Respondent 24–25. The month after the Government took title, this Court held in Carcieri v. Salazar, 555 U.S. 379, 382 (2009), that §465 authorizes the Secretary to take land into trust only for tribes that were “under federal jurisdiction” in 1934.[2] The District Court dismissed the suit without considering the merits (including the relevance of Carcieri), ruling that Patchak lacked prudential standing to challenge the Secretary’s acquisition of the Bradley Property. The court reasoned that the injuries Patchak alleged fell outside §465’s “zone of interests.” 646 F. Supp. 2d 72, 76 (DC 2009). The D. C. Circuit reversed that determination. See 632 F.3d 702, 704–707 (2011). The court also rejected the Secretary’s and the Band’s alternative argument that by virtue of the QTA, sovereign immunity barred the suit. See id., at 707–712. The latter ruling conflicted with decisions of three Circuits holding that the United States has immunity from suits like Patchak’s. See Neighbors for Rational Development, Inc. v. Norton, 379 F.3d 956, 961–962 (CA10 2004); Metropolitan Water Dist. of Southern Cal. v. United States, 830 F.2d 139, 143–144 (CA9 1987) (per curiam); Florida Dept. of Bus. Regulation v. Department of Interior, 768 F.2d 1248, 1253–1255 (CA11 1985). We granted certiorari to review both of the D. C. Circuit’s holdings, 565 U. S. ___ (2011), and we now affirm. II We begin by considering whether the United States’ sovereign immunity bars Patchak’s suit under the APA. That requires us first to look to the APA itself and then, for reasons we will describe, to the QTA. We conclude that the United States has waived its sovereign immunity from Patchak’s action. The APA generally waives the Federal Government’s immunity from a suit “seeking relief other than money damages and stating a claim that an agency or an officer or employee thereof acted or failed to act in an official capacity or under color of legal authority.” 5 U. S. C. §702. That waiver would appear to cover Patchak’s suit, which objects to official action of the Secretary and seeks only non-monetary relief. But the APA’s waiver of immunity comes with an important carve-out: The waiver does not apply “if any other statute that grants consent to suit expressly or impliedly forbids the relief which is sought” by the plaintiff. Ibid. That provision prevents plaintiffs from exploiting the APA’s waiver to evade limitations on suit contained in other statutes. The question thus becomes whether another statute bars Patchak’s demand for relief. The Government and Band contend that the QTA does so. The QTA authorizes (and so waives the Government’s sovereign immunity from) a particular type of action, known as a quiet title suit: a suit by a plaintiff asserting a “right, title, or interest” in real property that conflicts with a “right, title, or interest” the United States claims. 28 U. S. C. §2409a(d). The statute, however, contains an exception: The QTA’s authorization of suit “does not apply to trust or restricted Indian lands.” §2409a(a). According to the Government and Band, that limitation on quiet title suits satisfies the APA’s carve-out and so forbids Patchak’s suit. In the Band’s words, the QTA exception retains “the United States’ full immunity from suits seeking to challenge its title to or impair its legal interest in Indian trust lands.” Brief for Tribal Petitioner 18. Two hypothetical examples might help to frame consideration of this argument. First, suppose Patchak had sued under the APA claiming that he owned the Bradley Property and that the Secretary therefore could not take it into trust. The QTA would bar that suit, for reasons just suggested. True, it fits within the APA’s general waiver, but the QTA specifically authorizes quiet title actions (which this hypothetical suit is) except when they involve Indian lands (which this hypothetical suit does). In such a circumstance, a plaintiff cannot use the APA to end-run the QTA’s limitations. “[W]hen Congress has dealt in par- ticularity with a claim and [has] intended a specified remedy”—including its exceptions—to be exclusive, that is the end of the matter; the APA does not undo the judgment. Block v. North Dakota ex rel. Board of Univ. and School Lands, 461 U.S. 273, 286, n. 22 (1983) (quoting H. R. Rep. No. 94–1656, p. 13 (1976)). But now suppose that Patchak had sued under the APA claiming only that use of the Bradley Property was causing environmental harm, and raising no objection at all to the Secretary’s title. The QTA could not bar that suit because even though involving Indian lands, it asserts a grievance altogether different from the kind the statute concerns. Justice Scalia, in a former life as Assistant Attorney General, made this precise point in a letter to Congress about the APA’s waiver of immunity (which we hasten to add, given the author, we use not as legislative history, but only for its persuasive force). When a statute “is not addressed to the type of grievance which the plaintiff seeks to assert,” then the statute cannot prevent an APA suit. Id., at 28 (May 10, 1976, letter of Assistant Atty. Gen. A. Scalia).[3] We think that principle controls Patchak’s case: The QTA’s “Indian lands” clause does not render the Government immune because the QTA addresses a kind of grievance different from the one Patchak advances. As we will explain, the QTA—whose full name, recall, is the Quiet Title Act—concerns (no great surprise) quiet title actions. And Patchak’s suit is not a quiet title action, because although it contests the Secretary’s title, it does not claim any competing interest in the Bradley Property. That fact makes the QTA’s “Indian lands” limitation simply inapposite to this litigation. In reaching this conclusion, we need look no further than the QTA’s text. From its title to its jurisdictional grant to its venue provision, the Act speaks specifically and repeatedly of “quiet title” actions. See 86Stat. 1176 (“An Act [t]o permit suits to adjudicate certain real property quiet title actions”); 28 U. S. C. §1346(f) (giving district courts jurisdiction over “civil actions . . . to quiet title” to property in which the United States claims an interest); §1402(d) (setting forth venue for “[a]ny civil action . . . to quiet title” to property in which the United States claims an interest). That term is universally understood to refer to suits in which a plaintiff not only challenges someone else’s claim, but also asserts his own right to disputed property. See, e.g., Black’s Law Dictionary 34 (9th ed. 2009) (defining an “action to quiet title” as “[a] proceeding to establish a plaintiff’s title to land by compelling the adverse claimant to establish a claim or be forever estopped from asserting it”); Grable & Sons Metal Products, Inc. v. Darue Engineering & Mfg., 545 U.S. 308, 315 (2005) (“[T]he facts showing the plaintiffs’ title . . . are essential parts of the plaintiffs’ [quiet title] cause of action” (quoting Hopkins v. Walker, 244 U.S. 486, 490 (1917))). And the QTA’s other provisions make clear that the recurrent statutory term “quiet title action” carries its or- dinary meaning. The QTA directs that the complaint in such an action “shall set forth with particularity the nature of the right, title, or interest which the plaintiff claims in the real property.” 28 U. S. C. §2409a(d). If the plaintiff does not assert any such right (as Patchak does not), the statute cannot come into play.[4] Further, the QTA provides an option for the United States, if it loses the suit, to pay “just compensation,” rather than return the property, to the “person determined to be entitled” to it. §2409a(b). That provision makes perfect sense in a quiet title action: If the plaintiff is found to own the property, the Government can satisfy his claim through an award of money (while still retaining the land for its operations). But the provision makes no sense in a suit like this one, where Patchak does not assert a right to the property. If the United States loses the suit, an award of just compensation to the rightful owner (whoever and wherever he might be) could do nothing to satisfy Patchak’s claim.[5] In two prior cases, we likewise described the QTA as addressing suits in which the plaintiff asserts an ownership interest in Government-held property. In Block v. North Dakota ex rel. Board of Univ. and School Lands, 461 U.S. 273 (1982), we considered North Dakota’s claim to land that the United States viewed as its own. We held that the State could not circumvent the QTA’s statute of limitations by invoking other causes of action, among them the APA. See id., at 277–278, 286, n. 22. The crux of our reasoning was that Congress had enacted the QTA to address exactly the kind of suit North Dakota had brought. Prior to the QTA, we explained, “citizens asserting title to or the right to possession of lands claimed by the United States” had no recourse; by passing the statute, “Congress sought to rectify this state of affairs.” Id., at 282. Our decision reflected that legislative purpose: Congress, we held, “intended the QTA to provide the exclusive means by which adverse claimants could challenge the United States’ title to real property.” Id., at 286. We repeat: “adverse claimants,” meaning plaintiffs who themselves assert a claim to property antagonistic to the Federal Government’s. Our decision in United States v. Mottaz, 476 U.S. 834 (1986), is of a piece. There, we considered whether the QTA, or instead the Tucker Act or General Allotment Act, governed the plaintiff’s suit respecting certain allotments of land held by the United States. We thought the QTA the relevant statute because the plaintiff herself asserted title to the property. Our opinion quoted the plaintiff’s own description of her suit: “At no time in this proceeding did [the plaintiff] drop her claim for title. To the contrary, the claim for title is the essence and bottom line of [the plaintiff’s] case.” Id., at 842 (quoting Brief for Respondent in Mottaz, O. T. 1985, No. 546, p. 3). That fact, we held, brought the suit “within the [QTA’s] scope”: “What [the plaintiff] seeks is a declaration that she alone possesses valid title.” 476 U. S., at 842. So once again, we construed the QTA as addressing suits by adverse claimants. But Patchak is not an adverse claimant—and so the QTA (more specifically, its reservation of sovereign immunity from actions respecting Indian trust lands) cannot bar his suit. Patchak does not contend that he owns the Bradley Property, nor does he seek any relief corresponding to such a claim. He wants a court to strip the United States of title to the land, but not on the ground that it is his and not so that he can possess it. Patchak’s lawsuit therefore lacks a defining feature of a QTA action. He is not trying to disguise a QTA suit as an APA action to circumvent the QTA’s “Indian lands” exception. Rather, he is not bringing a QTA suit at all. He asserts merely that the Secretary’s decision to take land into trust violates a federal statute—a garden-variety APA claim. See 5 U. S. C. §§706(2)(A), (C) (“The reviewing court shall . . . hold unlawful and set aside agency action . . . not in accordance with law [or] in excess of statutory jurisdiction [or] authority”). Because that is true—because in then-Assistant Attorney General Scalia’s words, the QTA is “not addressed to the type of grievance which [Patchak] seeks to assert,” H. R. Rep. 94–1656, at 28—the QTA’s limitation of remedies has no bearing. The APA’s general waiver of sovereign immunity instead applies. The Band and Government, along with the dissent, object to this conclusion on three basic grounds. First, they contend that the QTA speaks more broadly than we have indicated, waiving immunity from suits “to adjudicate a disputed title to real property in which the United States claims an interest.” 28 U. S. C. §2409a(a). That language, the argument goes, encompasses all actions contesting the Government’s legal interest in land, regardless whether the plaintiff claims ownership himself. See Brief for Federal Petitioners 19–20; Reply Brief for Tribal Petitioner 4–6; post, at 8–9 (Sotomayor, J., dissenting). The QTA (not the APA) thus becomes the relevant statute after all—as to both its waiver and its “corresponding” reservation of immunity from suits involving Indian lands. Reply Brief for Tribal Petitioner 6. But the Band and Government can reach that result only by neglecting key words in the relevant provision. That sentence, more fully quoted, reads: “The United States may be named as a party defendant in a civil action under this section to adjudicate a disputed title to real property in which the United States claims an interest.” §2409a(a) (emphasis added). And as we have already noted, “this section”—§2409a—includes a host of indications that the “civil action” at issue is an ordinary quiet title suit: Just recall the section’s title (“Real property quiet title actions”), and its pleading requirements (the plaintiff “shall set forth with particularity the nature of the right, title, or interest which [he] claims”), and its permission to the Government to remedy an infraction by paying “just compensation.” Read with reference to all these provisions (as well as to the QTA’s contemporane ously enacted jurisdictional and venue sections), the waiver clause rebuts, rather than supports, the Band’s and the Government’s argument: That clause speaks not to any suit in which a plaintiff challenges the Government’s title, but only to an action in which the plaintiff also claims an interest in the property. The Band and Government next invoke cases holding that “when a statute provides a detailed mechanism for judicial consideration of particular issues at the behest of particular persons,” the statute may “impliedly preclude[ ]” judicial review “of those issues at the behest of other persons.” Block v. Community Nutrition Institute, 467 U.S. 340, 349 (1984); see United States v. Fausto, 484 U.S. 439, 455 (1988). Here, the Band and Government contend, the QTA’s specific authorization of adverse claimants’ suits creates a negative implication: non-adverse claimants like Patchak cannot challenge Government ownership of land under any other statute. See Reply Brief for Tribal Petitoner 7–10; Reply Brief for Federal Petitioners 7–9; see also post, at 3–4. The QTA, says the Band, thus “preempts [Patchak’s] more general remedies.” Brief for Tribal Petitioner 23 (internal quotation marks omitted). But we think that argument faulty, and the cited cases inapposite, for the reason already given: Patchak is bringing a different claim, seeking different relief, from the kind the QTA addresses. See supra, at 7–10. To see the point, consider a contrasting example. Suppose the QTA authorized suit only by adverse claimants who could assert a property interest of at least a decade’s duration. Then suppose an adverse claimant failing to meet that requirement (because, say, his claim to title went back only five years) brought suit under a general statute like the APA. We would surely bar that suit, citing the cases the Government and Band rely on; in our imaginary statute, Congress delineated the class of persons who could bring a quiet title suit, and that judgment would preclude others from doing so. But here, once again, Patchak is not bringing a quiet title action at all. He is not claiming to own the property, and he is not demanding that the court transfer the property to him. So to succeed in their argument, the Government and Band must go much further than the cited cases: They must say that in authorizing one person to bring one kind of suit seeking one form of relief, Congress barred another person from bringing another kind of suit seeking another form of relief. Presumably, that contention would extend only to suits involving similar subject matter—i.e., the Government’s ownership of property. But that commonality is not itself sufficient. We have never held, and see no cause to hold here, that some general similarity of subject matter can alone trigger a remedial statute’s preclusive effect. Last, the Band and Government argue that we should treat Patchak’s suit as we would an adverse claimant’s because they equally implicate the “Indian lands” exception’s policies. According to the Government, allowing challenges to the Secretary’s trust acquisitions would “pose significant barriers to tribes[’] . . . ability to promote investment and economic development on the lands.” Brief for Federal Petitioners 24. That harm is the same whether or not a plaintiff claims to own the land himself. Indeed, the Band argues that the sole difference in this suit cuts in its direction, because non-adverse claimants like Patchak have “the most remote injuries and indirect interests in the land.” Brief for Tribal Petitioner 13; see Reply Brief for Federal Petitioners 11–12; see also post, at 2, 7, 10.[6] That argument is not without force, but it must be addressed to Congress. In the QTA, Congress made a judgment about how far to allow quiet title suits—to a point, but no further. (The “no further” includes not only the “Indian lands” exception, but one for security interests and water rights, as well as a statute of limitations, a bar on jury trials, jurisdictional and venue constraints, and the just compensation option discussed earlier.) Perhaps Congress would—perhaps Congress should—make the identical judgment for the full range of lawsuits pertaining to the Government’s ownership of land. But that is not our call. The Band assumes that plaintiffs like Patchak have a lesser interest than those bringing quiet title actions, and so should be precluded a fortiori. But all we can say is that Patchak has a different interest. Whether it is lesser, as the Band argues, because not based on property rights; whether it is greater because implicating public interests; or whether it is in the end exactly the same—that is for Congress to tell us, not for us to tell Congress. As the matter stands, Congress has not assimilated to quiet title actions all other suits challenging the Government’s ownership of property. And so when a plaintiff like Patchak brings a suit like this one, it falls within the APA’s general waiver of sovereign immunity. III We finally consider the Band’s and the Government’s alternative argument that Patchak cannot bring this ac- tion because he lacks prudential standing. This Court has long held that a person suing under the APA must satisfy not only Article III’s standing requirements, but an additional test: The interest he asserts must be “arguably within the zone of interests to be protected or regulated by the statute” that he says was violated. Association of Data Processing Service Organizations, Inc. v. Camp, 397 U.S. 150, 153 (1970). Here, Patchak asserts that in taking title to the Bradley Property, the Secretary exceeded her authority under §465, which authorizes the acquisition of property “for the purpose of providing land for Indians.” And he alleges that this statutory violation will cause him economic, environmental, and aesthetic harm as a nearby property owner. See supra, at 3. The Government and Band argue that the relationship between §465 and Patchak’s asserted interests is insufficient. That is so, they contend, because the statute focuses on land acquisition, whereas Patchak’s interests relate to the land’s use as a casino. See Brief for Tribal Petitioner 46 (“The Secretary’s decision to put land into trust does not turn on any particular use of the land, gaming or otherwise[,] . . . [and] thus has no impact on [Patchak] or his asserted interests”); Brief for Federal Petitioners 34 (“[L]and may be taken into trust for a host of purposes that have noth- ing at all to do with gaming”). We find this argument unpersuasive. The prudential standing test Patchak must meet “is not meant to be especially demanding.” Clarke v. Securities Industry Assn., 479 U.S. 388, 399 (1987). We apply the test in keeping with Congress’s “evident intent” when enacting the APA “to make agency action presumptively reviewable.” Ibid. We do not require any “indication of congressional purpose to benefit the would-be plaintiff.” Id., at 399–400.[7] And we have always conspicuously included the word “arguably” in the test to indicate that the benefit of any doubt goes to the plaintiff. The test forecloses suit only when a plaintiff’s “interests are so marginally related to or inconsistent with the purposes implicit in the statute that it cannot reasonably be assumed that Congress intended to permit the suit.” Id., at 399. Patchak’s suit satisfies that standard, because §465 has far more to do with land use than the Government and Band acknowledge. Start with what we and others have said about §465’s context and purpose. As the leading treatise on federal Indian law notes, §465 is “the capstone” of the IRA’s land provisions. F. Cohen, Handbook of Federal Indian Law §15.07[1][a], p. 1010 (2005 ed.) (hereinafter Cohen). And those provisions play a key role in the IRA’s overall effort “to rehabilitate the Indian’s economic life,” Mescalero Apache Tribe v. Jones, 411 U.S. 145, 152 (1973) (internal quotation marks omitted). “Land forms the basis” of that “economic life,” providing the foundation for “tourism, manufacturing, mining, logging, . . . and gaming.” Cohen §15.01, at 965. Section 465 thus functions as a primary mechanism to foster Indian tribes’ economic development. As the D. C. Circuit explained in the MichGO litigation, the section “provid[es] lands sufficient to enable Indians to achieve self-support.” Michigan Gambling, 525 F. 3d, at 31 (internal quotation marks omitted); see Morton v. Mancari, 417 U.S. 535, 542 (1974) (noting the IRA’s economic aspect). So when the Secretary obtains land for Indians under §465, she does not do so in a vacuum. Rather, she takes title to properties with at least one eye directed toward how tribes will use those lands to support economic development. The Department’s regulations make this statutory concern with land use crystal clear. Those regulations permit the Secretary to acquire land in trust under §465 if the “land is necessary to facilitate tribal self-determination, economic development, or Indian housing.” 25 CFR §151.3(a)(3). And they require the Secretary to consider, in evaluating any acquisition, both “[t]he purposes for which the land will be used” and the “poten- tial conflicts of land use which may arise.” §§151.10(c), 151.10(f); see §151.11(a). For “off-reservation acquisitions” made “for business purposes”—like the Bradley Property—the regulations further provide that the tribe must “provide a plan which specifies the anticipated economic benefits associated with the proposed use.” §151.11(c). DOI’s regulations thus show that the statute’s implementation centrally depends on the projected use of a given property. The Secretary’s acquisition of the Bradley Property is a case in point. The Band’s application to the Secretary highlighted its plan to use the land for gaming purposes. See App. 41 (“[T]rust status for this Property is requested in order for the Tribe to acquire property on which it plans to conduct gaming”); id., at 61–62 (“The Tribe intends to . . . renovate the existing . . . building into a gaming fa- cility . . . . to offer Class II and/or Class III gaming”). Simi- larly, DOI’s notice of intent to take the land into trust announced that the land would “be used for the purpose of construction and operation of a gaming facility,” which the Department had already determined would meet the Indian Gaming Regulatory Act’s requirements. 70 Fed. Reg. 25596; 25 U. S. C. §§2701–2721. So from start to finish, the decision whether to acquire the Bradley Prop- erty under §465 involved questions of land use. And because §465’s implementation encompasses these issues, the interests Patchak raises—at least arguably— fall “within the zone . . . protected or regulated by the statute.” If the Government had violated a statute specifically addressing how federal land can be used, no one would doubt that a neighboring landowner would have prudential standing to bring suit to enforce the statute’s limits. The difference here, as the Government and Band point out, is that §465 specifically addresses only land ac- quisition. But for the reasons already given, decisions under the statute are closely enough and often enough entwined with considerations of land use to make that difference immaterial. As in this very case, the Secretary will typically acquire land with its eventual use in mind, after assessing potential conflicts that use might create. See 25 CFR §§151.10(c), 151.10(f), 151.11(a). And so neighbors to the use (like Patchak) are reasonable—indeed, predictable—challengers of the Secretary’s decisions: Their interests, whether economic, environmental, or aesthetic, come within §465’s regulatory ambit. * * * The QTA’s reservation of sovereign immunity does not bar Patchak’s suit. Neither does the doctrine of prudential standing. We therefore affirm the judgment of the D. C. Circuit, and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 Under the Indian Gaming Regulatory Act, 25 U. S. C. §§2701–2721, an Indian tribe may conduct gaming operations on “Indian lands,” §2710, which include lands “held in trust by the United States for the benefit of any Indian tribe,” §2703(4)(B). The application thus re-quested the Secretary to take the action necessary for the Band toopen a casino. 2 The merits of Patchak’s case are not before this Court. We therefore express no view on whether the Band was “under federal jurisdiction” in 1934, as Carcieri requires. Nor do we consider how that question relates to Patchak’s allegation that the Band was not “federally recognized” at the time. Cf. Carcieri, 555 U. S., at 397–399 (Breyer, J., concurring) (discussing this issue). 3 According to the dissent, we should look only to the kind of relief a plaintiff seeks, rather than the type of grievance he asserts, in deciding whether another statute bars an APA action. See post, at 6 (opinion of Sotomayor, J.). But the dissent’s test is inconsistent with the one we adopted in Block, which asked whether Congress had particularly dealt with a “claim.” See Block v. North Dakota ex rel. Board of Univ. and School Lands, 461 U.S. 273, 286, n. 22 (1983). And the dissent’s approach has no obvious limits. Suppose, for example, that Congress passed a statute authorizing a particular form of injunctive relief in a procurement contract suit except when the suit involved a “discretionary function” of a federal employee. Cf. 28 U. S. C. §2680(a). Under the dissent’s method, that exception would preclude any APA suit seeking that kind of injunctive relief if it involved a discretionary function, no matter what the nature of the claim. That implausible result demonstrates that limitations on relief cannot sensibly be un-derstood apart from the claims to which they attach. 4 The dissent contends that the QTA omits two other historical requirements for quiet title suits. See post, at 8. But many States had abandoned those requirements by the time the QTA was passed. See S. Rep. No. 92–575, p. 6 (1971) (noting “wide differences in State statutory and decisional law” on quiet title suits); Steadman, “Forgive the U. S. Its Trespasses?”: Land Title Disputes With the Sovereign—Present Remedies and Prospective Reforms, 1972 Duke L. J. 15, 48–49, and n. 152 (stating that cases had disputed whether a quiet title plaintiff needed to possess the land); Welch v. Kai, 4 Cal. App. 3d 374, 380–381, 84 Cal. Rptr. 619, 622–623 (1970) (allowing a quiet title action when the plaintiff claimed only an easement); Benson v. Fekete, 424 S.W.2d 729 (Mo. 1968) (en banc) (same). So Congress in enacting the QTA essentially chose one contemporaneous form of quiet title action. 5 The legislative history, for those who think it useful, further shows that the QTA addresses quiet title actions, as ordinarily conceived. The Senate Report states that the QTA aimed to alleviate the “[g]rave inequity” to private parties “excluded, without benefit of a recourse to the courts, from lands they have reason to believe are rightfully theirs.” S. Rep. No. 92–575, at 1. Similarly, the House Report notes that the history of quiet title actions “goes back to the Courts of England,” and provided as examples “a plaintiff whose title to land was continually being subjected to litigation in the law courts,” and “one who feared that an outstanding deed or other interest might cause a claim to be presented in the future.” H. R. Rep. No. 92–1559, p. 6 (1972). From top to bottom, these reports show that Congress thought itself to be authorizing bread-and-butter quiet title actions, in which a plaintiff asserts a right, title, or interest of his own in disputed land. 6 In a related vein, the dissent argues that our holding will undermine the QTA’s “Indian lands” exception by allowing adverse claimants to file APA complaints concealing their ownership interests or to recruit third parties to bring suit on their behalf. See post, at 9–11. But we think that concern more imaginary than real. We have trouble conceiving of a plausible APA suit that omits mention of an adverse claimant’s interest in property yet somehow leads to relief recognizing that very interest. 7 For this reason, the Band’s statement that Patchak is “not an Indian or tribal official seeking land” and does not “claim an interest in advancing tribal development,” Brief for Tribal Petitioner 42, is beside the point. The question is not whether §465 seeks to benefit Patchak; everyone can agree it does not. The question is instead, as the Band’s and the Government’s main argument acknowledges, whether issues of land use (arguably) fall within §465’s scope—because if they do, a neighbor complaining about such use may sue to enforce the statute’s limits. See infra this page and 16–17.

566.US.2011_10-1150

Although “laws of nature, natural phenomena, and abstract ideas” are not patentable subject matter under §101 of the Patent Act, Diamond v. Diehr, 450 U.S. 175, 185, “an application of a law of nature . . . to a known structure or process may [deserve] patent protection,” id., at 187. But to transform an unpatentable law of nature into a patent-eligible application of such a law, a patent must do more than simply state the law of nature while adding the words “apply it.” See, e.g., Gottschalk v. Benson, 409 U.S. 63, 71–72. It must limit its reach to a particular, inventive application of the law. Respondent, Prometheus Laboratories, Inc. (Prometheus), is the sole and exclusive licensee of the two patents at issue, which concern the use of thiopurine drugs to treat autoimmune diseases. When ingested, the body metabolizes the drugs, producing metabolites in the bloodstream. Because patients metabolize these drugs differently, doctors have found it difficult to determine whether a particular patient’s dose is too high, risking harmful side effects, or too low, and so likely ineffective. The patent claims here set forth processes embodying researchers’ findings that identify correlations between metabolite levels and likely harm or ineffectiveness with precision. Each claim recites (1) an “administering” step—instructing a doctor to administer the drug to his patient—(2) a “determining” step—telling the doctor to measure the resulting metabolite levels in the patient’s blood—and (3) a “wherein” step—describing the metabolite concentrations above which there is a likelihood of harmful side-effects and below which it is likely that the drug dosage is ineffective, and informing the doctor that metabolite concentrations above or below these thresholds “indicate a need” to decrease or increase (respectively) the drug dosage. Petitioners Mayo Collaborative Services and Mayo Clinic Rochester (Mayo) bought and used diagnostic tests based on Prometheus’ patents. But in 2004 Mayo announced that it intended to sell and market its own, somewhat different, diagnostic test. Prometheus sued Mayo contending that Mayo’s test infringed its patents. The District Court found that the test infringed the patents but granted summary judgment to Mayo, reasoning that the processes claimed by the patents effectively claim natural laws or natural phenomena—namely, the correlations between thiopurine metabolite levels and the toxicity and efficacy of thiopurine drugs—and therefore are not patentable. The Federal Circuit reversed, finding the processes to be patent eligible under the Circuit’s “machine or transformation test.” On remand from this Court for reconsideration in light of Bilski v. Kappos, 561 U. S. ___, which clarified that the “machine or transformation test” is not a definitive test of patent eligibility, id., at ___–___, the Federal Circuit reaffirmed its earlier conclusion. Held: Prometheus’ process is not patent eligible. Pp. 8–24. (a) Because the laws of nature recited by Prometheus’ patent claims—the relationships between concentrations of certain metabolites in the blood and the likelihood that a thiopurine drug dosage will prove ineffective or cause harm—are not themselves patentable, the claimed processes are not patentable unless they have additional features that provide practical assurance that the processes are genuine applications of those laws rather than drafting efforts designed to monopolize the correlations. The three additional steps in the claimed processes here are not themselves natural laws but neither are they sufficient to transform the nature of the claims. The “administering” step simply identifies a group of people who will be interested in the correlations, namely, doctors who used thiopurine drugs to treat patients suffering from autoimmune disorders. Doctors had been using these drugs for this purpose long before these patents existed. And a “prohibition against patenting abstract ideas ‘cannot be circumvented by attempting to limit the use of the formula to a particular technological environment.’ ” Bilski, supra, at ___. The “wherein” clauses simply tell a doctor about the relevant natural laws, adding, at most, a suggestion that they should consider the test results when making their treatment decisions. The “determining” step tells a doctor to measure patients’ metabolite levels, through whatever process the doctor wishes to use. Because methods for making such determinations were well known in the art, this step simply tells doctors to engage in well-understood, routine, conventional activity previously engaged in by scientists in the field. Such activity is normally not sufficient to transform an unpatentable law of nature into a patent-eligible application of such a law. Parker v. Flook, 437 U.S. 584, 590. Finally, considering the three steps as an ordered combination adds nothing to the laws of nature that is not already present when the steps are considered separately. Pp. 8–11. (b) A more detailed consideration of the controlling precedents reinforces this conclusion. Pp. 11–19. (1) Diehr and Flook, the cases most directly on point, both addressed processes using mathematical formulas that, like laws of nature, are not themselves patentable. In Diehr, the overall process was patent eligible because of the way the additional steps of the process integrated the equation into the process as a whole. 450 U. S., at 187. These additional steps transformed the process into an inventive application of the formula. But in Flook, the additional steps of the process did not limit the claim to a particular application, and the particular chemical processes at issue were all “well known,” to the point where, putting the formula to the side, there was no “inventive concept” in the claimed application of the formula. 437 U. S., at 594. Here, the claim presents a case for patentability that is weaker than Diehr’s patent-eligible claim and no stronger than Flook’s unpatentable one. The three steps add nothing specific to the laws of nature other than what is well-understood, routine, conventional activity, previously engaged in by those in the field. Pp. 11–13. (2) Further support for the view that simply appending conventional steps, specified at a high level of generality, to laws of nature, natural phenomena, and abstract ideas cannot make those laws, phenomena, and ideas patentable is provided in O’Reilly v. Morse, 15 How. 62, 114–115; Neilson v. Harford, Webster’s Patent Cases 295, 371; Bilski, supra, at ___–___; and Benson, supra, at 64, 65, 67. Pp. 14–16. (3) This Court has repeatedly emphasized a concern that patent law not inhibit future discovery by improperly tying up the use of laws of nature and the like. See, e.g., Benson, 409 U. S., at 67, 68. Rewarding with patents those who discover laws of nature might encourage their discovery. But because those laws and principles are “the basic tools of scientific and technological work,” id., at 67, there is a danger that granting patents that tie up their use will inhibit future innovation, a danger that becomes acute when a patented process is no more than a general instruction to “apply the natural law,” or otherwise forecloses more future invention than the underlying discovery could reasonably justify. The patent claims at issue implicate this concern. In telling a doctor to measure metabolite levels and to consider the resulting measurements in light of the correlations they describe, they tie up his subsequent treatment decision regardless of whether he changes his dosage in the light of the inference he draws using the correlations. And they threaten to inhibit the development of more refined treatment recommendations that combine Prometheus’ correlations with later discoveries. This reinforces the conclusion that the processes at issue are not patent eligible, while eliminating any temptation to depart from case law precedent. Pp. 16–19. (c) Additional arguments supporting Prometheus’ position—that the process is patent eligible because it passes the “machine or transformation test”; that, because the particular laws of nature that the claims embody are narrow and specific, the patents should be upheld; that the Court should not invalidate these patents under §101 because the Patent Act’s other validity requirements will screen out overly broad patents; and that a principle of law denying patent coverage here will discourage investment in discoveries of new diagnostic laws of nature—do not lead to a different conclusion. Pp. 19–24. 628 F.3d 1347, reversed. Breyer, J., delivered the opinion for a unanimous Court.

Section 101 of the Patent Act defines patentable subject matter. It says: “Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.” 35 U. S. C. §101. The Court has long held that this provision contains an important implicit exception. “[L]aws of nature, natural phenomena, and abstract ideas” are not patentable. Diamond v. Diehr, 450 U.S. 175, 185 (1981); see also Bilski v. Kappos, 561 U. S. ___, ___ (2010) (slip op., at 5); Diamond v. Chakrabarty, 447 U.S. 303, 309 (1980); Le Roy v. Tatham, 14 How. 156, 175 (1853); O’Reilly v. Morse, 15 How. 62, 112–120 (1854); cf. Neilson v. Harford, Webster’s Patent Cases 295, 371 (1841) (English case discussing same). Thus, the Court has written that “a new mineral discovered in the earth or a new plant found in the wild is not patentable subject matter. Likewise, Einstein could not patent his celebrated law that E=mc2; nor could Newton have patented the law of gravity. Such discoveries are ‘manifestations of . . . nature, free to all men and reserved exclusively to none.’ ” Chakrabarty, supra, at 309 (quoting Funk Brothers Seed Co. v. Kalo Inoculant Co., 333 U.S. 127, 130 (1948)). “Phenomena of nature, though just discovered, mental processes, and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work.” Gottschalk v. Benson, 409 U.S. 63, 67 (1972). And monopolization of those tools through the grant of a patent might tend to impede innovation more than it would tend to promote it. The Court has recognized, however, that too broad an interpretation of this exclusionary principle could eviscerate patent law. For all inventions at some level embody, use, reflect, rest upon, or apply laws of nature, natural phenomena, or abstract ideas. Thus, in Diehr the Court pointed out that “ ‘a process is not unpatentable simply because it contains a law of nature or a mathematical algorithm.’ ” 450 U. S., at 187 (quoting Parker v. Flook, 437 U.S. 584, 590 (1978)). It added that “an application of a law of nature or mathematical formula to a known structure or process may well be deserving of patent protection.” Diehr, supra, at 187. And it emphasized Justice Stone’s similar observation in Mackay Radio & Telegraph Co. v. Radio Corp. of America, 306 U.S. 86 (1939): “ ‘While a scientific truth, or the mathematical expression of it, is not a patentable invention, a novel and useful structure created with the aid of knowledge of scientific truth may be.’ ” 450 U. S., at 188 (quoting Mackay Radio, supra, at 94). See also Funk Brothers, supra, at 130 (“If there is to be invention from [a discovery of a law of nature], it must come from the application of the law of nature to a new and useful end”). Still, as the Court has also made clear, to transform an unpatentable law of nature into a patent-eligible application of such a law, one must do more than simply state the law of nature while adding the words “apply it.” See, e.g., Benson, supra, at 71–72. The case before us lies at the intersection of these basic principles. It concerns patent claims covering processes that help doctors who use thiopurine drugs to treat patients with autoimmune diseases determine whether a given dosage level is too low or too high. The claims purport to apply natural laws describing the relationships between the concentration in the blood of certain thiopurine metabolites and the likelihood that the drug dosage will be ineffective or induce harmful side-effects. We must determine whether the claimed processes have transformed these unpatentable natural laws into patent-eligible applications of those laws. We conclude that they have not done so and that therefore the processes are not patentable. Our conclusion rests upon an examination of the particular claims before us in light of the Court’s precedents. Those cases warn us against interpreting patent statutes in ways that make patent eligibility “depend simply on the draftsman’s art” without reference to the “principles underlying the prohibition against patents for [natural laws].” Flook, supra, at 593. They warn us against upholding patents that claim processes that too broadly preempt the use of a natural law. Morse, supra, at 112–120; Benson, supra, at 71–72. And they insist that a process that focuses upon the use of a natural law also contain other elements or a combination of elements, sometimes referred to as an “inventive concept,” sufficient to ensure that the patent in practice amounts to significantly more than a patent upon the natural law itself. Flook, supra, at 594; see also Bilski, supra, at ___ (slip op., at 14) (“[T]he prohibition against patenting abstract ideas ‘cannot be circumvented by attempting to limit the use of the formula to a particular technological environment’ or adding ‘insignificant postsolution activity’ ” (quoting Diehr, supra, at 191–192)). We find that the process claims at issue here do not satisfy these conditions. In particular, the steps in the claimed processes (apart from the natural laws themselves) involve well-understood, routine, conventional activity previously engaged in by researchers in the field. At the same time, upholding the patents would risk disproportionately tying up the use of the underlying nat- ural laws, inhibiting their use in the making of further discoveries. I A The patents before us concern the use of thiopurine drugs in the treatment of autoimmune diseases, such as Crohn’s disease and ulcerative colitis. When a patient ingests a thiopurine compound, his body metabolizes the drug, causing metabolites to form in his bloodstream. Because the way in which people metabolize thiopurine compounds varies, the same dose of a thiopurine drug affects different people differently, and it has been difficult for doctors to determine whether for a particular patient a given dose is too high, risking harmful side effects, or too low, and so likely ineffective. At the time the discoveries embodied in the patents were made, scientists already understood that the levels in a patient’s blood of certain metabolites, including, in particular, 6-thioguanine and its nucleotides (6–TG) and 6-methyl-mercaptopurine (6–MMP), were correlated with the likelihood that a particular dosage of a thiopurine drug could cause harm or prove ineffective. See U. S. Patent No. 6,355,623, col. 8, ll. 37–40, 2 App. 10. (“Previous studies suggested that measurement of 6–MP metabolite levels can be used to predict clinical efficacy and tol- erance to azathioprine or 6–MP” (citing Cuffari, Théorêt, Latour, & Seidman, 6-Mercaptopurine Metabolism in Crohn’s Disease: Correlation with Efficacy and Toxicity, 39 Gut 401 (1996))). But those in the field did not know the precise correlations between metabolite levels and likely harm or ineffectiveness. The patent claims at issue here set forth processes embodying researchers’ findings that identified these correlations with some precision. More specifically, the patents—U. S. Patent No. 6,355,623 (’623 patent) and U. S. Patent No. 6,680,302 (’302 patent)—embody findings that concentrations in a patient’s blood of 6–TG or of 6–MMP metabolite beyond a certain level (400 and 7000 picomoles per 8x108 red blood cells, respectively) indicate that the dosage is likely too high for the patient, while concentrations in the blood of 6–TG metabolite lower than a certain level (about 230 picomoles per 8x108 red blood cells) indicate that the dosage is likely too low to be effective. The patent claims seek to embody this research in a set of processes. Like the Federal Circuit we take as typical claim 1 of the ’623 Patent, which describes one of the claimed processes as follows: “A method of optimizing therapeutic efficacy for treatment of an immune-mediated gastrointestinal disorder, comprising: “(a) administering a drug providing 6-thioguanine to a subject having said immune-mediated gastrointestinal disorder; and “(b) determining the level of 6-thioguanine in said subject having said immune-mediated gastrointestinal disorder, “wherein the level of 6-thioguanine less than about 230 pmol per 8x108 red blood cells indicates a need to increase the amount of said drug subsequently administered to said subject and “wherein the level of 6-thioguanine greater than about 400 pmol per 8x108 red blood cells indicates a need to decrease the amount of said drug subsequently administered to said subject.” ’623 patent, col. 20, ll. 10–20, 2 App. 16. For present purposes we may assume that the other claims in the patents do not differ significantly from claim 1. B Respondent, Prometheus Laboratories, Inc. (Prometheus), is the sole and exclusive licensee of the ’623 and ’302 patents. It sells diagnostic tests that embody the processes the patents describe. For some time petitioners, Mayo Clinic Rochester and Mayo Collaborative Services (collectively Mayo), bought and used those tests. But in 2004 Mayo announced that it intended to begin using and selling its own test—a test using somewhat higher metabolite levels to determine toxicity (450 pmol per 8x108 for 6–TG and 5700 pmol per 8x108 for 6–MMP). Prometheus then brought this action claiming patent infringement. The District Court found that Mayo’s test infringed claim 7 of the ’623 patent. App. to Pet. for Cert. 110a–115a. In interpreting the claim, the court accepted Prometheus’ view that the toxicity-risk level numbers in Mayo’s test and the claim were too similar to render the tests significantly different. The number Mayo used (450) was too close to the number the claim used (400) to matter given appropriate margins of error. Id., at 98a–107a. The District Court also accepted Prometheus’ view that a doctor using Mayo’s test could violate the patent even if he did not actually alter his treatment decision in the light of the test. In doing so, the court construed the claim’s language, “indicates a need to decrease” (or “to increase”), as not limited to instances in which the doctor actually decreases (or increases) the dosage level where the test results suggest that such an adjustment is advisable. Id., at 107a–109a; see also Brief for Respondent i (describing claimed processes as methods “for improving . . . treatment . . . by using individualized metabolite measurements to inform the calibration of . . . dosages of . . . thiopurines” (emphasis added)). Nonetheless the District Court ultimately granted summary judgment in Mayo’s favor. The court reasoned that the patents effectively claim natural laws or natural phenomena—namely the correlations between thiopurine metabolite levels and the toxicity and efficacy of thiopurine drug dosages—and so are not patentable. App. to Pet. for Cert. 50a–83a. On appeal, the Federal Circuit reversed. It pointed out that in addition to these natural correlations, the claimed processes specify the steps of (1) “administering a [thiopurine] drug” to a patient and (2) “determining the [resulting metabolite] level.” These steps, it explained, involve the transformation of the human body or of blood taken from the body. Thus, the patents satisfied the Circuit’s “machine or transformation test,” which the court thought sufficient to “confine the patent monopoly within rather definite bounds,” thereby bringing the claims into compliance with §101. 581 F.3d 1336, 1345, 1346–1347 (2009) (internal quotation marks omitted). Mayo filed a petition for certiorari. We granted the petition, vacated the judgment, and remanded the case for reconsideration in light of Bilski, 561 U. S. ___, which clarified that the “machine or transformation test” is not a definitive test of patent eligibility, but only an important and useful clue. Id., at ___–___ (slip op., at 7–8). On remand the Federal Circuit reaffirmed its earlier conclusion. It thought that the “machine-or-transformation test,” understood merely as an important and useful clue, nonetheless led to the “clear and compelling conclusion . . . that the . . . claims . . . do not encompass laws of nature or preempt natural correlations.” 628 F.3d 1347, 1355 (2010). Mayo again filed a petition for certiorari, which we granted. II Prometheus’ patents set forth laws of nature—namely, relationships between concentrations of certain metabolites in the blood and the likelihood that a dosage of a thiopurine drug will prove ineffective or cause harm. Claim 1, for example, states that if the levels of 6–TG in the blood (of a patient who has taken a dose of a thiopurine drug) exceed about 400 pmol per 8x108 red blood cells, then the administered dose is likely to produce toxic side effects. While it takes a human action (the administration of a thiopurine drug) to trigger a manifestation of this relation in a particular person, the relation itself exists in principle apart from any human action. The relation is a consequence of the ways in which thiopurine compounds are metabolized by the body—entirely natural processes. And so a patent that simply describes that relation sets forth a natural law. The question before us is whether the claims do significantly more than simply describe these natural relations. To put the matter more precisely, do the patent claims add enough to their statements of the correlations to allow the processes they describe to qualify as patent-eligible processes that apply natural laws? We believe that the answer to this question is no. A If a law of nature is not patentable, then neither is a process reciting a law of nature, unless that process has additional features that provide practical assurance that the process is more than a drafting effort designed to monopolize the law of nature itself. A patent, for example, could not simply recite a law of nature and then add the instruction “apply the law.” Einstein, we assume, could not have patented his famous law by claiming a process consisting of simply telling linear accelerator operators to refer to the law to determine how much energy an amount of mass has produced (or vice versa). Nor could Archimedes have secured a patent for his famous principle of flotation by claiming a process consisting of simply telling boat builders to refer to that principle in order to determine whether an object will float. What else is there in the claims before us? The process that each claim recites tells doctors interested in the subject about the correlations that the researchers discovered. In doing so, it recites an “administering” step, a “determining” step, and a “wherein” step. These additional steps are not themselves natural laws but neither are they sufficient to transform the nature of the claim. First, the “administering” step simply refers to the relevant audience, namely doctors who treat patients with certain diseases with thiopurine drugs. That audience is a pre-existing audience; doctors used thiopurine drugs to treat patients suffering from autoimmune disorders long before anyone asserted these claims. In any event, the “prohibition against patenting abstract ideas ‘cannot be circumvented by attempting to limit the use of the formula to a particular technological environment.’ ” Bilski, supra, at ___ (slip op., at 14) (quoting Diehr, 450 U. S., at 191–192). Second, the “wherein” clauses simply tell a doctor about the relevant natural laws, at most adding a suggestion that he should take those laws into account when treating his patient. That is to say, these clauses tell the relevant audience about the laws while trusting them to use those laws appropriately where they are relevant to their decisionmaking (rather like Einstein telling linear accelerator operators about his basic law and then trusting them to use it where relevant). Third, the “determining” step tells the doctor to determine the level of the relevant metabolites in the blood, through whatever process the doctor or the laboratory wishes to use. As the patents state, methods for determining metabolite levels were well known in the art. ’623 patent, col. 9, ll. 12–65, 2 App. 11. Indeed, scientists routinely measured metabolites as part of their investigations into the relationships between metabolite levels and efficacy and toxicity of thiopurine compounds. ’623 patent, col. 8, ll. 37–40, id., at 10. Thus, this step tells doctors to engage in well-understood, routine, conventional activity previously engaged in by scientists who work in the field. Purely “conventional or obvious” “[pre]-solution activity” is normally not sufficient to transform an unpatentable law of nature into a patent-eligible application of such a law. Flook, 437 U. S., at 590; see also Bilski, 561 U. S., at ___ (slip op., at 14) (“[T]he prohibition against patenting abstract ideas ‘cannot be circumvented by’ . . . adding ‘insignificant post-solution activity’ ” (quoting Diehr, supra, at 191–192)). Fourth, to consider the three steps as an ordered combination adds nothing to the laws of nature that is not already present when the steps are considered separately. See Diehr, supra, at 188 (“[A] new combination of steps in a process may be patentable even though all the constituents of the combination were well known and in common use before the combination was made”). Anyone who wants to make use of these laws must first administer a thiopurine drug and measure the resulting metabolite concentrations, and so the combination amounts to nothing significantly more than an instruction to doctors to apply the applicable laws when treating their patients. The upshot is that the three steps simply tell doctors to gather data from which they may draw an inference in light of the correlations. To put the matter more suc- cinctly, the claims inform a relevant audience about certain laws of nature; any additional steps consist of well-understood, routine, conventional activity already engaged in by the scientific community; and those steps, when viewed as a whole, add nothing significant beyond the sum of their parts taken separately. For these reasons we believe that the steps are not sufficient to transform unpatentable natural correlations into patentable applications of those regularities. B 1 A more detailed consideration of the controlling precedents reinforces our conclusion. The cases most directly on point are Diehr and Flook, two cases in which the Court reached opposite conclusions about the patent eligibility of processes that embodied the equivalent of natural laws. The Diehr process (held patent eligible) set forth a method for molding raw, uncured rubber into various cured, molded products. The process used a known mathematical equation, the Arrhenius equation, to determine when (depending upon the temperature inside the mold, the time the rubber had been in the mold, and the thickness of the rubber) to open the press. It consisted in effect of the steps of: (1) continuously monitoring the temperature on the inside of the mold, (2) feeding the resulting numbers into a computer, which would use the Arrhenius equation to continuously recalculate the mold-opening time, and (3) configuring the computer so that at the appropriate moment it would signal “a device” to open the press. Diehr, 450 U. S., at 177–179. The Court pointed out that the basic mathematical equation, like a law of nature, was not patentable. But it found the overall process patent eligible because of the way the additional steps of the process integrated the equation into the process as a whole. Those steps included “installing rubber in a press, closing the mold, constantly determining the temperature of the mold, constantly re- calculating the appropriate cure time through the use of the formula and a digital computer, and automatically opening the press at the proper time.” Id., at 187. It nowhere suggested that all these steps, or at least the combination of those steps, were in context obvious, already in use, or purely conventional. And so the patentees did not “seek to pre-empt the use of [the] equation,” but sought “only to foreclose from others the use of that equation in conjunction with all of the other steps in their claimed process.” Ibid. These other steps apparently added to the formula something that in terms of patent law’s objectives had significance—they transformed the process into an inventive application of the formula. The process in Flook (held not patentable) provided a method for adjusting “alarm limits” in the catalytic conversion of hydrocarbons. Certain operating conditions (such as temperature, pressure, and flow rates), which are continuously monitored during the conversion process, signal inefficiency or danger when they exceed certain “alarm limits.” The claimed process amounted to an improved system for updating those alarm limits through the steps of: (1) measuring the current level of the variable, e.g., the temperature; (2) using an apparently novel mathematical algorithm to calculate the current alarm limits; and (3) adjusting the system to reflect the new alarm-limit values. 437 U. S., at 585–587. The Court, as in Diehr, pointed out that the basic mathematical equation, like a law of nature, was not patentable. But it characterized the claimed process as doing nothing other than “provid[ing] a[n unpatentable] formula for computing an updated alarm limit.” Flook, supra, at 586. Unlike the process in Diehr, it did not “explain how the variables used in the formula were to be selected, nor did the [claim] contain any disclosure relating to chemical processes at work or the means of setting off an alarm or adjusting the alarm limit.” Diehr, supra, at 192, n. 14; see also Flook, 437 U. S., at 586. And so the other steps in the process did not limit the claim to a particular application. Moreover, “[t]he chemical processes involved in catalytic conversion of hydrocarbons[,] . . . the practice of monitoring the chemical process variables, the use of alarm limits to trigger alarms, the notion that alarm limit values must be recomputed and readjusted, and the use of comput- ers for ‘automatic monitoring-alarming’ ” were all “well known,” to the point where, putting the formula to the side, there was no “inventive concept” in the claimed application of the formula. Id., at 594. “[P]ost-solution activity” that is purely “conventional or obvious,” the Court wrote, “can[not] transform an unpatentable principle into a patentable process.” Id., at 589, 590. The claim before us presents a case for patentability that is weaker than the (patent-eligible) claim in Diehr and no stronger than the (unpatentable) claim in Flook. Beyond picking out the relevant audience, namely those who administer doses of thiopurine drugs, the claim sim- ply tells doctors to: (1) measure (somehow) the current level of the relevant metabolite, (2) use particular (unpatentable) laws of nature (which the claim sets forth) to calculate the current toxicity/inefficacy limits, and (3) reconsider the drug dosage in light of the law. These instructions add nothing specific to the laws of nature other than what is well-understood, routine, conventional activity, previously engaged in by those in the field. And since they are steps that must be taken in order to apply the laws in question, the effect is simply to tell doctors to apply the law somehow when treating their patients. The process in Diehr was not so characterized; that in Flook was characterized in roughly this way. 2 Other cases offer further support for the view that simply appending conventional steps, specified at a high level of generality, to laws of nature, natural phenomena, and abstract ideas cannot make those laws, phenomena, and ideas patentable. This Court has previously discussed in detail an English case, Neilson, which involved a patent claim that posed a legal problem very similar to the problem now before us. The patent applicant there asserted a claim “for the improved application of air to produce heat in fires, forges, and furnaces, where a blowing apparatus is required. [The invention] was to be applied as follows: The blast or current of air produced by the blowing apparatus was to be passed from it into an air-vessel or receptacle made sufficiently strong to endure the blast; and through or from that vessel or receptacle by means of a tube, pipe, or aperture into the fire, the receptacle be kept artificially heated to a considerable temperature by heat externally applied.” Morse, 15 How., at 114–115. The English court concluded that the claimed process did more than simply instruct users to use the principle that hot air promotes ignition better than cold air, since it explained how the principle could be implemented in an inventive way. Baron Parke wrote (for the court): “It is very difficult to distinguish [Neilson’s claim] from the specification of a patent for a principle, and this at first created in the minds of some of the court much difficulty; but after full consideration, we think that the plaintiff does not merely claim a principle, but a machine embodying a principle, and a very valuable one. We think the case must be considered as if the principle being well known, the plaintiff had first invented a mode of applying it by a mechanical apparatus to furnaces; and his invention then consists in this—by interposing a receptacle for heated air between the blowing apparatus and the furnace. In this receptacle he directs the air to be heated by the application of heat externally to the receptacle, and thus he accomplishes the object of applying the blast, which was before of cold air, in a heated state to the furnace.” Neilson v. Harford, Webster’s Patent Cases, at 371. Thus, the claimed process included not only a law of nature but also several unconventional steps (such as inserting the receptacle, applying heat to the receptacle externally, and blowing the air into the furnace) that confined the claims to a particular, useful application of the principle. In Bilski the Court considered claims covering a process for hedging risks of price changes by, for example, contracting to purchase commodities from sellers at a fixed price, reflecting the desire of sellers to hedge against a drop in prices, while selling commodities to consumers at a fixed price, reflecting the desire of consumers to hedge against a price increase. One claim described the process; another reduced the process to a mathematical formula. 561 U. S., at ___–___ (slip op., at 2–3). The Court held that the described “concept of hedging” was “an unpatentable abstract idea.” Id., at ___ (slip op., at 15). The fact that some of the claims limited hedging to use in commodities and energy markets and specified that “well-known random analysis techniques [could be used] to help establish some of the inputs into the equation” did not undermine this conclusion, for “Flook established that limiting an abstract idea to one field of use or adding token postsolution components did not make the concept patentable.” Id., at ___, ___ (slip op., at 16, 15). Finally, in Benson the Court considered the patentability of a mathematical process for converting binary-coded decimal numerals into pure binary numbers on a general purpose digital computer. The claims “purported to cover any use of the claimed method in a general-purpose digital computer of any type.” 409 U. S., at 64, 65. The Court recognized that “ ‘a novel and useful structure created with the aid of knowledge of scientific truth’ ” might be patentable. Id., at 67 (quoting Mackay Radio, 306 U. S., at 94). But it held that simply implementing a mathematical principle on a physical machine, namely a computer, was not a patentable application of that principle. For the mathematical formula had “no substantial practical application except in connection with a digital computer.” Benson, supra, at 71. Hence the claim (like the claims before us) was overly broad; it did not differ significantly from a claim that just said “apply the algorithm.” 3 The Court has repeatedly emphasized this last mentioned concern, a concern that patent law not inhibit further discovery by improperly tying up the future use of laws of nature. Thus, in Morse the Court set aside as unpatentable Samuel Morse’s general claim for “ ‘the use of the motive power of the electric or galvanic current . . . however developed, for making or printing intelligible characters, letters, or signs, at any distances,’ ” 15 How., at 86. The Court explained: “For aught that we now know some future inventor, in the onward march of science, may discover a mode of writing or printing at a distance by means of the electric or galvanic current, without using any part of the process or combination set forth in the plaintiff’s specification. His invention may be less complicated—less liable to get out of order—less expensive in construction, and in its operation. But yet if it is covered by this patent the inventor could not use it, nor the public have the benefit of it without the permission of this patentee.” Id., at 113. Similarly, in Benson the Court said that the claims before it were “so abstract and sweeping as to cover both known and unknown uses of the [mathematical formula].” 409 U. S., at 67, 68. In Bilski the Court pointed out that to allow “petitioners to patent risk hedging would pre-empt use of this approach in all fields.” 561 U. S., at ___ (slip op., at 15). And in Flook the Court expressed concern that the claimed process was simply “a formula for computing an updated alarm limit,” which might “cover a broad range of potential uses.” 437 U. S., at 586. These statements reflect the fact that, even though rewarding with patents those who discover new laws of nature and the like might well encourage their discovery, those laws and principles, considered generally, are “the basic tools of scientific and technological work.” Benson, supra, at 67. And so there is a danger that the grant of patents that tie up their use will inhibit future innovation premised upon them, a danger that becomes acute when a patented process amounts to no more than an instruction to “apply the natural law,” or otherwise forecloses more future invention than the underlying discovery could reasonably justify. See generally Lemley, Risch, Sichelman, & Wagner, Life After Bilski, 63 Stan. L. Rev. 1315 (2011) (hereinafter Lemley) (arguing that §101 reflects this kind of concern); see also C. Bohannan & H. Hovenkamp, Creation without Restraint: Promoting Liberty and Rivalry in Innovation 112 (2012) (“One problem with [process] patents is that the more abstractly their claims are stated, the more difficult it is to determine precisely what they cover. They risk being applied to a wide range of situations that were not anticipated by the patentee”); W. Landes & R. Posner, The Economic Structure of Intellectual Property Law 305–306 (2003) (The exclusion from patent law of basic truths reflects “both . . . the enormous potential for rent seeking that would be created if property rights could be obtained in them and . . . the enormous transaction costs that would be imposed on would-be users [of those truths]”). The laws of nature at issue here are narrow laws that may have limited applications, but the patent claims that embody them nonetheless implicate this concern. They tell a treating doctor to measure metabolite levels and to consider the resulting measurements in light of the statistical relationships they describe. In doing so, they tie up the doctor’s subsequent treatment decision whether that treatment does, or does not, change in light of the inference he has drawn using the correlations. And they threaten to inhibit the development of more refined treatment recommendations (like that embodied in Mayo’s test), that combine Prometheus’ correlations with later discovered features of metabolites, human physiology or individual patient characteristics. The “determining” step too is set forth in highly general language covering all processes that make use of the correlations after measuring metabolites, including later discovered processes that measure metabolite levels in new ways. We need not, and do not, now decide whether were the steps at issue here less conventional, these features of the claims would prove sufficient to invalidate them. For here, as we have said, the steps add nothing of significance to the natural laws themselves. Unlike, say, a typical patent on a new drug or a new way of using an existing drug, the patent claims do not confine their reach to particular applications of those laws. The presence here of the basic underlying concern that these patents tie up too much future use of laws of nature simply reinforces our conclusion that the processes described in the patents are not patent eligible, while eliminating any temptation to depart from case law precedent. III We have considered several further arguments in support of Prometheus’ position. But they do not lead us to adopt a different conclusion. First, the Federal Circuit, in upholding the patent eligibility of the claims before us, relied on this Court’s determination that “[t]ransformation and reduction of an article ‘to a different state or thing’ is the clue to the patentability of a process claim that does not include particular machines.” Benson, supra, at 70–71 (emphasis added); see also Bilski, supra, at ___ (slip op., at 6–7); Diehr, 450 U. S., at 184; Flook, supra, at 588, n. 9; Cochrane v. Deener, 94 U.S. 780, 788 (1877). It reasoned that the claimed processes are therefore patent eligible, since they involve transforming the human body by administering a thiopurine drug and transforming the blood by analyzing it to determine metabolite levels. 628 F. 3d, at 1356–1357. The first of these transformations, however, is irrelevant. As we have pointed out, the “administering” step simply helps to pick out the group of individuals who are likely interested in applying the law of nature. See supra, at 9. And the second step could be satisfied without transforming the blood, should science develop a totally different system for determining metabolite levels that did not involve such a transformation. See supra, at 18. Regardless, in stating that the “machine-or-transformation” test is an “important and useful clue” to patentability, we have neither said nor implied that the test trumps the “law of nature” exclusion. Bilski, supra, at ___ (slip op., at 6–7) (emphasis added). That being so, the test fails here. Second, Prometheus argues that, because the particular laws of nature that its patent claims embody are narrow and specific, the patents should be upheld. Thus, it encourages us to draw distinctions among laws of nature based on whether or not they will interfere significantly with innovation in other fields now or in the future. Brief for Respondent 42–46; see also Lemley 1342–1344 (making similar argument). But the underlying functional concern here is a relative one: how much future innovation is foreclosed relative to the contribution of the inventor. See supra, at 17. A patent upon a narrow law of nature may not inhibit future research as seriously as would a patent upon Einstein’s law of relativity, but the creative value of the discovery is also considerably smaller. And, as we have previously pointed out, even a narrow law of nature (such as the one before us) can inhibit future research. See supra, at 17–18. In any event, our cases have not distinguished among different laws of nature according to whether or not the principles they embody are sufficiently narrow. See, e.g., Flook, 437 U.S. 584 (holding narrow mathematical formula unpatentable). And this is understandable. Courts and judges are not institutionally well suited to making the kinds of judgments needed to distinguish among differ- ent laws of nature. And so the cases have endorsed a bright-line prohibition against patenting laws of nature, mathematical formulas and the like, which serves as a somewhat more easily administered proxy for the underlying “building-block” concern. Third, the Government argues that virtually any step beyond a statement of a law of nature itself should transform an unpatentable law of nature into a potentially patentable application sufficient to satisfy §101’s demands. Brief for United States as Amicus Curiae. The Government does not necessarily believe that claims that (like the claims before us) extend just minimally beyond a law of nature should receive patents. But in its view, other statutory provisions—those that insist that a claimed process be novel, 35 U. S. C. §102, that it not be “obvious in light of prior art,” §103, and that it be “full[y], clear[ly], concise[ly], and exact[ly]” described, §112—can perform this screening function. In particular, it argues that these claims likely fail for lack of novelty under §102. This approach, however, would make the “law of nature” exception to §101 patentability a dead letter. The approach is therefore not consistent with prior law. The relevant cases rest their holdings upon section 101, not later sections. Bilski, 561 U. S. ___; Diehr, supra; Flook, supra; Benson, 409 U.S. 63. See also H. R. Rep. No. 1923, 82d Cong., 2d Sess., 6 (1952) (“A person may have ‘invented’ a machine or a manufacture, which may include anything under the sun that is made by man, but it is not necessarily patentable under section 101 unless the conditions of the title are fulfilled” (emphasis added)). We recognize that, in evaluating the significance of additional steps, the §101 patent-eligibility inquiry and, say, the §102 novelty inquiry might sometimes overlap. But that need not always be so. And to shift the patent-eligibility inquiry entirely to these later sections risks creating significantly greater legal uncertainty, while assuming that those sections can do work that they are not equipped to do. What role would laws of nature, including newly discovered (and “novel”) laws of nature, play in the Government’s suggested “novelty” inquiry? Intuitively, one would suppose that a newly discovered law of nature is novel. The Government, however, suggests in effect that the novelty of a component law of nature may be disregarded when evaluating the novelty of the whole. See Brief for United States as Amicus Curiae 27. But §§102 and 103 say nothing about treating laws of nature as if they were part of the prior art when applying those sections. Cf. Diehr, 450 U. S., at 188 (patent claims “must be considered as a whole”). And studiously ignoring all laws of nature when evaluating a patent application under §§102 and 103 would “make all inventions unpatentable because all inventions can be reduced to underlying principles of nature which, once known, make their implementation obvious.” Id., at 189, n. 12. See also Eisenberg, Wisdom of the Ages or Dead-Hand Control? Patentable Subject Matter for Diagnostic Methods After In re Bilski, 3 Case W. Res. J. L. Tech. & Internet 1, ___ (forthcoming, 2012) (manuscript, at 85–86, online at http://www.patentlyo.com/ files/eisenberg.wisdomordeadhand.patentlyo.pdf (as vis- ited Mar. 16, 2012, and available in Clerk of Court’s case file)); 2 D. Chisum, Patents §5.03[3] (2005). Section 112 requires only a “written description of the invention . . . in such full, clear, concise, and exact terms as to enable any person skilled in the art . . . to make and use the same.” It does not focus on the possibility that a law of nature (or its equivalent) that meets these conditions will nonetheless create the kind of risk that underlies the law of nature exception, namely the risk that a patent on the law would significantly impede future in- novation. See Lemley 1329–1332 (outlining differences between §§101 and 112); Eisenberg, supra, at ___ (manuscript, at 92–96) (similar). Compare Risch, Everything is Patentable, 75 Tenn. L. Rev. 591 (2008) (defending a minimalist approach to §101) with Lemley (reflecting Risch’s change of mind). These considerations lead us to decline the Government’s invitation to substitute §§102, 103, and 112 inquiries for the better established inquiry under §101. Fourth, Prometheus, supported by several amici, argues that a principle of law denying patent coverage here will interfere significantly with the ability of medical researchers to make valuable discoveries, particularly in the area of diagnostic research. That research, which includes research leading to the discovery of laws of nature, is expensive; it “ha[s] made the United States the world leader in this field”; and it requires protection. Brief for Respondent 52. Other medical experts, however, argue strongly against a legal rule that would make the present claims patent eligible, invoking policy considerations that point in the opposite direction. The American Medical Association, the American College of Medical Genetics, the American Hospital Association, the American Society of Human Genetics, the Association of American Medical Colleges, the Association for Molecular Pathology, and other medical organizations tell us that if “claims to exclusive rights over the body’s natural responses to illness and medical treatment are permitted to stand, the result will be a vast thicket of exclusive rights over the use of critical scientific data that must remain widely available if physicians are to provide sound medical care.” Brief for American College of Medical Genetics et al. as Amici Curiae 7; see also App. to Brief for Association Internationale pour la Protection de la Propriété Intellectuelle et al. as Amici Curiae A6, A16 (methods of medical treatment are not patentable in most of Western Europe). We do not find this kind of difference of opinion surprising. Patent protection is, after all, a two-edged sword. On the one hand, the promise of exclusive rights provides monetary incentives that lead to creation, invention, and discovery. On the other hand, that very exclusivity can impede the flow of information that might permit, indeed spur, invention, by, for example, raising the price of using the patented ideas once created, requiring potential users to conduct costly and time-consuming searches of existing patents and pending patent applications, and requiring the negotiation of complex licensing arrangements. At the same time, patent law’s general rules must govern inventive activity in many different fields of human endeavor, with the result that the practical effects of rules that reflect a general effort to balance these considerations may differ from one field to another. See Bohannan & Hovenkamp, Creation without Restraint, at 98–100. In consequence, we must hesitate before departing from established general legal rules lest a new protective rule that seems to suit the needs of one field produce unforeseen results in another. And we must recognize the role of Congress in crafting more finely tailored rules where necessary. Cf. 35 U. S. C. §§161–164 (special rules for plant patents). We need not determine here whether, from a policy perspective, increased protection for discoveries of diagnostic laws of nature is desirable. * * * For these reasons, we conclude that the patent claims at issue here effectively claim the underlying laws of nature themselves. The claims are consequently invalid. And the Federal Circuit’s judgment is reversed. It is so ordered.

565.US.2011_10-704

Shelly Kelly was afraid that she would be attacked by her boyfriend, Jerry Ray Bowen, while she moved out of her apartment. She therefore requested police protection. Two officers arrived, but they were called away to an emergency. As soon as the officers left, Bowen showed up at the apartment, yelled “I told you never to call the cops on me bitch!” and attacked Kelly, attempting to throw her over a second-story landing. After Kelly escaped to her car, Bowen pointed a sawed-off shotgun at her and threatened to kill her if she tried to leave. Kelly nonetheless sped away as Bowen fired five shots at the car, blowing out one of its tires. Kelly later met with Detective Curt Messerschmidt to discuss the incident. She described the attack in detail, mentioned that Bowen had previously assaulted her, that he had ties to the Mona Park Crips gang, and that he might be staying at the home of his former foster mother, Augusta Millender. Following this conversation, Messerschmidt conducted a detailed investigation, during which he confirmed Bowen’s connection to the Millenders’ home, verified his membership in two gangs, and learned that Bowen had been arrested and convicted for numerous violent and firearm-related offenses. Based on this investigation, Messerschmidt drafted an application for a warrant authorizing a search of the Millenders’ home for all firearms and ammunition, as well as evidence indicating gang membership. Messerschmidt included two affidavits in the warrant application. The first detailed his extensive law enforcement experience and his specialized training in gang-related crimes. The second, expressly incorporated into the search warrant, described the incident and explained why Messerschmidt believed there was probable cause for the search. It also requested that the warrant be endorsed for night service because of Bowen’s gang ties. Before submitting the application to a magistrate for approval, Messerschmidt had it reviewed by his supervisor, Sergeant Robert Lawrence, as well as a police lieutenant and a deputy district attorney. Messerschmidt then submitted the application to a magistrate, who issued the warrant. The ensuing search uncovered only Millender’s shotgun, a California Social Services letter addressed to Bowen, and a box of .45-caliber ammunition. The Millenders filed an action under 42 U. S. C. §1983 against petitioners Messerschmidt and Lawrence, alleging that the officers had subjected them to an unreasonable search in violation of the Fourth Amendment. The District Court granted summary judgment to the Millenders, concluding that the firearm and gang-material aspects of the search warrant were overbroad and that the officers were not entitled to qualified immunity from damages. The Ninth Circuit, sitting en banc, affirmed the denial of qualified immunity. The court held that the warrant’s authorization was unconstitutionally overbroad because the affidavits and warrant failed to establish probable cause that the broad categories of firearms, firearm-related material, and gang-related material were contraband or evidence of a crime, and that a reasonable officer would have been aware of the warrant’s deficiency. Held: The officers are entitled to qualified immunity. Pp. 8−19. (a) Qualified immunity “protects government officials ‘from liability for civil damages insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known.’ ” Pearson v. Callahan, 555 U.S. 223, 231. Where the alleged Fourth Amendment violation involves a search or seizure pursuant to a warrant, the fact that a neutral magistrate has issued a warrant is the clearest indication that the officers acted in an objectively reasonable manner, or in “objective good faith.” United States v. Leon, 468 U.S. 897, 922–923. Nonetheless, that fact does not end the inquiry into objective reasonableness. The Court has recognized an exception allowing suit when “it is obvious that no reasonably competent officer would have concluded that a warrant should issue.” Malley v. Briggs, 475 U.S. 335, 341. The “shield of immunity” otherwise conferred by the warrant, id., at 345, will be lost, for example, where the warrant was “based on an affidavit so lacking in indicia of probable cause as to render official belief in its existence entirely unreasonable.” Leon, 468 U. S., at 923. The threshold for establishing this exception is high. “[I]n the ordinary case, an officer cannot be expected to question the magistrate’s probable-cause determination” because “[i]t is the magistrate’s responsibility to determine whether the officer’s allegations establish probable cause and, if so, to issue a warrant comporting in form with the requirements of the Fourth Amendment.” Leon, supra, at 921. Pp. 8−10. (b) This case does not fall within that narrow exception. It would not be entirely unreasonable for an officer to believe that there was probable cause to search for all firearms and firearm-related materials. Under the circumstances set forth in the warrant, an officer could reasonably conclude that there was a “fair probability” that the sawed-off shotgun was not the only firearm Bowen owned, Illinois v. Gates, 462 U.S. 213, 238, and that Bowen’s sawed-off shotgun was illegal. Cf. 26 U. S. C. §§ 5845(a), 5861(d). Given Bowen’s possession of one illegal gun, his gang membership, willingness to use the gun to kill someone, and concern about the police, it would not be unreasonable for an officer to conclude that Bowen owned other illegal guns. An officer also could reasonably believe that seizure of firearms was necessary to prevent further assaults on Kelly. California law allows a magistrate to issue a search warrant for items “in the possession of any person with the intent to use them as a means of committing a public offense,” Cal. Penal Code Ann. §1524(a)(3), and the warrant application submitted by the officers specifically referenced this provision as a basis for the search. Pp. 10–12. (c) Regarding the warrant’s authorization to search for gang-related materials, a reasonable officer could view Bowen’s attack as motivated not by the souring of his romantic relationship with Kelly but by a desire to prevent her from disclosing details of his gang activity to the police. It would therefore not be unreasonable—based on the facts set out in the affidavit—for an officer to believe that evidence of Bowen’s gang affiliation would prove helpful in prosecuting him for the attack on Kelly, in supporting additional, related charges against Bowen for the assault, or in impeaching Bowen or rebutting his defenses. Moreover, even if this were merely a domestic dispute, a reasonable officer could still conclude that gang paraphernalia found at the Millenders’ residence could demonstrate Bowen’s control over the premises or his connection to other evidence found there. Pp. 12−16. (d) The fact that the officers sought and obtained approval of the warrant application from a superior and a deputy district attorney before submitting it to the magistrate provides further support for the conclusion that an officer could reasonably have believed that the scope of the warrant was supported by probable cause. A contrary conclusion would mean not only that Messerschmidt and Lawrence were “plainly incompetent” in concluding that the warrant was supported by probable cause, Malley, supra, at 341, but that their supervisor, the deputy district attorney, and the magistrate were as well. Pp. 16−18. (e) In holding that the warrant in this case was so obviously defective that no reasonable officer could have believed it to be valid, the court below erred in relying on Groh v. Ramirez, 540 U.S. 551. There, officers who carried out a warrant-approved search were not entitled to qualified immunity because the warrant failed to describe any of the items to be seized and “even a cursory reading of the warrant” would have revealed this defect. Id., at 557. Here, in contrast, any arguable defect would have become apparent only upon a close parsing of the warrant application, and a comparison of the supporting affidavit to the terms of the warrant to determine whether the affidavit established probable cause to search for all the items listed in the warrant. Unlike in Groh, any error here would not be one that “just a simple glance” would have revealed. Id. at 564. Pp. 18−19. 620 F.3d 1016, reversed. Roberts, C. J., delivered the opinion of the Court, in which Scalia, Kennedy, Thomas, Breyer, and Alito, JJ., joined. Breyer, J., filed a concurring opinion. Kagan, J., filed an opinion concurring in part and dissenting in part. Sotomayor, J., filed a dissenting opinion, in which Ginsburg, J., joined.

Petitioner police officers conducted a search of respondents’ home pursuant to a warrant issued by a neutral magistrate. The warrant authorized a search for all guns and gang-related material, in connection with the investigation of a known gang member for shooting at his ex-girlfriend with a pistol-gripped sawed-off shotgun, because she had “call[ed] the cops” on him. App. 56. Respondents brought an action seeking to hold the officers personally liable under 42 U. S. C. §1983, alleging that the search violated their Fourth Amendment rights because there was not sufficient probable cause to believe the items sought were evidence of a crime. In particular, respondents argued that there was no basis to search for all guns simply because the suspect owned and had used a sawed-off shotgun, and no reason to search for gang material because the shooting at the ex-girlfriend for “call[ing] the cops” was solely a domestic dispute. The Court of Appeals for the Ninth Circuit held that the warrant was invalid, and that the officers were not entitled to immu- nity from personal liability because this invalidity was so obvious that any reasonable officer would have recognized it, despite the magistrate’s approval. We disagree and reverse. I A Shelly Kelly decided to break off her romantic relationship with Jerry Ray Bowen and move out of her apartment, to which Bowen had a key. Kelly feared an attack from Bowen, who had previously assaulted her and had been convicted of multiple violent felonies. She therefore asked officers from the Los Angeles County Sheriff’s Department to accompany her while she gathered her things. Deputies from the Sheriff’s Department came to assist Kelly but were called away to respond to an emergency before the move was complete. As soon as the officers left, an enraged Bowen appeared at the bottom of the stairs to the apartment, yelling “I told you never to call the cops on me bitch!” App. 39, 56. Bowen then ran up the stairs to Kelly, grabbed her by her shirt, and tried to throw her over the railing of the second-story landing. When Kelly successfully resisted, Bowen bit her on the shoulder and attempted to drag her inside the apartment by her hair. Kelly again managed to escape Bowen’s grasp, and ran to her car. By that time, Bowen had retrieved a black sawed-off shotgun with a pistol grip. He ran in front of Kelly’s car, pointed the shotgun at her, and told Kelly that if she tried to leave he would kill her. Kelly leaned over, fully depressed the gas pedal, and sped away. Bowen fired at the car a total of five times, blowing out the car’s left front tire in the process, but Kelly managed to escape. Kelly quickly located police officers and reported the assault. She told the police what had happened—that Bowen had attacked her after becoming “angry because she had called the Sheriff’s Department”—and she mentioned that Bowen was “an active member of the ‘Mona Park Crips,’ ” a local street gang. Id., at 39. Kelly also provided the officers with photographs of Bowen. Detective Curt Messerschmidt was assigned to investigate the incident. Messerschmidt met with Kelly to obtain details of the assault and information about Bowen. Kelly described the attack and informed Messerschmidt that she thought Bowen was staying at his foster mother’s home at 2234 East 120th Street. Kelly also informed Messerschmidt of Bowen’s previous assaults on her and of his gang ties. Messerschmidt then conducted a background check on Bowen by consulting police records, California Department of Motor Vehicles records, and the “cal-gang” database. Based on this research, Messerschmidt confirmed Bowen’s connection to the 2234 East 120th Street address. He also confirmed that Bowen was an “active” member of the Mona Park Crips and a “secondary” member of the Dodge City Crips. Id., at 64. Finally, Messerschmidt learned that Bowen had been arrested and convicted for numerous violent and firearm-related offenses. Indeed, at the time of the investigation, Bowen’s “rapsheet” spanned over 17 printed pages, and indicated that he had been arrested at least 31 times. Nine of these arrests were for firearms offenses and six were for violent crimes, including three arrests for assault with a deadly weapon (firearm). Id., at 72–81. Messerschmidt prepared two warrants: one to authorize Bowen’s arrest and one to authorize the search of 2234 East 120th Street. An attachment to the search warrant described the property that would be the object of the search: “All handguns, rifles, or shotguns of any caliber, or any firearms capable of firing ammunition, or firearms or devices modified or designed to allow it [sic] to fire ammunition. All caliber of ammunition, miscellaneous gun parts, gun cleaning kits, holsters which could hold or have held any caliber handgun being sought. Any receipts or paperwork, showing the purchase, ownership, or possession of the handguns being sought. Any firearm for which there is no proof of ownership. Any firearm capable of firing or chambered to fire any caliber ammunition. “Articles of evidence showing street gang membership or affiliation with any Street Gang to include but not limited to any reference to ‘Mona Park Crips’, including writings or graffiti depicting gang membership, activity or identity. Articles of personal property tending to establish the identity of person [sic] in control of the premise or premises. Any photographs or photograph albums depicting persons, vehicles, weapons or locations, which may appear relevant to gang membership, or which may depict the item being sought and or believed to be evidence in the case being investigated on this warrant, or which may depict evidence of criminal activity. Additionally to include any gang indicia that would establish the persons being sought in this warrant, affiliation or membership with the ‘Mona Park Crips’ street gang.” Id., at 52. Two affidavits accompanied Messerschmidt’s warrant ap- plications. The first affidavit described Messerschmidt’s extensive law enforcement experience, including that he had served as a peace officer for 14 years, that he was then assigned to a “specialized unit” “investigating gang related crimes and arresting gang members for various violations of the law,” that he had been involved in “hundreds of gang related incidents, contacts, and or arrests” during his time on the force, and that he had “received specialized training in the field of gang related crimes” and training in “gang related shootings.” Id., at 53–54. The second affidavit—expressly incorporated into the search warrant—explained why Messerschmidt believed there was sufficient probable cause to support the warrant. That affidavit described the facts of the incident involving Kelly and Bowen in great detail, including the weapon used in the assault. The affidavit recounted that Kelly had identified Bowen as the assailant and that she thought Bowen might be found at 2234 East 120th Street. It also reported that Messerschmidt had “conducted an extensive background search on the suspect by utilizing departmental records, state computer records, and other police agency records,” and that from that information he had concluded that Bowen resided at 2234 East 120th Street. Id., at 58. The affidavit requested that the search warrant be endorsed for night service because “information provided by the victim and the cal-gang data base” indicated that Bowen had “gang ties to the Mona Park Crip gang” and that “night service would provide an added element of safety to the community as well as for the deputy personnel serving the warrant.” Id., at 59. The affidavit concluded by noting that Messerschmidt “believe[d] that the items sought” would be in Bowen’s possession and that “recovery of the weapon could be invaluable in the successful prosecution of the suspect involved in this case, and the curtailment of further crimes being committed.” Ibid. Messerschmidt submitted the warrants to his super- visors—Sergeant Lawrence and Lieutenant Ornales—for review. Deputy District Attorney Janet Wilson also reviewed the materials and initialed the search warrant, indicating that she agreed with Messerschmidt’s assessment of probable cause. Id., at 27, 47. Finally, Messerschmidt submitted the warrants to a magistrate. The magistrate approved the warrants and authorized night service. The search warrant was served two days later by a team of officers that included Messerschmidt and Lawrence. Sheriff’s deputies forced open the front door of 2234 East 120th Street and encountered Augusta Millender—a woman in her seventies—and Millender’s daughter and grandson. As instructed by the police, the Millenders went outside while the residence was secured but remained in the living room while the search was conducted. Bowen was not found in the residence. The search did, however, result in the seizure of Augusta Millender’s shotgun, a California Social Services letter addressed to Bowen, and a box of .45-caliber ammunition. Bowen was arrested two weeks later after Messerschmidt found him hiding under a bed in a motel room. B The Millenders filed suit in Federal District Court against the County of Los Angeles, the sheriff’s department, the sheriff, and a number of individual officers, including Messerschmidt and Lawrence. The complaint alleged, as relevant here, that the search warrant was invalid under the Fourth Amendment. It sought damages from Messerschmidt and Lawrence, among others. The parties filed cross motions for summary judgment on the validity of the search warrant. The District Court found the warrant defective in two respects. The District Court concluded that the warrant’s authorization to search for firearms was unconstitutionally overbroad because the “crime specified here was a physical assault with a very specific weapon”—a black sawed-off shotgun with a pistol grip—negating any need to “search for all firearms.” Millender v. County of Los Angeles, Civ. No. 05–2298 (CD Cal., Mar. 15, 2007), App. to Pet. for Cert. 106, 157, 2007 WL 7589200, *21. The court also found the warrant overbroad with respect to the search for gang-related materials, because there “was no evidence that the crime at issue was gang-related.” App. to Pet. for Cert. 157. As a result, the District Court granted summary judgment to the Millenders on their constitutional challenges to the firearm and gang material aspects of the search warrant. Id., at 160. The District Court also rejected the officers’ claim that they were entitled to qualified immunity from damages. Id., at 171. Messerschmidt and Lawrence appealed, and a divided panel of the Court of Appeals for the Ninth Circuit reversed the District Court’s denial of qualified immunity. 564 F.3d 1143 (2009). The court held that the officers were entitled to qualified immunity because “they reasonably relied on the approval of the warrant by a deputy district attorney and a judge.” Id., at 1145. The Court of Appeals granted rehearing en banc and affirmed the District Court’s denial of qualified immunity. 620 F.3d 1016 (CA9 2010). The en banc court concluded that the warrant’s authorization was unconstitutionally overbroad because the affidavit and the warrant failed to “establish[ ] probable cause that the broad categories of firearms, firearm-related material, and gang-related material described in the warrant were contraband or evidence of a crime.” Id., at 1033. In the en banc court’s view, “the deputies had probable cause to search for a single, identified weapon . . . . They had no probable cause to search for the broad class of firearms and firearm-related materials described in the warrant.” Id., at 1027. In addition, “[b]ecause the deputies failed to establish any link between gang-related materials and a crime, the warrant authorizing the search and seizure of all gang-related evidence [was] likewise invalid.” Id., at 1031. Concluding that “a reasonable officer in the deputies’ position would have been well aware of this deficiency,” the en banc court held that the officers were not entitled to qualified immunity. Id., at 1033–1035. There were two separate dissenting opinions. Judge Callahan determined that “the officers had probable cause to search for and seize any firearms in the home in which Bowen, a gang member and felon, was thought to reside.” Id., at 1036. She also concluded that “the officers reasonably relied on their superiors, the district attorney, and the magistrate to correct” any overbreadth in the warrant, and that the officers were entitled to qualified immunity because their actions were not objectively unreasonable. Id., at 1044, 1049. Judge Silverman also dissented, concluding that the “deputies’ belief in the validity of . . . the warrant was entirely reasonable” and that the “record [wa]s totally devoid of any evidence that the deputies acted other than in good faith.” Id., at 1050. Judge Tallman joined both dissents. We granted certiorari. 564 U. S. ___ (2011). II The Millenders allege that they were subjected to an unreasonable search in violation of the Fourth Amendment because the warrant authorizing the search of their home was not supported by probable cause. They seek damages from Messerschmidt and Lawrence for their roles in obtaining and executing this warrant. The validity of the warrant is not before us. The question instead is whether Messerschmidt and Lawrence are entitled to im- munity from damages, even assuming that the warrant should not have been issued. “The doctrine of qualified immunity protects government officials ‘from liability for civil damages insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known.’ ” Pearson v. Callahan, 555 U.S. 223, 231 (2009) (quoting Harlow v. Fitzgerald, 457 U.S. 800, 818 (1982)). Qualified immunity “gives government officials breathing room to make reasonable but mistaken judgments,” and “protects ‘all but the plainly incompetent or those who knowingly violate the law.’ ” Ashcroft v. al-Kidd, 563 U. S. ___, ___ (2011) (slip op., at 12) (quoting Malley v. Briggs, 475 U.S. 335, 341 (1986)). “[W]hether an official protected by qualified immunity may be held personally liable for an allegedly unlawful official action generally turns on the ‘objective legal reasonableness’ of the action, assessed in light of the legal rules that were ‘clearly established’ at the time it was taken.” Anderson v. Creighton, 483 U.S. 635, 639 (1987) (citation omitted). Where the alleged Fourth Amendment violation involves a search or seizure pursuant to a warrant, the fact that a neutral magistrate has issued a warrant is the clearest indication that the officers acted in an objectively reasonable manner or, as we have sometimes put it, in “objective good faith.” United States v. Leon, 468 U.S. 897, 922–923 (1984).[1] Nonetheless, under our precedents, the fact that a neutral magistrate has issued a warrant authorizing the allegedly unconstitutional search or seizure does not end the inquiry into objective reasonableness. Rather, we have recognized an exception allowing suit when “it is obvious that no reasonably competent officer would have concluded that a warrant should issue.” Malley, 475 U. S., at 341. The “shield of immunity” otherwise conferred by the warrant, id., at 345, will be lost, for example, where the warrant was “based on an affidavit so lacking in indicia of probable cause as to render official belief in its existence entirely unreasonable.” Leon, 468 U. S., at 923 (internal quotation marks omitted).[2] Our precedents make clear, however, that the threshold for establishing this exception is a high one, and it should be. As we explained in Leon, “[i]n the ordinary case, an officer cannot be expected to question the magistrate’s probable-cause determination” because “[i]t is the magistrate’s responsibility to determine whether the officer’s allegations establish probable cause and, if so, to issue a warrant comporting in form with the requirements of the Fourth Amendment.” Id., at 921; see also Malley, supra, at 346, n. 9 (“It is a sound presumption that the magistrate is more qualified than the police officer to make a probable cause determination, and it goes without saying that where a magistrate acts mistakenly in issuing a warrant but within the range of professional competence of a magistrate, the officer who requested the warrant cannot be held liable” (internal quotation marks and citation omitted)). III The Millenders contend, and the Court of Appeals held, that their case falls into this narrow exception. According to the Millenders, the officers “failed to provide any facts or circumstances from which a magistrate could properly conclude that there was probable cause to seize the broad classes of items being sought,” and “[n]o reasonable officer would have presumed that such a warrant was valid.” Brief for Respondents 27. We disagree. A With respect to the warrant’s authorization to search for and seize all firearms, the Millenders argue that “a reasonably well-trained officer would have readily perceived that there was no probable cause to search the house for all firearms and firearm-related items.” Id., at 32. Noting that “the affidavit indicated exactly what item was evidence of a crime—the ‘black sawed off shotgun with a pistol grip,’ ” they argue that “[n]o facts established that Bowen possessed any other firearms, let alone that such firearms (if they existed) were ‘contraband or evidence of a crime.’ ” Ibid. (quoting App. 56). Even if the scope of the warrant were overbroad in authorizing a search for all guns when there was information only about a specific one, that specific one was a sawed-off shotgun with a pistol grip, owned by a known gang member, who had just fired the weapon five times in public in an attempt to murder another person, on the asserted ground that she had “call[ed] the cops” on him. Id., at 56. Under these circumstances—set forth in the warrant—it would not have been unreasonable for an officer to conclude that there was a “fair probability” that the sawed-off shotgun was not the only firearm Bowen owned. Illinois v. Gates, 462 U.S. 213, 238 (1983). And it certainly would have been reasonable for an officer to assume that Bowen’s sawed-off shotgun was illegal. Cf. 26 U. S. C. §§5845(a), 5861(d). Evidence of one crime is not always evidence of several, but given Bowen’s possession of one illegal gun, his gang membership, his willingness to use the gun to kill someone, and his concern about the police, a reasonable officer could conclude that there would be additional illegal guns among others that Bowen owned.[3] A reasonable officer also could believe that seizure of the firearms was necessary to prevent further assaults on Kelly. California law allows a magistrate to issue a search warrant for items “in the possession of any person with the intent to use them as a means of committing a public offense,” Cal. Penal Code Ann. §1524(a)(3) (West 2011), and the warrant application submitted by the officers specifically referenced this provision as a basis for the search. App. 48. Bowen had already attempted to murder Kelly once with a firearm, and had yelled “I’ll kill you” as she tried to escape from him. Id., at 56–57. A reasonable officer could conclude that Bowen would make another attempt on Kelly’s life and that he possessed other firearms “with the intent to use them” to that end. Cal. Penal Code Ann. §1524(a)(3). Given the foregoing, it would not have been “entirely unreasonable” for an officer to believe, in the particular circumstances of this case, that there was probable cause to search for all firearms and firearm-related materials. Leon, supra, at 923 (internal quotation marks omitted). With respect to the warrant’s authorization to search for evidence of gang membership, the Millenders contend that “no reasonable officer could have believed that the affidavit presented to the magistrate contained a sufficient basis to conclude that the gang paraphernalia sought was contraband or evidence of a crime.” Brief for Respondents 28. They argue that “the magistrate [could not] have reasonably concluded, based on the affidavit, that Bowen’s gang membership had anything to do with the crime under investigation” because “[t]he affidavit described a ‘spousal assault’ that ensued after Kelly decided to end her ‘on going dating relationship’ with Bowen” and “[n]othing in that description suggests that the crime was gang-related.” Ibid. (quoting App. 55). This effort to characterize the case solely as a domes- tic dispute, however, is misleading. Cf. post, at 5 (Sotomayor, J., dissenting); post, at 2 (Kagan, J., concurring in part and dissenting in part). Messerschmidt began his affidavit in support of the warrant by explaining that he “has been investigating an assault with a deadly weapon incident” and elaborated that the crime was a “spousal assault and an assault with a deadly weapon.” App. 55 (emphasis added). The affidavit also stated that Bowen was “a known Mona Park Crip gang member” “based on information provided by the victim and the cal-gang database,”[4] and that he had attempted to murder Kelly after becoming enraged that she had “call[ed] the cops on [him].” Id., at 56, 58–59. A reasonable officer could certainly view Bowen’s attack as motivated not by the souring of his romantic relationship with Kelly but instead by a desire to prevent her from disclosing details of his gang activity to the police. She was, after all, no longer linked with him as a girlfriend; he had assaulted her in the past; and she had indeed called the cops on him. And, as the affidavit supporting the warrant made clear, Kelly had in fact given the police information about Bowen’s gang ties. Id., at 59.[5] It would therefore not have been unreasonable—based on the facts set out in the affidavit—for an officer to believe that evidence regarding Bowen’s gang affiliation would prove helpful in prosecuting him for the attack on Kelly. See Warden, Md. Penitentiary v. Hayden, 387 U.S. 294, 307 (1967) (holding that the Fourth Amendment allows a search for evidence when there is “probable cause . . . to believe that the evidence sought will aid in a particular apprehension or conviction”). Not only would such evidence help to establish motive, either apart from or in addition to any domestic dispute, it would also support the bringing of additional, related charges against Bowen for the assault. See, e.g., Cal. Penal Code Ann. §136.1(b)(1) (West 1999) (It is a crime to “attempt[ ] to prevent or dissuade another person who has been the victim of a crime or who is witness to a crime from . . . [m]aking any report of that victimization to any . . . law enforcement officer”).[6] In addition, a reasonable officer could believe that evidence demonstrating Bowen’s membership in a gang might prove helpful in impeaching Bowen or rebutting various defenses he could raise at trial. For example, evidence that Bowen had ties to a gang that uses guns such as the one he used to assault Kelly would certainly be relevant to establish that he had familiarity with or access to this type of weapon. Moreover, even if this were merely a domestic dispute, a reasonable officer could still conclude that gang paraphernalia found at the Millenders’ residence would aid in the prosecution of Bowen by, for example, demonstrating Bowen’s connection to other evidence found there. The warrant authorized a search for “any gang indicia that would establish the persons being sought in this warrant,” and “[a]rticles of personal property tending to establish the identity of [the] person in control of the premise or premises.” App. 52. Before the District Court, the Millenders “acknowledge[d] that evidence of who controlled the premises would be relevant if incriminating evidence were found and it became necessary to tie that evidence to a person, ” and the District Court approved that aspect of the warrant on this basis. App. to Pet. for Cert. 158–159 (internal quotation marks omitted). Given Bowen’s known gang affiliation, a reasonable officer could conclude that gang paraphernalia found at the residence would be an effective means of demonstrating Bowen’s control over the premises or his connection to evidence found there.[7] Whatever the use to which evidence of Bowen’s gang involvement might ultimately have been put, it would not have been “entirely unreasonable” for an officer to believe that the facts set out in the affidavit established a fair probability that such evidence would aid the prosecution of Bowen for the criminal acts at issue. Leon, 468 U. S., at 923 (internal quotation marks omitted). B Whether any of these facts, standing alone or taken together, actually establish probable cause is a question we need not decide. Qualified immunity “gives government officials breathing room to make reasonable but mistaken judgments.” al-Kidd, 563 U. S., at ___ (slip op., at 12). The officers’ judgment that the scope of the warrant was supported by probable cause may have been mistaken, but it was not “plainly incompetent.” Malley, 475 U. S., at 341. On top of all this, the fact that the officers sought and obtained approval of the warrant application from a superior and a deputy district attorney before submitting it to the magistrate provides further support for the conclusion that an officer could reasonably have believed that the scope of the warrant was supported by probable cause. Ibid. Before seeking to have the warrant issued by a magistrate, Messerschmidt conducted an extensive investigation into Bowen’s background and the facts of the crime. Based on this investigation, Messerschmidt prepared a detailed warrant application that truthfully laid out the pertinent facts. The only facts omitted—the offi- cers’ knowledge of Bowen’s arrest and conviction records, see supra, at 3—would only have strengthened the warrant. Messerschmidt then submitted the warrant application for review by Lawrence, another superior officer, and a deputy district attorney, all of whom approved the application without any apparent misgivings. Only after this did Messerschmidt seek the approval of a neutral magistrate, who issued the requested warrant. The officers thus “took every step that could reasonably be expected of them.” Massachusetts v. Sheppard, 468 U.S. 981, 989 (1984). In light of the foregoing, it cannot be said that “no officer of reasonable competence would have requested the warrant.” Malley, 475 U. S., at 346, n. 9. Indeed, a contrary conclusion would mean not only that Messerschmidt and Lawrence were “plainly incompetent,” id., at 341, but that their supervisor, the deputy district attorney, and the magistrate were as well. The Court of Appeals, however, gave no weight to the fact that the warrant had been reviewed and approved by the officers’ superiors, a deputy district attorney, and a neutral magistrate. Relying on Malley, the court held that the officers had an “independent responsibility to ensure there [was] at least a colorable argument for probable cause.” 620 F. 3d, at 1034. It explained that “[t]he deputies here had a responsibility to exercise their reasonable professional judgment,” and that “in circumstances such as these a neutral magistrate’s approval (and, a fortiori, a non-neutral prosecutor’s) cannot absolve an officer of liability.” Ibid. (citation omitted). We rejected in Malley the contention that an officer is automatically entitled to qualified immunity for seeking a warrant unsupported by probable cause, simply because a magistrate had approved the application. 475 U. S., at 345. And because the officers’ superior and the deputy district attorney are part of the prosecution team, their review also cannot be regarded as dispositive. But by holding in Malley that a magistrate’s approval does not automatically render an officer’s conduct reasonable, we did not suggest that approval by a magistrate or review by others is irrelevant to the objective reasonableness of the officers’ determination that the warrant was valid. Indeed, we expressly noted that we were not deciding “whether [the officer’s] conduct in [that] case was in fact objectively reasonable.” Id., at 345, n. 8. The fact that the officers secured these approvals is certainly pertinent in assessing whether they could have held a reasonable belief that the warrant was supported by probable cause. C In holding that the warrant in this case was so obviously defective that no reasonable officer could have believed it was valid, the court below relied heavily on our decision in Groh v. Ramirez, 540 U.S. 551 (2004), but that precedent is far afield. There, we held that officers who carried out a warrant-approved search were not entitled to qualified immunity because the warrant in question failed to describe the items to be seized at all. Id., at 557. We explained that “[i]n the portion of the form that called for a description of the ‘person or property’ to be seized, [the applicant] typed a description of [the target’s] two-story blue house rather than the alleged stockpile of firearms.” Id., at 554. Thus, the warrant stated nonsensically that “ ‘there is now concealed [on the specified premises] a certain person or property, namely [a] single dwelling residence two story in height which is blue in color and has two additions attached to the east.’ ” Id., at 554–555, n. 2 (bracketed material in original). Because “even a cursory reading of the warrant in [that] case—perhaps just a simple glance—would have revealed a glaring de-ficiency that any reasonable police officer would have known was constitutionally fatal,” id., at 564, we held that the officer was not entitled to qualified immunity. The instant case is not remotely similar. In contrast to Groh, any defect here would not have been obvious from the face of the warrant. Rather, any arguable defect would have become apparent only upon a close parsing of the warrant application, and a comparison of the affidavit to the terms of the warrant to determine whether the affidavit established probable cause to search for all the items listed in the warrant. This is not an error that “just a simple glance” would have revealed. Ibid. Indeed, unlike in Groh, the officers here did not merely submit their application to a magistrate. They also presented it for review by a superior officer, and a deputy district attorney, before submitting it to the magistrate. The fact that none of the officials who reviewed the application expressed concern about its validity demonstrates that any error was not obvious. Groh plainly does not control the result here. * * * The question in this case is not whether the magistrate erred in believing there was sufficient probable cause to support the scope of the warrant he issued. It is instead whether the magistrate so obviously erred that any reasonable officer would have recognized the error. The occasions on which this standard will be met may be rare, but so too are the circumstances in which it will be appropriate to impose personal liability on a lay officer in the face of judicial approval of his actions. Even if the warrant in this case were invalid, it was not so obviously lacking in probable cause that the officers can be con- sidered “plainly incompetent” for concluding otherwise. Malley, supra, at 341. The judgment of the Court of Appeals denying the officers qualified immunity must therefore be reversed. It is so ordered. Notes 1 Although Leon involved the proper application of the exclusionary rule to remedy a Fourth Amendment violation, we have held that “the same standard of objective reasonableness that we applied in the con-text of a suppression hearing in Leon defines the qualified immun-ity accorded an officer” who obtained or relied on an allegedly invalid warrant. Malley v. Briggs, 475 U.S. 335, 344 (1986) (citation omitted); Groh v. Ramirez, 540 U.S. 551, 565, n. 8 (2004). 2 The dissent relies almost entirely on facts outside the affidavit, including Messerschmidt’s deposition testimony, post, at 4, 11 (opinion of Sotomayor, J.), crime analysis forms, post, at 5, Kelly’s interview, post, at 5–6, and n. 5, Messerschmidt’s notes regarding Kelly’s interview, post, at 5–6, n. 5, and even several briefs filed in the District Court and the Court of Appeals, post, at 8–9, 12. In contrast, the dissent cites the probable cause affidavit itself only twice. See post, at 12. There is no contention before us that the affidavit was misleading in omitting any of the facts on which the dissent relies. Cf. Leon, 468 U. S., at 923. 3 The dissent caricatures our analysis as being that “because Bowen fired one firearm, it was reasonable for the police to conclude . . . that [he] must have possessed others,” post, at 10 (opinion of Sotomayor, J.). This simply avoids coming to grips with the facts of the crime at issue. 4 Although the cal-gang database states that information contained therein cannot be used to establish probable cause, see App. 64, the affidavit makes clear that Kelly also provided this information to Messerschmidt, id., at 59, as she did to the deputies who initially responded to the attack, id., at 39 (describing Kelly’s statement that Bowen was “an active member of the ‘Mona Park Crips’ ”). We therefore need not decide whether the cal-gang database’s disclaimer is relevant to Fourth Amendment analysis. 5 Contrary to the dissent’s suggestion, see post, at 5–6, n. 5 (opinionof Sotomayor, J.), the affidavit’s account of Bowen’s statements is consistent with other accounts of the confrontation, in particular the report prepared by the officers who spoke with Kelly immediately after the attack. See App. 39 (stating that when Bowen “appeared at the base of the stairs and began yelling at [Kelly,] [h]e was angry because she had called the Sheriff’s Department”). And at no point during this litigation has the accuracy of the affidavit’s account of the attack been called into question. 6 The dissent relies heavily on Messerschmidt’s deposition, in which he stated that Bowen’s crime was not a “gang crime.” See post, at 4–7. Messerschmidt’s belief about the nature of the crime, however, is not information he possessed but a conclusion he reached based on information known to him. See Anderson v. Creighton, 483 U.S. 635, 641 (1987). We have “eschew[ed] inquiries into the subjective beliefs of law enforcement officers who seize evidence pursuant to a subsequently invalidated warrant.” United States v. Leon, 468 U.S. 897, 922, n. 23 (1984); see also Harlow v. Fitzgerald, 457 U.S. 800, 815–819 (1982). In any event, as the dissent recognizes, the inquiry under our precedents is whether “a reasonably well-trained officer in petitioner’s position would have known that his affidavit failed to establish probable cause.” Malley, 475 U. S., at 345 (emphasis added). Messerschmidt’s own evaluation does not answer the question whether it would have been unreasonable for an officer to have reached a different conclusion from the facts in the affidavit. See n. 2, supra. 7 The Fourth Amendment does not require probable cause to believe evidence will conclusively establish a fact before permitting a search, but only “probable cause . . . to believe the evidence sought will aid in a particular apprehension or conviction.” Warden, Md. Penitentiary v. Hayden, 387 U.S. 294, 307 (1967) (emphasis added). Even if gang evidence might have turned out not to be conclusive because other members of the Millender household also had gang ties, see post, at 8 (opinion of Sotomayor, J.); post, at 2–3 (opinion of Kagan, J.), a reasonable officer could still conclude that evidence of gang membership would help show Bowen’s connection to the residence. Such evidence could, for example, have displayed Bowen’s gang moniker (“C Jay”)or could have been identified by Kelly as belonging to Bowen. SeeApp. 64.

567.US.2011_10-9646

In each of these cases, a 14-year-old was convicted of murder and sentenced to a mandatory term of life imprisonment without the possibility of parole. In No. 10−9647, petitioner Jackson accompanied two other boys to a video store to commit a robbery; on the way to the store, he learned that one of the boys was carrying a shotgun. Jackson stayed outside the store for most of the robbery, but after he entered, one of his co-conspirators shot and killed the store clerk. Arkansas charged Jackson as an adult with capital felony murder and aggravated robbery, and a jury convicted him of both crimes. The trial court imposed a statutorily mandated sentence of life imprisonment without the possibility of parole. Jackson filed a state habeas petition, arguing that a mandatory life-without-parole term for a 14-year-old violates the Eighth Amendment. Disagreeing, the court granted the State’s motion to dismiss. The Arkansas Supreme Court affirmed. In No. 10−9646, petitioner Miller, along with a friend, beat Miller’s neighbor and set fire to his trailer after an evening of drinking and drug use. The neighbor died. Miller was initially charged as a juvenile, but his case was removed to adult court, where he was charged with murder in the course of arson. A jury found Miller guilty, and the trial court imposed a statutorily mandated punishment of life without parole. The Alabama Court of Criminal Appeals affirmed, holding that Miller’s sentence was not overly harsh when compared to his crime, and that its mandatory nature was permissible under the Eighth Amendment. Held: The Eighth Amendment forbids a sentencing scheme that mandates life in prison without possibility of parole for juvenile homicide offenders. Pp. 6−27. (a) The Eighth Amendment’s prohibition of cruel and unusual punishment “guarantees individuals the right not to be subjected to excessive sanctions.” Roper v. Simmons, 543 U.S. 551, 560. That right “flows from the basic ‘precept of justice that punishment for crime should be graduated and proportioned’ ” to both the offender and the offense. Ibid. Two strands of precedent reflecting the concern with proportionate punishment come together here. The first has adopted categorical bans on sentencing practices based on mismatches between the culpability of a class of offenders and the severity of a penalty. See, e.g., Kennedy v. Louisiana, 554 U.S. 407. Several cases in this group have specially focused on juvenile offenders, because of their lesser culpability. Thus, Roper v. Simmons held that the Eighth Amendment bars capital punishment for children, and Graham v. Florida, 560 U. S. ___, concluded that the Amendment prohibits a sentence of life without the possibility of parole for a juvenile convicted of a nonhomicide offense. Graham further likened life without parole for juveniles to the death penalty, thereby evoking a second line of cases. In those decisions, this Court has required sentencing authorities to consider the characteristics of a defendant and the details of his offense before sentencing him to death. See, e.g., Woodson v. North Carolina, 428 U.S. 280 (plurality opinion). Here, the confluence of these two lines of precedent leads to the conclusion that mandatory life without parole for juveniles violates the Eighth Amendment. As to the first set of cases: Roper and Graham establish that children are constitutionally different from adults for sentencing purposes. Their “ ‘lack of maturity’ ” and “ ‘underdeveloped sense of responsibility’ ” lead to recklessness, impulsivity, and heedless risk-taking. Roper, 543 U. S., at 569. They “are more vulnerable . . . to negative influences and outside pressures,” including from their family and peers; they have limited “contro[l] over their own environment” and lack the ability to extricate themselves from horrific, crime-producing settings. Ibid. And because a child’s character is not as “well formed” as an adult’s, his traits are “less fixed” and his actions are less likely to be “evidence of irretrievabl[e] deprav[ity].” Id., at 570. Roper and Graham emphasized that the distinctive attributes of youth diminish the penological justifications for imposing the harshest sentences on juvenile offenders, even when they commit terrible crimes. While Graham’s flat ban on life without parole was for nonhomicide crimes, nothing that Graham said about children is crime-specific. Thus, its reasoning implicates any life-without-parole sentence for a juvenile, even as its categorical bar relates only to nonhomicide offenses. Most fundamentally, Graham insists that youth matters in determining the appropriateness of a lifetime of incarceration without the possibility of parole. The mandatory penalty schemes at issue here, however, prevent the sentencer from considering youth and from assessing whether the law’s harshest term of imprisonment proportionately punishes a juvenile offender. This contravenes Graham’s (and also Roper’s) foundational principle: that imposition of a State’s most severe penalties on juvenile offenders cannot proceed as though they were not children. Graham also likened life-without-parole sentences for juveniles to the death penalty. That decision recognized that life-without-parole sentences “share some characteristics with death sentences that are shared by no other sentences.” 560 U. S., at ___. And it treated life without parole for juveniles like this Court’s cases treat the death penalty, imposing a categorical bar on its imposition for nonhomicide offenses. By likening life-without-parole sentences for juveniles to the death penalty, Graham makes relevant this Court’s cases demanding individualized sentencing in capital cases. In particular, those cases have emphasized that sentencers must be able to consider the mitigating qualities of youth. In light of Graham’s reasoning, these decisions also show the flaws of imposing mandatory life-without-parole sentences on juvenile homicide offenders. Pp. 6−17. (b) The counterarguments of Alabama and Arkansas are unpersuasive. Pp. 18–27. (1) The States first contend that Harmelin v. Michigan, 501 U.S. 957, forecloses a holding that mandatory life-without-parole sentences for juveniles violate the Eighth Amendment. Harmelin declined to extend the individualized sentencing requirement to noncapital cases “because of the qualitative difference between death and all other penalties.” Id., at 1006 (Kennedy, J., concurring in part and concurring in judgment). But Harmelin had nothing to do with children, and did not purport to apply to juvenile offenders. Indeed, since Harmelin, this Court has held on multiple occasions that sentencing practices that are permissible for adults may not be so for children. See Roper, 543 U.S. 551; Graham, 560 U. S ___. The States next contend that mandatory life-without-parole terms for juveniles cannot be unconstitutional because 29 jurisdictions impose them on at least some children convicted of murder. In considering categorical bars to the death penalty and life without parole, this Court asks as part of the analysis whether legislative enactments and actual sentencing practices show a national consensus against a sentence for a particular class of offenders. But where, as here, this Court does not categorically bar a penalty, but instead requires only that a sentencer follow a certain process, this Court has not scrutinized or relied on legislative enactments in the same way. See, e.g., Sumner v. Schuman, 483 U.S. 66. In any event, the “objective indicia of society’s standards,” Graham, 560 U. S., at ___, that the States offer do not distinguish these cases from others holding that a sentencing practice violates the Eighth Amendment. Fewer States impose mandatory life-without-parole sentences on juvenile homicide offenders than authorized the penalty (life-without-parole for nonhomicide offenders) that this Court invalidated in Graham. And as Graham and Thompson v. Oklahoma, 487 U.S. 815, explain, simply counting legislative enactments can present a distorted view. In those cases, as here, the relevant penalty applied to juveniles based on two separate provisions: One allowed the transfer of certain juvenile offenders to adult court, while another set out penalties for any and all individuals tried there. In those circumstances, this Court reasoned, it was impossible to say whether a legislature had endorsed a given penalty for children (or would do so if presented with the choice). The same is true here. Pp. 18–25. (2) The States next argue that courts and prosecutors sufficiently consider a juvenile defendant’s age, as well as his background and the circumstances of his crime, when deciding whether to try him as an adult. But this argument ignores that many States use mandatory transfer systems. In addition, some lodge the decision in the hands of the prosecutors, rather than courts. And even where judges have transfer-stage discretion, it has limited utility, because the decisionmaker typically will have only partial information about the child or the circumstances of his offense. Finally, because of the limited sentencing options in some juvenile courts, the transfer decision may present a choice between a light sentence as a juvenile and standard sentencing as an adult. It cannot substitute for discretion at post-trial sentencing. Pp. 25−27. No. 10−9646, 63 So. 3d 676, and No. 10−9647, 2011 Ark. 49, ___ S. W. 3d ___, reversed and remanded. Kagan, J., delivered the opinion of the Court, in which Kennedy, Ginsburg, Breyer, and Sotomayor, JJ., joined. Breyer, J., filed a concurring opinion, in which Sotomayor, J., joined. Roberts, C. J., filed a dissenting opinion, in which Scalia, Thomas, and Alito, JJ., joined. Thomas, J., filed a dissenting opinion, in which Scalia, J., joined. Alito, J., filed a dissenting opinion, in which Scalia, J., joined. Notes 1 Together with No. 10–9647, Jackson v. Hobbs, Director, Arkansas Department of Correction, on certiorari to the Supreme Court of Arkansas.

The two 14-year-old offenders in these cases were convicted of murder and sentenced to life imprisonment without the possibility of parole. In neither case did the sentencing authority have any discretion to impose a different punishment. State law mandated that each juvenile die in prison even if a judge or jury would have thought that his youth and its attendant characteristics, along with the nature of his crime, made a lesser sentence (for example, life with the possibility of parole) more appropriate. Such a scheme prevents those meting out punishment from considering a juvenile’s “lessened culpability” and greater “capacity for change,” Graham v. Florida, 560 U. S. ___, ___ (2010) (slip op., at 17, 23), and runs afoul of our cases’ requirement of individualized sentencing for defendants facing the most serious penalties. We therefore hold that mandatory life without parole for those under the age of 18 at the time of their crimes violates the Eighth Amendment’s prohibition on “cruel and unusual punishments.” I A In November 1999, petitioner Kuntrell Jackson, then 14 years old, and two other boys decided to rob a video store. En route to the store, Jackson learned that one of the boys, Derrick Shields, was carrying a sawed-off shotgun in his coat sleeve. Jackson decided to stay outside when the two other boys entered the store. Inside, Shields pointed the gun at the store clerk, Laurie Troup, and demanded that she “give up the money.” Jackson v. State, 359 Ark. 87, 89, 194 S.W.3d 757, 759 (2004) (internal quotation marks omitted). Troup refused. A few moments later, Jackson went into the store to find Shields continuing to demand money. At trial, the parties disputed whether Jackson warned Troup that “[w]e ain’t playin’,” or instead told his friends, “I thought you all was playin’.” Id., at 91, 194 S. W. 3d, at 760 (internal quotation marks omitted). When Troup threatened to call the police, Shields shot and killed her. The three boys fled empty-handed. See id., at 89–92, 194 S. W. 3d, at 758–760. Arkansas law gives prosecutors discretion to charge 14-year-olds as adults when they are alleged to have committed certain serious offenses. See Ark. Code Ann. §9–27–318(c)(2) (1998). The prosecutor here exercised that authority by charging Jackson with capital felony murder and aggravated robbery. Jackson moved to transfer the case to juvenile court, but after considering the alleged facts of the crime, a psychiatrist’s examination, and Jackson’s juvenile arrest history (shoplifting and several incidents of car theft), the trial court denied the motion, and an appellate court affirmed. See Jackson v. State, No. 02–535, 2003 WL 193412, *1 (Ark. App., dateJan. 29, 2003); §§9–27–318(d), (e). A jury later convicted Jackson of both crimes. Noting that “in view of [the] verdict, there’s only one possible punishment,” the judge sentenced Jackson to life without parole. App. in No. 10–9647, p. 55 (hereinafter Jackson App.); see Ark. Code Ann. §5–4–104(b) (1997) (“A defendant convicted of capital murder or treason shall be sentenced to death or life imprisonment without parole”).[1] Jackson did not challenge the sentence on appeal, and the Arkansas Supreme Court affirmed the convictions. See 359 Ark. 87, 194 S.W.3d 757. Following Roper v. Simmons, 543 U.S. 551 (2005), in which this Court invalidated the death penalty for all juvenile offenders under the age of 18, Jackson filed a state petition for habeas corpus. He argued, based on Roper’s reasoning, that a mandatory sentence of life without parole for a 14-year-old also violates the Eighth Amendment. The circuit court rejected that argument and granted the State’s motion to dismiss. See Jackson App. 72–76. While that ruling was on appeal, this Court held in Graham v. Florida that life without parole violates the Eighth Amendment when imposed on juvenile nonhomicide offenders. After the parties filed briefs addressing that decision, the Arkansas Supreme Court affirmed the dismissal of Jackson’s petition. See Jackson v. Norris, 2011 Ark. 49, ___ S. W. 3d ___. The majority found that Roper and Graham were “narrowly tailored” to their contexts: “death-penalty cases involving a juvenile and life-imprisonment-without-parole cases for nonhomicide offenses involving a juvenile.” Id., at 5, ___ S. W. 3d, at ___. Two justices dissented. They noted that Jackson was not the shooter and that “any evidence of intent to kill was severely lacking.” Id., at 10, ___ S. W. 3d, at ___ (Danielson, J., dissenting). And they argued that Jackson’s mandatory sentence ran afoul of Graham’s admonition that “ ‘[a]n offender’s age is relevant to the Eighth Amendment, and criminal procedure laws that fail to take defendants’ youthfulness into account at all would be flawed.’ ” Id., at 10–11, ___ S. W. 3d, at ___ (quoting Graham, 560 U. S., at ___ (slip op., at 25)).[2] B Like Jackson, petitioner Evan Miller was 14 years old at the time of his crime. Miller had by then been in and out of foster care because his mother suffered from alcoholism and drug addiction and his stepfather abused him. Miller, too, regularly used drugs and alcohol; and he had attempted suicide four times, the first when he was six years old. See E. J. M. v. State, 928 So. 2d 1077, 1081 (Ala. Crim. App. 2004) (Cobb, J., concurring in result); App. in No. 10–9646, pp. 26–28 (hereinafter Miller App.). One night in 2003, Miller was at home with a friend, Colby Smith, when a neighbor, Cole Cannon, came to make a drug deal with Miller’s mother. See 6 Record in No. 10–9646, p. 1004. The two boys followed Cannon back to his trailer, where all three smoked marijuana and played drinking games. When Cannon passed out, Miller stole his wallet, splitting about $300 with Smith. Miller then tried to put the wallet back in Cannon’s pocket, but Cannon awoke and grabbed Miller by the throat. Smith hit Cannon with a nearby baseball bat, and once released, Miller grabbed the bat and repeatedly struck Cannon with it. Miller placed a sheet over Cannon’s head, told him “ ‘I am God, I’ve come to take your life,’ ” and delivered one more blow. Miller v. State, 63 So. 3d 676, 689 (Ala. Crim. App. 2010). The boys then retreated to Miller’s trailer, but soon decided to return to Cannon’s to cover up evidence of their crime. Once there, they lit two fires. Cannon eventually died from his injuries and smoke inhalation. See id., at 683–685, 689. Alabama law required that Miller initially be charged as a juvenile, but allowed the District Attorney to seek removal of the case to adult court. See Ala. Code §12–15–34 (1977). The D. A. did so, and the juvenile court agreed to the transfer after a hearing. Citing the nature of the crime, Miller’s “mental maturity,” and his prior juvenile offenses (truancy and “criminal mischief”), the Alabama Court of Criminal Appeals affirmed. E. J. M. v. State, No. CR–03–0915, pp. 5–7 (dateAug. 27, 2004) (unpublished memorandum).[3] The State accordingly charged Miller as an adult with murder in the course of arson. That crime (like capital murder in Arkansas) carries a mandatory minimum punishment of life without parole. See Ala. Code §§13A–5–40(9), 13A–6–2(c) (1982). Relying in significant part on testimony from Smith, who had pleaded to a lesser offense, a jury found Miller guilty. He was therefore sentenced to life without the possibility of parole. The Alabama Court of Criminal Appeals affirmed, ruling that life without parole was “not overly harsh when compared to the crime” and that the mandatory nature of the sentencing scheme was permissible under the Eighth Amendment. 63 So. 3d, at 690; see id., at 686–691. The Alabama Supreme Court denied review. We granted certiorari in both cases, see 565 U. S. ___ (2011) (No. 10–9646); 565 U. S. ___ (2011) (No. 10–9647), and now reverse. II The Eighth Amendment’s prohibition of cruel and un- usual punishment “guarantees individuals the right not to be subjected to excessive sanctions.” Roper, 543 U. S., at 560. That right, we have explained, “flows from the basic ‘precept of justice that punishment for crime should be graduated and proportioned’ ” to both the offender and the offense. Ibid. (quoting Weems v. United States, 217 U.S. 349, 367 (1910)). As we noted the last time we consid- ered life-without-parole sentences imposed on juveniles, “[t]he concept of proportionality is central to the Eighth Amendment.” Graham, 560 U. S., at ___ (slip op., at 8). And we view that concept less through a historical prism than according to “ ‘the evolving standards of decency that mark the progress of a maturing society.’ ” Estelle v. Gamble, 429 U.S. 97, 102 (1976) (quoting Trop v. Dulles, 356 U.S. 86, 101 (1958) (plurality opinion)). The cases before us implicate two strands of precedent reflecting our concern with proportionate punishment. The first has adopted categorical bans on sentencing practices based on mismatches between the culpability of a class of offenders and the severity of a penalty. See Graham, 560 U. S., at ___ (slip op., at 9–10) (listing cases). So, for example, we have held that imposing the death penalty for nonhomicide crimes against individuals, or imposing it on mentally retarded defendants, violates the Eighth Amendment. See Kennedy v. Louisiana, 554 U.S. 407 (2008); Atkins v. Virginia, 536 U.S. 304 (2002). Several of the cases in this group have specially focused on juvenile offenders, because of their lesser culpability. Thus, Roper held that the Eighth Amendment bars capital punishment for children, and Graham concluded that the Amendment also prohibits a sentence of life without the possibility of parole for a child who committed a nonhomicide offense. Graham further likened life without parole for juveniles to the death penalty itself, thereby evoking a second line of our precedents. In those cases, we have prohibited mandatory imposition of capital punishment, requiring that sentencing authorities consider the characteristics of a defendant and the details of his offense before sentencing him to death. See Woodson v. North Carolina, 428 U.S. 280 (1976) (plurality opinion); Lockett v. Ohio, 438 U.S. 586 (1978). Here, the confluence of these two lines of precedent leads to the conclusion that mandatory life-without-parole sentences for juveniles violate the Eighth Amendment.[4] To start with the first set of cases: Roper and Graham establish that children are constitutionally different from adults for purposes of sentencing. Because juveniles have diminished culpability and greater prospects for reform, we explained, “they are less deserving of the most severe punishments.” Graham, 560 U. S., at ___ (slip op., at 17). Those cases relied on three significant gaps between juveniles and adults. First, children have a “ ‘lack of maturity and an underdeveloped sense of responsibility,’ ” leading to recklessness, impulsivity, and heedless risk-taking. Roper, 543 U. S., at 569. Second, children “are more vulner- able . . . to negative influences and outside pressures,” including from their family and peers; they have limited “contro[l] over their own environment” and lack the ability to extricate themselves from horrific, crime-producing settings. Ibid. And third, a child’s character is not as “well formed” as an adult’s; his traits are “less fixed” and his actions less likely to be “evidence of irretrievabl[e] deprav[ity].” Id., at 570. Our decisions rested not only on common sense—on what “any parent knows”—but on science and social science as well. Id., at 569. In Roper, we cited studies showing that “ ‘[o]nly a relatively small proportion of adolescents’ ” who engage in illegal activity “ ‘develop entrenched patterns of problem behavior.’ ” Id., at 570 (quoting Steinberg & Scott, Less Guilty by Reason of Adolescence: Developmental Immaturity, Diminished Responsibility, and the Juvenile Death Penalty, 58 Am. Psychologist 1009, 1014 (2003)). And in Graham, we noted that “developments in psychology and brain science continue to show fundamental differences between juvenile and adult minds”—for example, in “parts of the brain involved in behavior control.” 560 U. S., at ___ (slip op., at 17).[5] We reasoned that those findings—of transient rashness, proclivity for risk, and inability to assess consequences—both lessened a child’s “moral culpability” and enhanced the prospect that, as the years go by and neurological development occurs, his “ ‘deficiencies will be reformed.’ ” Id., at ___ (slip op., at 18) (quoting Roper, 543 U. S., at 570). Roper and Graham emphasized that the distinctive at- tributes of youth diminish the penological justifications for imposing the harshest sentences on juvenile offenders, even when they commit terrible crimes. Because “ ‘[t]he heart of the retribution rationale’ ” relates to an offender’s blameworthiness, “ ‘the case for retribution is not as strong with a minor as with an adult.’ ” Graham, 560 U. S., at ___ (slip op., at 20–21) (quoting Tison v. Arizona, 481 U.S. 137, 149 (1987); Roper, 543 U. S., at 571). Nor can deterrence do the work in this context, because “ ‘the same characteristics that render juveniles less culpable than adults’ ”—their immaturity, recklessness, and impetuosity—make them less likely to consider potential punishment. Graham, 560 U. S., at ___ (slip op., at 21) (quoting Roper, 543 U. S., at 571). Similarly, incapacitation could not support the life-without-parole sentence in Graham: Deciding that a “juvenile offender forever will be a danger to society” would require “mak[ing] a judgment that [he] is incorrigible”—but “ ‘incorrigibility is inconsistent with youth.’ ” 560 U. S., at ___ (slip op., at 22) (quoting Workman v. Commonwealth, 429 S.W.2d 374, 378 (Ky. App. 1968)). And for the same reason, rehabilitation could not justify that sentence. Life without parole “forswears altogether the rehabilitative ideal.” Graham, 560 U. S., at ___ (slip op., at 23). It reflects “an irrevocable judgment about [an offender’s] value and place in society,” at odds with a child’s capacity for change. Ibid. Graham concluded from this analysis that life-without-parole sentences, like capital punishment, may violate the Eighth Amendment when imposed on children. To be sure, Graham’s flat ban on life without parole applied only to nonhomicide crimes, and the Court took care to distinguish those offenses from murder, based on both moral culpability and consequential harm. See id., at ___ (slip op., at 18). But none of what it said about children—about their distinctive (and transitory) mental traits and en- vironmental vulnerabilities—is crime-specific. Those features are evident in the same way, and to the same de- gree, when (as in both cases here) a botched robbery turns into a killing. So Graham’s reasoning implicates any life-without-parole sentence imposed on a juvenile, even as its categorical bar relates only to nonhomicide offenses. Most fundamentally, Graham insists that youth matters in determining the appropriateness of a lifetime of incarceration without the possibility of parole. In the circumstances there, juvenile status precluded a life-without-parole sentence, even though an adult could receive it for a similar crime. And in other contexts as well, the characteristics of youth, and the way they weaken rationales for punishment, can render a life-without-parole sentence disproportionate. Cf. id., at ___ (slip op., at 20–23) (generally doubting the penological justifications for imposing life without parole on juveniles). “An offender’s age,” we made clear in Graham, “is relevant to the Eighth Amendment,” and so “criminal procedure laws that fail to take defendants’ youthfulness into account at all would be flawed.” Id., at ___ (slip op., at 25). The Chief Justice, concurring in the judgment, made a similar point. Al- though rejecting a categorical bar on life-without-parole sentences for juveniles, he acknowledged “Roper’s conclusion that juveniles are typically less culpable than adults,” and accordingly wrote that “an offender’s juvenile status can play a central role” in considering a sentence’s proportionality. Id., at ___ (slip op., at 5–6); see id., at ___ (slip op., at 12) (Graham’s “youth is one factor, among others, that should be considered in deciding whether his punishment was unconstitutionally excessive”).[6] But the mandatory penalty schemes at issue here prevent the sentencer from taking account of these central considerations. By removing youth from the balance— by subjecting a juvenile to the same life-without-parole sentence applicable to an adult—these laws prohibit a sentencing authority from assessing whether the law’s harshest term of imprisonment proportionately punishes a juvenile offender. That contravenes Graham’s (and also Roper’s) foundational principle: that imposition of a State’s most severe penalties on juvenile offenders cannot proceed as though they were not children. And Graham makes plain these mandatory schemes’ defects in another way: by likening life-without-parole sentences imposed on juveniles to the death penalty itself. Life-without-parole terms, the Court wrote, “share some characteristics with death sentences that are shared by no other sentences.” 560 U. S., at ___ (slip op., at 19). Imprisoning an offender until he dies alters the remainder of his life “by a forfeiture that is irrevocable.” Ibid. (citing Solem v. Helm, 463 U.S. 277, 300–301 (1983)). And this lengthiest possible incarceration is an “especially harsh punishment for a juvenile,” because he will almost inevitably serve “more years and a greater percentage of his life in prison than an adult offender.” Graham, 560 U. S., at ___ (slip op., at 19–20). The penalty when imposed on a teenager, as compared with an older person, is therefore “the same . . . in name only.” Id., at ___ (slip op., at 20). All of that suggested a distinctive set of legal rules: In part because we viewed this ultimate penalty for juveniles as akin to the death penalty, we treated it similarly to that most severe punishment. We imposed a categorical ban on the sentence’s use, in a way unprecedented for a term of imprisonment. See id., at ___ (slip op., at 9); id., at ___ (Thomas, J., dissenting) (slip op., at 7) (“For the first time in its history, the Court declares an entire class of offenders immune from a noncapital sentence using the categorical approach it previously reserved for death penalty cases alone”). And the bar we adopted mirrored a proscription first established in the death penalty context—that the punishment cannot be imposed for any nonhomicide crimes against individuals. See Kennedy, 554 U.S. 407; Coker v. Georgia, 433 U.S. 584 (1977). That correspondence—Graham’s “[t]reat[ment] [of] juvenile life sentences as analogous to capital punishment,” 560 U. S., at ___ (Roberts, C. J., concurring in judgment) (slip op., at 5)—makes relevant here a second line of our precedents, demanding individualized sentencing when imposing the death penalty. In Woodson, 428 U.S. 280, we held that a statute mandating a death sentence for first-degree murder violated the Eighth Amendment. We thought the mandatory scheme flawed because it gave no significance to “the character and record of the individual offender or the circumstances” of the offense, and “exclud[ed] from consideration . . . the possibility of compassionate or mitigating factors.” Id., at 304. Subsequent decisions have elaborated on the requirement that capital defendants have an opportunity to advance, and the judge or jury a chance to assess, any mitigating factors, so that the death penalty is reserved only for the most culpable defendants committing the most serious offenses. See, e.g., Sumner v. Shuman, 483 U.S. 66, 74–76 (1987); Eddings v. Oklahoma, 455 U.S. 104, 110–112 (1982); Lockett, 438 U. S., at 597–609 (plurality opinion). Of special pertinence here, we insisted in these rulings that a sentencer have the ability to consider the “mitigating qualities of youth.” Johnson v. Texas, 509 U.S. 350, 367 (1993). Everything we said in Roper and Graham about that stage of life also appears in these decisions. As we observed, “youth is more than a chronological fact.” Eddings, 455 U. S., at 115. It is a time of immaturity, ir- responsibility, “impetuousness[,] and recklessness.” Johnson, 509 U. S., at 368. It is a moment and “condition of life when a person may be most susceptible to influence and to psychological damage.” Eddings, 455 U. S., at 115. And its “signature qualities” are all “transient.” Johnson, 509 U. S., at 368. Eddings is especially on point. There, a 16-year-old shot a police officer point-blank and killed him. We invalidated his death sentence because the judge did not consider evidence of his neglectful and violent family background (including his mother’s drug abuse and his father’s physical abuse) and his emotional disturbance. We found that evidence “particularly relevant”—more so than it would have been in the case of an adult offender. 455 U. S., at 115. We held: “[J]ust as the chronological age of a minor is itself a relevant mitigating factor of great weight, so must the background and mental and emotional development of a youthful defendant be duly considered” in assessing his culpability. Id., at 116. In light of Graham’s reasoning, these decisions too show the flaws of imposing mandatory life-without-parole sentences on juvenile homicide offenders. Such mandatory penalties, by their nature, preclude a sentencer from taking account of an offender’s age and the wealth of characteristics and circumstances attendant to it. Under these schemes, every juvenile will receive the same sentence as every other—the 17-year-old and the 14-year-old, the shooter and the accomplice, the child from a stable household and the child from a chaotic and abusive one. And still worse, each juvenile (including these two 14-year-olds) will receive the same sentence as the vast majority of adults committing similar homicide offenses—but really, as Graham noted, a greater sentence than those adults will serve.[7] In meting out the death penalty, the elision of all these differences would be strictly forbidden. And once again, Graham indicates that a similar rule should apply when a juvenile confronts a sentence of life (and death) in prison. So Graham and Roper and our individualized sentencing cases alike teach that in imposing a State’s harshest penalties, a sentencer misses too much if he treats every child as an adult. To recap: Mandatory life without parole for a juvenile precludes consideration of his chronological age and its hallmark features—among them, immaturity, impetuosity, and failure to appreciate risks and consequences. It prevents taking into account the family and home environment that surrounds him—and from which he cannot usually extricate himself—no matter how bru- tal or dysfunctional. It neglects the circumstances of the homicide offense, including the extent of his participation in the conduct and the way familial and peer pressures may have affected him. Indeed, it ignores that he might have been charged and convicted of a lesser offense if not for incompetencies associated with youth—for example, his inability to deal with police officers or prosecutors (including on a plea agreement) or his incapacity to assist his own attorneys. See, e.g., Graham, 560 U. S., at ___ (slip op., at 27) (“[T]he features that distinguish juveniles from adults also put them at a significant disadvantage in criminal proceedings”); J. D. B. v. North Carolina, 564 U. S. ___, ___ (2011) (slip op., at 5–6) (discussing children’s responses to interrogation). And finally, this mandatory punishment disregards the possibility of rehabilitation even when the circumstances most suggest it. Both cases before us illustrate the problem. Take Jackson’s first. As noted earlier, Jackson did not fire the bullet that killed Laurie Troup; nor did the State argue that he intended her death. Jackson’s conviction was instead based on an aiding-and-abetting theory; and the appellate court affirmed the verdict only because the jury could have believed that when Jackson entered the store, he warned Troup that “[w]e ain’t playin’,” rather than told his friends that “I thought you all was playin’.” See 359 Ark., at 90–92, 194 S. W. 3d, at 759–760; supra, at 2. To be sure, Jackson learned on the way to the video store that his friend Shields was carrying a gun, but his age could well have affected his calculation of the risk that posed, as well as his willingness to walk away at that point. All these circumstances go to Jackson’s culpability for the offense. See Graham, 560 U. S., at ___ (slip op., at 18) (“[W]hen compared to an adult murderer, a juvenile offender who did not kill or intend to kill has a twice diminished moral culpability”). And so too does Jackson’s family background and immersion in violence: Both his mother and his grandmother had previously shot other individuals. See Record in No. 10–9647, pp. 80–82. At the least, a sentencer should look at such facts before depriving a 14-year-old of any prospect of release from prison. That is true also in Miller’s case. No one can doubt that he and Smith committed a vicious murder. But they did it when high on drugs and alcohol consumed with the adult victim. And if ever a pathological background might have contributed to a 14-year-old’s commission of a crime, it is here. Miller’s stepfather physically abused him; his alcoholic and drug-addicted mother neglected him; he had been in and out of foster care as a result; and he had tried to kill himself four times, the first when he should have been in kindergarten. See 928 So. 2d, at 1081 (Cobb, J., concurring in result); Miller App. 26–28; supra, at 4. Nonetheless, Miller’s past criminal history was limited—two instances of truancy and one of “second-degree criminal mischief.” No. CR–03–0915, at 6 (unpublished memorandum). That Miller deserved severe punishment for killing Cole Cannon is beyond question. But once again, a sentencer needed to examine all these circumstances before concluding that life without any possibility of parole was the appropriate penalty. We therefore hold that the Eighth Amendment forbids a sentencing scheme that mandates life in prison without possibility of parole for juvenile offenders. Cf. Graham, 560 U. S., at ___ (slip op., at 24) (“A State is not required to guarantee eventual freedom,” but must provide “some meaningful opportunity to obtain release based on demonstrated maturity and rehabilitation”). By making youth (and all that accompanies it) irrelevant to imposition of that harshest prison sentence, such a scheme poses too great a risk of disproportionate punishment. Because that holding is sufficient to decide these cases, we do not consider Jackson’s and Miller’s alternative argument that the Eighth Amendment requires a categorical bar on life without parole for juveniles, or at least for those 14 and younger. But given all we have said in Roper, Graham, and this decision about children’s diminished culpability and heightened capacity for change, we think appropriate occasions for sentencing juveniles to this harshest possible penalty will be uncommon. That is especially so because of the great difficulty we noted in Roper and Graham of distinguishing at this early age between “the juvenile of- fender whose crime reflects unfortunate yet transient immaturity, and the rare juvenile offender whose crime reflects irreparable corruption.” Roper, 543 U. S., at 573; Graham, 560 U. S., at ___ (slip op., at 17). Although we do not foreclose a sentencer’s ability to make that judgment in homicide cases, we require it to take into account how children are different, and how those differences coun- sel against irrevocably sentencing them to a lifetime in prison.[8] III Alabama and Arkansas offer two kinds of arguments against requiring individualized consideration before sen- tencing a juvenile to life imprisonment without possi- bility of parole. The States (along with the dissents) first contend that the rule we adopt conflicts with aspects of our Eighth Amendment caselaw. And they next assert that the rule is unnecessary because individualized circumstances come into play in deciding whether to try a juvenile offender as an adult. We think the States are wrong on both counts. A The States (along with Justice Thomas) first claim that Harmelin v. Michigan, 501 U.S. 957 (1991), precludes our holding. The defendant in Harmelin was sentenced to a mandatory life-without-parole term for possessing more than 650 grams of cocaine. The Court upheld that pen- alty, reasoning that “a sentence which is not otherwise cruel and unusual” does not “becom[e] so simply because it is ‘mandatory.’ ” Id., at 995. We recognized that a different rule, requiring individualized sentencing, applied in the death penalty context. But we refused to extend that command to noncapital cases “because of the qualitative difference between death and all other penalties.” Ibid.; see id., at 1006 (Kennedy, J., concurring in part and concurring in judgment). According to Alabama, invalidating the mandatory imposition of life-without-parole terms on juveniles “would effectively overrule Harmelin.” Brief for Respondent in No. 10–9646, p. 59 (hereinafter Alabama Brief); see Arkansas Brief 39. We think that argument myopic. Harmelin had nothing to do with children and did not purport to apply its holding to the sentencing of juvenile offenders. We have by now held on multiple occasions that a sentencing rule permissible for adults may not be so for children. Capital punishment, our decisions hold, generally comports with the Eighth Amendment—except it cannot be imposed on children. See Roper, 543 U.S. 551; Thompson, 487 U.S. 815. So too, life without parole is permissible for nonhomicide offenses—except, once again, for children. See Graham, 560 U. S., at ___ (slip op., at 24). Nor are these sentencing decisions an oddity in the law. To the contrary, “ ‘[o]ur history is replete with laws and judicial recogni- tion’ that children cannot be viewed simply as miniature adults.” J. D. B., 564 U. S., at ___ (slip op., at 10–11) (quoting Eddings, 455 U. S., at 115–116, citing examples from criminal, property, contract, and tort law). So if (as Harmelin recognized) “death is different,” children are different too. Indeed, it is the odd legal rule that does not have some form of exception for children. In that context, it is no surprise that the law relating to society’s harshest punishments recognizes such a distinction. Cf. Graham, 560 U. S., at ___ (Roberts, C. J., concurring in judgment) (slip op., at 7) (“Graham’s age places him in a significantly different category from the defendan[t] in . . . Harmelin”). Our ruling thus neither overrules nor undermines nor con- flicts with Harmelin. Alabama and Arkansas (along with The Chief Jus- tice and Justice Alito) next contend that because many States impose mandatory life-without-parole sentences on juveniles, we may not hold the practice unconstitutional. In considering categorical bars to the death penalty and life without parole, we ask as part of the analysis whether “ ‘objective indicia of society’s standards, as expressed in legislative enactments and state practice,’ ” show a “national consensus” against a sentence for a particular class of offenders. Graham, 560 U. S., at ___ (slip op., at 10) (quoting Roper, 543 U. S., at 563). By our count, 29 jurisdictions (28 States and the Federal Government) make a life-without-parole term mandatory for some juveniles convicted of murder in adult court.[9] The States argue that this number precludes our holding. We do not agree; indeed, we think the States’ argument on this score weaker than the one we rejected in Graham. For starters, the cases here are different from the typical one in which we have tallied legislative enactments. Our decision does not categorically bar a penalty for a class of offenders or type of crime—as, for example, we did in Roper or Graham. Instead, it mandates only that a sentencer follow a certain process—considering an offender’s youth and attendant characteristics—before imposing a particular penalty. And in so requiring, our decision flows straightforwardly from our precedents: specifically, the principle of Roper, Graham, and our individualized sentencing cases that youth matters for purposes of meting out the law’s most serious punishments. When both of those circumstances have obtained in the past, we have not scrutinized or relied in the same way on legislative enactments. See, e.g., Sumner v. Shuman, 483 U.S. 66 (relying on Woodson’s logic to prohibit the mandatory death penalty for murderers already serving life without parole); Lockett, 438 U. S., at 602–608 (plurality opinion) (applying Woodson to require that judges and juries consider all mitigating evidence); Eddings, 455 U. S., at 110–117 (similar). We see no difference here. In any event, the “objective indicia” that the States offer do not distinguish these cases from others holding that a sentencing practice violates the Eighth Amendment. In Graham, we prohibited life-without-parole terms for juveniles committing nonhomicide offenses even though 39 jurisdictions permitted that sentence. See 560 U. S., at ___ (slip op., at 11). That is 10 more than impose life without parole on juveniles on a mandatory basis.[10] And in Atkins, Roper, and Thompson, we similarly banned the death penalty in circumstances in which “less than half” of the “States that permit[ted] capital punishment (for whom the issue exist[ed])” had previously chosen to do so. Atkins, 536 U. S., at 342 (Scalia, J., dissenting) (emphasis deleted); see id., at 313–315 (majority opinion); Roper, 543 U. S., at 564–565; Thompson, 487 U. S., at 826–827 (plurality opinion). So we are breaking no new ground in these cases.[11] Graham and Thompson provide special guidance, because they considered the same kind of statutes we do and explained why simply counting them would present a distorted view. Most jurisdictions authorized the death penalty or life without parole for juveniles only through the combination of two independent statutory provisions. One allowed the transfer of certain juvenile offenders to adult court, while another (often in a far-removed part of the code) set out the penalties for any and all individuals tried there. We reasoned that in those circumstances, it was impossible to say whether a legislature had endorsed a given penalty for children (or would do so if presented with the choice). In Thompson, we found that the statutes “t[old] us that the States consider 15-year-olds to be old enough to be tried in criminal court for serious crimes (or too old to be dealt with effectively in juvenile court), but t[old] us nothing about the judgment these States have made regarding the appropriate punishment for such youthful offenders.” 487 U. S., at 826, n. 24 (plurality opinion) (emphasis deleted); see also id., at 850 (O’Connor, J., concurring in judgment); Roper, 543 U. S., at 596, n. (O’Connor, J., dissenting). And Graham echoed that reasoning: Although the confluence of state laws “ma[de] life without parole possible for some juvenile nonhomicide offenders,” it did not “justify a judgment” that many States actually “intended to subject such offenders” to those sentences. 560 U. S., at ___ (slip op., at 16).[12] All that is just as true here. Almost all jurisdictions allow some juveniles to be tried in adult court for some kinds of homicide. See Dept. of Justice, H. Snyder & M. Sickmund, Juvenile Offenders and Victims: 2006 National Report 110–114 (hereinafter 2006 National Report). But most States do not have separate penalty provisions for those juvenile offenders. Of the 29 jurisdictions mandating life without parole for children, more than half do so by virtue of generally applicable penalty provisions, imposing the sentence without regard to age.[13] And indeed, some of those States set no minimum age for who may be transferred to adult court in the first instance, thus applying life-without-parole mandates to children of any age—be it 17 or 14 or 10 or 6.[14] As in Graham, we think that “underscores that the statutory eligibility of a juvenile offender for life without parole does not indicate that the penalty has been endorsed through deliberate, express, and full legislative consideration.” 560 U. S., at ___ (slip op., at 16). That Alabama and Arkansas can count to 29 by including these possibly (or probably) inadvertent legislative outcomes does not preclude our determination that mandatory life without parole for juveniles violates the Eighth Amendment. B Nor does the presence of discretion in some jurisdictions’ transfer statutes aid the States here. Alabama and Arkansas initially ignore that many States use mandatory transfer systems: A juvenile of a certain age who has committed a specified offense will be tried in adult court, regardless of any individualized circumstances. Of the 29 relevant jurisdictions, about half place at least some juvenile homicide offenders in adult court automatically, with no apparent opportunity to seek transfer to juvenile court.[15] Moreover, several States at times lodge this decision exclusively in the hands of prosecutors, again with no statutory mechanism for judicial reevaluation.[16] And those “prosecutorial discretion laws are usually silent regarding standards, protocols, or appropriate considerations for decisionmaking.” Dept. of Justice, Office of Juvenile Justice and Delinquency Prevention, P. Griffin, S. Addie, B. Adams, & K. Firestine, Trying Juveniles as Adults: An Analysis of State Transfer Laws and Reporting 5 (2011). Even when States give transfer-stage discretion to judges, it has limited utility. First, the decisionmaker typically will have only partial information at this early, pretrial stage about either the child or the circumstances of his offense. Miller’s case provides an example. As noted earlier, see n. 3, supra, the juvenile court denied Miller’s request for his own mental-health expert at the transfer hearing, and the appeals court affirmed on the ground that Miller was not then entitled to the protections and services he would receive at trial. See No. CR–03–0915, at 3–4 (unpublished memorandum). But by then, of course, the expert’s testimony could not change the sentence; whatever she said in mitigation, the mandatory life-without-parole prison term would kick in. The key mo- ment for the exercise of discretion is the transfer—and as Miller’s case shows, the judge often does not know then what she will learn, about the offender or the offense, over the course of the proceedings. Second and still more important, the question at transfer hearings may differ dramatically from the issue at a post-trial sentencing. Because many juvenile systems require that the offender be released at a particular age or after a certain number of years, transfer decisions often present a choice between extremes: light punishment as a child or standard sentencing as an adult (here, life without parole). In many States, for example, a child convicted in juvenile court must be released from custody by the age of 21. See, e.g., Ala. Code §12–15–117(a) (Cum. Supp. 2011); see generally 2006 National Report 103 (noting limitations on the length of juvenile court sanctions). Discretionary sentencing in adult court would provide different options: There, a judge or jury could choose, rather than a life-without-parole sentence, a lifetime prison term with the possibility of parole or a lengthy term of years. It is easy to imagine a judge deciding that a minor deserves a (much) harsher sentence than he would receive in juvenile court, while still not thinking life-without-parole appropriate. For that reason, the discretion available to a judge at the transfer stage cannot substitute for discretion at post-trial sentencing in adult court—and so cannot satisfy the Eighth Amendment. IV Graham, Roper, and our individualized sentencing decisions make clear that a judge or jury must have the opportunity to consider mitigating circumstances before imposing the harshest possible penalty for juveniles. By requiring that all children convicted of homicide receive lifetime incarceration without possibility of parole, regardless of their age and age-related characteristics and the nature of their crimes, the mandatory sentencing schemes before us violate this principle of proportionality, and so the Eighth Amendment’s ban on cruel and unusual punishment. We accordingly reverse the judgments of the Arkansas Supreme Court and Alabama Court of Criminal Appeals and remand the cases for further proceedings not inconsistent with this opinion. It is so ordered. Notes 1 Jackson was ineligible for the death penalty under Thompson v. Oklahoma, 487 U.S. 815 (1988) (plurality opinion), which held that capital punishment of offenders under the age of 16 violates the Eighth Amendment. 2 For the first time in this Court, Arkansas contends that Jackson’s sentence was not mandatory. On its view, state law then in effect allowed the trial judge to suspend the life-without-parole sentence and commit Jackson to the Department of Human Services for a “training-school program,” at the end of which he could be placed on probation. Brief for Respondent in No. 10–9647, pp. 36–37 (hereinafter Arkansas Brief) (citing Ark. Code Ann. §12–28–403(b)(2) (1999)). But Arkansas never raised that objection in the state courts, and they treated Jackson’s sentence as mandatory. We abide by that interpretation of state law. See, e.g., Mullaney v. Wilbur, 421 U.S. 684, 690–691 (1975). 3 The Court of Criminal Appeals also affirmed the juvenile court’s denial of Miller’s request for funds to hire his own mental expert for the transfer hearing. The court pointed out that under governing Alabama Supreme Court precedent, “the procedural requirements of a trial do not ordinarily apply” to those hearings. E. J. M. v. State, 928 So. 2d 1077 (2004) (Cobb, J., concurring in result) (internal quotation marks omitted). In a separate opinion, Judge Cobb agreed on the reigning precedent, but urged the State Supreme Court to revisit the question in light of transfer hearings’ importance. See id., at 1081 (“[A]lthough later mental evaluation as an adult affords some semblance of procedural due process, it is, in effect, too little, too late”). 4 The three dissenting opinions here each take issue with some or all of those precedents. See post, at 5–6 (opinion of Roberts, C. J.); post, at 1–6 (opinion of Thomas, J.); post, at 1–4 (opinion of Alito, J.). That is not surprising: their authors (and joiner) each dissented from some or all of those precedents. See, e.g., Kennedy, 554 U. S., at 447 (Alito, J., joined by Roberts, C. J., and Scalia and Thomas, JJ., dissenting); Roper, 543 U. S., at 607 (Scalia, J., joined by Thomas, J., dissenting); Atkins, 536 U. S., at 337 (Scalia, J., joined by Thomas, J., dissent-ing); Thompson, 487 U. S., at 859 ((Scalia, J., dissenting); Graham v. Collins, 506 U.S. 461, 487 (1993) (Thomas, J., concurring) (contending that Woodson was wrongly decided). In particular, each disagreed with the majority’s reasoning in Graham, which is the foundation stone of our analysis. See Graham, 560 U. S., at ___ (Roberts, C. J., concurring in judgment) (slip op., at 1); id., at ___ (Thomas, J., joined by Scalia and Alito, JJ., dissenting) (slip op., at 1–25); id., at ___ (Alito, J., dissenting) (slip op., at 1). While the dissents seek to relitigate old Eighth Amendment battles, repeating many arguments this Court has previously (and often) rejected, we apply the logic of Roper, Graham, and our individualized sentencing decisions to these two cases. 5 The evidence presented to us in these cases indicates that the science and social science supporting Roper’s and Graham’s conclusions have become even stronger. See, e.g., Brief for American Psychologi-cal Association et al. as Amici Curiae 3 (“[A]n ever-growing body of research in developmental psychology and neuroscience continues to confirm and strengthen the Court’s conclusions”); id., at 4 (“It is increasingly clear that adolescent brains are not yet fully mature in regions and systems related to higher-order executive functions such as impulse control, planning ahead, and risk avoidance”); Brief for J. Lawrence Aber et al. as Amici Curiae 12–28 (discussing post-Graham studies); id., at 26–27 (“Numerous studies post-Graham indicate that exposure to deviant peers leads to increased deviant behavior and is a consistent predictor of adolescent delinquency” (footnote omitted)). 6 In discussing Graham, the dissents essentially ignore all of this reasoning. See post, at 3–6 (opinion of Roberts, C. J.); post, at 4 (opinion of Alito, J.). Indeed, The Chief Justice ignores the points made in his own concurring opinion. The only part of Graham that the dissents see fit to note is the distinction it drew between homicide and nonhomicide offenses. See post, at 7–8 (opinion of Roberts, C. J.); post, at 4 (opinion of Alito, J.). But contrary to the dissents’ charge, our decision today retains that distinction: Graham established one rule (a flat ban) for nonhomicide offenses, while we set out a different one (individualized sentencing) for homicide offenses. 7 Although adults are subject as well to the death penalty in many jurisdictions, very few offenders actually receive that sentence. See, e.g., Dept. of Justice, Bureau of Justice Statistics, S. Rosenmerkel, M. Durose, & D. Farole, Felony Sentences in State Courts 2006—Statistical Tables, p. 28 (Table 4.4) (rev. Nov. 22, 2010). So in practice, the sentencing schemes at issue here result in juvenile homicide offenders receiving the same nominal punishment as almost all adults, even though the two classes differ significantly in moral culpability and capacity for change. 8 Given our holding, and the dissents’ competing position, we see a certain irony in their repeated references to 17-year-olds who have committed the “most heinous” offenses, and their comparison of those defendants to the 14-year-olds here. See post, at 2 (opinion of Roberts, C. J.) (noting the “17-year old [who] is convicted of deliberately murdering an innocent victim”); post, at 3 (“the most heinous murders”); post, at 7 (“the worst types of murder”); post, at 5 (opinion of Alito, J.) (warning the reader not to be “confused by the particulars” of these two cases); post, at 1 (discussing the “171∕2-year-old who sets off a bomb ina crowded mall”). Our holding requires factfinders to attend to exactly such circumstances—to take into account the differences among defendants and crimes. By contrast, the sentencing schemes that the dissents find permissible altogether preclude considering these factors. 9 The States note that 26 States and the Federal Government make life without parole the mandatory (or mandatory minimum) punishment for some form of murder, and would apply the relevant provision to 14-year-olds (with many applying it to even younger defendants). See Alabama Brief 17–18. In addition, life without parole is mandatory for older juveniles in Louisiana (age 15 and up) and Texas (age 17). See La. Child. Code Ann., Arts. 857(A), (B) (West Supp. 2012); La. Rev. Stat. Ann. §§14:30(C), 14:30.1(B) (West Supp. 2012); Tex. Family Code Ann. §§51.02(2)(A), 54.02(a)(2)(A) (West Supp. 2011); Tex. Penal Code Ann. §12.31(a) (West 2011). In many of these jurisdictions, life without parole is the mandatory punishment only for aggravated forms of murder. That distinction makes no difference to our analysis. We have consistently held that limiting a mandatory death penalty law to particular kinds of murder cannot cure the law’s “constitutional vice” of disregarding the “circumstances of the particular offense and the character and propensities of the offender.” Roberts v. Louisiana, 428 U.S. 325, 333 (1976) (plurality opinion); see Sumner v. Shuman, 483 U.S. 66 (1987). The same analysis applies here, for the same reasons. 10 In assessing indicia of societal standards, Graham discussed “ac-tual sentencing practices” in addition to legislative enactments, noting how infrequently sentencers imposed the statutorily available penalty. 560 U. S., at ___ (slip op., at 11). Here, we consider the constitutional-ity of mandatory sentencing schemes—which by definition remove a judge’s or jury’s discretion—so no comparable gap between legislation and practice can exist. Rather than showing whether sentencers consider life without parole for juvenile homicide offenders appropriate, the number of juveniles serving this sentence, see post, at 1, 3–4 (Roberts, C. J., dissenting), merely reflects the number who have com-mitted homicide in mandatory-sentencing jurisdictions. For the same reason, The Chief Justice’s comparison of ratios in this case and Gra-ham carries little weight. He contrasts the number of mandatorylife-without-parole sentences for juvenile murderers, relative to the number of juveniles arrested for murder, with “the corresponding number” of sentences in Graham (i.e., the number of life-without-parole sentences for juveniles who committed serious nonhomicide crimes, as compared to arrests for those crimes). Post, at 4. But because the mandatory nature of the sentences here necessarily makes them more common, The Chief Justice’s figures do not “correspon[d]” at all. The higher ratio is mostly a function of removing the sentencer’s discretion. Where mandatory sentencing does not itself account for the number of juveniles serving life-without-parole terms, the evidence we have of practice supports our holding. Fifteen jurisdictions make life without parole discretionary for juveniles. See Alabama Brief 25 (listing 12 States); Cal. Penal Code Ann. §190.5(b) (West 2008); Ind. Code §35–50–2–3(b) (2011); N. M. Stat. §§31–18–13(B), 31–18–14, 31–18–15.2 (2010). According to available data, only about 15% of all juvenile life-without-parole sentences come from those 15 jurisdictions, while 85% come from the 29 mandatory ones. See Tr. of Oral Arg. in No. 10–9646, p. 19; Human Rights Watch, State Distribution of Youth Offenders Serv-ing Juvenile Life Without Parole (JLWOP), dateOct. 2, 2009, online athttp: // www.hrw.org / news/2009 / 10/02 / state-distribution-juvenile-offenders-serving-juvenile-life-without-parole (as visited dateJune 21, 2012, and available in Clerk of Court’s case file). That figure indicates that when given the choice, sentencers impose life without parole on children relatively rarely. And contrary to The Chief Justice’s argument, see post, at 5, n. 2, we have held that when judges and juries do not often choose to impose a sentence, it at least should not be mandatory. See Woodson v. North Carolina, 428 U.S. 280, 295–296 (1976) (plurality opinion) (relying on the infrequency with which juries imposed the death penalty when given discretion to hold that its mandatory imposition violates the Eighth Amendment). 11 In response, The Chief Justice complains: “To say that a sentence may be considered unusual because so many legislatures approve it stands precedent on its head.” Post, at 5. To be clear: That description in no way resembles our opinion. We hold that the sentence violates the Eighth Amendment because, as we have exhaustively shown, it conflicts with the fundamental principles of Roper, Graham, and our individualized sentencing cases. We then show why the number of States imposing this punishment does not preclude our holding, and note how its mandatory nature (in however many States adopt it) makes use of actual sentencing numbers unilluminating. 12 The Chief Justice attempts to distinguish Graham on this point, arguing that there “the extreme rarity with which the sentence in question was imposed could suggest that legislatures did not really intend the inevitable result of the laws they passed.” Post, at 6. But neither Graham nor Thompson suggested such reasoning, presumably because the time frame makes it difficult to comprehend. Those cases considered what legislators intended when they enacted, at different moments, separate juvenile-transfer and life-without-parole provisions—by definition, before they knew or could know how many juvenile life-without-parole sentences would result. 13 See Ala. Code §§13A–5–45(f), 13A–6–2(c) (2005 and Cum. Supp. 2011); Ariz. Rev. Stat. Ann. §13–752 (West 2010), §41–1604.09(I) (West 2011); Conn. Gen. Stat. §53a–35a(1) (2011); Del. Code Ann., Tit. 11, §4209(a) (2007); Fla. Stat. §775.082(1) (2010); Haw. Rev. Stat. §706–656(1) (1993); Idaho Code §18–4004 (Lexis 2004); Mich. Comp. Laws Ann. §791.234(6)(a) (West Cum. Supp. 2012); Minn. Stat. Ann. §§609.106, subd. 2 (West 2009); Neb. Rev. Stat. §29–2522 (2008); N. H. Rev. Stat. Ann. §630:1–a (West 2007); 18 Pa. Cons. Stat. §§1102(a), (b), 61 Pa. Cons. Stat. §6137(a)(1) (Supp. 2012); S. D. Codified Laws §22-6-1(1) (2006), §24–15–4 (2004); Vt. Stat. Ann., Tit. 13, §2311(c)(2009); Wash. Rev. Code §10.95.030(1) (2010). 14 See Del. Code Ann., Tit. 10, §1010 (1999 and Cum. Supp. 2010), Tit. 11, §4209(a) (2007); Fla. Stat. §985.56 (2010), 775.082(1); Haw. Rev. Stat. §571–22(d) (1993), §706–656(1); Idaho Code §§20–508, 20–509 (Lexis Cum. Supp. 2012), §18–4004; Mich. Comp. Laws Ann. §712A.2d (West 2009), §791.234(6)(a); Neb. Rev. Stat. §§43–247, 29–2522 (2008); 42 Pa. Cons. Stat. §6355(e) (2000), 18 Pa. Cons. Stat. §1102. Other States set ages between 8 and 10 as the minimum for transfer, thus exposing those young children to mandatory life without parole. See S. D. Codified Laws §§26–8C–2, 26–11–4 (2004), §22–6–1 (age 10); Vt. Stat. Ann., Tit. 33, §5204 (2011 Cum. Supp.), Tit. 13, §2311(a) (2009) (age 10); Wash. Rev. Code §§9 A. 04.050, 13.40.110 (2010), §10.95.030 (age 8). 15 See Ala. Code §12–15–204(a) (Cum. Supp. 2011); Ariz. Rev. Stat. Ann. §13–501(A) (West Cum. Supp. 2011); Conn. Gen. Stat. §46b–127 (2011); Ill. Comp. Stat. ch. 705, §§405/5–130(1)(a), (4)(a) (West 2010); La. Child. Code Ann., Art. 305(A) (West Cum. Supp. 2012); Mass. Gen. Laws, ch. 119, §74 (West 2010); Mich. Comp. Laws Ann. §712A.2(a) (West 2002); Minn. Stat. Ann. §260B.007, subd. 6(b) (West Cum. Supp. 2011), §260B.101, subd. 2 (West 2007); Mo. Rev. Stat. §§211.021(1), (2) (2011); N. C. Gen. Stat. Ann. §§7B–1501(7), 7B–1601(a), 7B–2200 (Lexis 2011); N. H. Rev. Stat. Ann. §169–B:2(IV) (West Cum. Supp. 2011), §169–B:3 (West 2010); Ohio Rev. Code Ann. §2152.12(A)(1)(a) (Lexis 2011); Tex. Family Code Ann. §51.02(2); Va. Code Ann. §§16.1–241(A), 16.1–269.1(B), (D) (Lexis 2010). 16 Fla. Stat. Ann. §985.557(1) (West Supp. 2012); Mich. Comp. Laws Ann. §712A.2(a)(1); Va. Code Ann. §§16.1–241(A), 16.1–269.1(C), (D).

565.US.2011_10-1195

Consumer complaints about abuses of telephone technology—for example, computerized calls to private homes—prompted Congress to pass the Telephone Consumer Protection Act of 1991 (TCPA or Act), 47 U. S. C. §227. Congress determined that federal legislation was needed because telemarketers, by operating interstate, were escaping state-law prohibitions on intrusive nuisance calls. The Act bans certain invasive telemarketing practices and directs the Federal Communications Commission (FCC) to prescribe implementing regulations. It authorizes States to bring civil actions to enjoin prohibited practices and recover damages on their residents’ behalf, 47 U. S. C. A. §227(g)(1) (Supp. 2011), and provides that jurisdiction over these state-initiated suits lies exclusively in the U. S. district courts, §227(g)(2). It also permits a private person to seek redress for violations of the Act or regulations “in an appropriate court of [a] State,” “if [such an action is] otherwise permitted by the laws or rules of court of [that] State.” 47 U. S. C. §§227(b)(3), (c)(5). Petitioner Mims filed a damages action in Federal District Court, alleging that respondent Arrow, seeking to collect a debt, violated the TCPA by repeatedly using an automatic telephone dialing system or prerecorded or artificial voice to call Mims’s cellular phone without his consent. Mims invoked the court’s “federal question” jurisdiction, i.e., its authority to adjudicate claims “arising under the . . . laws . . . of the United States,” 28 U. S. C. §1331. The District Court, affirmed by the Eleventh Circuit, dismissed Mims’s complaint for want of subject-matter jurisdiction, concluding that the TCPA had vested jurisdiction over private actions exclusively in state courts. Held: The TCPA’s permissive grant of jurisdiction to state courts does not deprive the U. S. district courts of federal-question jurisdiction over private TCPA suits. Pp. 7–18. (a) Because federal law creates the right of action and provides the rules of decision, Mims’s TCPA claim, in §1331’s words, plainly “aris[es] under” the “laws . . . of the United States.” Arrow agrees that this action arises under federal law, but urges that Congress vested exclusive adjudicatory authority over private TCPA actions in state courts. In cases “arising under” federal law, there is a presumption of concurrent state-court jurisdiction, rebuttable if “Congress affirmatively ousts the state courts of jurisdiction over a particular federal claim.” Tafflin v. Levitt, 493 U.S. 455, 458–459. Arrow acknowledges the presumption, but maintains that §1331 creates no converse presumption in favor of federal-court jurisdiction. Instead, Arrow urges, the TCPA, a later, more specific statute, displaces §1331, an earlier, more general prescription. Section 1331 is not swept away so easily. The principle that district courts possess federal-question jurisdiction under §1331 when federal law creates a private right of action and furnishes the substantive rules of decision endures unless Congress divests federal courts of their §1331 adjudicatory authority. See, e.g., Verizon Md. Inc. v. Public Serv. Comm’n of Md., 535 U.S. 635, 642. Accordingly, the District Court retains §1331 jurisdiction over Mims’s complaint unless the TCPA, expressly or by fair implication, excludes federal-court adjudication. See id., at 644. Pp. 7–10. (b) Arrow’s arguments do not persuade this Court that Congress eliminated §1331 jurisdiction over private TCPA actions. Title 47 U. S. C. §227(b)(3)’s language may be state-court oriented, but “the grant of jurisdiction to one court does not, of itself, imply that the jurisdiction is to be exclusive,” United States v. Bank of New York & Trust Co., 296 U.S. 463, 479. Nothing in §227(b)(3)’s permissive language makes state-court jurisdiction exclusive, or otherwise purports to oust federal courts of their §1331 jurisdiction. The provision does not state that a private plaintiff may bring a TCPA action “only” or “exclusively” in state court. In contrast, 47 U. S. C. A. §227(g)(2) (Supp. 2011) vests “exclusive jurisdiction” over state-initiated TCPA suits in the federal courts. Section 227(g)(2)’s exclusivity prescription “reinforce[s] the conclusion that [ 47 U. S. C. §227(b)(3)’s] silence . . . leaves the jurisdictional grant of §1331 untouched. For where otherwise applicable jurisdiction was meant to be excluded, it was excluded expressly.” Verizon Md., 535 U. S., at 644. Arrow argues that Congress had no reason to provide for a private action “in an appropriate [state] court,” §227(b)(3), if it did not mean to make the state forum exclusive, for state courts would have concurrent jurisdiction even if Congress had said nothing at all. But, as already noted, Congress had simultaneously made federal-court jurisdiction exclusive in TCPA enforcement actions brought by state authorities, see 47 U. S. C. A. §227(g)(2) (Supp. 2011), and may simply have wanted to avoid any argument that federal jurisdiction was also exclusive for private actions. Moreover, by providing that private actions may be brought in state court “if otherwise permitted by the laws or rules of court of [the] State,” 47 U. S. C. §227(b)(3), Congress arguably gave States leeway they would otherwise lack to decide whether to entertain TCPA claims. Arrow further asserts that making state-court jurisdiction over §227(b)(3) claims exclusive serves Congress’ objective of enabling States to control telemarketers whose interstate operations evaded state law. Even so, jurisdiction conferred by 28 U. S. C. §1331 should hold firm against “mere implication flowing from subsequent legislation.” Colorado River Water Conservation Dist. v. United States, 424 U.S. 800, 808, 809, n. 15. Furthermore, had Congress sought only to fill a gap in the States’ enforcement capabilities, it could have provided that out-of-state telemarketing calls directed into a State would be subject to the receiving State’s laws. Instead, Congress enacted detailed, uniform, federal substantive prescriptions and provided for a regulatory regime administered by a federal agency. Arrow’s reliance on a statement by Senator Hollings, the TCPA’s sponsor, is misplaced. The remarks nowhere mention federal-court jurisdiction or otherwise suggest that 47 U. S. C. §227(b)(3) is intended to divest federal courts of authority over TCPA claims. Even if Hollings and other TCPA supporters expected private actions to proceed solely in state courts, their expectation would not control this Court’s judgment on §1331’s compass. Arrow’s arguments that federal courts will be inundated by $500-per-violation TCPA claims or that defendants could use federal-court removal to force small-claims-court plaintiffs to abandon suit seem more imaginary than real. Pp. 10–18. 421 Fed. Appx. 920, reversed and remanded. Ginsburg, J., delivered the opinion for a unanimous Court.

This case concerns enforcement, through private suits, of the Telephone Consumer Protection Act of 1991 (TCPA or Act), 47 U. S. C. §227. Voluminous consumer complaints about abuses of telephone technology—for example, computerized calls dispatched to private homes—prompted Congress to pass the TCPA. Congress de- termined that federal legislation was needed because telemarketers, by operating interstate, were escaping state-law prohibitions on intrusive nuisance calls. The Act bans certain practices invasive of privacy and directs the Federal Communications Commission (FCC or Commission) to prescribe implementing regulations. It authorizes States to bring civil actions to enjoin prohibited practices and to recover damages on their residents’ behalf. The Commission must be notified of such suits and may intervene in them. Jurisdiction over state-initiated TCPA suits, Congress provided, lies exclusively in the U. S. district courts. Congress also provided for civil actions by private parties seeking redress for violations of the TCPA or of the Commission’s implementing regulations. Petitioner Marcus D. Mims, complaining of multiple violations of the Act by respondent Arrow Financial Services, LLC (Arrow), a debt-collection agency, commenced an action for damages against Arrow in the U. S. District Court for the Southern District of Florida. Mims invoked the court’s “federal question” jurisdiction, i.e., its authority to adjudicate claims “arising under the . . . laws . . . of the United States,” 28 U. S. C. §1331. The District Court, affirmed by the U. S. Court of Appeals for the Eleventh Circuit, dismissed Mims’s complaint for want of subject-matter jurisdiction. Both courts relied on Congress’ specification, in the TCPA, that a private person may seek redress for violations of the Act (or of the Commission’s regulations thereunder) “in an appropriate court of [a] State,” “if [such an action is] otherwise permitted by the laws or rules of court of [that] State.” 47 U. S. C. §§227(b)(3), (c)(5). The question presented is whether Congress’ provision for private actions to enforce the TCPA renders state courts the exclusive arbiters of such actions. We have long recognized that “[a] suit arises under the law that creates the cause of action.” American Well Works Co. v. Layne & Bowler Co., 241 U. S. 257, 260 (1916) . Beyond doubt, the TCPA is a federal law that both creates the claim Mims has brought and supplies the substantive rules that will govern the case. We find no convincing reason to read into the TCPA’s permissive grant of jurisdiction to state courts any barrier to the U. S. district courts’ exercise of the general federal-question jurisdiction they have possessed since 1875. See Act of Mar. 3, 1875, §1, 18Stat. 470; 13D C. Wright, A. Miller, E. Cooper, & R. Freer, Federal Practice and Procedure §3561, p. 163 (3d ed. 2008) (hereinafter Wright & Miller). We hold, therefore, that federal and state courts have concurrent jurisdiction over private suits arising under the TCPA. I A In enacting the TCPA, Congress made several findings relevant here. “Unrestricted telemarketing,” Congress determined, “can be an intrusive invasion of privacy.” TCPA, 105 Stat. 2394, note following 47 U. S. C. §227 (Congressional Findings) (internal quotation marks omitted). In particular, Congress reported, “[m]any consumers are outraged over the proliferation of intrusive, nuisance [telemarketing] calls to their homes.” Ibid. (internal quotation marks omitted). “[A]utomated or prerecorded telephone calls” made to private residences, Congress found, were rightly regarded by recipients as “an invasion of privacy.” Ibid. (internal quotation marks omitted). Although over half the States had enacted statutes restricting telemarketing, Congress believed that federal law was needed because “telemarketers [could] evade [state-law] prohibitions through interstate operations.” Ibid. (internal quotation marks omitted). See also S. Rep. No. 102–178, p. 3 (1991) (“[B]ecause States do not have jurisdiction over interstate calls[,] [m]any States have expressed a desire for Federal legislation . . . .”). [ 1 ] Subject to exceptions not pertinent here, the TCPA prin- cipally outlaws four practices. First, the Act makes it unlawful to use an automatic telephone dialing system or an artificial or prerecorded voice message, without the prior express consent of the called party, to call any emergency telephone line, hospital patient, pager, cellular telephone, or other service for which the receiver is charged for the call. See 47 U. S. C. §227(b)(1)(A). Sec-ond, the TCPA forbids using artificial or prerecorded voice messages to call residential telephone lines with- out prior express consent. §227(b)(1)(B). Third, the Act proscribes sending unsolicited advertisements to fax machines. §227(b)(1)(C). Fourth, it bans using automatic telephone dialing systems to engage two or more of a business’ telephone lines simultaneously. §227(b)(1)(D). [ 2 ] The TCPA delegates authority to the FCC to ban ar- tificial and prerecorded voice calls to businesses, §227(b)(2)(A), and to exempt particular types of calls from the law’s requirements, §§227(b)(2)(B), (C). The Act also directs the FCC to prescribe regulations to protect the privacy of residential telephone subscribers, possibly through the creation of a national “do not call” system. §227(c). [ 3 ] Congress provided complementary means of enforcing the Act. State Attorneys General may “bring a civil action on behalf of [State] residents,” if the Attorney General “has reason to believe that any person has engaged . . . in a pattern or practice” of violating the TCPA or FCC regulations thereunder. 47 U. S. C. A. §227(g)(1) (Supp. 2011). “The district courts of the United States . . . have exclusive jurisdiction” over all TCPA actions brought by State Attorneys General. §227(g)(2). The Commission may intervene in such suits. §227(g)(3). [ 4 ] Title 47 U. S. C. §227(b)(3), captioned “Private right of action,” provides: “A person or entity may, if otherwise permitted by the laws or rules of court of a State, bring in an appropriate court of that State— “(A) an action based on a violation of this subsection or the regulations prescribed under this subsection to enjoin such violation, “(B) an action to recover for actual monetary loss from such a violation, or to receive $500 in damages for each such violation, whichever is greater, or “(C) both such actions. “If the court finds that the defendant willfully or knowingly violated this subsection or the regulations prescribed under this subsection, the court may, in its discretion, increase the amount of the award to an amount equal to not more than 3 times the amount available under subparagraph (B) of this paragraph.” A similar provision authorizes a private right of action for a violation of the FCC’s implementing regulations. [ 5 ] B Mims, a Florida resident, alleged that Arrow, seeking to collect a debt, repeatedly used an automatic telephone dialing system or prerecorded or artificial voice to call Mims’s cellular phone without his consent. Commencing suit in the U. S. District Court for the Southern District of Florida, Mims charged that Arrow “willfully or knowingly violated the TCPA.” App. 14. He sought declaratory relief, a permanent injunction, and damages. Id., at 18–19. The District Court held that it lacked subject-matter jurisdiction over Mims’s TCPA claim. Under Eleventh Circuit precedent, the District Court explained, federal-question jurisdiction under 28 U. S. C. §1331 was unavailable “because Congress vested jurisdiction over [private actions under] the TCPA exclusively in state courts.” Civ. No. 09–22347 (SD Fla., Apr. 1, 2010), App. to Pet. for Cert. 4a–5a (citing Nicholson v. Hooters of Augusta, Inc., 136 F. 3d 1287 (CA11 1998)). Adhering to Circuit precedent, the U. S. Court of Appeals for the Eleventh Circuit affirmed. 421 Fed. Appx. 920, 921 (2011) (quoting Nicholson, 136 F. 3d, at 1287–1288 (“Congress granted state courts exclusive jurisdiction over private actions under the [TCPA].”)). We granted certiorari, 564 U. S. ___ (2011), to resolve a split among the Circuits as to whether Congress granted state courts exclusive jurisdiction over private actions brought under the TCPA. Compare Murphey v. Lanier, 204 F. 3d 911, 915 (CA9 2000) (U. S. district courts lack federal-question jurisdiction over private TCPA actions), ErieNet, Inc. v. Velocity Net, Inc., 156 F. 3d 513, 519 (CA3 1998) (same), Foxhall Realty Law Offices, Inc. v. Telecommunications Premium Servs., Ltd., 156 F. 3d 432, 434 (CA2 1998) (same), Nicholson, 136 F. 3d, at 1287–1288, Chair King, Inc. v. Houston Cellular Corp., 131 F. 3d 507, 514 (CA5 1997) (same), and International Science & Technology Inst. v. Inacom Communications, Inc., 106 F. 3d 1146, 1158 (CA4 1997) (same), with Charvat v. EchoStar Satellite, LLC, 630 F. 3d 459, 463–465 (CA6 2010) (U. S. district courts have federal-question jurisdiction over private TCPA actions), Brill v. Countrywide Home Loans, Inc., 427 F. 3d 446, 447 (CA7 2005) (same), and ErieNet, 156 F. 3d, at 521 (Alito, J., dissenting) (same). We now hold that Congress did not deprive federal courts of federal-question jurisdiction over private TCPA suits. II Federal courts, though “courts of limited jurisdiction,” Kokkonen v. Guardian Life Ins. Co. of America, 511 U. S. 375, 377 (1994) , in the main “have no more right to decline the exercise of jurisdiction which is given, then to usurp that which is not given.” Cohens v. Virginia, 6 Wheat. 264 (1821). Congress granted federal courts general federal-question jurisdiction in 1875. See Act of Mar. 3, 1875, §1, 18Stat. 470. [ 6 ] As now codified, the law provides: “The district courts shall have original jurisdiction of all civil actions arising under the Constitution, laws, or treaties of the United States.” 28 U. S. C. §1331. The statute originally included an amount-in-controversy requirement, set at $500. See Act of Mar. 3, 1875, §1, 18Stat. 470. Recognizing the responsibility of federal courts to decide claims, large or small, arising under federal law, Congress in 1980 eliminated the amount-in-controversy requirement in federal-question (but not diversity) cases. See Federal Question Jurisdictional Amendments Act of 1980, 94Stat. 2369 (amending 28 U. S. C. §1331). See also H. R. Rep. No. 96–1461, p. 1 (1980). [ 7 ] Apart from deletion of the amount-in-controversy requirement, the general federal-question provision has remained essentially unchanged since 1875. See 13D Wright & Miller 163. Because federal law creates the right of action and provides the rules of decision, Mims’s TCPA claim, in 28 U. S. C. §1331’s words, plainly “aris[es] under” the “laws . . . of the United States.” As already noted, supra, at 2, “[a] suit arises under the law that creates the cause of action.” American Well Works, 241 U. S., at 260. Al- though courts have described this formulation as “more useful for inclusion than for . . . exclusion,” Merrell Dow Pharmaceuticals Inc. v. Thompson, 478 U. S. 804, 809, n. 5 (1986) (quoting T. B. Harms Co. v. Eliscu, 339 F. 2d 823, 827 (CA2 1964)), there is no serious debate that a federally created claim for relief is generally a “sufficient con- dition for federal-question jurisdiction.” Grable & Sons Metal Products, Inc. v. Darue Engineering & Mfg., 545 U. S. 308, 317 (2005) . [ 8 ] Arrow agrees that this action arises under federal law, see Tr. of Oral Arg. 27, but urges that Congress vested exclusive adjudicatory authority over private TCPA actions in state courts. In cases “arising under” federal law, we note, there is a “deeply rooted presumption in favor of concurrent state court jurisdiction,” rebuttable if “Congress affirmatively ousts the state courts of jurisdiction over a particular federal claim.” Tafflin v. Levitt, 493 U. S. 455 –459 (1990). E.g., 28 U. S. C. §1333 (“The district courts shall have original jurisdiction, exclusive of the courts of the States, of: (1) Any civil case of admiralty or maritime jurisdiction . . . .”). The presumption of concurrent state-court jurisdiction, we have recognized, can be overcome “by an explicit statutory directive, by unmistakable implication from legislative history, or by a clear incompatibility between state-court jurisdiction and federal interests.” Gulf Offshore Co. v. Mobil Oil Corp., 453 U. S. 473, 478 (1981) . Arrow readily acknowledges the presumption of con- current state-court jurisdiction, but maintains that 28 U. S. C. §1331 creates no converse presumption in favor of federal-court jurisdiction. Instead, Arrow urges, the TCPA, a later, more specific statute, displaces §1331, an earlier, more general prescription. See Tr. of Oral Arg. 28–29; Brief for Respondent 31. Section 1331, our decisions indicate, is not swept away so easily. As stated earlier, see supra, at 8, when federal law creates a private right of action and furnishes the substantive rules of decision, the claim arises under federal law, and district courts possess federal-question jurisdiction under §1331. [ 9 ] That principle endures unless Congress divests federal courts of their §1331 adjudica- tory authority. See, e.g., Verizon Md. Inc. v. Public Serv. Comm’n of Md., 535 U. S. 635, 642 (2002) (Nothing in 47 U. S. C. §252(e)(6) “divest[s] the district courts of their authority under 28 U. S. C. §1331 to review the [state agency’s] order for compliance with federal law.”); K mart Corp. v. Cartier, Inc., 485 U. S. 176 –183 (1988) (“The District Court would be divested of [§1331] jurisdiction . . . if this action fell within one of several specific grants of exclusive jurisdiction to the Court of International Trade [under 28 U. S. C. §1581(a) or §1581(i)(3)].”). “[D]ivestment of district court jurisdiction” should be found no more readily than “divestmen[t] of state court jurisdiction,” given “the longstanding and explicit grant of federal question jurisdiction in 28 U. S. C. §1331.” ErieNet, 156 F. 3d, at 523 (Alito, J., dissenting); see Gonell, Note, Statutory Interpretation of Federal Jurisdictional Statutes: Jurisdiction of the Private Right of Action under the TCPA, 66 Ford. L. Rev. 1895, 1929–1930 (1998). Accordingly, the District Court retains §1331 jurisdiction over Mims’s complaint unless the TCPA, expressly or by fair implication, excludes federal-court adjudication. See Verizon Md., 535 U. S., at 644; Gonell, supra, at 1929 (Jurisdiction over private TCPA actions “is proper under §1331 unless Congress enacted a partial repeal of §1331 in the TCPA.”). III Arrow’s arguments do not persuade us that Congress has eliminated §1331 jurisdiction over private actions under the TCPA. The language of the TCPA—“A person or entity may, if otherwise permitted by the laws or rules of court of a State, bring [an action] in an appropriate court of that State,” 47 U. S. C. §227(b)(3)—Arrow asserts, is uniquely state-court oriented. See Brief for Respondent 13. That may be, but “[i]t is a general rule that the grant of jurisdiction to one court does not, of itself, imply that the jurisdiction is to be exclusive.” United States v. Bank of New York & Trust Co., 296 U. S. 463, 479 (1936) . Nothing in the permissive language of §227(b)(3) makes state-court jurisdiction exclusive, or otherwise purports to oust federal courts of their 28 U. S. C. §1331 jurisdiction over federal claims. See, e.g., Verizon Md., 535 U. S., at 643 (“[N]othing in 47 U. S. C. §252(e)(6) purports to strip [§1331] jurisdiction.”). Cf. Yellow Freight System, Inc. v. Donnelly, 494 U. S. 820, 823 (1990) (Title VII’s language—“[e]ach United States district court . . . shall have jurisdiction of actions brought under this subchapter,” 42 U. S. C. §2000e–5(f)(3)—does not “ous[t] state courts of their presumptive jurisdiction.” (internal quotation marks omitted)). Congress may indeed provide a track for a federal claim exclusive of §1331. See, e.g., 42 U. S. C. §405(h) (“No action . . . shall be brought under [§1331] to recover on any claim arising under [Title II of the Social Security Act].”); Weinberger v. Salfi, 422 U. S. 749 –757 (1975). Congress has done nothing of that sort here, however. Title 47 U. S. C. §227(b)(3) does not state that a private plaintiff may bring an action under the TCPA “only” in state court, or “exclusively” in state court. The absence of such a statement contrasts with the Act’s instruction on suits instituted by State Attorneys General. As earlier noted, see supra, at 4, 47 U. S. C. A. §227(g)(2) (Supp. 2011) vests “exclusive jurisdiction over [such] actions” in “[t]he district courts of the United States.” [ 10 ] Section 227(g)(2)’s exclusivity prescription “reinforce[s] the conclusion that [ 47 U. S. C. §227(b)(3)’s] silence . . . leaves the jurisdictional grant of §1331 untouched. For where otherwise applicable jurisdiction was meant to be excluded, it was excluded expressly.” Verizon Md., 535 U. S., at 644; see ErieNet, 156 F. 3d, at 522 (Alito, J., dissenting) (“[47 U. S. C. A. §227(g)(2) (Supp. 2011)] reveals that, while drafting the TCPA, Congress knew full well how to grant exclusive jurisdiction with mandatory language. The most natural interpretation of Congress’ failure to use similar language in [47 U. S. C. §]227(b)(3) is that Congress did not intend to grant exclusive jurisdiction in that section.”); Brill, 427 F. 3d, at 451 (“[47 U. S. C. A. §227(g)(2) (Supp. 2011)] is explicit about exclusivity, while [47 U. S. C.] §227(b)(3) is not; the natural inference is that the state forum mentioned in §227(b)(3) is optional rather than mandatory.”). [ 11 ] Arrow urges that Congress would have had no reason to provide for a private action “in an appropriate [state] court,” §227(b)(3), if it did not mean to make the state forum exclusive. Had Congress said nothing at all about bringing private TCPA claims in state courts, Arrow observes, those courts would nevertheless have concurrent jurisdiction. See supra, at 9. True enough, but Con- gress had simultaneously provided for TCPA enforcement actions by state authorities, 47 U. S. C. A. §227(g) (Supp. 2011), and had made federal district courts exclusively competent in such cases, §227(g)(2). Congress may simply have wanted to avoid any argument that in private actions, as in actions brought by State Attorneys General, “federal jurisdiction is exclusive.” Brill, 427 F. 3d, at 451 (emphasis deleted) (citing Yellow Freight, 494 U. S. 820 (holding, after 26 years of litigation, that claims under the Civil Rights Act of 1964 may be resolved in state as well as federal courts) and Tafflin, 493 U. S. 455 (holding, after 20 years of litigation, that claims under RICO may be resolved in state as well as federal courts)). Moreover, by providing that private actions may be brought in state court “if otherwise permitted by the laws or rules of court of [the] State,” 47 U. S. C. §227(b)(3), Congress arguably gave States leeway they would otherwise lack to “decide for [themselves] whether to entertain claims under the [TCPA],” Brill, 427 F. 3d, at 451. See Brief for Respondent 16 (Congress “le[ft] States free to decide what TCPA claims to allow.”). [ 12 ] Making state-court jurisdiction over §227(b)(3) claims exclusive, Arrow further asserts, “fits hand in glove with [Congress’] objective”: enabling States to control telemarketers whose interstate operations evaded state law. Id., at 15. Even so, we have observed, jurisdiction conferred by 28 U. S. C. §1331 should hold firm against “mere implication flowing from subsequent legislation.” Colorado River Water Conservation Dist. v. United States, 424 U. S. 800, 808, 809, n. 15 (1976) (quoting Rosencrans v. United States, 165 U. S. 257, 262 (1897) ). We are not persuaded, moreover, that Congress sought only to fill a gap in the States’ enforcement capabilities. Had Congress so limited its sights, it could have passed a statute providing that out-of-state telemarketing calls directed into a State would be subject to the laws of the receiving State. Congress did not enact such a law. Instead, it enacted detailed, uniform, federal substantive prescriptions and provided for a regulatory regime administered by a federal agency. See 47 U. S. C. §227. TCPA liability thus depends on violation of a federal statutory requirement or an FCC regulation, §§227(b)(3)(A), (c)(5), not on a violation of any state substantive law. The federal interest in regulating telemarketing to “protec[t] the privacy of individuals” while “permit[ting] legitimate [commercial] practices,” 105Stat. 2394, note following 47 U. S. C. §227 (Congressional Findings) (internal quotation marks omitted), is evident from the regu- latory role Congress assigned to the FCC. See, e.g., §227(b)(2) (delegating to the FCC authority to exempt calls from the Act’s reach and prohibit calls to businesses). Congress’ design would be less well served if consumers had to rely on “the laws or rules of court of a State,” §227(b)(3), or the accident of diversity jurisdiction, [ 13 ] to gain redress for TCPA violations. Arrow emphasizes a statement made on the Senate floor by Senator Hollings, the TCPA’s sponsor: “Computerized calls are the scourge of modern civilization. They wake us up in the morning; they in- terrupt our dinner at night; they force the sick and elderly out of bed; they hound us until we want to rip the telephone right out of the wall. . . . . . “The substitute bill contains a private right-of-action provision that will make it easier for consumers to recover damages from receiving these computerized calls. The provision would allow consumers to bring an action in State court against any entity that violates the bill. The bill does not, because of constitutional constraints, dictate to the States which court in each State shall be the proper venue for such an action, as this is a matter for State legislators to determine. Nevertheless, it is my hope that States will make it as easy as possible for consumers to bring such actions, preferably in small claims court. . . . “Small claims court or a similar court would allow the consumer to appear before the court without an attorney. The amount of damages in this legislation is set to be fair to both the consumer and the telemarketer. However, it would defeat the purposes of the bill if the attorneys’ costs to consumers of bringing an action were greater than the potential damages. I thus expect that the States will act reasonably in permitting their citizens to go to court to enforce this bill.” 137 Cong. Rec. 30821–30822 (1991). This statement does not bear the weight Arrow would place on it. First, the views of a single legislator, even a bill’s sponsor, are not controlling. Consumer Product Safety Comm’n v. GTE Sylvania, Inc., 447 U. S. 102, 118 (1980) . Second, Senator Hollings did not mention federal-court jurisdiction or otherwise suggest that 47 U. S. C. §227(b)(3) is intended to divest federal courts of authority to hear TCPA claims. Hollings no doubt believed that mine-run TCPA claims would be pursued most expeditiously in state small-claims court. [ 14 ] But one cannot glean from his statement any expectation that those courts, or state courts generally, would have exclusive jurisdiction over private actions alleging violations of the Act or of the FCC’s implementing regulations. Third, even if we agreed with Arrow that Senator Hollings expected private TCPA actions to proceed solely in state courts, and even if other supporters shared his view, that expectation would not control our judgment on 28 U. S. C. §1331’s compass. Cf. Yellow Freight, 494 U. S., at 826 (“persuasive showing that most legislators, judges, and administrators . . . involved in the enactment, amendment, enforcement, and interpretation of Title VII expected that such litiga- tion would be processed exclusively in federal courts” did not overcome presumption of concurrent state-court jurisdiction). Among its arguments for state-court exclusivity, Arrow raises a concern about the impact on federal courts were we to uphold §1331 jurisdiction over private actions under the TCPA. “[G]iven the enormous volume of telecommunications presenting potential claims,” Arrow projects, federal courts could be inundated by $500-per-violation TCPA claims. Brief for Respondent 33. “Moreover, if plaintiffs are free to bring TCPA claims in federal court under §1331, then defendants sued in state court would be equally free to remove those cases to federal court under 28 U. S. C. §1441.” Id., at 22–23. Indeed, Arrow suggests, defendants could use removal as a mechanism to force small-claims-court plaintiffs to abandon suit rather than “figh[t] it out” in the “more expensive federal forum.” Id., at 23. Arrow’s floodgates argument assumes “a shocking degree of noncompliance” with the Act, Reply Brief 11, and seems to us more imaginary than real. The current fed- eral district court civil filing fee is $350. 28 U. S. C. §1914(a). How likely is it that a party would bring a $500 claim in, or remove a $500 claim to, federal court? Lexis and Westlaw searches turned up 65 TCPA claims removed to federal district courts in Illinois, Indiana, and Wisconsin since the Seventh Circuit held, in October 2005, that the Act does not confer exclusive jurisdiction on state courts. All 65 cases were class actions, not individual cases removed from small-claims court. [ 15 ] There were also 26 private TCPA claims brought initially in federal district courts; of those, 24 were class actions. IV Nothing in the text, structure, purpose, or legislative history of the TCPA calls for displacement of the federal-question jurisdiction U. S. district courts ordinarily have under 28 U. S. C. §1331. In the absence of direction from Congress stronger than any Arrow has advanced, we apply the familiar default rule: Federal courts have §1331 jurisdiction over claims that arise under federal law. Because federal law gives rise to the claim for relief Mims has stated and specifies the substantive rules of decision, the Eleventh Circuit erred in dismissing Mims’s case for lack of subject-matter jurisdiction. * * * For the reasons stated, the judgment of the United States Court of Appeals for the Eleventh Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 In general, the Communications Act of 1934 grants to the Fed-eral Communications Commission (FCC or Commission) authority to regulate interstate telephone communications and reserves to the States authority to regulate intrastate telephone communications. See Louisiana Pub. Serv. Comm’n v. FCC, –370 (1986). 2 In 2010, Congress amended the statute to prohibit an additional practice: the manipulation of caller-identification information. See Truth in Caller ID Act of 2009, Pub. L. 111–331, . This legislation inserted a new subsection (e) into and redesignated the former subsections (e), (f), and (g) as subsections (f), (g), and (h), respectively. Ibid. While the new subsection (e) does not bear on this case and is not here discussed, our citations of subsection (g) refer to the current, redesignated statutory text. 3 The National Do Not Call Registry is currently managed by the Federal Trade Commission. See (2006 ed., Supp. IV); 16 CFR §310.4(b)(1)(iii) (2011). 4 The TCPA envisions civil actions instituted by the Commission for violations of the implementing regulations. See 47 U. S. C. A. §227(g)(7) (Supp. 2011). The Commission may also seek forfeiture penalties for willful or repeated failure to comply with the Act or regulations. (2006 ed. and Supp. IV), §504(a) (2006 ed.). 5 Title , also captioned “Private right of action,” provides: “A person who has received more than one telephone call within any 12-month period by or on behalf of the same entity in violation of the regulations prescribed under this subsection may, if otherwise permitted by the laws or rules of court of a State bring in an appropriate court of that State— “(A) an action based on a violation of the regulations prescribed under this subsection to enjoin such violation, “(B) an action to recover for actual monetary loss from such a violation, or to receive up to $500 in damages for each such violation, whichever is greater, or “(C) both such actions.” 6 Congress had previously granted general federal-question jurisdiction to federal courts, but the grant was short lived. See Steffel v. Thompson, (describing Midnight Judges Act of 1801, §11, , repealed by Act of Mar. 8, 1802, §1, ). 7 At the time it was repealed, the amount-in-controversy requirement in federal-question cases had reached $10,000. See Act of July 25, 1958, . Currently, the amount in controversy in diversity cases must exceed $75,000. See . 8 For a rare exception to the rule that a federal cause of action suffices to ground federal-question jurisdiction, see Shoshone Mining Co. v. Rutter, , discussed in R. Fallon, J. Manning, D. Meltzer, & D. Shapiro, Hart and Wechsler’s The Federal Courts and the Federal System, 784–785 (6th ed. 2009). In Shoshone Mining, we held that a suit for a federal mining patent did not arise under federal law for jurisdictional purposes because “the right of possession” in controversy could be determined by “local rules or customs, or state statutes,” 177 U. S., at 509, or “may present simply a question of fact,” id., at 510. Here, by contrast, the TCPA not only creates the claim for relief and designates the remedy; critically, the Act and regulations thereunder supply the governing substantive law. 9 Even when a right of action is created by state law, if the claim requires resolution of significant issues of federal law, the case may arise under federal law for purposes. See Grable & Sons Metal Products, Inc. v. Darue Engineering & Mfg., . 10 “How strange it would be,” the Seventh Circuit observed, “to make federal courts the exclusive forum for suits by the states, while making state courts the exclusive forum for suits by private plaintiffs.” Brill v. Countrywide Home Loans, Inc., 427 F. 3d 446, 451 (2005). 11 For TCPA actions brought by State Attorneys General, or “an[other] official or agency designated by a State,” 47 U. S. C. A. §227(g)(1) (Supp. 2011), Arrow points out, Congress specifically addressed venue, service of process, §227(g)(4), and potential conflicts between federal and state enforcement efforts, §227(g)(7). No similar prescriptions appear in the section on private actions, , for this obvious reason: “[As] the general rules governing venue and service of process in the district courts are well established, see ; Fed. Rules Civ. Proc. 4, 4.1, there was no need for Congress to reiterate them in section 227(b)(3). The fact that venue and service of process are discussed in [47 U. S. C. A. §227(g)(4) (Supp. 2011)] and not [47 U. S. C. §]227(b)(3) simply indicates that Congress wished to make adjustments to the general rules in the former section and not the latter. As for the conflict provision that appears in section [47 U. S. C. A. §227(g) (Supp. 2011)] but not [47 U. S. C. §]227(b)(3), it is hardly surprising that Congress would be concerned about agency conflicts in the section of the TCPA dealing with official state enforcement efforts but not in the section governing private lawsuits.” ErieNet, Inc. v. Velocity Net, Inc., 156 F. 3d 513, 523 (CA3 1998) (Alito, J., dissenting). 12 The Supremacy Clause declares federal law the “supreme law of the land,” and state courts must enforce it “in the absence of a valid excuse.” Howlett v. Rose, . “An excuse that is inconsistent with or violates federal law is not a valid excuse: The Supremacy Clause forbids state courts to dissociate themselves from federal law because of disagreement with its content or a refusal to recognize the superior authority of its source.” Id., at 371. Without the “if otherwise permitted” language, , there is little doubt that state courts would be obliged to hear TCPA claims. See Testa v. Katt, . 13 Although all courts of appeals to have considered the question have held that the TCPA does not bar district courts from exercising diver-sity jurisdiction under , see, e.g., Gottlieb v. Carnival Corp., 436 F. 3d 335 (CA2 2006), at oral argument, Arrow’s counsel maintained that diversity jurisdiction “should go, too,” Tr. of Oral Arg. 39. Were we to accept Arrow’s positions that diversity and federal-question jurisdiction are unavailable, and that state courts may refuse to hear TCPA claims, residents of States that choose not to hear TCPA claims would have no forum in which to sue. 14 The complaint in this very case, we note, could not have been brought in small-claims court. Mims alleged some 12 calls, and sought treble damages ($1,500) for each. See App. 9–14; Tr. of Oral Arg. 12. The amount he sought to recover far exceeded the $5,000 ceiling on claims a Florida small-claims court can adjudicate. See Fla. Small Claims Rule 7.010(b) (rev. ed. 2011). 15 When Congress wants to make federal claims instituted in state court nonremovable, it says just that. See Breuer v. Jim’s Concrete of Brevard, Inc., –697 (2003) (quoting, e.g., (“A civil action in any State court against a railroad or its receivers or trustees, [arising under §§51–60 of Title 45,] may not be removed to any district court of the United States.”) and (“[N]o case arising under this subchapter and brought in any State court of competent jurisdiction shall be removed to any court of the United States.”)).

565.US.2011_10-1104

Respondent Pollard sought damages from employees at a privately run federal prison in California, claiming that they had deprived him of adequate medical care in violation of the Eighth Amendment’s prohibition against cruel and unusual punishment. The Federal District Court dismissed the complaint, ruling that the Eighth Amendment does not imply an action under Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388, against a privately managed prison’s personnel. The Ninth Circuit reversed. Held: Because in the circumstance of this case, state tort law authorizes adequate alternative damages actions—providing both significant deterrence and compensation—no Bivens remedy can be implied here. Pp. 3−12. (a) Wilkie v. Robbins, 551 U.S. 537, fairly summarizes the basic considerations the Court applies here. In deciding whether to recognize a Bivens remedy, a court must first ask “whether any alternative, existing process for protecting the [constitutionally recognized] interest amounts to a convincing reason for the Judicial Branch to refrain from providing a new and freestanding” damages remedy. Even absent an alternative, “a Bivens remedy is a subject of judgment: ‘the federal courts must make the kind of remedial determination that is appropriate for a common-law tribunal, paying particular heed . . . to any special factors counselling hesitation before authorizing a new kind of federal litigation.’ ” Id., at 550. In Bivens itself, the Court held that the Fourth Amendment implicitly authorized a court to order federal agents to pay damages to a person injured by the agents’ violation of the Amendment’s strictures, 403 U. S., at 389, noting that the Fourth Amendment prohibited conduct that state law might permit, id., at 392–393, and that the interests protected on the one hand by state “trespass” and “invasion of privacy” laws and on the other hand by the Fourth Amendment “may be inconsistent or even hostile,” id., at 394. It also stated that “[h]istorically, damages have been regarded as the ordinary remedy for an invasion of personal interests in liberty,” id., at 395, and found “no special factors counselling hesitation in the absence of affirmative action by Congress.” Id., at 396. Bivens actions were allowed in Davis v. Passman, 442 U.S. 228, for a Fifth Amendment due process claim involving gender-based employment discrimination, and in Carlson v. Green, 446 U.S. 14, for an Eighth Amendment claim based on federal government officials’ “deliberat[e] indifferen[ce]” to a federal prisoner’s medical needs, id., at 16, n. 1, 17. Since Carlson, this Court has declined to imply a Bivens action in several different instances. See, e.g., Bush v. Lucas, 462 U.S. 367, Correctional Services Corp. v. Malesko, 534 U.S. 61. Applying Wilkie’s approach here, Pollard cannot assert a Bivens claim, primarily because his Eighth Amendment claim focuses on a kind of conduct that typically falls within the scope of traditional state tort law. And in the case of a privately employed defendant, state tort law provides an “alternative, existing process” capable of protecting the constitutional interests at stake. Wilkie, 551 U. S., at 550. The existence of that alternative remedy constitutes a “convincing reason for the Judicial Branch to refrain from providing a new and freestanding” damages remedy. Ibid. Pp. 3−7. (b) Pollard’s contrary arguments are rejected. First, he claims that Carlson authorizes an Eighth Amendment-based Bivens action here, but Carlson involved government, not privately employed, personnel. The potential existence of an “adequate alternative, existing process” differs dramatically for public and private employees, as prisoners ordinarily can bring state tort actions against private employees, but not against public ones. Second, Pollard’s argument that this Court should consider only whether federal laws provide adequate alternative remedies because of the “vagaries” of state tort law, Carlson, supra, at 23, was rejected in Malesko, supra, at 72−73. Third, Pollard claims that state tort law does not provide remedies adequate to protect the constitutional interests at issue here, but California, like every other State (as far as the Court is aware), has tort law that provides for negligence actions for claims such as his. That the state law may prove less generous than would a Bivens action does not render the state law inadequate, and state remedies and a potential Bivens remedy need not be perfectly congruent. Fourth, Pollard argues that there may be similar Eighth Amendment claims that state tort law does not cover, but he offers no supporting cases. The possibility of a future case, where an Eighth Amendment claim or state law differs significantly from those at issue, provides insufficient grounds for reaching a different conclusion here. Pp. 7−12. 607 F.3d 583 and 629 F.3d 843, reversed. Breyer, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Thomas, Alito, Sotomayor, and Kagan, JJ., joined. Scalia, J., filed a concurring opinion, in which Thomas, J., joined. Ginsburg, J., filed a dissenting opinion.

The question is whether we can imply the existence of an Eighth Amendment-based damages action (a Bivens action) against employees of a privately operated federal prison. See generally Bivens v. Six Unknown Fed. Narcotics Agents, 403 U. S. 388, 389 (1971) (“[V]iolation of [the Fourth Amendment] by a federal agent . . . gives rise to a cause of action for damages” against a Federal Government employee). Because we believe that in the circumstances present here state tort law authorizes adequate alternative damages actions—actions that provide both significant deterrence and compensation—we cannot do so. See Wilkie v. Robbins, 551 U. S. 537, 550 (2007) (no Bivens action where “alternative, existing” processes provide adequate protection). I Richard Lee Pollard was a prisoner at a federal facility operated by a private company, the Wackenhut Corrections Corporation. In 2002 he filed a pro se complaint in federal court against several Wackenhut employees, who (now) include a security officer, a food-services supervisor, and several members of the medical staff. As the Federal Magistrate Judge interpreted Pollard’s complaint, he claimed that these employees had deprived him of adequate medical care, had thereby violated the Eighth Amendment’s prohibition against “cruel and unusual” punishment, and had caused him injury. He sought damages. Pollard said that a year earlier he had slipped on a cart left in the doorway of the prison’s butcher shop. The prison medical staff took x rays, thought he might have fractured both elbows, brought him to an outside clinic for further orthopedic evaluation, and subsequently arranged for surgery. In particular, Pollard claimed: (1) Despite his having told a prison guard that he could not extend his arm, the guard forced him to put on a jumpsuit (to travel to the outside clinic), causing him “the most excruciating pain,” App. 32; (2) During several visits to the outside clinic, prison guards made Pollard wear arm restraints that were connected in a way that caused him continued pain; (3) Prison medical (and other) personnel failed to follow the outside clinic’s instructions to put Pollard’s left elbow in a posterior splint, failed to provide necessary physical therapy, and failed to conduct necessary studies, including nerve conduction studies; (4) At times when Pollard’s arms were in casts or similarly disabled, prison officials failed to make alternative arrangements for him to receive meals, with the result that (to avoid “being humiliated” in the general food service area, id., at 35) Pollard had to auction off personal items to obtain funds to buy food at the commissary; (5) Prison officials deprived him of basic hygienic care to the point where he could not bathe for two weeks; (6) Prison medical staff provided him with insufficient medicine, to the point where he was in pain and could not sleep; and (7) Prison officials forced him to return to work before his injuries had healed. After concluding that the Eighth Amendment did not provide for a Bivens action against a privately managed prison’s personnel, the Magistrate Judge recommended that the District Court dismiss Pollard’s complaint. The District Court did so. But on appeal the Ninth Circuit found that the Eighth Amendment provided Pollard with a Bivens action, and it reversed the District Court. Pollard v. The GEO Group, Inc., 607 F. 3d 583, 603, as amended, 629 F. 3d 843, 868 (CA9 2010). The defendants sought certiorari. And, in light of a split among the Courts of Appeals, we granted the petition. Com- pare ibid. (finding an Eighth Amendment Bivens action where prisoner sues employees of a privately operated federal prison), with, e.g., Alba v. Montford, 517 F. 3d 1249, 1254–1256 (CA11 2008) (no Bivens action available), and Holly v. Scott, 434 F. 3d 287, 288 (CA4 2006) (same). II Recently, in Wilkie v. Robbins, supra, we rejected a claim that the Fifth Amendment impliedly authorized a Bivens action that would permit landowners to obtain damages from government officials who unconstitutionally interfere with their exercise of property rights. After reviewing the Court’s earlier Bivens cases, the Court stated: “[T]he decision whether to recognize a Bivens remedy may require two steps. In the first place, there is the question whether any alternative, existing process for protecting the [constitutionally recognized] interest amounts to a convincing reason for the Judicial Branch to refrain from providing a new and freestanding remedy in damages. . . . But even in the absence of an alternative, a Bivens remedy is a subject of judgment: ‘the federal courts must make the kind of remedial determination that is appropriate for a common-law tribunal, paying particular heed, how- ever, to any special factors counselling hesitation be- fore authorizing a new kind of federal litigation.’ ” 551 U. S., at 550 (quoting Bush v. Lucas, 462 U. S. 367, 378 (1983) ). These standards seek to reflect and to reconcile the Court’s reasoning set forth in earlier cases. In Bivens itself the Court held that the Fourth Amendment implicitly authorized a court to order federal agents to pay damages to a person injured by the agents’ violation of the Amendment’s constitutional strictures. 403 U. S., at 389. The Court noted that “ ‘where federally protected rights have been invaded,’ ”courts can “ ‘adjust their remedies so as to grant the necessary relief.’ ” Id., at 392 (quoting Bell v. Hood, 327 U. S. 678, 684 (1946) ). See also Correctional Services Corp. v. Malesko, 534 U. S. 61, 66 (2001) (“authority to imply a new constitutional tort” anchored within general “ ‘arising under’ ” jurisdiction). It pointed out that the Fourth Amendment prohibited, among other things, conduct that state law might permit (such as the conduct at issue in that very case). Bivens, 403 U. S., at 392–393. It added that the interests protected on the one hand by state “trespass” and “invasion of privacy” laws and on the other hand by the Fourth Amendment’s guarantees “may be inconsistent or even hostile.” Id., at 394. It stated that “[h]istorically, damages have been regarded as the ordinary remedy for an invasion of personal interests in liberty.” Id., at 395. And it found “no special factors counselling hesitation in the absence of affirmative action by Congress.” Id., at 396. In Davis v. Passman, 442 U. S. 228 (1979) , the Court considered a former congressional employee’s claim for damages suffered as a result of her employer’s unconstitutional discrimination based on gender. The Court found a damages action implicit in the Fifth Amendment’s Due Process Clause. Id., at 248–249. In doing so, the Court emphasized the unavailability of “other alternative forms of judicial relief.” Id., at 245. And the Court noted that there was “no evidence” that Congress (or the Constitution) intended to foreclose such a remedy. Id., at 247. In Carlson v. Green, 446 U. S. 14 (1980) , the Court considered a claim for damages brought by the estate of a federal prisoner who (the estate said) had died as the result of government officials’ “deliberat[e] indifferen[ce]” to his medical needs—indifference that violated the Eighth Amendment. Id., at 16, n. 1, 17 (citing Estelle v. Gamble, 429 U. S. 97 (1976) ). The Court implied an action for damages from the Eighth Amendment. 446 U. S., at 17–18. It noted that state law offered the particular plaintiff no meaningful damages remedy. Id., at 17, n. 4. Although the estate might have brought a damages claim under the Federal Tort Claims Act, the defendant in any such lawsuit was the employer, namely the United States, not the individual officers who had committed the violation. Id., at 21. A damages remedy against an individual officer, the Court added, would prove a more effective deterrent. Ibid. And, rather than leave compensation to the “vagaries” of state tort law, a federal Bivens action would provide “uniform rules.” 446 U. S., at 23. Since Carlson, the Court has had to decide in several different instances whether to imply a Bivens action. And in each instance it has decided against the existence of such an action. These instances include: (1) A federal employee’s claim that his federal employer dismissed him in violation of the First Amendment, Bush, supra, at 386–388 (congressionally created federal civil service procedures provide meaningful redress); (2) A claim by military personnel that military superiors violated various constitutional provisions, Chappell v. Wallace, 462 U. S. 296 –300 (1983) (special factors related to the military counsel against implying a Bivens action), see also United States v. Stanley, 483 U. S. 669 –684 (1987) (similar); (3) A claim by recipients of Social Security disability benefits that benefits had been denied in violation of the Fifth Amendment, Schweiker v. Chilicky, 487 U. S. 412, 414, 425 (1988) (elaborate administrative scheme provides meaningful alternative remedy); (4) A former bank employee’s suit against a federal banking agency, claiming that he lost his job due to agency action that violated the Fifth Amendment’s Due Process Clause, FDIC v. Meyer, 510 U. S. 471 –486 (1994) (no Bivens actions against government agencies rather than particular individuals who act unconstitutionally); (5) A prisoner’s Eighth Amendment-based suit against a private corporation that managed a federal prison, Ma-lesko, 534 U. S., at 70–73 (to permit suit against the employer-corporation would risk skewing relevant incentives; at the same time, the ability of a prisoner to bring state tort law damages action against private individual defendants means that the prisoner does not “lack effective remedies,” id., at 72). Although the Court, in reaching its decisions, has not always similarly emphasized the same aspects of the cases, Wilkie fairly summarizes the basic considerations that underlie those decisions. 551 U. S., at 550. We consequently apply its approach here. And we conclude that Pollard cannot assert a Bivens claim. That is primarily because Pollard’s Eighth Amendment claim focuses upon a kind of conduct that typically falls within the scope of traditional state tort law. And in the case of a privately employed defendant, state tort law provides an “alternative, existing process” capable of protecting the constitutional interests at stake. 551 U. S., at 550. The existence of that alternative here constitutes a “convincing reason for the Judicial Branch to refrain from providing a new and freestanding remedy in dam- ages.” Ibid. Our reasoning is best understood if we set forth and explain why we reject Pollard’s arguments to the contrary. III Pollard (together with supporting amici) asks us to imply a Bivens action for four basic reasons—none of which we find convincing. First, Pollard argues that this Court has already decided in Carlson that a federal prisoner may bring an Eighth Amendment-based Bivens action against prison personnel; and we need do no more than simply apply Carlson’s holding here. Carlson, however, was a case in which a federal prisoner sought damages from personnel employed by the government, not personnel employed by a private firm. 446 U. S., at 25. And for present purposes that fact—of employment status—makes a critical difference. For one thing, the potential existence of an adequate “alternative, existing process” differs dramatically in the two sets of cases. Prisoners ordinarily cannot bring state-law tort actions against employees of the Federal Government. See 28 U. S. C. §§2671, 2679(b)(1) (Westfall Act) (substituting United States as defendant in tort action against federal employee); Osborn v. Haley, 549 U. S. 225, 238, 241 (2007) (Westfall Act immunizes federal employee through removal and substitution of United States as defendant). But prisoners ordinarily can bring state-law tort actions against employees of a private firm. Infra, at 9–10. For another thing, the Court specifically rejected Justice Stevens’ somewhat similar suggestion in his dissenting opinion in Malesko, namely that a prisoner’s suit against a private prison-management firm should fall within Carlson’s earlier holding because such a firm, like a federal employee, is a “federal agent.” Compare Malesko, 534 U. S., at 70, and n. 4 (majority opinion), with id., at 76–77, 82 (dissenting opinion). In rejecting the dissent’s suggestion, the Court explained that the context in Malesko was “fundamentally different” from the contexts at issue in earlier cases, including Carlson. 534 U. S., at 70. That difference, the Court said, reflected in part the nature of the defendant, i.e., a corporate employer rather than an individual employee, ibid., and in part reflected the existence of alternative “effective” state tort remedies, id., at 72–73. This last-mentioned factor makes it difficult to square Pollard’s argument with Malesko’s reasoning. Second, Pollard argues that, because of the “vagaries” of state tort law, Carlson, 446 U. S., at 23, we should consider only whether federal law provides adequate alternative remedies. See id., at 18–19, 23 (considering adequacy of federal remedies); see also, e.g., Schweiker, supra, at 423 (similar); Bush, 462 U. S., at 378 (similar). But cf. Carlson, supra, at 24 (“ ‘[R]elevant Indiana statute would not permit survival of the [state tort] claim’ ”). This argument flounders, however, on the fact that the Court rejected it in Malesko. Compare 534 U. S., at 72–73 (majority opinion), with id., at 79–80 (Stevens, J., dissenting) (making similar suggestion). State tort law, after all, can help to deter constitutional violations as well as to provide compensation to a violation’s victim. And it is consequently unsurprising that several cases have considered the adequacy or inadequacy of state-law remedies when determining whether to imply a Bivens remedy. See, e.g., Bivens, 403 U. S., at 394 (state tort law “inconsistent or even hostile” to Fourth Amendment); Davis, 442 U. S., at 245, n. 23 (noting no state-law remedy available); cf. Malesko, supra, at 70 (noting that the Court has implied Bivens action only where any alternative remedy against individual officers was “nonexistent” or where plaintiff “lacked any alternative remedy” at all). Third, Pollard argues that state tort law does not provide remedies adequate to protect the constitutional interests at issue here. Pollard’s claim, however, is a claim for physical or related emotional harm suffered as a result of aggravated instances of the kind of conduct that state tort law typically forbids. That claim arose in California, where state tort law provides for ordinary negligence actions, for actions based upon “want of ordinary care or skill,” for actions for “negligent failure to diagnose or treat,” and for actions based upon the failure of one with a custodial duty to care for another to protect that other from “ ‘unreasonable risk of physical harm.’ ” See Cal. Civ. Code Ann. §§1714(a), 1714.8(a) (West 2009 and Supp. 2012); Giraldo v. California Dept. of Corrections and Rehabilitation, 168 Cal. App. 4th 231, 248, 85 Cal. Rptr. 3d 371, 384 (2008) (quoting Haworth v. State, 60 Haw. 557, 562, 592 P. 2d 820, 824 (1979)). California courts have specifically applied this law to jailers, including private operators of prisons. Giraldo, supra, at 252, 85 Cal. Rptr. 3d, at 387 (“[J]ailers owe prisoners a duty of care to protect them from foreseeable harm”); see also Lawson v. Superior Ct., 180 Cal. App. 4th 1372, 1389–1390, 1397, 103 Cal. Rptr. 3d 834, 849–850, 855 (2010) (same). Moreover, California’s tort law basically reflects general principles of tort law present, as far as we can tell, in the law of every State. See Restatement (Second) of Torts §§314A(4), 320 (1963–1964). We have found specific authority indicating that state law imposes general tort duties of reasonable care (including medical care) on prison employees in every one of the eight States where privately managed secure federal facilities are currently lo- cated. See Dept. of Justice, Federal Bureau of Prisions, Weekly Population Report (Dec 22, 2011), http:// www.bop.gov/locations/weekly_report.jsp (listing States) (as visited Dec. 29, 2011, and available in Clerk of Court’s case file); Thomas v. Williams, 105 Ga. App. 321, 326, 124 S. E. 2d 409, 412–413 (1962) (In Georgia, “ ‘sheriff owes to a prisoner placed in his custody a duty to keep the prisoner safely and free from harm, to render him medical aid when necessary, and to treat him humanely and refrain from oppressing him’ ”); Giraldo, supra, at 248, 85 Cal. Rptr. 3d, at 384 (California, same); Farmer v. State ex rel. Russell, 224 Miss. 96, 105, 79 So. 2d 528, 531 (1955) (Mississippi, same); Doe v. Albuquerque, 96 N. M. 433, 438, 631 P. 2d 728, 733 (App. 1981) (New Mexico, same); Multiple Claimants v. North Carolina Dept. of Health and Human Servs., 176 N. C. App. 278, 280, 626 S. E. 2d 666, 668 (2006) (North Carolina, same); Clemets v. Heston, 20 Ohio App. 3d 132, 135–136, 485 N. E. 2d 287, 291 (1985) (Ohio, same); Williams v. Syed, 782 A. 2d 1090, 1093–1094 (Pa. Commw. 2001) (Pennsylvania, same); Salazar v. Collins, 255 S. W. 3d 191, 198–200 (Tex. App. 2008) (Texas, same); see also Schellenger, 14 A. L. R. 2d 353, §2[a] (Later Case Service and Supp. 2011) (same). But cf. Miss. Code. Ann. §11–46–9(1)(m) (Supp. 2011) (statute forbidding such actions against State—though not private—employees); N. Y. Correc. Law Ann. §§24 (West 2003), 121 (2011 Cum. Supp.) (similar). We note, as Pollard points out, that state tort law may sometimes prove less generous than would a Bivens action, say, by capping damages, see Cal. Civ. Code Ann. §3333.2(b) (West 1997), or by forbidding recovery for emotional suffering unconnected with physical harm, see 629 F. 3d, at 864, or by imposing procedural obstacles, say, initially requiring the use of expert administrative panels in medical malpractice cases, see, e.g., Me. Rev. Stat. Ann., Tit. 24, §2853, (Supp. 2010); Mass. Gen. Laws, ch. 231, §60B (West 2010). But we cannot find in this fact sufficient basis to determine state law inadequate. State-law remedies and a potential Bivens remedy need not be perfectly congruent. See Bush, supra, at 388 (administrative remedies adequate even though they “do not provide complete relief”). Indeed, federal law as well as state law contains limitations. Prisoners bringing federal lawsuits, for example, ordinarily may not seek damages for mental or emotional injury unconnected with physical injury. See 42 U. S. C. §1997e(e). And Bivens actions, even if more generous to plaintiffs in some respects, may be less generous in others. For example, to show an Eighth Amendment violation a prisoner must typically show that a defendant acted, not just negligently, but with “deliberate indifference.” Farmer v. Brennan, 511 U. S. 825, 834 (1994) . And a Bivens plaintiff, unlike a state tort law plaintiff, normally could not apply principles of respondeat superior and thereby obtain recovery from a defendant’s potentially deep-pocketed employer. See Ash- croft v. Iqbal, 556 U. S. 662, 676 (2009) . Rather, in principle, the question is whether, in general, state tort law remedies provide roughly similar incentives for potential defendants to comply with the Eighth Amendment while also providing roughly similar compensation to victims of violations. The features of the two kinds of actions just mentioned suggest that, in practice, the answer to this question is “yes.” And we have found nothing here to convince us to the contrary. Fourth, Pollard argues that there “may” be similar kinds of Eighth Amendment claims that state tort law does not cover. But Pollard does not convincingly show that there are such cases. Compare Brief for Respondent Pollard 32 (questioning the availability of state tort remedies for “prisoners [who] suffer attacks by other inmates, preventable suicides, or the denial of heat, ventilation or movement”), with Giraldo, supra, at 248–249, 85 Cal Rptr. 3d, at 384–385 (courts have long held that prison officials must protect, e.g., transgender inmate from foreseeable harm by other inmates), and Restatement (Second) of Torts §§314A(4), 320. Regardless, we concede that we cannot prove a negative or be totally certain that the features of state tort law relevant here will universally prove to be, or remain, as we have described them. Nonetheless, we are certain enough about the shape of present law as applied to the kind of case before us to leave different cases and different state laws to another day. That is to say, we can decide whether to imply a Bivens action in a case where an Eighth Amendment claim or state law differs significantly from those at issue here when and if such a case arises. The possibility of such a different future case does not provide sufficient grounds for reaching a different conclusion here. For these reasons, where, as here, a federal prisoner seeks damages from privately employed personnel working at a privately operated federal prison, where the conduct allegedly amounts to a violation of the Eighth Amendment, and where that conduct is of a kind that typically falls within the scope of traditional state tort law (such as the conduct involving improper medical care at issue here), the prisoner must seek a remedy under state tort law. We cannot imply a Bivens remedy in such a case. The judgment of the Ninth Circuit is reversed. So ordered.

566.US.2011_10-444

Respondent Frye was charged with driving with a revoked license. Because he had been convicted of the same offense three times before, he was charged, under Missouri law, with a felony carrying a maximum 4-year prison term. The prosecutor sent Frye’s counsel a letter, offering two possible plea bargains, including an offer to reduce the charge to a misdemeanor and to recommend, with a guilty plea, a 90-day sentence. Counsel did not convey the offers to Frye, and they expired. Less than a week before Frye’s preliminary hearing, he was again arrested for driving with a revoked license. He subsequently pleaded guilty with no underlying plea agreement and was sentenced to three years in prison. Seeking postconviction relief in state court, he alleged his counsel’s failure to inform him of the earlier plea offers denied him the effective assistance of counsel, and he testified that he would have pleaded guilty to the misdemeanor had he known of the offer. The court denied his motion, but the Missouri appellate court reversed, holding that Frye met both of the requirements for showing a Sixth Amendment violation under Strickland v. Washington, 466 U.S. 668. Specifically, the court found that defense counsel had been ineffective in not communicating the plea offers to Frye and concluded that Frye had shown that counsel’s deficient performance caused him prejudice because he pleaded guilty to a felony instead of a misdemeanor. Held: 1. The Sixth Amendment right to effective assistance of counsel extends to the consideration of plea offers that lapse or are rejected. That right applies to “all ‘critical’ stages of the criminal proceedings.” Montejo v. Louisiana, 556 U.S. 778, 786. Hill v. Lockhart, 474 U.S. 52, established that Strickland’s two-part test governs ineffective-assistance claims in the plea bargain context. There, the defendant had alleged that his counsel had given him inadequate advice about his plea, but he failed to show that he would have proceeded to trial had he received the proper advice. 474 U. S., at 60. In Padilla v. Kentucky, 559 U. S. ___, where a plea offer was set aside because counsel had misinformed the defendant of its immigration consequences, this Court made clear that “the negotiation of a plea bargain is a critical” stage for ineffective-assistance purposes, id., at ___, and rejected the argument made by the State in this case that a knowing and voluntary plea supersedes defense counsel’s errors. The State attempts to distinguish Hill and Padilla from the instant case. It notes that Hill and Padilla concerned whether there was ineffective assistance leading to acceptance of a plea offer, a process involving a formal court appearance with the defendant and all counsel present, while no formal court proceedings are involved when a plea offer has lapsed or been rejected; and it insists that there is no right to receive a plea offer in any event. Thus, the State contends, it is unfair to subject it to the consequences of defense counsel’s inadequacies when the opportunities for a full and fair trial, or for a later guilty plea albeit on less favorable terms, are preserved. While these contentions are neither illogical nor without some persuasive force, they do not suffice to overcome the simple reality that 97 percent of federal convictions and 94 percent of state convictions are the result of guilty pleas. Plea bargains have become so central to today’s criminal justice system that defense counsel must meet responsibilities in the plea bargain process to render the adequate assistance of counsel that the Sixth Amendment requires at critical stages of the criminal process. Pp. 3–8. 2. As a general rule, defense counsel has the duty to communicate formal prosecution offers to accept a plea on terms and conditions that may be favorable to the accused. Any exceptions to this rule need not be addressed here, for the offer was a formal one with a fixed expiration date. Standards for prompt communication and consultation recommended by the American Bar Association and adopted by numerous state and federal courts, though not determinative, serve as important guides. The prosecution and trial courts may adopt measures to help ensure against late, frivolous, or fabricated claims. First, a formal offer’s terms and processing can be documented. Second, States may require that all offers be in writing. Third, formal offers can be made part of the record at any subsequent plea proceeding or before trial to ensure that a defendant has been fully advised before the later proceedings commence. Here, as the result of counsel’s deficient performance, the offers lapsed. Under Strickland, the question then becomes what, if any, prejudice resulted from the breach of duty. Pp. 8–11. 3. To show prejudice where a plea offer has lapsed or been rejected because of counsel’s deficient performance, defendants must demonstrate a reasonable probability both that they would have accepted the more favorable plea offer had they been afforded effective assistance of counsel and that the plea would have been entered without the prosecution’s canceling it or the trial court’s refusing to accept it, if they had the authority to exercise that discretion under state law. This application of Strickland to uncommunicated, lapsed pleas does not alter Hill’s standard, which requires a defendant complaining that ineffective assistance led him to accept a plea offer instead of going to trial to show “a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” 474 U. S., at 59. Hill correctly applies in the context in which it arose, but it does not provide the sole means for demonstrating prejudice arising from counsel’s deficient performance during plea negotiations. Because Frye argues that with effective assistance he would have accepted an earlier plea offer as opposed to entering an open plea, Strickland’s inquiry into whether “the result of the proceeding would have been different,” 466 U. S., at 694, requires looking not at whether the defendant would have proceeded to trial but at whether he would have accepted the earlier plea offer. He must also show that, if the prosecution had the discretion to cancel the plea agreement or the trial court had the discretion to refuse to accept it, there is a reasonable probability neither the prosecution nor the trial court would have prevented the offer from being accepted or implemented. This further showing is particularly important because a defendant has no right to be offered a plea, see Weatherford v. Bursey, 429 U.S. 545, 561, nor a federal right that the judge accept it, Santobello v. New York, 404 U.S. 257, 262. Missouri, among other States, appears to give the prosecution some discretion to cancel a plea agreement; and the Federal Rules of Criminal Procedure, some state rules, including Missouri’s, and this Court’s precedents give trial courts some leeway to accept or reject plea agreements. Pp. 11–13. 4. Applying these standards here, the Missouri court correctly concluded that counsel’s failure to inform Frye of the written plea offer before it expired fell below an objective reasonableness standard, but it failed to require Frye to show that the plea offer would have been adhered to by the prosecution and accepted by the trial court. These matters should be addressed by the Missouri appellate court in the first instance. Given that Frye’s new offense for driving without a license occurred a week before his preliminary hearing, there is reason to doubt that the prosecution would have adhered to the agreement or that the trial court would have accepted it unless they were required by state law to do so. Pp. 13–15. 311 S.W.3d 350, vacated and remanded. Kennedy, J., delivered the opinion of the Court, in which Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined. Scalia, J., filed a dissenting opinion, in which Roberts, C. J., and Thomas and Alito, JJ., joined.

The Sixth Amendment, applicable to the States by the terms of the Fourteenth Amendment, provides that the ac- cused shall have the assistance of counsel in all criminal prosecutions. The right to counsel is the right to effective assistance of counsel. See Strickland v. Washington, 466 U.S. 668, 686 (1984). This case arises in the context of claimed ineffective assistance that led to the lapse of a prosecution offer of a plea bargain, a proposal that offered terms more lenient than the terms of the guilty plea entered later. The initial question is whether the consti- tutional right to counsel extends to the negotiation and consideration of plea offers that lapse or are rejected. If there is a right to effective assistance with respect to those offers, a further question is what a defendant must dem- onstrate in order to show that prejudice resulted from counsel’s deficient performance. Other questions relating to ineffective assistance with respect to plea offers, including the question of proper remedies, are considered in a second case decided today. See Lafler v. Cooper, post, at 3–16. I In August 2007, respondent Galin Frye was charged with driving with a revoked license. Frye had been convicted for that offense on three other occasions, so the State of Missouri charged him with a class D felony, which carries a maximum term of imprisonment of four years. See Mo. Rev. Stat. §§302.321.2, 558.011.1(4) (2011). On November 15, the prosecutor sent a letter to Frye’s counsel offering a choice of two plea bargains. App. 50. The prosecutor first offered to recommend a 3-year sentence if there was a guilty plea to the felony charge, without a recommendation regarding probation but with a recommendation that Frye serve 10 days in jail as so-called “shock” time. The second offer was to reduce the charge to a misdemeanor and, if Frye pleaded guilty to it, to recommend a 90-day sentence. The misdemeanor charge of driving with a revoked license carries a maximum term of imprisonment of one year. 311 S.W.3d 350, 360 (Mo. App. 2010). The letter stated both offers would expire on December 28. Frye’s attorney did not advise Frye that the offers had been made. The offers expired. Id., at 356. Frye’s preliminary hearing was scheduled for January 4, 2008. On December 30, 2007, less than a week before the hearing, Frye was again arrested for driving with a re- voked license. App. 47–48, 311 S. W. 3d, at 352–353. At the January 4 hearing, Frye waived his right to a preliminary hearing on the charge arising from the August 2007 arrest. He pleaded not guilty at a subsequent arraignment but then changed his plea to guilty. There was no underlying plea agreement. App. 5, 13, 16. The state trial court accepted Frye’s guilty plea. Id., at 21. The prosecutor recommended a 3-year sentence, made no recommendation regarding probation, and requested 10 days shock time in jail. Id., at 22. The trial judge sentenced Frye to three years in prison. Id., at 21, 23. Frye filed for postconviction relief in state court. Id., at 8, 25–29. He alleged his counsel’s failure to inform him of the prosecution’s plea offer denied him the effective assistance of counsel. At an evidentiary hearing, Frye testified he would have entered a guilty plea to the misdemeanor had he known about the offer. Id., at 34. A state court denied the postconviction motion, id., at 52–57, but the Missouri Court of Appeals reversed, 311 S.W.3d 350. It determined that Frye met both of the requirements for showing a Sixth Amendment violation under Strickland. First, the court determined Frye’s counsel’s performance was deficient because the “record is void of any evidence of any effort by trial counsel to communicate the Offer to Frye during the Offer window.” 311 S. W. 3d, at 355, 356 (emphasis deleted). The court next concluded Frye had shown his counsel’s deficient performance caused him prejudice because “Frye pled guilty to a felony instead of a misdemeanor and was subject to a maximum sentence of four years instead of one year.” Id., at 360. To implement a remedy for the violation, the court deemed Frye’s guilty plea withdrawn and remanded to allow Frye either to insist on a trial or to plead guilty to any offense the prosecutor deemed it appropriate to charge. This Court granted certiorari. 562 U. S. ___ (2011). II A It is well settled that the right to the effective assistance of counsel applies to certain steps before trial. The “ Sixth Amendment guarantees a defendant the right to have counsel present at all ‘critical’ stages of the criminal proceedings.” Montejo v. Louisiana, 556 U.S. 778, 786 (2009) (quoting United States v. Wade, 388 U.S. 218, 227–228 (1967)). Critical stages include arraignments, postindictment interrogations, postindictment lineups, and the entry of a guilty plea. See Hamilton v. Alabama, 368 U.S. 52 (1961) (arraignment); Massiah v. United States, 377 U.S. 201 (1964) (postindictment interrogation); Wade, supra (postindictment lineup); Argersinger v. Hamlin, 407 U.S. 25 (1972) (guilty plea). With respect to the right to effective counsel in plea negotiations, a proper beginning point is to discuss two cases from this Court considering the role of counsel in advising a client about a plea offer and an ensuing guilty plea: Hill v. Lockhart, 474 U.S. 52 (1985); and Padilla v. Kentucky, 559 U. S. ___(2010). Hill established that claims of ineffective assistance of counsel in the plea bargain context are governed by the two-part test set forth in Strickland. See Hill, supra, at 57. As noted above, in Frye’s case, the Missouri Court of Appeals, applying the two part test of Strickland, determined first that defense counsel had been ineffective and second that there was resulting prejudice. In Hill, the decision turned on the second part of the Strickland test. There, a defendant who had entered a guilty plea claimed his counsel had misinformed him of the amount of time he would have to serve before he became eligible for parole. But the defendant had not alleged that, even if adequate advice and assistance had been given, he would have elected to plead not guilty and proceed to trial. Thus, the Court found that no prejudice from the inadequate advice had been shown or alleged. Hill, supra, at 60. In Padilla, the Court again discussed the duties of counsel in advising a client with respect to a plea offer that leads to a guilty plea. Padilla held that a guilty plea, based on a plea offer, should be set aside because counsel misinformed the defendant of the immigration consequences of the conviction. The Court made clear that “the negotiation of a plea bargain is a critical phase of litigation for purposes of the Sixth Amendment right to effective assistance of counsel.” 559 U. S., at ___ (slip op., at 16). It also rejected the argument made by petitioner in this case that a knowing and voluntary plea supersedes errors by defense counsel. Cf. Brief for Respondent in Padilla v. Kentucky, O. T. 2009, No. 08–651, p. 27 (arguing Sixth Amendment’s assurance of effective assistance “does not extend to collateral aspects of the prosecution” because “knowledge of the consequences that are collateral to the guilty plea is not a prerequisite to the entry of a knowing and intelligent plea”). In the case now before the Court the State, as petitioner, points out that the legal question presented is different from that in Hill and Padilla. In those cases the claim was that the prisoner’s plea of guilty was invalid because counsel had provided incorrect advice pertinent to the plea. In the instant case, by contrast, the guilty plea that was accepted, and the plea proceedings concerning it in court, were all based on accurate advice and information from counsel. The challenge is not to the advice pertaining to the plea that was accepted but rather to the course of legal representation that preceded it with respect to other potential pleas and plea offers. To give further support to its contention that the instant case is in a category different from what the Court considered in Hill and Padilla, the State urges that there is no right to a plea offer or a plea bargain in any event. See Weatherford v. Bursey, 429 U.S. 545, 561 (1977). It claims Frye therefore was not deprived of any legal benefit to which he was entitled. Under this view, any wrongful or mistaken action of counsel with respect to earlier plea offers is beside the point. The State is correct to point out that Hill and Padilla concerned whether there was ineffective assistance leading to acceptance of a plea offer, a process involving a formal court appearance with the defendant and all counsel present. Before a guilty plea is entered the defendant’s understanding of the plea and its consequences can be established on the record. This affords the State substantial protection against later claims that the plea was the result of inadequate advice. At the plea entry proceedings the trial court and all counsel have the opportunity to establish on the record that the defendant understands the process that led to any offer, the advantages and disadvantages of accepting it, and the sentencing consequences or possibilities that will ensue once a conviction is entered based upon the plea. See, e.g., Fed. Rule Crim. Proc. 11; Mo. Sup. Ct. Rule 24.02 (2004). Hill and Padilla both illustrate that, nevertheless, there may be instances when claims of ineffective assistance can arise after the conviction is entered. Still, the State, and the trial court itself, have had a substantial opportunity to guard against this contingency by establishing at the plea entry proceeding that the defendant has been given proper advice or, if the advice received appears to have been inadequate, to remedy that deficiency before the plea is accepted and the conviction entered. When a plea offer has lapsed or been rejected, however, no formal court proceedings are involved. This underscores that the plea-bargaining process is often in flux, with no clear standards or timelines and with no judicial supervision of the discussions between prosecution and defense. Indeed, discussions between client and defense counsel are privileged. So the prosecution has little or no notice if something may be amiss and perhaps no capacity to intervene in any event. And, as noted, the State insists there is no right to receive a plea offer. For all these reasons, the State contends, it is unfair to subject it to the consequences of defense counsel’s inadequacies, especially when the opportunities for a full and fair trial, or, as here, for a later guilty plea albeit on less favorable terms, are preserved. The State’s contentions are neither illogical nor without some persuasive force, yet they do not suffice to overcome a simple reality. Ninety-seven percent of federal con- victions and ninety-four percent of state convictions are the result of guilty pleas. See Dept. of Justice, Bureau of Justice Statistics, Sourcebook of Criminal Justice Sta- tistics Online, Table 5.22.2009, http://www.albany.edu/ sourcebook/pdf/t5222009.pdf (all Internet materials as visited Mar. 1, 2012, and available in Clerk of Court’s case file); Dept. of Justice, Bureau of Justice Statistics, S. Rosenmerkel, M. Durose, & D. Farole, Felony Sentences in State Courts, 2006-Statistical Tables, p. 1 (NCJ226846, rev. Nov. 2010), http://bjs.ojp.usdoj.gov/content/pub/pdf/ fssc06st.pdf; Padilla, supra, at ___ (slip op., at 15) (recognizing pleas account for nearly 95% of all criminal convictions). The reality is that plea bargains have become so central to the administration of the criminal justice system that defense counsel have responsibilities in the plea bargain process, responsibilities that must be met to render the adequate assistance of counsel that the Sixth Amendment requires in the criminal process at critical stages. Because ours “is for the most part a system of pleas, not a system of trials,” Lafler, post, at 11, it is insufficient simply to point to the guarantee of a fair trial as a backstop that inoculates any errors in the pretrial process. “To a large extent . . . horse trading [between prosecutor and defense counsel] determines who goes to jail and for how long. That is what plea bargaining is. It is not some adjunct to the criminal justice system; it is the criminal justice system.” Scott & Stuntz, Plea Bargaining as Contract, 101 Yale L. J. 1909, 1912 (1992). See also Barkow, Separation of Powers and the Criminal Law, 58 Stan. L. Rev. 989, 1034 (2006) (“[Defendants] who do take their case to trial and lose receive longer sentences than even Congress or the prosecutor might think appropriate, because the longer sentences exist on the books largely for bargaining purposes. This often results in individuals who accept a plea bargain receiving shorter sentences than other individuals who are less morally culpable but take a chance and go to trial” (footnote omitted)). In today’s criminal justice system, therefore, the negotiation of a plea bargain, rather than the unfolding of a trial, is almost always the critical point for a defendant. To note the prevalence of plea bargaining is not to criticize it. The potential to conserve valuable prosecutorial resources and for defendants to admit their crimes and receive more favorable terms at sentencing means that a plea agreement can benefit both parties. In order that these benefits can be realized, however, criminal defendants require effective counsel during plea negotiations. “Anything less . . . might deny a defendant ‘effective representation by counsel at the only stage when legal aid and advice would help him.’ ” Massiah, 377 U. S., at 204 (quoting Spano v. New York, 360 U.S. 315, 326 (1959) (Douglas, J., concurring)). B The inquiry then becomes how to define the duty and responsibilities of defense counsel in the plea bargain process. This is a difficult question. “The art of negotiation is at least as nuanced as the art of trial advocacy and it presents questions farther removed from immediate judicial supervision.” Premo v. Moore, 562 U. S. ___, ___ (2011) (slip op., at 8–9). Bargaining is, by its nature, defined to a substantial degree by personal style. The alternative courses and tactics in negotiation are so individual that it may be neither prudent nor practicable to try to elaborate or define detailed standards for the proper discharge of defense counsel’s participation in the process. Cf. ibid. This case presents neither the necessity nor the occasion to define the duties of defense counsel in those respects, however. Here the question is whether defense counsel has the duty to communicate the terms of a formal offer to accept a plea on terms and conditions that may result in a lesser sentence, a conviction on lesser charges, or both. This Court now holds that, as a general rule, defense counsel has the duty to communicate formal offers from the prosecution to accept a plea on terms and conditions that may be favorable to the accused. Any exceptions to that rule need not be explored here, for the offer was a formal one with a fixed expiration date. When defense counsel allowed the offer to expire without advising the defendant or allowing him to consider it, defense counsel did not render the effective assistance the Constitution requires. Though the standard for counsel’s performance is not determined solely by reference to codified standards of professional practice, these standards can be important guides. The American Bar Association recommends defense counsel “promptly communicate and explain to the defendant all plea offers made by the prosecuting attorney,” ABA Standards for Criminal Justice, Pleas of Guilty 14–3.2(a) (3d ed. 1999), and this standard has been adopted by numerous state and federal courts over the last 30 years. See, e.g., Davie v. State, 381 S. C. 601, 608–609, 675 S.E.2d 416, 420 (2009); Cottle v. State, 733 So. 2d 963, 965–966 (Fla. 1999); Becton v. Hun, 205 W. Va. 139, 144, 516 S.E.2d 762, 767 (1999); Harris v. State, 875 S.W.2d 662, 665 (Tenn. 1994); Lloyd v. State, 258 Ga. 645, 648, 373 S.E.2d 1, 3 (1988); United States v. Rodriguez Rodriguez, 929 F.2d 747, 752 (CA1 1991) (per curiam); Pham v. United States, 317 F.3d 178, 182 (CA2 2003); United States ex rel. Caruso v. Zelinsky, 689 F.2d 435, 438 (CA3 1982); Griffin v. United States, 330 F.3d 733, 737 (CA6 2003); Johnson v. Duckworth, 793 F.2d 898, 902 (CA7 1986); United States v. Blaylock, 20 F.3d 1458, 1466 (CA9 1994); cf. Diaz v. United States, 930 F.2d 832, 834 (CA11 1991). The standard for prompt communication and consultation is also set out in state bar professional standards for attorneys. See, e.g., Fla. Rule Regulating Bar 4–1.4 (2008); Ill. Rule Prof. Conduct 1.4 (2011); Kan. Rule Prof. Conduct 1.4 (2010); Ky. Sup. Ct. Rule 3.130, Rule Prof. Conduct 1.4 (2011); Mass. Rule Prof. Conduct 1.4 (2011–2012); Mich. Rule Prof. Conduct 1.4 (2011). The prosecution and the trial courts may adopt some measures to help ensure against late, frivolous, or fabricated claims after a later, less advantageous plea offer has been accepted or after a trial leading to conviction with resulting harsh consequences. First, the fact of a formal offer means that its terms and its processing can be documented so that what took place in the negotiation process becomes more clear if some later inquiry turns on the conduct of earlier pretrial negotiations. Second, States may elect to follow rules that all offers must be in writing, again to ensure against later misunderstandings or fabricated charges. See N. J. Ct. Rule 3:9–1(b) (2012) (“Any plea offer to be made by the prosecutor shall be in writing and forwarded to the defendant’s attorney”). Third, formal offers can be made part of the record at any subsequent plea proceeding or before a trial on the merits, all to ensure that a defendant has been fully advised before those further proceedings commence. At least one State often follows a similar procedure before trial. See Brief for National Association of Criminal Defense Lawyers et al. as Amici Curiae 20 (discussing hearings in Arizona conducted pursuant to State v. Donald, 198 Ariz. 406, 10 P.3d 1193 (App. 2000)); see also N. J. Ct. Rules 3:9–1(b), (c) (requiring the prosecutor and defense counsel to discuss the case prior to the arraignment/status conference including any plea offers and to report on these discussions in open court with the defendant present); In re Alvernaz, 2 Cal. 4th 924, 938, n. 7, 830 P.2d 747, 756, n. 7 (1992) (encouraging parties to “memorialize in some fashion prior to trial (1) the fact that a plea bargain offer was made, and (2) that the defendant was advised of the offer [and] its precise terms, . . . and (3) the defendant’s response to the plea bargain offer”); Brief for Center on the Administration of Criminal Law, New York University School of Law as Amicus Curiae 25–27. Here defense counsel did not communicate the formal offers to the defendant. As a result of that deficient performance, the offers lapsed. Under Strickland, the question then becomes what, if any, prejudice resulted from the breach of duty. C To show prejudice from ineffective assistance of counsel where a plea offer has lapsed or been rejected because of counsel’s deficient performance, defendants must demonstrate a reasonable probability they would have accepted the earlier plea offer had they been afforded effective assistance of counsel. Defendants must also demonstrate a reasonable probability the plea would have been entered without the prosecution canceling it or the trial court refusing to accept it, if they had the authority to exercise that discretion under state law. To establish prejudice in this instance, it is necessary to show a reasonable probability that the end result of the criminal process would have been more favorable by reason of a plea to a lesser charge or a sentence of less prison time. Cf. Glover v. United States, 531 U.S. 198, 203 (2001) (“[A]ny amount of [additional] jail time has Sixth Amendment significance”). This application of Strickland to the instances of an uncommunicated, lapsed plea does nothing to alter the standard laid out in Hill. In cases where a defendant complains that ineffective assistance led him to accept a plea offer as opposed to proceeding to trial, the defendant will have to show “a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” Hill, 474 U. S., at 59. Hill was correctly decided and applies in the context in which it arose. Hill does not, however, provide the sole means for demonstrating prejudice arising from the deficient performance of counsel during plea negotiations. Unlike the defendant in Hill, Frye argues that with effective assistance he would have accepted an earlier plea offer (limiting his sentence to one year in prison) as opposed to entering an open plea (exposing him to a maximum sentence of four years’ imprisonment). In a case, such as this, where a defendant pleads guilty to less favorable terms and claims that ineffective assistance of counsel caused him to miss out on a more favorable earlier plea offer, Strickland’s inquiry into whether “the result of the proceeding would have been different,” 466 U. S., at 694, requires looking not at whether the defendant would have proceeded to trial absent ineffective assistance but whether he would have accepted the offer to plead pursuant to the terms earlier proposed. In order to complete a showing of Strickland prejudice, defendants who have shown a reasonable probability they would have accepted the earlier plea offer must also show that, if the prosecution had the discretion to cancel it or if the trial court had the discretion to refuse to accept it, there is a reasonable probability neither the prosecution nor the trial court would have prevented the offer from being accepted or implemented. This further showing is of particular importance because a defendant has no right to be offered a plea, see Weatherford, 429 U. S., at 561, nor a federal right that the judge accept it, Santobello v. New York, 404 U.S. 257, 262 (1971). In at least some States, including Missouri, it appears the prosecution has some discretion to cancel a plea agreement to which the defendant has agreed, see, e.g., 311 S. W. 3d, at 359 (case below); Ariz. Rule Crim. Proc. 17.4(b) (Supp. 2011). The Federal Rules, some state rules including in Missouri, and this Court’s precedents give trial courts some leeway to accept or reject plea agreements, see Fed. Rule Crim. Proc. 11(c)(3); see Mo. Sup. Ct. Rule 24.02(d)(4); Boykin v. Alabama, 395 U.S. 238, 243–244 (1969). It can be assumed that in most jurisdictions prosecutors and judges are familiar with the boundaries of acceptable plea bargains and sentences. So in most instances it should not be difficult to make an objective assessment as to whether or not a particular fact or intervening circumstance would suffice, in the normal course, to cause prosecutorial withdrawal or judicial nonapproval of a plea bargain. The determination that there is or is not a reasonable probability that the outcome of the proceeding would have been different absent counsel’s errors can be conducted within that framework. III These standards must be applied to the instant case. As regards the deficient performance prong of Strickland, the Court of Appeals found the “record is void of any evidence of any effort by trial counsel to communicate the [formal] Offer to Frye during the Offer window, let alone any evidence that Frye’s conduct interfered with trial counsel’s ability to do so.” 311 S. W. 3d, at 356. On this record, it is evident that Frye’s attorney did not make a meaningful attempt to inform the defendant of a written plea offer before the offer expired. See supra, at 2. The Missouri Court of Appeals was correct that “counsel’s representation fell below an objective standard of reasonableness.” Strickland, supra, at 688. The Court of Appeals erred, however, in articulating the precise standard for prejudice in this context. As noted, a defendant in Frye’s position must show not only a reasonable probability that he would have accepted the lapsed plea but also a reasonable probability that the prosecution would have adhered to the agreement and that it would have been accepted by the trial court. Frye can show he would have accepted the offer, but there is strong reason to doubt the prosecution and the trial court would have permitted the plea bargain to become final. There appears to be a reasonable probability Frye would have accepted the prosecutor’s original offer of a plea bargain if the offer had been communicated to him, because he pleaded guilty to a more serious charge, with no promise of a sentencing recommendation from the prosecutor. It may be that in some cases defendants must show more than just a guilty plea to a charge or sentence harsher than the original offer. For example, revelations between plea offers about the strength of the prosecution’s case may make a late decision to plead guilty insufficient to demonstrate, without further evidence, that the defendant would have pleaded guilty to an earlier, more generous plea offer if his counsel had reported it to him. Here, however, that is not the case. The Court of Appeals did not err in finding Frye’s acceptance of the less favorable plea offer indicated that he would have accepted the earlier (and more favorable) offer had he been apprised of it; and there is no need to address here the showings that might be required in other cases. The Court of Appeals failed, however, to require Frye to show that the first plea offer, if accepted by Frye, would have been adhered to by the prosecution and accepted by the trial court. Whether the prosecution and trial court are required to do so is a matter of state law, and it is not the place of this Court to settle those matters. The Court has established the minimum requirements of the Sixth Amendment as interpreted in Strickland, and States have the discretion to add procedural protections under state law if they choose. A State may choose to preclude the prosecution from withdrawing a plea offer once it has been accepted or perhaps to preclude a trial court from rejecting a plea bargain. In Missouri, it appears “a plea offer once accepted by the defendant can be withdrawn without re- course” by the prosecution. 311 S. W. 3d, at 359. The ex- tent of the trial court’s discretion in Missouri to reject a plea agreement appears to be in some doubt. Compare id., at 360, with Mo. Sup. Ct. Rule 24.02(d)(4). We remand for the Missouri Court of Appeals to consider these state-law questions, because they bear on the federal question of Strickland prejudice. If, as the Missouri court stated here, the prosecutor could have canceled the plea agreement, and if Frye fails to show a reasonable probability the prosecutor would have adhered to the agreement, there is no Strickland prejudice. Likewise, if the trial court could have refused to accept the plea agreement, and if Frye fails to show a reasonable probability the trial court would have accepted the plea, there is no Strickland prejudice. In this case, given Frye’s new offense for driving without a license on December 30, 2007, there is reason to doubt that the prosecution would have adhered to the agreement or that the trial court would have accepted it at the January 4, 2008, hearing, unless they were required by state law to do so. It is appropriate to allow the Missouri Court of Appeals to address this question in the first instance. The judgment of the Missouri Court of Appeals is vacated, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered.

566.US.2011_11-88

While visiting the West Bank, Azzam Rahim, a naturalized United States citizen, allegedly was arrested by Palestinian Authority intelligence officers, imprisoned, tortured, and ultimately killed. Rahim’s relatives, petitioners here, sued the Palestinian Authority and the Palestinian Liberation Organization under the Torture Victim Protection Act of 1991 (TVPA), which authorizes a cause of action against “[a]n individual” for acts of torture and extrajudicial killing committed under authority or color of law of any foreign nation. 106Stat. 73, note following 28 U. S. C. §1350. The District Court dismissed the suit, concluding, as relevant here, that the TVPA’s authorization of suit against “[a]n individual” extended liability only to natural persons. The United States Court of Appeals for the District of Columbia Circuit affirmed. Held: As used in the TVPA, the term “individual” encompasses only natural persons. Consequently, the Act does not impose liability against organizations. Pp. 2–11. (a) The ordinary, everyday meaning of “individual” refers to a human being, not an organization, and Congress in the normal course does not employ the word any differently. The Dictionary Act defines “person” to include certain artificial entities “as well as individuals,” 1 U. S. C. §1, thereby marking “individual” as distinct from artificial entities. Federal statutes routinely distinguish between an “individual” and an organizational entity. See, e.g., 7 U. S. C. §§92(k), 511. And the very Congress that passed the TVPA defined “person” in a separate Act to include “any individual or entity.” 18 U. S. C. §2331(3). Pp. 2–5. (b) Before a word will be assumed to have a meaning broader than or different from its ordinary meaning, Congress must give some indication that it intended such a result. There are no such indications in the TVPA. To the contrary, the statutory context confirms that Congress in the Act created a cause of action against natural persons alone. The Act’s liability provision uses the word “individual” five times in the same sentence: once to refer to the perpetrator and four times to refer to the victim. See TVPA §2(a). Since only a natural person can be a victim of torture or extrajudicial killing, it is difficult to conclude that Congress used “individual” four times in the same sentence to refer to a natural person and once to refer to a natural person and any nonsovereign organization. In addition, the TVPA holds perpetrators liable for extrajudicial killing to “any person who may be a claimant in an action for wrongful death.” See TVPA §2(a)(2). “Persons” often has a broader meaning in the law than “individual,” and frequently includes non-natural persons. Construing “individual” in the Act to encompass solely natural persons credits Congress’ use of disparate terms. Pp. 5–6. (c) Petitioners’ counterarguments are unpersuasive. Pp. 6–11. (1) Petitioners dispute that the plain text of the TVPA requires this Court’s result. First, they rely on definitions that frame “individual” in nonhuman terms, emphasizing the idea of “oneness,” but these definitions make for an awkward fit in the context of the TVPA. Next they claim that federal tort statutes uniformly provide for liability against organizations, a convention they maintain is common to the legal systems of other nations. But while “Congress is understood to legislate against a background of common-law adjudicatory principles,” Astoria Fed. Sav. & Loan Assn. v. Solimino, 501 U.S. 104, 108, Congress plainly evinced its intent in the TVPA not to subject organizations to liability. Petitioners next argue that the TVPA’s scope of liability should be construed to conform with other federal statutes they claim provide civil remedies to victims of torture or extrajudicial killing. But none of the statutes petitioners cite employs the term “individual,” as the TVPA, to describe the covered defendant. Finally, although petitioners rightly note that the TVPA contemplates liability against officers who do not personally execute the torture or extrajudicial killing, it does not follow that the Act embraces liability against nonsovereign organizations. Pp. 6–8. (2) Petitioners also contend that legislative history supports their broad reading of “individual,” but “reliance on legislative history is unnecessary in light of the statute’s unambiguous language.” Milavetz, Gallop & Milavetz, P. A. v. United States, 559 U. S. ___, ___. In any event, the history supports this Court’s interpretation. Pp. 8–10. (3) Finally, petitioners argue that precluding organizational liability may foreclose effective remedies for victims and their relatives. This purposive argument simply cannot overcome the force of the plain text. Moreover, Congress appeared well aware of the limited nature of the cause of action it established in the TVPA. Pp. 10–11. 634 F.3d 604, affirmed. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, Ginsburg, Breyer, Alito, and Kagan, JJ., joined, and in which Scalia, J., joined except as to Part III–B. Breyer, J., filed a concurring opinion.

o legislation pursues its purposes at all costs,” Rodriguez v. United States, 480 U.S. 522, 525–526 (1987) (per curiam), and petitioners’ purposive argument simply cannot overcome the force of the plain text. We add only that Congress appeared well aware of the limited nature of the cause of action it established in the Act. See, e.g., 138 Cong. Rec. 4177 (1992) (remarks of Sen. Simpson) (noting that “as a practical matter, this legislation will result in a very small number of cases”); 137 Cong. Rec. 2671 (1991) (remarks of Sen. Specter) (“Let me emphasize that the bill is a limited measure. It is estimated that only a few of these lawsuits will ever be brought”). * * * The text of the TVPA convinces us that Congress did not extend liability to organizations, sovereign or not. There are no doubt valid arguments for such an extension. But Congress has seen fit to proceed in more modest steps in the Act, and it is not the province of this Branch to do otherwise. The judgment of the United States Court of Appeals for the District of Columbia Circuit is affirmed. It is so ordered. Notes 1 * Justice Scalia joins this opinion except as to Part III–B. 2 Respondents also argued before the District Court that the TVPA’s requirement that acts be committed under authority or color of law ofa foreign nation was not met. Neither the District Court nor Court of Appeals addressed the argument, and we offer no opinion on its merits. 3 Compare Aziz v. Alcolac, Inc., 658 F.3d 388 (CA4 2011) (TVPA excludes corporate defendants from liability); Mohamad v. Rajoub, 634 F.3d 604 (CADC 2011) (TVPA liability limited to natural persons); Bowoto v. Chevron Corp., 621 F.3d 1116 (CA9 2010) (same as Aziz), with Sinaltrainal v. Coca Cola Co., 578 F.3d 1252, 1264, n. 13 (CA11 2009) (TVPA liability extends to corporate defendants). 4 The parties debate whether estates, or other nonnatural persons, in fact may be claimants in a wrongful-death action. We think the debate largely immaterial. Regardless of whether jurisdictions today allow for such actions, Congress’ use of the broader term evidences an intent to accommodate that possibility. 5 Petitioners’ separate contention that the TVPA must be construedin light of international agreements prohibiting torture and extrajudicial killing fails for similar reasons. Whatever the scope of those agree-ments, the TVPA does not define “individual” by reference to them,and principles they elucidate cannot overcome the statute’s text. The same is true of petitioners’ suggestion that Congress in the TVPA imported a “specialized usage” of the word “individual” in international law. Brief for Petitioners 6. There is no indication in the text of the statute or legislative history that Congress knew of any such specialized usage of the term, much less intended to import it into the Act.

567.US.2011_11-393

In 2010, Congress enacted the Patient Protection and Affordable Care Act in order to increase the number of Americans covered by health insurance and decrease the cost of health care. One key provision is the individual mandate, which requires most Americans to maintain “minimum essential” health insurance coverage. 26 U. S. C. §5000A. For individuals who are not exempt, and who do not receive health insurance through an employer or government program, the means of satisfying the requirement is to purchase insurance from a private company. Beginning in 2014, those who do not comply with the mandate must make a “[s]hared responsibility payment” to the Federal Government. §5000A(b)(1). The Act provides that this “penalty” will be paid to the Internal Revenue Service with an individual’s taxes, and “shall be assessed and collected in the same manner” as tax penalties. §§5000A(c), (g)(1). Another key provision of the Act is the Medicaid expansion. The current Medicaid program offers federal funding to States to assist pregnant women, children, needy families, the blind, the elderly, and the disabled in obtaining medical care. 42 U. S. C. §1396d(a). The Affordable Care Act expands the scope of the Medicaid program and increases the number of individuals the States must cover. For example, the Act requires state programs to provide Medicaid coverage by 2014 to adults with incomes up to 133 percent of the federal poverty level, whereas many States now cover adults with children only if their income is considerably lower, and do not cover childless adults at all. §1396a(a)(10)(A)(i)(VIII). The Act increases federal funding to cover the States’ costs in expanding Medicaid coverage. §1396d(y)(1). But if a State does not comply with the Act’s new coverage requirements, it may lose not only the federal funding for those requirements, but all of its federal Medicaid funds. §1396c. Twenty-six States, several individuals, and the National Federation of Independent Business brought suit in Federal District Court, challenging the constitutionality of the individual mandate and the Medicaid expansion. The Court of Appeals for the Eleventh Circuit upheld the Medicaid expansion as a valid exercise of Congress’s spending power, but concluded that Congress lacked authority to enact the individual mandate. Finding the mandate severable from the Act’s other provisions, the Eleventh Circuit left the rest of the Act intact. Held: The judgment is affirmed in part and reversed in part. 648 F.3d 1235, affirmed in part and reversed in part. 1. Chief Justice Roberts delivered the opinion of the Court with respect to Part II, concluding that the Anti-Injunction Act does not bar this suit. The Anti-Injunction Act provides that “no suit for the purpose of restraining the assessment or collection of any tax shall be maintained in any court by any person,” 26 U. S. C. §7421(a), so that those subject to a tax must first pay it and then sue for a refund. The present challenge seeks to restrain the collection of the shared responsibility payment from those who do not comply with the individual mandate. But Congress did not intend the payment to be treated as a “tax” for purposes of the Anti-Injunction Act. The Affordable Care Act describes the payment as a “penalty,” not a “tax.” That label cannot control whether the payment is a tax for purposes of the Constitution, but it does determine the application of the Anti-Injunction Act. The Anti-Injunction Act therefore does not bar this suit. Pp. 11–15. 2. Chief Justice Roberts concluded in Part III–A that the individual mandate is not a valid exercise of Congress’s power under the Commerce Clause and the Necessary and Proper Clause. Pp. 16–30. (a) The Constitution grants Congress the power to “regulate Commerce.” Art. I, §8, cl. 3 (emphasis added). The power to regulate commerce presupposes the existence of commercial activity to be regulated. This Court’s precedent reflects this understanding: As expansive as this Court’s cases construing the scope of the commerce power have been, they uniformly describe the power as reaching “activity.” E.g., United States v. Lopez, 514 U.S. 549, 560. The individual mandate, however, does not regulate existing commercial activity. It instead compels individuals to become active in commerce by purchasing a product, on the ground that their failure to do so affects interstate commerce. Construing the Commerce Clause to permit Congress to regulate individuals precisely because they are doing nothing would open a new and potentially vast domain to congressional authority. Congress already possesses expansive power to regulate what people do. Upholding the Affordable Care Act under the Commerce Clause would give Congress the same license to regulate what people do not do. The Framers knew the difference between doing something and doing nothing. They gave Congress the power to regulate commerce, not to compel it. Ignoring that distinction would undermine the principle that the Federal Government is a government of limited and enumerated powers. The individual mandate thus cannot be sustained under Congress’s power to “regulate Commerce.” Pp. 16–27. (b) Nor can the individual mandate be sustained under the Necessary and Proper Clause as an integral part of the Affordable Care Act’s other reforms. Each of this Court’s prior cases upholding laws under that Clause involved exercises of authority derivative of, and in service to, a granted power. E.g., United States v. Comstock, 560 U. S. ___. The individual mandate, by contrast, vests Congress with the extraordinary ability to create the necessary predicate to the exercise of an enumerated power and draw within its regulatory scope those who would otherwise be outside of it. Even if the individual mandate is “necessary” to the Affordable Care Act’s other reforms, such an expansion of federal power is not a “proper” means for making those reforms effective. Pp. 27–30. 3. Chief Justice Roberts concluded in Part III–B that the individual mandate must be construed as imposing a tax on those who do not have health insurance, if such a construction is reasonable. The most straightforward reading of the individual mandate is that it commands individuals to purchase insurance. But, for the reasons explained, the Commerce Clause does not give Congress that power. It is therefore necessary to turn to the Government’s alternative argument: that the mandate may be upheld as within Congress’s power to “lay and collect Taxes.” Art. I, §8, cl. 1. In pressing its taxing power argument, the Government asks the Court to view the mandate as imposing a tax on those who do not buy that product. Because “every reasonable construction must be resorted to, in order to save a statute from unconstitutionality,” Hooper v. California, 155 U.S. 648, 657, the question is whether it is “fairly possible” to interpret the mandate as imposing such a tax, Crowell v. Benson, 285 U.S. 22, 62. Pp. 31–32. 4. Chief Justice Roberts delivered the opinion of the Court with respect to Part III–C, concluding that the individual mandate may be upheld as within Congress’s power under the Taxing Clause. Pp. 33–44. (a) The Affordable Care Act describes the “[s]hared responsibility payment” as a “penalty,” not a “tax.” That label is fatal to the application of the Anti-Injunction Act. It does not, however, control whether an exaction is within Congress’s power to tax. In answering that constitutional question, this Court follows a functional approach, “[d]isregarding the designation of the exaction, and viewing its substance and application.” United States v. Constantine, 296 U.S. 287, 294. Pp. 33–35. (b) Such an analysis suggests that the shared responsibility payment may for constitutional purposes be considered a tax. The payment is not so high that there is really no choice but to buy health insurance; the payment is not limited to willful violations, as penalties for unlawful acts often are; and the payment is collected solely by the IRS through the normal means of taxation. Cf. Bailey v. Drexel Furniture Co., 259 U.S. 20, 36–37. None of this is to say that payment is not intended to induce the purchase of health insurance. But the mandate need not be read to declare that failing to do so is unlawful. Neither the Affordable Care Act nor any other law attaches negative legal consequences to not buying health insurance, beyond requiring a payment to the IRS. And Congress’s choice of language—stating that individuals “shall” obtain insurance or pay a “penalty”—does not require reading §5000A as punishing unlawful conduct. It may also be read as imposing a tax on those who go without insurance. See New York v. United States, 505 U.S. 144, 169–174. Pp. 35–40. (c) Even if the mandate may reasonably be characterized as a tax, it must still comply with the Direct Tax Clause, which provides: “No Capitation, or other direct, Tax shall be laid, unless in Proportion to the Census or Enumeration herein before directed to be taken.” Art. I, §9, cl. 4. A tax on going without health insurance is not like a capitation or other direct tax under this Court’s precedents. It therefore need not be apportioned so that each State pays in proportion to its population. Pp. 40–41. 5. Chief Justice Roberts, joined by Justice Breyer and Justice Kagan, concluded in Part IV that the Medicaid expansion violates the Constitution by threatening States with the loss of their existing Medicaid funding if they decline to comply with the expansion. Pp. 45–58. (a) The Spending Clause grants Congress the power “to pay the Debts and provide for the . . . general Welfare of the United States.” Art. I, §8, cl. 1. Congress may use this power to establish cooperative state-federal Spending Clause programs. The legitimacy of Spending Clause legislation, however, depends on whether a State voluntarily and knowingly accepts the terms of such programs. Pennhurst State School and Hospital v. Halderman, 451 U.S. 1, 17. “[T]he Constitution simply does not give Congress the authority to require the States to regulate.” New York v. United States, 505 U.S. 144, 178. When Congress threatens to terminate other grants as a means of pressuring the States to accept a Spending Clause program, the legislation runs counter to this Nation’s system of federalism. Cf. South Dakota v. Dole, 483 U.S. 203, 211. Pp. 45–51. (b) Section 1396c gives the Secretary of Health and Human Services the authority to penalize States that choose not to participate in the Medicaid expansion by taking away their existing Medicaid funding. 42 U. S. C. §1396c. The threatened loss of over 10 percent of a State’s overall budget is economic dragooning that leaves the States with no real option but to acquiesce in the Medicaid expansion. The Government claims that the expansion is properly viewed as only a modification of the existing program, and that this modification is permissible because Congress reserved the “right to alter, amend, or repeal any provision” of Medicaid. §1304. But the expansion accomplishes a shift in kind, not merely degree. The original program was designed to cover medical services for particular categories of vulnerable individuals. Under the Affordable Care Act, Medicaid is transformed into a program to meet the health care needs of the entire nonelderly population with income below 133 percent of the poverty level. A State could hardly anticipate that Congress’s reservation of the right to “alter” or “amend” the Medicaid program included the power to transform it so dramatically. The Medicaid expansion thus violates the Constitution by threatening States with the loss of their existing Medicaid funding if they decline to comply with the expansion. Pp. 51–55. (c) The constitutional violation is fully remedied by precluding the Secretary from applying §1396c to withdraw existing Medicaid funds for failure to comply with the requirements set out in the expansion. See §1303. The other provisions of the Affordable Care Act are not affected. Congress would have wanted the rest of the Act to stand, had it known that States would have a genuine choice whether to participate in the Medicaid expansion. Pp. 55–58. 6. Justice Ginsburg, joined by Justice Sotomayor, is of the view that the Spending Clause does not preclude the Secretary from withholding Medicaid funds based on a State’s refusal to comply with the expanded Medicaid program. But given the majority view, she agrees with The Chief Justice’s conclusion in Part IV–B that the Medicaid Act’s severability clause, 42 U. S. C. §1303, determines the appropriate remedy. Because The Chief Justice finds the withholding—not the granting—of federal funds incompatible with the Spending Clause, Congress’ extension of Medicaid remains available to any State that affirms its willingness to participate. Even absent §1303’s command, the Court would have no warrant to invalidate the funding offered by the Medicaid expansion, and surely no basis to tear down the ACA in its entirety. When a court confronts an unconstitutional statute, its endeavor must be to conserve, not destroy, the legislation. See, e.g., Ayotte v. Planned Parenthood of Northern New Eng., 546 U.S. 320, 328–330. Pp. 60–61. Roberts, C. J., announced the judgment of the Court and delivered the opinion of the Court with respect to Parts I, II, and III–C, in which Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined; an opinion with respect to Part IV, in which Breyer and Kagan, JJ., joined; and an opinion with respect to Parts III–A, III–B, and III–D. Ginsburg, J., filed an opinion concurring in part, concurring in the judgment in part, and dissenting in part, in which Sotomayor, J., joined, and in which Breyer and Kagan, JJ., joined as to Parts I, II, III, and IV. Scalia, Kennedy, Thomas, and Alito, JJ., filed a dissenting opinion. Thomas, J., filed a dissenting opinion. Notes 1 Together with No. 11–398, Department of Health and Human Services et al. v. Florida et al., and No. 11–400, Florida et al. v. Department of Health and Human Services et al., also on certiorari to the same court.

of the Court with respect to Parts I, II, and III–C, an opinion with respect to Part IV, in which Justice Breyer and Justice Kagan join, and an opinion with respect to Parts III–A, III–B, and III–D. Today we resolve constitutional challenges to two provisions of the Patient Protection and Affordable Care Act of 2010: the individual mandate, which requires individuals to purchase a health insurance policy providing a minimum level of coverage; and the Medicaid expansion, which gives funds to the States on the condition that they provide specified health care to all citizens whose income falls below a certain threshold. We do not consider whether the Act embodies sound policies. That judgment is entrusted to the Nation’s elected leaders. We ask only whether Congress has the power under the Constitution to enact the challenged provisions. In our federal system, the National Government possesses only limited powers; the States and the people retain the remainder. Nearly two centuries ago, Chief Justice Marshall observed that “the question respecting the extent of the powers actually granted” to the Federal Government “is perpetually arising, and will probably continue to arise, as long as our system shall exist.” McCulloch v. Maryland, 4 Wheat. 316, 405 (1819). In this case we must again determine whether the Constitution grants Congress powers it now asserts, but which many States and individuals believe it does not possess. Resolving this controversy requires us to examine both the limits of the Government’s power, and our own limited role in policing those boundaries. The Federal Government “is acknowledged by all to be one of enumerated powers.” Ibid. That is, rather than granting general authority to perform all the conceiv-able functions of government, the Constitution lists, or enumerates, the Federal Government’s powers. Congress may, for example, “coin Money,” “establish Post Offices,” and “raise and support Armies.” Art. I, §8, cls. 5, 7, 12. The enumeration of powers is also a limitation of pow- ers, because “[t]he enumeration presupposes something not enumerated.” Gibbons v. Ogden, 9 Wheat. 1, 195 (1824). The Constitution’s express conferral of some powers makes clear that it does not grant others. And the Federal Government “can exercise only the powers granted to it.” McCulloch, supra, at 405. Today, the restrictions on government power foremost in many Americans’ minds are likely to be affirmative pro-hibitions, such as contained in the Bill of Rights. These affirmative prohibitions come into play, however, only where the Government possesses authority to act in the first place. If no enumerated power authorizes Congress to pass a certain law, that law may not be enacted, even if it would not violate any of the express prohibitions in the Bill of Rights or elsewhere in the Constitution. Indeed, the Constitution did not initially include a Bill of Rights at least partly because the Framers felt the enu-meration of powers sufficed to restrain the Government. As Alexander Hamilton put it, “the Constitution is itself, in every rational sense, and to every useful purpose, a bill of rights.” The Federalist No. 84, p. 515 (C. Ros-siter ed. 1961). And when the Bill of Rights was ratified, it made express what the enumeration of powers necessarily implied: “The powers not delegated to the United States by the Constitution . . . are reserved to the States respectively, or to the people.” U. S. Const., Amdt. 10. The Federal Government has expanded dramatically over the past two centuries, but it still must show that a constitutional grant of power authorizes each of its actions. See, e.g., United States v. Comstock, 560 U. S. ___ (2010). The same does not apply to the States, because the Con-stitution is not the source of their power. The Consti-tution may restrict state governments—as it does, for example, by forbidding them to deny any person the equal protection of the laws. But where such prohibitions do not apply, state governments do not need constitutional au-thorization to act. The States thus can and do perform many of the vital functions of modern government—punishing street crime, running public schools, and zoning property for development, to name but a few—even though the Constitution’s text does not authorize any government to do so. Our cases refer to this general power of governing, possessed by the States but not by the Federal Government, as the “police power.” See, e.g., United States v. Morrison, 529 U.S. 598, 618–619 (2000). “State sovereignty is not just an end in itself: Rather, federalism secures to citizens the liberties that derive from the diffusion of sovereign power.” New York v. United States, 505 U.S. 144, 181 (1992) (internal quotation marks omitted). Because the police power is controlled by 50 different States instead of one national sovereign, the facets of governing that touch on citizens’ daily lives are normally administered by smaller governments closer to the governed. The Framers thus ensured that powers which “in the ordinary course of affairs, concern the lives, liberties, and properties of the people” were held by governments more local and more accountable than a dis- tant federal bureaucracy. The Federalist No. 45, at 293 (J. Madison). The independent power of the States also serves as a check on the power of the Federal Government: “By denying any one government complete jurisdiction over all the concerns of public life, federalism protects the liberty of the individual from arbitrary power.” Bond v. United States, 564 U. S. ___, ___ (2011) (slip op., at 9–10). This case concerns two powers that the Constitution does grant the Federal Government, but which must be read carefully to avoid creating a general federal authority akin to the police power. The Constitution authorizes Congress to “regulate Commerce with foreign Nations, and among the several States, and with the Indian Tribes.” Art. I, §8, cl. 3. Our precedents read that to mean that Congress may regulate “the channels of interstate commerce,” “persons or things in interstate commerce,” and “those activities that substantially affect interstate commerce.” Morrison, supra, at 609 (internal quotation marks omitted). The power over activities that substantially affect interstate commerce can be expansive. That power has been held to authorize federal regulation of such seem-ingly local matters as a farmer’s decision to grow wheat for himself and his livestock, and a loan shark’s extor-tionate collections from a neighborhood butcher shop. See Wickard v. Filburn, 317 U.S. 111 (1942); Perez v. United States, 402 U.S. 146 (1971). Congress may also “lay and collect Taxes, Duties, Imposts and Excises, to pay the Debts and provide for the common Defence and general Welfare of the United States.” U. S. Const., Art. I, §8, cl. 1. Put simply, Congress may tax and spend. This grant gives the Federal Government considerable influence even in areas where it cannot directly regulate. The Federal Government may enact a tax on an activity that it cannot authorize, forbid, or otherwise control. See, e.g., License Tax Cases, 5 Wall. 462, 471 (1867). And in exercising its spending power, Congress may offer funds to the States, and may condition those offers on compliance with specified conditions. See, e.g., College Savings Bank v. Florida Prepaid Postsecondary Ed. Expense Bd., 527 U.S. 666, 686 (1999). These offers may well induce the States to adopt policies that the Federal Government itself could not impose. See, e.g., South Dakota v. Dole, 483 U.S. 203, 205–206 (1987) (conditioning federal highway funds on States raising their drinking age to 21). The reach of the Federal Government’s enumerated powers is broader still because the Constitution authorizes Congress to “make all Laws which shall be necessary and proper for carrying into Execution the foregoing Powers.” Art. I, §8, cl. 18. We have long read this provision to give Congress great latitude in exercising its powers: “Let the end be legitimate, let it be within the scope of the constitution, and all means which are appropriate, which are plainly adapted to that end, which are not prohibited, but consist with the letter and spirit of the constitution, are constitutional.” McCulloch, 4 Wheat., at 421. Our permissive reading of these powers is explained in part by a general reticence to invalidate the acts of the Nation’s elected leaders. “Proper respect for a co-ordinate branch of the government” requires that we strike down an Act of Congress only if “the lack of constitutional authority to pass [the] act in question is clearly demonstrated.” United States v. Harris, 106 U.S. 629, 635 (1883). Members of this Court are vested with the authority to interpret the law; we possess neither the expertise nor the prerogative to make policy judgments. Those decisions are entrusted to our Nation’s elected leaders, who can be thrown out of office if the people disagree with them. It is not our job to protect the people from the consequences of their political choices. Our deference in matters of policy cannot, however, become abdication in matters of law. “The powers of the legislature are defined and limited; and that those lim- its may not be mistaken, or forgotten, the constitution is written.” Marbury v. Madison, 1 Cranch 137, 176 (1803). Our respect for Congress’s policy judgments thus can never extend so far as to disavow restraints on federal power that the Constitution carefully constructed. “The peculiar circumstances of the moment may render a measure more or less wise, but cannot render it more or less constitutional.” Chief Justice John Marshall, A Friend of the Constitution No. V, Alexandria Gazette, July 5, 1819, in John Marshall’s Defense of McCulloch v. Maryland 190–191 (G. Gunther ed. 1969). And there can be no question that it is the responsibility of this Court to enforce the limits on federal power by striking down acts of Congress that transgress those limits. Marbury v. Madison, supra, at 175–176. The questions before us must be considered against the background of these basic principles. I In 2010, Congress enacted the Patient Protection and Affordable Care Act, 124Stat. 119. The Act aims to increase the number of Americans covered by health in-surance and decrease the cost of health care. The Act’s 10 titles stretch over 900 pages and contain hundreds of provisions. This case concerns constitutional challenges to two key provisions, commonly referred to as the individual mandate and the Medicaid expansion. The individual mandate requires most Americans to maintain “minimum essential” health insurance coverage. 26 U. S. C. §5000A. The mandate does not apply to some individuals, such as prisoners and undocumented aliens. §5000A(d). Many individuals will receive the required cov-erage through their employer, or from a government program such as Medicaid or Medicare. See §5000A(f). But for individuals who are not exempt and do not receive health insurance through a third party, the means of satisfying the requirement is to purchase insurance from a private company. Beginning in 2014, those who do not comply with the mandate must make a “[s]hared responsibility payment” to the Federal Government. §5000A(b)(1). That payment, which the Act describes as a “penalty,” is calculated as a percentage of household income, subject to a floor based on a specified dollar amount and a ceiling based on the average annual premium the individual would have to pay for qualifying private health insurance. §5000A(c). In 2016, for example, the penalty will be 2.5 percent of an individ-ual’s household income, but no less than $695 and no more than the average yearly premium for insurance that covers 60 percent of the cost of 10 specified services (e.g., prescription drugs and hospitalization). Ibid.; 42 U. S. C. §18022. The Act provides that the penalty will be paid to the Internal Revenue Service with an individual’s taxes, and “shall be assessed and collected in the same manner” as tax penalties, such as the penalty for claiming too large an income tax refund. 26 U. S. C. §5000A(g)(1). The Act, however, bars the IRS from using several of its normal enforcement tools, such as criminal prosecutions and levies. §5000A(g)(2). And some individuals who are subject to the mandate are nonetheless exempt from the penalty—for example, those with income below a certain threshold and members of Indian tribes. §5000A(e). On the day the President signed the Act into law, Florida and 12 other States filed a complaint in the Federal District Court for the Northern District of Florida. Those plaintiffs—who are both respondents and petitioners here, depending on the issue—were subsequently joined by 13 more States, several individuals, and the National Fed-eration of Independent Business. The plaintiffs alleged, among other things, that the individual mandate provisions of the Act exceeded Congress’s powers under Article I of the Constitution. The District Court agreed, holding that Congress lacked constitutional power to enact the individual mandate. 780 F. Supp. 2d 1256 (ND Fla. 2011). The District Court determined that the individual mandate could not be severed from the remainder of the Act, and therefore struck down the Act in its entirety. Id., at 1305–1306. The Court of Appeals for the Eleventh Circuit affirmed in part and reversed in part. The court affirmed the District Court’s holding that the individual mandate exceeds Congress’s power. 648 F.3d 1235 (2011). The panel unanimously agreed that the individual mandate did not impose a tax, and thus could not be authorized by Congress’s power to “lay and collect Taxes.” U. S. Const., Art. I, §8, cl. 1. A majority also held that the individual mandate was not supported by Congress’s power to “regulate Commerce . . . among the several States.” Id., cl. 3. According to the majority, the Commerce Clause does not empower the Federal Government to order individuals to engage in commerce, and the Government’s efforts to cast the individual mandate in a different light were unpersuasive. Judge Marcus dissented, reasoning that the individual mandate regulates economic activity that has a clear effect on interstate commerce. Having held the individual mandate to be unconstitutional, the majority examined whether that provision could be severed from the remainder of the Act. The ma-jority determined that, contrary to the District Court’s view, it could. The court thus struck down only the individual mandate, leaving the Act’s other provisions intact. 648 F. 3d, at 1328. Other Courts of Appeals have also heard challenges to the individual mandate. The Sixth Circuit and the D. C. Circuit upheld the mandate as a valid exercise of Congress’s commerce power. See Thomas More Law Center v. Obama, 651 F.3d 529 (CA6 2011); Seven-Sky v. Holder, 661 F.3d 1 (CADC 2011). The Fourth Circuit determined that the Anti-Injunction Act prevents courts from considering the merits of that question. See Liberty Univ., Inc. v. Geithner, 671 F.3d 391 (2011). That statute bars suits “for the purpose of restraining the assessment or collection of any tax.” 26 U. S. C. §7421(a). A majority of the Fourth Circuit panel reasoned that the individual mandate’s penalty is a tax within the meaning of the Anti-Injunction Act, because it is a financial assessment collected by the IRS through the normal means of taxation. The majority therefore determined that the plaintiffs could not challenge the individual mandate until after they paid the penalty.[1] The second provision of the Affordable Care Act directly challenged here is the Medicaid expansion. Enacted in 1965, Medicaid offers federal funding to States to assist pregnant women, children, needy families, the blind, the elderly, and the disabled in obtaining medical care. See 42 U. S. C. §1396a(a)(10). In order to receive that funding, States must comply with federal criteria governing matters such as who receives care and what services are provided at what cost. By 1982 every State had chosen to participate in Medicaid. Federal funds received through the Medicaid program have become a substantial part of state budgets, now constituting over 10 percent of most States’ total revenue. The Affordable Care Act expands the scope of the Medicaid program and increases the number of individuals the States must cover. For example, the Act requires state programs to provide Medicaid coverage to adults with incomes up to 133 percent of the federal poverty level, whereas many States now cover adults with children only if their income is considerably lower, and do not cover childless adults at all. See §1396a(a)(10)(A)(i)(VIII). The Act increases federal funding to cover the States’ costs in expanding Medicaid coverage, although States will bear a portion of the costs on their own. §1396d(y)(1). If a State does not comply with the Act’s new coverage requirements, it may lose not only the federal funding for those requirements, but all of its federal Medicaid funds. See §1396c. Along with their challenge to the individual mandate, the state plaintiffs in the Eleventh Circuit argued that the Medicaid expansion exceeds Congress’s constitutional powers. The Court of Appeals unanimously held that the Medicaid expansion is a valid exercise of Congress’s power under the Spending Clause. U. S. Const., Art. I, §8, cl. 1. And the court rejected the States’ claim that the threatened loss of all federal Medicaid funding violates the Tenth Amendment by coercing them into complying with the Medicaid expansion. 648 F. 3d, at 1264, 1268. We granted certiorari to review the judgment of the Court of Appeals for the Eleventh Circuit with respect to both the individual mandate and the Medicaid expansion. 565 U. S. ___ (2011). Because no party supports the Eleventh Circuit’s holding that the individual mandate can be completely severed from the remainder of the Affordable Care Act, we appointed an amicus curiae to defend that aspect of the judgment below. And because there is a reasonable argument that the Anti-Injunction Act deprives us of jurisdiction to hear challenges to the individ-ual mandate, but no party supports that proposition, we appointed an amicus curiae to advance it.[2] II Before turning to the merits, we need to be sure we have the authority to do so. The Anti-Injunction Act provides that “no suit for the purpose of restraining the assessment or collection of any tax shall be maintained in any court by any person, whether or not such person is the per- son against whom such tax was assessed.” 26 U. S. C. §7421(a). This statute protects the Government’s ability to collect a consistent stream of revenue, by barring litigation to enjoin or otherwise obstruct the collection of taxes. Because of the Anti-Injunction Act, taxes can ordinarily be challenged only after they are paid, by suing for a refund. See Enochs v. Williams Packing & Nav. Co., 370 U.S. 1, 7–8 (1962). The penalty for not complying with the Affordable Care Act’s individual mandate first becomes enforceable in 2014. The present challenge to the mandate thus seeks to restrain the penalty’s future collection. Amicus contends that the Internal Revenue Code treats the penalty as a tax, and that the Anti-Injunction Act therefore bars this suit. The text of the pertinent statutes suggests otherwise. The Anti-Injunction Act applies to suits “for the purpose of restraining the assessment or collection of any tax.” §7421(a) (emphasis added). Congress, however, chose to describe the “[s]hared responsibility payment” imposed on those who forgo health insurance not as a “tax,” but as a “penalty.” §§5000A(b), (g)(2). There is no immediate reason to think that a statute applying to “any tax” would apply to a “penalty.” Congress’s decision to label this exaction a “penalty” rather than a “tax” is significant because the Affordable Care Act describes many other exactions it creates as “taxes.” See Thomas More, 651 F. 3d, at 551. Where Congress uses certain language in one part of a statute and different language in another, it is generally presumed that Congress acts intentionally. See Russello v. United States, 464 U.S. 16, 23 (1983). Amicus argues that even though Congress did not label the shared responsibility payment a tax, we should treat it as such under the Anti-Injunction Act because it functions like a tax. It is true that Congress cannot change whether an exaction is a tax or a penalty for constitutional pur-poses simply by describing it as one or the other. Congress may not, for example, expand its power under the Taxing Clause, or escape the Double Jeopardy Clause’s constraint on criminal sanctions, by labeling a severe financial pun-ishment a “tax.” See Bailey v. Drexel Furniture Co., 259 U.S. 20, 36–37 (1922); Department of Revenue of Mont. v. Kurth Ranch, 511 U.S. 767, 779 (1994). The Anti-Injunction Act and the Affordable Care Act, however, are creatures of Congress’s own creation. How they relate to each other is up to Congress, and the best evidence of Congress’s intent is the statutory text. We have thus applied the Anti-Injunction Act to statutorily described “taxes” even where that label was inaccurate. See Bailey v. George, 259 U.S. 16 (1922) (Anti-Injunction Act applies to “Child Labor Tax” struck down as exceeding Congress’s taxing power in Drexel Furniture). Congress can, of course, describe something as a penalty but direct that it nonetheless be treated as a tax for purposes of the Anti-Injunction Act. For example, 26 U. S. C. §6671(a) provides that “any reference in this title to ‘tax’ imposed by this title shall be deemed also to refer to the penalties and liabilities provided by” subchapter 68B of the Internal Revenue Code. Penalties in subchapter 68B are thus treated as taxes under Title 26, which includes the Anti-Injunction Act. The individual mandate, however, is not in subchapter 68B of the Code. Nor does any other provision state that references to taxes in Title 26 shall also be “deemed” to apply to the individual mandate. Amicus attempts to show that Congress did render the Anti-Injunction Act applicable to the individual mandate, albeit by a more circuitous route. Section 5000A(g)(1) spec-ifies that the penalty for not complying with the man- date “shall be assessed and collected in the same manner as an assessable penalty under subchapter B of chapter 68.” Assessable penalties in subchapter 68B, in turn, “shall be assessed and collected in the same manner as taxes.” §6671(a). According to amicus, by directing that the penalty be “assessed and collected in the same man-ner as taxes,” §5000A(g)(1) made the Anti-Injunction Act applicable to this penalty. The Government disagrees. It argues that §5000A(g)(1) does not direct courts to apply the Anti-Injunction Act, because §5000A(g) is a directive only to the Secretary of the Treasury to use the same “ ‘methodology and procedures’ ” to collect the penalty that he uses to collect taxes. Brief for United States 32–33 (quoting Seven-Sky, 661 F. 3d, at 11). We think the Government has the better reading. As it observes, “Assessment” and “Collection” are chapters of the Internal Revenue Code providing the Secretary author-ity to assess and collect taxes, and generally specifying the means by which he shall do so. See §6201 (assess-ment authority); §6301 (collection authority). Section 5000A(g)(1)’s command that the penalty be “assessed and collected in the same manner” as taxes is best read as referring to those chapters and giving the Secretary the same authority and guidance with respect to the penalty. That interpretation is consistent with the remainder of §5000A(g), which instructs the Secretary on the tools he may use to collect the penalty. See §5000A(g)(2)(A) (barring criminal prosecutions); §5000A(g)(2)(B) (prohibiting the Secretary from using notices of lien and levies). The Anti-Injunction Act, by contrast, says nothing about the procedures to be used in assessing and collecting taxes. Amicus argues in the alternative that a different section of the Internal Revenue Code requires courts to treat the penalty as a tax under the Anti-Injunction Act. Section 6201(a) authorizes the Secretary to make “assessments of all taxes (including interest, additional amounts, additions to the tax, and assessable penalties).” (Emphasis added.) Amicus contends that the penalty must be a tax, because it is an assessable penalty and §6201(a) says that taxes include assessable penalties. That argument has force only if §6201(a) is read in isolation. The Code contains many provisions treating taxes and assessable penalties as distinct terms. See, e.g., §§860(h)(1), 6324A(a), 6601(e)(1)–(2), 6602, 7122(b). There would, for example, be no need for §6671(a) to deem “tax” to refer to certain assessable penalties if the Code al- ready included all such penalties in the term “tax.” Indeed, amicus’s earlier observation that the Code requires assessable penalties to be assessed and collected “in the same manner as taxes” makes little sense if assessable penalties are themselves taxes. In light of the Code’s consistent distinction between the terms “tax” and “assessable penalty,” we must accept the Government’s in-terpretation: §6201(a) instructs the Secretary that his authority to assess taxes includes the authority to assess penalties, but it does not equate assessable penalties to taxes for other purposes. The Affordable Care Act does not require that the penalty for failing to comply with the individual mandate be treated as a tax for purposes of the Anti-Injunction Act. The Anti-Injunction Act therefore does not apply to this suit, and we may proceed to the merits. III The Government advances two theories for the proposition that Congress had constitutional authority to enact the individual mandate. First, the Government argues that Congress had the power to enact the mandate under the Commerce Clause. Under that theory, Congress may order individuals to buy health insurance because the failure to do so affects interstate commerce, and could un-dercut the Affordable Care Act’s other reforms. Second, the Government argues that if the commerce power does not support the mandate, we should nonetheless uphold it as an exercise of Congress’s power to tax. According to the Government, even if Congress lacks the power to direct individuals to buy insurance, the only effect of the individual mandate is to raise taxes on those who do not do so, and thus the law may be upheld as a tax. A The Government’s first argument is that the individual mandate is a valid exercise of Congress’s power under the Commerce Clause and the Necessary and Proper Clause. According to the Government, the health care market is characterized by a significant cost-shifting problem. Everyone will eventually need health care at a time and to an extent they cannot predict, but if they do not have insurance, they often will not be able to pay for it. Because state and federal laws nonetheless require hospitals to provide a certain degree of care to individuals without regard to their ability to pay, see, e.g., 42 U. S. C. §1395dd; Fla. Stat. Ann. §395.1041, hospitals end up receiving compensation for only a portion of the services they provide. To recoup the losses, hospitals pass on the cost to insurers through higher rates, and insurers, in turn, pass on the cost to policy holders in the form of higher premiums. Congress estimated that the cost of uncompensated care raises family health insurance premiums, on average, by over $1,000 per year. 42 U. S. C. §18091(2)(F). In the Affordable Care Act, Congress addressed the problem of those who cannot obtain insurance coverage because of preexisting conditions or other health issues. It did so through the Act’s “guaranteed-issue” and “community- rating” provisions. These provisions together prohibit insurance companies from denying coverage to those with such conditions or charging unhealthy individuals higher premiums than healthy individuals. See §§300gg, 300gg–1, 300gg–3, 300gg–4. The guaranteed-issue and community-rating reforms do not, however, address the issue of healthy individuals who choose not to purchase insurance to cover potential health care needs. In fact, the reforms sharply exacerbate that problem, by providing an incentive for individuals to delay purchasing health insurance until they become sick, relying on the promise of guaranteed and affordable coverage. The reforms also threaten to impose massive new costs on insurers, who are required to accept unhealthy individuals but prohibited from charging them rates necessary to pay for their coverage. This will lead insurers to significantly increase premiums on everyone. See Brief for America’s Health Insurance Plans et al. as Amici Curiae in No. 11–393 etc. 8–9. The individual mandate was Congress’s solution to these problems. By requiring that individuals purchase health insurance, the mandate prevents cost-shifting by those who would otherwise go without it. In addition, the mandate forces into the insurance risk pool more healthy individuals, whose premiums on average will be higher than their health care expenses. This allows insurers to subsidize the costs of covering the unhealthy individuals the reforms require them to accept. The Government claims that Congress has power under the Commerce and Necessary and Proper Clauses to enact this solution. 1 The Government contends that the individual mandate is within Congress’s power because the failure to pur-chase insurance “has a substantial and deleterious effect on interstate commerce” by creating the cost-shifting prob-lem. Brief for United States 34. The path of our Commerce Clause decisions has not always run smooth, see United States v. Lopez, 514 U.S. 549, 552–559 (1995), but it is now well established that Congress has broad authority under the Clause. We have recognized, for example, that “[t]he power of Congress over interstate commerce is not confined to the regulation of commerce among the states,” but extends to activities that “have a substantial effect on interstate commerce.” United States v. Darby, 312 U.S. 100, 118–119 (1941). Congress’s power, more-over, is not limited to regulation of an activity that by itself substantially affects interstate commerce, but also extends to activities that do so only when aggregated with similar activities of others. See Wickard, 317 U. S., at 127–128. Given its expansive scope, it is no surprise that Congress has employed the commerce power in a wide variety of ways to address the pressing needs of the time. But Congress has never attempted to rely on that power to compel individuals not engaged in commerce to purchase an unwanted product.[3] Legislative novelty is not nec-essarily fatal; there is a first time for everything. But sometimes “the most telling indication of [a] severe con-stitutional problem . . . is the lack of historical precedent” for Congress’s action. Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U. S. ___, ___ (2010) (slip op., at 25) (internal quotation marks omitted). At the very least, we should “pause to consider the implications of the Government’s arguments” when confronted with such new conceptions of federal power. Lopez, supra, at 564. The Constitution grants Congress the power to “regulate Commerce.” Art. I, §8, cl. 3 (emphasis added). The power to regulate commerce presupposes the existence of commercial activity to be regulated. If the power to “regulate” something included the power to create it, many of the provisions in the Constitution would be superfluous. For example, the Constitution gives Congress the power to “coin Money,” in addition to the power to “regulate the Value thereof.” Id., cl. 5. And it gives Congress the power to “raise and support Armies” and to “provide and maintain a Navy,” in addition to the power to “make Rules for the Government and Regulation of the land and naval Forces.” Id., cls. 12–14. If the power to regulate the armed forces or the value of money included the power to bring the subject of the regulation into existence, the specific grant of such powers would have been unnecessary. The language of the Constitution reflects the natural understanding that the power to regulate assumes there is already something to be regulated. See Gibbons, 9 Wheat., at 188 (“[T]he enlightened patriots who framed our constitution, and the people who adopted it, must be understood to have employed words in their natural sense, and to have intended what they have said”).[4] Our precedent also reflects this understanding. As expansive as our cases construing the scope of the commerce power have been, they all have one thing in common: They uniformly describe the power as reaching “activity.” It is nearly impossible to avoid the word when quoting them. See, e.g., Lopez, supra, at 560 (“Where economic activity substantially affects interstate commerce, legislation regulating that activity will be sustained”); Perez, 402 U. S., at 154 (“Where the class of activities is regulated and that class is within the reach of federal power, the courts have no power to excise, as trivial, individual instances of the class” (emphasis in original; internal quotation marks omitted)); Wickard, supra, at 125 (“[E]ven if appellee’s activity be local and though it may not be regarded as commerce, it may still, whatever its nature, be reached by Congress if it exerts a substantial economic effect on interstate commerce”); NLRB v. Jones & Laughlin Steel Corp., 301 U.S. 1, 37 (1937) (“Although activities may be intrastate in character when separately considered, if they have such a close and substantial relation to interstate commerce that their control is essential or appropriate to protect that commerce from burdens and obstructions, Congress cannot be denied the power to exercise that control”); see also post, at 15, 25–26, 28, 32 (Ginsburg, J., concurring in part, concurring in judgment in part, and dissenting in part).[5] The individual mandate, however, does not regulate existing commercial activity. It instead compels individ-uals to become active in commerce by purchasing a product, on the ground that their failure to do so affects interstate commerce. Construing the Commerce Clause to permit Con-gress to regulate individuals precisely because they are doing nothing would open a new and potentially vast do-main to congressional authority. Every day individuals do not do an infinite number of things. In some cases they decide not to do something; in others they simply fail to do it. Allowing Congress to justify federal regulation by pointing to the effect of inaction on commerce would bring countless decisions an individual could potentially make within the scope of federal regulation, and—under the Government’s theory—empower Congress to make those decisions for him. Applying the Government’s logic to the familiar case of Wickard v. Filburn shows how far that logic would carry us from the notion of a government of limited powers. In Wickard, the Court famously upheld a federal penalty im-posed on a farmer for growing wheat for consumption on his own farm. 317 U. S., at 114–115, 128–129. That amount of wheat caused the farmer to exceed his quota under a program designed to support the price of wheat by limiting supply. The Court rejected the farmer’s argument that growing wheat for home consumption was beyond the reach of the commerce power. It did so on the ground that the farmer’s decision to grow wheat for his own use allowed him to avoid purchasing wheat in the market. That decision, when considered in the aggregate along with sim-ilar decisions of others, would have had a substantial ef-fect on the interstate market for wheat. Id., at 127–129. Wickard has long been regarded as “perhaps the most far reaching example of Commerce Clause authority over intrastate activity,” Lopez, 514 U. S., at 560, but the Government’s theory in this case would go much further. Under Wickard it is within Congress’s power to regulate the market for wheat by supporting its price. But price can be supported by increasing demand as well as by decreasing supply. The aggregated decisions of some consumers not to purchase wheat have a substantial effect on the price of wheat, just as decisions not to purchase health insurance have on the price of insurance. Congress can therefore command that those not buying wheat do so, just as it argues here that it may command that those not buying health insurance do so. The farmer in Wickard was at least actively engaged in the production of wheat, and the Government could regulate that activity because of its effect on commerce. The Government’s theory here would effectively override that limitation, by establishing that individuals may be regulated under the Commerce Clause whenever enough of them are not doing something the Government would have them do. Indeed, the Government’s logic would justify a mandatory purchase to solve almost any problem. See Seven-Sky, 661 F. 3d, at 14–15 (noting the Government’s inability to “identify any mandate to purchase a product or ser- vice in interstate commerce that would be unconstitu-tional” under its theory of the commerce power). To consider a different example in the health care market, many Americans do not eat a balanced diet. That group makes up a larger percentage of the total population than those without health insurance. See, e.g., Dept. of Agriculture and Dept. of Health and Human Services, Dietary Guidelines for Americans 1 (2010). The failure of that group to have a healthy diet increases health care costs, to a greater extent than the failure of the uninsured to purchase insurance. See, e.g., Finkelstein, Trogdon, Cohen, & Dietz, Annual Medical Spending Attributable to Obesity: Payer- and Service-Specific Estimates, 28 Health Affairs w822 (2009) (detailing the “undeniable link between ris-ing rates of obesity and rising medical spending,” and esti-mating that “the annual medical burden of obesity has risen to almost 10 percent of all medical spending and could amount to $147 billion per year in 2008”). Those in-creased costs are borne in part by other Americans who must pay more, just as the uninsured shift costs to the insured. See Center for Applied Ethics, Voluntary Health Risks: Who Should Pay?, 6 Issues in Ethics 6 (1993) (noting “overwhelming evidence that individuals with unhealthy habits pay only a fraction of the costs associated with their behaviors; most of the expense is borne by the rest of society in the form of higher insurance premiums, government expenditures for health care, and disability benefits”). Congress addressed the insurance problem by ordering everyone to buy insurance. Under the Gov-ernment’s theory, Congress could address the diet problem by ordering everyone to buy vegetables. See Dietary Guidelines, supra, at 19 (“Improved nutrition, appropriate eating behaviors, and increased physical activity have tre-mendous potential to . . . reduce health care costs”). People, for reasons of their own, often fail to do things that would be good for them or good for society. Those failures—joined with the similar failures of others—can readily have a substantial effect on interstate commerce. Under the Government’s logic, that authorizes Congress to use its commerce power to compel citizens to act as the Government would have them act. That is not the country the Framers of our Constitution envisioned. James Madison explained that the Commerce Clause was “an addition which few oppose and from which no apprehensions are entertained.” The Federalist No. 45, at 293. While Congress’s authority under the Commerce Clause has of course expanded with the growth of the national economy, our cases have “always recognized that the power to regulate commerce, though broad indeed, has limits.” Maryland v. Wirtz, 392 U.S. 183, 196 (1968). The Government’s theory would erode those limits, permitting Congress to reach beyond the natural extent of its authority, “everywhere extending the sphere of its activity and drawing all power into its impetuous vortex.” The Federalist No. 48, at 309 (J. Madison). Congress already enjoys vast power to regulate much of what we do. Accepting the Government’s theory would give Congress the same license to regulate what we do not do, fundamentally changing the relation between the citizen and the Federal Government.[6] To an economist, perhaps, there is no difference between activity and inactivity; both have measurable economic effects on commerce. But the distinction between doing something and doing nothing would not have been lost on the Framers, who were “practical statesmen,” not metaphysical philosophers. Industrial Union Dept., AFL–CIO v. American Petroleum Institute, 448 U.S. 607, 673 (1980) (Rehnquist, J., concurring in judgment). As we have ex-plained, “the framers of the Constitution were not mere visionaries, toying with speculations or theories, but practical men, dealing with the facts of political life as they understood them, putting into form the government they were creating, and prescribing in language clear and intelligible the powers that government was to take.” South Carolina v. United States, 199 U.S. 437, 449 (1905). The Framers gave Congress the power to regulate commerce, not to compel it, and for over 200 years both our decisions and Congress’s actions have reflected this understanding. There is no reason to depart from that understanding now. The Government sees things differently. It argues that because sickness and injury are unpredictable but unavoidable, “the uninsured as a class are active in the market for health care, which they regularly seek and obtain.” Brief for United States 50. The individual mandate “merely regulates how individuals finance and pay for that active participation—requiring that they do so through insurance, rather than through attempted self-insurance with the back-stop of shifting costs to others.” Ibid. The Government repeats the phrase “active in the market for health care” throughout its brief, see id., at 7, 18, 34, 50, but that concept has no constitutional significance. An individual who bought a car two years ago and may buy another in the future is not “active in the car market” in any pertinent sense. The phrase “active in the market” cannot obscure the fact that most of those regulated by the individual mandate are not currently engaged in any commercial activity involving health care, and that fact is fatal to the Government’s effort to “regulate the uninsured as a class.” Id., at 42. Our precedents recognize Congress’s power to regulate “class[es] of activities,” Gonzales v. Raich, 545 U.S. 1, 17 (2005) (emphasis added), not classes of individuals, apart from any activity in which they are engaged, see, e.g., Perez, 402 U. S., at 153 (“Petitioner is clearly a member of the class which engages in ‘extortionate credit transactions’ . . .” (emphasis deleted)). The individual mandate’s regulation of the uninsured as a class is, in fact, particularly divorced from any link to existing commercial activity. The mandate primarily affects healthy, often young adults who are less likely to need significant health care and have other priorities for spending their money. It is precisely because these individuals, as an actuarial class, incur relatively low health care costs that the mandate helps counter the effect of forcing insurance companies to cover others who impose greater costs than their premiums are allowed to reflect. See 42 U. S. C. §18091(2)(I) (recognizing that the mandate would “broaden the health insurance risk pool to include healthy individuals, which will lower health insurance premiums”). If the individual mandate is targeted at a class, it is a class whose commercial inactivity rather than activity is its defining feature. The Government, however, claims that this does not matter. The Government regards it as sufficient to trigger Congress’s authority that almost all those who are uninsured will, at some unknown point in the future, engage in a health care transaction. Asserting that “[t]here is no temporal limitation in the Commerce Clause,” the Government argues that because “[e]veryone subject to this regulation is in or will be in the health care market,” they can be “regulated in advance.” Tr. of Oral Arg. 109 (Mar. 27, 2012). The proposition that Congress may dictate the conduct of an individual today because of prophesied future ac-tivity finds no support in our precedent. We have said that Congress can anticipate the effects on commerce of an eco-nomic activity. See, e.g., Consolidated Edison Co. v. NLRB, 305 U.S. 197 (1938) (regulating the labor practices of utility companies); Heart of Atlanta Motel, Inc. v. United States, 379 U.S. 241 (1964) (prohibiting discrimination by hotel operators); Katzenbach v. McClung, 379 U.S. 294 (1964) (prohibiting discrimination by restaurant owners). But we have never permitted Congress to anticipate that activity itself in order to regulate individuals not currently engaged in commerce. Each one of our cases, including those cited by Justice Ginsburg, post, at 20–21, involved preexisting economic activity. See, e.g., Wickard, 317 U. S., at 127–129 (producing wheat); Raich, supra, at 25 (growing marijuana). Everyone will likely participate in the markets for food, clothing, transportation, shelter, or energy; that does not authorize Congress to direct them to purchase particular products in those or other markets today. The Commerce Clause is not a general license to regulate an individual from cradle to grave, simply because he will predictably engage in particular transactions. Any police power to regulate individuals as such, as opposed to their activities, remains vested in the States. The Government argues that the individual mandate can be sustained as a sort of exception to this rule, because health insurance is a unique product. According to the Government, upholding the individual mandate would not justify mandatory purchases of items such as cars or broccoli because, as the Government puts it, “[h]ealth in-surance is not purchased for its own sake like a car or broccoli; it is a means of financing health-care consumption and covering universal risks.” Reply Brief for United States 19. But cars and broccoli are no more purchased for their “own sake” than health insurance. They are purchased to cover the need for transportation and food. The Government says that health insurance and health care financing are “inherently integrated.” Brief for United States 41. But that does not mean the compelled purchase of the first is properly regarded as a regulation of the second. No matter how “inherently integrated” health insurance and health care consumption may be, they are not the same thing: They involve different transactions, entered into at different times, with different providers. And for most of those targeted by the mandate, significant health care needs will be years, or even decades, away. The proximity and degree of connection between the mandate and the subsequent commercial activity is too lack-ing to justify an exception of the sort urged by the Gov- ernment. The individual mandate forces individuals into commerce precisely because they elected to refrain from commercial activity. Such a law cannot be sus- tained under a clause authorizing Congress to “regulate Commerce.” 2 The Government next contends that Congress has the power under the Necessary and Proper Clause to enact the individual mandate because the mandate is an “integral part of a comprehensive scheme of economic regulation”—the guaranteed-issue and community-rating insurance reforms. Brief for United States 24. Under this argument, it is not necessary to consider the effect that an individual’s inactivity may have on interstate commerce; it is enough that Congress regulate commercial activity in a way that requires regulation of inactivity to be effective. The power to “make all Laws which shall be necessary and proper for carrying into Execution” the powers enumerated in the Constitution, Art. I, §8, cl. 18, vests Congress with authority to enact provisions “incidental to the [enumerated] power, and conducive to its beneficial exercise,” McCulloch, 4 Wheat., at 418. Although the Clause gives Congress authority to “legislate on that vast mass of incidental powers which must be involved in the con-stitution,” it does not license the exercise of any “great substantive and independent power[s]” beyond those specifi-cally enumerated. Id., at 411, 421. Instead, the Clause is “ ‘merely a declaration, for the removal of all uncertainty, that the means of carrying into execution those [powers] otherwise granted are included in the grant.’ ” Kinsella v. United States ex rel. Singleton, 361 U.S. 234, 247 (1960) (quoting VI Writings of James Madison 383 (G. Hunt ed. 1906)). As our jurisprudence under the Necessary and Proper Clause has developed, we have been very deferential to Congress’s determination that a regulation is “necessary.” We have thus upheld laws that are “ ‘convenient, or use-ful’ or ‘conducive’ to the authority’s ‘beneficial exercise.’ ” Comstock, 560 U. S., at ___ (slip op., at 5) (quoting McCulloch, supra, at 413, 418). But we have also carried out our responsibility to declare unconstitutional those laws that undermine the structure of government established by the Constitution. Such laws, which are not “consist[ent] with the letter and spirit of the constitution,” McCulloch, supra, at 421, are not “proper [means] for carrying into Execution” Congress’s enumerated powers. Rather, they are, “in the words of The Federalist, ‘merely acts of usurpation’ which ‘deserve to be treated as such.’ ” Printz v. United States, 521 U.S. 898, 924 (1997) (alterations omitted) (quoting The Federalist No. 33, at 204 (A. Hamilton)); see also New York, 505 U. S., at 177; Comstock, supra, at ___ (slip op., at 5) (Kennedy, J., concurring in judgment) (“It is of fundamental importance to consider whether essential attributes of state sovereignty are compromised by the assertion of federal power under the Necessary and Proper Clause . . .”). Applying these principles, the individual mandate cannot be sustained under the Necessary and Proper Clause as an essential component of the insurance reforms. Each of our prior cases upholding laws under that Clause involved exercises of authority derivative of, and in service to, a granted power. For example, we have upheld provisions permitting continued confinement of those already in federal custody when they could not be safely released, Comstock, supra, at ___ (slip op., at 1–2); criminaliz- ing bribes involving organizations receiving federal funds, Sabri v. United States, 541 U.S. 600, 602, 605 (2004); and tolling state statutes of limitations while cases are pending in federal court, Jinks v. Richland County, 538 U.S. 456, 459, 462 (2003). The individual mandate, by con-trast, vests Congress with the extraordinary ability to create the necessary predicate to the exercise of an enumerated power. This is in no way an authority that is “narrow in scope,” Comstock, supra, at ___ (slip op., at 20), or “incidental” to the exercise of the commerce power, McCulloch, supra, at 418. Rather, such a conception of the Necessary and Proper Clause would work a substantial expansion of federal authority. No longer would Congress be limited to regulating under the Commerce Clause those who by some preexisting activity bring themselves within the sphere of federal regulation. Instead, Congress could reach beyond the natural limit of its authority and draw within its regulatory scope those who otherwise would be outside of it. Even if the individual mandate is “necessary” to the Act’s insurance reforms, such an expansion of federal power is not a “proper” means for making those reforms effective. The Government relies primarily on our decision in Gonzales v. Raich. In Raich, we considered “comprehensive legislation to regulate the interstate market” in marijuana. 545 U. S., at 22. Certain individuals sought an exemption from that regulation on the ground that they engaged in only intrastate possession and consumption. We denied any exemption, on the ground that marijuana is a fungible commodity, so that any marijuana could be readily diverted into the interstate market. Congress’s attempt to regulate the interstate market for marijuana would therefore have been substantially undercut if it could not also regulate intrastate possession and consumption. Id., at 19. Accordingly, we recognized that “Congress was acting well within its authority” under the Necessary and Proper Clause even though its “regulation ensnare[d] some purely intrastate activity.” Id., at 22; see also Perez, 402 U. S., at 154. Raich thus did not involve the exercise of any “great substantive and independent power,” McCulloch, supra, at 411, of the sort at issue here. Instead, it concerned only the constitutionality of “individual applications of a concededly valid statutory scheme.” Raich, supra, at 23 (emphasis added). Just as the individual mandate cannot be sustained as a law regulating the substantial effects of the failure to purchase health insurance, neither can it be upheld as a “necessary and proper” component of the insurance re-forms. The commerce power thus does not authorize the mandate. Accord, post, at 4–16 (joint opinion of Scalia, Kennedy, Thomas, and Alito, JJ., dissenting). B That is not the end of the matter. Because the Commerce Clause does not support the individual mandate, it is necessary to turn to the Government’s second argument: that the mandate may be upheld as within Congress’s enumerated power to “lay and collect Taxes.” Art. I, §8, cl. 1. The Government’s tax power argument asks us to view the statute differently than we did in considering its commerce power theory. In making its Commerce Clause argument, the Government defended the mandate as a regulation requiring individuals to purchase health in-surance. The Government does not claim that the taxing power allows Congress to issue such a command. Instead, the Government asks us to read the mandate not as ordering individuals to buy insurance, but rather as imposing a tax on those who do not buy that product. The text of a statute can sometimes have more than one possible meaning. To take a familiar example, a law that reads “no vehicles in the park” might, or might not, ban bicycles in the park. And it is well established that if a statute has two possible meanings, one of which violates the Constitution, courts should adopt the meaning that does not do so. Justice Story said that 180 years ago: “No court ought, unless the terms of an act rendered it unavoidable, to give a construction to it which should involve a violation, however unintentional, of the constitution.” Parsons v. Bedford, 3 Pet. 433, 448–449 (1830). Justice Holmes made the same point a century later: “[T]he rule is settled that as between two possible interpretations of a statute, by one of which it would be unconstitutional and by the other valid, our plain duty is to adopt that which will save the Act.” Blodgett v. Holden, 275 U.S. 142, 148 (1927) (concurring opinion). The most straightforward reading of the mandate is that it commands individuals to purchase insurance. After all, it states that individuals “shall” maintain health insurance. 26 U. S. C. §5000A(a). Congress thought it could enact such a command under the Commerce Clause, and the Government primarily defended the law on that basis. But, for the reasons explained above, the Commerce Clause does not give Congress that power. Under our precedent, it is therefore necessary to ask whether the Government’s alternative reading of the statute—that it only imposes a tax on those without insurance—is a reasonable one. Under the mandate, if an individual does not maintain health insurance, the only consequence is that he must make an additional payment to the IRS when he pays his taxes. See §5000A(b). That, according to the Government, means the mandate can be regarded as establishing a condition—not owning health insurance—that triggers a tax—the required payment to the IRS. Under that theory, the mandate is not a legal command to buy insurance. Rather, it makes going without insurance just another thing the Government taxes, like buying gasoline or earning income. And if the mandate is in effect just a tax hike on certain taxpayers who do not have health insurance, it may be within Congress’s constitutional power to tax. The question is not whether that is the most natural interpretation of the mandate, but only whether it is a “fairly possible” one. Crowell v. Benson, 285 U.S. 22, 62 (1932). As we have explained, “every reasonable construction must be resorted to, in order to save a statute from unconstitutionality.” Hooper v. California, 155 U.S. 648, 657 (1895). The Government asks us to interpret the mandate as imposing a tax, if it would otherwise violate the Constitution. Granting the Act the full measure of deference owed to federal statutes, it can be so read, for the reasons set forth below. C The exaction the Affordable Care Act imposes on those without health insurance looks like a tax in many respects. The “[s]hared responsibility payment,” as the statute entitles it, is paid into the Treasury by “taxpayer[s]” when they file their tax returns. 26 U. S. C. §5000A(b). It does not apply to individuals who do not pay federal income taxes because their household income is less than the filing threshold in the Internal Revenue Code. §5000A(e)(2). For taxpayers who do owe the payment, its amount is determined by such familiar factors as taxable income, number of dependents, and joint filing status. §§5000A(b)(3), (c)(2), (c)(4). The requirement to pay is found in the Internal Revenue Code and enforced by the IRS, which—as we previously explained—must assess and collect it “in the same manner as taxes.” Supra, at 13–14. This process yields the essential feature of any tax: it produces at least some revenue for the Government. United States v. Kahriger, 345 U.S. 22, 28, n. 4 (1953). Indeed, the payment is expected to raise about $4 billion per year by 2017. Congressional Budget Office, Payments of Penalties for Being Uninsured Under the Patient Protection and Affordable Care Act (Apr. 30, 2010), in Selected CBO Publications Related to Health Care Legislation, 2009–2010, p. 71 (rev. 2010). It is of course true that the Act describes the payment as a “penalty,” not a “tax.” But while that label is fatal to the application of the Anti-Injunction Act, supra, at 12–13, it does not determine whether the payment may be viewed as an exercise of Congress’s taxing power. It is up to Con-gress whether to apply the Anti-Injunction Act to any particular statute, so it makes sense to be guided by Congress’s choice of label on that question. That choice does not, however, control whether an exaction is within Congress’s constitutional power to tax. Our precedent reflects this: In 1922, we decided two challenges to the “Child Labor Tax” on the same day. In the first, we held that a suit to enjoin collection of the so-called tax was barred by the Anti-Injunction Act. George, 259 U. S., at 20. Congress knew that suits to obstruct taxes had to await payment under the Anti-Injunction Act; Congress called the child labor tax a tax; Congress therefore intended the Anti-Injunction Act to apply. In the second case, however, we held that the same exaction, although labeled a tax, was not in fact authorized by Con-gress’s taxing power. Drexel Furniture, 259 U. S., at 38. That constitutional question was not controlled by Congress’s choice of label. We have similarly held that exactions not labeled taxes nonetheless were authorized by Congress’s power to tax. In the License Tax Cases, for example, we held that federal licenses to sell liquor and lottery tickets—for which the licensee had to pay a fee—could be sustained as exercises of the taxing power. 5 Wall., at 471. And in New York v. United States we upheld as a tax a “surcharge” on out-of-state nuclear waste shipments, a portion of which was paid to the Federal Treasury. 505 U. S., at 171. We thus ask whether the shared responsibility payment falls within Congress’s taxing power, “[d]isregarding the designa-tion of the exaction, and viewing its substance and application.” United States v. Constantine, 296 U.S. 287, 294 (1935); cf. Quill Corp. v. North Dakota, 504 U.S. 298, 310 (1992) (“[M]agic words or labels” should not “disable an otherwise constitutional levy” (internal quotation marks omitted)); Nelson v. Sears, Roebuck & Co., 312 U.S. 359, 363 (1941) (“In passing on the constitutionality of a tax law, we are concerned only with its practical operation, not its definition or the precise form of descriptive words which may be applied to it” (internal quotation marks omitted)); United States v. Sotelo, 436 U.S. 268, 275 (1978) (“That the funds due are referred to as a ‘penalty’ . . . does not alter their essential character as taxes”).[7] Our cases confirm this functional approach. For example, in Drexel Furniture, we focused on three practical characteristics of the so-called tax on employing child laborers that convinced us the “tax” was actually a penalty. First, the tax imposed an exceedingly heavy burden—10 percent of a company’s net income—on those who employed children, no matter how small their infraction. Second, it imposed that exaction only on those who knowingly employed underage laborers. Such scienter require-ments are typical of punitive statutes, because Congress often wishes to punish only those who intentionally break the law. Third, this “tax” was enforced in part by the Department of Labor, an agency responsible for pun-ishing violations of labor laws, not collecting revenue. 259 U. S., at 36–37; see also, e.g., Kurth Ranch, 511 U. S., at 780–782 (considering, inter alia, the amount of the exaction, and the fact that it was imposed for violation of a separate criminal law); Constantine, supra, at 295 (same). The same analysis here suggests that the shared responsibility payment may for constitutional purposes be considered a tax, not a penalty: First, for most Americans the amount due will be far less than the price of insurance, and, by statute, it can never be more.[8] It may often be a reasonable financial decision to make the payment rather than purchase insurance, unlike the “prohibitory” financial punishment in Drexel Furniture. 259 U. S., at 37. Second, the individual mandate contains no scienter requirement. Third, the payment is collected solely by the IRS through the normal means of taxation—except that the Service is not allowed to use those means most suggestive of a punitive sanction, such as criminal prosecution. See §5000A(g)(2). The reasons the Court in Drexel Furniture held that what was called a “tax” there was a penalty support the conclusion that what is called a “penalty” here may be viewed as a tax.[9] None of this is to say that the payment is not intended to affect individual conduct. Although the payment will raise considerable revenue, it is plainly designed to expand health insurance coverage. But taxes that seek to influence conduct are nothing new. Some of our earliest federal taxes sought to deter the purchase of imported manufactured goods in order to foster the growth of domestic industry. See W. Brownlee, Federal Taxation in America 22 (2d ed. 2004); cf. 2 J. Story, Commentaries on the Constitution of the United States §962, p. 434 (1833) (“the taxing power is often, very often, applied for other purposes, than revenue”). Today, federal and state taxes can compose more than half the retail price of cigarettes, not just to raise more money, but to encourage people to quit smoking. And we have upheld such obviously regulatory measures as taxes on selling marijuana and sawed-off shotguns. See United States v. Sanchez, 340 U.S. 42, 44–45 (1950); Sonzinsky v. United States, 300 U.S. 506, 513 (1937). Indeed, “[e]very tax is in some measure regula-tory. To some extent it interposes an economic impediment to the activity taxed as compared with others not taxed.” Sonzinsky, supra, at 513. That §5000A seeks to shape decisions about whether to buy health insurance does not mean that it cannot be a valid exercise of the taxing power. In distinguishing penalties from taxes, this Court has explained that “if the concept of penalty means anything, it means punishment for an unlawful act or omission.” United States v. Reorganized CF&I Fabricators of Utah, Inc., 518 U.S. 213, 224 (1996); see also United States v. La Franca, 282 U.S. 568, 572 (1931) (“[A] penalty, as the word is here used, is an exaction imposed by statute as punishment for an unlawful act”). While the individual mandate clearly aims to induce the purchase of health insurance, it need not be read to declare that failing to do so is unlawful. Neither the Act nor any other law attaches negative legal consequences to not buying health insurance, beyond requiring a payment to the IRS. The Government agrees with that reading, confirming that if someone chooses to pay rather than obtain health insurance, they have fully complied with the law. Brief for United States 60–61; Tr. of Oral Arg. 49–50 (Mar. 26, 2012). Indeed, it is estimated that four million people each year will choose to pay the IRS rather than buy insurance. See Congressional Budget Office, supra, at 71. We would expect Congress to be troubled by that prospect if such conduct were unlawful. That Congress apparently regards such extensive failure to comply with the mandate as tolerable suggests that Congress did not think it was creating four million outlaws. It suggests instead that the shared responsibility payment merely imposes a tax citizens may lawfully choose to pay in lieu of buying health insurance. The plaintiffs contend that Congress’s choice of language—stating that individuals “shall” obtain insurance or pay a “penalty”—requires reading §5000A as punishing unlawful conduct, even if that interpretation would ren-der the law unconstitutional. We have rejected a similar argument before. In New York v. United States we examined a statute providing that “ ‘[e]ach State shall be responsible for providing . . . for the disposal of . . . low-level radioactive waste.’ ” 505 U. S., at 169 (quoting 42 U. S. C. §2021c(a)(1)(A)). A State that shipped its waste to another State was exposed to surcharges by the receiving State, a portion of which would be paid over to the Federal Government. And a State that did not adhere to the statutory scheme faced “[p]enalties for failure to comply,” including increases in the surcharge. §2021e(e)(2); New York, 505 U. S., at 152–153. New York urged us to read the statute as a federal command that the state legislature enact legislation to dispose of its waste, which would have violated the Constitution. To avoid that outcome, we interpreted the statute to impose only “a series of incentives” for the State to take responsibility for its waste. We then sustained the charge paid to the Federal Government as an exercise of the taxing power. Id., at 169–174. We see no insurmountable obstacle to a similar approach here.[10] The joint dissenters argue that we cannot uphold §5000A as a tax because Congress did not “frame” it as such. Post, at 17. In effect, they contend that even if the Constitution permits Congress to do exactly what we interpret this statute to do, the law must be struck down because Congress used the wrong labels. An example may help illustrate why labels should not control here. Suppose Congress enacted a statute providing that every taxpayer who owns a house without energy efficient windows must pay $50 to the IRS. The amount due is adjusted based on factors such as taxable income and joint filing status, and is paid along with the taxpayer’s income tax return. Those whose income is below the filing threshold need not pay. The required payment is not called a “tax,” a “penalty,” or anything else. No one would doubt that this law imposed a tax, and was within Congress’s power to tax. That conclusion should not change simply because Congress used the word “penalty” to describe the pay-ment. Interpreting such a law to be a tax would hardly “[i]mpos[e] a tax through judicial legislation.” Post, at 25. Rather, it would give practical effect to the Legislature’s enactment. Our precedent demonstrates that Congress had the power to impose the exaction in §5000A under the taxing power, and that §5000A need not be read to do more than impose a tax. That is sufficient to sustain it. The “question of the constitutionality of action taken by Congress does not depend on recitals of the power which it undertakes to exercise.” Woods v. Cloyd W. Miller Co., 333 U.S. 138, 144 (1948). Even if the taxing power enables Congress to impose a tax on not obtaining health insurance, any tax must still comply with other requirements in the Constitution. Plaintiffs argue that the shared responsibility payment does not do so, citing Article I, §9, clause 4. That clause provides: “No Capitation, or other direct, Tax shall be laid, unless in Proportion to the Census or Enumeration herein before directed to be taken.” This requirement means that any “direct Tax” must be apportioned so that each State pays in proportion to its population. According to the plaintiffs, if the individual mandate imposes a tax, it is a direct tax, and it is unconstitutional because Congress made no effort to apportion it among the States. Even when the Direct Tax Clause was written it was unclear what else, other than a capitation (also known as a “head tax” or a “poll tax”), might be a direct tax. See Springer v. United States, 102 U.S. 586, 596–598 (1881). Soon after the framing, Congress passed a tax on ownership of carriages, over James Madison’s objection that it was an unapportioned direct tax. Id., at 597. This Court upheld the tax, in part reasoning that apportioning such a tax would make little sense, because it would have required taxing carriage owners at dramatically different rates depending on how many carriages were in their home State. See Hylton v. United States, 3 Dall. 171, 174 (1796) (opinion of Chase, J.). The Court was unanimous, and those Justices who wrote opinions either directly asserted or strongly suggested that only two forms of taxation were direct: capitations and land taxes. See id., at 175; id., at 177 (opinion of Paterson, J.); id., at 183 (opinion of Iredell, J.). That narrow view of what a direct tax might be per-sisted for a century. In 1880, for example, we explained that “direct taxes, within the meaning of the Constitution, are only capitation taxes, as expressed in that instrument, and taxes on real estate.” Springer, supra, at 602. In 1895, we expanded our interpretation to include taxes on personal property and income from personal property, in the course of striking down aspects of the federal income tax. Pollock v. Farmers’ Loan & Trust Co., 158 U.S. 601, 618 (1895). That result was overturned by the Sixteenth Amendment, although we continued to consider taxes on personal property to be direct taxes. See Eisner v. Macom-ber, 252 U.S. 189, 218–219 (1920). A tax on going without health insurance does not fall within any recognized category of direct tax. It is not a capitation. Capitations are taxes paid by every person, “without regard to property, profession, or any other circumstance.” Hylton, supra, at 175 (opinion of Chase, J.) (emphasis altered). The whole point of the shared responsibility payment is that it is triggered by specific cir-cumstances—earning a certain amount of income but not obtaining health insurance. The payment is also plainly not a tax on the ownership of land or personal property. The shared responsibility payment is thus not a direct tax that must be apportioned among the several States. There may, however, be a more fundamental objection to a tax on those who lack health insurance. Even if only a tax, the payment under §5000A(b) remains a burden that the Federal Government imposes for an omission, not an act. If it is troubling to interpret the Commerce Clause as authorizing Congress to regulate those who abstain from commerce, perhaps it should be similarly troubling to permit Congress to impose a tax for not doing something. Three considerations allay this concern. First, and most importantly, it is abundantly clear the Constitution does not guarantee that individuals may avoid taxation through inactivity. A capitation, after all, is a tax that every- one must pay simply for existing, and capitations are expressly contemplated by the Constitution. The Court today holds that our Constitution protects us from federal regulation under the Commerce Clause so long as we ab-stain from the regulated activity. But from its creation, the Constitution has made no such promise with respect to taxes. See Letter from Benjamin Franklin to M. Le Roy (Nov. 13, 1789) (“Our new Constitution is now established . . . but in this world nothing can be said to be certain, except death and taxes”). Whether the mandate can be upheld under the Commerce Clause is a question about the scope of federal authority. Its answer depends on whether Congress can exercise what all acknowledge to be the novel course of directing individuals to purchase insurance. Congress’s use of the Taxing Clause to encourage buying something is, by contrast, not new. Tax incentives already promote, for example, purchasing homes and professional educations. See 26 U. S. C. §§163(h), 25A. Sustaining the mandate as a tax depends only on whether Congress has properly exercised its taxing power to encourage purchasing health insurance, not whether it can. Upholding the individual mandate under the Taxing Clause thus does not recognize any new federal power. It determines that Congress has used an existing one. Second, Congress’s ability to use its taxing power to influence conduct is not without limits. A few of our cases policed these limits aggressively, invalidating punitive exactions obviously designed to regulate behavior otherwise regarded at the time as beyond federal authority. See, e.g., United States v. Butler, 297 U.S. 1 (1936); Drexel Furniture, 259 U.S. 20. More often and more recently we have declined to closely examine the regulatory motive or effect of revenue-raising measures. See Kahriger, 345 U. S., at 27–31 (collecting cases). We have nonetheless maintained that “ ‘there comes a time in the extension of the penalizing features of the so-called tax when it loses its character as such and becomes a mere penalty with the characteristics of regulation and punishment.’ ” Kurth Ranch, 511 U. S., at 779 (quoting Drexel Furniture, supra, at 38). We have already explained that the shared responsibility payment’s practical characteristics pass muster as a tax under our narrowest interpretations of the taxing power. Supra, at 35–36. Because the tax at hand is within even those strict limits, we need not here decide the precise point at which an exaction becomes so punitive that the taxing power does not authorize it. It remains true, however, that the “ ‘power to tax is not the power to destroy while this Court sits.’ ” Oklahoma Tax Comm’n v. Texas Co., 336 U.S. 342, 364 (1949) (quoting Panhandle Oil Co. v. Mississippi ex rel. Knox, 277 U.S. 218, 223 (1928) (Holmes, J., dissenting)). Third, although the breadth of Congress’s power to tax is greater than its power to regulate commerce, the taxing power does not give Congress the same degree of control over individual behavior. Once we recognize that Congress may regulate a particular decision under the Commerce Clause, the Federal Government can bring its full weight to bear. Congress may simply command individ-uals to do as it directs. An individual who disobeys may be subjected to criminal sanctions. Those sanctions can include not only fines and imprisonment, but all the attendant consequences of being branded a criminal: deprivation of otherwise protected civil rights, such as the right to bear arms or vote in elections; loss of employment opportunities; social stigma; and severe disabilities in other controversies, such as custody or immigration disputes. By contrast, Congress’s authority under the taxing power is limited to requiring an individual to pay money into the Federal Treasury, no more. If a tax is properly paid, the Government has no power to compel or punish individuals subject to it. We do not make light of the se-vere burden that taxation—especially taxation motivated by a regulatory purpose—can impose. But imposition of a tax nonetheless leaves an individual with a lawful choice to do or not do a certain act, so long as he is willing to pay a tax levied on that choice.[11] The Affordable Care Act’s requirement that certain in-dividuals pay a financial penalty for not obtaining health insurance may reasonably be characterized as a tax. Be-cause the Constitution permits such a tax, it is not our role to forbid it, or to pass upon its wisdom or fairness. D Justice Ginsburg questions the necessity of rejecting the Government’s commerce power argument, given that §5000A can be upheld under the taxing power. Post, at 37. But the statute reads more naturally as a command to buy insurance than as a tax, and I would uphold it as a command if the Constitution allowed it. It is only because the Commerce Clause does not authorize such a command that it is necessary to reach the taxing power question. And it is only because we have a duty to construe a stat-ute to save it, if fairly possible, that §5000A can be interpreted as a tax. Without deciding the Commerce Clause question, I would find no basis to adopt such a saving construction. The Federal Government does not have the power to order people to buy health insurance. Section 5000A would therefore be unconstitutional if read as a command. The Federal Government does have the power to impose a tax on those without health insurance. Section 5000A is therefore constitutional, because it can reasonably be read as a tax. IV A The States also contend that the Medicaid expansion exceeds Congress’s authority under the Spending Clause. They claim that Congress is coercing the States to adopt the changes it wants by threatening to withhold all of a State’s Medicaid grants, unless the State accepts the new expanded funding and complies with the conditions that come with it. This, they argue, violates the basic principle that the “Federal Government may not compel the States to enact or administer a federal regulatory program.” New York, 505 U. S., at 188. There is no doubt that the Act dramatically increases state obligations under Medicaid. The current Medicaid program requires States to cover only certain discrete categories of needy individuals—pregnant women, children, needy families, the blind, the elderly, and the dis-abled. 42 U. S. C. §1396a(a)(10). There is no mandatory coverage for most childless adults, and the States typically do not offer any such coverage. The States also enjoy considerable flexibility with respect to the coverage levels for parents of needy families. §1396a(a)(10)(A)(ii). On average States cover only those unemployed parents who make less than 37 percent of the federal poverty level, and only those employed parents who make less than 63 percent of the poverty line. Kaiser Comm’n on Medicaid and the Uninsured, Performing Under Pressure 11, and fig. 11 (2012). The Medicaid provisions of the Affordable Care Act, in contrast, require States to expand their Medicaid programs by 2014 to cover all individuals under the age of 65 with incomes below 133 percent of the federal poverty line. §1396a(a)(10)(A)(i)(VIII). The Act also establishes a new “[e]ssential health benefits” package, which States must provide to all new Medicaid recipients—a level sufficient to satisfy a recipient’s obligations under the individual man-date. §§1396a(k)(1), 1396u–7(b)(5), 18022(b). The Af-fordable Care Act provides that the Federal Government will pay 100 percent of the costs of covering these newly eligible individuals through 2016. §1396d(y)(1). In the following years, the federal payment level gradually decreases, to a minimum of 90 percent. Ibid. In light of the expansion in coverage mandated by the Act, the Federal Government estimates that its Medicaid spending will in-crease by approximately $100 billion per year, nearly 40 percent above current levels. Statement of Douglas W. Elmendorf, CBO’s Analysis of the Major Health Care Legislation Enacted in March 2010, p. 14, Table 2 (Mar. 30, 2011). The Spending Clause grants Congress the power “to pay the Debts and provide for the . . . general Welfare of the United States.” U. S. Const., Art. I, §8, cl. 1. We have long recognized that Congress may use this power to grant federal funds to the States, and may condition such a grant upon the States’ “taking certain actions that Congress could not require them to take.” College Savings Bank, 527 U. S., at 686. Such measures “encourage a State to regulate in a particular way, [and] influenc[e] a State’s policy choices.” New York, supra, at 166. The con-ditions imposed by Congress ensure that the funds are used by the States to “provide for the . . . general Welfare” in the manner Congress intended. At the same time, our cases have recognized limits on Congress’s power under the Spending Clause to secure state compliance with federal objectives. “We have repeatedly characterized . . . Spending Clause legislation as ‘much in the nature of a contract.’ ” Barnes v. Gorman, 536 U.S. 181, 186 (2002) (quoting Pennhurst State School and Hospital v. Halderman, 451 U.S. 1, 17 (1981)). The legitimacy of Congress’s exercise of the spending power “thus rests on whether the State voluntarily and knowingly accepts the terms of the ‘contract.’ ” Pennhurst, supra, at 17. Respecting this limitation is critical to ensuring that Spending Clause legislation does not undermine the status of the States as independent sovereigns in our fed-eral system. That system “rests on what might at first seem a counterintuitive insight, that ‘freedom is enhanced by the creation of two governments, not one.’ ” Bond, 564 U. S., at ___ (slip op., at 8) (quoting Alden v. Maine, 527 U.S. 706, 758 (1999)). For this reason, “the Constitution has never been understood to confer upon Congress the ability to require the States to govern according to Congress’ instructions.” New York, supra, at 162. Otherwise the two-government system established by the Framers would give way to a system that vests power in one central government, and individual liberty would suffer. That insight has led this Court to strike down fed- eral legislation that commandeers a State’s legislative or administrative apparatus for federal purposes. See, e.g., Printz, 521 U. S., at 933 (striking down federal legisla- tion compelling state law enforcement officers to perform federally mandated background checks on handgun purchasers); New York, supra, at 174–175 (invalidating provisions of an Act that would compel a State to either take title to nuclear waste or enact particular state waste regulations). It has also led us to scrutinize Spending Clause legislation to ensure that Congress is not using financial inducements to exert a “power akin to undue influence.” Steward Machine Co. v. Davis, 301 U.S. 548, 590 (1937). Congress may use its spending power to create incentives for States to act in accordance with federal policies. But when “pressure turns into compulsion,” ibid., the legislation runs contrary to our system of federalism. “[T]he Constitution simply does not give Congress the authority to require the States to regulate.” New York, 505 U. S., at 178. That is true whether Congress directly commands a State to regulate or indirectly coerces a State to adopt a federal regulatory system as its own. Permitting the Federal Government to force the States to implement a federal program would threaten the political accountability key to our federal system. “[W]here the Federal Government directs the States to regulate, it may be state officials who will bear the brunt of public disapproval, while the federal officials who devised the regu-latory program may remain insulated from the electoral ramifications of their decision.” Id., at 169. Spending Clause programs do not pose this danger when a State has a legitimate choice whether to accept the federal conditions in exchange for federal funds. In such a situation, state officials can fairly be held politically accountable for choosing to accept or refuse the federal offer. But when the State has no choice, the Federal Government can achieve its objectives without accountability, just as in New York and Printz. Indeed, this danger is heightened when Congress acts under the Spending Clause, because Congress can use that power to implement federal policy it could not impose directly under its enumerated powers. We addressed such concerns in Steward Machine. That case involved a federal tax on employers that was abated if the businesses paid into a state unemployment plan that met certain federally specified conditions. An employer sued, alleging that the tax was impermissibly “driv[ing] the state legislatures under the whip of economic pressure into the enactment of unemployment compensation laws at the bidding of the central government.” 301 U. S., at 587. We acknowledged the danger that the Federal Government might employ its taxing power to exert a “power akin to undue influence” upon the States. Id., at 590. But we observed that Congress adopted the challenged tax and abatement program to channel money to the States that would otherwise have gone into the Federal Treasury for use in providing national unemployment services. Congress was willing to direct businesses to instead pay the money into state programs only on the condition that the money be used for the same purposes. Predicating tax abatement on a State’s adoption of a particular type of un-employment legislation was therefore a means to “safeguard [the Federal Government’s] own treasury.” Id., at 591. We held that “[i]n such circumstances, if in no others, inducement or persuasion does not go beyond the bounds of power.” Ibid. In rejecting the argument that the federal law was a “weapon[ ] of coercion, destroying or impairing the autonomy of the states,” the Court noted that there was no reason to suppose that the State in that case acted other than through “her unfettered will.” Id., at 586, 590. Indeed, the State itself did “not offer a suggestion that in passing the unemployment law she was affected by duress.” Id., at 589. As our decision in Steward Machine confirms, Congress may attach appropriate conditions to federal taxing and spending programs to preserve its control over the use of federal funds. In the typical case we look to the States to defend their prerogatives by adopting “the simple expedient of not yielding” to federal blandishments when they do not want to embrace the federal policies as their own. Massachusetts v. Mellon, 262 U.S. 447, 482 (1923). The States are separate and independent sovereigns. Sometimes they have to act like it. The States, however, argue that the Medicaid expansion is far from the typical case. They object that Congress has “crossed the line distinguishing encouragement from coercion,” New York, supra, at 175, in the way it has structured the funding: Instead of simply refusing to grant the new funds to States that will not accept the new conditions, Congress has also threatened to withhold those States’ existing Medicaid funds. The States claim that this threat serves no purpose other than to force unwilling States to sign up for the dramatic expansion in health care coverage effected by the Act. Given the nature of the threat and the programs at issue here, we must agree. We have upheld Congress’s authority to condition the receipt of funds on the States’ complying with restrictions on the use of those funds, because that is the means by which Congress ensures that the funds are spent according to its view of the “general Welfare.” Conditions that do not here govern the use of the funds, however, cannot be justified on that ba- sis. When, for example, such conditions take the form of threats to terminate other significant independent grants, the conditions are properly viewed as a means of pressuring the States to accept policy changes. In South Dakota v. Dole, we considered a challenge to a federal law that threatened to withhold five percent of a State’s federal highway funds if the State did not raise its drinking age to 21. The Court found that the condition was “directly related to one of the main purposes for which highway funds are expended—safe interstate travel.” 483 U. S., at 208. At the same time, the condition was not a restriction on how the highway funds—set aside for specific highway improvement and maintenance efforts—were to be used. We accordingly asked whether “the financial inducement offered by Congress” was “so coercive as to pass the point at which ‘pressure turns into compulsion.’ ” Id., at 211 (quoting Steward Machine, supra, at 590). By “financial inducement” the Court meant the threat of losing five percent of highway funds; no new money was offered to the States to raise their drinking ages. We found that the inducement was not impermissibly coercive, because Congress was offering only “relatively mild encouragement to the States.” Dole, 483 U. S., at 211. We observed that “all South Dakota would lose if she adheres to her chosen course as to a suitable minimum drinking age is 5%” of her highway funds. Ibid. In fact, the federal funds at stake constituted less than half of one percent of South Dakota’s budget at the time. See Nat. Assn. of State Budget Officers, The State Expenditure Report 59 (1987); South Dakota v. Dole, 791 F.2d 628, 630 (CA8 1986). In consequence, “we conclude[d] that [the] encouragement to state action [was] a valid use of the spending power.” Dole, 483 U. S., at 212. Whether to accept the drinking age change “remain[ed] the prerogative of the States not merely in theory but in fact.” Id., at 211–212. In this case, the financial “inducement” Congress has chosen is much more than “relatively mild encouragement”—it is a gun to the head. Section 1396c of the Medicaid Act provides that if a State’s Medicaid plan does not comply with the Act’s requirements, the Secretary of Health and Human Services may declare that “further payments will not be made to the State.” 42 U. S. C. §1396c. A State that opts out of the Affordable Care Act’s expansion in health care coverage thus stands to lose not merely “a relatively small percentage” of its existing Medicaid funding, but all of it. Dole, supra, at 211. Medicaid spending accounts for over 20 percent of the average State’s total budget, with federal funds covering 50 to 83 percent of those costs. See Nat. Assn. of State Budget Officers, Fiscal Year 2010 State Expenditure Report, p. 11, Table 5 (2011); 42 U. S. C. §1396d(b). The Federal Government estimates that it will pay out approximately $3.3 trillion between 2010 and 2019 in order to cover the costs of pre-expansion Medicaid. Brief for United States 10, n. 6. In addition, the States have developed intricate statutory and administrative regimes over the course of many decades to implement their objectives under existing Medicaid. It is easy to see how the Dole Court could conclude that the threatened loss of less than half of one percent of South Dakota’s budget left that State with a “prerogative” to reject Congress’s desired policy, “not merely in theory but in fact.” 483 U. S., at 211–212. The threatened loss of over 10 percent of a State’s overall budget, in contrast, is economic dragooning that leaves the States with no real option but to acquiesce in the Medicaid expansion.[12] Justice Ginsburg claims that Dole is distinguishable because here “Congress has not threatened to withhold funds earmarked for any other program.” Post, at 47. But that begs the question: The States contend that the expansion is in reality a new program and that Congress is forcing them to accept it by threatening the funds for the existing Medicaid program. We cannot agree that existing Medicaid and the expansion dictated by the Affordable Care Act are all one program simply because “Congress styled” them as such. Post, at 49. If the expansion is not properly viewed as a modification of the existing Medicaid program, Congress’s decision to so title it is irrelevant.[13] Here, the Government claims that the Medicaid expansion is properly viewed merely as a modification of the ex-isting program because the States agreed that Congress could change the terms of Medicaid when they signed on in the first place. The Government observes that the Social Security Act, which includes the original Medicaid provisions, contains a clause expressly reserving “[t]he right to alter, amend, or repeal any provision” of that statute. 42 U. S. C. §1304. So it does. But “if Congress intends to impose a condition on the grant of federal moneys, it must do so unambiguously.” Pennhurst, 451 U. S., at 17. A State confronted with statutory language reserving the right to “alter” or “amend” the pertinent provisions of the Social Security Act might reasonably assume that Congress was entitled to make adjustments to the Medicaid program as it developed. Congress has in fact done so, sometimes conditioning only the new funding, other times both old and new. See, e.g., Social Security Amendments of 1972, 86Stat. 1381–1382, 1465 (extending Med-icaid eligibility, but partly conditioning only the new funding); Omnibus Budget Reconciliation Act of 1990, §4601, 104Stat. 1388–166 (extending eligibility, and conditioning old and new funds). The Medicaid expansion, however, accomplishes a shift in kind, not merely degree. The original program was de-signed to cover medical services for four particular cat-egories of the needy: the disabled, the blind, the elderly, and needy families with dependent children. See 42 U. S. C. §1396a(a)(10). Previous amendments to Medicaid eligibility merely altered and expanded the boundaries of these categories. Under the Affordable Care Act, Medicaid is transformed into a program to meet the health care needs of the entire nonelderly population with income below 133 percent of the poverty level. It is no longer a program to care for the neediest among us, but rather an element of a comprehensive national plan to provide universal health insurance coverage.[14] Indeed, the manner in which the expansion is structured indicates that while Congress may have styled the expansion a mere alteration of existing Medicaid, it recognized it was enlisting the States in a new health care program. Congress created a separate funding provision to cover the costs of providing services to any person made newly eligible by the expansion. While Congress pays 50 to 83 percent of the costs of covering individuals currently enrolled in Medicaid, §1396d(b), once the expansion is fully implemented Congress will pay 90 percent of the costs for newly eligible persons, §1396d(y)(1). The conditions on use of the different funds are also distinct. Congress mandated that newly eligible persons receive a level of coverage that is less comprehensive than the traditional Medicaid benefit package. §1396a(k)(1); see Brief for United States 9. As we have explained, “[t]hough Congress’ power to legislate under the spending power is broad, it does not include surprising participating States with postacceptance or ‘retroactive’ conditions.” Pennhurst, supra, at 25. A State could hardly anticipate that Congress’s reservation of the right to “alter” or “amend” the Medicaid program included the power to transform it so dramatically. Justice Ginsburg claims that in fact this expansion is no different from the previous changes to Medicaid, such that “a State would be hard put to complain that it lacked fair notice.” Post, at 56. But the prior change she dis-cusses—presumably the most dramatic alteration she could find—does not come close to working the transformation the expansion accomplishes. She highlights an amendment requiring States to cover pregnant women and increasing the number of eligible children. Ibid. But this modification can hardly be described as a major change in a program that—from its inception—provided health care for “families with dependent children.” Previous Medicaid amendments simply do not fall into the same category as the one at stake here. The Court in Steward Machine did not attempt to “fix the outermost line” where persuasion gives way to coercion. 301 U. S., at 591. The Court found it “[e]nough for present purposes that wherever the line may be, this statute is within it.” Ibid. We have no need to fix a line either. It is enough for today that wherever that line may be, this statute is surely beyond it. Congress may not simply “conscript state [agencies] into the national bureaucratic army,” FERC v. Mississippi, 456 U.S. 742, 775 (1982) (O’Connor, J., concurring in judgment in part and dissenting in part), and that is what it is attempting to do with the Medicaid expansion. B Nothing in our opinion precludes Congress from offering funds under the Affordable Care Act to expand the availability of health care, and requiring that States accepting such funds comply with the conditions on their use. What Congress is not free to do is to penalize States that choose not to participate in that new program by taking away their existing Medicaid funding. Section 1396c gives the Secretary of Health and Human Services the authority to do just that. It allows her to withhold all “further [Medicaid] payments . . . to the State” if she determines that the State is out of compliance with any Medicaid requirement, including those contained in the expansion. 42 U. S. C. §1396c. In light of the Court’s holding, the Secretary cannot apply §1396c to withdraw existing Medicaid funds for failure to comply with the requirements set out in the expansion. That fully remedies the constitutional violation we have identified. The chapter of the United States Code that contains §1396c includes a severability clause confirming that we need go no further. That clause specifies that “[i]f any provision of this chapter, or the application thereof to any person or circumstance, is held invalid, the remainder of the chapter, and the application of such provision to other persons or circumstances shall not be affected thereby.” §1303. Today’s holding does not affect the continued ap-plication of §1396c to the existing Medicaid program. Nor does it affect the Secretary’s ability to withdraw funds pro-vided under the Affordable Care Act if a State that has chosen to participate in the expansion fails to comply with the requirements of that Act. This is not to say, as the joint dissent suggests, that we are “rewriting the Medicaid Expansion.” Post, at 48. Instead, we determine, first, that §1396c is unconstitutional when applied to withdraw existing Medicaid funds from States that decline to comply with the expansion. We then follow Congress’s explicit textual instruction to leave unaffected “the remainder of the chapter, and the application of [the challenged] provision to other persons or circumstances.” §1303. When we invalidate an application of a statute because that application is unconstitutional, we are not “rewriting” the statute; we are merely enforcing the Constitution. The question remains whether today’s holding affects other provisions of the Affordable Care Act. In considering that question, “[w]e seek to determine what Congress would have intended in light of the Court’s constitutional holding.” United States v. Booker, 543 U.S. 220, 246 (2005) (internal quotation marks omitted). Our “touchstone for any decision about remedy is legislative intent, for a court cannot use its remedial powers to circum- vent the intent of the legislature.” Ayotte v. Planned Parenthood of Northern New Eng., 546 U.S. 320, 330 (2006) (internal quotation marks omitted). The question here is whether Congress would have wanted the rest of the Act to stand, had it known that States would have a genuine choice whether to participate in the new Medicaid expansion. Unless it is “evident” that the answer is no, we must leave the rest of the Act intact. Champlin Refining Co. v. Corporation Comm’n of Okla., 286 U.S. 210, 234 (1932). We are confident that Congress would have wanted to preserve the rest of the Act. It is fair to say that Congress assumed that every State would participate in the Medicaid expansion, given that States had no real choice but to do so. The States contend that Congress enacted the rest of the Act with such full participation in mind; they point out that Congress made Medicaid a means for satisfying the mandate, 26 U. S. C. §5000A(f)(1)(A)(ii), and enacted no other plan for providing coverage to many low-income individuals. According to the States, this means that the entire Act must fall. We disagree. The Court today limits the financial pressure the Secretary may apply to induce States to accept the terms of the Medicaid expansion. As a practical matter, that means States may now choose to reject the expansion; that is the whole point. But that does not mean all or even any will. Some States may indeed decline to participate, either because they are unsure they will be able to afford their share of the new funding obligations, or because they are unwilling to commit the administrative resources necessary to support the expansion. Other States, however, may voluntarily sign up, finding the idea of expanding Medicaid coverage attractive, particularly given the level of federal funding the Act offers at the outset. We have no way of knowing how many States will accept the terms of the expansion, but we do not believe Congress would have wanted the whole Act to fall, simply because some may choose not to participate. The other reforms Congress enacted, after all, will remain “fully operative as a law,” Champlin, supra, at 234, and will still function in a way “consistent with Congress’ basic objectives in enacting the statute,” Booker, supra, at 259. Confident that Congress would not have intended anything different, we conclude that the rest of the Act need not fall in light of our constitutional holding. * * * The Affordable Care Act is constitutional in part and unconstitutional in part. The individual mandate cannot be upheld as an exercise of Congress’s power under the Commerce Clause. That Clause authorizes Congress to regulate interstate commerce, not to order individuals to engage in it. In this case, however, it is reasonable to con-strue what Congress has done as increasing taxes on those who have a certain amount of income, but choose to go without health insurance. Such legislation is within Con-gress’s power to tax. As for the Medicaid expansion, that portion of the Affordable Care Act violates the Constitution by threatening existing Medicaid funding. Congress has no authority to order the States to regulate according to its instructions. Congress may offer the States grants and require the States to comply with accompanying conditions, but the States must have a genuine choice whether to accept the offer. The States are given no such choice in this case: They must either accept a basic change in the nature of Medicaid, or risk losing all Medicaid funding. The remedy for that constitutional violation is to preclude the Federal Government from imposing such a sanction. That remedy does not require striking down other portions of the Affordable Care Act. The Framers created a Federal Government of limited powers, and assigned to this Court the duty of enforcing those limits. The Court does so today. But the Court does not express any opinion on the wisdom of the Affordable Care Act. Under the Constitution, that judgment is reserved to the people. The judgment of the Court of Appeals for the Eleventh Circuit is affirmed in part and reversed in part. It is so ordered. Notes 1 The Eleventh Circuit did not consider whether the Anti-Injunction Act bars challenges to the individual mandate. The District Court had determined that it did not, and neither side challenged that holding on appeal. The same was true in the Fourth Circuit, but that court examined the question sua sponte because it viewed the Anti-Injunction Act as a limit on its subject matter jurisdiction. See Liberty Univ., 671 F. 3d, at 400–401. The Sixth Circuit and the D. C. Circuit considered the question but determined that the Anti-Injunction Act did not apply. See Thomas More, 651 F. 3d, at 539–540 (CA6); Seven-Sky, 661 F. 3d, at 5–14 (CADC). 2 We appointed H. Bartow Farr III to brief and argue in support of the Eleventh Circuit’s judgment with respect to severability, and Robert A. Long to brief and argue the proposition that the Anti-Injunction Act bars the current challenges to the individual mandate. 565 U. S. ___ (2011). Both amici have ably discharged their assigned responsibilities. 3 The examples of other congressional mandates cited by Justice Ginsburg, post, at 35, n. 10 (opinion concurring in part, concurring in judgment in part, and dissenting in part), are not to the contrary. Each of those mandates—to report for jury duty, to register for the draft, to purchase firearms in anticipation of militia service, to exchange gold currency for paper currency, and to file a tax return—are based on constitutional provisions other than the Commerce Clause. See Art. I, §8, cl. 9 (to “constitute Tribunals inferior to the supreme Court”); id., cl. 12 (to “raise and support Armies”); id., cl. 16 (to “provide for organizing, arming, and disciplining, the Militia”); id., cl. 5 (to “coin Money”); id., cl. 1 (to “lay and collect Taxes”). 4 Justice Ginsburg suggests that “at the time the Constitution was framed, to ‘regulate’ meant, among other things, to require action.” Post, at 23 (citing Seven-Sky v. Holder, 661 F.3d 1, 16 (CADC 2011); brackets and some internal quotation marks omitted). But to reach this conclusion, the case cited by Justice Ginsburg relied on a dictionary in which “[t]o order; to command” was the fifth-alternative definition of “to direct,” which was itself the second-alternative definition of “to regulate.” See Seven-Sky, supra, at 16 (citing S. Johnson, Dictionary of the English Language (4th ed. 1773) (reprinted 1978)). It is unlikely that the Framers had such an obscure meaning in mind when they used the word “regulate.” Far more commonly, “[t]o regulate” meant “[t]o adjust by rule or method,” which presupposes something to adjust. 2 Johnson, supra, at 1619; see also Gibbons, 9 Wheat., at 196 (defining the commerce power as the power “to prescribe the rule by which commerce is to be governed”). 5 Justice Ginsburg cites two eminent domain cases from the 1890s to support the proposition that our case law does not “toe the activity versus inactivity line.” Post, at 24–25 (citing Monongahela Nav. Co. v. United States, 148 U.S. 312, 335–337 (1893), and Cherokee Nation v. Southern Kansas R. Co., 135 U.S. 641, 657–659 (1890)). The fact that the Fifth Amendment requires the payment of just compensation when the Government exercises its power of eminent domain does not turn the taking into a commercial transaction between the landowner and the Government, let alone a government-compelled transaction between the landowner and a third party. 6 In an attempt to recast the individual mandate as a regulation of commercial activity, Justice Ginsburg suggests that “[a]n individual who opts not to purchase insurance from a private insurer can be seen as actively selecting another form of insurance: self-insurance.” Post, at 26. But “self-insurance” is, in this context, nothing more than a description of the failure to purchase insurance. Individuals are no more “activ[e] in the self-insurance market” when they fail to purchase insurance, ibid., than they are active in the “rest” market when doing nothing. 7 Sotelo, in particular, would seem to refute the joint dissent’s contention that we have “never” treated an exaction as a tax if it was denominated a penalty. Post, at 20. We are not persuaded by the dissent’s attempt to distinguish Sotelo as a statutory construction case from the bankruptcy context. Post, at 17, n. 5. The dissent itself treats the question here as one of statutory interpretation, and indeed also relies on a statutory interpretation case from the bankruptcy context. Post, at 23 (citing United States v. Reorganized CF&I Fabricators of Utah, Inc., 518 U.S. 213, 224 (1996)). 8 In 2016, for example, individuals making $35,000 a year are expected to owe the IRS about $60 for any month in which they do not have health insurance. Someone with an annual income of $100,000 a year would likely owe about $200. The price of a qualifying insurance policy is projected to be around $400 per month. See D. Newman, CRS Report for Congress, Individual Mandate and Related Information Re-quirements Under PPACA 7, and n. 25 (2011). 9 We do not suggest that any exaction lacking a scienter requirement and enforced by the IRS is within the taxing power. See post, at 23–24 (joint opinion of Scalia, Kennedy, Thomas, and Alito, JJ., dissenting). Congress could not, for example, expand its authority to impose criminal fines by creating strict liability offenses enforced by the IRS rather than the FBI. But the fact the exaction here is paid like a tax, to the agency that collects taxes—rather than, for example, exacted by Department of Labor inspectors after ferreting out willful malfeasance—suggests that this exaction may be viewed as a tax. 10 The joint dissent attempts to distinguish New York v. United States on the ground that the seemingly imperative language in that case was in an “introductory provision” that had “no legal consequences.” Post, at 19. We did not rely on that reasoning in New York. See 505 U. S., at 169–170. Nor could we have. While the Court quoted only the broad statement that “[e]ach State shall be responsible” for its waste, that language was implemented through operative provisions that also use the words on which the dissent relies. See 42 U. S. C. §2021e(e)(1) (entitled “Requirements for non-sited compact regions and non-member States” and directing that those entities “shall comply with the following requirements”); §2021e(e)(2) (describing “Penalties for failure to comply”). The Court upheld those provisions not as lawful commands, but as “incentives.” See 505 U. S., at 152–153, 171–173. 11 Of course, individuals do not have a lawful choice not to pay a tax due, and may sometimes face prosecution for failing to do so (although not for declining to make the shared responsibility payment, see 26 U. S. C. §5000A(g)(2)). But that does not show that the tax restricts the lawful choice whether to undertake or forgo the activity on which the tax is predicated. Those subject to the individual mandate may lawfully forgo health insurance and pay higher taxes, or buy health insurance and pay lower taxes. The only thing they may not lawfully do is not buy health insurance and not pay the resulting tax. 12 Justice Ginsburg observes that state Medicaid spending will increase by only 0.8 percent after the expansion. Post, at 43. That not only ignores increased state administrative expenses, but also assumes that the Federal Government will continue to fund the expansion at the current statutorily specified levels. It is not unheard of, however, for the Federal Government to increase requirements in such a manner as to impose unfunded mandates on the States. More importantly, the size of the new financial burden imposed on a State is irrelevant in analyzing whether the State has been coerced into accepting that burden. “Your money or your life” is a coercive proposition, whether you have a single dollar in your pocket or $500. 13 Nor, of course, can the number of pages the amendment occu-pies, or the extent to which the change preserves and works withinthe existing program, be dispositive. Cf. post, at 49–50 (opinion of Ginsburg, J.). Take, for example, the following hypothetical amendment: “All of a State’s citizens are now eligible for Medicaid.” That change would take up a single line and would not alter any “operational aspect[ ] of the program” beyond the eligibility requirements. Post, at 49. Yet it could hardly be argued that such an amendment was a permissible modification of Medicaid, rather than an attempt to foist an entirely new health care system upon the States. 14 Justice Ginsburg suggests that the States can have no objection to the Medicaid expansion, because “Congress could have repealed Medicaid [and,] [t]hereafter, . . . could have enacted Medicaid II, a new program combining the pre-2010 coverage with the expanded coverage required by the ACA.” Post, at 51; see also post, at 38. But it would certainly not be that easy. Practical constraints would plainly inhibit, if not preclude, the Federal Government from repealing the existing program and putting every feature of Medicaid on the table for political reconsideration. Such a massive undertaking would hardly be “ritualistic.” Ibid. The same is true of Justice Ginsburg’s suggestion that Congress could establish Medicaid as an exclusively federal program. Post, at 44.

565.US.2011_10-224

The Federal Meat Inspection Act (FMIA), 21 U. S. C. §601 et seq., regulates a broad range of activities at slaughterhouses to ensure the safety of meat and the humane handling of animals. The Department of Agriculture’s Food Safety and Inspection Service (FSIS), which administers the FMIA, has issued extensive regulations to govern the inspection of animals and meat, as well as other aspects of slaughterhouses’ operations and facilities. See 9 CFR §300.1 et seq. The FSIS inspection procedure begins with an “ante-mortem” inspection of each animal brought to a slaughterhouse. If, at that inspection, a nonambulatory animal is found to suffer from a particularly severe disease or condition, it must be classified as “U. S. Condemned” and killed apart from the slaughtering facilities where food is produced. §§309.3, 311.1 et seq. Nonambulatory animals that are not condemned are classified as “U. S. Suspect.” §309.2(b). They are set apart, specially monitored, and “slaughtered separately from other livestock.” §309.2(n). Following slaughter, an inspector decides at a “post-mortem” examination which parts, if any, of the suspect animal’s carcass may be processed into food for humans. See 9 CFR pts. 310, 311. FSIS regulations additionally prescribe methods for handling animals humanely at all stages of the slaughtering process, 9 CFR pt. 313, including specific provisions for the humane treatment of nonambulatory animals, 9 CFR 313.2(d). The FMIA’s preemption clause, §678, precludes states from imposing requirements that are “within the scope” of the FMIA, relate to slaughterhouse “premises, facilities and operations,” and are “in addition to, or different than those made under” the FMIA. In 2008, California amended its penal code to provide that no slaughterhouse shall “buy, sell, or receive a nonambulatory animal”; “process, butcher, or sell meat or products of nonambulatory animals for human consumption”; or “hold a nonambulatory animal without taking immediate action to humanely euthanize the animal.” §§599f(a)–(c). Petitioner National Meat Association (NMA), a trade association representing meatpackers and processors, sued to enjoin enforcement of §599f against swine slaughterhouses, arguing that the FMIA preempts application of the state law. The District Court agreed, and granted the NMA a preliminary injunction. The Ninth Circuit reversed, holding that §599f is not preempted because it regulates only “the kind of animal that may be slaughtered,” not the inspection or slaughtering process itself. Held: The FMIA expressly preempts §599f’s application against federally inspected swine slaughterhouses. Pp. 6−14. (a) The FMIA’s preemption clause sweeps widely, and so blocks the applications of §599f challenged here. The clause prevents a State from imposing any additional or different―even if nonconflicting―requirements that fall within the FMIA’s scope and concern slaughterhouse facilities or operations. Section 599f imposes additional or different requirements on swine slaughterhouses: Where under federal law a slaughterhouse may take one course of action in handling a nonambulatory pig, under state law the slaughterhouse must take another. For example, when a pig becomes injured and thus nonambulatory sometime after delivery to a slaughterhouse, §599f(c) prohibits the slaughterhouse from “hold[ing]” the pig without immediately euthanizing it; and §599f(b) prohibits the slaughterhouse from “process[ing]” or “butcher[ing]” the animal to make food. By contrast, the FMIA and its regulations allow a slaughterhouse to hold (without euthanizing) any nonambulatory animal that has not been condemned, and to process and butcher such an animal’s meat, subject to an FSIS official’s approval at post-mortem inspection. Similarly, when a pig is nonambulatory at the time of delivery, §599f(a) prohibits a slaughterhouse from “receiv[ing]” or “buy[ing]” the pig. But the FMIA and its regulations expressly allow slaughterhouses to purchase nonambulatory pigs. See 21 U. S. C. §644; 9 CFR §325.20(c). And the FSIS inspection regime clearly contemplates that slaughterhouses will receive nonambulatory animals. So §599f substitutes a new regulatory regime for the one the FMIA prescribes. Respondent Humane Society argues that §599f(a)’s ban on purchasing nonambulatory animals escapes preemption because it would not be preempted if applied to purchases occurring off slaughterhouse premises. But the record does not disclose whether §599f(a) ever applies beyond the slaughterhouse gate, much less how an application of that kind would affect a slaughterhouse’s operations. Moreover, even if the State could regulate off-site purchases, it does not follow that on-site purchases would escape preemption, because the FMIA’s preemption clause expressly focuses on slaughterhouse “premises, facilities and operations.” And while the Humane Society is correct that the FMIA does not normally regulate slaughterhouse sales activities, §599f’s sales ban serves to regulate how slaughterhouses must handle nonambulatory pigs on their premises. Its effect is to make sure that slaughterhouses remove nonambulatory pigs from the production process. It is therefore preempted by the FMIA. Pp. 6−10. (b) Also rejected is the broad argument that §599f’s challenged provisions fall outside the FMIA’s scope because they exclude a class of animals from the slaughtering process, while the FMIA extends only to “animals that are going to be turned into meat.” In fact, the FMIA regulates animals on slaughterhouse premises that will never be turned into meat. For example, the Act’s implementing regulations exclude many classes of animals from the slaughtering process, e.g., swine with hog cholera, 9 CFR §309.5(a). The argument that §599f’s exclusion avoids the FMIA’s scope because it is designed to ensure the humane treatment of pigs, rather than meat safety, misunderstands the FMIA’s scope. The FMIA addresses not just food safety, but humane treatment, as well. See, e.g., 21 U. S. C. §§603, 610(b). Pp. 11−14. 599 F.3d 1093, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court.

The Federal Meat Inspection Act (FMIA or Act), 21 U. S. C. §601 et seq., regulates the inspection, handling, and slaughter of livestock for human consumption. We consider here whether the FMIA expressly preempts a California law dictating what slaughterhouses must do with pigs that cannot walk, known in the trade as nonambulatory pigs. We hold that the FMIA forecloses the challenged applications of the state statute. I A The FMIA regulates a broad range of activities at slaughterhouses to ensure both the safety of meat and the humane handling of animals. [ 1 ] First enacted in 1906, after Upton Sinclair’s muckraking novel The Jungle sparked an uproar over conditions in the meatpacking industry, the Act establishes “an elaborate system of inspecti[ng]” live animals and carcasses in order “to prevent the shipment of impure, unwholesome, and unfit meat and meat-food products.” Pittsburgh Melting Co. v. Totten, 248 U. S. 1 –5 (1918). And since amended in 1978, see 92Stat. 1069, the FMIA requires all slaughterhouses to comply with the standards for humane handling and slaughter of animals set out in the Humane Methods of Slaughter Act of 1958, (HMSA), 72Stat. 862, 7 U. S. C. §1901 et seq., which originally applied only to slaughterhouses selling meat to the Federal Government. The Department of Agriculture’s Food Safety and Inspection Service (FSIS) has responsibility for administering the FMIA to promote its dual goals of safe meat and humane slaughter. Over the years, the FSIS has issued extensive regulations to govern the inspection of animals and meat, as well as other aspects of slaughterhouses’ operations and facilities. See 9 CFR §300.1 et seq. (2011). The FSIS employs about 9,000 inspectors, veterinarians, and investigators to implement its inspection regime and enforce its humane-handling requirements. See Hearings on 2012 Appropriations before the Subcommittee on Agriculture of the House Committee on Appropriations, 112th Cong., 1st Sess., pt. 1B, p. 921 (2011). In fiscal year 2010, those personnel examined about 147 million head of livestock and carried out more than 126,000 “humane handling verification procedures.” Id., at 942–943. The FSIS’s inspection procedure begins with an “ante-mortem” examination of each animal brought to a slaughterhouse. See 9 CFR §309.1. If the inspector finds no evidence of disease or injury, he approves the animal for slaughter. If, at the other end of the spectrum, the inspector sees that an animal is dead or dying, comatose, suffering from a high fever, or afflicted with a serious disease or condition, he designates the animal as “U. S. Condemned.” See §309.3; §311.1 et seq. (listing diseases requiring condemnation). A condemned animal (if not already dead) must be killed apart from the slaughtering facilities where food is produced, and no part of its carcass may be sold for human consumption. See §309.13(a); 21 U. S. C. §610(c). The inspector also has an intermediate option: If he determines that an animal has a less severe condition—or merely suspects the animal of having a disease meriting condemnation—he classifies the animal as “U. S. Suspect.” See 9 CFR §309.2. That category includes all nonambulatory animals not found to require condemnation. [ 2 ] See §309.2(b). Suspect livestock must be “set apart,” specially monitored, and (if not reclassified because of a change in condition) “slaughtered separately from other livestock.” §309.2(n). Following slaughter, an inspector decides at a “post-mortem” examination which parts, if any, of the suspect animal’s carcass may be processed into food for humans. See 9 CFR pts. 310, 311. The regulations implementing the FMIA additionally prescribe methods for handling animals humanely at all stages of the slaughtering process. Those rules apply from the moment a truck carrying livestock “enters, or is in line to enter,” a slaughterhouse’s premises. Humane Handling and Slaughter of Livestock, FSIS Directive 6900.2, ch. II(I) (rev. Aug. 15, 2011). And they include specific provisions for the humane treatment of animals that cannot walk. See 9 CFR §313.2(d). Under the regulations, slaughterhouse employees may not drag conscious, nonambulatory animals, see §313.2(d)(2), and may move them only with “equipment suitable for such purposes,” §313.2(d)(3). Similarly, employees must place nonambulatory animals, as well as other sick and disabled livestock, in covered pens sufficient to protect the animals from “adverse climatic conditions.” See §313.2(d)(1); §313.1(c). The FMIA contains an express preemption provision, at issue here, addressing state laws on these and similar matters. That provision’s first sentence reads: “Requirements within the scope of this [Act] with respect to premises, facilities and operations of any establishment at which inspection is provided under . . . this [Act] which are in addition to, or different than those made under this [Act] may not be imposed by any State.” 21 U. S. C. §678. [ 3 ] B In 2008, the Humane Society of the United States released an undercover video showing workers at a slaughterhouse in California dragging, kicking, and electro-shocking sick and disabled cows in an effort to move them. The video led the Federal Government to institute the largest beef recall in U. S. history in order to prevent consumption of meat from diseased animals. Of greater relevance here, the video also prompted the California legislature to strengthen a pre-existing statute governing the treatment of nonambulatory animals and to apply that statute to slaughterhouses regulated under the FMIA. See National Meat Assn. v. Brown, 599 F. 3d 1093, 1096 (CA9 2010). As amended, the California law—§599f of the state penal code—provides in relevant part: “(a) No slaughterhouse, stockyard, auction, market agency, or dealer shall buy, sell, or receive a nonambulatory animal. “(b) No slaughterhouse shall process, butcher, or sell meat or products of nonambulatory animals for human consumption. “(c) No slaughterhouse shall hold a nonambulatory animal without taking immediate action to humanely euthanize the animal.” Cal. Penal Code Ann. §599f (West 2010). The maximum penalty for violating any of these prohibitions is one year in jail and a $20,000 fine. See §599f(h). Petitioner National Meat Association (NMA) is a trade association representing meatpackers and processors, in-cluding operators of swine slaughterhouses. It sued to enjoin the enforcement of §599f against those slaughterhouses, principally on the ground that the FMIA preempts application of the state law. [ 4 ] The District Court granted the NMA’s motion for a preliminary injunction, reasoning that §599f is expressly preempted because it requires swine “to be handled in a manner other than that prescribed by the FMIA” and its regulations. App. to Pet. for Cert. 36a. But the United States Court of Appeals for the Ninth Circuit vacated the injunction. According to that court, the FMIA does not expressly preempt §599f because the state law regulates only “the kind of animal that may be slaughtered,” and not the inspection or slaughtering process itself. 599 F. 3d, at 1098. We granted certiorari, 564 U. S. __ (2011), and now reverse. II The FMIA’s preemption clause sweeps widely—and in so doing, blocks the applications of §599f challenged here. The clause prevents a State from imposing any additional or different—even if non-conflicting—requirements that fall within the scope of the Act and concern a slaughterhouse’s facilities or operations. And at every turn §599f imposes additional or different requirements on swine slaughterhouses: It compels them to deal with nonambulatory pigs on their premises in ways that the federal Act and regulations do not. In essence, California’s statute substitutes a new regulatory scheme for the one the FSIS uses. Where under federal law a slaughterhouse may take one course of action in handling a nonambulatory pig, under state law the slaughterhouse must take another. Consider first what the two statutes tell a slaughterhouse to do when (as not infrequently occurs) a pig becomes injured and thus nonambulatory sometime after delivery to the slaughterhouse. [ 5 ] Section 599f(c) prohibits the slaughterhouse from “hold[ing]” such an animal “without taking immediate action to humanely euthanize” it. And §599f(b) provides that no part of the animal’s carcass may be “process[ed]” or “butcher[ed]” to make food. By contrast, under the FMIA and its regulations, a slaughterhouse may hold (without euthanizing) any nonambulatory pig that has not been condemned. See supra, at 3. And the slaughterhouse may process or butcher such an animal’s meat for human consumption, subject to an FSIS official’s approval at a post-mortem inspection. See ibid. The State’s proscriptions thus exceed the FMIA’s. To be sure, nothing in the federal Act requires what the state law forbids (or forbids what the state law requires); California is right to note that “[t]he FMIA does not mandate that ‘U. S. Suspect’ [nonambulatory] animals . . . be placed into the human food production process.” Brief for State Respondents 31. But that is irrelevant, because the FMIA’s preemption clause covers not just conflicting, but also different or additional state requirements. It therefore precludes California’s effort in §§599f(b) and (c) to im-pose new rules, beyond any the FSIS has chosen to adopt, on what a slaughterhouse must do with a pig that be-comes nonambulatory during the production process. Similarly, consider how the state and federal laws address what a slaughterhouse should do when a pig is non-ambulatory at the time of delivery, usually because of harsh transportation conditions. [ 6 ] Section 599f(a) of the California law bars a slaughterhouse from “receiv[ing]” or “buy[ing]” such a pig, thus obligating the slaughterhouse to refuse delivery of the animal. [ 7 ] But that directive, too, deviates from any imposed by federal law. A regulation issued under the FMIA specifically authorizes slaughterhouses to buy disabled or diseased animals (including nonambulatory swine), by exempting them from a general prohibition on such purchases. See 9 CFR §325.20(c). And other regulations contemplate that slaughterhouses will in fact take, rather than refuse, receipt of nonambulatory swine. Recall that the FMIA’s regulations provide for the inspection of all pigs at delivery, see supra, at 2—in the case of nonambulatory pigs, often right on the truck, see Humane Handling and Slaughter of Livestock, FSIS Directive 6900.2, ch. II(I). They further instruct slaughterhouses to kill and dispose of any nonambulatory pigs labeled “condemned,” and to slaughter separately those marked “suspect.” See supra, at 3. In short, federal law establishes rules for handling and slaughtering nonam-bulatory pigs brought to a slaughterhouse, rather than ordering them returned to sender. So §599f(a) and the FMIA require different things of a slaughterhouse confronted with a delivery truck containing nonambula- tory swine. The former says “do not receive or buy them”; the latter does not. The Humane Society counters that at least §599f(a)’s ban on buying nonambulatory animals escapes preemption because that provision applies no matter when or where a purchase takes place. The argument proceeds in three steps: (1) §599f(a)’s ban covers purchases of non-ambulatory pigs made prior to delivery, away from the slaughterhouse itself (say, at a farm or auction); (2) the State may regulate such offsite purchases because they do not involve a slaughterhouse’s “premises, facilities and operations,” which is a condition of preemption under the FMIA; and (3) no different result should obtain just because a slaughterhouse structures its swine purchases to occur at delivery, on its own property. See Brief for Non-State Respondents 43–45. But this argument fails on two grounds. First, its preliminary steps have no foundation in the record. Until a stray comment at oral argument, see Tr. of Oral Arg. 50, neither the State nor the Humane Society had disputed the NMA’s assertion that slaughterhouses buy pigs at delivery (or still later, upon successful ante-mortem inspection). See Brief for Petitioner 46, n. 18; Brief for Non-State Respondents 44; Brief for State Respondents 16, n. 5. Nor had the parties presented evidence that a significant number of pigs become nonambulatory before shipment, when any offsite purchases would occur. The record therefore does not disclose whether §599f(a)’s ban on purchase ever applies beyond the slaughterhouse gate, much less how an application of that kind would affect a slaughterhouse’s operations. And because that is so, we have no basis for deciding whether the FMIA would preempt it. Second, even assuming that a State could regulate offsite purchases, the concluding step of the Humane Society’s argument would not follow. The FMIA’s preemption clause expressly focuses on “premises, facilities and operations”—at bottom, the slaughtering and processing of animals at a given location. So the distinction between a slaughterhouse’s site-based activities and its more far-flung commercial dealings is not, as the Humane Society contends, an anomaly that courts should strain to avoid. It is instead a fundamental feature of the FMIA’s preemption clause. For that reason, the Humane Society’s stronger argument concerns California’s effort to regulate the last stage of a slaughterhouse’s business—the ban in §599f(b) on “sell[ing] meat or products of nonambulatory animals for human consumption.” The Government acknowledges that the FMIA’s preemption clause does not usually foreclose “state regulation of the commercial sales activities of slaughterhouses.” Brief for United States as Amicus Curiae 17. And the Humane Society asserts, in line with that general rule, that §599f(b)’s ban on sales does not regulate a slaughterhouse’s “operations” because it kicks in only after they have ended: Once meat from a slaughtered pig has passed a post-mortem inspection, the Act “is not concerned with whether or how it is ever actually sold.” Brief for Non-State Respondents 45. At most, the Humane Society claims, §599f(b)’s ban on sales offers an “incentiv[e]” to a slaughterhouse to take nonambulatory pigs out of the meat production process. Id., at 46. And California may so “motivate[]” an operational choice without running afoul of the FMIA’s preemption provision. Ibid. (quoting Bates v. Dow Agrosciences LLC, 544 U. S. 431, 443 (2005) ). But this argument mistakes how the prohibition on sales operates within §599f as a whole. The sales ban is a criminal proscription calculated to help implement and enforce each of the section’s other regulations—its prohibition of receipt and purchase, its bar on butchering and processing, and its mandate of immediate euthanasia. The idea—and the inevitable effect—of the provision is to make sure that slaughterhouses remove nonambulatory pigs from the production process (or keep them out of the process from the beginning) by criminalizing the sale of their meat. That, we think, is something more than an “incentive” or “motivat[or]”; the sales ban instead functions as a command to slaughterhouses to structure their operations in the exact way the remainder of §599f mandates. And indeed, if the sales ban were to avoid the FMIA’s preemption clause, then any State could impose any regulation on slaughterhouses just by framing it as a ban on the sale of meat produced in whatever way the State disapproved. That would make a mockery of the FMIA’s preemption provision. Cf. Engine Mfrs. Assn. v. South Coast Air Quality Management Dist., 541 U. S. 246, 255 (2004) (stating that it “would make no sense” to allow state regulations to escape preemption because they addressed the purchase, rather than manufacture, of a federally regulated product). Like the rest of §599f, the sales ban regulates how slaughterhouses must deal with non-ambulatory pigs on their premises. The FMIA therefore preempts it for all the same reasons. III California’s and the Humane Society’s broadest argument against preemption maintains that all of §599f’s challenged provisions fall outside the “scope” of the FMIA because they exclude a class of animals from the slaughtering process. See 21 U. S. C. §678 (preempting certain requirements “within the scope of this [Act]”). According to this view, the Act (and the FSIS’s authority under it) extends only to “animals that are going to be turned into meat,” Tr. of Oral Arg. 28—or to use another phrase, animals that will “be slaughtered . . . for purposes of human food production,” Brief for State Respondents 19 (emphasis deleted). Section 599f avoids the scope of the Act, respondents claim, by altogether removing nonambulatory pigs from the slaughtering process. [ 8 ] The Ninth Circuit accepted this argument, analogizing §599f to state laws upheld in two other Circuits banning the slaughter of horses for human consumption. 599 F. 3d, at 1098 (discussing Cavel Int’l., Inc. v. Madigan, 500 F. 3d 551 (CA7 2007), and Empacadora de Carnes de Fresnillo, S. A. de C.V. v. Curry, 476 F. 3d 326 (CA5 2007)). According to the Court of Appeals, “states are free to decide which animals may be turned into meat.” 599 F. 3d, at 1098, 1099. We think not. The FMIA’s scope includes not only “animals that are going to be turned into meat,” but animals on a slaughterhouse’s premises that will never suffer that fate. The Act’s implementing regulations themselves exclude many classes of animals from the slaughtering process. Swine with hog cholera, for example, are disqualified, see 9 CFR §309.5(a); so too are swine and other livestock “affected with anthrax,” §309.7(a). Indeed, the federal regulations prohibit the slaughter of any nonambulatory cattle for human consumption. See §309.3(e). As these examples demonstrate, one vital function of the Act and its regulations is to ensure that some kinds of livestock delivered to a slaughterhouse’s gates will not be turned into meat. Under federal law, nonambulatory pigs are not among those excluded animals. But that is to say only that §599f’s requirements differ from those of the FMIA—not that §599f’s requirements fall outside the FMIA’s scope. Nor are respondents right to suggest that §599f’s exclusion avoids the FMIA’s scope because it is designed to ensure the humane treatment of pigs, rather than the safety of meat. See, e.g., Brief for State Respondents 29; Brief for Non-State Respondents 39–40. That view misunderstands the authority—and indeed responsibility—that the FMIA gives to federal officials. Since 1978, when Congress incorporated the HMSA’s standards, the FMIA has required slaughterhouses to follow prescribed methods of humane handling, so as to minimize animals’ pain and suffering. See 21 U. S. C. §§603(b), 610(b); supra, at 2–4. A violation of those standards is a crime, see §676, and the Secretary of Agriculture can suspend inspections at—and thus effectively shut down—a slaughterhouse that dis-obeys them, see §§603(b), 610(c). To implement the Act’s humane-handling provisions, the FSIS has issued detailed regulations, see 9 CFR pt. 313, including some specifically addressing animals that cannot walk, see §§313.2(d), 313.1(c). Those rules, as earlier noted, apply throughout the time an animal is on a slaughterhouse’s premises, from the moment a delivery truck pulls up to the gate. See supra, at 3–4. So the FMIA addresses not just food safety, but humane treatment as well. Even California conceded at oral argument that the FSIS could issue regulations under the FMIA, similar to §599f, mandating the euthanasia of nonambulatory swine. [ 9 ] See Tr. of Oral Arg. 46–47. If that is so—and it is, because of the FSIS’s authority over humane-handling methods—then §599f’s requirements must fall within the FMIA’s scope. The Circuit decisions upholding state bans on slaughtering horses, on which the Ninth Circuit relied, do not demand any different conclusion. We express no view on those decisions, except to say that the laws sustained there differ from §599f in a significant respect. A ban on butchering horses for human consumption works at a remove from the sites and activities that the FMIA most directly governs. When such a ban is in effect, no horses will be delivered to, inspected at, or handled by a slaughterhouse, because no horses will be ordered for purchase in the first instance. But §599f does not and cannot work in that way. As earlier noted, many nonambulatory pigs become disabled either in transit to or after arrival at a slaughterhouse. See supra, at 6–9, and nn. 5–6. So even with §599f in effect, a swine slaughterhouse will encounter nonambulatory pigs. In that circumstance, §599f tells the slaughterhouse what to do with those animals. Unlike a horse slaughtering ban, the statute thus reaches into the slaughterhouse’s facilities and affects its daily activities. And in so doing, the California law runs smack into the FMIA’s regulations. So whatever might be said of other bans on slaughter, §599f imposes requirements within—and indeed at the very heart of—the FMIA’s scope. [ 10 ] IV The FMIA regulates slaughterhouses’ handling and treatment of nonambulatory pigs from the moment of their delivery through the end of the meat production process. California’s §599f endeavors to regulate the same thing, at the same time, in the same place—except by imposing different requirements. The FMIA expressly preempts such a state law. Accordingly, we reverse the judgment of the Ninth Circuit, and remand this case for further proceedings consistent with this opinion. It is so ordered. Notes 1 The FMIA applies to all slaughterhouses producing meat for in-terstate and foreign commerce. See 21 U. S. C. §§601(h), 603(a). The FMIA also regulates slaughterhouses serving an exclusively intrastate market in any State that does not administer an inspection system with “requirements at least equal to those” of the Act. §661(c)(1). Because California has chosen not to adopt such an inspection program, the FMIA governs all slaughterhouses in the State (except for any limited to “custom slaughtering for personal, household, guest, and employee uses,” §623(a)). 2 The FSIS’s regulations define “non-ambulatory disabled livestock” as “livestock that cannot rise from a recumbent position or that cannot walk, including, but not limited to, those with broken appendages, severed tendons or ligaments, nerve paralysis, fractured vertebral column, or metabolic conditions.” §309.2(b). 3 The preemption provision also includes a saving clause, which states that the Act “shall not preclude any State . . . from making requirement[s] or taking other action, consistent with this [Act], with respect to any other matters regulated under this [Act].” ; see n. 10, infra. 4 The Humane Society intervened to defend §599f in the District Court. See Motion to Intervene in No. 08–1963 (ED Cal.), Record, Doc. 46. The organization continues as a respondent in this Court. 5 The percentage of pigs becoming nonambulatory after delivery varies by slaughterhouse from 0.1 percent to over 1 percent. See McGlone, Fatigued Pigs: The Final Link, Pork Magazine 14 (Mar. 2006). About 100 million pigs are slaughtered each year in the United States, see Dept. of Agriculture, National Agricultural Statistics Service, Livestock Slaughter 13 (Jan. 2011), so those percentages work out to between 100,000 and 1,000,000 pigs. 6 According to one estimate, almost half of one percent of the pigs slaughtered annually in the United States become nonambulatory during the trip from farm to slaughterhouse. See National Pork Board, Transport Quality Assurance Handbook 25 (Version 4, 2010). About half that many die during transport. See ibid. 7 Section 599f(a) also bans “sell[ing]” nonambulatory animals. But because slaughterhouses (unlike other entities referenced in the provision) do not typically sell live animals, that prohibition is not at issue in this case. The statute’s distinct ban on selling meat from nonambula-tory animals that have been slaughtered is discussed infra, at 9–10. 8 California’s brief sometimes casts its argument in terms of the “operations” language of the FMIA’s preemption clause (although the State appeared to abandon this phrasing at oral argument). In this version of the claim, California contends that the “operations” of a slaughterhouse are only those “of federal concern,” and that excluding a class of animals from the slaughtering process does not impinge on such operations. Brief for State Respondents 20, n. 9; see also id., at 20–21. We see no real difference between saying that a categorical exclusion of animals does not implicate “operations of federal concern” and saying that it does not fall within the scope of the Act. Accordingly, our answer to both forms of the argument is the same. 9 Indeed, the FSIS recently solicited comment on a rulemaking petition that would require all nonambulatory disabled livestock, including swine, to be humanely euthanized. See 76 Fed. Reg. 6572 (2011). The FSIS has taken no further action on that petition. 10 We finally reject California’s argument, see Brief for State Respondents 20, that our reading of the FMIA’s preemption provision renders its saving clause insignificant. That clause provides that States may regulate slaughterhouses as to “other matters,” not addressed in the express preemption clause, as long as those laws are “consistent with” the FMIA. . So, for example, the Government acknowledges that state laws of general application (workplace safety regulations, building codes, etc.) will usually apply to slaughterhouses. See Tr. of Oral Arg. 22. Moreover, because the FMIA’s express preemption provision prevents States from imposing only “addition[al]” or “different” requirements, §678, States may exact civil or criminal penalties for animal cruelty or other conduct that also violates the FMIA. See §678; cf. Bates v. Dow Agrosciences, LLC, (holding that a preemption clause barring state laws “in addition to or different” from a federal Act does not interfere with an “equivalent” state provision). Although the FMIA preempts much state law involving slaughterhouses, it thus leaves some room for the States to regulate.

565.US.2011_10-8974

Around 3 a.m. on August 15, 2008, the Nashua, New Hampshire Police Department received a call reporting that an African-American male was trying to break into cars parked in the lot of the caller’s apartment building. When an officer responding to the call asked eyewitness Nubia Blandon to describe the man, Blandon pointed to her kitchen window and said the man she saw breaking into the car was standing in the parking lot, next to a police officer. Petitioner Barion Perry’s arrest followed this identification. Before trial, Perry moved to suppress Blandon’s identification on the ground that admitting it at trial would violate due process. The New Hampshire trial court denied the motion. To determine whether due process prohibits the introduction of an out-of-court identification at trial, the Superior Court said, this Court’s decisions instruct a two-step inquiry: The trial court must first decide whether the police used an unnecessarily suggestive identification procedure; if they did, the court must next consider whether that procedure so tainted the resulting identification as to render it unreliable and thus inadmissible. Perry’s challenge, the court found, failed at step one, for Blandon’s identification did not result from an unnecessarily suggestive procedure employed by the police. A jury subsequently convicted Perry of theft by unauthorized taking. On appeal, Perry argued that the trial court erred in requiring an initial showing that police arranged a suggestive identification procedure. Suggestive circumstances alone, Perry contended, suffice to require court evaluation of the reliability of an eyewitness identification before allowing it to be presented to the jury. The New Hampshire Supreme Court rejected Perry’s argument and affirmed his conviction. Held: The Due Process Clause does not require a preliminary judicial inquiry into the reliability of an eyewitness identification when the identification was not procured under unnecessarily suggestive circumstances arranged by law enforcement. Pp. 6–19. (a) The Constitution protects a defendant against a conviction based on evidence of questionable reliability, not by prohibiting introduction of the evidence, but by affording the defendant means to persuade the jury that the evidence should be discounted as unworthy of credit. Only when evidence “is so extremely unfair that its admission violates fundamental conceptions of justice,” Dowling v. United States, 493 U.S. 342, 352 (internal quotation marks omitted), does the Due Process Clause preclude its admission. Contending that the Due Process Clause is implicated here, Perry relies on a series of decisions involving police-arranged identification procedures. See Stovall v. Denno, 388 U.S. 293; Simmons v. United States, 390 U.S. 377; Foster v. California, 394 U.S. 440; Neil v. Biggers, 409 U.S. 188; and Manson v. Brathwaite, 432 U.S. 98. These cases detail the approach appropriately used to determine whether due process requires suppression of an eyewitness identification tainted by police arrangement. First, due process concerns arise only when law enforcement officers use an identification procedure that is both suggestive and unnecessary. Id., at 107, 109; Biggers, 409 U. S., at 198. Even when the police use such a procedure, however, suppression of the resulting identification is not the inevitable consequence. Brathwaite, 432 U. S., at 112–113; Biggers, 409 U. S., at 198–199. Instead, due process requires courts to assess, on a case-by-case basis, whether improper police conduct created a “substantial likelihood of misidentification.” Id., at 201. “[R]eliability [of the eyewitness identification] is the linchpin” of that evaluation. Brathwaite, 432 U. S., at 114. Where the “indicators of [a witness’] ability to make an accurate identification” are “outweighed by the corrupting effect” of law enforcement suggestion, the identification should be suppressed. Id., at 114, 116. Otherwise, the identification, assuming no other barrier to its admission, should be submitted to the jury. Pp. 6–10. (b) Perry argues that it was mere happenstance that all of the cases in the Stovall line involved improper police action. The rationale underlying this Court’s decisions, Perry asserts, calls for a rule requiring trial judges to prescreen eyewitness evidence for reliability any time an identification is made under suggestive circumstances. This Court disagrees. If “reliability is the linchpin” of admissibility under the Due Process Clause, Brathwaite, 432 U. S., at 114, Perry contends, it should not matter whether law enforcement was responsible for creating the suggestive circumstances that marred the identification. This argument removes Brathwaite’s statement from its mooring, attributing to it a meaning that a fair reading of the opinion does not bear. The due process check for reliability, Brathwaite made plain, comes into play only after the defendant establishes improper police conduct. Perry’s contention also ignores a key premise of Brathwaite: A primary aim of excluding identification evidence obtained under unnecessarily suggestive circumstances is to deter law enforcement use of improper procedures in the first place. This deterrence rationale is inapposite in cases, like Perry’s, where there is no improper police conduct. Perry also places significant weight on United States v. Wade, 388 U.S. 218, describing it as a decision not anchored to improper police conduct. But the risk of police rigging was the very danger that prompted the Court in Wade to extend a defendant’s right to counsel to cover postindictment lineups and showups. Perry’s position would also open the door to judicial preview, under the banner of due process, of most, if not all, eyewitness identifications. There is no reason why an identification made by an eyewitness with poor vision or one who harbors a grudge against the defendant, for example, should be regarded as inherently more reliable than Blandon’s identification here. Even if this Court could, as Perry contends, distinguish “suggestive circumstances” from other factors bearing on the reliability of eyewitness evidence, Perry’s limitation would still involve trial courts, routinely, in preliminary examinations, for most eyewitness identifications involve some element of suggestion. Pp. 10–14. (c) In urging a broadly applicable rule, Perry maintains that eyewitness identifications are uniquely unreliable. The fallibility of eyewitness evidence does not, without the taint of improper state conduct, warrant a due process rule requiring a trial court to screen the evidence for reliability before allowing the jury to assess its creditworthiness. The Court’s unwillingness to adopt such a rule rests, in large part, on its recognition that the jury, not the judge, traditionally determines the reliability of evidence. It also takes account of other safeguards built into the adversary system that caution juries against placing undue weight on eyewitness testimony of questionable reliability. These protections include the defendant’s Sixth Amendment rights to counsel and to confront and cross-examine the eyewitness, eyewitness-specific instructions warning juries to take care in appraising identification evidence, and state and federal rules of evidence permitting trial judges to exclude relevant evidence if its probative value is substantially outweighed by its prejudicial impact or potential for misleading the jury. Many of these safeguards were availed of by Perry’s defense. Given the safeguards generally applicable in criminal trials, the introduction of Blandon’s eyewitness testimony, without a preliminary judicial assessment of its reliability, did not render Perry’s trial fundamentally unfair. Pp. 14–18. Affirmed. Ginsburg, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Thomas, Breyer, Alito, and Kagan, JJ., joined. Thomas, J., filed a concurring opinion. Sotomayor, J., filed a dissenting opinion.

In our system of justice, fair trial for persons charged with criminal offenses is secured by the Sixth Amendment, which guarantees to defendants the right to counsel, compulsory process to obtain defense witnesses, and the opportunity to cross-examine witnesses for the prosecution. Those safeguards apart, admission of evidence in state trials is ordinarily governed by state law, and the reliability of relevant testimony typically falls within the province of the jury to determine. This Court has recognized, in addition, a due process check on the admission of eyewitness identification, applicable when the police have arranged suggestive circumstances leading the witness to identify a particular person as the perpetrator of a crime. An identification infected by improper police influence, our case law holds, is not automatically excluded. Instead, the trial judge must screen the evidence for reliability pretrial. If there is “a very substantial likelihood of irreparable misidentification,” Simmons v. United States, 390 U. S. 377, 384 (1968) , the judge must disallow presentation of the evidence at trial. But if the indicia of reliability are strong enough to outweigh the corrupting effect of the police-arranged suggestive circumstances, the identification evidence ordinarily will be admitted, and the jury will ultimately determine its worth. We have not extended pretrial screening for reliability to cases in which the suggestive circumstances were not arranged by law enforcement officers. Petitioner requests that we do so because of the grave risk that mistaken identification will yield a miscarriage of justice. [ 1 ] Our decisions, however, turn on the presence of state action and aim to deter police from rigging identification procedures, for example, at a lineup, showup, or photograph array. When no improper law enforcement activity is involved, we hold, it suffices to test reliability through the rights and opportunities generally designed for that purpose, notably, the presence of counsel at postindictment lineups, vigorous cross-examination, protective rules of evi- dence, and jury instructions on both the fallibility of eyewitness identification and the requirement that guilt be proved beyond a reasonable doubt. I A Around 3 a.m. on August 15, 2008, Joffre Ullon called the Nashua, New Hampshire, Police Department and reported that an African-American male was trying to break into cars parked in the lot of Ullon’s apartment building. Officer Nicole Clay responded to the call. Upon arriving at the parking lot, Clay heard what “sounded like a metal bat hitting the ground.” App. 37a–38a. She then saw petitioner Barion Perry standing between two cars. Perry walked toward Clay, holding two car-stereo amplifiers in his hands. A metal bat lay on the ground behind him. Clay asked Perry where the amplifiers came from. “[I] found them on the ground,” Perry responded. Id., at 39a. Meanwhile, Ullon’s wife, Nubia Blandon, woke her neighbor, Alex Clavijo, and told him she had just seen someone break into his car. Clavijo immediately went downstairs to the parking lot to inspect the car. He first observed that one of the rear windows had been shattered. On further inspection, he discovered that the speakers and amplifiers from his car stereo were missing, as were his bat and wrench. Clavijo then approached Clay and told her about Blandon’s alert and his own subsequent observations. By this time, another officer had arrived at the scene. Clay asked Perry to stay in the parking lot with that officer, while she and Clavijo went to talk to Blandon. Clay and Clavijo then entered the apartment building and took the stairs to the fourth floor, where Blandon’s and Clavijo’s apartments were located. They met Blandon in the hallway just outside the open door to her apartment. Asked to describe what she had seen, Blandon stated that, around 2:30 a.m., she saw from her kitchen window a tall, African-American man roaming the parking lot and looking into cars. Eventually, the man circled Clavijo’s car, opened the trunk, and removed a large box. [ 2 ] Clay asked Blandon for a more specific description of the man. Blandon pointed to her kitchen window and said the person she saw breaking into Clavijo’s car was standing in the parking lot, next to the police officer. Perry’s arrest followed this identification. About a month later, the police showed Blandon a photographic array that included a picture of Perry and asked her to point out the man who had broken into Clavijo’s car. Blandon was unable to identify Perry. B Perry was charged in New Hampshire state court with one count of theft by unauthorized taking and one count of criminal mischief. [ 3 ] Before trial, he moved to suppress Blandon’s identification on the ground that admitting it at trial would violate due process. Blandon witnessed what amounted to a one-person showup in the parking lot, Perry asserted, which all but guaranteed that she would identify him as the culprit. Id., at 15a–16a. The New Hampshire Superior Court denied the motion. Id., at 82a–88a. To determine whether due process prohibits the introduction of an out-of-court identification at trial, the Superior Court said, this Court’s decisions instruct a two-step inquiry. First, the trial court must decide whether the police used an unnecessarily suggestive identification procedure. Id., at 85a. If they did, the court must next consider whether the improper identification procedure so tainted the resulting identification as to render it unreliable and therefore inadmissible. Ibid. (citing Neil v. Biggers, 409 U. S. 188 (1972) , and Manson v. Brathwaite, 432 U. S. 98 (1977) ). Perry’s challenge, the Superior Court concluded, failed at step one: Blandon’s identification of Perry on the night of the crime did not result from an unnecessarily suggestive procedure “manufacture[d] . . . by the police.” App. 86a–87a. Blandon pointed to Perry “spontaneously,” the court noted, “without any inducement from the police.” Id., at 85a–86a. Clay did not ask Blandon whether the man standing in the parking lot was the man Blandon had seen breaking into Clavijo’s car. Ibid. Nor did Clay ask Blandon to move to the window from which she had observed the break-in. Id., at 86a. The Superior Court recognized that there were reasons to question the accuracy of Blandon’s identification: the parking lot was dark in some locations; Perry was standing next to a police officer; Perry was the only African-American man in the vicinity; and Blandon was unable, later, to pick Perry out of a photographic array. Id., at 86a–87a. But “[b]ecause the police procedures were not unnecessarily suggestive,” the court ruled that the reliability of Blandon’s testimony was for the jury to consider. Id., at 87a. At the ensuing trial, Blandon and Clay testified to Blandon’s out-of-court identification. The jury found Perry guilty of theft and not guilty of criminal mischief. On appeal, Perry repeated his challenge to the admissibility of Blandon’s out-of-court identification. The trial court erred, Perry contended, in requiring an initial showing that the police arranged the suggestive identification procedure. Suggestive circumstances alone, Perry argued, suffice to trigger the court’s duty to evaluate the reliability of the resulting identification before allowing presentation of the evidence to the jury. The New Hampshire Supreme Court rejected Perry’s argument and affirmed his conviction. Id., at 9a–11a. Only where the police employ suggestive identification techniques, that court held, does the Due Process Clause require a trial court to assess the reliability of identification evidence before permitting a jury to consider it. Id., at 10a–11a. We granted certiorari to resolve a division of opinion on the question whether the Due Process Clause requires a trial judge to conduct a preliminary assessment of the reliability of an eyewitness identification made under suggestive circumstances not arranged by the police. 563 U. S. ___ (2011). [ 4 ] II A The Constitution, our decisions indicate, protects a de- fendant against a conviction based on evidence of questionable reliability, not by prohibiting introduction of the evidence, but by affording the defendant means to persuade the jury that the evidence should be discounted as unworthy of credit. Constitutional safeguards available to defendants to counter the State’s evidence include the Sixth Amendment rights to counsel, Gideon v. Wainwright, 372 U. S. 335 –345 (1963); compulsory process, Taylor v. Illinois, 484 U. S. 400 –409 (1988); and confrontation plus cross-examination of witnesses, Delaware v. Fensterer, 474 U. S. 15 –20 (1985) (per curiam). Apart from these guarantees, we have recognized, state and federal statutes and rules ordinarily govern the admissibility of evidence, and juries are assigned the task of determining the reliability of the evidence presented at trial. See Kansas v. Ventris, 556 U. S. 586 , n. (2009) (“Our legal system . . . is built on the premise that it is the province of the jury to weigh the credibility of competing witnesses.”). Only when evidence “is so extremely unfair that its admission violates fundamental conceptions of justice,” Dowling v. United States, 493 U. S. 342, 352 (1990) (internal quotation marks omitted), have we imposed a constraint tied to the Due Process Clause. See, e.g., Napue v. Illinois, 360 U. S. 264, 269 (1959) (Due process prohibits the State’s “knowin[g] use [of] false evidence,” because such use violates “any concept of ordered liberty.”). Contending that the Due Process Clause is implicated here, Perry relies on a series of decisions involving police-arranged identification procedures. In Stovall v. Denno, 388 U. S. 293 (1967) , first of those decisions, a witness identified the defendant as her assailant after police officers brought the defendant to the witness’ hospital room. Id., at 295. At the time the witness made the identification, the defendant—the only African-American in the room—was handcuffed and surrounded by police officers. Ibid. Although the police-arranged showup was undeniably suggestive, the Court held that no due process violation occurred. Id., at 302. Crucial to the Court’s decision was the procedure’s necessity: The witness was the only person who could identify or exonerate the defendant; the witness could not leave her hospital room; and it was uncertain whether she would live to identify the defendant in more neutral circumstances. Ibid. A year later, in Simmons v. United States, 390 U. S. 377 (1968) , the Court addressed a due process challenge to police use of a photographic array. When a witness identifies the defendant in a police-organized photo lineup, the Court ruled, the identification should be suppressed only where “the photographic identification procedure was so [unnecessarily] suggestive as to give rise to a very substantial likelihood of irreparable misidentification.” Id., at 384–385. Satisfied that the photo array used by Federal Bureau of Investigation agents in Simmons was both necessary and unlikely to have led to a mistaken identification, the Court rejected the defendant’s due process challenge to admission of the identification. Id., at 385–386. In contrast, the Court held in Foster v. California, 394 U. S. 440 (1969) , that due process required the exclusion of an eyewitness identification obtained through police-arranged procedures that “made it all but inevitable that [the witness] would identify [the defendant].” Id., at 443. Synthesizing previous decisions, we set forth in Neil v. Biggers, 409 U. S. 188 (1972) , and reiterated in Manson v. Brathwaite, 432 U. S. 98 (1977) , the approach appropri- ately used to determine whether the Due Process Clause requires suppression of an eyewitness identification tainted by police arrangement. The Court emphasized, first, that due process concerns arise only when law enforcement officers use an identification procedure that is both suggestive and unnecessary. Id., at 107, 109; Biggers, 409 U. S., at 198. Even when the police use such a procedure, the Court next said, suppression of the resulting identification is not the inevitable consequence. Brathwaite, 432 U. S., at 112–113; Biggers, 409 U. S., at 198–199. A rule requiring automatic exclusion, the Court reasoned, would “g[o] too far,” for it would “kee[p] evidence from the jury that is reliable and relevant,” and “may result, on occasion, in the guilty going free.” Brathwaite, 432 U. S., at 112; see id., at 113 (when an “identification is reliable despite an unnecessarily suggestive [police] identification procedure,” automatic exclusion “is a Draconian sanction,” one “that may frustrate rather than promote justice”). Instead of mandating a per se exclusionary rule, the Court held that the Due Process Clause requires courts to assess, on a case-by-case basis, whether improper police conduct created a “substantial likelihood of misidentifi- cation.” Biggers, 409 U. S., at 201; see Brathwaite, 432 U. S., at 116. “[R]eliability [of the eyewitness identification] is the linchpin” of that evaluation, the Court stated in Brathwaite. Id., at 114. Where the “indicators of [a witness’] ability to make an accurate identification” are “outweighed by the corrupting effect” of law enforcement suggestion, the identification should be suppressed. Id., at 114, 116. Otherwise, the evidence (if admissible in all other respects) should be submitted to the jury. [ 5 ] Applying this “totality of the circumstances” approach, id., at 110, the Court held in Biggers that law enforcement’s use of an unnecessarily suggestive showup did not require suppression of the victim’s identification of her assailant. 409 U. S., at 199–200. Notwithstanding the improper procedure, the victim’s identification was reliable: She saw her assailant for a considerable period of time under adequate light, provided police with a detailed de- scription of her attacker long before the showup, and had “no doubt” that the defendant was the person she had seen. Id., at 200 (internal quotation marks omitted). Similarly, the Court concluded in Brathwaite that police use of an unnecessarily suggestive photo array did not require exclusion of the resulting identification. 432 U. S., at 114–117. The witness, an undercover police officer, viewed the defendant in good light for several minutes, provided a thorough description of the suspect, and was certain of his identification. Id., at 115. Hence, the “indicators of [the witness’] ability to make an accurate identification [were] hardly outweighed by the corrupting effect of the challenged identification.” Id., at 116. B Perry concedes that, in contrast to every case in the Stovall line, law enforcement officials did not arrange the suggestive circumstances surrounding Blandon’s identification. See Brief for Petitioner 34; Tr. of Oral Arg. 5 (counsel for Perry) (“[W]e do not allege any manipulation or intentional orchestration by the police.”). He contends, however, that it was mere happenstance that each of the Stovall cases involved improper police action. The rationale underlying our decisions, Perry asserts, supports a rule requiring trial judges to prescreen eyewitness evidence for reliability any time an identification is made under suggestive circumstances. We disagree. Perry’s argument depends, in large part, on the Court’s statement in Brathwaite that “reliability is the linchpin in determining the admissibility of identification testimony.” 432 U. S., at 114. If reliability is the linchpin of admissibility under the Due Process Clause, Perry maintains, it should make no difference whether law enforcement was responsible for creating the suggestive circumstances that marred the identification. Perry has removed our statement in Brathwaite from its mooring, and thereby attributes to the statement a meaning a fair reading of our opinion does not bear. As just explained, supra, at 8–9, the Brathwaite Court’s reference to reliability appears in a portion of the opinion concerning the appropriate remedy when the police use an unnecessarily suggestive identification procedure. The Court adopted a judicial screen for reliability as a course preferable to a per se rule requiring exclusion of identification evidence whenever law enforcement officers employ an improper procedure. The due process check for reliability, Brathwaite made plain, comes into play only after the defendant establishes improper police conduct. The very purpose of the check, the Court noted, was to avoid depriving the jury of identification evidence that is reliable, notwithstanding improper police conduct. 432 U. S., at 112–113. [ 6 ] Perry’s contention that improper police action was not essential to the reliability check Brathwaite required is echoed by the dissent. Post, at 3–4. Both ignore a key premise of the Brathwaite decision: A primary aim of ex- cluding identification evidence obtained under unnecessarily suggestive circumstances, the Court said, is to deter law enforcement use of improper lineups, showups, and photo arrays in the first place. See 432 U. S., at 112. Alerted to the prospect that identification evidence improperly obtained may be excluded, the Court reasoned, police officers will “guard against unnecessarily suggestive procedures.” Ibid. This deterrence rationale is inapposite in cases, like Perry’s, in which the police engaged in no improper conduct. Coleman v. Alabama, 399 U. S. 1 (1970) , another decision in the Stovall line, similarly shows that the Court has linked the due process check, not to suspicion of eyewitness testimony generally, but only to improper police arrangement of the circumstances surrounding an identification. The defendants in Coleman contended that a witness’ in-court identifications violated due process, because a pretrial stationhouse lineup was “so unduly prejudicial and conducive to irreparable misidentification as fatally to taint [the later identifications].” 399 U. S., at 3 (plurality opinion). The Court rejected this argument. Id., at 5–6 (plurality opinion), 13–14 (Black, J., concurring), 22, n. 2 (Burger, C. J., dissenting), 28, n. 2 (Stewart, J., dissenting). No due process violation occurred, the plurality explained, because nothing “the police said or did prompted [the witness’] virtually spontaneous identification of [the defendants].” Id., at 6. True, Coleman was the only person in the lineup wearing a hat, the plurality noted, but “nothing in the record show[ed] that he was required to do so.” Ibid. See also Colorado v. Connelly, 479 U. S. 157, 163, 167 (1986) (Where the “crucial element of police overreaching” is missing, the admissibility of an allegedly unreliable confession is “a matter to be governed by the evidentiary laws of the forum, . . . and not by the Due Process Clause.”). Perry and the dissent place significant weight on United States v. Wade, 388 U. S. 218 (1967) , describing it as a decision not anchored to improper police conduct. See Brief for Petitioner 12, 15, 21–22, 28; post, at 2–4, 8–10. In fact, the risk of police rigging was the very danger to which the Court responded in Wade when it recognized a defendant’s right to counsel at postindictment, police-organized identification procedures. 388 U. S., at 233, 235–236. “[T]he confrontation compelled by the State between the accused and the victim or witnesses,” the Court began, “is peculiarly riddled with innumerable dangers and variable factors which might seriously, even crucially, derogate from a fair trial.” Id., at 228 (emphasis added). “A major factor contributing to the high incidence of miscarriage of justice from mistaken identification,” the Court continued, “has been the degree of suggestion inherent in the manner in which the prosecution presents the suspect to witnesses for pretrial identification.” Ibid. (emphasis added). To illustrate the improper suggestion it was concerned about, the Court pointed to police-designed lineups where “all in the lineup but the suspect were known to the identifying witness, . . . the other participants in [the] lineup were grossly dissimilar in appearance to the suspect, . . . only the suspect was required to wear distinctive clothing which the culprit allegedly wore, . . . the witness is told by the police that they have caught the culprit after which the defendant is brought before the witness alone or is viewed in jail, . . . the suspect is pointed out before or during a lineup, . . . the participants in the lineup are asked to try on an article of clothing which fits only the suspect.” Id., at 233 (footnotes omitted). Beyond genuine debate, then, prevention of unfair police practices prompted the Court to extend a defendant’s right to counsel to cover postindictment lineups and showups. Id., at 235. Perry’s argument, reiterated by the dissent, thus lacks support in the case law he cites. Moreover, his position would open the door to judicial preview, under the banner of due process, of most, if not all, eyewitness identifications. External suggestion is hardly the only factor that casts doubt on the trustworthiness of an eyewitness’ testimony. As one of Perry’s amici points out, many other factors bear on “the likelihood of misidentification,” post, at 9—for example, the passage of time between exposure to and identification of the defendant, whether the witness was under stress when he first encountered the suspect, how much time the witness had to observe the suspect, how far the witness was from the suspect, whether the suspect carried a weapon, and the race of the suspect and the witness. Brief for American Psychological Association as Amicus Curiae 9–12. There is no reason why an iden- tification made by an eyewitness with poor vision, for ex- ample, or one who harbors a grudge against the defendant, should be regarded as inherently more reliable, less of a “threat to the fairness of trial,” post, at 14, than the identification Blandon made in this case. To embrace Perry’s view would thus entail a vast enlargement of the reach of due process as a constraint on the admission of evidence. Perry maintains that the Court can limit the due process check he proposes to identifications made under “suggestive circumstances.” Tr. of Oral Arg. 11–14. Even if we could rationally distinguish suggestiveness from other factors bearing on the reliability of eyewitness evidence, Perry’s limitation would still involve trial courts, routinely, in preliminary examinations. Most eyewitness identifications involve some element of suggestion. Indeed, all in-court identifications do. Out-of-court identifications volunteered by witnesses are also likely to involve suggestive circumstances. For example, suppose a witness identifies the defendant to police officers after seeing a photograph of the defendant in the press captioned “theft suspect,” or hearing a radio report implicating the defendant in the crime. Or suppose the witness knew that the defendant ran with the wrong crowd and saw him on the day and in the vicinity of the crime. Any of these circumstances might have “suggested” to the witness that the defendant was the person the witness observed committing the crime. C In urging a broadly applicable due process check on eyewitness identifications, Perry maintains that eyewitness identifications are a uniquely unreliable form of evidence. See Brief for Petitioner 17–22 (citing studies showing that eyewitness misidentifications are the leading cause of wrongful convictions); Brief for American Psychological Association as Amicus Curiae 14–17 (describing research indicating that as many as one in three eyewitness identifications is inaccurate). See also post, at 14–17. We do not doubt either the importance or the fallibility of eyewitness identifications. Indeed, in recognizing that defendants have a constitutional right to counsel at postindictment police lineups, we observed that “the annals of criminal law are rife with instances of mistaken identification.” Wade, 388 U. S., at 228. We have concluded in other contexts, however, that the potential unreliability of a type of evidence does not alone render its introduction at the defendant’s trial fundamentally unfair. See, e.g., Ventris, 556 U. S., at 594, n. (declining to “craft a broa[d] exclusionary rule for uncorroborated statements obtained [from jailhouse snitches],” even though “rewarded informant testimony” may be inherently untrustworthy); Dowling, 493 U. S., at 353 (rejecting ar- gument that the introduction of evidence concerning acquitted conduct is fundamentally unfair because such evidence is “inherently unreliable”). We reach a similar conclusion here: The fallibility of eyewitness evidence does not, without the taint of improper state conduct, warrant a due process rule requiring a trial court to screen such evidence for reliability before allowing the jury to assess its creditworthiness. Our unwillingness to enlarge the domain of due process as Perry and the dissent urge rests, in large part, on our recognition that the jury, not the judge, traditionally de- termines the reliability of evidence. See supra, at 7. We also take account of other safeguards built into our adversary system that caution juries against placing undue weight on eyewitness testimony of questionable reliability. These protections include the defendant’s Sixth Amendment right to confront the eyewitness. See Maryland v. Craig, 497 U. S. 836, 845 (1990) (“The central concern of the Confrontation Clause is to ensure the reliability of the evidence against a criminal defendant.”). Another is the defendant’s right to the effective assistance of an attorney, who can expose the flaws in the eyewitness’ testimony during cross-examination and focus the jury’s attention on the fallibility of such testimony during opening and closing arguments. Eyewitness-specific jury instructions, which many federal and state courts have adopted, [ 7 ] likewise warn the jury to take care in appraising identification evidence. See, e.g., United States v. Telfaire, 469 F. 2d 552, 558–559 (CADC 1972) (per curiam) (D. C. Circuit Model Jury Instructions) (“If the identification by the witness may have been influenced by the circumstances under which the defendant was presented to him for identification, you should scrutinize the identification with great care.”). See also Ventris, 556 U. S., at 594, n. (citing jury instructions that informed jurors about the unreliability of uncorroborated jailhouse-informant testimony as a reason to resist a ban on such testimony); Dowling, 493 U. S., at 352–353. The constitutional requirement that the government prove the defendant’s guilt beyond a reasonable doubt also impedes convictions based on dubious identification evidence. State and federal rules of evidence, moreover, permit trial judges to exclude relevant evidence if its probative value is substantially outweighed by its prejudicial impact or potential for misleading the jury. See, e.g., Fed. Rule Evid. 403; N. H. Rule Evid. 403 (2011). See also Tr. of Oral Arg. 19–22 (inquiring whether the standard Perry seeks differs materially from the one set out in Rule 403). In appropriate cases, some States also permit defendants to present expert testimony on the hazards of eyewitness identification evidence. See, e.g., State v. Clopten, 2009 UT 84, A33, 223 P. 3d 1103, 1113 (“We expect . . . that in cases involving eyewitness identification of strangers or near-strangers, trial courts will routinely admit expert testimony [on the dangers of such evidence].”). Many of the safeguards just noted were at work at Perry’s trial. During her opening statement, Perry’s court-appointed attorney cautioned the jury about the vulnerability of Blandon’s identification. App. 115a (Blandon, “the eyewitness that the State needs you to believe[,] can’t pick [Perry] out of a photo array. How carefully did she really see what was going on? . . . How well could she really see him?”). While cross-examining Blandon and Officer Clay, Perry’s attorney constantly brought up the weaknesses of Blandon’s identification. She highlighted: (1) the significant distance between Blandon’s window and the parking lot, id., at 226a; (2) the lateness of the hour, id., at 225a; (3) the van that partly obstructed Blandon’s view, id., at 226a; (4) Blandon’s concession that she was “so scared [she] really didn’t pay attention” to what Perry was wearing, id., at 233a; (5) Blandon’s inability to describe Perry’s facial features or other identifying marks, id., at 205a, 233a–235a; (6) Blandon’s failure to pick Perry out of a photo array, id., at 235a; and (7) Perry’s position next to a uniformed, gun-bearing police officer at the moment Blandon made her identification, id., at 202a–205a. Perry’s counsel reminded the jury of these frailties during her summation. Id., at 374a–375a (Blandon “wasn’t able to tell you much about who she saw . . . . She couldn’t pick [Perry] out of a lineup, out of a photo array . . . . [Blandon said] [t]hat guy that was with the police officer, that’s who was circling. Again, think about the context with the guns, the uniforms. Powerful, powerful context clues.”). After closing arguments, the trial court read the jury a lengthy instruction on identification testimony and the factors the jury should consider when evaluating it. Id., at 399a–401a. The court also instructed the jury that the defendant’s guilt must be proved beyond a reasonable doubt, id., at 390a, 392a, 395a–396a, and specifically cautioned that “one of the things the State must prove [beyond a reasonable doubt] is the identification of the defendant as the person who committed the offense,” id., at 398a–399a. Given the safeguards generally applicable in criminal trials, protections availed of by the defense in Perry’s case, we hold that the introduction of Blandon’s eyewitness testimony, without a preliminary judicial assessment of its reliability, did not render Perry’s trial fundamentally unfair. * * * For the foregoing reasons, we agree with the New Hampshire courts’ appraisal of our decisions. See supra, at 4–5. Finding no convincing reason to alter our precedent, we hold that the Due Process Clause does not require a preliminary judicial inquiry into the reliability of an eyewitness identification when the identification was not procured under unnecessarily suggestive circum- stances arranged by law enforcement. Accordingly, the judgment of the New Hampshire Supreme Court is Affirmed. Notes 1 The dissent, too, appears to urge that all suggestive circumstances raise due process concerns warranting a pretrial ruling. See post, at 6, 9, 14–17. Neither Perry nor the dissent, however, points to a single case in which we have required pretrial screening absent a police-arranged identification procedure. Understandably so, for there are no such cases. Instead, the dissent surveys our decisions, heedless of the police arrangement that underlies every one of them, and inventing a “longstanding rule,” post, at 6, that never existed. Nor are we, as the dissent suggests, imposing a mens rea requirement, post, at 1, 7, or otherwise altering our precedent in any way. As our case law makes clear, what triggers due process concerns is police use of an unnecessarily suggestive identification procedure, whether or not they intended the arranged procedure to be suggestive. 2 The box, which Clay found on the ground near where she first encountered Perry, contained car-stereo speakers. App. 177a–178a. 3 The theft charge was based on the taking of items from Clavijo’s car, while the criminal mischief count was founded on the shattering of Clavijo’s car window. 4 Compare United States v. Bouthot, 878 F. 2d 1506, 1516 (CA1 1989) (Due process requires federal courts to “scrutinize all suggestive identification procedures, not just those orchestrated by the police.”); Dunnigan v. Keane, 137 F. 3d 117, 128 (CA2 1998) (same); Thigpen v. Cory, 804 F. 2d 893, 895 (CA6 1986) (same), with United States v. Kimberlin, 805 F. 2d 210, 233 (CA7 1986) (Due process check is required only in cases involving improper state action.); United States v. Zeiler, 470 F. 2d 717, 720 (CA3 1972) (same); State v. Addison, 160 N. H. 792, 801, 8 A. 3d 118, 125 (2010) (same); State v. Reid, 91 S. W. 3d 247, 272 (Tenn. 2002) (same); State v. Nordstrom, 200 Ariz. 229, 241, 25 P. 3d 717, 729 (2001) (same); Semple v. State, 271 Ga. 416, 417–418, 519 S. E. 2d 912, 914–915 (1999) (same); Harris v. State, 619 N. E. 2d 577, 581 (Ind. 1993) (same); State v. Pailon, 590 A. 2d 858, 862–863 (R. I. 1991) (same); Commonwealth v. Colon-Cruz, 408 Mass. 533, 541–542, 562 N. E. 2d 797, 805 (1990) (same); State v. Brown, 38 Ohio St. 3d 305, 310–311, 528 N. E. 2d 523, 533 (1988) (same); Wilson v. Commonwealth, 695 S. W. 2d 854, 857 (Ky. 1985) (same). 5 Among “factors to be considered” in evaluating a witness’ “ability to make an accurate identification,” the Court listed: “the opportunity of the witness to view the criminal at the time of the crime, the witness’ degree of attention, the accuracy of his prior description of the criminal, the level of certainty demonstrated at the confrontation, and the time between the crime and the confrontation.” Manson v. Brathwaite, (citing Neil v. Biggers, –200 (1972)). 6 The Court’s description of the question presented in Brathwaite assumes that improper state action occurred: “[Does] the Due Process Clause of the compe[l] the exclusion, in a state criminal trial, apart from any consideration of reliability, of pretrial identification evidence obtained by a police procedure that was both suggestive and unnecessary.” 432 U. S., at 99. 7 See Model Crim. Jury Instr. No. 4.15 (CA3 2009); United States v. Holley, 502 F. 2d 273, 277–278 (CA4 1974); Pattern Crim. Jury Instr. No. 1.29 (CA5 2001); Pattern Crim. Jury Instr. No. 7.11 (CA6 2011); Fed. Crim. Jury Instr. No. 3.08 (CA7 1999); Model Crim. Jury Instr. for the District Courts No. 4.08 (CA8 2011); Model Crim. Jury Instr. No. 4.11 (CA9 2010); Crim. Pattern Jury Instr. No. 1.29 (CA10 2011); Pattern Jury Instr. (Crim. Cases) Spec. Instr. No. 3 (CA11 2010); Rev. Ariz. Jury Instr., Crim., No. 39 (3d ed. 2008); 1 Judicial Council of Cal. Crim. Jury Instr. No. 315 (Summer 2011); Conn. Crim. Jury Instr. 2.6–4 (2007); 2 Ga. Suggested Pattern Jury Instr. (Crim. Cases) No. 1.35.10 (4th ed. 2011); Ill. Pattern Jury Instr., Crim., No. 3.15 (Supp. 2011); Pattern Instr., Kan. 3d, Crim., No. 52.20 (2011); 1 Md. Crim. Jury Instr. & Commentary §§2.56, 2.57(A), 2.57(B) (3d ed. 2009 and Supp. 2010); Mass. Crim. Model Jury Instr. No. 9.160 (2009); 10 Minn. Jury Instr. Guides, Crim., No. 3.19 (Supp. 2006); N. H. Crim. Jury Instr. No. 3.06 (1985); N. Y. Crim. Jury Instr. “Identification—One Witness” and “Identification—Witness Plus” (2d ed. 2011); Okla. Uniform Jury Instr., Crim., No. 9–19 (Supp. 2000); 1 Pa. Suggested Standard Crim. Jury Instr. No. 4.07B (2d ed. 2010); Tenn. Pattern Jury Instr., Crim., No. 42.05 (15th ed. 2011); Utah Model Jury Instr. CR404 (2d ed. 2010); Model Instructions from the Vt. Crim. Jury Instr. Comm. Nos. CR5–601, CR5–605 (2003); W. Va. Crim. Jury Instr. No. 5.05 (6th ed. 2003).

565.US.2011_10-218

Petitioner PPL Montana, LLC (PPL), owns and operates hydroelectric facilities in Montana. Ten of its facilities are located on riverbeds underlying segments of the Missouri, Madison, and Clark Fork Rivers. Five hydroelectric dams on the Upper Missouri River are along the Great Falls reach, including on the three tallest waterfalls; and PPL’s two other dams on that river are in canyons on the Stubbs Ferry stretch. These, together with two dams located in steep canyons on the Madison River, are called the Missouri-Madison project. The Thompson Falls project is a facility on the Clark Fork River. Both projects are licensed by the Federal Energy Regulatory Commission. PPL’s facilities have existed for many decades, some for over a century. Until recently, Montana, though aware of the projects’ existence, sought no rent for use of the riverbeds. Instead, the understanding of PPL and the United States is that PPL has paid rents to the United States. In 2003, parents of Montana schoolchildren filed a federal suit, claiming that PPL’s facilities were on riverbeds that were state owned and part of Montana’s school trust lands. The State joined the suit and, for the first time, sought rents from PPL for its use of the riverbeds. That case was dismissed, and PPL and other power companies filed a state-court suit, claiming that Montana was barred from seeking compensation for PPL’s riverbed use. Montana counterclaimed, contending that under the equal-footing doctrine it owns the riverbeds and can charge rent for their use. The trial court granted Montana summary judgment as to navigability for purposes of determining riverbed title and ordered PPL to pay Montana $41 million in rent for riverbed use between 2000 and 2007. The Montana Supreme Court affirmed. Adopting a liberal construction of the navigability test, it discounted this Court’s approach of considering the navigability of particular river segments for purposes of determining whether a State acquired title to the riverbeds underlying those segments at the time of statehood. Instead, the Montana court declared the river stretches in question to be short interruptions of navigability that were insufficient as a matter of law to find nonnavigability, since traffic had circumvented those stretches by portage. Based on evidence of present-day, recreational use of the Madison River, the court found that river navigable as a matter of law at the time of statehood. Held: The Montana Supreme Court’s ruling that Montana owns and may charge for use of the riverbeds at issue was based on an infirm legal understanding of this Court’s rules of navigability for title under the equal-footing doctrine. Pp. 10–26. (a) The rule that the States, in their capacity as sovereigns, hold “title in the soil of rivers really navigable,” Shively v. Bowlby, 152 U.S. 1, 31, has federal constitutional significance under the equal-footing doctrine. Pursuant to that doctrine, upon its date of statehood, a State gains title within its borders to the beds of waters then navigable. It may allocate and govern those lands according to state law subject only to the United States’ power “to control such waters for purposes of navigation in interstate and foreign commerce.” United States v. Oregon, 295 U.S. 1, 14. The United States retains title vested in it before statehood to land beneath waters not then navigable. To be navigable for purposes of title under the equal-footing doctrine, rivers must be “navigable in fact,” meaning “they are used, or are susceptible of being used, . . . as highways for commerce, over which trade and travel are or may be conducted in the customary modes of trade and travel on water.” The Daniel Ball, 10 Wall. 557, 563. This formulation has been used to determine questions of waterbed title under the equal-footing doctrine. See United States v. Utah, 283 U.S. 64, 76. Pp. 10–14. (b) The Montana Supreme Court erred in its treatment of the question of river segments and portage. To determine riverbed title under the equal-footing doctrine, this Court considers the river on a segment-by-segment basis to assess whether the segment of the river, under which the riverbed in dispute lies, is navigable or not. See, e.g., Utah, supra, at 77. The State Supreme Court erred in discounting this well-settled approach. A key justification for sovereign ownership of navigable riverbeds is that a contrary rule would allow private riverbed owners to erect improvements on the riverbeds that could interfere with the public’s right to use the waters as a highway for commerce. Because commerce could not have occurred on segments nonnavigable at the time of statehood, there is no reason to deem those segments owned by the State under the equal-footing doctrine. Practical considerations also support segmentation. Physical conditions affecting navigability vary over the length of a river and provide a means to determine appropriate start points and end points for disputed segments. A segment approach is also consistent with the manner in which private parties seek to establish riverbed title. Montana cannot suggest that segmentation is inadministrable when the state courts managed to apportion the underlying riverbeds for purposes of determining their value and PPL’s corresponding rents. The State Supreme Court’s view that the segment-by-segment approach does not apply to short interruptions of navigability is not supported by this Court’s Utah decision. Even if the law might find some nonnavigable segments so minimal that they merit treatment as part of a longer, navigable reach, it is doubtful that the segments in this case would meet that standard. Applying its “short interruptions” approach, the State Supreme Court found the Great Falls reach navigable because it could be managed by way of land route portage, as done by Lewis and Clark. But a portage of even one day would demonstrate the need to bypass a nonnavigable river segment. Thus, the State Supreme Court was wrong to conclude, with respect to the Great Falls reach and other disputed stretches, that portages were insufficient to defeat a navigability finding. In most cases, they are, because they require transportation over land rather than over the water. This is the case at least as to the Great Falls reach. In reaching a contrary conclusion, the State Supreme Court misapplied The Montello, 20 Wall. 430. There, portage was considered in determining whether a river was part of a channel of interstate commerce for federal regulatory purposes. The Montello does not control the outcome where the quite different concerns of the riverbed title context apply. Portages may defeat navigability for title purposes, and do so with respect to the Great Falls reach. Montana does not dispute that overland portage was necessary to traverse that reach, and the trial court noted the waterfalls had never been navigated. The Great Falls reach, at least from the head of the first waterfall to the foot of the last, is not navigable for purposes of riverbed title under the equal-footing doctrine. There is also a significant likelihood that some of the other river stretches in dispute fail this federal navigability test. The ultimate decision as to these other disputed river stretches is to be determined, in the first instance, by the Montana courts on remand, which should assess the relevant evidence in light of the principles discussed here. Pp. 14–21. (c) The Montana Supreme Court further erred as a matter of law in relying on evidence of present-day, primarily recreational use of the Madison River. Navigability must be assessed as of the time of statehood, and it concerns a river’s usefulness for “ ‘trade and travel.’ ” Utah, 283 U. S., at 75–76. River segments are navigable if they “ ‘[were]’ ” used and if they “ ‘[were] susceptible of being used’ ” as highways of commerce at the time of statehood. Id., at 76. Evidence of recreational use and poststatehood evidence may bear on susceptibility of commercial use at the time of statehood. See id., at 82–83. In order for present-day use to have a bearing on navigability at statehood, (1) the watercraft must be meaningfully similar to those in customary use for trade and travel at the time of statehood, and (2) the river’s poststatehood condition may not be materially different from its physical condition at statehood. The State Supreme Court offered no indication that it made these necessary findings. Pp. 21–24. (d) Because this analysis is sufficient to require reversal here, the Court declines to decide whether the State Supreme Court also erred as to the burden of proof regarding navigability. P. 24. (e) Montana’s suggestion that denying the State title to the disputed riverbeds will undermine the public trust doctrine—which concerns public access to the waters above those beds for navigation, fishing, and other recreational uses—underscores its misapprehension of the equal-footing and public trust doctrines. Unlike the equal-footing doctrine, which is the constitutional foundation for the navigability rule of riverbed title, the scope of the public trust over waters within the State’s borders is a matter of state law, subject to federal regulatory power. Pp. 24–25. (f) This Court does not reach the question whether, by virtue of Montana’s sovereignty, neither laches nor estoppel could apply to bar the State’s claim. Still, the reliance by PPL and its predecessors in title on the State’s long failure to assert title to the riverbeds is some evidence supporting the conclusion that the river segments over those beds were nonnavigable for purposes of the equal-footing doctrine. Pp. 25–26. 2010 MT 64, 355 Mont. 402, 229 P.3d 421, reversed and remanded. Kennedy, J., delivered the opinion for a unanimous Court.

This case concerns three rivers which flow through Montana and then beyond its borders. The question is whether discrete, identifiable segments of these rivers in Montana were nonnavigable, as federal law defines that concept for purposes of determining whether the State acquired title to the riverbeds underlying those segments, when the State entered the Union in 1889. Montana contends that the rivers must be found navigable at the disputed locations. From this premise, the State asserts that in 1889 it gained title to the disputed riverbeds under the constitutional equal-footing doctrine. Based on its title claims, Montana sought compensation from PPL Montana, LLC, a power company, for its use of the riverbeds for hydroelectric projects. The Montana courts granted summary judgment on title to Montana, awarding it $41 million in rent for the riverbeds for the period from 2000 to 2007 alone. That judgment must be reversed. I The three rivers in question are the Missouri River, the Madison River, and the Clark Fork River. The Missouri and the Madison are on the eastern side of the Continental Divide. The Madison flows into the Missouri, which then continues at length to its junction with the Mississippi River. The Clark Fork River is on the western side of the Continental Divide. Its waters join the Columbia River system that flows into the Pacific Ocean. Each river shall be described in somewhat more detail. A The Missouri River originates in Montana and traverses seven States before a point just north of St. Louis where it joins the Mississippi. 19 Encyclopedia Americana 270 (int’l ed. 2006). If considered with the continuous path formed by certain streams that provide the Missouri River’s headwaters, the Missouri is over 2,500 miles long, the longest river in the United States. Ibid. The Missouri River’s basin (the land area drained by the river) is the second largest in the Nation, surpassed only by the Mississippi River basin of which it is a part. Rivers of North America 427 (A. Benke & C. Cushing eds. 2005) (hereinafter Rivers of North America). As a historical matter, the river shifted and flooded often, and contained many sandbars, islands, and unstable banks. Id., at 432–433. The river was once described as one of the most “variable beings in creation,” as “inconstant [as] the action of the jury,” Sioux City Register (Mar. 28, 1868); and its high quantity of downstream sediment flow spawned its nickname, the “Big Muddy,” Rivers of North America 433. The upstream part of the Missouri River in Montana, known as the Upper Missouri River, is better charac- terized as rocky rather than muddy. While one usually thinks of the Missouri River as flowing generally south, as indeed it does beginning in North Dakota, the Upper Missouri in Montana flows north from its principal headwaters at Three Forks, which is located about 4,000 feet above sea level in the Rocky Mountain area of southwestern Montana. It descends through scenic mountain terrain including the deep gorge at the Gates of the Mountains; turns eastward through the Great Falls reach, cascading over a roughly 10-mile stretch of cataracts and rapids over which the river drops more than 400 feet; and courses swiftly to Fort Benton, a 19th-century fur trading post, before progressing farther east into North Dakota and on to the Great Plains. 19 Encyclopedia Americana, supra, at 270; 8 New Encyclopaedia Britannica 190 (15th ed. 2007) (hereinafter Encyclopaedia Britannica); 2 Columbia Gazetteer of the World 2452 (2d ed. 2008) (here- inafter Columbia Gazetteer); F. Warner, Montana and the Northwest Territory 75 (1879). In 1891, just after Montana became a State, the Upper Missouri River above Fort Benton was “seriously obstructed by numerous rapids and rocks,” and the 168-mile portion flowing eastward “[f]rom Fort Benton to Carroll, Mont., [was] called the rocky river.” Annual Report of the Chief of Engineers, U. S. Army (1891), in 2 H. R. Exec. Doc. No. 1, 52d Cong., 1st Sess., pt. 2, pp. 275–276 (1891) (hereinafter H. R. Exec. Doc.). The Great Falls exemplify the rocky, rapid character of the Upper Missouri. They consist of five cascade-like waterfalls located over a stretch of the Upper Missouri leading downstream from the city of Great Falls in midwestern Montana. The waterfall farthest downstream, and the one first encountered by Meriwether Lewis and William Clark when they led their remarkable expedition through the American West in 1805, is the eponymous “Great Falls,” the tallest of the five falls at 87 feet. W. Clark, Dear Brother: Letters of William Clark to Jonathan Clark 109, n. 5 (J. Holmberg ed. 2002) (hereinafter Dear Brother). Lewis recorded observations of this “sublimely grand specticle”: “[T]he whole body of water passes with incredible swiftness. . . . over a precipice of at least eighty feet . . . . [T]he irregular and somewhat projecting rocks below receives the water . . . and brakes it into a perfect white foam which assumes a thousand forms in a moment sometimes flying up in jets . . . [that] are scarcely formed before large roling bodies of the same beaten and foaming water is thrown over and conceals them. . . . [T]he [rainbow] reflection of the sun on the sprey or mist . . . adds not a little to the beauty of this majestically grand senery.” The Lewis and Clark Journals: An American Epic of Discovery 129 (G. Moulton ed. 2003) (hereinafter Lewis and Clark Journals); The Journals of Lewis and Clark 136–138 (B. DeVoto ed. 1981). If one proceeds alongside the river upstream from Great Falls, as Lewis did in scouting the river for the expedition, the other four falls in order are “Crooked Falls” (19 feet high); “Rainbow Falls” (48 feet), which Lewis called “one of the most bea[u]tifull objects in nature”; “Colter Falls” (7 feet), and “Black Eagle Falls” (26 feet). See Lewis and Clark Journals 131–132; Dear Brother 109, n. 5; P. Cut- right, Lewis & Clark: Pioneering Naturalists 154–156 (2003). Despite the falls’ beauty, Lewis could see that their steep cliffs and swift waters would impede progress on the river, which had been the expedition’s upstream course for so many months. The party proceeded over a more circuitous land route by means of portage, circumventing the Great Falls and their surrounding reach of river before returning to travel upon the river about a month later. See Lewis and Clark Journals 126–152. The Upper Missouri River, both around and further upstream of the Great Falls, shares the precipitous and fast-moving character of the falls themselves. As it moves downstream over the Great Falls reach, a 17-mile stretch that begins somewhat above the head of Black Eagle Falls, the river quickly descends about 520 feet in elevation, see Montana Power Co. v. Federal Power Comm’n, 185 F.2d 491 (CADC 1950); 2010 MT 64, ¶¶29–30, 108–109, 355 Mont. 402, 416, 442, 229 P.3d 421, 433, 449, dropping over 400 feet within 10 miles from the first rapid to the foot of Great Falls, Parker, Black Eagle Falls Dam, 27 Transactions of the Am. Soc. of Civil Engineers 56 (1892). In 1879, that stretch was a “constant succession of rapids and falls.” Warner, supra, at 75; see also 9 The Journals of the Lewis & Clark Expedition 171 (G. Moulton ed. 1995) (hereinafter Journals of the Lewis & Clark Expedition) (“a continued rapid the whole way for 17 miles”). Lewis noted the water was so swift over the area that buffalo were swept over the cataracts in “considerable quantities” and were “instantly crushed.” Lewis and Clark Journals 136–137. Well above the Great Falls reach, the Stubbs Ferry stretch of the river from Helena to Cascade also had steep gradient and was “much obstructed by rocks and dangerous rapids.” Report of the Secretary of War, 2 H. R. Doc. No. 2, 54th Cong., 1st Sess., pt. 1, p. 301 (1895). B The second river to be considered is the Madison, one of the Missouri River’s headwater tributaries. Named by Lewis and Clark for then-Secretary of State James Madison, the Madison River courses west out of the Northern Rocky Mountains of Wyoming and Montana in what is now Yellowstone National Park, then runs north and merges with the Jefferson and Gallatin Rivers at Three Forks, Montana, to form the Upper Missouri. Lewis and Clark Journals 158; Rivers of North America 459; 7 En- cyclopaedia Britannica 658; 2 Columbia Gazetteer 2242. Along its path, the Madison River flows through two lakes artificially created by dams built in canyons: Hebgen Lake and Ennis Lake. Federal Writers’ Project of the Work Projects Administration, Montana: A State Guide Book 356 (J. Stahlberg ed. 1949); R. Aarstad, E. Arguimbau, E. Baumler, C. Porsild, & B. Shovers, Montana Place Names from Alzada to Zortman: A Montana Historical Society Guide 166 (2009). C The third river at issue in this case is the Clark Fork. That river, which consists in large part of “long, narrow streams confined by mountainous terrain,” rises at an ele- vation of about 5,000 feet in the Silver Bow Mountains of southwestern Montana. 3 Encyclopaedia Britannica 352; Dept. of Interior, U. S. Geological Survey, J. Stevens & F. Henshaw, Surface Water Supply of the United States, 1907–8, Water-Supply Paper 252, pp. 81–82 (1910). The river flows northward for about 40 miles; turns northwest for a stretch; then turns abruptly northeast for a short stint, by which time it has descended nearly 2,500 feet in altitude. It then resumes a northwestward course until it empties into Lake Pend Oreille in northern Idaho, out of which flows a tributary to the Columbia River of the Pacific Northwest. Ibid.; 1 Columbia Gazetteer 816. The Clark Fork is “one of the wildest and most picturesque streams in the West,” marked by “many waterfalls and boxed gorges.” Federal Writers’ Projects of the Works Progress Administration, Idaho: A Guide in Word and Picture 230 (2d ed. 1950). Lewis and Clark knew of the Clark Fork River but did not try to navigate it, in part because the absence of salmon in one of its tributaries made Lewis believe “ ‘there must be a considerable fall in [the river] below.’ ” H. Fritz, The Lewis and Clark Expedition 38–39 (2004). This was correct, for shortly before the Clark Fork exits to Idaho from the northwest corner of Montana, “the waters of the river dash madly along their rocky bed,” dropping over 30 feet in a half-mile as they rush over falls and rapids including a “foaming waterfall” now known as Thompson Falls. O. Rand, A Vacation Excursion: From Massachusetts Bay to Puget Sound 176–177 (1884); C. Kirk, A History of the Montana Power Company 231 (2008). II Petitioner PPL Montana, LLC (PPL), owns and operates hydroelectric facilities that serve Montana residents and businesses. Ten of its facilities are built upon riverbeds underlying segments of the Upper Missouri, Madison, and Clark Fork Rivers. It is these beds to which title is disputed. On the Upper Missouri River, PPL has seven hydroelectric dams. Five of them are along the Great Falls reach, including on the three tallest falls; and the other two are in canyons upstream on the Stubbs Ferry stretch. See K. Robison, Cascade County and Great Falls 56 (2011); Aarstad et al., supra, at 125, 119, 145–146. On the Madison River, two hydroelectric dams are located in steep canyons. On the Clark Fork River, a hydroelectric facility is constructed on the Thompson Falls. The dams on the Upper Missouri and Madison are called the Missouri-Madison project. The Thompson Falls facility is called the Thompson Falls project. Both projects are licensed by the Federal Energy Regulatory Commission. PPL acquired them in 1999 from its predecessor, the Montana Power Company. 355 Mont., at 405–406, 229 P. 3d, at 426. PPL’s power facilities have existed at their locations for many decades, some for over a century. See Robison, supra, at 40 (Black Eagle Falls dam constructed by 1891). Until recently, these facilities were operated without title-based objection by the State of Montana. The State was well aware of the facilities’ existence on the riverbeds—indeed, various Montana state agencies had participated in federal licensing proceedings for these hydroelectric projects. See, e.g., Montana Power Co., 8 F. P. C. 751, 752 (1949) (Thompson Falls project); Montana Power Co., 27 FERC ¶62,097, pp. 63,188–63,189 (1984) (Ryan Dam of Missouri-Madison project). Yet the State did not seek, and accordingly PPL and its predecessor did not pay, compensation for use of the riverbeds. 355 Mont., at 406, 229 P. 3d, at 427. Instead, the understanding of PPL and the United States is that PPL has been paying rents to the United States for use of those riverbeds, as well as for use of river uplands flooded by PPL’s projects. Reply Brief for Petitioner 4; App. to Supp. Brief for Petitioner 4–5; Brief for United States as Amicus Curiae 3, n. 3. In 2003, parents of Montana schoolchildren sued PPL in the United States District Court for the District of Montana, arguing that PPL had built its facilities on riverbeds that were state owned and part of Montana’s school trust lands. 355 Mont., at 406, 229 P. 3d, at 426. Prompted by the litigation, the State joined the lawsuit, for the first time seeking rents for PPL’s riverbed use. The case was dismissed in September 2005 for lack of diversity juris- diction. Dolan v. PPL Montana, LLC, No. 9:03–cv–167 (D Mont., Sept. 27, 2005). PPL and two other power companies sued the State of Montana in the First Judicial District Court of Montana, arguing that the State was barred from seeking compensation for use of the riverbeds. 355 Mont., at 407–408, 229 P. 3d, at 427–428. By counterclaim, the State sought a declaration that under the equal-footing doctrine it owns the riverbeds used by PPL and can charge rent for their use. Id., at 408, 229 P. 3d, at 428. The Montana trial court granted summary judgment to Montana as to navigability for purposes of determining riverbed title. Id., at 408–409, 413–414, 229 P. 3d, at 428, 431–432; App. to Pet. for Cert. 143. The court decided that the State owned the riverbeds. 355 Mont., at 428–429, 229 P. 3d, at 440. The court ordered PPL to pay $40,956,180 in rent for use of the riverbeds between 2000 and 2007. Id., at 431–432, 229 P. 3d, at 442–443. Whether a lease for future periods would commence, and, if so, at what rental rate, seems to have been left to the discretion of the Montana Board of Land Commissioners. App. to Pet. for Cert. 128–129. In a decision by a divided court, the Montana Supreme Court affirmed. 355 Mont., at 461–462, 229 P. 3d, at 460–461; id., at 462, 229 P. 3d, at 461 (dissenting opinion). The court reasoned from the background principle that “navigability for title purposes is very liberally construed.” Id., at 438, 229 P. 3d, at 446. It dismissed as having “limited applicability” this Court’s approach of assessing the navigability of the disputed segment of the river rather than the river as a whole. Id., at 441–442, 229 P. 3d, at 448–449. The Montana court accepted that certain relevant stretches of the rivers were not navigable but declared them “merely short interruptions” insufficient as a matter of law to find nonnavigability, since traffic had circumvented those stretches by overland portage. Id., at 438, 442, 229 P. 3d, at 446, 449. Placing extensive reliance upon evidence of present-day use of the Madison River, the court found that river navigable as a matter of law at the time of statehood. Id., at 439, 229 P. 3d, at 447. Justice Rice dissented. Id., at 462, 229 P. 3d, at 461. He stated that “courts are not to assume an entire river is navigable merely because certain reaches of the river are navigable.” Id., at 464, 229 P. 3d, at 462. The majority erred, he wrote, in rejecting the “section-by-section approach” and “declaring, as a matter of law, that the reaches claimed by PPL to be non-navigable are simply too ‘short’ to matter,” when in fact PPL’s evidence showed the “disputed reaches of the rivers were, at the time of statehood, non-navigable.” Id., at 463–466, 476–477, 229 P. 3d, at 462–464, 470. This Court granted certiorari, 564 U. S. ___ (2011), and now reverses the judgment. III A PPL contends the opinion of the Montana Supreme Court is flawed in three respects: first, the court’s failure to consider with care the navigability of the particular river segments to which title is disputed, and its disregard of the necessary overland portage around some of those segments; second, its misplaced reliance upon evidence of present-day, recreational use; and third, what the state court itself called its liberal construction of the navigability test, which did not place the burden of proof upon the State to show navigability. Brief for Petitioner 26. The United States as amicus is in substantial agreement with PPL’s arguments, although it offers a more extended dis- cussion with respect to evidence of present-day, recreational use. Brief for United States 27–33. It is appropriate to begin the analysis by discussing the legal principles that control the case. B The rule that the States, in their capacity as sovereigns, hold title to the beds under navigable waters has origins in English common law. See Shively v. Bowlby, 152 U.S. 1, 13 (1894). A distinction was made in England between waters subject to the ebb and flow of the tide (royal rivers) and nontidal waters (public highways). With respect to royal rivers, the Crown was presumed to hold title to the riverbed and soil, but the public retained the right of passage and the right to fish in the stream. With respect to public highways, as the name suggests, the public also retained the right of water passage; but title to the riverbed and soil, as a general matter, was held in private ownership. Riparian landowners shared title, with each owning from his side to the center thread of the stream, as well as the exclusive right to fish there. See Idaho v. Coeur d’Alene Tribe of Idaho, 521 U.S. 261, 285 (1997) (summarizing J. Angell, A Treatise on the Common Law in Relation to Water-Courses 14–18 (1824)); 3 J. Kent, Commentaries on American Law 528–529 (9th ed. 1858). While the tide-based distinction for bed title was the initial rule in the 13 Colonies, after the Revolution American law moved to a different standard. Some state courts came early to the conclusion that a State holds presumptive title to navigable waters whether or not the waters are subject to the ebb and flow of the tide. See, e.g., Carson v. Blazer, 2 Binn. 475 (Pa. 1810); Executors of Cates v. Wadlington, 12 S. C. L. 580 (1822); Wilson v. Forbes, 13 N. C. 30 (1828); Bullock v. Wilson, 2 Port. 436 (Ala. 1835); Elder v. Burrus, 25 Tenn. 358 (1845). The tidal rule of “navigability” for sovereign ownership of riverbeds, while perhaps appropriate for England’s dominant coastal geography, was ill suited to the United States with its vast number of major inland rivers upon which navigation could be sustained. See L. Houck, Law of Navigable Rivers 26–27, 31–35 (1868); Packer v. Bird, 137 U.S. 661, 667–669 (1891). By the late 19th century, the Court had recognized “the now prevailing doctrine” of state sovereign “title in the soil of rivers really navigable.” Shively, supra, at 31; see Barney v. Keokuk, 94 U.S. 324, 336 (1877) (“In this country, as a general thing, all waters are deemed navigable which are really so”). This title rule became known as “navigability in fact.” The rule for state riverbed title assumed federal constitutional significance under the equal-footing doctrine. In 1842, the Court declared that for the 13 original States, the people of each State, based on principles of sovereignty, “hold the absolute right to all their navigable waters and the soils under them,” subject only to rights surrendered and powers granted by the Constitution to the Federal Government. Martin v. Lessee of Waddell, 16 Pet. 367, 410 (1842). In a series of 19th-century cases, the Court determined that the same principle applied to States later admitted to the Union, because the States in the Union are coequal sovereigns under the Constitution. See, e.g., Lessee of Pollard v. Hagan, 3 How. 212, 228–229 (1845); Knight v. United States Land Assn., 142 U.S. 161, 183 (1891); Shively, supra, at 26–31; see United States v. Texas, 339 U.S. 707, 716 (1950). These precedents are the basis for the equal-footing doctrine, under which a State’s title to these lands was “conferred not by Congress but by the Constitution itself.” Oregon ex rel. State Land Bd. v. Corvallis Sand & Gravel Co., 429 U.S. 363, 374 (1977). It follows that any ensuing questions of navigability for de- termining state riverbed title are governed by federal law. See, e.g., United States v. Utah, 283 U.S. 64, 75 (1931); United States v. Oregon, 295 U.S. 1, 14 (1935). The title consequences of the equal-footing doctrine can be stated in summary form: Upon statehood, the State gains title within its borders to the beds of waters then navigable (or tidally influenced, see Phillips Petroleum Co. v. Mississippi, 484 U.S. 469 (1988), although that is not relevant in this case). It may allocate and govern those lands according to state law subject only to “the paramount power of the United States to control such waters for purposes of navigation in interstate and foreign commerce.” Oregon, supra, at 14; see Montana v. United States, 450 U.S. 544, 551 (1981); United States v. Holt State Bank, 270 U.S. 49, 54 (1926). The United States retains any title vested in it before statehood to any land beneath waters not then navigable (and not tidally influenced), to be transferred or licensed if and as it chooses. See Utah, supra, at 75; Oregon, supra, at 14. Returning to the “navigability in fact” rule, the Court has explained the elements of this test. A basic formulation of the rule was set forth in The Daniel Ball, 10 Wall. 557 (1871), a case concerning federal power to regulate navigation: “Those rivers must be regarded as public navigable rivers in law which are navigable in fact. And they are navigable in fact when they are used, or are susceptible of being used, in their ordinary condition, as highways for commerce, over which trade and travel are or may be conducted in the customary modes of trade and travel on water.” Id., at 563. The Daniel Ball formulation has been invoked in considering the navigability of waters for purposes of assessing federal regulatory authority under the Constitution, and the application of specific federal statutes, as to the waters and their beds. See, e.g., ibid.; The Montello, 20 Wall. 430, 439 (1874); United States v. Appalachian Elec. Power Co., 311 U.S. 377, 406, and n. 21 (1940) (Federal Power Act); Rapanos v. United States, 547 U.S. 715, 730–731 (2006) (plurality opinion) (Clean Water Act); id., at 761 (Kennedy, J., concurring in judgment) (same). It has been used as well to determine questions of title to water beds under the equal-footing doctrine. See Utah, supra, at 76; Oklahoma v. Texas, 258 U.S. 574, 586 (1922); Holt State Bank, supra, at 56. It should be noted, however, that the test for navigability is not applied in the same way in these distinct types of cases. Among the differences in application are the following. For state title under the equal-footing doctrine, naviga- bility is determined at the time of statehood, see Utah, supra, at 75, and based on the “natural and ordinary con- dition” of the water, see Oklahoma, supra, at 591. In contrast, admiralty jurisdiction extends to water routes made navigable even if not formerly so, see, e.g., Ex parte Boyer, 109 U.S. 629, 631–632 (1884) (artificial canal); and federal regulatory authority encompasses waters that only recently have become navigable, see, e.g., Philadelphia Co. v. Stimson, 223 U.S. 605, 634–635 (1912), were once navigable but are no longer, see Economy Light & Power Co. v. United States, 256 U.S. 113, 123–124 (1921), or are not navigable and never have been but may become so by reasonable improvements, see Appalachian Elec. Power Co., supra, at 407–408. With respect to the federal commerce power, the inquiry regarding navigation historically focused on interstate commerce. See The Daniel Ball, supra, at 564. And, of course, the commerce power extends beyond navigation. See Kaiser Aetna v. United States, 444 U.S. 164, 173–174 (1979). In contrast, for title purposes, the inquiry depends only on navigation and not on interstate travel. See Utah, supra, at 76. This list of differences is not exhaustive. Indeed, “[e]ach application of [the Daniel Ball] test . . . is apt to uncover variations and refinements which require further elaboration.” Ap- palachian Elec. Power Co., supra, at 406. IV A The primary flaw in the reasoning of the Montana Supreme Court lies in its treatment of the question of river segments and overland portage. To determine title to a riverbed under the equal-footing doctrine, this Court considers the river on a segment- by-segment basis to assess whether the segment of the river, under which the riverbed in dispute lies, is navigable or not. In United States v. Utah, for example, the Court noted, “the controversy relates only to the sections of the rivers which are described in the complaint, and the Master has limited his findings and conclusions as to navigability accordingly. The propriety of this course, in view of the physical characteristics of the streams, is apparent. Even where the navigability of a river, speaking generally, is a matter of common knowledge, and hence one of which judicial notice may be taken, it may yet be a question, to be determined upon evidence, how far navigability extends.” 283 U. S., at 77. The Court went on to conclude, after reciting and assessing the evidence, that the Colorado River was navigable for its first roughly 4-mile stretch, nonnavigable for the next roughly 36-mile stretch, and navigable for its remaining 149 miles. Id., at 73–74, 79–81, 89. The Court noted the importance of determining “the exact point at which navigability may be deemed to end.” Id., at 90. Similarly, in Brewer-Elliott Oil & Gas Co. v. United States, 260 U.S. 77, 85 (1922), the Court examined the segment of the Arkansas River that ran along the Osage Indian Reservation, assessing whether the Arkansas River was “navigable in fact at the locus in quo.” The Court concluded that the United States originally, and the Osages as its grantees, unequivocally held title to the riverbeds because the Arkansas River “is and was not navigable at the place where the river bed lots, here in controversy, are.” Id., at 86. The Court found the segment of river along the reservation to be nonnavigable even though a segment of the river that began further downstream was navigable. Ibid. See also Oklahoma, supra, at 583, 584, 587–588, 589–591 (noting that “how far up the streams navigability extended was not known”; assessing separately the segments of the Red River above and below its confluence with the Washita River within Oklahoma’s borders; and concluding that neither segment, and hence “no part of the river within Oklahoma,” was navigable). The Montana Supreme Court discounted the segment-by-segment approach of this Court’s cases, calling it “a piecemeal classification of navigability—with some stretches declared navigable, and others declared non-navigable.” 355 Mont., at 440–442, 229 P. 3d, at 448–449. This was error. The segment-by-segment approach to navigability for title is well settled, and it should not be disregarded. A key justification for sovereign ownership of navigable riverbeds is that a contrary rule would allow private riverbed owners to erect improvements on the riverbeds that could interfere with the public’s right to use the waters as a highway for commerce. While the Federal Government and States retain regulatory power to protect public navigation, allocation to the State of the beds underlying navigable rivers reduces the possibility of conflict between private and public interests. See Utah, supra, at 82–83; Packer, 137 U. S., at 667. By contrast, segments that are nonnavigable at the time of statehood are those over which commerce could not then occur. Thus, there is no reason that these segments also should be deemed owned by the State under the equal-footing doctrine. Practical considerations also support segmentation. Physical conditions that affect navigability often vary significantly over the length of a river. This is particularly true with longer rivers, which can traverse vastly different terrain and the flow of which can be affected by varying local climates. The Missouri River provides an excellent example: Between its headwaters and mouth, it runs for over 2,000 miles out of steep mountains, through canyons and upon rocky beds, over waterfalls and rapids, and across sandy plains, capturing runoff from snow melt and farmland rains alike. These shifts in physical conditions provide a means to determine appropriate start points and end points for the segment in question. Topographical and geographical indicators may assist. See, e.g., Utah, supra, at 77–80 (gradient changes); Oklahoma, 258 U. S., at 589 (location of tributary providing additional flow). A segment approach to riverbed title allocation under the equal-footing doctrine is consistent with the manner in which private parties seek to establish riverbed title. For centuries, where title to the riverbed was not in the sovereign, the common-law rule for allocating riverbed title among riparian landowners involved apportionment defined both by segment (each landowner owns bed and soil along the length of his land adjacent) and thread (each landowner owns bed and soil to the center of the stream). See J. Angell, A Treatise on the Law of Watercourses 18 (6th ed. 1869); Tyler v. Wilkinson, 24 F. Cas. 472, 474 (No. 14,312) (CC RI 1827) (Story, J.). Montana, moreover, cannot suggest that segmentation is inadministrable when the state courts managed to divide up and apportion the underlying riverbeds for purposes of determining their value and the corresponding rents owed by PPL. The Montana Supreme Court, relying upon Utah, de- cided that the segment-by-segment approach is inapplicable here because it “does not apply to ‘short interruption[s] of navigability in a stream otherwise navigable.’ ” 355 Mont., at 442, 229 P. 3d, at 449 (quoting Utah, 283 U. S., at 77). This was mistaken. In Utah, this Court noted in pass- ing that the facts of the case concerned “long reaches with particular characteristics of navigability or non-navigability” rather than “short interruption[s].” Id., at 77. The Court in Utah did not say the case would have a different outcome if a “short interruption” were concerned. Ibid. Even if the law might find some nonnavigable segments so minimal that they merit treatment as part of a longer, navigable reach for purposes of title under the equal-footing doctrine, it is doubtful that any of the segments in this case would meet that standard, and one—the Great Falls reach—certainly would not. As an initial matter, the kinds of considerations that would define a de minimis exception to the segment-by-segment approach would be those related to principles of ownership and title, such as inadministrability of parcels of exceedingly small size, or worthlessness of the parcels due to overdivision. See Heller, The Tragedy of the Anticommons, 111 Harv. L. Rev. 621, 682–684 (1998) (explaining that dividing prop- erty into square-inch parcels, could, absent countervail- ing legal mechanisms, “paralyze the alienability of scarce resources . . . or diminish their value too drastically”). An analysis of segmentation must be sensibly applied. A comparison of the nonnavigable segment’s length to the overall length of the stream, for instance, would be simply irrelevant to the issue at hand. A number of the segments at issue here are both discrete, as defined by physical features characteristic of navigability or nonnavigability, and substantial, as a matter of administrability for title purposes. This is best illustrated by the Great Falls reach, which is 17 miles long and has distinct drops including five waterfalls and continuous rapids in between. There is plenty of reason to doubt that reach’s navigability based on the presence of the series of falls. There is also reason to think that title to that segment of bed would not be worthless or inadministrable. Indeed, the State sought and was awarded rent in the amount of $41 million for PPL’s various hydroelectric facilities attached to the riverbeds, half of which are along the Great Falls reach. Applying its “short interruptions” approach, the Montana Supreme Court decided that the Great Falls reach was navigable because it could be managed by way of land route portage. 355 Mont., at 440, 442, 229 P. 3d, at 447, 449. The court noted in particular the portage of Lewis and Clark’s expedition. Ibid. Yet that very portage reveals the problem with the Montana Supreme Court’s analysis. Leaving behind their larger boats, Lewis and Clark transported their supplies and some small canoes about 18 miles over land, which took at least 11 days and probably more. See Lewis and Clark Journals 126–152; 9 Journals of the Lewis & Clark Expedition 173; Dear Brother 109. Even if portage were to take travelers only one day, its significance is the same: it demonstrates the need to bypass the river segment, all because that part of the river is nonnavigable. Thus, the Montana Supreme Court was wrong to state, with respect to the Great Falls reach and other stretches of the rivers in question, that portages “are not sufficient to defeat a finding of navigability.” 355 Mont., at 438, 229 P. 3d, at 446. In most cases, they are, because they require transportation over land rather than over the water. This is such a case, at least as to the Great Falls reach. In reaching its conclusion that the necessity of portage does not undermine navigability, the Montana Supreme Court misapplied this Court’s decision in The Montello, 20 Wall. 430. See 355 Mont., at 438, 229 P. 3d, at 446. The consideration of portage in The Montello was for a different purpose. The Court did not seek to determine whether the river in question was navigable for title purposes but instead whether it was navigable for purposes of determining whether boats upon it could be regulated by the Federal Government. 20 Wall., at 439, 445. The primary focus in The Montello was not upon navigability in fact but upon whether the river was a “navigable water of the United States.” Id., at 439, 443. The latter inquiry is doctrinally distinct. It turns upon whether the river “forms by itself, or by its connection with other waters, a continued highway over which commerce is, or may be, carried with other States or foreign countries in the customary modes in which such commerce is conducted by water.” Id., at 439 (citing The Daniel Ball, 10 Wall. 557). It is language similar to “continued highway” that Montana urges the Court to import into the title context in lieu of the Court’s established segmentation approach. Brief for Respondent 42–43, n. 16. The Montello reasonably concluded that the portages required in that case did not prevent the river from being part of a channel of interstate commerce. Portages continued that channel because goods could be successfully transported interstate, in part upon the waters in question. This provided sufficient basis to regulate steamboats at places where those boats could and did, in fact, navigate portions of the river. 20 Wall., at 445. Here, by contrast, the question regards ownership of the bed under river segments that the Montana Supreme Court, by calling them “interruptions in the navigation,” 355 Mont., at 442, 229 P. 3d, at 449, acknowledges were nonnavigable. The reasoning and the inquiry of The Montello does not control the outcome where the quite different concerns of the riverbed title context apply. Having clarified that portages may defeat navigability for title purposes, and do so with respect to the Great Falls reach, the Court sees no evidence in the record that could demonstrate that the Great Falls reach was navigable. Montana does not dispute that overland portage was necessary to traverse that reach. Indeed, the State admits “the falls themselves were not passable by boat at statehood.” Brief for Respondent 10. And the trial court noted the falls had never been navigated. App. to Pet. for Cert. 137. Based on these statements, this Court now con- cludes, contrary to the Montana Supreme Court’s decision, that the 17-mile Great Falls reach, at least from the head of the first waterfall to the foot of the last, is not navigable for purposes of riverbed title under the equal-footing doctrine. This Court also determines, based on evidence in the record, that there is a significant likelihood that some of the other river stretches in dispute also fail the federal test of navigability for the purpose of determining title. For example, as to the disputed segment of the Clark Fork River, the Montana Supreme Court incorrectly stated the sole evidence for nonnavigability “consists of conclusory statements . . . without any specific factual support.” 355 Mont., at 440, 229 P. 3d, at 448. In fact, PPL introduced a report of the U. S. Army Corps of Engineers from 1891, two years after Montana’s date of statehood, documenting that the portion of the Clark Fork river between Missoula and Lake Pend Oreille (which includes the location of PPL’s Thompson Falls facility) had a fall of about 1,100 feet in 250 miles and “is a mountain torrential stream, full of rocks, rapids, and falls, . . . utterly unnavigable, and in- capable of being made navigable except at an enormous cost.” 2 H. R. Exec. Doc., pt. 5, at 3250; see App. 379–380 (Docket No. 169). The report based its conclusions on various failed attempts to navigate the river. It found the Thompson Falls “a complete obstruction to navigation” and the river around that area “exceedingly rapid, rough, and full of rocks.” 2 H. R. Exec. Doc., pt. 5, at 3251. This was consistent with a 1910 Federal District Court decree. The decree adjudicated a title dispute between two private parties over the riverbed near and under Thompson Falls and declared the river at that place “was and is a non-navigable stream incapable of carrying the products of the country in the usual manner of water transportation.” Steele v. Donlan, Equity No. 950 (CC D Mont., July 19, 1910), p. 1; see App. 380–381 (Docket No. 169). While the ultimate decision as to this and the other disputed river stretches is to be determined, in the first instance, by the Montana courts upon remand, the relevant evidence should be assessed in light of the principles discussed in this opinion. B The Montana Supreme Court further erred as a matter of law in its reliance upon the evidence of present-day, primarily recreational use of the Madison River. Error is not inherent in a court’s consideration of such evidence, but the evidence must be confined to that which shows the river could sustain the kinds of commercial use that, as a realistic matter, might have occurred at the time of statehood. Navigability must be assessed as of the time of statehood, and it concerns the river’s usefulness for “ ‘trade and travel,’ ” rather than for other purposes. See Utah, 283 U. S., at 75–76. Mere use by initial explorers or trappers, who may have dragged their boats in or alongside the river despite its nonnavigability in order to avoid getting lost, or to provide water for their horses and themselves, is not itself enough. See Oregon, 295 U. S., at 20–21 (evidence that “trappers appear to have waded or walked” through the river, dragging their boats rather than floating them, had “no bearing on navigability”). True, river segments are navigable not only if they “[were] used,” but also if they “[were] susceptible of being used,” as highways of commerce at the time of statehood. Utah, supra, at 76 (internal quotation marks omitted). Evidence of recreational use, depending on its nature, may bear upon susceptibility of commercial use at the time of statehood. See Appalachian Elec. Power Co., 311 U. S., at 416 (“[P]ersonal or private use by boats demonstrates the availability of the stream for the simpler types of commercial navigation”); Utah, 283 U. S., at 82 (fact that actual use has “been more of a private nature than of a public, commercial sort . . . cannot be regarded as controlling”). Similarly, poststatehood evidence, depending on its nature, may show susceptibility of use at the time of statehood. See id., at 82–83 (“[E]xtensive and continued [historical] use for commercial purposes” may be the “most persuasive” form of evidence, but the “crucial question” is the potential for such use at the time of statehood, rather than “the mere manner or extent of actual use”). Evidence of present-day use may be considered to the extent it informs the historical determination whether the river segment was susceptible of use for commercial navigation at the time of statehood. For the susceptibility analysis, it must be determined whether trade and travel could have been conducted “in the customary modes of trade and travel on water,” over the relevant river segment “in [its] natural and ordinary condition.” Id., at 76 (internal quotation marks omitted). At a minimum, therefore, the party seeking to use present-day evidence for title purposes must show: (1) the watercraft are meaningfully similar to those in customary use for trade and travel at the time of statehood; and (2) the river’s poststatehood condition is not materially different from its physical con- dition at statehood. See also Oregon, supra, at 18 (find- ing that scientific and historical evidence showed that the physical condition of particular water bodies had not varied substantially since statehood in a way that might affect navigation). If modern watercraft permit navigability where the historical watercraft would not, or if the river has changed in ways that substantially improve its navigability, then the evidence of present-day use has little or no bearing on navigability at statehood. The Montana Supreme Court opinion offered no indication that it made these necessary findings. The court concluded the evidence of present-day use of the Madison was probative of its susceptibility of use at statehood, but there is no apparent basis for its conclusion. 355 Mont., at 442–443, 438–439, 229 P. 3d, at 449, 446–447. The court did not find the watercraft similar to those used at the time of statehood, and the State’s evidence of present-day use for recreational fishing did not indicate what types of boats are now used. App. 46–48. Modern recreational fishing boats, including inflatable rafts and lightweight canoes or kayaks, may be able to navigate waters much more shallow or with rockier beds than the boats customarily used for trade and travel at statehood. As to the river’s physical condition, the Montana Supreme Court did not assess with care PPL’s evidence about changes to the river’s flow and the location and pattern of its channel since statehood. The affidavit of PPL’s expert in fluvial geomorphology—the study of river-related landforms—at least suggests that as a result of PPL’s dams, the river has become “less torrential” in high flow periods and less shallow in low flow periods. App. 575–577 (Docket No. 170). Thus, the river may well be easier to navigate now than at statehood. The Montana Supreme Court altogether ignored the expert’s reasoning about the past condition of the river’s channels and the significance of that information for navigability. Further, contrary to the Montana Supreme Court’s suggestion, the expert’s affidavit was not mere evidence of change in “seasonal variations” of water depth. 355 Mont., at 440, 229 P. 3d, at 448. It provided meaningful evidence that the river’s conditions had changed since statehood in ways that made present-day navigation of the river easier in all seasons than it was at the relevant time. While the Montana court was correct that a river need not be susceptible of navigation at every point during the year, neither can that susceptibility be so brief that it is not a commercial reality. Against this background, the present-day recreational use of the river did not bear on navigability for purposes of title under the equal-footing doctrine. The Montana Supreme Court’s reliance upon the State’s evidence of present-day, recreational use, at least without further inquiry, was wrong as a matter of law. C The above analysis is sufficient to require reversal of the grant of summary judgment to Montana. Therefore, the Court declines to decide whether the Montana Supreme Court further erred as to the burden of proof regarding navigability. D As a final contention, the State of Montana suggests that denying the State title to the riverbeds here in dispute will undermine the public trust doctrine, which concerns public access to the waters above those beds for purposes of navigation, fishing, and other recreational uses. Brief for Respondent 20, 24–26. This suggestion underscores the State’s misapprehension of the equal footing and public trust doctrines. The public trust doctrine is of ancient origin. Its roots trace to Roman civil law and its principles can be found in the English common law on public navigation and fishing rights over tidal lands and in the state laws of this country. See Coeur d’Alene, 521 U. S., at 284–286; Illinois Central R. Co. v. Illinois, 146 U.S. 387, 458 (1892); D. Slade, Putting the Public Trust Doctrine to Work 3–8, 15–24 (1990); see, e.g., National Audubon Soc. v. Superior Court of Alpine Cty., 33 Cal. 3d 419, 433–441, 658 P.2d 709, 718–724 (1983); Arnold v. Mundy, 6 N. J. L. 1, 9–10 (1821). Unlike the equal-footing doctrine, however, which is the constitutional foundation for the navigability rule of riverbed title, the public trust doctrine remains a matter of state law, see Coeur d’Alene, supra, at 285 (Illinois Central, a Supreme Court public trust case, was “ ‘necessarily a statement of Illinois law’ ”); Appleby v. City of New York, 271 U.S. 364, 395 (1926) (same), subject as well to the federal power to regulate vessels and navigation under the Commerce Clause and admiralty power. While equal-footing cases have noted that the State takes title to the navigable waters and their beds in trust for the public, see Shively, 152 U. S., at 49, 15–17, 24, 46, the contours of that public trust do not depend upon the Constitution. Under accepted principles of federalism, the States retain residual power to determine the scope of the public trust over waters within their borders, while federal law determines riverbed title under the equal-footing doctrine. V As the litigation history of this case shows, Montana filed its claim for riverbed rent over a century after the first of the dams was built upon the riverbeds. Montana had not sought compensation before then, despite its full awareness of PPL’s hydroelectric projects and despite the State’s own participation in the projects’ federal licensing process. While this Court does not reach the question, it may be that by virtue of the State’s sovereignty, neither laches nor estoppel could apply in a strict sense to bar the State’s much belated claim. Still, the reliance by PPL and its predecessors in title upon the State’s long failure to assert title is some evidence to support the conclusion that the river segments were nonnavigable for purposes of the equal-footing doctrine. The Montana Supreme Court’s ruling that Montana owns and may charge for use of riverbeds across the State was based upon an infirm legal understanding of this Court’s rules of navigability for title under the equal-footing doctrine. As the Court said in Brewer-Elliott, “It is not for a State by courts or legislature, in dealing with the general subject of beds or streams, to adopt a retroactive rule for determining navigability which . . . would enlarge what actually passed to the State, at the time of her admission, under the constitutional rule of equality here invoked.” 260 U. S., at 88. * * * The judgment of the Montana Supreme Court is reversed, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered.

566.US.2011_11-166

To finance the purchase of a commercial property and associated renovation and construction costs, petitioners (debtors) obtained a secured loan from an investment fund, for which respondent (Bank) serves as trustee. The debtors ultimately became insolvent, and sought relief under Chapter 11 of the Bankruptcy Code. Pursuant to 11 U. S. C. §1129(b)(2)(A), the debtors sought to confirm a “cramdown” bankruptcy plan over the Bank’s objection. That plan proposed selling substantially all of the debtors’ property at an auction, and using the sale proceeds to repay the Bank. Under the debtors’ proposed auction procedures, the Bank would not be permitted to bid for the property using the debt it is owed to offset the purchase price, a practice known as “credit-bidding.” The Bankruptcy Court denied the debtors’ request, concluding that the auction procedures did not comply with §1129(b)(2)(A)’s requirements for cramdown plans. The Seventh Circuit affirmed, holding that §1129(b)(2)(A) does not permit debtors to sell an encumbered asset free and clear of a lien without permitting the lienholder to credit-bid. Held: The debtors may not obtain confirmation of a Chapter 11 cramdown plan that provides for the sale of collateral free and clear of the Bank’s lien, but does not permit the Bank to credit-bid at the sale. Pp. 3–10. (a) A Chapter 11 plan proposed over the objection of a “class of secured claims” must meet one of three requirements in order to be deemed “fair and equitable,” and therefore confirmable. The secured creditor may retain its lien on the property and receive deferred cash payments, §1129(b)(2)(A)(i); the debtors may sell the property free and clear of the lien, “subject to section 363(k)”—which permits the creditor to credit-bid at the sale—and provide the creditor with a lien on the sale proceeds, §1129(b)(2)(A)(ii); or the plan may provide the secured creditor with the “indubitable equivalent” of its claim, §1129(b)(2)(A)(iii). Here, the debtors proposed to sell their property free and clear of the Bank’s liens and repay the Bank with the sale proceeds, as contemplated by clause (ii). Because the debtors’ auction procedures do not permit the Bank to credit-bid, however, the proposed sale cannot satisfy the requirements of clause (ii). The debtors claim their plan can instead satisfy clause (iii) by providing the Bank with the “indubitable equivalent” of its secured claim, in the form of cash generated by the auction. The debtors’ reading of §1129(b)(2)(A), under which clause (iii) permits precisely what clause (ii) proscribes, is hyperliteral and contrary to common sense. “[I]t is a commonplace of statutory construction that the specific governs the general.” Morales v. Trans World Airlines, Inc., 504 U.S. 374, 384. Here, where general and specific authorizations exist side-by-side, the general/specific canon avoids rendering superfluous a specific provision that is swallowed by the general one. See D. Ginsberg & Sons, Inc. v. Popkin, 285 U.S. 204, 208. As applied to §1129(b)(2)(A), the canon provides that the “general language” of clause (iii), “although broad enough to include it, will not be held to apply to a matter specifically dealt with” in clause (ii). 285 U. S., at 208. Although the canon can be overcome by other textual indications of statutory meaning, the debtors point to none here. Pp. 3–8. (b) None of the debtors’ objections to this approach is valid. Pp. 8–9. 651 F.3d 642, affirmed. Scalia, J., delivered the opinion of the Court, in which all other Members joined, except Kennedy, J., who took no part in the decision of the case.

We consider whether a Chapter 11 bankruptcy plan may be confirmed over the objection of a secured creditor pursuant to 11 U. S. C. §1129(b)(2)(A) if the plan provides for the sale of collateral free and clear of the creditor’s lien, but does not permit the creditor to “credit-bid” at the sale. I In 2007, petitioners RadLAX Gateway Hotel, LLC, and RadLAX Gateway Deck, LLC (hereinafter debtors), purchased the Radisson Hotel at Los Angeles International Airport, together with an adjacent lot on which the debtors planned to build a parking structure. To finance the purchase, the renovation of the hotel, and construction of the parking structure, the debtors obtained a $142 million loan from Longview Ultra Construction Loan Investment Fund, for which respondent Amalgamated Bank (hereinafter creditor or Bank) serves as trustee. The lenders obtained a blanket lien on all of the debtors’ assets to secure the loan. Completing the parking structure proved more expensive than anticipated, and within two years the debtors had run out of funds and were forced to halt construction. By August 2009, they owed more than $120 million on the loan, with over $1 million in interest accruing every month and no prospect for obtaining additional funds to complete the project. Both debtors filed voluntary petitions for relief under Chapter 11 of the Bankruptcy Code. A Chapter 11 bankruptcy is implemented according to a “plan,” typically proposed by the debtor, which divides claims against the debtor into separate “classes” and specifies the treatment each class will receive. See 11 U. S. C. §1123. Generally, a bankruptcy court may confirm a Chapter 11 plan only if each class of creditors affected by the plan consents. See §1129(a)(8). Section 1129(b) creates an exception to that general rule, per- mitting confirmation of nonconsensual plans—commonly known as “cramdown” plans—if “the plan does not discriminate unfairly, and is fair and equitable, with respect to each class of claims or interests that is impaired under, and has not accepted, the plan.” Section 1129(b)(2)(A), which we review in further depth below, establishes criteria for determining whether a cramdown plan is “fair and equitable” with respect to secured claims like the Bank’s. In 2010, the RadLAX debtors submitted a Chapter 11 plan to the United States Bankruptcy Court for the Northern District of Illinois. The plan proposed to dissolve the debtors and to sell substantially all of their assets pursuant to procedures set out in a contemporaneously filed “Sale and Bid Procedures Motion.” Specifically, the debtors sought to auction their assets to the highest bidder, with the initial bid submitted by a “stalking horse”—a potential purchaser who was willing to make an advance bid of $47.5 million.[1] The sale proceeds would be used to fund the plan, primarily by repaying the Bank. Of course the Bank itself might wish to obtain the property if the alternative would be receiving auction proceeds that fall short of the property’s full value. Under the debtors’ proposed auction procedures, however, the Bank would not be permitted to bid for the property using the debt it is owed to offset the purchase price, a practice known as “credit-bidding.” Instead, the Bank would be forced to bid cash. Correctly anticipating that the Bank would object to this arrangement, the debtors sought to confirm their plan under the cramdown provisions of §1129(b)(2)(A). The Bankruptcy Court denied the debtors’ Sale and Bid Procedures Motion, concluding that the proposed auction procedures did not comply with §1129(b)(2)(A)’s requirements for cramdown plans. In re River Road Hotel Partners, LLC, Case No. 09 B 30029 (ND Ill., Oct. 5, 2010), App. to Pet. for Cert. 40a. The Bankruptcy Court certified an appeal directly to the United States Court of Appeals for the Seventh Circuit. That court accepted the certifi- cation and affirmed, holding that §1129(b)(2)(A) does not permit debtors to sell an encumbered asset free and clear of a lien without permitting the lienholder to credit-bid. River Road Hotel Partners, LLC, et al. v. Amalgamated Bank, 651 F.3d 642 (2011). We granted certiorari. 565 U. S. ___ (2011). II A A Chapter 11 plan confirmed over the objection of a “class of secured claims” must meet one of three requirements in order to be deemed “fair and equitable” with respect to the nonconsenting creditor’s claim. The plan must provide: “(i)(I) that the holders of such claims retain the liens securing such claims, whether the property subject to such liens is retained by the debtor or transferred to another entity, to the extent of the allowed amount of such claims; and (II) that each holder of a claim of such class receive on account of such claim deferred cash payments totaling at least the allowed amount of such claim, of a value, as of the effective date of the plan, of at least the value of such holder’s interest in the estate’s interest in such property; “(ii) for the sale, subject to section 363(k) of this title, of any property that is subject to the liens securing such claims, free and clear of such liens, with such liens to attach to the proceeds of such sale, and the treatment of such liens on proceeds under clause (i) or (iii) of this subparagraph; or “(iii) for the realization by such holders of the indubitable equivalent of such claims.” 11 U. S. C. §1129(b)(2)(A). Under clause (i), the secured creditor retains its lien on the property and receives deferred cash payments. Under clause (ii), the property is sold free and clear of the lien, “subject to section 363(k),” and the creditor receives a lien on the proceeds of the sale. Section 363(k), in turn, provides that “unless the court for cause orders otherwise the holder of such claim may bid at such sale, and, if the holder of such claim purchases such property, such holder may offset such claim against the purchase price of such property”—i.e., the creditor may credit-bid at the sale, up to the amount of its claim.[2] Finally, under clause (iii), the plan provides the secured creditor with the “indubitable equivalent” of its claim. The debtors in this case have proposed to sell their property free and clear of the Bank’s liens, and to repay the Bank using the sale proceeds—precisely, it would seem, the disposition contemplated by clause (ii). Yet since the debtors’ proposed auction procedures do not permit the Bank to credit-bid, the proposed sale cannot satisfy the requirements of clause (ii).[3] Recognizing this problem, the debtors instead seek plan confirmation pursuant to clause (iii), which—unlike clause (ii)—does not expressly foreclose the possibility of a sale without credit-bidding. According to the debtors, their plan can satisfy clause (iii) by ultimately providing the Bank with the “indubitable equivalent” of its secured claim, in the form of cash generated by the auction. We find the debtors’ reading of §1129(b)(2)(A)—under which clause (iii) permits precisely what clause (ii) proscribes—to be hyperliteral and contrary to common sense. A well established canon of statutory interpretation succinctly captures the problem: “[I]t is a commonplace of statutory construction that the specific governs the general.” Morales v. Trans World Airlines, Inc., 504 U.S. 374, 384 (1992). That is particularly true where, as in §1129(b)(2)(A), “Congress has enacted a comprehensive scheme and has deliberately targeted specific problems with specific solutions.” Varity Corp. v. Howe, 516 U.S. 489, 519 (1996) (Thomas, J., dissenting); see also HCSC-Laundry v. United States, 450 U.S. 1, 6 (1981) (per curiam) (the specific governs the general “particularly when the two are interrelated and closely positioned, both in fact being parts of [the same statutory scheme]”). The general/specific canon is perhaps most frequently applied to statutes in which a general permission or prohibition is contradicted by a specific prohibition or permission. To eliminate the contradiction, the specific provision is construed as an exception to the general one. See, e.g., Morton v. Mancari, 417 U.S. 535, 550–551 (1974). But the canon has full application as well to statutes such as the one here, in which a general authorization and a more limited, specific authorization exist side-by-side. There the canon avoids not contradiction but the superfluity of a specific provision that is swallowed by the general one, “violat[ing] the cardinal rule that, if possible, effect shall be given to every clause and part of a statute.” D. Ginsberg & Sons, Inc. v. Popkin, 285 U.S. 204, 208 (1932). The terms of the specific authorization must be complied with. For example, in the last cited case a provision of the Bankruptcy Act prescribed in great detail the procedures governing the arrest and detention of bankrupts about to leave the district in order to avoid examination. The Court held that those prescriptions could not be avoided by relying upon a general provision of the Act authoriz- ing bankruptcy courts to “ ‘make such orders, issue such process, and enter such judgments in addition to those spe- cifically provided for as may be necessary for the enforcement of the provisions of [the] Act.’ ” Id., at 206 (quoting Bankruptcy Act of 1898, §2(15), 30Stat. 546). The Court said that “[g]eneral language of a statutory provision, although broad enough to include it, will not be held to apply to a matter specifically dealt with in another part of the same enactment.” 285 U. S., at 208. We recently quoted that language approvingly in Bloate v. United States, 559 U. S. ___, ___ (2010) (slip op., at 10). Or as we said in a much earlier case: “It is an old and familiar rule that, where there is, in the same statute, a particular enactment, and also a general one, which, in its most comprehensive sense, would include what is embraced in the former, the particular enactment must be operative, and the general enactment must be taken to affect only such cases within its general language as are not within the provisions of the particular enactment. This rule applies wherever an act contains general provisions and also special ones upon a subject, which, standing alone, the general provisions would include.” United States v. Chase, 135 U.S. 255, 260 (1890) (citations and internal quotation marks omitted). Here, clause (ii) is a detailed provision that spells out the requirements for selling collateral free of liens, while clause (iii) is a broadly worded provision that says nothing about such a sale. The general/specific canon explains that the “general language” of clause (iii), “although broad enough to include it, will not be held to apply to a matter specifically dealt with” in clause (ii). D. Ginsberg & Sons, Inc., supra, at 208. Of course the general/specific canon is not an absolute rule, but is merely a strong indication of statutory meaning that can be overcome by textual indications that point in the other direction. The debtors point to no such indi- cation here. One can conceive of a statutory scheme in which the specific provision embraced within a general one is not superfluous, because it creates a so-called safe harbor. The debtors effectively contend that that is the case here—clause (iii) (“indubitable equivalent”) being the general rule, and clauses (i) and (ii) setting forth procedures that will always, ipso facto, establish an “indubitable equivalent,” with no need for judicial evaluation. But the structure here would be a surpassingly strange manner of accomplishing that result—which would normally be achieved by setting forth the “indubitable equivalent” rule first (rather than last), and establishing the two safe harbors as provisos to that rule. The structure here suggests, to the contrary, that (i) is the rule for plans under which the creditor’s lien remains on the property, (ii) is the rule for plans under which the property is sold free and clear of the creditor’s lien, and (iii) is a residual provision covering dispositions under all other plans—for example, one under which the creditor receives the property itself, the “indubitable equivalent” of its secured claim. Thus, debtors may not sell their property free of liens under §1129(b)(2)(A) without allowing lienholders to credit-bid, as required by clause (ii). B None of the debtors’ objections to this approach is valid. The debtors’ principal textual argument is that §1129(b)(2)(A) “unambiguously provides three distinct options for confirming a Chapter 11 plan over the objection of a secured creditor.” Brief for Petitioners 15 (capitalization and bold typeface removed). With that much we agree; the three clauses of §1129(b)(2)(A) are connected by the disjunctive “or.” The debtors contend that our interpretation of §1129(b)(2)(A) “transforms ‘or’ into ‘and.’ ” Reply Brief for Petitioners 3. But that is not so. The question here is not whether debtors must comply with more than one clause, but rather which one of the three they must satisfy. Debtors seeking to sell their property free of liens under §1129(b)(2)(A) must satisfy the requirements of clause (ii), not the requirements of both clauses (ii) and (iii). The debtors make several arguments against applying the general/specific canon. They contend that clause (ii) is no more specific than clause (iii), because the former provides a procedural protection to secured creditors (credit-bidding) while the latter provides a substantive protection (indubitable equivalence). As a result, they say, clause (ii) is not “a limiting subset” of clause (iii), which (according to their view) application of the general/specific canon requires. Brief for Petitioners 30–31; Reply Brief for Petitioners 5–6. To begin with, we know of no authority for the proposition that the canon is confined to situations in which the entirety of the specific provision is a “subset” of the general one. When the conduct at issue falls within the scope of both provisions, the specific presumptively governs, whether or not the specific provision also applies to some conduct that falls outside the general. In any case, we think clause (ii) is entirely a subset. Clause (iii) applies to all cramdown plans, which include all of the plans within the more narrow category described in clause (ii).[4] That its requirements are “substantive” whereas clause (ii)’s are “procedural” is quite beside the point. What counts for application of the general/specific canon is not the nature of the provisions’ prescriptions but their scope. Finally, the debtors contend that the Court of Appeals conflated approval of bid procedures with plan confirmation. Brief for Petitioners 39. They claim the right to pursue their auction now, leaving it for the Bankruptcy Judge to determine, at the confirmation stage, whether the resulting plan (funded by auction proceeds) provides the Bank with the “indubitable equivalent” of its secured claim. Under our interpretation of §1129(b)(2)(A), how- ever, that approach is simply a nonstarter. As a matter of law, no bid procedures like the ones proposed here could satisfy the requirements of §1129(b)(2)(A), and the distinction between approval of bid procedures and plan confirmation is therefore irrelevant. III The parties debate at some length the purposes of the Bankruptcy Code, pre-Code practices, and the merits of credit-bidding. To varying extents, some of those debates also occupied the attention of the Courts of Appeals that considered the question presented here. See, e.g., In re Philadelphia Newspapers, LLC, 599 F.3d 298, 314–317 (CA3 2010); id., at 331–337 (Ambro, J., dissenting). But nothing in the generalized statutory purpose of protecting secured creditors can overcome the specific manner of that protection which the text of §1129(b)(2)(A) contains. As for pre-Code practices, they can be relevant to the interpretation of an ambiguous text, but we find no textual ambiguity here. And the pros and cons of credit-bidding are for the consideration of Congress, not the courts. The Bankruptcy Code standardizes an expansive (and sometimes unruly) area of law, and it is our obligation to interpret the Code clearly and predictably using well established principles of statutory construction. See United States v. Ron Pair Enterprises, Inc., 489 U.S. 235, 240–241 (1989). Under that approach, this is an easy case. Because the RadLAX debtors may not obtain confirmation of a Chapter 11 cramdown plan that provides for the sale of collateral free and clear of the Bank’s lien, but does not permit the Bank to credit-bid at the sale, we affirm the judgment of the Court of Appeals. It is so ordered. Justice Kennedy took no part in the decision of this case. Notes 1 In a later proposal, the stalking-horse bid increased to $55 million. The precise amount of the bid is not relevant here. 2 The ability to credit-bid helps to protect a creditor against the risk that its collateral will be sold at a depressed price. It enables the creditor to purchase the collateral for what it considers the fair market price (up to the amount of its security interest) without committing additional cash to protect the loan. That right is particularly important for the Federal Government, which is frequently a secured creditor in bankruptcy and which often lacks appropriations authority to throw good money after bad in a cash-only bankruptcy auction. 3 Title 11 U. S. C. §363(k)—and by extension clause (ii)—provides an exception to the credit-bidding requirement if “the court for cause orders otherwise.” The Bankruptcy Court found that there was no “cause” to deny credit-bidding in this case, and the debtors have not appealed that disposition. 4 We are speaking here about whether clause (ii) is a subset for purposes of determining whether the canon applies. As we have described earlier, after applying the canon—ex post, so to speak—it ceases to be a subset, governing a situation to which clause (iii) will no longer be deemed applicable.

566.US.2011_10-788

Respondent, the chief investigator for a district attorney’s office, testified at grand jury proceedings that resulted in petitioner’s indictment. After the indictments were dismissed, petitioner brought an action under 42 U. S. C. §1983, alleging that respondent had conspired to present and did present false testimony to the grand jury. The Federal District Court denied respondent’s motion to dismiss on immunity grounds, but the Eleventh Circuit reversed, holding that respondent had absolute immunity from a §1983 claim based on his grand jury testimony. Held: A witness in a grand jury proceeding is entitled to the same absolute immunity from suit under §1983 as a witness who testifies at trial. Pp. 3–18. (a) Section 1983, which derives from §1 of the Civil Rights Act of 1871, was not meant to effect a radical departure from ordinary tort law and the common-law immunities applicable in tort suits. See, e.g., Burns v. Reed, 500 U.S. 478, 484. This interpretation of §1983 has been reaffirmed by the Court time and again. Thus, the Court looks to the common law for guidance in determining the scope of the immunities available in actions brought under §1983. See Kalina v. Fletcher, 522 U.S. 118, 123. Taking a “functional approach,” see, e.g., Forrester v. White, 484 U.S. 219, 224, the Court identifies those governmental functions that were historically viewed as so important and vulnerable to interference by means of litigation that some form of absolute immunity from civil liability was needed to ensure that they are “ ‘performed with independence and without fear of consequences,’ ” Pierson v. Ray, 386 U.S. 547, 554. The Court’s functional approach is tied to the common law’s identification of functions meriting the protection of absolute immunity, but the Court’s precedents have not mechanically duplicated the precise scope of the absolute immunity the common law provided to protect those functions. For example, it was common in 1871 for cases to be prosecuted by private parties, who did not enjoy absolute immunity from suit. But as the prosecutorial function was increasingly assumed by public officials, common-law courts held that public prosecutors, unlike their private predecessors, were absolutely immune from the types of tort claims that an aggrieved or vengeful criminal defendant was most likely to assert. This adaptation of prosecutorial immunity accommodated the special needs of public, as opposed to private, prosecutors. Thus, when the issue of prosecutorial immunity under §1983 reached this Court in Imbler v. Pachtman, 424 U.S. 409, the Court did not simply apply the scope of immunity recognized by common-law courts as of 1871 but instead relied substantially on post-1871 cases extending broad immunity to public prosecutors sued for common-law torts. Neither has the Court suggested that §1983 is simply a federalized amalgamation of pre-existing common-law claims. The new federal claim created by §1983 differs in important ways from pre-existing common-law torts. Accordingly, both the scope of the new tort and the scope of the absolute immunity available in §1983 actions differ in some respects from the common law. Pp. 3―9. (b) A trial witness sued under §1983 enjoys absolute immunity from any claim based on his testimony. Briscoe v. LaHue, 460 U.S. 352. Without absolute immunity, the truth-seeking process would be impaired as witnesses might be reluctant to testify, and even a witness who took the stand “might be inclined to shade his testimony in favor of the potential plaintiff” for “fear of subsequent liability.” Id., at 333. These factors apply with equal force to grand jury witnesses. In both contexts, a witness’ fear of retaliatory litigation may deprive the tribunal of critical evidence. And in neither context is the deterrent of potential civil liability needed to prevent false testimony because other sanctions, chiefly prosecution for perjury, provide a sufficient deterrent. For the reasons identified in Briscoe, supra, at 342–344, there is no reason to distinguish law enforcement witnesses from lay witnesses in §1983 actions. And the rule that a grand jury witness has absolute immunity from any §1983 claim based on the witness’ testimony may not be circumvented by claiming that a grand jury witness conspired to present false testimony, or by using evidence of the witness’ testimony to support any other §1983 claim concerning the initiation or maintenance of a prosecution. Were it otherwise, a criminal defendant turned civil plaintiff could reframe a claim to attack the preparatory activity—such as a preliminary discussion in which the witness relates the substance of his intended testimony—rather than the absolutely immune actions themselves. Pp. 9−12. (c) Petitioner’s main argument is that under Malley v. Briggs, 475 U.S. 335, 340−341, and Kalina v. Fletcher, 522 U.S. 118, 131, grand jury witnesses who are “complaining witnesses” are not entitled to absolute immunity. But at the time §1983’s predecessor was enacted, a “complaining witness” was a party who procured an arrest and initiated a criminal prosecution. A “complaining witness” might testify, either before a grand jury or at trial, but testifying was not a necessary characteristic of a “complaining witness.” Thus, testifying, whether before a grand jury or at trial, was not the distinctive function performed by a “complaining witness.” A “complaining witness” cannot be held liable for perjurious trial testimony, see Briscoe, 460 U. S., at 326, and there is no more reason why a “complaining witness” should be subject to liability for testimony before a grand jury. Once the distinctive function performed by a “complaining witness” is understood, it is apparent that a law enforcement officer who testifies before a grand jury is not comparable to a “complaining witness” because it is not the officer who makes the critical decision to press criminal charges, but the prosecutor. It would be anomalous to permit a police officer testifying before a grand jury to be sued for maliciously procuring an unjust prosecution when it is the prosecutor, who is shielded by absolute immunity, who is actually responsible for the decision to initiate a prosecution. Petitioner also contends that the deterrent effect of civil liability is more needed in grand jury proceedings because trial witnesses face cross-examination. But the force of that argument is more than offset by the problem that allowing such civil actions would create—subversion of grand jury secrecy, which is essential to the proper functioning of the grand jury system. See United States v. Sells Engineering, Inc., 463 U.S. 418, 424. And finally, contrary to petitioner’s suggestion, recognizing absolute immunity for grand jury witnesses does not create an insupportable distinction between States that use grand juries and States that permit felony prosecutions to be brought by complaint or information. Most States that do not require an indictment for felonies provide a preliminary hearing at which witnesses testify, and the lower courts have held that preliminary hearing witnesses are protected by the same immunity accorded grand jury witnesses. Pp. 12−18. 611 F.3d 828, affirmed. Alito, J., delivered the opinion for a unanimous Court.

This case requires us to decide whether a “complaining witness” in a grand jury proceeding is entitled to the same immunity in an action under 42 U. S. C. §1983 as a witness who testifies at trial. We see no sound reason to draw a distinction for this purpose between grand jury and trial witnesses. I Petitioner Charles Rehberg, a certified public accountant, sent anonymous faxes to several recipients, including the management of a hospital in CityplaceAlbany, country-regionGeorgia, criticizing the hospital’s management and activities. In response, the local district attorney’s office, with the assistance of its chief investigator, respondent James Paulk, launched a criminal investigation of petitioner, allegedly as a favor to the hospital’s leadership. Respondent testified before a grand jury, and petitioner was then indicted for aggravated assault, burglary, and six counts of making harassing telephone calls. The indictment charged that petitioner had assaulted a hospital physician, Dr. James Hotz, after unlawfully entering the doctor’s home. Petitioner challenged the sufficiency of the indictment, and it was dismissed. A few months later, respondent returned to the grand jury, and petitioner was indicted again, this time for assaulting Dr. Hotz on August 22, 2004, and for making harassing phone calls. On this occasion, both the doctor and respondent testified. Petitioner challenged the suf- ficiency of this second indictment, claiming that he was “nowhere near Dr. Hotz” on the date in question and that “[t]here was no evidence whatsoever that [he] committed an assault on anybody.” 611 F.3d 828, 836 (CA11 2010). Again, the indictment was dismissed. While the second indictment was still pending, respondent appeared before a grand jury for a third time, and yet another indictment was returned. Petitioner was charged with assault and making harassing phone calls. This final indictment was ultimately dismissed as well. Petitioner then brought this action against respondent under Rev. Stat. §1979, 42 U. S. C. §1983. Petitioner alleged that respondent conspired to present and did present false testimony to the grand jury. Respondent moved to dismiss, arguing, among other things, that he was entitled to absolute immunity for his grand jury testimony. The United States District Court for the Middle District of Georgia denied respondent’s motion to dismiss, but the Court of Appeals reversed, holding, in accordance with Circuit precedent, that respondent was absolutely immune from a §1983 claim based on his grand jury testimony. The Court of Appeals noted petitioner’s allegation that respondent was the sole “complaining witness” before the grand jury, but the Court of Appeals declined to recognize a “complaining witness” exception to its precedent on grand jury witness immunity. See 611 F. 3d, at 839–840. “[A]llowing civil suits for false grand jury testimony,” the court reasoned, “would . . . emasculate the confidential nature of grand jury testimony, and eviscerate the traditional absolute immunity for witness testimony in judi- cial proceedings.” Id., at 840. The court went on to hold that respondent was entitled to absolute immunity, not only with respect to claims based directly on his grand jury testimony, but also with respect to the claim that he conspired to present such testimony. Id., at 841. To allow liability to be predicated on the alleged conspiracy, the court concluded, “ ‘would be to permit through the back door what is prohibited through the front.’ ” Ibid. (quoting Jones v. Cannon, 174 F.3d 1271, 1289 (CA11 1999)). We granted certiorari to resolve a Circuit conflict regarding the immunity of a “complaining witness” in a grand jury proceeding, 562 U. S. ___ (2011), and we now affirm. II Section 1983, which derives from §1 of the Civil Rights Act of 1871, 17Stat. 13, creates a private right of action to vindicate violations of “rights, privileges, or immunities secured by the Constitution and laws” of the United States. Under the terms of the statute, “ ‘[e]very person’ who acts under color of state law to deprive another of a constitutional right [is] answerable to that person in a suit for damages.” Imbler v. Pachtman, 424 U.S. 409, 417 (1976) (citing 42 U. S. C. §1983). A Despite the broad terms of §1983, this Court has long recognized that the statute was not meant to effect a radical departure from ordinary tort law and the common-law immunities applicable in tort suits. See, e.g., Burns v. Reed, 500 U.S. 478, 484 (1991). More than 60 years ago, in Tenney v. Brandhove, 341 U.S. 367 (1951), the Court held that §1983 did not abrogate the long-established absolute immunity enjoyed by legislators for actions taken within the legitimate sphere of legislative authority. Immunities “well grounded in history and reason,” the Court wrote, were not somehow eliminated “by covert inclusion in the general language” of §1983. Id., at 376. This interpretation has been reaffirmed by the Court time and again and is now an entrenched feature of our §1983 jurisprudence. See, e.g., Pierson v. Ray, 386 U.S. 547, 554–555 (1967) (“The legislative record gives no clear indication that Congress meant to abolish wholesale all common-law immunities. Accordingly, this Court held . . . that the immunity of legislators for acts within the legislative role was not abolished. The immunity of judges for acts within the judicial role is equally well established, and we presume that Congress would have specifically so provided had it wished to abolish the doctrine”); Imbler, supra, at 418 (statute must “be read in harmony with general principles of tort immunities and defenses rather than in derogation of them”); Procunier v. Navarette, 434 U.S. 555, 561 (1978) (“Although the Court has recognized that in enacting §1983 Congress must have intended to expose state officials to damages liability in some circumstances, the section has been consistently construed as not intending wholesale revocation of the common-law immunity afforded government officials”); Briscoe v. LaHue, 460 U.S. 325, 330 (1983) (“ ‘It is by now well settled that the tort liability created by §1983 cannot be understood in a historical vacuum. . . . One important assumption underlying the Court’s decisions in this area is that members of the 42d Congress were familiar with common-law principles, including defenses previously recognized in ordinary tort litigation, and that they likely intended these common-law principles to obtain, absent specific provisions to the contrary’ ” (quoting Newport v. Fact Concerts, Inc., 453 U.S. 247, 258 (1981)); Pulliam v. Allen, 466 U.S. 522, 529 (1984) (“The starting point in our own analysis is the common law. Our cases have proceeded on the assumption that common-law principles of . . . immunity were incorporated into our judicial system and that they should not be abrogated absent clear legislative intent to do so”). B Recognizing that “Congress intended [§1983] to be construed in the light of common-law principles,” the Court has looked to the common law for guidance in determining the scope of the immunities available in a §1983 action. Kalina v. Fletcher, 522 U.S. 118, 123 (1997). We do not simply make our own judgment about the need for immunity. We have made it clear that it is not our role “to make a freewheeling policy choice,” Malley v. Briggs, 475 U.S. 335, 342 (1986), and that we do not have a license to create immunities based solely on our view of sound pol- icy, see Tower v. Glover, 467 U.S. 914, 922–923 (1984). Instead, we conduct “a considered inquiry into the immunity historically accorded the relevant official at common law and the interests behind it.” Imbler, supra, at 421. We take what has been termed a “functional approach.” See Forrester v. White, 484 U.S. 219, 224 (1988); Burns, supra, at 486. We consult the common law to identify those governmental functions that were historically viewed as so important and vulnerable to interference by means of litigation that some form of absolute immunity from civil liability was needed to ensure that they are performed “ ‘with independence and without fear of consequences.’ ” Pierson, supra, at 554 (quoting Bradley v. Fisher, 13 Wall. 335, 350, n. ‡ (1872)). Taking this approach, we have identified the following functions that are absolutely immune from liability for damages under §1983: actions taken by legislators within the legitimate scope of legislative authority, see Tenney, supra; actions taken by judges within the legitimate scope of judicial authority, see Pierson, supra; actions taken by prosecutors in their role as advocates, see Imbler, 424 U. S., at 430–431; and the giving of testimony by witnesses at trial, see Briscoe, supra. By contrast, the Court has found no absolute immunity for the acts of the chief executive officer of a State, the senior and subordinate officers of a State’s National Guard, the president of a state university, see Scheuer v. Rhodes, 416 U.S. 232, 247–248 (1974); school board members, see Wood v. Strickland, 420 U.S. 308, 318 (1975); the superintendent of a state hospital, see O’Connor v. Donaldson, 422 U.S. 563, 577 (1975); police officers, see Pierson, supra, at 555; prison officials and officers, Procunier, supra, at 561; and private co-conspirators of a judge, see Dennis v. Sparks, 449 U.S. 24, 27 (1980). C While the Court’s functional approach is tied to the common law’s identification of the functions that merit the protection of absolute immunity, the Court’s precedents have not mechanically duplicated the precise scope of the absolute immunity that the common law provided to pro- tect those functions. See, e.g., Burns, 500 U. S., at 493 (“ ‘[T]he precise contours of official immunity’ need not mirror the immunity at common law” (quoting Anderson v. Creighton, 483 U.S. 635, 645 (1987))). This approach is illustrated by the Court’s analysis of the absolute immunity enjoyed today by public prosecutors. When §1983’s predecessor was enacted in 1871, it was common for criminal cases to be prosecuted by private parties. See, e.g., Stewart v. Sonneborn, 98 U.S. 187, 198 (1879) (Bradley, J., dissenting) (“[E]very man in the community, if he has probable cause for prosecuting another, has a perfect right, by law, to institute such prosecution, subject only, in the case of private prosecutions, to the penalty of paying the costs if he fails in his suit”). And private prosecutors, like private plaintiffs in civil suits, did not enjoy absolute immunity from suit. See Malley, 475 U. S., at 340–341, and n. 3 (citing cases). Instead, “the generally accepted rule” was that a private complainant who procured an arrest or prosecution could be held liable in an action for malicious prosecution if the complainant acted with malice and without probable cause. See id., at 340–341; see also Briscoe, 460 U. S., at 351 (Marshall, J., dissenting) (“Both English and American courts routinely permitted plaintiffs to bring actions alleging that the de- fendant had made a false and malicious accusation of a felony to a magistrate or other judicial officer”); Wheeler v. Nesbitt, 24 How. 544, 550 (1861) (“Undoubtedly, every person who puts the criminal law in force maliciously, and without any reasonable or probable cause, commits a wrongful act; and if the accused is thereby prejudiced, either in his person or property, the injury and loss so sustained constitute the proper foundation of an action to recover compensation”); Dinsman v. Wilkes, 12 How. 390, 402 (1852) (no immunity “where a party had maliciously, and without probable cause, procured the plaintiff to be indicted or arrested for an offence of which he was not guilty”). In the decades after the adoption of the 1871 Civil Rights Act, however, the prosecutorial function was increasingly assumed by public officials, and common-law courts held that public prosecutors, unlike their private predecessors, were absolutely immune from the types of tort claims that an aggrieved or vengeful criminal defendant was most likely to assert, namely, claims for malicious prosecution or defamation. See Imbler, supra, at 441–442 (White, J., concurring in judgment); Kalina, supra, at 124, n. 11 (noting that cases “decided after 1871 . . . granted a broader immunity to public prosecutors than had been available in malicious prosecution actions against private persons who brought prosecutions at early common law”); see also Burns, supra, at 505 (Scalia, J., concurring in judgment in part and dissenting in part) (noting that the “common-law tradition of prosecutorial immunity . . . developed much later than 1871”). This adaptation of prosecutorial immunity accommo- dated the special needs of public, as opposed to private, prosecutors. Because the daily function of a public prosecutor is to bring criminal charges, tort claims against public prosecutors “could be expected with some frequency, for a defendant often will transform his resentment at being prosecuted into the ascription of improper and malicious actions to the State’s advocate.” Imbler, 424 U. S., at 425. Such “harassment by unfounded litigation would cause a deflection of the prosecutor’s energies from his public duties,” and would result in a severe interference with the administration of an important public office. Id., at 423. Constant vulnerability to vexatious litigation would give rise to the “possibility that [the prosecutor] would shade his decisions instead of exercising the independence of judgment required by his public trust.” Ibid. Thus, when the issue of prosecutorial immunity un- der §1983 reached this Court in Imbler, the Court did not simply apply the scope of immunity recognized by common-law courts as of 1871 but instead placed substantial reliance on post-1871 cases extending broad immunity to public prosecutors sued for common-law torts. While the Court has looked to the common law in determining the scope of the absolute immunity available under §1983, the Court has not suggested that §1983 is simply a federalized amalgamation of pre-existing common-law claims, an all-in-one federal claim encompassing the torts of assault, trespass, false arrest, defamation, malicious prosecution, and more. The new federal claim created by §1983 differs in important ways from those pre-existing torts. It is broader in that it reaches constitutional and statutory violations that do not correspond to any previously known tort. See Kalina, 522 U. S., at 123. But it is narrower in that it applies only to tortfeasors who act under color of state law. See Briscoe, supra, at 329. Section 1983 “ha[s] no precise counterpart in state law. . . . [I]t is the purest coincidence when state statutes or the common law provide for equivalent remedies; any analogies to those causes of action are bound to be imperfect.” Wilson v. Garcia, 471 U.S. 261, 272 (1985) (internal quotation marks and citation omitted). Thus, both the scope of the new tort and the scope of the absolute immunity available in §1983 actions differ in some respects from the common law. III A At common law, trial witnesses enjoyed a limited form of absolute immunity for statements made in the course of a judicial proceeding: They had complete immunity against slander and libel claims, even if it was alleged that the statements in question were maliciously false. Kalina, supra, at 133 (Scalia, J., concurring) (citing F. Hilliard, Law of Torts 319 (1866)); see Briscoe, supra, at 351 (Marshall, J., dissenting); Burns, 500 U. S., at 501 (opinion of Scalia, J.). In Briscoe, however, this Court held that the immunity of a trial witness sued under §1983 is broader: In such a case, a trial witness has absolute immunity with respect to any claim based on the witness’ testimony. When a witness is sued because of his testimony, the Court wrote, “ ‘the claims of the individual must yield to the dictates of public policy.’ ” 460 U. S., at 332–333 (quoting Calkins v. Sumner, 13 Wis. 193, 197 (1860)). Without absolute immunity for witnesses, the Court concluded, the truth-seeking process at trial would be impaired. Witnesses “might be reluctant to come forward to testify,” and even if a witness took the stand, the witness “might be inclined to shade his testimony in favor of the potential plaintiff” for “fear of subsequent liability.” 460 U. S., at 333. The factors that justify absolute immunity for trial witnesses apply with equal force to grand jury witnesses. In both contexts, a witness’ fear of retaliatory litigation may deprive the tribunal of critical evidence. And in neither context is the deterrent of potential civil liability needed to prevent perjurious testimony. In Briscoe, the Court concluded that the possibility of civil liability was not needed to deter false testimony at trial because other sanctions—chiefly prosecution for perjury—provided a sufficient deterrent. Id., at 342. Since perjury before a grand jury, like perjury at trial, is a serious criminal offense, see, e.g., 18 U. S. C. §1623(a), there is no reason to think that this deterrent is any less effective in preventing false grand jury testimony. B Neither is there any reason to distinguish law enforcement witnesses from lay witnesses. In Briscoe, it was argued that absolute immunity was not needed for police-officer witnesses, but the Court refused to draw that distinction. The Court wrote: “When a police officer appears as a witness, he may reasonably be viewed as acting like any other witness sworn to tell the truth—in which event he can make a strong claim to witness immunity; alternatively, he may be regarded as an official performing a critical role in the judicial process, in which event he may seek the benefit afforded to other governmental participants in the same proceeding. Nothing in the language of the statute suggests that such a witness belongs in a narrow, special category lacking protection against damages suits.” 460 U. S., at 335–336 (footnote omitted). See also id., at 342 (“A police officer on the witness stand performs the same functions as any other witness”). The Briscoe Court rebuffed two arguments for distinguishing between law enforcement witnesses and lay witnesses for immunity purposes: first, that absolute im- munity is not needed for law enforcement witnesses because they are less likely to be intimidated by the threat of suit and, second, that such witnesses should not be shielded by absolute immunity because false testimony by a police officer is likely to be more damaging than false testimony by a lay witness. See ibid. The Court observed that there are other factors not applicable to lay witnesses that weigh in favor of extending absolute immunity to police officer witnesses. First, police officers testify with some frequency. Id., at 343. “Police officers testify in scores of cases every year,” the Court noted, “and defendants often will transform resentment at being convicted into allegations of perjury by the State’s official witnesses.” Ibid. If police officer witnesses were routinely forced to defend against claims based on their testimony, their “ ‘energy and attention would be diverted from the pressing duty of enforcing the criminal law.’ ” Id., at 343–344 (quoting Imbler, 424 U. S., at 425). Second, a police officer witness’ potential liability, if conditioned on the exoneration of the accused, could influence decisions on appeal and collateral relief. 460 U. S., at 344. Needless to say, such decisions should not be influenced by the likelihood of a subsequent civil rights action. But the possibility that a decision favorable to the accused might subject a police officer witness to liability would create the “ ‘risk of injecting extraneous concerns’ ” into appellate review and postconviction proceedings. Ibid. (quoting Imbler, supra, at 428, n. 27). In addition, law enforcement witnesses face the possibility of sanctions not applicable to lay witnesses, namely, loss of their jobs and other employment-related sanctions. For these reasons, we conclude that grand jury wit- nesses should enjoy the same immunity as witnesses at trial. This means that a grand jury witness has absolute immunity from any §1983 claim based on the witness’ testimony. In addition, as the Court of Appeals held, this rule may not be circumvented by claiming that a grand jury witness conspired to present false testimony or by using evidence of the witness’ testimony to support any other §1983 claim concerning the initiation or maintenance of a prosecution. Were it otherwise, “a criminal defendant turned civil plaintiff could simply reframe a claim to attack the preparation instead of the absolutely immune actions themselves.” Buckley v. Fitzsimmons, 509 U.S. 259, 283 (1993) (Kennedy, J., concurring in part and dissenting in part); see also Dykes v. Hosemann, 776 F.2d 942, 946 (CA11 1985) (per curiam) (“[J]udges, on mere allegations of conspiracy or prior agreement, could be hauled into court and made to defend their judicial acts, the precise result judicial immunity was designed to avoid”). In the vast majority of cases involving a claim against a grand jury witness, the witness and the prose- cutor conducting the investigation engage in preparatory activity, such as a preliminary discussion in which the witness relates the substance of his intended testimony. We decline to endorse a rule of absolute immunity that is so easily frustrated.[1] IV A Petitioner’s main argument is that our cases, chiefly Malley and Kalina, already establish that a “complaining witness” is not shielded by absolute immunity. See Brief for Petitioner 17–22. In those cases, law enforcement officials who submitted affidavits in support of applications for arrest warrants were denied absolute immunity because they “performed the function of a complaining witness.” Kalina, 522 U. S., at 131; see Malley, 475 U. S., at 340–341. Relying on these cases, petitioner contends that certain grand jury witnesses—namely, those who qualify as “complaining witnesses”—are not entitled to absolute immunity. Petitioner’s argument is based on a fundamental misunderstanding of the distinctive function played by a “complaining witness” during the period when §1983’s predecessor was enacted. At that time, the term “complaining witness” was used to refer to a party who procured an arrest and initiated a criminal prosecution, see Kalina, 522 U. S., at 135 (Scalia, J., concurring). A “complaining witness” might not actually ever testify, and thus the term “ ‘witness’ in ‘complaining witness’ is misleading.” Ibid. See also Malley, supra, at 340 (complaining witness “procure[s] the issuance of an arrest warrant by submitting a complaint”); Wyatt v. Cole, 504 U.S. 158, 164–165 (1992) (complaining witness “set[s] the wheels of government in motion by instigating a legal action”). It is true that a mid-19th century complaining witness might testify, either before a grand jury or at trial. But testifying was not a necessary characteristic of a “complaining witness.” See M. Newell, Malicious Prosecution 368 (1892). Nor have we been presented with evidence that witnesses who did no more than testify before a grand jury were regarded as complaining witnesses and were successfully sued for malicious prosecution. See Tr. of Oral Arg. 14–15, 24–25. In sum, testifying, whether before a grand jury or at trial, was not the distinctive function performed by a complaining witness. It is clear—and petitioner does not contend otherwise—that a complaining witness cannot be held liable for perjurious trial testimony. Briscoe, 460 U. S., at 326. And there is no more reason why a complaining witness should be subject to liability for testi- mony before a grand jury. Once the distinctive function performed by a “complaining witness” is understood, it is apparent that a law enforcement officer who testifies before a grand jury is not at all comparable to a “complaining witness.” By testifying before a grand jury, a law enforcement officer does not perform the function of applying for an arrest warrant; nor does such an officer make the critical decision to initiate a prosecution. It is of course true that a detective or case agent who has performed or supervised most of the investigative work in a case may serve as an important witness in the grand jury proceeding and may very much want the grand jury to return an indictment. But such a witness, unlike a complaining witness at common law, does not make the decision to press criminal charges. Instead, it is almost always a prosecutor who is responsible for the decision to present a case to a grand jury, and in many jurisdictions, even if an indictment is handed up, a prosecution cannot proceed unless the prosecutor signs the indictment.[2] It would thus be anomalous to permit a police officer who testifies before a grand jury to be sued for maliciously procuring an unjust prosecution when it is the prosecutor, who is shielded by absolute immunity, who is actually responsible for the decision to prosecute. See Albright v. Oliver, 510 U.S. 266, 279, n. 5 (1994) (Ginsburg, J., concurring) (the prosecutor is the “principal player in carrying out a prosecution”); see ibid. (“[T]he star player is exonerated, but the supporting actor is not”).[3] Precisely because no grand jury witness has the power to initiate a prosecution, petitioner is unable to provide a workable standard for determining whether a particular grand jury witness is a “complaining witness.” Here, respondent was the only witness to testify in two of the three grand jury sessions that resulted in indictments. But where multiple witnesses testify before a grand jury, identifying the “complaining witness” would often be difficult. Petitioner suggests that a “complaining witness” is “someone who sets the prosecution in motion.” Tr. of Oral Arg. 8; see Reply Brief for Petitioner 15. And petitioner maintains that the same distinction made at common law between complaining witnesses and other witnesses applies in §1983 actions. See id., at 14–16. But, as we have explained, a complaining witness played a dis- tinctive role, and therefore even when a “complaining witness” testified, there was a clear basis for distinguishing between the “complaining witness” and other wit- nesses. Because no modern grand jury witness plays a comparable role, petitioner’s proposed test would be of little use. Consider a case in which the case agent or lead detective testifies before the grand jury and provides a wealth of background information and then a cooperating witness appears and furnishes critical incriminating testimony. Or suppose that two witnesses each provide essential testimony regarding different counts of an indictment or different elements of an offense. In these cases, which witnesses would be “complaining witnesses” and thus vulnerable to suit based on their testimony? B Petitioner contends that the deterrent effect of civil liability is more needed in the grand jury context because trial witnesses are exposed to cross-examination, which is designed to expose perjury. See Brief for Petitioner 21, 25–26. This argument overlooks the fact that a critical grand jury witness is likely to testify again at trial and may be cross-examined at that time. But in any event, the force of petitioner’s argument is more than offset by a special problem that would be created by allowing civil actions against grand jury witnesses—subversion of grand jury secrecy. “ ‘We consistently have recognized that the proper functioning of our grand jury system depends upon the secrecy of grand jury proceedings.’ ” United States v. Sells Engineering, Inc., 463 U.S. 418, 424 (1983) (quoting Douglas Oil Co. v. Petrol Stops Northwest, 441 U.S. 211, 218–219 (1979)). “ ‘[I]f preindictment proceedings were made public, many prospective witnesses would be hesitant to come forward voluntarily, knowing that those against whom they testify would be aware of that testimony. Moreover, witnesses who appeared before the grand jury would be less likely to testify fully and frankly, as they would be open to retribution.’ ” 463 U. S., at 424. Allowing §1983 actions against grand jury witnesses would compromise this vital secrecy. If the testimony of witnesses before a grand jury could provide the basis for, or could be used as evidence supporting, a §1983 claim, the identities of grand jury witnesses could be discovered by filing a §1983 action and moving for the disclosure of the transcript of grand jury proceedings. Especially in cases involving violent criminal organizations or other subjects who might retaliate against adverse grand jury witnesses, the threat of such disclosure might seriously undermine the grand jury process. C Finally, contrary to petitioner’s suggestion, recognizing absolute immunity for grand jury witnesses does not create an insupportable distinction between States that use grand juries and those that do not. Petitioner argues that it would make no sense to distinguish for purposes of §1983 immunity between prosecutions initiated by the return of a grand jury indictment and those initiated by the filing of a complaint or information, and he notes that 26 States permit felony prosecutions to be brought by information. Brief for Petitioner 23–24. But petitioner draws the wrong analogy. In States that permit felony prosecutions to be initiated by information, the closest analog to a grand jury witness is a witness at a preliminary hearing. Most of the States that do not require an indictment for felonies provide a preliminary hearing at which witnesses testify. See LaFave §14.2(d), at 304, and n. 47, 307, and n. 60. The lower courts have held that witnesses at a preliminary hearing are protected by the same immunity accorded grand jury witnesses, see, e.g., Brice v. Nkaru, 220 F.3d 233, 239, n. 6 (CA4 2000); Curtis v. Bembenek, 48 F.3d 281, 284–285 (CA7 1995) (citing cases), and petitioner does not argue otherwise, see Tr. of Oral Arg. 51. * * * For these reasons, we hold that a grand jury witness is entitled to the same immunity as a trial witness. Accordingly, the judgment of the Court of Appeals for the Eleventh Circuit is Affirmed. Notes 1 Of course, we do not suggest that absolute immunity extends to all activity that a witness conducts outside of the grand jury room. For example, we have accorded only qualified immunity to law enforcement officials who falsify affidavits, see Kalina v. Fletcher, 522 U.S. 118, 129–131 (1997); Malley v. Briggs, 475 U.S. 335, 340–345 (1986), and fabricate evidence concerning an unsolved crime, see Buckley, 509 U. S., at 272–276. 2 The federal courts have concluded uniformly that Rule 7(c) of the Federal Rules of Criminal Procedure, providing that an indictment “must be signed by an attorney for the government,” precludes federal grand juries from issuing an indictment without the prosecutor’s signature, signifying his or her approval. See 4 W. LaFave, J. Israel,N. King, & O. Kerr, Criminal Procedure §15.1(d) (3d ed. 2007) (hereinafter LaFave). However, in some jurisdictions, the grand jury may return an indictment and initiate a prosecution without the prosecutor’s signature, but such cases are rare. See 1 S. Beale, W. Bryson, J. Felman, & M. Elston, Grand Jury Law and Practice, p. 4–76, and n. 2 (2d ed. 2001). 3 Petitioner says there is no reason to distinguish between a person who goes to the police to swear out a criminal complaint and a person who testifies to facts before a grand jury for the same purpose and with the same effect. Brief for Petitioner 2, 23. But this is like saying thata bicycle and an F-16 are the same thing. Even if the functions are similar as a general matter, the entities are quite different. Grand juries, by tradition, statute, and sometimes constitutional mandate, have a status and entitlement to information that absolute immunity furthers. See, e.g., Imbler v. Pachtman, 424 U.S. 409, 423, n. 20 (1976) (“It is the functional comparability of their judgments to those ofthe judge that has resulted in both grand jurors and prosecutors be-ing referred to as ‘quasi-judicial’ officers, and their immunities being termed ‘quasi-judicial’ as well”); see also United States v. Sells Engineering, Inc., 463 U.S. 418, 423 (1983) (“The grand jury has always occupied a high place as an instrument of justice in our system of criminal law—so much so that it is enshrined in the Constitution”). Our holding today supports the functioning of the grand jury system. The importance of the grand jury cannot be underestimated: In the federal system and many States, see LaFave §15.1(d), a felony cannot be charged without the consent of community representatives, a vital protection from unwarranted prosecutions.

566.US.2011_11-262

Petitioners Reichle and Doyle were members of a Secret Service detail protecting Vice President Richard Cheney while he greeted members of the public at a shopping mall. Agent Doyle overheard respondent Howards, who was speaking into his cell phone, state that he “was going to ask [the Vice President] how many kids he’s killed today.” Doyle and other agents observed Howards enter the line to meet the Vice President, tell the Vice President that his “policies in Iraq are disgusting,” and touch the Vice President’s shoulder as the Vice President was leaving. After being briefed by Doyle, Agent Reichle interviewed and then arrested Howards, who was charged with harassment. After that charge was dismissed, Howards brought an action against petitioners and others under 42 U. S. C. §1983 and Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388. Howards claimed that he was arrested and searched without probable cause, in violation of the Fourth Amendment, and that the arrest violated the First Amendment because it was made in retaliation for Howards’ criticism of the Vice President. Petitioners moved for summary judgment on the ground that they were entitled to qualified immunity, but the Federal District Court denied the motion. On appeal, the Tenth Circuit reversed the immunity ruling with respect to the Fourth Amendment claim because petitioners had probable cause to arrest Howards, but the court affirmed with regard to the First Amendment claim. In doing so, the court rejected petitioners’ argument that, under Hartman v. Moore, 547 U.S. 250, probable cause to arrest defeats a First Amendment retaliatory arrest claim. It concluded instead that Hartman applied only to retaliatory prosecution claims and thus did not upset prior Tenth Circuit precedent holding that a retaliatory arrest violates the First Amendment even if supported by probable cause. Held: Petitioners are entitled to qualified immunity because, at the time of Howards’ arrest, it was not clearly established that an arrest supported by probable cause could give rise to a First Amendment violation. Pp. 5−12. (a) Courts may grant qualified immunity on the ground that a purported right was not “clearly established” by prior case law. Pearson v. Callahan, 555 U.S. 223, 236. To be clearly established, a right must be sufficiently clear “that every ‘reasonable official would [have understood] that what he is doing violates that right.’ ” Ashcroft v. al-Kidd, 563 U. S. ___, ___. Pp. 5−6. (b) The “clearly established” standard is not satisfied here. This Court has never recognized a First Amendment right to be free from a retaliatory arrest that is supported by probable cause; nor was such a right otherwise clearly established at the time of Howards’ arrest. P. 6. (c) At that time, Hartman’s impact on the Tenth Circuit’s precedent was far from clear. Although Hartman’s facts involved only a retaliatory prosecution, reasonable law enforcement officers could have questioned whether its rule also applied to arrests. First, Hartman was decided against a legal backdrop that treated retaliatory arrest claims and retaliatory prosecution claims similarly. It resolved a Circuit split concerning the impact of probable cause on retaliatory prosecution claims, but some of the conflicting cases involved both retaliatory prosecution and retaliatory arrest claims and made no distinction between the two when considering the relevance of probable cause. Second, a reasonable official could have interpreted Hartman’s rationale to apply to retaliatory arrests. Like in retaliatory prosecution cases, evidence of the presence or absence of probable cause for the arrest will be available in virtually all retaliatory arrest cases, and the causal link between the defendant’s alleged retaliatory animus and the plaintiff’s injury may be tenuous. Finally, decisions from other Circuits in the wake of Hartman support the conclusion that, for qualified immunity purposes, it was at least arguable at the time of Howards’ arrest that Hartman extended to retaliatory arrests. Pp. 7−12. 634 F.3d 1131, reversed and remanded. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Alito, and Sotomayor, JJ., joined. Ginsburg, J., filed an opinion concurring in the judgment, in which Breyer, J., joined. Kagan, J., took no part in the consideration or decision of the case.

This case requires us to decide whether two federal law enforcement agents are immune from suit for allegedly ar- resting a suspect in retaliation for his political speech, when the agents had probable cause to arrest the suspect for committing a federal crime. I On June 16, 2006, Vice President Richard Cheney vis- ited a shopping mall in Beaver Creek, Colorado. A Secret Service protective detail accompanied the Vice President. Petitioners Gus Reichle and Dan Doyle were members of that detail. Respondent Steven Howards was also at the mall. He was engaged in a cell phone conversation when he noticed the Vice President greeting members of the public. Agent Doyle overheard Howards say, during this conversation, “ ‘I’m going to ask [the Vice President] how many kids he’s killed today.’ ” Brief for Petitioners 4. Agent Doyle told two other agents what he had heard, and the three of them began monitoring Howards more closely. Agent Doyle watched Howards enter the line to meet the Vice President. When Howards approached the Vice President, he told him that his “ ‘policies in Iraq are disgusting.’ ” Ibid. The Vice President simply thanked Howards and moved along, but Howards touched the Vice President’s shoulder as the Vice President departed.[1] Howards then walked away. Several agents observed Howards’ encounter with the Vice President. The agents determined that Agent Reichle, who coordinated the protective intelligence team respon- sible for interviewing individuals suspected of violat- ing the law, should question Howards. Agent Reichle had not personally heard Howards’ comments or seen his con- tact with the Vice President, but Agent Doyle briefed Agent Reichle on what had happened. Agent Reichle approached Howards, presented his badge and identified himself, and asked to speak with him. Howards refused and attempted to walk away. At that point, Agent Reichle stepped in front of Howards and asked if he had assaulted the Vice President. Pointing his finger at Agent Reichle, Howards denied assaulting the Vice President and told Agent Reichle, “if you don’t want other people sharing their opinions, you should have him [the Vice President] avoid public places.” Howards v. McLaughlin, 634 F.3d 1131, 1137 (CA10 2011) (internal quotation marks omitted). During this exchange, Agent Reichle also asked Howards whether he had touched the Vice President. Howards falsely denied doing so. After confirming that Agent Doyle had indeed seen Howards touch the Vice President, Reichle arrested Howards. The Secret Service transferred Howards to the custody of the local sheriff’s department. Howards was charged by local officials with harassment in violation of state law. The charge was eventually dismissed. II Howards brought this action in the United States District Court for the District of Colorado under Rev. Stat. §1979, 42 U. S. C. §1983, and Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388 (1971).[2] Howards alleged that he was arrested and searched without probable cause, in violation of the Fourth Amendment. Howards also al- leged that he was arrested in retaliation for criticizing the Vice President, in violation of the First Amendment. Petitioners Reichle and Doyle moved for summary judgment on the ground that they were entitled to qualified immunity. The District Court denied the motion. See App. to Pet. for Cert. 46–61. On interlocutory appeal, a divided panel of the United States Court of Appeals for the Tenth Circuit affirmed in part and reversed in part. 634 F.3d 1131. The Court of Appeals held that petitioners enjoyed qualified immunity with respect to Howards’ Fourth Amendment claim. The court concluded that petitioners had probable cause to arrest Howards for making a materially false statement to a federal official in violation of 18 U. S. C. §1001 because he falsely denied touching the Vice President. 634 F. 3d, at 1142. Thus, the court concluded that neither Howards’ arrest nor search incident to the arrest violated the Fourth Amendment.[3] Id., at 1142–1143. However, the Court of Appeals denied petitioners qualified immunity from Howards’ First Amendment claim. The court first determined that Howards had established a material factual dispute regarding whether petitioners were substantially motivated by Howards’ speech when they arrested him. Id., at 1144–1145. The court then rejected petitioners’ argument that, under this Court’s decision in Hartman v. Moore, 547 U.S. 250 (2006), probable cause to arrest defeats a First Amendment claim of retaliatory arrest. The court concluded that Hartman established such a rule only for retaliatory prosecution claims and, therefore, did not upset prior Tenth Circuit precedent clearly establishing that a retaliatory arrest violates the First Amendment even if supported by probable cause. 634 F. 3d, at 1148. Judge Paul Kelly dissented from the court’s denial of qualified immunity. He would have held that when Howards was arrested, it was not clearly established that an arrest supported by probable cause could violate the First Amendment. In Judge Kelly’s view, Hartman called into serious question the Tenth Circuit’s prior precedent on retaliatory arrests. 634 F. 3d, at 1151. He noted that other Circuits had applied Hartman to retaliatory arrests and that there was a “strong argument” in favor of doing so. 634 F. 3d, at 1151–1152. We granted certiorari on two questions: whether a First Amendment retaliatory arrest claim may lie despite the presence of probable cause to support the arrest, and whether clearly established law at the time of Howards’ arrest so held. See 565 U. S. ___ (2011). If the answer to either question is “no,” then the agents are entitled to qualified immunity. We elect to address only the second question. We conclude that, at the time of Howards’ arrest, it was not clearly established that an arrest supported by probable cause could violate the First Amendment. We, therefore, reverse the judgment of the Court of Appeals denying petitioners qualified immunity.[4] III Qualified immunity shields government officials from civil damages liability unless the official violated a statutory or constitutional right that was clearly established at the time of the challenged conduct. See Ashcroft v. al-Kidd, 563 U. S. ___, ___ (2011) (slip op., at 3). In Pearson v. Callahan, 555 U.S. 223, 236 (2009), we held that courts may grant qualified immunity on the ground that a purported right was not “clearly established” by prior case law, without resolving the often more difficult question whether the purported right exists at all. Id., at 227. This approach comports with our usual reluctance to decide constitutional questions unnecessarily. Id., at 241; see also Camreta v. Greene, 563 U. S. ___, ___ – ___ (2011) (slip op., at 9–10); al-Kidd, 563 U. S., at ___ (slip op., at 3). To be clearly established, a right must be sufficiently clear “that every ‘reasonable official would [have understood] that what he is doing violates that right.’ ” Id., at ___ (slip op., at 9) (quoting Anderson v. Creighton, 483 U.S. 635, 640 (1987)). In other words, “existing precedent must have placed the statutory or constitutional question beyond debate.” 563 U. S., at ___ (slip op., at 9). This “clearly established” standard protects the balance between vindication of constitutional rights and government officials’ effective performance of their duties by ensuring that officials can “ ‘reasonably . . . anticipate when their conduct may give rise to liability for damages.’ ” Anderson, supra, at 639 (quoting Davis v. Scherer, 468 U.S. 183, 195 (1984)). The “clearly established” standard is not satisfied here. This Court has never recognized a First Amendment right to be free from a retaliatory arrest that is supported by probable cause; nor was such a right otherwise clearly established at the time of Howards’ arrest. A Howards contends that our cases have “settled” the rule that, “ ‘as a general matter[,] the First Amendment prohibits government officials from subjecting an individual to retaliatory actions’ ” for his speech. See Brief for Respondent 39 (quoting Hartman, supra, at 256). But we have previously explained that the right allegedly violated must be established, “ ‘not as a broad general proposition,’ ” Brosseau v. Haugen, 543 U.S. 194, 198 (2004) (per curiam), but in a “particularized” sense so that the “contours” of the right are clear to a reasonable official, Anderson, supra, at 640. Here, the right in question is not the general right to be free from retaliation for one’s speech, but the more specific right to be free from a retaliatory arrest that is otherwise supported by probable cause. This Court has never held that there is such a right.[5] B We next consider Tenth Circuit precedent. Assuming ar- guendo that controlling Court of Appeals’ authority could be a dispositive source of clearly established law in the circumstances of this case, the Tenth Circuit’s cases do not satisfy the “clearly established” standard here. Relying on DeLoach v. Bevers, 922 F.2d 618 (1990), and Poole v. County of Otero, 271 F.3d 955 (2001), the Court of Appeals concluded that, at the time of Howards’ arrest, its precedent had clearly established the unlawfulness of an arrest in retaliation for the exercise of First Amendment rights, irrespective of probable cause. In DeLoach, a case involving both a retaliatory arrest and a retaliatory prosecution, the court held that “[a]n act taken in retaliation for the exercise of a constitutionally protected right is actionable under §1983 even if the act, when taken for a different reason, would have been proper.” 922 F. 2d, at 620 (internal quotation marks omitted). In Poole, a sub- sequent retaliatory prosecution case, the court relied on DeLoach for the proposition that a plaintiff’s illegal conduct is “not relevant to his First Amendment claim.” 271 F. 3d, at 961. The Court of Appeals acknowledged that Poole was abrogated by this Court’s subsequent decision in Hartman v. Moore, 547 U.S. 250, which held that a plaintiff cannot state a claim of retaliatory prosecution in violation of the First Amendment if the charges were supported by probable cause. But the Court of Appeals determined that Hartman’s no-probable-cause requirement did not extend to claims of retaliatory arrest and therefore did not disturb its prior precedent in DeLoach. Accordingly, the court concluded, “when Mr. Howards was arrested it was clearly established that an arrest made in retaliation of an individual’s First Amendment rights is unlawful, even if the arrest is supported by probable cause.” 634 F. 3d, at 1148. We disagree. At the time of Howards’ arrest, Hartman’s impact on the Tenth Circuit’s precedent governing retal- iatory arrests was far from clear. Although the facts of Hartman involved only a retaliatory prosecution, reason- able officers could have questioned whether the rule of Hartman also applied to arrests. Hartman was decided against a legal backdrop that treated retaliatory arrest and prosecution claims similarly. Hartman resolved a split among the Courts of Appeals about the relevance of probable cause in retaliatory prosecution suits, but some of the conflicting court of appeals cases involved both an arrest and a prosecution that were alleged to be retaliation for the exercise of First Amendment rights. See 547 U. S., at 255–256, 259, n. 6 (citing Mozzochi v. Borden, 959 F.2d 1174 (CA2 1992); Singer v. Fulton Cty. Sheriff, 63 F.3d 110 (CA2 1995); Keenan v. Tejeda, 290 F.3d 252 (CA5 2002); Wood v. Kesler, 323 F.3d 872 (CA11 2003)). Those cases made no distinction between claims of retaliatory arrest and claims of retaliatory prosecution when considering the relevance of prob- able cause. See Mozzochi, supra, at 1179–1180; Singer, supra, at 120; Keenan, supra, at 260; Wood, supra, at 883. Indeed, the close relationship between retaliatory arrest and prosecution claims is well demonstrated by the Tenth Circuit’s own decision in DeLoach. DeLoach, too, involved allegations of both retaliatory arrest and retaliatory prosecution, and the Tenth Circuit analyzed the two claims as one. 922 F. 2d, at 620–621. A reasonable official also could have interpreted Hartman’s rationale to apply to retaliatory arrests. Hartman first observed that, in retaliatory prosecution cases, evidence showing whether there was probable cause for the charges would always be “available and apt to prove or disprove retaliatory causation.” 547 U. S., at 261. In this Court’s view, the presence of probable cause, while not a “guarantee” that retaliatory motive did not cause the prosecution, still precluded any prima facie inference that retaliatory motive was the but-for cause of the plaintiff’s injury. Id., at 265. This was especially true because, as Hartman next emphasized, retaliatory prosecution claims involve particularly attenuated causation between the de- fendant’s alleged retaliatory animus and the plaintiff’s injury. Id., at 259–261. In a retaliatory prosecution case, the key defendant is typically not the prosecutor who made the charging decision that injured the plaintiff, because prosecutors enjoy absolute immunity for their decisions to prosecute. Rather, the key defendant is the person who allegedly prompted the prosecutor’s decision. Thus, the intervening decision of the third-party prosecutor widens the causal gap between the defendant’s animus and the plaintiff’s injury. Id., at 261–263. Like retaliatory prosecution cases, evidence of the presence or absence of probable cause for the arrest will be available in virtually every retaliatory arrest case. Such evidence could be thought similarly fatal to a plaintiff’s claim that animus caused his arrest, given that retaliatory arrest cases also present a tenuous causal connection between the defendant’s alleged animus and the plaintiff’s injury. An officer might bear animus toward the content of a suspect’s speech. But the officer may decide to arrest the suspect because his speech provides evidence of a crime or suggests a potential threat. See, e.g., Wayte v. United States, 470 U.S. 598, 612–613 (1985) (noting that letters of protest written to the Selective Service, in which the author expressed disagreement with the draft, “provided strong, perhaps conclusive evidence” of the nonregistrant’s intent not to comply—one of the elements of the offense” of willful failure to register for the draft). Like retaliatory prosecution cases, then, the connection between alleged animus and injury may be weakened in the arrest context by a police officer’s wholly legitimate consideration of speech. To be sure, we do not suggest that Hartman’s rule in fact extends to arrests. Nor do we suggest that every as- pect of Hartman’s rationale could apply to retaliatory arrests. Hartman concluded that the causal connection in retaliatory prosecution cases is attenuated because those cases necessarily involve the animus of one person and the injurious action of another, 547 U. S., at 262, but in many retaliatory arrest cases, it is the officer bearing the al- leged animus who makes the injurious arrest. Moreover, Hartman noted that, in retaliatory prosecution cases, the causal connection between the defendant’s animus and the prosecutor’s decision is further weakened by the “presumption of regularity accorded to prosecutorial decisionmaking.” Id., at 263. That presumption does not apply here. Nonetheless, the fact remains that, for qualified immunity purposes, at the time of Howards’ arrest it was at least arguable that Hartman’s rule extended to retaliatory arrests.[6] Decisions from other Federal Courts of Appeals in the wake of Hartman support this assessment. Shortly before Howards’ arrest, the Sixth Circuit held that Hartman required a plaintiff alleging a retaliatory arrest to show that the defendant officer lacked probable cause. See Barnes v. Wright, 449 F.3d 709, 720 (2006) (reasoning that the Hartman “rule sweeps broadly”). That court’s treatment of Hartman confirms that the inapplicability of Hartman to arrests would not have been clear to a reasonable officer when Howards was arrested. Moreover, since Howards’ arrest, additional Courts of Appeals have concluded that Hartman’s no-probable-cause requirement extends to retaliatory arrests. See, e.g., McCabe v. Parker, 608 F.3d 1068, 1075 (CA8 2010); Phillips v. Irvin, 222 Fed. Appx. 928, 929 (CA11 2007) (per curiam). As we have previously observed, “[i]f judges thus disagree on a constitutional question, it is unfair to subject police to money damages for picking the losing side of the controversy.” Wilson v. Layne, 526 U.S. 603, 618 (1999).[7] * * * Hartman injected uncertainty into the law governing retaliatory arrests, particularly in light of Hartman’s rationale and the close relationship between retaliatory arrest and prosecution claims. This uncertainty was only confirmed by subsequent appellate decisions that disagreed over whether the reasoning in Hartman applied similarly to retaliatory arrests. Accordingly, when Howards was arrested it was not clearly established that an arrest supported by probable cause could give rise to a First Amendment violation. Petitioners Reichle and Doyle are thus entitled to qualified immunity. The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Notes 1 The parties dispute the manner of the touch. Howards described it as an open-handed pat, while several Secret Service agents described it as a forceful push. This dispute does not affect our analysis. 2 Howards named several Secret Service agents as defendants, but only Agents Reichle and Doyle are petitioners here. We address only those parts of the lower courts’ decisions that involve petitioners Reichle and Doyle. 3 Howards does not challenge the Court of Appeals’ probable-cause determination. 4 This Court has recognized an implied cause of action for damages against federal officials for Fourth Amendment violations. See Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388 (1971). We have never held that Bivens extends to First Amendment claims. See Ashcroft v. Iqbal, 556 U.S. 662, 675 (2009) (assuming without deciding that a First Amendment free exercise claim is actionable under Bivens); Bush v. Lucas, 462 U.S. 367, 368 (1983) (refusing to extend Bivens to a First Amendment speech claim involving federal employment). We need not (and do not) decide here whether Bivens extends to First Amendment retaliatory arrest claims. 5 The Court of Appeals’ reliance on Whren v. United States, 517 U.S. 806 (1996), was misplaced. There, we held that a traffic stop supported by probable cause did not violate the Fourth Amendment regardlessof the officer’s actual motivations, but we explained that the Equal Protection Clause would prohibit an officer from selectively enforcing the traffic laws based on race. Id., at 813. Citing Whren, the Court of Appeals noted that “[i]t is well established that an act which is lawful under the Fourth Amendment may still violate other provisions of the Constitution.” Howards v. McLaughlin, 634 F.3d 1131, 1149, n. 15 (CA10 2011). But, again, we do not define clearly established law at such a “high level of generality.” Ashcroft v. al-Kidd, 563 U. S. ___, ___ (2011) (slip op., at 10). Whren’s discussion of the Fourteenth Amendment does not indicate, much less “clearly establish,” that an arrest supported by probable cause could nonetheless violate the First Amendment. 6 Howards argues that petitioners violated his clearly established First Amendment right even if Hartman’s rule applies equally to retaliatory arrests. According to Howards, Hartman did not hold that a prosecution violates the First Amendment only when it is unsupported by probable cause. Rather, Howards argues, Hartman made probable cause relevant only to a plaintiff’s ability to recover damages for a First Amendment violation. See Brief for Respondent 37–41. We need not resolve whether Hartman is best read as defining the scope of the First Amendment right or as simply establishing a prerequisite for recovery. Nor need we decide whether that distinction matters. It suffices, for qualified immunity purposes, that the answer would not have been clear to a reasonable official when Howards was arrested. 7 Indeed, the Tenth Circuit itself has applied Hartman outside the context of retaliatory prosecution. See McBeth v. Himes, 598 F.3d 708, 719 (2010) (requiring the absence of probable cause in the context of a claim alleging that government officials suspended a business license in retaliation for the exercise of First Amendment rights).

565.US.2011_10-6549

The federal Sex Offender Registration and Notification Act (Act) requires convicted sex offenders to provide state governments with, and to update, information, e.g., names and current addresses, for state and federal sex offender registries. It is a crime if a person who is “required to register under [the Act]” and who “travels in interstate . . . commerce” knowingly “fails to register or update a registration.” 18 U. S. C. §2250(a). The Act defines “sex offender” to include offenders who were convicted before the Act’s effective date, 42 U. S. C. §16911(1), and says that “the Attorney General shall have the authority to specify the applicability of the [registration] requirements” to pre-Act offenders, §16913(d). The Act, which seeks to make more uniform and effective a patchwork of pre-Act federal and 50 state registration systems, became law in July 2006. In February 2007, the Attorney General promulgated an Interim Rule specifying that the Act applies to all pre-Act offenders. He has since promulgated further rules, regulations, and specifications. Petitioner Reynolds, a pre-Act offender, registered in Missouri in 2005 but moved to Pennsylvania in September 2007 without updating the Missouri registration or registering in Pennsylvania. He was indicted for failing to meet the Act’s registration requirements between September 16 and October 16, 2007. He moved to dismiss the indictment on the ground that the Act was not applicable to pre-Act offenders during that time, arguing that the Attorney General’s February 2007 Interim Rule was invalid because it violated the Constitution’s “nondelegation” doctrine and the Administrative Procedure Act’s notice and comment requirements. The District Court rejected on the merits of Reynolds’ legal attack on the Interim Rule, but the Third Circuit rejected his argument without reaching the merits, concluding that the Act’s registration requirements applied to pre-Act offenders even in the absence of a rule by the Attorney General. Thus, it found, the Interim Rule’s validity made no legal difference in the outcome. Held: The Act does not require pre-Act offenders to register before the Attorney General validly specifies that the Act’s registration provisions apply to them. Pp. 6–13. (a) This conclusion is supported by a natural reading of the Act’s text, which consists of four statements. Statement One says that “[a] sex offender shall register, and keep the registration current.” Statement Two says that, generally, the offender must initially register before completing his “sentence of imprisonment.” Statement Three says that the sex offender must update a registration within three business days of any change of “name, residence, employment, or student status.” Statement Four says that “[t]he Attorney General shall have the authority to specify the applicability of the requirements . . . to sex offenders convicted before the enactment of” the Act. §16913. Read naturally, the Fourth Statement modifies the First. It deals specifically with a subset (pre-Act offenders) of the First Statement’s broad general class (all sex offenders) and thus should control the Act’s application to that subset. See Gozlon-Peretz v. United States, 498 U.S. 395, 407. Also, by giving the Attorney General authority to specify the Act’s “applicability,” not its “nonapplicability,” the Fourth Statement is more naturally read to confer authority to apply the Act, not authority to make exceptions. This reading efficiently resolves what may have been Congress’ concern about the practical problems of applying the new registration requirements to a large number of pre-Act offenders, which could have been expensive and might not have proved feasible to do immediately. It might have thought that such concerns warranted different treatment for different categories of pre-Act offenders. And it could have concluded that it was efficient and desirable to ask the Justice Department, charged with responsibility for implementation, to examine pre-Act offender problems and to apply the new requirements accordingly. This reading also takes Congress to have filled potential lacunae (created by related Act provisions) in a manner consistent with basic criminal law principles. The Second Statement, e.g., requires a sex offender to register before completing his prison term, but says nothing about when a pre-Act offender who has left prison is to register. An Attorney General ruling could diminish such uncertainties, helping to eliminate the kind of vagueness and uncertainty that criminal law must seek to avoid. Pp. 6–9. (b) The Government’s three principal contrary arguments—that the Court’s reading conflicts with the Act’s purpose of establishing a national registration system that includes pre-Act offenders; that the Court’s reading could lead to an absurdly long implementation delay; and that the Act should be read to apply the requirements immediately and on their own to all pre-Act offenders to avoid the possibility that the Attorney General, who has, but is not required to use, “the authority to specify” requirements, might take no action—are unpersuasive. Some lower courts have read the Attorney General’s authority to apply only to pre-Act sex offenders who are unable to comply with the statute’s “initial registration” requirements, but that is not what the Act says. Pp. 9–13. (c) Because the Act’s registration requirements do not apply to pre-Act offenders until the Attorney General so specifies, the question whether the Attorney General’s Interim Rule is a valid specification matters in this case. P. 13. 380 Fed. Appx. 125, reversed and remanded. Breyer, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, Alito, Sotomayor, and Kagan, JJ., joined. Scalia, J., filed a dissenting opinion, in which Ginsburg, J., joined.

The federal Sex Offender Registration and Notification Act (Act), 120Stat. 590, 42 U. S. C. §16901 et seq. (2006 ed. and Supp. III), requires those convicted of certain sex crimes to provide state governments with (and to update) information, such as names and current addresses, for inclusion on state and federal sex offender registries. §§16912(a), 16913–16914, 16919(a) (2006 ed.). The Act makes it a crime for a person who is “required to regis- ter” under the Act and who “travels in interstate or foreign commerce” knowingly to “fai[l] to register or update a reg- istration . . . .” 18 U. S. C. §2250(a). The question be- fore us concerns the date on which this federal registra- tion requirement took effect with respect to sex offenders convicted before the Act became law. The Act defines the term “sex offender” as including these pre-Act offenders. 42 U. S. C. §16911(1); see Carr v. United States, 560 U. S. ___, ___ (2010) (slip op., at 7). It says that “[a] sex offender shall register.” §16913(a). And it further says that “[t]he Attorney General shall have the authority to specify the applicability of the [registration] requirements . . . to sex offenders convicted before the enactment of this chapter . . . .” §16913(d) (emphasis added). In our view, these provisions, read together, mean that the Act’s registration requirements do not apply to pre-Act offenders until the Attorney General specifies that they do apply. We reverse a Court of Appeals determination that, in effect, holds the contrary. I A The new federal Act reflects Congress’ awareness that pre-Act registration law consisted of a patchwork of fed- eral and 50 individual state registration systems. See 73 Fed. Reg. 38045 (2008). The Act seeks to make those systems more uniform and effective. It does so by repealing several earlier federal laws that also (but less effectively) sought uniformity; by setting forth comprehensive registration-system standards; by making federal funding contingent on States’ bringing their systems into compliance with those standards; by requiring both state and federal sex offenders to register with relevant jurisdictions (and to keep registration information current); and by creating federal criminal sanctions applicable to those who violate the Act’s registration requirements. 18 U. S. C. §2250(a) (criminal provision); 42 U. S. C. §§16911(10), 16913–16916 (2006 ed. and Supp. III) (registration requirements); §16925 (federal funding); §129, 120Stat. 600 (repeal of earlier laws). The Act’s criminal penalty applies to “[w]ho[m]ever . . . is required to register under [the Act].” 18 U. S. C. §2250(a). It says that such a person (a federal sex offender or a nonfederal sex offender who travels in interstate commerce) must not knowingly fail “to register or update a registration as required by [the Act].” Ibid. (emphasis added); see Appendix, infra, at 14. The relevant registration requirements are set forth in an Act provision that states: “Registry requirements for sex offenders “(a) In general “A sex offender [defined to include any offender who was convicted of a sex offense] shall register, and keep the registration current, in each jurisdiction where the offender resides, where the offender is an employee, and where the offender is a student. . . . “(b) Initial registration “The sex offender shall initially register [either] before completing a sentence of imprisonment with respect to the offense giving rise to the registration re- quirement; or [for those not sentenced to prison] not later than 3 business days after being sentenced . . . . “(c) Keeping the registration current “A sex offender shall [update his registration within] 3 business days after each change of name, residence, employment, or student status [by] appear[ing] in person in at least 1 jurisdiction involved . . . and inform[ing] that jurisdiction of all [relevant] changes . . . . “(d) Initial registration of sex offenders unable to comply with subsection (b) “The Attorney General shall have the authority to specify the applicability of the [registration] requirements . . . to sex offenders convicted before the en- actment of this chapter or its implementation in a particular jurisdiction, and to prescribe rules for the registration of any such sex offenders and for other categories of sex offenders who are unable to comply with subsection (b).” 42 U. S. C. §16913 (emphasis added). The new Act became law on July 27, 2006. On February 28, 2007, the Attorney General promulgated an Interim Rule specifying that “[t]he requirements of [the Act] apply to all sex offenders, including sex offenders convicted of the offense for which registration is required prior to the enactment of that Act.” 72 Fed. Reg. 8897 (codified at 28 CFR §72.3). Subsequently, the Attorney General promulgated further rules, regulations, and specifications. See 73 Fed. Reg. 38030 (2008); 75 Fed. Reg. 81849 (2010); 76 Fed. Reg. 1630 (2011). The present case focuses upon the applicability of the Act’s registration requirements to pre-Act offenders during the period between (1) July 27, 2006 (when the Act took effect) and (2) the moment when the Attorney General promulgated a valid rule specifying the registration requirements’ ap- plicability, namely, February 28, 2007 (or a later date if the February 28 specification was invalid). B Billy Joe Reynolds, the petitioner, is a pre-Act offender. He was convicted of a Missouri sex offense in October 2001; he served four years in prison; he was released in July 2005; he then registered as a Missouri sex offender; but he moved to Pennsylvania in September 2007 without updating his Missouri registration information (as Missouri law required) and without registering in Pennsylvania. A federal grand jury indicted him, charging him with, between September 16 and October 16, 2007, having “knowingly failed to register and update a registration as required by [the Act].” App. 13; see 18 U. S. C. §2250(a). In the Government’s view, Reynolds’ failure to update his address information when he moved to Pennsylvania violated the requirement that a “sex offender” update registration information within “3 business days after each change of . . . residence.” 42 U. S. C. §16913(c). Reynolds moved to dismiss the indictment on the ground that in September and October 2007 the Act’s reg- istration requirements had not yet become applicable to pre-Act offenders. He conceded that the Act had become law earlier (namely, in July 2006), and he conceded that the Attorney General had already (in February 2007) promulgated an Interim Rule specifying that the Act’s registration requirements were applicable to pre-Act offenders. But he claimed that the Interim Rule was invalid because it violated both the Constitution’s “nondel- egation” doctrine and the Administrative Procedure Act’s (APA) requirement for “good cause” to promulgate a rule without “notice and comment” (as the Attorney General had done). See A. L. A. Schechter Poultry Corp. v. United States, 295 U.S. 495, 529 (1935) (nondelegation doctrine); 5 U. S. C. §§553(b)(3)(B), (d)(3) (APA). Because the Interim Rule is invalid, he added, the law must treat him like a pre-Act offender who traveled interstate and violated the Act’s registration requirements before the Attorney General specified their applicability. The District Court rejected on the merits Reynolds’ legal attack on the Interim Rule. But the Court of Appeals rejected Reynolds’ argument without reaching those merits. 380 Fed. Appx. 125 (2010). That court thought that the Act’s registration requirements apply to pre-Act offenders such as Reynolds (who was subject to a pre-existing state-law registration requirement) from the date of the new law’s enactment—even in the absence of any rule or regulation by the Attorney General specifying that the new registration requirements apply. That being so, the validity of the Interim Rule could make no legal dif- ference, for the Act required Reynolds to follow the new federal registration requirements regardless of any rulemaking. The Courts of Appeals have reached different conclusions about whether the Act’s registration requirements apply to pre-Act offenders prior to the time that the Attorney General specifies their applicability, i.e., from July 2006 until at least February 2007. Six Circuits have held that the Act’s registration requirements do not apply to pre-Act offenders unless and until the Attorney General so specifies. United States v. Johnson, 632 F.3d 912, 922–927 (CA5 2011); United States v. Valverde, 628 F.3d 1159, 1162–1164 (CA9 2010); United States v. Cain, 583 F.3d 408, 414–419 (CA6 2009); United States v. Hatcher, 560 F.3d 222, 226–229 (CA4 2009); United States v. Dixon, 551 F.3d 578, 585 (CA7 2008); United States v. Madera, 528 F.3d 852, 856–859 (CA11 2008) (per curiam). Five Circuits have held that they apply from the date of the Act’s enactment, and prior to any such specification, at least with respect to pre-Act offenders who had already registered under state law. United States v. Fuller, 627 F.3d 499, 506 (CA2 2010); United States v. DiTomasso, 621 F.3d 17, 24 (CA1 2010); United States v. Shenandoah, 595 F.3d 151, 163 (CA3 2010); United States v. Hinckley, 550 F.3d 926, 932 (CA10 2008); United States v. May, 535 F.3d 912, 918–919 (CA8 2008). In light of this split, we agreed to consider the question. II A The question before us is whether the Act requires pre-Act offenders to register before the Attorney General validly specifies that the Act’s registration provisions ap- ply to them. We believe that it does not. For one thing, a natural reading of the textual language supports our conclusion. The text consists of four statements. See supra, at 3. Statement One says that “[a] sex offender shall register, and keep the registration current.” Statement Two says that a sex offender must initially register before completing his “sentence of imprisonment” (or, if the sentence does not involve imprisonment, within three days of conviction). Statement Three says that the sex offender must update a registration within three business days of any change of “name, residence, employment, or student status.” Statement Four says that “[t]he Attorney General shall have the authority to specify the applicability of the requirements of this subchapter to sex offenders convicted before the enactment of this chapter.” Read naturally, the Fourth Statement modifies the First. It specifically deals with a subset (pre-Act offenders) of a broad general class (all sex offenders) to which the First Statement applies. And it therefore should control the Act’s application to that subset. See Gozlon-Peretz v. United States, 498 U.S. 395, 407 (1991) (specific statutory provision normally controls over one of more general application); see also Bloate v. United States, 559 U. S. ___, ___ (2010) (slip op., at 10) (same). At the same time, the Fourth Statement says that the Attorney General has authority to specify the Act’s “applicability,” not its “nonapplicability.” And it consequently is more naturally read as conferring the authority to apply the Act, not the authority to make exceptions. That is how we normally understand a term such as “authority to specify” in the context of applying new rules to persons already governed by pre-existing rules. If, for example, the Major League Baseball Players Association and the team owners agreed that the Commissioner of Baseball “shall have the authority to specify the applicability” to the major leagues of the more stringent minor league drug testing policy, we should think that the minor league policy would not apply unless and until the Commissioner so specified. For another thing, this reading of the Act efficiently resolves what Congress may well have thought were practical problems arising when the Act sought to apply the new registration requirements to pre-Act offenders. The problems arise out of the fact that the Act seeks to make more uniform a patchwork of pre-existing state systems. Doing so could require newly registering or re-registering “a large number” of pre-Act offenders. That effort could prove expensive. And it might not prove feasible to do so immediately. See 73 Fed. Reg. 38063 (recognizing these problems). Congress’ concern about these problems is re- flected in the Act’s providing the States with three years to bring their systems into compliance with federal standards while permitting the Attorney General to extend that 3-year grace period to five years. 42 U. S. C. §16924. These same considerations might have warranted different federal registration treatment of different categories of pre-Act offenders. Cf. 73 Fed. Reg. 38035–38036, and 38046–38047 (final Department of Justice guidelines allowing States to meet Act requirements without registering certain categories of pre-Act offenders); 76 Fed. Reg. 1635–1636 (supplemental guidelines allowing the same). At least Congress might well have so thought. And consequently, Congress might well have looked for a solution. Asking the Department of Justice, charged with respon- sibility for implementation, to examine these pre-Act offender problems and to apply the new registration requirements accordingly could have represented one efficient and desirable solution (though we express no view on Reynolds’ related constitutional claim). Cf. 42 U. S. C. §§16912(b), 16914(a)(7), (b)(7), 16919, 16941, 16945 (granting the Attorney General authority to administer various aspects of the Act). And that is just the solution that the Act’s language says that Congress adopted. Finally, our reading of the Act takes Congress to have filled potential lacunae (created by related Act provisions) in a manner consistent with basic background principles of criminal law. The Second Statement, for example, says that a sex offender must register before completing his prison term, but the provision says nothing about when a pre-Act offender who completed his prison term pre-Act must register. Although a state pre-Act offender could not be prosecuted until he traveled interstate, there is no interstate requirement for a federal pre-Act offender. And to apply the Act to either of these pre-Act offenders from the date of enactment would require reading into the statute, silent on the point, some kind of unsaid equivalent (e.g., registering or updating within a “reasonable time” or “within three days of first post-Act travel in interstate commerce” or “as preexisting state law requires”). Pre-Act offenders, aware of such complexities, lacunae, and difficulties, might, on their own, reach different conclusions about whether, or how, the new registration requirements applied to them. A ruling from the Attorney General, however, could diminish or eliminate those uncertainties, thereby helping to eliminate the very kind of vagueness and uncertainty that criminal law must seek to avoid. Cf., e.g., United States v. Lanier, 520 U.S. 259, 266 (1997) (noting that “the canon of strict construction of criminal statutes, or rule of lenity, ensures fair warning by so resolving ambiguity in a criminal statute as to apply it only to conduct clearly covered”). B The Government makes three principal arguments to the contrary. First, it says that our interpretation of the Act conflicts with one basic statutory purpose, namely, the “establish[ment of] a comprehensive national system for the registration of [sex] offenders,” 42 U. S. C. §16901, that includes offenders who committed their offenses before the Act became law. The Act reflects that purpose when it defines “sex offender” broadly to include any “individual who was convicted of a sex offense.” §16911(1). And we have recognized that purpose in stating that, in general, the Act’s criminal provisions apply to any pre-Act offender required to register under the Act who later travels interstate and fails to register. See Carr, 560 U. S., at ___ (slip op., at 7). The Act’s history also reveals that many of its supporters placed considerable importance upon the registration of pre-Act offenders. See, e.g., H. R. Rep. No. 109–218, pt. 1, p. 24 (2005) (H. R. Rep.) (“[Twenty] percent of sexual offenders are ‘lost,’ and there is a strong public interest in finding them and having them register with current information to mitigate the risks of additional crimes against children”); 152 Cong. Rec. 15333 (2006) (statement of Sen. Cantwell) (“Child sex offenders have exploited this stunning lack of uniformity, and the consequences have been tragic. Twenty percent of the Nation’s 560,000 sex offenders are ‘lost’ because State offender registry programs are not coordinated well enough”); id., at 15338 (statement of Sen. Kyl) (“There currently are over 100,000 sex offenders in this country who are required to register but are ‘off the system.’ They are not registered. The penalties in this bill should be adequate to ensure that these individuals register”); id., at 13050 (statement of Sen. Frist) (“There are currently 550,000 registered sex offenders in the U. S. and at least 100,000 of them are missing from the system. Every day that we don’t have this national sex offender registry, these missing sex predators are out there somewhere”). The difficulty with the Government’s argument, how- ever, is that it overstates the need for instantaneous registration of pre-Act offenders. Our different reading, we concede, involves implementation delay. But that delay need not be long (the Attorney General issued his Interim Rule 217 days after the effective date of the new law). And that delay can be justified by the need to accommodate other Act-related interests. See supra, at 7–9. Second, the Government suggests that our reading leads to an absurd result. As it points out, the Fourth Statement grants the Attorney General the “authority to specify” the registration requirements’ applicability not only to pre-Act offenders but also to those convicted prior to the “implementation” of the new Act “in a particular jurisdiction.” Some jurisdictions might not implement the Act for up to five years. See 42 U. S. C. §16924; see also Dept. of Justice, Office of Justice Programs, Justice Department Finds 24 Jurisdictions Have Substantially Implemented SORNA Requirements (July 28, 2011) (stating that as of July 28, 2011, 14 States had implemented the Act’s requirements), http://www.ojp.usdoj.gov/newsroom/ pressreleases/2011/SMART_PR-072811.htm (all Internet materials as visited Jan. 19, 2012, and available in Clerk of Court’s case file). Yet, the Government concludes, it is absurd to believe that Congress would have desired so long a delay in the application of its new registration requirements. The problem with this argument, however, is that reading the two categories similarly (a matter which we need not decide) would not require a long delay in applying the registration requirements to post-Act offenders who committed a crime in a jurisdiction that is slow to implement the new requirements. At most, that reading would require the Attorney General to promulgate a rule applicable to all preimplementation offenders. That rule could specify that the Act’s preregistration provisions apply to some or to all those offenders. And it could do so quickly, well before a jurisdiction implements the Act’s requirements. Indeed, the Attorney General’s Interim Rule and the Department of Justice’s final guidelines, both issued before any jurisdiction implemented the Act’s requirements, state that the Act’s requirements apply to “all sex offenders,” including all preimplementation offenders. See 72 Fed. Reg. 8897 (codified at 28 CFR §72.3); 73 Fed. Reg. 38036; cf. Dept. of Justice, Office of Justice Programs, Justice Department Announces First Two Jurisdictions to Implement Sex Offender Registration and Notifica- tion Act (Sept. 23, 2009), http://www.ojp.usdoj.gov/newsroom/ pressreleases/2009/SMART09154.htm. Third, the Government argues against our interpretation on the ground that the Act says only that the At- torney General “shall have the authority to specify the applicability” of the Act’s registration requirements to pre-Act offenders; it does not say that he “shall specify” or otherwise require him to do so. The Act’s language, the Government continues, consequently gives the Attorney General the power not to specify anything; that power is inconsistent with Congress’ intent to ensure the speedy registration of thousands of “lost” pre-Act offenders, supra, at 10; and we can avoid this result only by reading the Act’s registration requirements as applying immedi- ately and on their own to all pre-Act offenders (though the Attorney General would have the power to make exceptions). This argument bases too much upon too little. There is no reason to believe that Congress feared that the Attorney General would refuse to apply the new requirements to pre-Act offenders. See, e.g., H. R. Rep., at 23–24; Protecting Our Nation’s Children from Sexual Predators and Violent Criminals: What Needs To Be Done? Hearing before the Subcommittee on Crime, Terrorism, and Homeland Security of the House Committee on the Judiciary, 109th Cong., 1st Sess., 4–13 (2005); Office of the Press Sec’y, The White House, President Signs H. R. 4472, the Adam Walsh Child Protection and Safety Act of 2006 (July 27, 2006), http://georgewbush-whitehouse.archives.gov/ news/releases/2006/07/20060727-6.html. And there was no need for a mandatory requirement to avoid that unrealistic possibility. There is consequently no need to read the language unnaturally as giving the Attorney General the authority only to make exceptions from an implicit (unstated) rule that would otherwise apply the new registration requirements to all pre-Act offenders across the board and immediately. Finally, we note that some lower courts have read the Attorney General’s specification authority as applying only to those pre-Act sex offenders unable to comply with the statute’s “initial registration” requirements. See 42 U. S. C. §16913(b). That, however, is not what the statute says. Rather, its Fourth Statement, §16913(d), says that the Attorney General has the authority (1) to specify the applicability of the registration requirements to pre-Act (and preimplementation) offenders, “and ” (2) to prescribe rules for their registration, “and ” (3) to prescribe registration rules for other categories of sex offenders who are unable to comply with the initial registration requirements. See supra, at 3. The word “and” means that the Attorney General’s authority extends beyond those pre-Act “sex offenders who are unable to comply” with the initial registration requirements. III For these reasons, we conclude that the Act’s registration requirements do not apply to pre-Act offenders until the Attorney General so specifies. Whether the Attorney General’s Interim Rule sets forth a valid specification consequently matters in the case before us. And we reverse the Third Circuit’s judgment to the contrary. We remand the case for further proceedings consistent with this opinion. So ordered. APPENDIX 18 U. S. C. §2250(a) “In general.—Whoever— “(1) is required to register under the Sex Offender Registration and Notification Act; “(2)(A) is a sex offender as defined for the purposes of the Sex Offender Registration and Notification Act by reason of a conviction under Federal law (including the Uniform Code of Military Justice), the law of the District of Columbia, Indian tribal law, or the law of any territory or possession of the United States; or “(B) travels in interstate or foreign commerce, or enters or leaves, or resides in, Indian country; and “(3) knowingly fails to register or update a registration as required by the Sex Offender Registration and Notification Act; “shall be fined under this title or imprisoned not more than 10 years, or both.” 42 U. S. C. §16913 “Registry requirements for sex offenders “(a) In general “A sex offender shall register, and keep the registration current, in each jurisdiction where the offender resides, where the offender is an employee, and where the offender is a student. For initial registration purposes only, a sex offender shall also register in the jurisdiction in which convicted if such jurisdiction is different from the jurisdiction of residence. “(b) Initial registration “The sex offender shall initially register—(1) before completing a sentence of imprisonment with respect to the offense giving rise to the registration requirement; or (2) not later than 3 business days after being sentenced for that offense, if the sex offender is not sentenced to a term of imprisonment. “(c) Keeping the registration current “A sex offender shall, not later than 3 business days after each change of name, residence, employment, or student status, appear in person in at least 1 jurisdiction involved pursuant to subsection (a) and inform that jurisdiction of all changes in the information required for that offender in the sex offender registry. That jurisdiction shall immediately provide that information to all other jurisdictions in which the offender is required to register. “(d) Initial registration of sex offenders unable to comply with subsection (b) “The Attorney General shall have the authority to specify the applicability of the requirements of this subchapter to sex offenders convicted before the enactment of this chapter or its implementation in a particular jurisdiction, and to prescribe rules for the registration of any such sex offenders and for other categories of sex offenders who are unable to comply with subsection (b). “(e) State penalty for failure to comply “Each jurisdiction, other than a Federally recognized Indian tribe, shall provide a criminal penalty that includes a maximum term of imprisonment that is greater than 1 year for the failure of a sex offender to comply with the requirements of this subchapter.”

566.US.2011_10-1062

The Clean Water Act prohibits “the discharge of any pollutant by any person,” 33 U. S. C. §1311, without a permit, into “navigable waters,” §1344. Upon determining that a violation has occurred, the Environmental Protection Agency (EPA) may either issue a compliance order or initiate a civil enforcement action. §1319(a)(3). The resulting civil penalty may not “exceed [$37,500] per day for each violation.” §1319(d). The Government contends that the amount doubles to $75,000 when the EPA prevails against a person who has been issued a compliance order but has failed to comply. The Sacketts, petitioners here, received a compliance order from the EPA, which stated that their residential lot contained navigable waters and that their construction project violated the Act. The Sacketts sought declarative and injunctive relief in the Federal District Court, contending that the compliance order was “arbitrary [and] capricious” under the Administrative Procedure Act (APA), 5 U. S. C. §706(2)(A), and that it deprived them of due process in violation of the Fifth Amendment. The District Court dismissed the claims for want of subject-matter jurisdiction. The Ninth Circuit affirmed, concluding that the Clean Water Act precluded pre-enforcement judicial review of compliance orders and that such preclusion did not violate due process. Held: The Sacketts may bring a civil action under the APA to challenge the issuance of the EPA’s order. Pp. 4–10. (a) The APA provides for judicial review of “final agency action for which there is no other adequate remedy in a court.” 5 U. S. C. §704. The compliance order here has all the hallmarks of APA finality. Through it, the EPA “determined” “rights or obligations,” Bennett v. Spear, 520 U.S. 154, 178, requiring the Sacketts to restore their property according to an agency-approved plan and to give the EPA access. Also, “legal consequences . . . flow” from the order, ibid., which, according to the Government’s litigating position, exposes the Sacketts to double penalties in future enforcement proceedings. The order also severely limits their ability to obtain a permit for their fill from the Army Corps of Engineers, see 33 U. S. C. §1344; 33 CFR §326.3(e)(1)(iv). Further, the order’s issuance marks the “consummation” of the agency’s decisionmaking process, Bennett, supra, at 178, for the EPA’s findings in the compliance order were not subject to further agency review. The Sacketts also had “no other adequate remedy in a court,” 5 U. S. C. §704. A civil action brought by the EPA under 33 U. S. C. §1319 ordinarily provides judicial review in such cases, but the Sacketts cannot initiate that process. And each day they wait, they accrue additional potential liability. Applying to the Corps of Engineers for a permit and then filing suit under the APA if that permit is denied also does not provide an adequate remedy for the EPA’s action. Pp. 4–6. (b) The Clean Water Act is not a statute that “preclude[s] judicial review” under the APA, 5 U. S. C. §701(a)(1). The APA creates a “presumption favoring judicial review of administrative action.” Block v. Community Nutrition Institute, 467 U.S. 340, 349. While this presumption “may be overcome by inferences of intent drawn from the statutory scheme as a whole,” ibid., the Government’s arguments do not support an inference that the Clean Water Act’s statutory scheme precludes APA review. Pp. 7–10. 622 F.3d 1139, reversed and remanded. Scalia, J., delivered the opinion for a unanimous Court. Ginsburg, J., and Alito, J., filed concurring opinions.

We consider whether Michael and Chantell Sackett may bring a civil action under the Administrative Procedure Act, 5 U. S. C. §500 et seq., to challenge the issuance by the Environmental Protection Agency (EPA) of an administrative compliance order under §309 of the Clean Water Act, 33 U. S. C. §1319. The order asserts that the Sacketts’ property is subject to the Act, and that they have violated its provisions by placing fill material on the property; and on this basis it directs them immediately to restore the property pursuant to an EPA work plan. I The Clean Water Act prohibits, among other things, “the discharge of any pollutant by any person,” §1311, without a permit, into the “navigable waters,” §1344—which the Act defines as “the waters of the United States,” §1362(7). If the EPA determines that any person is in violation of this restriction, the Act directs the agency either to issue a compliance order or to initiate a civil enforcement action. §1319(a)(3). When the EPA prevails in a civil action, the Act provides for “a civil penalty not to exceed [$37,500] per day for each violation.”[1] §1319(d). And according to the Government, when the EPA prevails against any person who has been issued a compliance order but has failed to comply, that amount is increased to $75,000—up to $37,500 for the statutory violation and up to an additional $37,500 for violating the compliance order. The particulars of this case flow from a dispute about the scope of “the navigable waters” subject to this enforcement regime. Today we consider only whether the dispute may be brought to court by challenging the compliance order—we do not resolve the dispute on the merits. The reader will be curious, however, to know what all the fuss is about. In United States v. Riverside Bayview Homes, Inc., 474 U.S. 121 (1985), we upheld a regulation that construed “the navigable waters” to include “freshwater wetlands,” id., at 124, themselves not actually navigable, that were adjacent to navigable-in-fact waters. Later, in Solid Waste Agency of Northern Cook Cty. v. Army Corps of Engineers, 531 U.S. 159 (2001), we held that an abandoned sand and gravel pit, which “seasonally ponded” but which was not adjacent to open water, id., at 164, was not part of the navigable waters. Then most recently, in Rapanos v. United States, 547 U.S. 715 (2006), we considered whether a wetland not adjacent to navigable-in-fact waters fell within the scope of the Act. Our answer was no, but no one rationale commanded a majority of the Court. In his separate opinion, The Chief Justice expressed the concern that interested parties would lack guidance “on precisely how to read Congress’ limits on the reach of the Clean Water Act” and would be left “to feel their way on a case-by-case basis.” Id., at 758 (concurring opinion). The Sacketts are interested parties feeling their way. They own a 2∕3-acre residential lot in Bonner County, Idaho. Their property lies just north of Priest Lake, but is separated from the lake by several lots containing permanent structures. In preparation for constructing a house, the Sacketts filled in part of their lot with dirt and rock. Some months later, they received from the EPA a compliance order. The order contained a number of “Findings and Conclusions,” including the following: “1.4 [The Sacketts’ property] contains wetlands within the meaning of 33 C. F. R. §328.4(8)(b); the wetlands meet the criteria for jurisdictional wetlands in the 1987 ‘Federal Manual for Identifying and Delineating Jurisdictional Wetlands.’ “1.5 The Site’s wetlands are adjacent to Priest Lake within the meaning of 33 C. F. R. §328.4(8)(c). Priest Lake is a ‘navigable water’ within the meaning of section 502(7) of the Act, 33 U. S. C. §1362(7), and ‘waters of the United States’ within the meaning of 40 C. F. R. §232.2. “1.6 In April and May, 2007, at times more fully known to [the Sacketts, they] and/or persons acting on their behalf discharged fill material into wetlands at the Site. [They] filled approximately one half acre. . . . . . “1.9 By causing such fill material to enter waters of the United States, [the Sacketts] have engaged, and are continuing to engage, in the ‘discharge of pollutants’ from a point source within the meaning of sections 301 and 502(12) of the Act, 33 U. S. C. §§1311 and 1362(12). . . . . . “1.11 [The Sacketts’] discharge of pollutants into waters of the United States at the Site without [a] permit constitutes a violation of section 301 of the Act, 33 U. S. C. §1311.” App. 19–20. On the basis of these findings and conclusions, the order directs the Sacketts, among other things, “immediately [to] undertake activities to restore the Site in accordance with [an EPA-created] Restoration Work Plan” and to “pro- vide and/or obtain access to the Site . . . [and] access to all records and documentation related to the conditions at the Site . . . to EPA employees and/or their designated representatives.” Id., at 21–22, ¶¶2.1, 2.7. The Sacketts, who do not believe that their property is subject to the Act, asked the EPA for a hearing, but that request was denied. They then brought this action in the United States District Court for the District of Idaho, seeking declaratory and injunctive relief. Their complaint contended that the EPA’s issuance of the compliance order was “arbitrary [and] capricious” under the Administrative Procedure Act (APA), 5 U. S. C. §706(2)(A), and that it deprived them of “life, liberty, or property, without due process of law,” in violation of the Fifth Amendment. The District Court dismissed the claims for want of subject- matter jurisdiction, and the United States Court of Appeals for the Ninth Circuit affirmed, 622 F.3d 1139 (2010). It concluded that the Act “preclude[s] pre-enforcement judicial review of compliance orders,” id., at 1144, and that such preclusion does not violate the Fifth Amendment’s due process guarantee, id., at 1147. We granted certiorari. 564 U. S. ___ (2011). II The Sacketts brought suit under Chapter 7 of the APA, which provides for judicial review of “final agency action for which there is no other adequate remedy in a court.” 5 U. S. C. §704. We consider first whether the compliance order is final agency action. There is no doubt it is agency action, which the APA defines as including even a “failure to act.” §§551(13), 701(b)(2). But is it final? It has all of the hallmarks of APA finality that our opinions establish. Through the order, the EPA “ ‘determined’ ” “ ‘rights or ob- ligations.’ ” Bennett v. Spear, 520 U.S. 154, 178 (1997) (quoting Port of Boston Marine Terminal Assn. v. Re- deriaktiebolaget Transatlantic, 400 U.S. 62, 71 (1970)). By reason of the order, the Sacketts have the legal obligation to “restore” their property according to an agency-approved Restoration Work Plan, and must give the EPA access to their property and to “records and documentation related to the conditions at the Site.” App. 22, ¶2.7. Also, “ ‘legal consequences . . . flow’ ” from issuance of the order. Bennett, supra, at 178 (quoting Marine Terminal, supra, at 71). For one, according to the Government’s current litigating position, the order exposes the Sacketts to double penalties in a future enforcement proceeding.[2] It also severely limits the Sacketts’ ability to obtain a permit for their fill from the Army Corps of Engineers, see 33 U. S. C. §1344. The Corps’ regulations provide that, once the EPA has issued a compliance order with respect to certain property, the Corps will not process a permit application for that property unless doing so “is clearly appropriate.” 33 CFR §326.3(e)(1)(iv) (2011).[3] The issuance of the compliance order also marks the “ ‘consummation’ ” of the agency’s decisionmaking process. Bennett, supra, at 178 (quoting Chicago & Southern Air Lines, Inc. v. Waterman S. S. Corp., 333 U.S. 103, 113 (1948)). As the Sacketts learned when they unsuccessfully sought a hearing, the “Findings and Conclusions” that the compliance order contained were not subject to further agency review. The Government resists this conclusion, pointing to a portion of the order that invited the Sacketts to “engage in informal discussion of the terms and requirements” of the order with the EPA and to inform the agency of “any allegations [t]herein which [they] believe[d] to be inaccurate.” App. 22–23, ¶2.11. But that confers no entitlement to further agency review. The mere possibility that an agency might reconsider in light of “informal discussion” and invited contentions of inaccuracy does not suffice to make an otherwise final agency action nonfinal. The APA’s judicial review provision also requires that the person seeking APA review of final agency action have “no other adequate remedy in a court,” 5 U. S. C. §704. In Clean Water Act enforcement cases, judicial review ordinarily comes by way of a civil action brought by the EPA under 33 U. S. C. §1319. But the Sacketts cannot initiate that process, and each day they wait for the agency to drop the hammer, they accrue, by the Government’s telling, an additional $75,000 in potential liability. The other possible route to judicial review—applying to the Corps of Engineers for a permit and then filing suit under the APA if a permit is denied—will not serve either. The remedy for denial of action that might be sought from one agency does not ordinarily provide an “adequate remedy” for ac- tion already taken by another agency. The Government, to its credit, does not seriously contend that other available remedies alone foreclose review under §704. Instead, the Government relies on §701(a)(1) of the APA, which excludes APA review “to the extent that [other] statutes preclude judicial review.” The Clean Water Act, it says, is such a statute. III Nothing in the Clean Water Act expressly precludes judicial review under the APA or otherwise. But in determining “[w]hether and to what extent a particular statute precludes judicial review,” we do not look “only [to] its express language.” Block v. Community Nutrition Institute, 467 U.S. 340, 345 (1984). The APA, we have said, creates a “presumption favoring judicial review of administrative action,” but as with most presumptions, this one “may be overcome by inferences of intent drawn from the statutory scheme as a whole.” Id., at 349. The Government offers several reasons why the statutory scheme of the Clean Water Act precludes review. The Government first points to 33 U. S. C. §1319(a)(3), which provides that, when the EPA “finds that any person is in violation” of certain portions of the Act, the agency “shall issue an order requiring such person to comply with [the Act], or . . . shall bring a civil action [to enforce the Act].” The Government argues that, because Congress gave the EPA the choice between a judicial proceeding and an administrative action, it would undermine the Act to allow judicial review of the latter. But that argument rests on the question-begging premise that the relevant difference between a compliance order and an enforcement proceeding is that only the latter is subject to judicial review. There are eminently sound reasons other than insulation from judicial review why compliance orders are useful. The Government itself suggests that they “provid[e] a means of notifying recipients of potential vio- lations and quickly resolving the issues through volun- tary compliance.” Brief for Respondents 39. It is entirely consistent with this function to allow judicial review when the recipient does not choose “voluntary compliance.” The Act does not guarantee the EPA that issuing a compliance order will always be the most effective choice. The Government also notes that compliance orders are not self-executing, but must be enforced by the agency in a plenary judicial action. It suggests that Congress therefore viewed a compliance order “as a step in the deliberative process[,] . . . rather than as a coercive sanction that itself must be subject to judicial review.” Id., at 38. But the stocktickerAPA provides for judicial review of all final agency actions, not just those that impose a self-executing sanction. And it is hard for the Government to defend its claim that the issuance of the compliance order was just “a step in the deliberative process” when the agency rejected the Sacketts’ attempt to obtain a hearing and when the next step will either be taken by the Sacketts (if they comply with the order) or will involve judicial, not administrative, deliberation (if the EPA brings an enforcement action). As the text (and indeed the very name) of the compliance order makes clear, the EPA’s “deliberation” over whether the Sacketts are in violation of the Act is at an end; the agency may still have to deliberate over whether it is confident enough about this conclusion to initiate litigation, but that is a separate subject. The Government further urges us to consider that Congress expressly provided for prompt judicial review, on the administrative record, when the EPA assesses administrative penalties after a hearing, see §1319(g)(8), but did not expressly provide for review of compliance orders. But if the express provision of judicial review in one section of a long and complicated statute were alone enough to over- come the APA’s presumption of reviewability for all final agency action, it would not be much of a presumption at all. The cases on which the Government relies simply are not analogous. In Block v. Community Nutrition Institute, supra, we held that the Agricultural Marketing Agreement Act of 1937, which expressly allowed milk handlers to obtain judicial review of milk market orders, precluded review of milk market orders in suits brought by milk consumers. 467 U. S., at 345–348. Where a statute provides that particular agency action is reviewable at the instance of one party, who must first exhaust administrative remedies, the inference that it is not reviewable at the instance of other parties, who are not subject to the administrative process, is strong. In United States v. Erika, Inc., 456 U.S. 201 (1982), we held that the Medicare statute, which expressly provided for judicial review of awards under Part A, precluded review of awards under Part B. Id., at 206–208. The strong parallel between the award provisions in Part A and Part B of the Medicare statute does not exist between the issuance of a compliance order and the assessment of administrative penalties under the Clean Water Act. And in United States v. Fausto, 484 U.S. 439 (1988), we held that the Civil Service Reform Act, which expressly excluded certain “nonpreference” employees from the statute’s review scheme, precluded review at the instance of those employees in a separate Claims Court action. Id., at 448–449. Here, there is no suggestion that Congress has sought to exclude compliance-order recipients from the Act’s review scheme; quite to the contrary, the Government’s case is premised on the notion that the Act’s primary review mechanisms are open to the Sacketts. Finally, the Government notes that Congress passed the Clean Water Act in large part to respond to the inefficiency of then-existing remedies for water pollution. Compliance orders, as noted above, can obtain quick remediation through voluntary compliance. The Government warns that the EPA is less likely to use the orders if they are subject to judicial review. That may be true—but it will be true for all agency actions subjected to judicial review. The APA’s presumption of judicial review is a repudiation of the principle that efficiency of regulation conquers all. And there is no reason to think that the Clean Water Act was uniquely designed to enable the strong-arming of regulated parties into “voluntary compliance” without the opportunity for judicial review—even judicial review of the question whether the regulated party is within the EPA’s jurisdiction. Compliance orders will remain an effective means of securing prompt voluntary compliance in those many cases where there is no substantial basis to question their validity. * * * We conclude that the compliance order in this case is final agency action for which there is no adequate remedy other than APA review, and that the Clean Water Act does not preclude that review. We therefore reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 The original statute set a penalty cap of $25,000 per violation per day. The Federal Civil Penalties Inflation Adjustment Act of 1990, 104Stat. 890, note following 28 U. S. C. §2461, as amended by the Debt Collection Improvement Act of 1996, §3720E, 110Stat. 1321–373, note following 28 U. S. C. §2461, p. 1315 (Amendment), authorizes the EPA to adjust that maximum penalty for inflation. On the basis of that authority, the agency has raised the cap to $37,500. See 74 Fed. Reg. 626, 627 (2009). 2 We do not decide today that the Government’s position is correct, but assume the consequences of the order to be what the Government asserts. 3 The regulation provides this consequence for “enforcement litigation that has been initiated by other Federal . . . regulatory agencies.” 33 CFR §326.3(e)(1)(iv) (2011). The Government acknowledges, however, that EPA’s issuance of a compliance order is considered by the Corps to fall within the provision. Brief for Respondents 31. Here again, we take the Government at its word without affirming that it represents a proper interpretation of the regulation.

567.US.2011_11-551

The Indian Self-Determination and Education Assistance Act (ISDA) directs the Secretary of the Interior to enter into contracts with willing tribes under which they will provide services such as education and law enforcement that the Federal Government otherwise would have provided. It requires the Secretary to contract to pay the “full amount” of “contract support costs,” 45 U. S. C. §§450j–1(a)(2), (g), subject to the availability of appropriations, §450j–1(b). In the event of a contractual breach, tribal contractors are entitled to seek money damages under the Contract Disputes Act. In Fiscal Years (FYs) 1994 to 2001, respondent Tribes contracted with the Secretary to provide services. During each of those FYs, Congress appropriated sufficient funds to pay any individual tribal contractor’s contract support costs in full but did not appropriate enough to pay all tribal contractors collectively. Unable to pay every contractor in full, the Secretary paid the Tribes on a uniform, pro rata basis. Respondents sued under the Contract Disputes Act for breach of contract. The District Court granted the Government summary judgment. The Tenth Circuit reversed, finding the Government liable to each contractor for the full contract amount. Held: The Government must pay each Tribe’s contract support costs in full. Pp. 5−18. (a) In Cherokee Nation of Okla. v. Leavitt, 543 U.S. 631, this Court considered the Government’s promise to pay contract support costs in ISDA self-determination contracts that made the Government’s obligation “subject to the availability of appropriations,” id., at 634−637. The Government contended that Congress appropriated inadequate funds to fulfill its contractual obligations to the Tribes, while meeting the agency’s competing fiscal priorities. Because Congress appropriated sufficient legally unrestricted funds to pay the contracts, however, the Court held that the Government was obligated to pay those costs in full absent “something special about the promises,” id., at 637–638. That conclusion followed directly from well-established principles of Government contracting law: When a Government contractor is one of several persons to be paid out of a larger appropriation sufficient in itself to pay the contractor, the Government is responsible to the contractor for the full amount due under the contract, even if the agency exhausts the appropriation in service of other permissible ends. See Ferris v. United States, 27 Ct. Cl. 542, 546. That is so “even if an agency’s total lump-sum appropriation is insufficient to pay all” of its contracts. Cherokee Nation, 543 U. S., at 637. This principle safeguards both the expectations of Government contractors and the long-term fiscal interests of the United States. Contractors need not keep track of agencies’ shifting priorities and competing obligations; rather, they may trust that the Government will honor its contractual promises. And the rule furthers “the Government’s own long-run interest as a reliable contracting partner in the myriad workaday transaction of its agencies.” United States v. Winstar Corp., 518 U.S. 839, 883. Pp. 5–8. (b) The principles underlying Cherokee Nation and Ferris control here. Once “Congress has appropriated sufficient legally unrestricted funds to pay the contracts at issue, the Government normally cannot back out of a promise on grounds of ‘insufficient appropriations,’ even if the contract uses language such as ‘subject to the availability of appropriations,’ and even if an agency’s total lump-sum appropriation is insufficient to pay all the contracts the agency has made.” Cherokee Nation, 543 U. S., at 637. That condition is satisfied here, because Congress made sufficient funds available to pay any individual contractor in full. Pp. 8−10. (c) The Government attempts to distinguish Ferris and Cherokee Nation on the ground that they involved unrestricted, lump-sum appropriations, while Congress here appropriated “not to exceed” a certain amount for contract support costs. The effect of the appropriations in each case, however, was identical: the agency remained free to allocate funds among multiple contractors, so long as the contracts served the purpose Congress identified. The “not to exceed” language still has legal effect; it prevents the Secretary from reprogramming other funds to pay contract support costs, thereby protecting funds that Congress envisioned for other Bureau of Indian Affairs programs. Section 450j–1(b), which specifies that the Secretary is not required to reduce funding for one tribe’s programs to make funds available to another tribe, does not warrant a different result. Consistent with ordinary Government contracting principles, that language merely underscores the Secretary’s discretion to allocate funds among tribes. It does not alter the Government’s legal obligation when the Secretary fails to pay. The Government’s remaining counterarguments are unpersuasive. First, it suggests that the Secretary could violate the Anti-Deficiency Act, which prevents federal officers from making or authorizing an expenditure or obligation exceeding an amount available in an appropriation. That Act applies only to government officials, however, and does not affect the rights of citizens contracting with the Government. Second, the Government argues that permitting respondents to recover from the Judgment Fund would circumvent Congress’ intent to cap total expenditures for contract support costs. But ISDA expressly provides that tribal contractors may sue for “money damages” under the Contract Disputes Act, and any ensuing judgments are payable from the Judgment Fund. See Cherokee Nation, 543 U. S., at 642. Third, the Government invokes cases in which courts have rejected contractors’ attempts to recover for amounts beyond the maximum appropriated by Congress for a particular purpose. See, e.g., Sutton v. United States, 256 U.S. 575. However, Sutton involved a specific line-item appropriation for an amount beyond which the sole contractor could not recover. This case involves several contractors, each of whom contracted within the lump-sum amount Congress appropriated for all contractors. Unlike the sole contractor in Sutton, they cannot reasonably be expected to know how much remained available of Congress’ lump-sum appropriation. Finally, the Government claims that legislative history suggests that Congress approved of pro rata distribution, but “indicia in committee reports and other legislative history as to how funds should or are expected to be spent do not establish any legal requirement on the agency.” Lincoln v. Vigil, 508 U.S. 182, 192. Pp. 11−17. (d) This case is the product of two decisions in some tension: Congress required the Secretary to accept every qualifying ISDA contract, promising “full” funding for all contract support costs, but then appropriated insufficient funds to pay in full each tribal contractor. Responsibility for the resolution of that situation, however, is committed to Congress. Pp. 17−18. 644 F.3d 1054, affirmed. Sotomayor, J., delivered the opinion of the Court, in which Scalia, Kennedy, Thomas, and Kagan, JJ., joined. Roberts, C. J., filed a dissenting opinion, in which Ginsburg, Breyer, and Alito, JJ., joined.

The Indian Self-Determination and Education Assistance Act (ISDA), 25 U. S. C. §450 et seq., directs the Secretary of the Interior to enter into contracts with willing tribes, pursuant to which those tribes will provide services such as education and law enforcement that otherwise would have been provided by the Federal Government. ISDA mandates that the Secretary shall pay the full amount of “contract support costs” incurred by tribes in performing their contracts. At issue in this case is whether the Government must pay those costs when Congress appropriates sufficient funds to pay in full any individual contractor’s contract support costs, but not enough funds to cover the aggregate amount due every contractor. Consistent with longstanding principles of Government contracting law, we hold that the Government must pay each tribe’s contract support costs in full. I A Congress enacted ISDA in 1975 in order to achieve “maximum Indian participation in the direction of educational as well as other Federal services to Indian communities so as to render such services more responsive to the needs and desires of those communities.” 25 U. S. C. §450a(a). To that end, the Act directs the Secretary of the Interior, “upon the request of any Indian tribe . . . to enter into a self-determination contract . . . to plan, conduct, and administer” health, education, economic, and social programs that the Secretary otherwise would have administered. §450f(a)(1). As originally enacted, ISDA required the Government to provide contracting tribes with an amount of funds equiv- alent to those that the Secretary “would have other- wise provided for his direct operation of the programs.” §106(h), 88Stat. 2211. It soon became apparent that this secretarial amount failed to account for the full costs to tribes of providing services. Because of “concern with Government’s past failure adequately to reimburse tribes’ indirect administrative costs,” Cherokee Nation of Okla. v. Leavitt, 543 U.S. 631, 639 (2005), Congress amended ISDA to require the Secretary to contract to pay the “full amount” of “contract support costs” related to each self-determination contract, §§450j–1(a)(2), (g).[1] The Act also provides, however, that “[n]otwithstanding any other provision in [ISDA], the provision of funds under [ISDA] is subject to the availability of appropriations.” §450j–1(b). Congress included a model contract in ISDA and directed that each tribal self-determination contract “shall . . . contain, or incorporate [it] by reference.” §450l(a)(1). The model contract specifies that “ ‘[s]ubject to the availability of appropriations, the Secretary shall make avail- able to the Contractor the total amount specified in the annual funding agreement’ ” between the Secretary and the tribe. §450l(c), (model agreement §1(b)(4)). That amount “ ‘shall not be less than the applicable amount determined pursuant to [§450j–1(a)],’ ” which includes contract support costs. Ibid.; §450j–1(a)(2). The contract indicates that “ ‘[e]ach provision of [ISDA] and each provision of this Contract shall be liberally construed for the benefit of the Contractor . . . .’ ” §450l(c), (model agreement §1(a)(2)). Finally, the Act makes clear that if the Government fails to pay the amount contracted for, then tribal contractors are entitled to pursue “money dam- ages” in accordance with the Contract Disputes Act. §450m–1(a). B During Fiscal Years (FYs) 1994 to 2001, respondent Tribes contracted with the Secretary of the Interior to provide services such as law enforcement, environmental protection, and agricultural assistance. The Tribes fully performed. During each FY, Congress appropriated a total amount to the Bureau of Indian Affairs (BIA) “for the operation of Indian programs.” See, e.g., Department of the Interior and Related Agencies Appropriations Act, 2000, 113Stat. 1501A–148. Of that sum, Congress provided that “not to exceed [a particular amount] shall be available for payments to tribes and tribal organiza- tions for contract support costs” under ISDA. E.g., ibid. Thus, in FY 2000, for example, Congress appropriated $1,670,444,000 to the BIA, of which “not to exceed $120,229,000” was allocated for contract support costs. Ibid. During each relevant FY, Congress appropriated sufficient funds to pay in full any individual tribal contractor’s contract support costs. Congress did not, however, appropriate sufficient funds to cover the contract support costs due all tribal contractors collectively. Between FY 1994 and 2001, appropriations covered only between 77% and 92% of tribes’ aggregate contract support costs. The extent of the shortfall was not revealed until each fiscal year was well underway, at which point a tribe’s performance of its contractual obligations was largely complete. See 644 F.3d 1054, 1061 (CA10 2011). Lacking funds to pay each contractor in full, the Secretary paid tribes’ contract support costs on a uniform, pro rata basis. Tribes responded to these shortfalls by reducing ISDA services to tribal members, diverting tribal resources from non-ISDA programs, and forgoing opportunities to contract in furtherance of Congress’ self-determination objective. GAO, V. Rezendes, Indian Self-Determination Act: Shortfalls in Indian Contract Support Costs Need to Be Addressed 3–4 (GAO/RCED–99–150, 2009). Respondent Tribes sued for breach of contract pursuant to the Contract Disputes Act, 41 U. S. C. §§601–613, alleging that the Government failed to pay the full amount of contract support costs due from FY 1994 through 2001, as required by ISDA and their contracts. The United States District Court for the District of New Mexico granted summary judgment for the Government. A divided panel of the United States Court of Appeals for the Tenth Circuit reversed. The court reasoned that Congress made sufficient appropriations “legally available” to fund any individual tribal contractor’s contract support costs, and that the Government’s contractual commitment was therefore binding. 644 F. 3d, at 1063–1065. In such cases, the Court of Appeals held that the Government is liable to each contractor for the full contract amount. Judge Hartz dissented, contending that Congress intended to set a maximum limit on the Government’s liability for contract support costs. We granted certiorari to resolve a split among the Courts of Appeals, 565 U. S. ___ (2012), and now affirm.[2] II A In evaluating the Government’s obligation to pay tribes for contract support costs, we do not write on a clean slate. Only seven years ago, in Cherokee Nation, we also con- sidered the Government’s promise to pay contract sup- port costs in ISDA self-determination contracts that made the Government’s obligation “subject to the availability of appropriations.” 543 U. S., at 634–637. For each FY at issue, Congress had appropriated to the Indian Health Service (IHS) a lump sum between $1.277 and $1.419 billion, “far more than the [contract support cost] amounts” due under the Tribes’ individual contracts. Id., at 637; see id., at 636 (Cherokee Nation and Shoshone-Paiute Tribes filed claims seeking $3.4 and $3.5 million, respectively). The Government contended, however, that Congress had appropriated inadequate funds to enable the IHS to pay the Tribes’ contract support costs in full, while meeting all of the agency’s competing fiscal priorities. As we explained, that did not excuse the Government’s responsibility to pay the Tribes. We stressed that the Government’s obligation to pay contract support costs should be treated as an ordinary contract promise, noting that ISDA “uses the word ‘contract’ 426 times to describe the nature of the Government’s promise.” Id., at 639. As even the Government conceded, “in the case of ordinary contracts . . . ‘if the amount of an unrestricted appropriation is sufficient to fund the contract, the contractor is entitled to payment even if the agency has allocated the funds to another purpose or assumes other obligations that exhaust the funds.’ ” Id., at 641. It followed, therefore, that absent “something special about the promises at issue,” the Government was obligated to pay the Tribes’ contract support costs in full. Id., at 638. We held that the mere fact that ISDA self-determination contracts are made “subject to the availability of appropriations” did not warrant a special rule. Id., at 643 (internal quotation marks omitted). That commonplace provision, we explained, is ordinarily satisfied so long as Congress appropriates adequate legally unrestricted funds to pay the contracts at issue. See ibid. Because Congress made sufficient funds legally available to the agency to pay the Tribes’ contracts, it did not matter that the BIA had allocated some of those funds to serve other purposes, such that the remainder was insufficient to pay the Tribes in full. Rather, we agreed with the Tribes that “as long as Congress has appropriated sufficient legally unrestricted funds to pay the contracts at issue,” the Government’s promise to pay was binding. Id., at 637–638. Our conclusion in Cherokee Nation followed directly from well-established principles of Government contracting law. When a Government contractor is one of several persons to be paid out of a larger appropriation sufficient in itself to pay the contractor, it has long been the rule that the Government is responsible to the contractor for the full amount due under the contract, even if the agency exhausts the appropriation in service of other permissible ends. See Ferris v. United States, 27 Ct. Cl. 542, 546 (1892); Dougherty v. United States, 18 Ct. Cl. 496, 503 (1883); see also 2 GAO, Principles of Federal Appropriations Law, p. 6–17 (2d ed. 1992) (hereinafter GAO Redbook).[3] That is so “even if an agency’s total lump-sum appropriation is insufficient to pay all the contracts the agency has made.” Cherokee Nation, 543 U. S., at 637. In such cases, “[t]he United States are as much bound by their contracts as are individuals.” Lynch v. United States, 292 U.S. 571, 580 (1934) (internal quotation marks omitted). Although the agency itself cannot disburse funds beyond those appropriated to it, the Government’s “valid obligations will remain enforceable in the courts.” GAO Redbook, p. 6–17. This principle safeguards both the expectations of Government contractors and the long-term fiscal interests of the United States. For contractors, the Ferris rule reflects that when “a contract is but one activity under a larger appropriation, it is not reasonable to expect the contractor to know how much of that appropriation remains available for it at any given time.” GAO Redbook, p. 6–18. Contractors are responsible for knowing the size of the pie, not how the agency elects to slice it. Thus, so long as Congress appropriates adequate funds to cover a prospective contract, contractors need not keep track of agencies’ shifting priorities and competing obligations; rather, they may trust that the Government will honor its contractual promises. Dougherty, 18 Ct. Cl., at 503. In such cases, if an agency overcommits its funds such that it cannot fulfill its contractual commitments, even the Government has acknowledged that “[t]he risk of over-obligation may be found to fall on the agency,” not the contractor. Brief for Federal Parties in Cherokee Nation v. Leavitt, O. T. 2004, No. 02–1472 et al., p. 24 (hereinafter Brief for Federal Parties). The rule likewise furthers “the Government’s own long-run interest as a reliable contracting partner in the myr- iad workaday transaction of its agencies.” United States v. Winstar Corp., 518 U.S. 839, 883 (1996) (plurality opinion). If the Government could be trusted to fulfill its promise to pay only when more pressing fiscal needs did not arise, would-be contractors would bargain warily—if at all—and only at a premium large enough to account for the risk of nonpayment. See, e.g., Logue, Tax Transitions, Opportunistic Retroactivity, and the Benefits of Government Precommitment, 94 Mich. L. Rev. 1129, 1146 (1996). In short, contracting would become more cumbersome and expensive for the Government, and willing partners more scarce. B The principles underlying Cherokee Nation and Ferris dictate the result in this case. Once “Congress has appropriated sufficient legally unrestricted funds to pay the contracts at issue, the Government normally cannot back out of a promise to pay on grounds of ‘insufficient appropriations,’ even if the contract uses language such as ‘subject to the availability of appropriations,’ and even if an agency’s total lump-sum appropriation is insufficient to pay all the contracts the agency has made.” Cherokee Nation, 543 U. S., at 637; see also id., at 638 (“[T]he Government denies none of this”). That condition is satisfied here. In each FY between 1994 and 2001, Congress appropriated to the BIA a lump-sum from which “not to exceed” between $91 and $125 million was allocated for contract support costs, an amount that exceeded the sum due any tribal contractor. Within those constraints, the ability to direct those funds was “ ‘committed to agency discretion by law.’ ” Lincoln v. Vigil, 508 U.S. 182, 193 (1993) (quoting 5 U. S. C. §701(a)(2)). Nothing, for instance, prevented the BIA from paying in full respondent Ramah Navajo Chapter’s contract support costs rather than other tribes’, whether based on its greater need or simply because it sought payment first.[4] See International Union, United Auto., Aerospace & Agricultural Implement Workers of Am. v. Donovan, 746 F.2d 855, 861 (CADC 1984) (Scalia, J.) (“A lump-sum appropriation leaves it to the recipient agency (as a matter of law, at least) to distribute the funds among some or all of the permissible objects as it sees fit”). And if there was any doubt that that general rule applied here, ISDA’s statutory language itself makes clear that the BIA may allocate funds to one tribe at the expense of another. See §450j–1(b) (“[T]he Secretary is not required to reduce funding for programs, projects, or activities serving a tribe to make funds available to another tribe or tribal or- ganization under this [Act]”). The upshot is that the funds appropriated by Congress were legally available to pay any individual tribal contractor in full. See 1 GAO Redbook, p. 4–6 (3d ed. 2004). The Government’s contractual promise to pay each tribal contractor the “full amount of funds to which the contractor [was] entitled,” §450j–1(g), was therefore binding. We have expressly rejected the Government’s argument that “the tribe should bear the risk that a total lump-sum appropriation (though sufficient to cover its own contracts) will not prove sufficient to pay all similar contracts.” Cherokee Nation, 543 U. S., at 638. Rather, the tribal contractors were entitled to rely on the Government’s promise to pay because they were “not chargeable with knowledge” of the BIA’s administration of Congress’ appropriation, “nor [could their] legal rights be affected or impaired by its maladministration or by its diversion.” Ferris, 27 Ct. Cl., at 546. As in Cherokee Nation, we decline the Government’s invitation to ascribe “special, rather than ordinary” meaning to the fact that ISDA makes contracts “subject to the availability of appropriations.”[5] 543 U. S., at 644. Under our previous interpretation of that language, that condition was satisfied here because Congress appropriated adequate funds to pay in full any individual contractor. It is important to afford that language a “uniform interpretation” in this and comparable statutes, “lest legal uncertainty undermine contractors’ confidence that they will be paid, and in turn increase the cost to the Government of purchasing goods and services.” Ibid. It would be particularly anomalous to read the statutory language differently here. Contracts made under ISDA specify that “ ‘[e]ach provision of the [ISDA] and each provision of this Contract shall be liberally construed for the benefit of the Contractor. . . .’ ” §450l(c), (model agreement §1(a)(2)). The Government, in effect, must demonstrate that its reading is clearly required by the statutory language. Accordingly, the Government cannot back out of its contractual promise to pay each Tribe’s full contract support costs. III A The Government primarily seeks to distinguish this case from Cherokee Nation and Ferris on the ground that Congress here appropriated “not to exceed” a given amount for contract support costs, thereby imposing an express cap on the total funds available. See Brief for Petitioners 26, 49. The Government argues, on this basis, that Ferris and Cherokee Nation involved “contracts made against the back- drop of unrestricted, lump-sum appropriations,” while this case does not. See Brief for Petitioners 49, 26. That premise, however, is inaccurate. In Ferris, Congress appropriated “[f]or improving Delaware River below Bridesburg, Pennsylvania, forty-five thousand dollars.” 20Stat. 364. As explained in the Government’s own appropriations law handbook, the “not to exceed” language at issue in this case has an identical meaning to the quoted language in Ferris. See GAO Redbook, p. 6–5 (“Words like ‘not to exceed’ are not the only way to establish a maximum limitation. If the appropriation includes a specific amount for a particular object (such as ‘For Cuban cigars, $100’), then the appropriation is a maximum which may not be exceeded”). The appropriation in Cherokee Nation took a similar form. See, e.g., 108Stat. 2527–2528 (“For expenses necessary to carry out . . . ISDA [and certain other enumerated Acts], $1,713,052,000”). There is no ba- sis, therefore, for distinguishing the class of appropria- tion in those cases from this one. In each case, the agency remained free to allocate funds among multiple contractors, so long as the contracts served the purpose Congress identified. This result does not leave the “not to exceed” language in Congress’ appropriation without legal effect. To the contrary, it prevents the Secretary from reprogramming other funds to pay contract support costs—thereby protecting funds that Congress envisioned for other BIA programs, including tribes that choose not to enter ISDA contracts. But when an agency makes competing contractual commitments with legally available funds and then fails to pay, it is the Government that must bear the fiscal consequences, not the contractor. B The dissent attempts to distinguish this case from Cherokee Nation and Ferris on different grounds, relying on §450j–1(b)’s proviso that “the Secretary is not required to reduce funding for programs, projects, or activities serv- ing a tribe to make funds available to another tribe.” In the dissent’s view, that clause establishes that each dol- lar allocated by the Secretary reduces the amount of appropriations legally available to pay other contractors. In effect, the dissent understands §450j–1(b) to make the legal availability of appropriations turn on the Secretary’s expenditures rather than the sum allocated by Congress. That interpretation, which is inconsistent with ordinary principles of Government contracting law, is improbable. We have explained that Congress ordinarily controls the availability of appropriations; the agency controls whether to make funds from that appropriation available to pay a contractor. See Cherokee Nation, 543 U. S., at 642–643. The agency’s allocation choices do not affect the Government’s liability in the event of an underpayment. See id., at 641 (when an “ ‘unrestricted appropriation is sufficient to fund the contract, the contractor is entitled to payment even if the agency has allocated the funds to another purpose’ ”).[6] In Cherokee Nation, we found those ordinary principles generally applicable to ISDA. See id., at 637–646. We also found no evidence that Congress intended that “the tribe should bear the risk that a total lump-sum appropriation (though sufficient to cover its own contracts) will not prove sufficient to pay all similar contracts.” Id., at 638 (citing Brief for Federal Parties 23–25). The dissent’s reading, by contrast, would impose precisely that regime. See post, at 4–5. The better reading of §450j–1(b) accords with ordinary Government contracting principles. As we explained, su- pra, at 9, the clause underscores the Secretary’s discre- tion to allocate funds among tribes, but does not alter the Government’s legal obligation when the agency fails to pay. That reading gives full effect to the clause’s text, which addresses the “amount of funds provided,” and specifies that the Secretary is not required to reduce funding for one tribe to make “funds available” to another. 450j–1(b). Indeed, even the Government acknowledges the clause governs the Secretary’s discretion to distribute funds. See Brief for Petitioners 52 (pursuant to §450j–1(b), the Secretary was not obligated to pay tribes’ “contract support costs on a first-come, first-served basis, but had the authority to distribute the available money among all tribal contractors in an equitable fashion”). At minimum, the fact that we, the court below, the Government, and the Tribes do not share the dissent’s reading of §450j–1(b) is strong evidence that its inter- pretation is not, as it claims, “unambiguous[ly]” correct. Post, at 7 (opinion of Roberts, C. J.). Because ISDA is con- strued in favor of tribes, that conclusion is fatal to the dissent. C The remaining counterarguments are unpersuasive. First, the Government suggests that today’s holding could cause the Secretary to violate the Anti-Deficiency Act, which prevents federal officers from “mak[ing] or authoriz[ing] an expenditure or obligation exceeding an amount available in an appropriation.” 31 U. S. C. §1341(a)(1)(A). But a predecessor version of that Act was in place when Ferris and Dougherty were decided, see GAO Redbook, pp. 6–9 to 6–10, and the Government did not prevail there. As Dougherty explained, the Anti-Deficiency Act’s requirements “apply to the official, but they do not affect the rights in this court of the citizen honestly contracting with the Government.” 18 Ct. Cl., at 503; see also Ferris, 27 Ct. Cl., at 546 (“An appropriation per se merely imposes limitations upon the Government’s own agents; . . . but its insufficiency does not pay the Government’s debts, nor cancel its obligations”).[7] Second, the Government argues that Congress could not have intended for respondents to recover from the Judgment Fund, 31 U. S. C. §1304, because that would allow the Tribes to circumvent Congress’ intent to cap total expenditures for contract support costs.[8] That contention is puzzling. Congress expressly provided in ISDA that tribal contractors were entitled to sue for “money dam- ages” under the Contract Disputes Act upon the Government’s failure to pay, 25 U. S. C. §§450m–1(a), (d), and judgments against the Government under that Act are payable from the Judgment Fund, 41 U. S. C. §7108(a).[9] Indeed, we cited the Contract Disputes Act, Judgment Fund, and Anti-Deficiency Act in Cherokee Nation, explaining that if the Government commits its appropriations in a manner that leaves contractual obligations unfulfilled, “the contractor [is] free to pursue appropriate legal remedies arising because the Government broke its contractual promise.” 543 U. S., at 642. Third, the Government invokes cases in which courts have rejected contractors’ attempts to recover for amounts beyond the maximum appropriated by Congress for a particular purpose. See, e.g., Sutton v. United States, 256 U.S. 575 (1921). In Sutton, for instance, Congress made a specific line-item appropriation of $23,000 for the completion of a particular project. Id., at 577. We held that the sole contractor engaged to complete that project could not recover more than that amount for his work. The Ferris and Sutton lines of cases are distinguishable, however. GAO Redbook, p. 6–18. “[I]t is settled that contractors paid from a general appropriation are not barred from recovering for breach of contract even though the appropriation is exhausted,” but that “under a specific line-item appropriation, the answer is different.” Ibid.[10] The different results “follo[w] logically from the old maxim that ignorance of the law is no excuse.” Ibid. “If Congress appropriates a specific dollar amount for a particular contract, that amount is specified in the appropriation act and the contractor is deemed to know it.” Ibid. This case is far different. Hundreds of tribes entered into thousands of independent contracts, each for amounts well within the lump sum appropriated by Congress to pay contract support costs. Here, where each Tribe’s “contract is but one activity under a larger appropriation, it is not reasonable to expect [each] contractor to know how much of that appropriation remain[ed] available for it at any given time.” Ibid.; see also Ferris, 27 Ct. Cl., at 546. Finally, the Government argues that legislative history suggests that Congress approved of the distribution of available funds on a uniform, pro rata basis. But “a fundamental principle of appropriations law is that where Congress merely appropriates lump-sum amounts without statutorily restricting what can be done with those funds, a clear inference arises that it does not intend to impose legally binding restrictions.” Lincoln, 508 U. S., at 192 (internal quotation marks omitted). “[I]ndicia in committee reports and other legislative history as to how the funds should or are expected to be spent do not establish any legal requirements on the agency.” Ibid. (internal quotation marks omitted). An agency’s discretion to spend appropriated funds is cabined only by the “text of the appropriation,” not by Congress’ expectations of how the funds will be spent, as might be reflected by legislative history. Int’l Union, UAW, 746 F. 2d, at 860–861. That principle also reflects the same ideas underlying Ferris. If a contractor’s right to payment varied based on a future court’s uncertain interpretation of legislative history, it would increase the Government’s cost of contracting. Cf. Cherokee Nation, 543 U. S., at 644. That long-run expense would likely far exceed whatever money might be saved in any individual case. IV As the Government points out, the state of affairs resulting in this case is the product of two congressional decisions which the BIA has found difficult to reconcile. On the one hand, Congress obligated the Secretary to accept every qualifying ISDA contract, which includes a promise of “full” funding for all contract support costs. On the other, Congress appropriated insufficient funds to pay in full each tribal contractor. The Government’s frustration is understandable, but the dilemma’s resolution is the responsibility of Congress. Congress is not short of options. For instance, it could reduce the Government’s financial obligation by amending ISDA to remove the statutory mandate compelling the BIA to enter into self-determination contracts, or by giving the BIA flexibility to pay less than the full amount of contract support costs. It could also pass a moratorium on the formation of new self-determination contracts, as it has done before. See §328, 112Stat. 2681–291 to 292. Or Congress could elect to make line-item appropriations, allocating funds to cover tribes’ contract support costs on a contractor-by-contractor basis. On the other hand, Con- gress could appropriate sufficient funds to the BIA to meet the tribes’ total contract support cost needs. Indeed, there is some evidence that Congress may do just that. See H. R. Rep. No. 112–151, p. 42 (2011) (“The Committee believes that the Bureau should pay all contract support costs for which it has contractually agreed and directs the Bureau to include the full cost of the contract support obligations in its fiscal year 2013 budget submission”). The desirability of these options is not for us to say. We make clear only that Congress has ample means at hand to resolve the situation underlying the Tribes’ suit. Any one of the options above could also promote transparency about the Government’s fiscal obligations with respect to ISDA’s directive that contract support costs be paid in full. For the period in question, however, it is the Govern- ment—not the Tribes—that must bear the consequences of Congress’ decision to mandate that the Government enter into binding contracts for which its appropriation was sufficient to pay any individual tribal contractor, but “insufficient to pay all the contracts the agency has made.” Cherokee Nation, 543 U. S., at 637. The judgment of the Court of Appeals is affirmed. It is so ordered. Notes 1 As defined by ISDA, contract support costs “shall consist of an amount for the reasonable costs for activities which must be carried on by a tribal organization as a contractor to ensure compliance with the terms of the contract and prudent management, but which . . . (A) normally are not carried on by the respective Secretary in his direct operation of the program; or (B) are provided by the Secretary in support of the contracted program from resources other than those under contract.” §450j–1(a)(2). Such costs include overhead administrative costs, as well as expenses such as federally mandated audits and liability insurance. See Cherokee Nation of Okla., 543 U. S., at 635. 2 Compare 644 F.3d 1054 (case below), with Arctic Slope Native Assn., Ltd. v. Sebelius, 629 F.3d 1296 (CA Fed. 2010) (no liability to pay total contract support costs beyond cap in appropriations Act). 3 In Ferris, for instance, Congress appropriated $45,000 for the improvement of the Delaware River below Bridesburg, Pennsylvania. Act of Mar. 3, 1879, ch. 181, 20Stat. 364. The Government contracted with Ferris for $37,000 to dredge the river. Halfway through Ferris’ performance of his contract, the United States Army Corps of Engineers ran out of money to pay Ferris, having used $17,000 of the appropriation to pay for other improvements. Nonetheless, the Court of Claims found that Ferris could recover for the balance of his contract. As the court explained, the appropriation “merely impose[d] limitations upon the Government’s own agents; . . . its insufficiency [did] not pay the Government’s debts, nor cancel its obligations, nor defeat the rights of other parties.” 27 Ct. Cl., at 546; see also Dougherty, 18 Ct. Cl., at 503 (rejecting Government’s argument that a contractor could not recover upon similar facts because the “appropriation had, at the time of the purchase, been covered by other contracts”). 4 Indeed, the Indian Health Service once allocated its appropriations for new ISDA contracts on a first-come, first-serve basis. See Dept. of Health and Human Services, Indian Self-Determination Memorandum No. 92–2, p. 4 (Feb. 27, 1992). 5 The Government’s reliance on this statutory language is particularly curious because it suggests it is superfluous. See Brief for Petitioners 30–31 (it is “unnecessary” to specify that contracts are “subject to the availability of appropriations” (internal quotation marks omitted));see also Reply Brief for Petitioners 7 (“[A]ll government contracts are contingent upon the appropriations provided by Congress”). 6 The dissent’s view notwithstanding, it is beyond question that Congress appropriated sufficient unrestricted funds to pay any contractor in full. The dissent’s real argument is that §450j–1(b) reverses the applicability of the Ferris rule to ISDA, so that the Secretary’s allocation of funds to one contractor reduces the legal availability of funds to others. See post, at 4 (opinion of Roberts, C. J.) (“that the Secretary could have allocated the funds to [a] tribe is irrelevant. What matters is what the Secretary does, and once he allocates the funds to one tribe, they are not available to another”). We are not persuaded that §450j–1(b) was intended to enact that radical departure from ordinary Government contracting principles. Indeed, Congress has spoken clearly and directly when limiting the Government’s total contractual liability to an amount appropriated in similar schemes; that it did not do so here further counsels against the dissent’s reading. See, e.g., 25 U. S. C. §2008(j)(2) (“[i]f the total amount of funds necessary to provide grants to tribes . . . for a fiscal year exceeds the amount of funds appropriated . . . , the Secretary shall reduce the amount of each grant [pro rata]”). 7 We have some doubt whether a Government employee would violate the Anti-Deficiency Act by obeying an express statutory command to enter a contract, as was the case here. But we need not decide the question, for this case concerns only the contractual rights of tribal contractors, not the consequences of entering into such contracts for agency employees. 8 The Judgment Fund is a “permanent, indefinite appropriation” enacted by Congress to pay final judgments against the United States when, inter alia, “[p]ayment may not legally be made from any other source of funds.” 31 CFR §256.1 (2011). 9 For that reason, the Government’s reliance on Office of Personnel Management v. Richmond, 496 U.S. 414 (1990), is misplaced. In Richmond, we held that the Appropriations Clause does not permit plaintiffs to recover money for Government-caused injuries for which Congress “appropriated no money.” Id., at 424. Richmond, however, indicated that the Appropriations Clause is no bar to recovery in a case like this one, in which “the express terms of a specific statute” establish “a substantive right to compensation” from the Judgment Fund. Id.,at 432. 10 Of course, “[t]he terms ‘lump-sum’ and ‘line-item’ are relative concepts.” GAO Redbook, p. 6–165. For example, an appropriation for building two ships “could be viewed as a line-item appropriation in relation to the broader ‘Shipbuilding and Conversion’ category, but it was also a lump-sum appropriation in relation to the two specific vessels included.” Ibid. So long as a contractor does not seek payment beyond the amount Congress made legally available for a given purpose, “[t]his factual distinction does not affect the legal principle.” Ibid. See also In re Newport News Shipbuilding & Dry Dock Co., 55 Comp. Gen. 812 (1976).

566.US.2011_10-7387

When petitioner Setser was indicted in a Texas court on drug charges, the State also moved to revoke the probation term that he was then serving for another drug offense. At about the same time, Setser pleaded guilty to federal drug charges. The Federal District Court imposed a 151-month sentence to run consecutively to any state sentence imposed for the probation violation, but concurrently with any state sentence imposed on the new drug charge. While Setser’s federal appeal was pending, the state court sentenced him to 5 years for the probation violation and 10 years for the drug charge, but ordered the sentences to be served concurrently. The Fifth Circuit affirmed the federal sentence, holding that the District Court had authority to order a sentence consecutive to an anticipated state sentence, and that Setser’s sentence was reasonable, even if the state court’s decision made it unclear exactly how to administer it. Held: 1. The District Court had discretion to order that Setser’s federal sentence run consecutively to his anticipated state sentence for the probation violation. Pp. 2–12. (a) Judges have traditionally had broad discretion in selecting whether the sentences they impose will run concurrently or consecutively with respect to other sentences that they impose, or that have been imposed in other proceedings, including state proceedings, see Oregon v. Ice, 555 U.S. 160, 168–169. The statutory text and structure do not foreclose a district court’s exercise of this discretion with respect to anticipated state sentences. The Sentencing Reform Act of 1984 addresses the concurrent-vs.-consecutive decision, but not the situation here, since the District Court did not impose “multiple terms of imprisonment . . . at the same time,” and Setser was not “already subject to” the state sentences at issue, 18 U. S. C. §3584(a). This does not mean, as Setser and the Government claim, that the District Court lacked authority to act as it did and that the Bureau of Prisons is to make the concurrent-vs.-consecutive decision after the federal sentence has been imposed. Section 3621(b), from which the Bureau claims to derive this authority, says nothing about concurrent or consecutive sentences. And it is more natural to read §3584(a) as leaving room for the exercise of judicial discretion in situations not covered than it is to read §3621(b) as giving the Bureau what amounts to sentencing authority. Setser’s arguments to the contrary are unpersuasive. Pp. 2–8. (b) None of the other objections raised by Setser and the Government requires a different result. Pp. 8–12. 2. The state court’s subsequent decision to make the state sentences run concurrently does not establish that the Federal District Court imposed an unreasonable sentence. The difficulty here arises not from the federal-court sentence—which is to run concurrently with one state sentence and consecutively with another—but from the state court’s decision. Deciding which of the District Court’s dispositions should prevail under these circumstances is a problem, but it does not show the District Court’s sentence to be unlawful. The reasonableness standard for reviewing federal sentences asks whether the district court abused its discretion, see Gall v. United States, 552 U.S. 38, 46, but Setser identifies no flaw in the District Court’s decisionmaking process, nor anything available at the time of sentencing that the court failed to consider. Where late-onset facts make it difficult, or even impossible, to implement the sentence, the Bureau of Prisons may determine, in the first instance, how long the District Court’s sentence authorizes it to continue Setser’s confinement, subject to the potential for judicial review. Pp. 12–14. 607 F.3d 128, affirmed. Scalia, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Alito, Sotomayor, and Kagan, JJ., joined. Breyer, J., filed a dissenting opinion, in which Kennedy and Ginsburg, JJ., joined.

We consider whether a district court, in sentencing a de- fendant for a federal offense, has authority to order that the federal sentence be consecutive to an anticipated state sentence that has not yet been imposed. I When officers of the Lubbock Police Department ar- rested petitioner Monroe Setser for possessing methamphetamine, he was already serving a 5-year term of probation imposed by a Texas court for another drug offense. Setser was indicted in state court for possession with intent to deliver a controlled substance, and the State also moved to revoke his term of probation. As often happens in drug cases, the federal authorities also got involved. A federal grand jury indicted Setser for possessing with intent to distribute 50 grams or more of methamphetamine, 21 U. S. C. §841(a)(1), (b)(1)(A)(viii), and he pleaded guilty. Before the federal sentencing hearing, a probation officer calculated the applicable Guidelines range to be 121 to 151 months’ imprisonment. Citing precedent from the United States Court of Appeals for the Fifth Circuit, United States v. Brown, 920 F.2d 1212 (1991) (per curiam), he indicated that the District Court had discretion to make Setser’s sentence either concurrent with or consecutive to any sentence anticipated in the separate state-court pro- ceedings. Setser objected, arguing that the District Court lacked such authority. The court nevertheless made the sentence of 151 months that it imposed consecutive to any state sentence imposed for probation violation, but concurrent with any state sentence imposed on the new drug charge. Setser appealed. While Setser’s appeal was pending, the state court sentenced him to a prison term of 5 years for probation violation and 10 years on the new drug charge. It ordered that these sentences be served concurrently. Setser then made before the Court of Appeals, in addition to the argument that the District Court had no authority to order a consecutive sentence, the argument that his federal sentence was unreasonable because it was impossible to implement in light of the concurrent state sentences. The Court of Appeals for the Fifth Circuit affirmed. 607 F.3d 128 (2010). Following its earlier Brown decision, the court held that the District Court did have authority to order a consecutive sentence. 607 F. 3d, at 131–132. It also held that Setser’s sentence was reasonable, even if it was “ ‘partially foiled’ ” by the state court’s decision. Id., at 132–133. We granted certiorari, 564 U. S. ___ (2011), and appointed an amicus curiae to brief and argue this case in support of the judgment below, 564 U. S. ___ (2011). II Before proceeding further, it is important to be clear about what is at issue. Setser does not contend that his federal sentence must run concurrently with both state sentences imposed after his federal sentencing hearing. He acknowledges that someone must answer “the consecutive versus concurrent question,” Brief for Petitioner 27, and decide how the state and federal sentences will fit together. The issue here is who will make that decision, which in turn determines when that decision is made. One possible answer, and the one the Fifth Circuit gave, is that the decision belongs to the Federal District Court at the federal sentencing hearing. The concurrent-vs.-consecutive decision has been addressed by §212(a) of the Sentencing Reform Act of 1984, 18 U. S. C. §3584, reproduced in full as Appendix A, infra. The first subsection of that provision, which says when concurrent and consecutive sentences may be imposed, and specifies which of those dispositions will be assumed in absence of indication by the sentencing judge, does not cover the situation here. It addresses only “multiple terms of imprisonment . . . imposed . . . at the same time” and “a term of imprisonment . . . imposed on a defendant who is already subject to an undischarged term of imprisonment.” §3584(a). Here the state sentence is not imposed at the same time as the federal sentence, and the defendant was not already subject to that state sentence. Setser, supported by the Government, argues that, be- cause §3584(a) does not cover this situation, the District Court lacked authority to act as it did; and that the concurrent-vs.-consecutive decision is therefore to be made by the Bureau of Prisons at any time after the federal sen- tence has been imposed. The Bureau of Prisons is said to derive this authority from 18 U. S. C. §3621(b) (2006 ed. and Supp. IV), reproduced in full as Appendix B, infra. On its face, this provision says nothing about concurrent or consecutive sentences, but the Government explains its position as follows: Section 3621(b) gives the Bureau the authority to order that a prisoner serve his federal sentence in any suitable prison facility “whether maintained by the Federal Government or otherwise.” The Bureau may therefore order that a prisoner serve his federal sentence in a state prison. Thus, when a person subject to a federal sentence is serving a state sentence, the Bureau may designate the state prison as the place of impris- onment for the federal sentence—effectively making the two sentences concurrent—or decline to do so—effec- tively making them consecutive.[1] Based on §§3584(a) and 3621(b), Setser and the Government argue that the concurrent-vs.-consecutive decision, under the circumstances presented here, is committed exclusively to the Bureau of Prisons. It is fundamental that we construe statutes governing the jurisdiction of the federal courts in light of “the common-law background against which the statutes . . . were enacted,” New Orleans Public Service, Inc. v. Council of City of New Orleans, 491 U.S. 350, 359 (1989), and the same approach is appropriate here, where the issue concerns a matter of discretion traditionally committed to the Judiciary. Judges have long been understood to have discretion to select whether the sentences they impose will run concurrently or consecutively with respect to other sentences that they impose, or that have been imposed in other proceedings, including state proceedings. See Oregon v. Ice, 555 U.S. 160, 168–169 (2009). And a large majority of the federal appellate courts addressing the question have recognized a similar authority in the context here, where a federal judge anticipates a state sentence that has not yet been imposed. See Salley v. United States, 786 F.2d 546, 547 (CA2 1986); Anderson v. United States, 405 F.2d 492, 493 (CA10 1969) (per curiam); United States ex rel. Lester v. Parker, 404 F.2d 40, 41–42 (CA3 1968) (per curiam); United States v. Kanton, 362 F.2d 178, 179–180 (CA7 1966) (per curiam); but see United States v. Eastman, 758 F.2d 1315, 1317 (CA9 1985)[2]. We find nothing in the Sentencing Reform Act, or in any other provision of law, to show that Congress foreclosed the exercise of district courts’ sentencing discretion in these circumstances. Setser’s main contention is that §3584(a) has this effect. But that provision cannot sustain the weight that Setser asks it to bear. In essence, he reads the first sentence in §3584(a) to say that “terms [of imprisonment] may run concurrently or consecutively” only “[i]f multiple terms of imprisonment are imposed . . . at the same time, or if a term of imprisonment is imposed on a defendant who is already subject to an undischarged term of imprisonment.” Since the District Court was not imposing the state sentence and since it was not already imposed, the sentence could not be ordered to run consecutively. But if the text is exclusive—if the addition of only is correct—the provision forbids not only the imposition of consecutive sentences, but the imposition of concurrent ones as well. And yet, as Setser acknowledges, it must be one or the other; someone must decide the issue. Setser’s response is that, read in context, the sentence speaks only to district courts. Under the circumstances at issue here, he says, the federal and state sentences still might run either concurrently or consecutively, but just not at the discretion of the District Court. That is an odd parsing of the text, which makes no distinction between the district court and the Bureau of Prisons. The placement of §3584 does indeed suggest that it is directed at district courts—but that is likely because Congress contemplated that only district courts would have the authority to make the concurrent-vs.-consecutive decision, not because Congress meant to leave the Bureau unfettered. Indeed, the Bureau already follows the other directives in §3584(a). See Brief for United States 35. For example, if the district court imposes multiple terms of imprisonment at the same time, but fails to address the concurrent-vs.-consecutive issue, the terms “run concurrently,” §3584(a), and the Bureau is not free to use its “place of imprisonment” authority to achieve a different result.[3] The Latin maxim on which Setser relies—expressio unius est exclusio alterius—might have application here if the provision in question were a conferral of authority on district courts. Giving sentencing authority in only specified circumstances could be said to imply that it is withheld in other circumstances. Section 3584, however, is framed not as a conferral of authority but as a limitation of authority that already exists (and a specification of what will be assumed when the exercise of that authority is ambiguous). It reads not “District courts shall have authority to impose multiple terms of imprisonment on a defendant at the same time, etc.” but rather “If multiple terms of imprisonment are imposed on a defendant at the same time, [etc.]”—quite clearly assuming that such au- thority already exists. The mere acknowledgment of the existence of certain pre-existing authority (and regulation of that authority) in no way implies a repeal of other pre-existing authority. And that is especially true when there is an obvious reason for selecting the instances of pre-existing authority that are addressed—to wit, that they are the examples of sentencing discretion most frequently encountered. Moreover, expressio unius est exclusio alterius is a double-edged sword. Setser thinks it suggests that, because §3584(a) recognizes judicial discretion in scenario A and scenario B, there is no such discretion in scenario C. But the same maxim shows much more convincingly why §3621(b) cannot be read to give the Bureau of Prisons exclusive authority to make the sort of decision committed to the district court in §3584(a). When §3584(a) specific- ally addresses decisions about concurrent and consecutive sentences, and makes no mention of the Bureau’s role in the process, the implication is that no such role exists. And that conclusion is reinforced by application of the same maxim (properly, in this instance) to §3621(b)—which is a conferral of authority on the Bureau of Prisons, but does not confer authority to choose between concurrent and consecutive sentences. Put to the choice, we believe it is much more natural to read §3584(a) as not containing an implied “only,” leaving room for the exercise of judicial discretion in the situations not covered, than it is to read §3621(b) as giving the Bureau of Prisons what amounts to sentencing authority. III None of the other objections to this approach raised by Setser and the Government require a different result. Our decision today follows the interpretive rule they invoke, that we must “give effect . . . to every clause and word” of the Act. United States v. Menasche, 348 U.S. 528, 538–539 (1955) (internal quotation marks omitted). The first sentence in §3584(a) addresses the most common situations in which the decision between concurrent and consecutive sentences must be made: where two sentences are imposed at the same time, and where a sentence is imposed subsequent to a prior sentence that has not yet been fully served. It says that the district court has discretion whether to make the sentences concurrent or consecutive, except that it may not make consecutive a sentence for “an attempt” and a sentence for an “offense that was the sole objective of the attempt.” And the last two sentences of §3584(a) say what will be assumed in those two common situations if the court does not specify that the sentence is concurrent or consecutive. Giving those dispositions full effect does not demand that we regard them as eliminating sentencing discretion in other situations. Setser and the Government both suggest that, because §3584(b) directs courts to consider the sentencing factors in §3553(a) in making these decisions, and because some of those factors will be difficult to apply with respect to anticipated sentences, the Act cannot be read to allow judicial discretion in these circumstances. One cannot be sure that the sentence imposed is “sufficient, but not greater than necessary,” §3553(a), the argument goes, if one does not know how long it will actually be. But the district judge faces the same uncertainty if the concurrent-vs.-consecutive decision is left for later resolution by the Bureau of Prisons; he does not know, for example, whether the 5-year sentence he imposes will be an actual five years or will be simply swallowed within another sentence. To be sure, the Bureau of Prisons, if it waits to decide the matter until after the state court has imposed its sentence, will know for sure what sentences it is dealing with. But the Bureau is not charged with applying §3553(a). The factors that guide the agency’s “place of imprisonment” decision do include “the nature and circumstances of the offense” and “the history and characteristics of the pris- oner,” §3621(b)(2), (b)(3) (2006 ed.)—factors that are, to be sure, relevant to sentencing but also relevant to selection of the place of confinement; but they also include factors that make little, if any, sense in the sentencing context, such as “the resources of the facility contemplated” and whether the state facility “meets minimum standards of health and habitability,” §3621(b), (b)(1). (These factors confirm our view that §3621 is not a sentencing provision but a place-of-confinement provision.) It is much more natural for a judge to apply the §3553(a) factors in making all concurrent-vs.-consecutive decisions, than it is for some such decisions to be made by a judge applying §3553(a) factors and others by the Bureau of Prisons applying §3621(b) factors. The final objection is that principles of federalism and good policy do not allow a district court to make the concurrent-vs.-consecutive decision when it does not have before it all of the information about the anticipated state sentence. As for principles of federalism, it seems to us they cut in precisely the opposite direction. In our American system of dual sovereignty, each sovereign—whether the Federal Government or a State—is responsible for “the administration of [its own] criminal justice syste[m].” Ice, 555 U. S., at 170. If a prisoner like Setser starts in state custody, serves his state sentence, and then moves to federal custody, it will always be the Federal Government—whether the district court or the Bureau of Prisons—that decides whether he will receive credit for the time served in state custody. And if he serves his federal sentence first, the State will decide whether to give him credit against his state sentences without being bound by what the district court or the Bureau said on the matter. Given this framework, it is always more respectful of the State’s sovereignty for the district court to make its decision up front rather than for the Bureau of Prisons to make the decision after the state court has acted. That way, the state court has all of the information before it when it acts.[4] The Government’s position does not promote the States’ interest—just the interests of the Bureau of Prisons. As for good policy: The basic claim of Setser, the Government, and the dissent is that when it comes to sentencing, later is always better because the decisionmaker has more information. See, e.g., post, at 7 (“[A] sentencing judge typically needs detailed information when constructing a multiple-count or multiple-conviction Guideline sentence”). That is undoubtedly true, but when that desideratum is applied to the statutory structure before us here it is overwhelmed by text, by our tradition of judicial sentencing,[5] and by the accompanying desideratum that sentencing not be left to employees of the same Department of Justice that conducts the prosecution.[6] Moreover, when the district court’s failure to “anticipat[e] developments that take place after the first sentencing,” Brief for United States 29, produces unfairness to the defendant, the Act provides a mechanism for relief. Section 3582(c)(1)(A) provides that a district court, “upon motion of the Director of the Bureau of Prisons, may reduce the term of imprisonment . . . after considering the factors set forth in section 3553(a) to the extent that they are applicable, if it finds that . . . extraordinary and compelling reasons warrant such a reduction [or that the defendant meets other criteria for relief].” IV Setser argues that, even if the District Court’s consecutive order was consistent with §3584(a), it made his sentence impossible to implement and therefore unreasonable under the Act, see United States v. Booker, 543 U.S. 220, 261–262 (2005),[7] in light of the State’s decision to make his sentences concurrent. We think not. There is nothing unreasonable—let alone inherently impossible—about the sentence itself. Setser is ordered to serve a 151-month term in federal custody, and that sentence should run concurrently with one state sentence and consecutively with another. The difficulty arises not from the sentence, but from the state court’s decision to make both state sentences concurrent. Which of the District Court’s dispositions should prevail: that his federal sentence run consecutively to the state sentence on the parole revocation charge, or that his federal sentence run concurrently with the state sentence on the new drug charge? If the federal sentence is added to the state sentence it will not be concurrent with the new drug charge, and if it is merged in the state sentence it will not be consecutive to the parole revocation charge. This is indeed a problem, but not, we think, one that shows the District Court’s sentence to be unlawful. The reasonableness standard we apply in reviewing federal sentences asks whether the district court abused its discretion. See Gall v. United States, 552 U.S. 38, 46 (2007). Setser identifies no flaw in the District Court’s decisionmaking process, nor anything available at the time of sentencing that the District Court failed to consider. That a sentence is thwarted does not mean that it was unreasonable. If a district court ordered, as a term of supervised release, that a defendant maintain a steady job, but a subsequent disability rendered gainful employment infeasible, we doubt that one would call the original sentence an abuse of discretion. There will often be late-onset facts that materially alter a prisoner’s position and that make it difficult, or even impossible, to implement his sentence. This is where the Bureau of Prisons comes in—which ultimately has to determine how long the District Court’s sentence authorizes it to continue Setser’s confinement. Setser is free to urge the Bureau to credit his time served in state court based on the District Court’s judgment that the federal sentence run concurrently with the state sentence for the new drug charges. If the Bureau initially declines to do so, he may raise his claim through the Bureau’s Administrative Remedy Program. See 28 CFR §542.10 et seq. (2011). And if that does not work, he may seek a writ of habeas corpus. See 28 U. S. C. §2241. We express no view on whether those proceedings would be successful. * * * Because it was within the District Court’s discretion to order that Setser’s sentence run consecutively to his anticipated state sentence in the probation revocation proceeding; and because the state court’s subsequent decision to make that sentence concurrent with its other sentence does not establish that the District Court abused its discretion by imposing an unreasonable sentence; we affirm the judgment of the Court of Appeals. It is so ordered. APPENDIXES A 18 U. S. C. §3584 “Multiple sentences of imprisonment “(a) Imposition of Concurrent or Consecutive Terms.—If multiple terms of imprisonment are imposed on a defendant at the same time, or if a term of imprisonment is imposed on a defendant who is already subject to an undischarged term of imprisonment, the terms may run concurrently or consecutively, except that the terms may not run consecutively for an attempt and for another offense that was the sole objective of the attempt. Multiple terms of imprisonment imposed at the same time run concurrently unless the court orders or the statute mandates that the terms are to run consecutively. Multiple terms of imprisonment imposed at different times run consecutively unless the court orders that the terms are to run concurrently. “(b) Factors to Be Considered in Imposing Concurrent or Consecutive Terms.—The court, in determining whether the terms imposed are to be ordered to run concurrently or consecutively, shall consider, as to each offense for which a term of imprisonment is being imposed, the factors set forth in section 3553(a). “(c) Treatment of Multiple Sentence as an Aggregate.—Multiple terms of imprisonment ordered to run consecutively or concurrently shall be treated for administrative purposes as a single, aggregate term of imprisonment.” B 18 U. S. C. §3621(b) (2006 ed. and Supp. IV) “Place of imprisonment.—The Bureau of Prisons shall designate the place of the prisoner’s imprisonment. The Bureau may designate any available penal or correctional facility that meets minimum standards of health and habitability established by the Bureau, whether maintained by the Federal Government or otherwise and whether within or without the judicial district in which the person was convicted, that the Bureau determines to be appropriate and suitable, considering— “(1) the resources of the facility contemplated; “(2) the nature and circumstances of the offense; “(3) the history and characteristics of the prisoner; “(4) any statement by the court that imposed the sentence— “(A) concerning the purposes for which the sentence to imprisonment was determined to be warranted; or “(B) recommending a type of penal or correctional facility as appropriate; and “(5) any pertinent policy statement issued by the Sentencing Commission pursuant to section 994(a)(2) of title 28. “In designating the place of imprisonment or making transfers under this subsection, there shall be no favoritism given to prisoners of high social or economic status. The Bureau may at any time, having regard for the same matters, direct the transfer of a prisoner from one penal or correctional facility to another. The Bureau shall make available appropriate substance abuse treatment for each prisoner the Bureau determines has a treatable condition of substance addiction or abuse. Any order, recommendation, or request by a sentencing court that a convicted person serve a term of imprisonment in a community corrections facility shall have no binding effect on the authority of the Bureau under this section to determine or change the place of imprisonment of that person.” Notes 1 The Bureau of Prisons sometimes makes this designation while the prisoner is in state custody and sometimes makes a nunc pro tunc designation once the prisoner enters federal custody. 2 The dissent is incorrect to say, post, at 7–8 (opinion of Breyer, J.), that only the Second Circuit, in Salley held to that effect. So did the Seventh Circuit in Kanton and the Tenth Circuit in Anderson. The dissent says that Anderson addressed only the question “whether a federal sentence runs from the date of its imposition or from the date of entry into federal custody,” post, at 7–8. That is true enough (and it is true of Kanton as well); but answering that question in a manner that upheld the consecutive federal sentence (i.e., it runs from the date of entry into federal custody) necessarily upheld the sentencing court’s authority to impose the consecutive federal sentence. In fact, Ander-son confronted and specifically rejected the defendant’s argument that “ ‘[n]o court has the authority to impose a sentence consecutive to something that does not exist,’ ” 405 F. 2d, at 493. And, finally, so did the Third Circuit in Lester. The dissent says that Lester addressed only the question “whether a sentence was insufficiently certain for pur-poses of due process,” post, at 8. But that was the defendant’s princi-pal reason (as it appears also to be the dissent’s principal reason) for asserting that the sentencing court had no authority to impose a consecutive sentence. And the Third Circuit rejected not only that reason but “[o]ther arguments advanced by [the defendant] ” attacking the consecutive sentence, 404 F. 2d, at 42. The only contrary federal appellate decision rendered before the Sentencing Reform Act took effect relied upon 18 U. S. C. §4082 (1982 ed.) (the predecessor of §3621) and §3568 (1982 ed.) (repealed by 98Stat. 1987), which provided that a federal sentence “shall commence to run from the date on which such person is received” into federal custody. See United States v. Eastman, 758 F.2d 1315, 1317 (CA9 1985). 3 The Government contends that the Bureau applies the default rules in §3584(a) “[a]s a matter of discretion” but is not “ ‘bound’ ” by that subsection. Reply Brief for United States 15, n. 5. We think it implausible that the effectiveness of those rules—of §3584(a)’s prescription, for example, that “[m]ultiple terms of imprisonment imposed at different times run consecutively unless the court orders that the terms are to run concurrently”—depends upon the “discretion” of the Bureau. 4 Setser notes that the text of §3584(a) does not distinguish between state and federal sentences. If a district court can enter a consecutive sentencing order in advance of an anticipated state sentence, he asks, what is to stop it from issuing such an order in advance of an antici-pated federal sentence? It could be argued that §3584(a) impliedly prohibits such an order because it gives that decision to the federal court that sentences the defendant when the other sentence is “already” imposed—and does not speak (of course) to what a state court must do when a sentence has already been imposed. It suffices to say, however, that this question is not before us. 5 To support its view that Congress authorized the Bureau to make concurrent-vs.-consecutive decisions, the dissent relies on the fact that the Executive long had what is effectively sentencing authority in its ability to grant or deny parole. That is a particularly curious power for the dissent to rely upon, inasmuch as most of the dissent discusses (in great detail) the Sentencing Reform Act, whose principal objective was to eliminate the Executive’s parole power. Curiouser still is the dissent’s invocation of the Guidelines system, which “tell[s] the sentencing judge how, through the use of partially concurrent and partially consecutive sentences, to build a total sentence that meets the Guidelines’ requirements.” Post, at 4. These “instructions,” ibid., do not cover yet-to-be-imposed sentences, the dissent says, because “the sentencing judge normally does not yet know enough about the behavior that underlies (or will underlie)” such a sentence. Post, at 5. That explains, perhaps, why the Guidelines’ “instructions” to judges do not cover them. But why do not the Guidelines “instruct” the Bureau of Prisons how to conduct its concurrent/consecutive sentencing? If the reason is (as we suspect) that the Sentencing Commission does not have, or does not believe it has, authority to “instruct” the Bureau of Prisons, the dissent’s entire argument based upon what it calls “the purposes and the mechanics of the SRA’s sentencing system,” post, at 6, falls apart. Yet-to-be-imposed sentences are not within the system at all, and we are simply left with the question whether judges or the Bureau of Prisonsis responsible for them. For the reasons we have given, we think it is judges. 6 Of course, a district court should exercise the power to impose anticipatory consecutive (or concurrent) sentences intelligently. In some situations, a district court may have inadequate information and may forbear, but in other situations, that will not be the case. 7 We have never had occasion to decide whether reasonableness review under Booker applies to a court’s decision that a federal sentence should run concurrently with or consecutively to another sentence. The Courts of Appeals, however, generally seem to agree that such review applies. See, e.g., United States v. Padilla, 618 F.3d 643, 647 (CA7 2010) (per curiam); United States v. Matera, 489 F.3d 115, 123–124 (CA2 2007). For purpose of the present case we assume, without deciding, that it does.

565.US.2011_10-8145

Petitioner Juan Smith was convicted of first-degree murder based on the testimony of a single eyewitness. During state postconviction relief proceedings, Smith obtained police files containing statements by the eyewitness contradicting his testimony. Smith argued that the prosecution’s failure to disclose those statements violated Brady v. Maryland, 373 U.S. 83. Brady held that due process bars a State from withholding evidence that is favorable to the defense and material to the defendant’s guilt or punishment. See id., at 87. The state trial court rejected Smith’s Brady claim, and the Louisiana Court of Appeal and Louisiana Supreme Court denied review. Held: Brady requires that Smith’s conviction be reversed. The State does not dispute that the eyewitness’s statements were favorable to Smith and that those statements were not disclosed to Smith. Under Brady, evidence is material if there is a “reasonable probability that, had the evidence been disclosed, the result of the proceeding would have been different.” Cone v. Bell, 556 U.S. 449, 469–470. A “reasonable probability” means that the likelihood of a different result is great enough to “undermine[ ] confidence in the outcome of the trial.” Kyles v. Whitley, 514 U.S. 419, 434. Evidence impeaching an eyewitness’s testimony may not be material if the State’s other evidence is strong enough to sustain confidence in the verdict. United States v. Agurs, 427 U.S. 97, 112–113, and n. 21. Here, however, the eyewitness’s testimony was the only evidence linking Smith to the crime, and the eyewitness’s undisclosed statements contradicted his testimony. The eyewitness’s statements were plainly material, and the State’s failure to disclose those statements to the defense thus violated Brady. Pp. 2–4. Reversed and remanded. Roberts, C. J., delivered the opinion of the Court, in which Scalia, Kennedy, Ginsburg, Breyer, Alito, Sotomayor, and Kagan, JJ., joined. Thomas, J., filed a dissenting opinion.

The State of Louisiana charged petitioner Juan Smith with killing five people during an armed robbery. At Smith’s trial a single witness, Larry Boatner, linked Smith to the crime. Boatner testified that he was socializing at a friend’s house when Smith and two other gunmen entered the home, demanded money and drugs, and shortly thereafter began shooting, resulting in the death of five of Boatner’s friends. In court Boatner identified Smith as the first gunman to come through the door. He claimed that he had been face to face with Smith during the initial moments of the robbery. No other witnesses and no physical evidence implicated Smith in the crime. The jury convicted Smith of five counts of first-degree murder. The Louisiana Court of Appeal affirmed Smith’s conviction. State v. Smith, 797 So. 2d 193 (2001). The Louisiana Supreme Court denied review, as did this Court. 2001–2416 (La. 9/13/02), 824 So. 2d 1189; 537 U.S. 1201 (2003) . Smith then sought postconviction relief in the state courts. As part of his effort, Smith obtained files from the police investigation of his case, including those of the lead investigator, Detective John Ronquillo. Ronquillo’s notes contain statements by Boatner that conflict with his testimony identifying Smith as a perpetrator. The notes from the night of the murder state that Boatner “could not . . . supply a description of the perpetrators other then [sic] they were black males.” App. 252–253. Ronquillo also made a handwritten account of a conversation he had with Boatner five days after the crime, in which Boatner said he “could not ID anyone because [he] couldn’t see faces” and “would not know them if [he] saw them.” Id., at 308. And Ronquillo’s typewritten report of that conversation states that Boatner told Ronquillo he “could not identify any of the perpetrators of the murder.” Id., at 259–260. Smith requested that his conviction be vacated, arguing, inter alia, that the prosecution’s failure to disclose Ronquillo’s notes violated this Court’s decision in Brady v. Maryland, 373 U.S. 83 (1963) . The state trial court rejected Smith’s Brady claim, and the Louisiana Court of Appeal and Louisiana Supreme Court denied review. We granted certiorari, 564 U. S. ___ (2011), and now reverse. Under Brady, the State violates a defendant’s right to due process if it withholds evidence that is favorable to the defense and material to the defendant’s guilt or punishment. See 373 U. S., at 87. The State does not dispute that Boatner’s statements in Ronquillo’s notes were fa- vorable to Smith and that those statements were not dis- closed to him. The sole question before us is thus whether Boatner’s statements were material to the determination of Smith’s guilt. We have explained that “evidence is ‘material’ within the meaning of Brady when there is a reasonable probability that, had the evidence been disclosed, the result of the proceeding would have been different.” Cone v. Bell, 556 U.S. 449 –470 (2009). A reasonable probability does not mean that the defendant “would more likely than not have received a different verdict with the evidence,” only that the likelihood of a different result is great enough to “undermine[] confidence in the outcome of the trial.” Kyles v. Whitley, 514 U.S. 419, 434 (1995) (internal quotation marks omitted). We have observed that evidence impeaching an eyewitness may not be material if the State’s other evidence is strong enough to sustain confidence in the verdict. See United States v. Agurs, 427 U.S. 97 –113, and n. 21 (1976). That is not the case here. Boatner’s testimony was the only evidence linking Smith to the crime. And Boatner’s undisclosed statements directly contradict his testimony: Boatner told the jury that he had “[n]o doubt” that Smith was the gunman he stood “face to face” with on the night of the crime, but Ronquillo’s notes show Boatner saying that he “could not ID anyone because [he] couldn’t see faces” and “would not know them if [he] saw them.” App. 196, 200, 308. Boatner’s undisclosed statements were plainly material. The State and the dissent advance various reasons why the jury might have discounted Boatner’s undisclosed statements. They stress, for example, that Boatner made other remarks on the night of the murder indicating that he could identify the first gunman to enter the house, but not the others. That merely leaves us to speculate about which of Boatner’s contradictory declarations the jury would have believed. The State also contends that Boatner’s statements made five days after the crime can be explained by fear of retaliation. Smith responds that the record contains no evidence of any such fear. Again, the State’s argument offers a reason that the jury could have disbelieved Boatner’s undisclosed statements, but gives us no confidence that it would have done so. The police files that Smith obtained in state postconviction proceedings contain other evidence that Smith contends is both favorable to him and material to the verdict. Because we hold that Boatner’s undisclosed statements alone suffice to undermine confidence in Smith’s conviction, we have no need to consider his arguments that the other undisclosed evidence also requires reversal under Brady. The judgment of the Orleans Parish Criminal District Court of Louisiana is reversed, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered.

567.US.2011_11-94

Petitioner Southern Union Company was convicted by a jury in federal court on one count of violating the Resource Conservation and Recovery Act of 1976 (RCRA) for having knowingly stored liquid mercury without a permit at a subsidiary’s facility “on or about September 19, 2002 to October 19, 2004.” Violations of the RCRA are punishable by, inter alia, a fine of not more than $50,000 for each day of violation. 42 U. S. C. §6928(d). At sentencing, the probation office calculated a maximum fine of $38.1 million, on the basis that Southern Union violated the RCRA for each of the 762 days from September 19, 2002, through October 19, 2004. Southern Union argued that imposing any fine greater than the 1-day penalty of $50,000 would be unconstitutional under Apprendi v. New Jersey, 530 U.S. 466, which holds that the Sixth Amendment’s jury-trial guarantee requires that any fact (other than the fact of a prior conviction) that increases the maximum punishment authorized for a particular crime be proved to a jury beyond a reasonable doubt. Southern Union contended that, based on the jury verdict and the District Court’s instructions, the only violation the jury necessarily found was for one day. The District Court held that Apprendi applies to criminal fines, but concluded from the “content and context of the verdict all together” that the jury found a 762-day violation. The court therefore set a maximum potential fine of $38.1 million, from which it imposed a fine of $6 million and a “community service obligation” of $12 million. On appeal, the First Circuit disagreed with the District Court that the jury necessarily found a violation of 762 days. But the First Circuit affirmed the sentence because it held that Apprendi does not apply to criminal fines. Held: The rule of Apprendi applies to the imposition of criminal fines. Pp. 3−16. (a) Apprendi’s rule is “rooted in longstanding common-law practice,” Cunningham v. California, 549 U.S. 270, 281, and preserves the “historic jury function” of “determining whether the prosecution has proved each element of an offense beyond a reasonable doubt,” Oregon v. Ice, 555 U.S. 160, 163. This Court has repeatedly affirmed Apprendi’s rule by applying it to a variety of sentencing schemes that allow judges to find facts that increase a defendant’s maximum authorized sentence. See Cunningham, 549 U. S., at 274−275; United States v. Booker, 543 U.S. 220, 226–227; Blakely v. Washington, 542 U.S. 296, 299–300; Ring v. Arizona, 536 U.S. 584, 588–589; Ap-prendi, 530 U. S., at 468–469. While the punishments at stake in these cases were imprisonment or a death sentence, there is no principled basis under Apprendi to treat criminal fines differently. Apprendi’s “core concern”—to reserve to the jury “the determination of facts that warrant punishment for a specific statutory offense,” Ice, 555 U. S., at 170—applies whether the sentence is a criminal fine or imprisonment or death. Criminal fines, like these other forms of punishment, are penalties inflicted by the sovereign for the commission of offenses. Fines were by far the most common form of noncapital punishment in colonial America and they continue to be frequently imposed today. And, the amount of a fine, like the maximum term of imprisonment or eligibility for the death penalty, is often determined by reference to particular facts. The Government argues that fines are less onerous than incarceration and the death sentence and therefore should be exempt from Apprendi. But where a fine is substantial enough to trigger the Sixth Amendment’s jury-trial guarantee, Apprendi applies in full. Pp. 3−8. (b) The “historical role of the jury at common law,” which informs the “scope of the constitutional jury right,” Ice, 555 U. S., at 170, supports applying Apprendi to criminal fines. To be sure, judges in the colonies and during the founding era had much discretion in determining whether to impose a fine and in what amount. But the exercise of such discretion is fully consistent with Apprendi, which permits courts to impose “judgment within the range prescribed by statute.” 530 U. S., at 481 (emphasis in original). The more salient question is what role the jury played in prosecutions for offenses that pegged the amount of a fine to the determination of specified facts. A review of both state and federal decisions discloses that the predominant practice was for such facts to be alleged in the indictment and proved to the jury. The rule that juries must determine facts that set a fine’s maximum amount is an application of the “two longstanding tenets of common-law criminal jurisprudence” on which Apprendi is based: first, “the ‘truth of every accusation’ against a defendant ‘should afterwards be confirmed by the unanimous suffrage of twelve of his equals and neighbours.’ ” Blakely, 542 U. S., at 301. And second, “ ‘an accusation which lacks any particular fact which the law makes essential to the punishment is . . . no accusation within the requirements of the common law, and is no accusation in reason.’ ” Ibid. Contrary to the Government’s contentions, neither United States v. Murphy, 16 Pet. 203, nor United States v. Tyler, 7 Cranch 285, overcomes the ample historical evidence that juries routinely found facts that set maximum criminal fines. Pp. 8−14. (c) The Government’s remaining arguments, echoed by the dissent, are unpersuasive. The Government claims that facts relevant to a fine’s amount typically quantify the harm caused by the defendant’s offense, and do not define a separate set of acts for punishment. The Government contends that only the latter determination implicates Apprendi’s concerns. But this argument rests on the rejected assumption that, in determining the maximum punishment for an offense, there is a constitutionally significant difference between a fact that is an “element” of the offense and one that is a “sentencing factor.” Further, the facts the District Court found in imposing a fine on Southern Union are not fairly characterized as merely quantifying the harm the company caused. The Government also argues that applying Apprendi to criminal fines will prevent States and the Federal Government from enact- ing statutes that calibrate the amount of a fine to a defendant’s culp-ability. But legislatures are free to enact such statutes, so long as the statutes are administered in conformance with the Sixth Amendment. Finally, the Government contends that requiring juries to determine facts related to fines will cause confusion, prejudice defendants, or be impractical. These policy arguments rehearse those made by the dissents in our prior Apprendi cases. They must be rejected because the rule the Government espouses is unconstitutional. In addition, because Apprendi is now more than a decade old, the reliance interests underlying the Government’s arguments are by this point attenuated. Pp. 14−16. 630 F.3d 17, reversed and remanded. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Thomas, Ginsburg, and Kagan, JJ., joined. Breyer, J., filed a dissenting opinion, in which Kennedy and Alito, JJ., joined.

The Sixth Amendment reserves to juries the determination of any fact, other than the fact of a prior conviction, that increases a criminal defendant’s maximum potential sentence. Apprendi v. New Jersey, 530 U.S. 466 (2000); Blakely v. Washington, 542 U.S. 296 (2004). We have applied this principle in numerous cases where the sentence was imprisonment or death. The question here is whether the same rule applies to sentences of criminal fines. We hold that it does. I Petitioner Southern Union Company is a natural gas distributor. Its subsidiary stored liquid mercury, a hazardous substance, at a facility in Pawtucket, Rhode Is-land. In September 2004, youths from a nearby apartment complex broke into the facility, played with the mer-cury, and spread it around the facility and complex. The complex’s residents were temporarily displaced during the cleanup and most underwent testing for mercury poisoning. In 2007, a grand jury indicted Southern Union on multiple counts of violating federal environmental statutes. As relevant here, the first count alleged that the company knowingly stored liquid mercury without a permit at the Pawtucket facility “[f]rom on or about September 19, 2002 until on or about October 19, 2004,” App. 104, in viola- tion of the Resource Conservation and Recovery Act of 1976 (RCRA). See 90Stat. 2812, as amended, 42 U. S. C. §6928(d)(2)(A). A jury convicted Southern Union on this count following a trial in the District Court for the District of Rhode Island. The verdict form stated that Southern Union was guilty of unlawfully storing liquid mercury “on or about September 19, 2002 to October 19, 2004.” App. 140. Violations of the RCRA are punishable by, inter alia, “a fine of not more than $50,000 for each day of violation.” §6928(d). At sentencing, the probation office set a maximum fine of $38.1 million, on the basis that Southern Union violated the RCRA for each of the 762 days from September 19, 2002, through October 19, 2004. Southern Union objected that this calculation violated Apprendi because the jury was not asked to determine the precise duration of the violation. The company noted that the ver-dict form listed only the violation’s approximate start date (i.e., “on or about”), and argued that the court’s instructions permitted conviction if the jury found even a 1-day violation. Therefore, Southern Union maintained, the only violation the jury necessarily found was for one day, and imposing any fine greater than the single-day penalty of $50,000 would require factfinding by the court, in contravention of Apprendi. The Government acknowledged the jury was not asked to specify the duration of the violation, but argued that Apprendi does not apply to criminal fines. The District Court disagreed and held that Apprendi applies. But the court concluded from the “content and context of the verdict all together” that the jury found a 762-day violation. App. to Pet. for Cert. 46a. The court therefore set a maximum potential fine of $38.1 million, from which it imposed a fine of $6 million and a “community service obligatio[n]” of $12 million. App. 154. On appeal, the United States Court of Appeals for the First Circuit rejected the District Court’s conclusion that the jury necessarily found a violation of 762 days. 630 F.3d 17, 36 (2010). But the Court of Appeals affirmed the sentence because it also held, again in contrast to the District Court, that Apprendi does not apply to criminal fines. 630 F. 3d, at 33–36. Other Circuits have reached the opposite conclusion. See United States v. Pfaff, 619 F.3d 172 (CA2 2010) (per curiam); United States v. LaGrou Distribution Sys., Inc., 466 F.3d 585 (CA7 2006). We granted certiorari to resolve the conflict, 565 U. S. ___ (2011), and now reverse. II A This case requires us to consider the scope of the Sixth Amendment right of jury trial, as construed in Apprendi. Under Apprendi, “[o]ther than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt.” 530 U. S., at 490. The “ ‘statutory maximum’ for Apprendi purposes is the maximum sentence a judge may impose solely on the basis of the facts reflected in the jury verdict or admitted by the defendant.” Blakely, 542 U. S., at 303 (emphasis deleted). Thus, while judges may exercise discretion in sentencing, they may not “inflic[t] punishment that the jury’s verdict alone does not allow.” Id., at 304. Apprendi’s rule is “rooted in longstanding common-law practice.” Cunningham v. California, 549 U.S. 270, 281 (2007). It preserves the “historic jury function” of “determining whether the prosecution has proved each element of an offense beyond a reasonable doubt.” Oregon v. Ice, 555 U.S. 160, 163 (2009). We have repeatedly affirmed this rule by applying it to a variety of sentencing schemes that allowed judges to find facts that increased a defendant’s maximum authorized sentence. See Cunningham, 549 U. S., at 274–275 (elevated “upper term” of impris-onment); United States v. Booker, 543 U.S. 220, 226– 227, 233–234 (2005) (increased imprisonment range for defendant under then-mandatory Federal Sentencing Guidelines); Blakely, 542 U. S., at 299–300 (increased im-prisonment above statutorily prescribed “standard range”); Ring v. Arizona, 536 U.S. 584, 588–589 (2002) (death penalty authorized upon finding existence of aggravating factors); Apprendi, 530 U. S., at 468–469 (extended term of imprisonment based on violation of a “hate crime” statute). While the punishments at stake in those cases were imprisonment or a death sentence, we see no principled basis under Apprendi for treating criminal fines differ-ently. Apprendi’s “core concern” is to reserve to the jury “the determination of facts that warrant punishment for a specific statutory offense.” Ice, 555 U. S., at 170. That concern applies whether the sentence is a criminal fine or imprisonment or death. Criminal fines, like these other forms of punishment, are penalties inflicted by the sovereign for the commission of offenses. Fines were by far the most common form of noncapital punishment in colonial America.[1] They are frequently imposed today, especially upon organizational defendants who cannot be imprisoned.[2] And the amount of a fine, like the maximum term of imprisonment or eligibility for the death penalty, is of-ten calculated by reference to particular facts. Sometimes, as here, the fact is the duration of a statutory violation;[3] under other statutes it is the amount of the defendant’s gain or the victim’s loss, or some other factor.[4] In all such cases, requiring juries to find beyond a reasonable doubt facts that determine the fine’s maximum amount is necessary to implement Apprendi’s “animating principle”: the “preservation of the jury’s historic role as a bulwark between the State and the accused at the trial for an alleged offense.” Ice, 555 U. S., at 168. In stating Apprendi’s rule, we have never distinguished one form of punishment from another. Instead, our decisions broadly prohibit judicial factfinding that increases maximum criminal “sentence[s],” “penalties,” or “punishment[s]”—terms that each undeniably embrace fines. E.g., Blakely, 542 U. S., at 304; Apprendi, 530 U. S., at 490; Ring, 536 U. S., at 589. The Government objects, however, that fines are less onerous than incarceration and the death sentence. The Government notes that Apprendi itself referred to the physical deprivation of liberty that imprisonment occasions, see 530 U. S., at 484, and that we have placed more weight on imprisonment than on fines when construing the scope of the Sixth Amendment rights to counsel and jury trial. See Blanton v. North Las Vegas, 489 U.S. 538, 542–543 (1989) (jury trial); Scott v. Illinois, 440 U.S. 367, 373–374 (1979) (counsel). Therefore, the Government concludes, fines categorically “do not implicate” the “primary concerns motivating Apprendi.” Brief for United States 23–25. This argument fails because its conclusion does not fol-low from its premise. Where a fine is so insubstantial that the underlying offense is considered “petty,” the Sixth Amendment right of jury trial is not triggered, and no Apprendi issue arises. See, e.g., Muniz v. Hoffman, 422 U.S. 454, 477 (1975) ($10,000 fine imposed on labor union does not entitle union to jury trial); see also Blanton, 489 U. S., at 541 (no jury trial right for “petty” offenses, as measured by the “severity of the maximum authorized penalty” (internal quotation marks omitted)). The same, of course, is true of offenses punishable by relatively brief terms of imprisonment—these, too, do not entitle a defendant to a jury trial. See id., at 543 (establishing a rebuttable presumption that offenses punishable by six months’ imprisonment or less are petty); Duncan v. Louisiana, 391 U.S. 145, 159–162 (1968). But not all fines are insubstantial, and not all offenses punishable by fines are petty. See, e.g., Mine Workers v. Bagwell, 512 U.S. 821, 838, n. 5 (1994) (criminal contempt fine of $52 million imposed on union “unquestionably is a serious contempt sanction” that triggers right of jury trial). The federal twice-the-gain-or-loss statute, in particular, see 18 U. S. C. §3571(d), has been used to obtain substantial judgments against organizational defendants. See, e.g., Amended Judgment in United States v. LG Display Co., Ltd., No. 08–CR–803–SI (ND Cal.), pp. 1–2 ($400 million fine for conviction of single count of violating Sherman Antitrust Act); Judgment in United States v. Siemens Aktiengesellschaft, No. 08–CR–367–RJL (D DC), pp. 1–2, 5 ($448.5 million fine for two violations of Foreign Corrupt Practices Act); United States Sentencing Commission, 2010 Annual Report, ch. 5, p. 38 (noting fine of $1.195 billion imposed on pharmaceutical corporation for violations of food and drug laws). And, where the defendant is an individual, a large fine may “engender ‘a significant infringement of personal freedom.’ ” Blanton, 489 U. S., at 542 (quoting Frank v. United States, 395 U.S. 147, 151 (1969)); see also 18 U. S. C. §3572(a)(2) (requiring court to consider “the burden that the fine will impose upon the defendant” in determining whether to impose a fine and in what amount). The Government thus asks the wrong question by comparing the severity of criminal fines to that of other punishments. So far as Apprendi is concerned, the relevant question is the significance of the fine from the perspective of the Sixth Amendment’s jury trial guarantee. Where a fine is substantial enough to trigger that right, Apprendi applies in full. As we said in Cunningham, “Asking whether a defendant’s basic jury-trial right is preserved, though some facts essential to punishment are reserved for determination by the judge, . . . is the very inquiry Apprendi’s ‘bright-line rule’ was designed to exclude.” 549 U. S., at 291. This case is exemplary. The RCRA subjects Southern Union to a maximum fine of $50,000 for each day of violation. 42 U. S. C. §6928(d). The Government does not deny that, in light of the seriousness of that punishment, the company was properly accorded a jury trial. And the Government now concedes the District Court made factual findings that increased both the “potential and actual” fine the court imposed. Brief for United States 28. This is exactly what Apprendi guards against: judicial factfinding that enlarges the maximum punishment a defendant faces beyond what the jury’s verdict or the defendant’s admissions allow. B In concluding that the rule of Apprendi does not apply to criminal fines, the Court of Appeals relied on our decision in Ice. Ice addressed the question whether, when a defendant is convicted of multiple offenses, Apprendi forbids judges to determine facts that authorize the imposition of consecutive sentences. 555 U. S., at 164. In holding that Apprendi does not, Ice emphasized that juries historically played no role in deciding whether sentences should run consecutively or concurrently. See 555 U. S., at 168–169. The Court of Appeals reasoned that juries were similarly uninvolved in setting criminal fines. 630 F. 3d, at 35.[5] The Court of Appeals was correct to examine the histor-ical record, because “the scope of the constitutional jury right must be informed by the historical role of the jury at common law.” Ice, 555 U. S., at 170. See also, e.g., Blakely, 542 U. S., at 301–302; Apprendi, 530 U. S., at 477–484. But in our view, the record supports applying Apprendi to criminal fines. To be sure, judges in the col-onies and during the founding era “possessed a great deal of discretion” in determining whether to impose a fine and in what amount. Lillquist 640–641; see also Preyer 350. Often, a fine’s range “was apparently without limit except insofar as it was within the expectation on the part of the court that it would be paid.” Ibid. For some other offenses, the maximum fine was capped by statute. See, e.g., id., at 333 (robbery, larceny, burglary, and other offenses punishable in Massachusetts Bay Colony “by fines of up to £5”); Act of Feb. 28, 1803, ch. 9, §7, 2Stat. 205 (any consul who gives a false certificate shall “forfeit and pay a fine not exceeding ten thousand dollars, at the discretion of the court”); K. Stith & J. Cabranes, Fear of Judging: Sentencing Guidelines in the Federal Courts 9 (1998) (describing federal practice). The exercise of such sentencing discretion is fully consistent with Apprendi, which permits courts to impose “judgment within the range prescribed by statute.” 530 U. S., at 481 (emphasis in original). Nor, a fortiori, could there be an Apprendi violation where no maximum is prescribed. Indeed, in surveying the historical record that formed the basis of our holding in Apprendi, we specifi-cally considered the English practice with respect to fines, which, as was true of many colonial offenses, made sentencing largely “dependent upon judicial discretion.” See id., at 480, n. 7; see also Jones v. United States, 526 U.S. 227, 244–245 (1999); 4 W. Blackstone, Commentaries on the Laws of England 372–373 (1769) (hereinafter Blackstone). And even then, as the dissent acknowledges, post, at 11–12 (opinion of Breyer, J.), there is authority suggesting that English juries were required to find facts that determined the authorized pecuniary punishment. See 1 T. Starkie, A Treatise on Criminal Pleading 187–188 (1814) (In cases “where the offence, or its defined measure of punishment, depends upon” property’s specific value, the value “must be proved precisely as it is laid [in the indictment], and any variance will be fatal”); see also id., at 188 (“[I]n the case of usury, where the judgment depends upon the quantum taken, the usurious contract must be averred according to the fact; and a variance from it, in evidence, would be fatal, because the penalty is apportioned to the value” (emphasis in original)); 2 W. Hawkins, A Treatise of the Pleas of the Crown, ch. 25, §75, pp. 234–235 (3d ed. 1739) (doubting whether “it be need-ful to set forth the Value of the Goods in an Indictment of Trespass for any other Purpose than to aggravate the Fine”). In any event, the salient question here is what role the jury played in prosecutions for offenses that did peg the amount of a fine to the determination of specified facts—often, the value of damaged or stolen property. See Apprendi, 530 U. S., at 502, n. 2 (Thomas, J., concurring). Our review of state and federal decisions discloses that the predominant practice was for such facts to be alleged in the indictment and proved to the jury. See, e.g., Commonwealth v. Smith, 1 Mass. 245, 247 (1804) (declining to award judgment of treble damages for all stolen items in larceny prosecution when indictment alleged value of only some of the items); Clark v. People, 2 Ill. 117, 120–121 (1833) (arson indictment must allege value of destroyed building because statute imposed “a fine equal in value to the property burned”); State v. Garner, 8 Port. 447, 448 (Ala. 1839) (same in malicious mischief prosecution where punishment was fine “not exceeding four fold the value of the property injured or destroyed”); Ritchey v. State, 7 Blackf. 168, 169 (Ind. 1844) (same in arson prosecution because, “[i]n addition to imprisonment in the penitentiary, the guilty person is liable to a fine not exceeding double the value of the property destroyed”); Hope v. Commonwealth, 50 Mass. 134, 137 (1845) (the “value of the property alleged to be stolen must be set forth in the indictment” in part because “[o]ur statutes . . . prescribe the punishment for larceny, with reference to the value of the property stolen”); State v. Goodrich, 46 N. H. 186, 188 (1865) (“It may also be suggested, that, in the case of simple larceny, the respondent may be sentenced to pay the owner of the goods stolen, treble the value thereof, which is an additional reason for requiring the [value of the stolen items] to be stated [in the indictment]”); United States v. Woodruff, 68 F. 536, 538 (Kan. 1895) (“[T]he defendant is entitled to his constitutional right of trial by jury” to ascertain “the exact sum for which a fine may be imposed”).[6] The rule that juries must determine facts that set a fine’s maximum amount is an application of the “two longstanding tenets of common-law criminal jurisprudence” on which Apprendi is based: First, “the ‘truth of every accusation’ against a defendant ‘should afterwards be confirmed by the unanimous suffrage of twelve of his equals and neighbours.’ ” Blakely, 542 U. S., at 301 (quoting 4 Blackstone 343). And second, “ ‘an accusation which lacks any particular fact which the law makes essential to the punishment is . . . no accusation within the requirements of the common law, and it is no accusation in reason.’ ” 542 U. S., at 301–302 (quoting 1 J. Bishop, Criminal Procedure §87, p. 55 (2d ed. 1872)). Indeed, Bishop’s leading treatise on criminal procedure specifically identified cases involving fines as evidence of the proposition that “the indictment must, in order to inform the court what punishment to inflict, contain an averment of every particular thing which enters into the punishment.” Id., §540, at 330 (discussing Clark and Garner). This principle, Bishop explained, “pervades the entire system of the adjudged law of criminal procedure. It is not made apparent to our understandings by a single case only, but by all the cases.” Criminal Procedure §81, at 51. See also Ap-prendi, 530 U. S., at 510–511 (Thomas, J., concurring) (explaining that Bishop grounded this principle in “well-established common-law practice . . . and in the provisions of Federal and State Constitutions guaranteeing notice of an accusation in all criminal cases, indictment by a grand jury for serious crimes, and trial by jury”). As counterevidence that juries historically did not determine facts relevant to criminal fines, the Government points to two decisions from this Court. One is United States v. Murphy, 16 Pet. 203 (1842), which considered whether an interested witness was competent to testify in a larceny prosecution brought under a provision of the Crimes Act of 1790. Murphy’s only relevance to this case is that the Crimes Act authorized a fine of up to four times the value of the stolen property, and the Court remarked that “the fine is, as to its amount, purely in the discretion of the Court.” Id., at 209. But this statement is best read as permitting the court to select a fine from within the maximum authorized by jury-found facts—a practice, as noted, that accords with Apprendi. Such a reading is consistent with the fact that the indictment in Murphy alleged the value of the stolen items, see 16 Pet., at 207–208, and with the practice of contemporary courts addressing the same statute, see United States v. Holland, 26 F. Cas. 343, 345 (No. 15,378) (CC SDNY 1843) (trial court instructs jury “to assess the value of the property taken” in order to determine maximum fine); Pye v. United States, 20 F. Cas. 99 (No. 11,488) (CC DC 1842) (value of stolen items alleged in indictment). The Government and dissent place greater reliance on United States v. Tyler, 7 Cranch 285 (1812). But like Murphy, this decision involved no constitutional question. Rather, it construed a federal embargo statute that imposed a fine of four times the value of the property intended to be exported. The indictment identified the property at issue as “pearl-ashes,” but the jury’s guilty verdict re-ferred instead to “ ‘pot-ashes [that] were worth two hundred and eighty dollars.’ ” Tyler, 7 Cranch, at 285.[7] The question was whether the discrepancy rendered the verdict “not sufficiently certain as to the value of the property charged in the indictment,” i.e., pearl-ashes. Ibid. The Court held that the discrepancy was immaterial, on the ground that “under this law, no valuation by the jury was necessary to enable the Circuit Court to impose the proper fine.” Ibid. The Court’s reasoning is somewhat opaque, but appears to rest on the text of the embargo statute, which directed that the defendant “shall, upon conviction, be . . . fined a sum by the Court.” Ibid. In any event, nothing in the decision purports to construe the Sixth Amendment. And, insofar as Tyler reflects prevailing practice, it bears noting that both the indictment and ver-dict identified the value of the property at issue. See Tr. 2 in Tyler, 7 Cranch 285, reprinted in Appellate Case Files of the Supreme Court of the United States, 1792–1831, National Archives Microfilm Publications No. 214 (1962), roll 18 (indictment: “nineteen barrels of pearlashes, which were then and there of the value of six hundred dollars”). Whatever the precise meaning of this decision, it does not outweigh the ample historical evidence showing that juries routinely found facts that set the maximum amounts of fines. III The Government’s remaining arguments, echoed by the dissent (see post, at 23–28), are unpersuasive. The Government first submits that, when it comes to fines, “the judicially found facts typically involve only quantifying the harm caused by the defendant’s offense”—for example, how long did the violation last, or how much money did the defendant gain (or the victim lose)?—“as opposed to de-fining a separate set of acts for punishment.” Brief for United States 25. Only the latter determination, the Government contends, implicates Apprendi’s concerns. This argument has two defects. First, it rests on an assumption that Apprendi and its progeny have uniformly rejected: that in determining the maximum punishment for an offense, there is a constitutionally significant difference between a fact that is an “element” of the offense and one that is a “sentencing factor.” See, e.g., 530 U. S., at 478; Ring, 536 U. S., at 605. Second, we doubt the coherence of this distinction. This case proves the point. Under 42 U. S. C. §6928(d), the fact that will ultimately determine the maximum fine Southern Union faces is the number of days the company violated the statute. Such a finding is not fairly characterized as merely “quantifying the harm” Southern Union caused. Rather, it is a determination that for each given day, the Government has proved that Southern Union committed all of the acts constituting the offense. The Government next contends that applying Apprendi to fines will prevent States and the Federal Government from enacting statutes that, like §6928(d), calibrate fines to a defendant’s culpability, thus providing just punishment and reducing unwarranted sentencing disparity. But the Government presents a false choice. As was true in our prior Apprendi cases, and remains so here, legislatures are free to enact statutes that constrain judges’ discretion in sentencing—Apprendi requires only that such provisions be administered in conformance with the Sixth Amendment. Last, the Government argues that requiring juries to determine facts related to fines will cause confusion (because expert testimony might be needed to guide the inquiry); or prejudice the defendant (who might have to deny violating a statute while simultaneously arguing that any violation was minimal); or be impractical (at least when the relevant facts are unknown or unknowable until the trial is completed).[8] These arguments rehearse those made by the dissents in our prior Apprendi cases. See Booker, 543 U. S., at 329 (Breyer, J., dissenting in part); Blakely, 542 U. S., at 318–320 (O’Connor, J., dissenting); id., at 330–340 (Breyer, J., dissenting); Apprendi, 530 U. S., at 555–559 (same). Here, as there, they must be rejected. For even if these predictions are ac-curate, the rule the Government espouses is unconstitutional. That “should be the end of the matter.” Blakely, 542 U. S., at 313. But here there is particular reason to doubt the strength of these policy concerns. Apprendi is now more than a decade old. The reliance interests that underlie many of the Government’s arguments are by this point attenuated. Nor, in our view, does applying Apprendi’s rule to criminal fines mark an unexpected extension of the doctrine. Most Circuits to have addressed the issue already embrace this position, see Pfaff, 619 F. 3d, at 175–176; LaGrou Distribution Sys., 466 F. 3d, at 594; United States v. Yang, 144 Fed. Appx. 521, 524 (CA6 2005), as did the Government prior to Ice, see Brief in Opposition 11, n. 2. In light of the reasons given in this opinion, the dramatic departure from precedent would be to hold criminal fines exempt from Apprendi. * * * We hold that the rule of Apprendi applies to the imposition of criminal fines. The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 See Preyer, Penal Measures in the American Colonies: An Overview, 26 Am. J. Legal Hist. 326, 350 (1982) (hereinafter Preyer); see also Lillquist, The Puzzling Return of Jury Sentencing: Misgivings About Apprendi, 82 N. C. L. Rev. 621, 640–641 (2004) (hereinafter Lillquist); Browning-Ferris Industries of Vt., Inc. v. Kelco Disposal, Inc., 492 U.S. 257, 290 (1989) (O’Connor, J., concurring in part and dissenting inpart) (fines were “the preferred penal sanction” in England by the 17th century). “Imprisonment,” in contrast, “although provided for as a punishment in some colonies, was not a central feature of criminal punishment until a later time.” Preyer 329; see also Lillquist 641–643. 2 In 2011, a fine was imposed on 9.0% of individual defendants and on 70.6% of organizational defendants in the federal system. See United States Sentencing Commission, 2011 Annual Report, ch. 5, pp. 34, 40. 3 See, e.g., 12 U. S. C. §1467a(i)(1); 15 U. S. C. §717t(b); 16 U. S. C. §825o(b); Cal. Health & Safety Code Ann. §25515(a) (West Supp. 2012); Colo. Rev. Stat. Ann. §§25–7–122.1(1)(b) and (c) (2011); Mass. Gen. Laws, ch. 21, §34C (West 2010); N. J. Stat. Ann. §13:1E–99.89(f) (West Supp. 2012). 4 See, e.g., 18 U. S. C. §3571(d) (fine “not more than the greater of twice the gross gain or twice the gross loss”); Fla. Stat. §775.083(1)(f) (2010) (same); Tex. Parks & Wild. Code Ann. §12.410(c) (West 2002) (same); see also 18 U. S. C. §645 (fine for embezzlement by officers of United States courts of up to twice the value of the money embezzled); §201(b) (fine for bribery of public officials of up to three times the value of the bribe). 5 Ice also stated in dicta that applying Apprendi to consecutive-versus-concurrent sentencing determinations might imperil a variety of sentencing decisions judges commonly make, including “the imposition of statutorily prescribed fines.” 555 U. S., at 171. The Court of Appeals read this statement to mean that Apprendi does not apply to criminal fines. 630 F. 3d, at 34. We think the statement is at most ambiguous, and more likely refers to the routine practice of judges’ imposing fines from within a range authorized by jury-found facts. Such a practice poses no problem under Apprendi because the penalty does not exceed what the jury’s verdict permits. See 530 U. S., at 481. In any event, our statement in Ice was unnecessary to the judgment and is not binding. Central Va. Community College v. Katz, 546 U.S. 356, 363 (2006). 6 The dissent believes these decisions are inapposite because some of them arose in States that authorized juries, rather than judges, to im-pose sentence. See post, at 18–20. But this fact was not the basis ofthe decisions; rather, the courts required value to be alleged and proved to the jury because “the extent of the punishment . . . depend[s] upon the value of the property consumed or injured.” Ritchey, 7 Blackf.,at 169; see also, e.g., Clark, 2 Ill., at 120–121 (same). And as Bishop explained, this requirement of proof originated not from a unique fea-ture of jury sentencing, but from longstanding common-law princi-ples—a point to which the dissent notably does not respond. 1 J. Bishop, Criminal Procedure §§81, 540 (2d ed. 1872). See infra, at 12. Nor, for that matter, do larceny cases “presen[t] a special circumstance.” Post, at 20. Such decisions invoked the same reasoning as the other cases just mentioned. See, e.g., Hope, 50 Mass., at 137 (value must be proved because, among other things, “[o]ur statutes . . . prescribe the punishment for larceny . . . with reference to the value of the property stolen”); Goodrich, 46 N. H., at 188 (same). Bishop made this point explicit: “[Value] must be alleged wherever it is an element tobe considered by the court in determining the punishment, and it is immaterial whether the particular crime is larceny or any other crime.” Criminal Procedure §541, at 331 (footnote omitted and emphasis added). At the end of the day, the only evidence the dissent musters that judges found fine-enhancing facts are United States v. Tyler, 7 Cranch 285 (1812), and one lower-court decision restating Tyler’s holding. See post, at 15–17. We address Tyler below. See infra, at13–14. 7 We will not keep the reader in suspense: pot-ash and pearl-ash are alkaline salts of differing causticity that “for a long time . . . [were] amongst the most valuable articles of manufacture and commerce”in parts of early America. D. Townsend, Principles and Observations Applied to the Manufacture and Inspection of Pot and Pearl Ashes 3 (1793). See also Board of Trustees of Leland Stanford Junior Univ. v. Roche Molecular Systems, Inc., 563 U. S. __, __ (2011) (slip op., at 6). 8 In this vein, the dissent speculates that today’s decision may “nudg[e] our [criminal justice] system” further in favor of plea bargains at the expense of jury trials. Post, at 28. But groups representing the interests of defendants—whom the dissent’s rule purportedly favors—tell us the opposite is true. See Brief for Chamber of Commerce of the United States of America et al. as Amici Curiae 5 (“[E]xempting criminal fines from Apprendi makes innocent defendants more likely to plead guilty”).

566.US.2011_10-1472

Title 28 U. S. C. §1920, as amended by the Court Interpreters Act, includes “compensation of interpreters” among the costs that may be awarded to prevailing parties in federal-court lawsuits. §1920(6). In this case, the District Court awarded costs to respondent as the prevailing party in a civil action instituted by petitioner. The award included the cost of translating from Japanese to English certain documents that respondent used in preparing its defense. The Ninth Circuit affirmed, concluding that §1920(6) covers the cost of translating documents as well as the cost of translating live speech. Held: Because the ordinary meaning of “interpreter” is someone who translates orally from one language to another, the category “compensation of interpreters” in §1920(6) does not include the cost of document translation. Pp. 3−15. (a) Section 1920 reflects the substance of an 1853 Act that specified for the first time what costs are allowable in federal court. That provision defines the term “costs” as used in Federal Rule of Civil Procedure 54(d), which gives courts the discretion to award costs to prevailing parties. Crawford Fitting Co. v. J. T. Gibbons, Inc., 482 U.S. 437, 441. As originally configured, §1920 contained five categories of taxable costs, but in 1978, Congress enacted the Court Interpreters Act, which added a sixth category that includes “compensation of interpreters.” §1920(6). Pp. 3−5. (b) Because the term “interpreter” is not defined in the Court Interpreters Act or in any other relevant statutory provision, it must be given its ordinary meaning. Asgrow Seed Co. v. Winterboer, 513 U.S. 179, 187. When Congress passed that Act in 1978, many dictionaries defined “interpreter” as one who translates spoken, as opposed to written, language. Pre-1978 legal dictionaries also generally defined “interpreter” and “interpret” in terms of oral translation. Respondent relies almost exclusively on a version of Webster’s Third New International Dictionary that defined “interpreter” as “one that translates; esp: a person who translates orally for parties conversing in different tongues.” Although the sense divider esp (for especially) indicates that the most common meaning of the term is one “who translates orally,” that meaning is subsumed within the more general definition “one that translates.” That a definition is broad enough to encompass one sense of a word does not establish, however, that the word is ordinarily understood in that sense. See Mallard v. United States Dist. Court for Southern Dist. of Iowa, 490 U.S. 296, 301. Although all relevant dictionaries defined “interpreter” at the time of the statute’s enactment as including persons who translate orally, only a handful defined the word broadly enough to encompass translators of written materials. Notably, the Oxford English Dictionary, one of the most authoritative, recognized that “interpreter” can mean one who translates writings, but it expressly designated that meaning as obsolete. Any definition of a word that is absent from many dictionaries and is deemed obsolete in others is hardly a common or ordinary meaning. Given this survey of relevant dictionaries, the ordinary meaning of “interpreter” does not include those who translate writings. Nothing in the Court Interpreters Act or in §1920 hints that Congress intended to go beyond this ordinary meaning. If anything, the statutory context suggests that “interpreter” includes only those who translate orally. See 28 U. S. C. §1827. Moreover, Congress’ use of technical terminology reflects the distinction in relevant professional literature between interpreters, who are used for oral conversations, and translators, who are used for written communications. Pp. 5−11. (c) No other tool of construction compels a departure from the ordinary meaning of “interpreter.” This Court has never held that Rule 54(d) creates a presumption in favor of the broadest possible reading of the costs enumerated in §1920. To the contrary, the Court has made clear that the “discretion granted by Rule 54(d) is not a power to evade” the specific categories of costs set forth by Congress, Crawford Fitting, supra, at 442, but “is solely a power to decline to tax, as costs, the items enumerated in §1920,” ibid. This Court’s conclusion is in keeping with the narrow bounds of taxable costs, which are limited by statute and modest in scope. Respondent’s extratextual arguments―that documentary evidence is no less important than testimonial evidence and that some translation tasks are not entirely oral or entirely written―are more properly directed at Congress. In any event, neither argument is so compelling that Congress must have intended to dispense with the ordinary meaning of “interpreter” in §1920(6). Pp. 12−15. 633 F.3d 1218, vacated and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Scalia, Kennedy, Thomas, and Kagan, JJ., joined. Ginsburg, J., filed a dissenting opinion, in which Breyer and Sotomayor, JJ., joined.

The costs that may be awarded to prevailing parties in lawsuits brought in federal court are set forth in 28 U. S. C. §1920. The Court Interpreters Act amended that statute to include “compensation of interpreters.” §1920(6); see also §7, 92Stat. 2044. The question pre- sented in this case is whether “compensation of interpreters” covers the cost of translating documents. Because the ordinary meaning of the word “interpreter” is a person who translates orally from one language to another, we hold that “compensation of interpreters” is limited to the cost of oral translation and does not include the cost of document translation. I This case arises from a personal injury action brought by petitioner Kouichi Taniguchi, a professional baseball player in Japan, against respondent Kan Pacific Saipan, Ltd., the owner of a resort in the Northern Mariana Islands. Petitioner was injured when his leg broke through a wooden deck during a tour of respondent’s resort prop- erty. Initially, petitioner said that he needed no medical attention, but two weeks later, he informed respondent that he had suffered cuts, bruises, and torn ligaments from the accident. Due to these alleged injuries, he claimed damages for medical expenses and for lost income from contracts he was unable to honor. After discovery concluded, both parties moved for summary judgment. The United States District Court for the Northern Ma- riana Islands granted respondent’s motion on the ground that petitioner offered no evidence that respondent knew of the defective deck or otherwise failed to exercise reasonable care. In preparing its defense, respondent paid to have various documents translated from Japanese to English. After the District Court granted summary judgment in respondent’s favor, respondent submitted a bill for those costs. Over petitioner’s objection, the District Court awarded the costs to respondent as “compensation of interpreters” under §1920(6). Explaining that interpreter services “can- not be separated into ‘translation’ and ‘interpretation,’ ” App. to Pet. for Cert. 25a, the court held that costs for document translation “fal[l] within the meaning of ‘compensation of an interpreter,’ ” ibid. Finding that it was necessary for respondent to have the documents translated in order to depose petitioner, the court con- cluded that the translation services were properly taxed as costs. The United States Court of Appeals for the Ninth Circuit affirmed both the District Court’s grant of summary judgment and its award of costs. The court rejected petitioner’s argument that the cost of document translation services is not recoverable as “compensation of interpreters.” The court explained that “the word ‘interpreter’ can reasonably encompass a ‘translator,’ both according to the dictionary definition and common usage of these terms, which does not always draw precise distinctions between foreign language interpretations involving live speech versus written documents.” 633 F.3d 1218, 1221 (2011). “More importantly,” the court stressed, this construction of the statute “is more compatible with Rule 54 of the Fed- eral Rules of Civil Procedure, which includes a decided preference for the award of costs to the prevailing party.” Ibid. The court thus concluded that “the prevailing party should be awarded costs for services required to interpret either live speech or written documents into a familiar language, so long as interpretation of the items is necessary to the litigation.” Id., at 1221–1222. Because there is a split among the Courts of Appeals on this issue,[1] we granted certiorari. 564 U. S. ___ (2011). II A Although the taxation of costs was not allowed at common law, it was the practice of federal courts in the early years to award costs in the same manner as the courts of the relevant forum State. Alyeska Pipeline Service Co. v. Wilderness Society, 421 U.S. 240, 247–248 (1975). In 1793, Congress enacted a statute that authorized the awarding of certain costs to prevailing parties based on state law: “That there be allowed and taxed in the supreme, circuit and district courts of the United States, in favour of the parties obtaining judgments therein, such compensation for their travel and attendance, and for attornies and counsellors’ fees . . . as are allowed in the supreme or superior courts of the respective states.” Act of Mar. 1, 1793, §4, 1Stat. 333. Although twice reenacted, this provision expired in 1799. Alyeska Pipeline, supra, at 248, n. 19; Crawford Fitting Co. v. J. T. Gibbons, Inc., 482 U.S. 437, 439 (1987). Yet even in the absence of express legislative authorization, the practice of referring to state rules for the taxation of costs persisted. See Alyeska Pipeline, 421 U. S., at 250. Not until 1853 did Congress enact legislation specifying the costs allowable in federal court. Id., at 251. The impetus for a uniform federal rule was largely the consequence of two developments. First, a “great diversity in practice among the courts” had emerged. Ibid. Second, “losing litigants were being unfairly saddled with exorbitant fees for the victor’s attorney.” Ibid. Against this backdrop, Congress passed the 1853 Fee Act, which we have described as a “far-reaching Act specifying in detail the nature and amount of the taxable items of cost in the federal courts.” Id., at 251–252. The substance of this Act was transmitted through the Revised Statutes of 1874 and the Judicial Code of 1911 to the Revised Code of 1948, where it was codified, “without any apparent intent to change the controlling rules,” as 28 U. S. C. §1920. 421 U. S., at 255. Federal Rule of Civil Procedure 54(d) gives courts the discretion to award costs to prevailing parties. That Rule provides in relevant part: “Unless a federal statute, these rules, or a court order provides otherwise, costs—other than attorney’s fees—should be allowed to the prevailing party.” Rule 54(d)(1). We have held that “§1920 defines the term ‘costs’ as used in Rule 54(d).” Crawford Fitting, 482 U. S., at 441. In so doing, we rejected the view that “the discretion granted by Rule 54(d) is a separate source of power to tax as costs expenses not enumerated in §1920.” Ibid. As originally configured, §1920 contained five categories of taxable costs: (1) “[f]ees of the clerk and marshal”; (2) “[f]ees of the court reporter for all or any part of the steno- graphic transcript necessarily obtained for use in the case”; (3) “[f]ees and disbursements for printing and witnesses”; (4) “[f]ees for exemplification and copies of papers necessarily obtained for use in the case”; and (5) “[d]ocket fees under section 1923 of this title.” 62Stat. 955. In 1978, Congress enacted the Court Interpreters Act, which amended §1920 to add a sixth category: “Compensation of court appointed experts, compensation of interpreters, and salaries, fees, expenses, and costs of special interpretation services under section 1828 of this title.” 28 U. S. C. §1920(6); see also §7, 92Stat. 2044. We are concerned here with this sixth category, specifically the item of tax- able costs identified as “compensation of interpreters.” B To determine whether the item “compensation of interpreters” includes costs for document translation, we must look to the meaning of “interpreter.” That term is not defined in the Court Interpreters Act or in any other relevant statutory provision. When a term goes undefined in a statute, we give the term its ordinary meaning. Asgrow Seed Co. v. Winterboer, 513 U.S. 179, 187 (1995). The question here is: What is the ordinary meaning of “interpreter”? Many dictionaries in use when Congress enacted the Court Interpreters Act in 1978 defined “interpreter” as one who translates spoken, as opposed to written, language. The American Heritage Dictionary, for instance, defined the term as “[o]ne who translates orally from one language into another.” American Heritage Dictionary 685 (1978). The Scribner-Bantam English Dictionary defined the related word “interpret” as “to translate orally.” Scribner-Bantam English Dictionary 476 (1977). Similarly, the Random House Dictionary defined the intransitive form of “interpret” as “to translate what is said in a foreign language.” Random House Dictionary of the English Language 744 (1973) (emphasis added). And, notably, the Oxford English Dictionary defined “interpreter” as “[o]ne who translates languages,” but then divided that definition into two senses: “a. [a] translator of books or writings,” which it designated as obsolete, and “b. [o]ne who translates the communications of persons speaking different languages; spec. one whose office it is to do so orally in the presence of the persons; a dragoman.” 5 Oxford English Dictionary 416 (1933); see also Concise Oxford Dictionary of Current English 566 (6th ed. 1976) (“One who interprets; one whose office it is to translate the words of persons speaking different languages, esp. orally in their presence”); Chambers Twentieth Century Dictionary 686 (1973) (“one who translates orally for the benefit of two or more parties speaking different languages: . . . a translator (obs.)”). Pre-1978 legal dictionaries also generally defined the words “interpreter” and “interpret” in terms of oral translation. The then-current edition of Black’s Law Dictionary, for example, defined “interpreter” as “[a] person sworn at a trial to interpret the evidence of a foreigner . . . to the court,” and it defined “interpret” in relevant part as “to translate orally from one tongue to another.” Black’s Law Dictionary 954, 953 (rev. 4th ed. 1968); see also W. Anderson, A Dictionary of Law 565 (1888) (“One who translates the testimony of witnesses speaking a foreign tongue, for the benefit of the court and jury”); 1 B. Abbott, Dictionary of Terms and Phrases Used in American or English Jurisprudence 639 (1878) (“one who restates the testimony of a witness testifying in a foreign tongue, to the court and jury, in their language”). But see Ballentine’s Law Dictionary 655, 654 (3d ed. 1969) (defining “interpreter” as “[o]ne who interprets, particularly one who interprets words written or spoken in a foreign language,” and “interpret” as “to translate from a foreign language”). Against these authorities, respondent relies almost exclusively on Webster’s Third New International Dictionary (hereinafter Webster’s Third). The version of that dictionary in print when Congress enacted the Court Interpreters Act defined “interpreter” as “one that translates; esp: a person who translates orally for parties conversing in different tongues.” Webster’s Third 1182 (1976).[2] The sense divider esp (for especially) indicates that the most common meaning of the term is one “who translates orally,” but that meaning is subsumed within the more general definition “one that translates.” See 12,000 Words: A Supplement to Webster’s Third 15a (1986) (explaining that esp “is used to introduce the most common meaning included in the more general preceding definition”). For respondent, the general definition suf- fices to establish that the term “interpreter” ordinarily includes persons who translate the written word. Explaining that “the word ‘interpreter’ can reasonably encompass a ‘translator,’ ” the Court of Appeals reached the same conclusion. 633 F. 3d, at 1221. We disagree. That a definition is broad enough to encompass one sense of a word does not establish that the word is ordinarily understood in that sense. See Mallard v. United States Dist. Court for Southern Dist. of Iowa, 490 U.S. 296, 301 (1989) (relying on the “most common meaning” and the “ordinary and natural signification” of the word “request,” even though it may sometimes “double for ‘demand’ or ‘command’ ”). The fact that the definition of “interpreter” in Webster’s Third has a sense divider denoting the most common usage suggests that other usages, although acceptable, might not be common or ordinary. It is telling that all the dictionaries cited above defined “interpreter” at the time of the statute’s enactment as including persons who translate orally, but only a handful defined the word broadly enough to encompass translators of written material. See supra, at 5–7. Although the Oxford English Dictionary, one of the most authoritative on the English language, recognized that “interpreter” can mean one who translates writings, it expressly designated that meaning as obsolete. See supra, at 6. Were the meaning of “interpreter” that respondent advocates truly common or ordinary, we would expect to see more support for that meaning. We certainly would not expect to see it designated as obsolete in the Oxford English Dictionary. Any definition of a word that is absent from many dictionaries and is deemed obsolete in others is hardly a common or ordinary meaning. Based on our survey of the relevant dictionaries, we conclude that the ordinary or common meaning of “interpreter” does not include those who translate writings. Instead, we find that an interpreter is normally understood as one who translates orally from one language to another. This sense of the word is far more natural. As the Seventh Circuit put it: “Robert Fagles made famous translations into English of the Iliad, the Odyssey, and the Aeneid, but no one would refer to him as an English-language ‘interpreter’ of these works.” Extra Equipamentos E Exportação Ltda. v. Case Corp., 541 F.3d 719, 727 (2008). To be sure, the word “interpreter” can encompass persons who translate documents, but because that is not the ordinary meaning of the word, it does not control unless the context in which the word appears indicates that it does. Nothing in the Court Interpreters Act or in §1920, however, even hints that Congress intended to go beyond the ordinary meaning of “interpreter” and to embrace the broadest possible meaning that the definition of the word can bear. If anything, the statutory context suggests the opposite: that the word “interpreter” applies only to those who translate orally. As previously mentioned, Congress en- acted §1920(6) as part of the Court Interpreters Act. The main provision of that Act is §2(a), codified in 28 U. S. C. §§1827 and 1828. See 92Stat. 2040–2042. Particularly relevant here is §1827. As it now reads, that statute provides for the establishment of “a program to facilitate the use of certified and otherwise qualified interpreters in judicial proceedings instituted by the United States.” §1827(a). Subsection (d) directs courts to use an interpreter in any criminal or civil action instituted by the United States if a party or witness “speaks only or primarily a language other than the English language” or “suffers from a hearing impairment” “so as to inhibit such party’s comprehension of the proceedings or communication with counsel or the presiding judicial officer, or so as to inhibit such witness’ comprehension of questions and the presentation of such testimony.” §1827(d)(1).[3] As originally enacted, subsection (k) mandated that the “interpretation provided by certified interpreters . . . shall be in the consecutive mode except that the presiding judicial officer . . . may authorize a simultaneous or summary interpretation.” §1827(k) (1976 ed., Supp. II); see also 92Stat. 2042. In its current form, subsection (k) provides that interpretation “shall be in the simultaneous mode for any party . . . and in the consecutive mode for witnesses,” unless the court directs otherwise. The simultaneous, consecutive, and summary modes are all methods of oral interpretation and have nothing to do with the translation of writings.[4] Taken together, these provisions are a strong contextual clue that Congress was dealing only with oral translation in the Court Interpreters Act and that it intended to use the term “interpreter” throughout the Act in its ordinary sense as someone who translates the spoken word. As we have said before, it is a “ ‘normal rule of statutory construction’ that ‘identical words used in different parts of the same act are intended to have the same meaning.’ ” Gustafson v. Alloyd Co., 513 U.S. 561, 570 (1995) (quoting Department of Revenue of Ore. v. ACF Industries, Inc., 510 U.S. 332, 342 (1994)).[5] The references to technical terminology in the Court Interpreters Act further suggest that Congress used “interpreter” in a technical sense, and it is therefore significant that relevant professional literature draws a line between “interpreters,” who “are used for oral conversations,” and “translators,” who “are used for written communications.” Zazueta, supra n. 4, at 477; see also M. Frankenthaler, Skills for Bilingual Legal Personnel 67 (1982) (“While the translator deals with the written word, the interpreter is concerned with the spoken language”); Brislin, Introduction, in Translation: Applications and Research 1 (R. Brislin ed. 1976) (explaining that when both terms are used together, translation “refers to the processing [of] written input, and interpretation to the processing of oral input” (emphasis deleted)); J. Herbert, Interpreter’s Handbook 1 (2d ed. 1952) (“In the present-day jargon of international organisations, the words translate, translations, translator are used when the immediate result of the work is a written text; and the words interpret, interpreter, interpretation when it is a speech delivered orally”). That Congress specified “interpreters” but not “translators” is yet another signal that it intended to limit §1920(6) to the costs of oral, instead of written, translation.[6] In sum, both the ordinary and technical meanings of “interpreter,” as well as the statutory context in which the word is found, lead to the conclusion that §1920(6) does not apply to translators of written materials.[7] C No other rule of construction compels us to depart from the ordinary meaning of “interpreter.” The Court of Appeals reasoned that a broader meaning is “more compat- ible with Rule 54 of the Federal Rules of Civil Procedure, which includes a decided preference for the award of costs to the prevailing party.” 633 F. 3d, at 1221. But we have never held that Rule 54(d) creates a presumption of statutory construction in favor of the broadest possible reading of the costs enumerated in §1920. To the contrary, we have made clear that the “discretion granted by Rule 54(d) is not a power to evade” the specific categories of costs set forth by Congress. Crawford Fitting, 482 U. S., at 442. “Rather,” we have said, “it is solely a power to decline to tax, as costs, the items enumerated in §1920.” Ibid. Rule 54(d) thus provides no sound basis for casting aside the ordinary meaning of the various items enumerated in the costs statute, including the ordinary meaning of “interpreter.” Our decision is in keeping with the narrow scope of taxable costs. “Although ‘costs’ has an everyday meaning synonymous with ‘expenses,’ the concept of taxable costs under Rule 54(d) is more limited and represents those expenses, including, for example, court fees, that a court will assess against a litigant.” 10 C. Wright, A. Miller, & M. Kane, Federal Practice and Procedure §2666, pp. 202–203 (3d ed. 1998) (hereinafter Wright & Miller). Taxable costs are limited to relatively minor, incidental expenses as is evident from §1920, which lists such items as clerk fees, court reporter fees, expenses for printing and witnesses, expenses for exemplification and copies, docket fees, and compensation of court-appointed experts. Indeed, “the assessment of costs most often is merely a clerical matter that can be done by the court clerk.” Hairline Creations, Inc. v. Kefalas, 664 F.2d 652, 656 (CA7 1981). Taxable costs are a fraction of the nontaxable expenses borne by litigants for attorneys, experts, consultants, and investigators. It comes as little surprise, therefore, that “costs almost always amount to less than the successful litigant’s total expenses in connection with a lawsuit.” 10 Wright & Miller §2666, at 203. Because taxable costs are limited by statute and are modest in scope, we see no compelling reason to stretch the ordinary meaning of the cost items Congress authorized in §1920. As for respondent’s extratextual arguments, they are more properly directed at Congress. Respondent contends that documentary evidence is no less important than testimonial evidence and that it would be anomalous to require the losing party to cover translation costs for spoken words but not for written words. Brief for Respondent 20. Respondent also observes that some translation tasks are not entirely oral or entirely written. Id., at 20–24. One task, called “ ‘sight translation,’ ” involves the oral translation of a document. Id., at 21. Another task involves the written translation of speech. Ibid. And a third task, called “ ‘document comparison,’ ” involves comparing documents in the source and target language to verify that the two are identical. Id., at 21–22. Respondent argues that a narrow definition cannot account for these variations and that a bright-line definition of “interpreter” as someone who translates spoken and written words would avoid complication and provide a simple, administrable rule for district courts. Neither of these arguments convinces us that Congress must have intended to dispense with the ordinary meaning of “interpreter” in §1920(6). First, Congress might have distinguished between oral and written translation out of a concern that requiring losing parties to bear the potentially sizable costs of translating discovery docu- ments, as opposed to the more limited costs of oral tes- timony, could be too burdensome and possibly unfair, especially for litigants with limited means. Cf. Fleischmann Distilling Corp. v. Maier Brewing Co., 386 U.S. 714, 718 (1967) (noting the argument “that since litigation is at best uncertain one should not be penalized for merely defending or prosecuting a lawsuit, and that the poor might be unjustly discouraged from instituting actions to vindicate their rights if the penalty for losing included the fees of their opponents’ counsel”). Congress might also have concluded that a document translator is more akin to an expert or consultant retained by a party to decipher documentary evidence—like, for instance, a forensic accountant—than to an interpreter whose real-time oral translation services are necessary for communication between litigants, witnesses, and the court.[8] Second, respondent has not shown that any of the hybrid translation/interpretation tasks to which it points actually arise with overwhelming frequency or that the problem of drawing the line between taxable and nontax- able costs in such cases will vex the trial courts. It certainly has not shown that any such problems will be more troublesome than the task of sifting through translated discovery documents to ascertain which can be taxed as necessary to the litigation. In any event, the present case does not present a hybrid situation; it involves purely written translation, which falls outside the tasks performed by an “interpreter” as that term is ordinarily understood. * * * Because the ordinary meaning of “interpreter” is someone who translates orally from one language to another, we hold that the category “compensation of interpreters” in §1920(6) does not include costs for document translation. We therefore vacate the judgment of the United States Court of Appeals for the Ninth Circuit and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 Compare BDT Products, Inc. v. Lexmark Int’l, Inc., 405 F.3d 415, 419 (CA6 2005) (holding that document translation costs are taxable under §1920(6) because the “definition of interpret expressly includes to ‘translate into intelligible or familiar language’ ” (quoting Webster’s Third New International Dictionary 1182 (1981))), with Extra Equipamentos E Exportação Ltda. v. Case Corp., 541 F.3d 719, 727–728 (CA7 2008) (holding that document translation costs are not taxable under §1920(6) because an interpreter is “normally understood [as] a person who translates living speech from one language to another”). 2 A handful of other contemporaneous dictionaries used a similar formulation. See Funk & Wagnalls New Comprehensive International Dictionary of the English Language 665 (1977) (“One who interprets or translates; specifically, one who serves as oral translator between people speaking different languages”); 1 World Book Dictionary 1103 (C. Barnhart & R. Barnhart eds. 1977) (“a person whose business is translating, especially orally, from a foreign language”); Cassell’s English Dictionary 617 (4th ed. 1969) (“One who interprets, esp. one employed to translate orally to persons speaking a foreign language”). 3 This provision remains substantially the same as it appeared when first enacted. See 28 U. S. C. §1827(d)(1) (1976 ed., Supp. II); see also 92Stat. 2040. 4 The simultaneous mode requires the interpreter “to interpret and to speak contemporaneously with the individual whose communication is being translated.” H. R. Rep. No. 95–1687, p. 8 (1978). The consecutive mode requires the speaker whose communication is being translated to pause so that the interpreter can “convey the testimony given.” Ibid. And the summary mode “allow[s] the interpreter to condense and distill the speech of the speaker.” Ibid.; see generally Zazueta, Attorneys Guide to the Use of Court Interpreters, 8 U. C. D. L. Rev. 471, 477–478 (1975). 5 The dissent agrees that context should help guide our analysis, but instead of looking to the Court Interpreters Act, it looks to “the practice of federal courts both before and after §1920(6)’s enactment.” Post, at 4 (opinion of Ginsburg, J.). The practice of federal courts after the Act’s enactment tells us nothing about what Congress intended at the time of enactment. And federal court practice before the Act under other provisions of §1920 tells us little, if anything, about what Congress intended when it added subsection (6). We think the statutory context in which the word “interpreter” appears is a more reliable guide to its meaning. 6 Some provisions within the United States Code use both “inter-preter” and “translator” together, thus implying that Congress understands the terms to have the distinct meanings described above. See, e.g., 8 U. S. C. §1555(b) (providing that appropriations for the Immigration and Naturalization Service “shall be available for payment of . . . interpreters and translators who are not citizens of the United States”); 28 U. S. C. §530C(b)(1)(I) (providing that Department of Justice funds may be used for “[p]ayment of interpreters and translators who are not citizens of the United States”). 7 Our conclusion is buttressed by respondent’s concession at oral argument that there is no provision in the United States Code where it is clear that the word extends to those who translate documents. Tr. of Oral Arg. 39; see also Brief for Petitioner 32 (“And the Code is wholly devoid of any corresponding definition of ‘interpreter’ extending to the translation of written documents”). As respondent acknowledged, either the word is used in a context that strongly suggests it applies only to oral translation or its meaning is unclear. See Tr. of OralArg. 38. 8 The dissent contends that document translation, no less than oral translation, is essential “to equip the parties to present their case clearly and the court to decide the merits intelligently.” Post, at 5. But a document translator is no more important than an expert or consultant in making sense of otherwise incomprehensible documentary evidence, yet expenses for experts and consultants are generally not taxable as costs. To be sure, forgoing document translation can impair a litigant’s case, but document translation is not indispensable, in the way oral translation is, to the parties’ ability to communicate with each other, with witnesses, and with the court.

567.US.2011_11-210

The Stolen Valor Act makes it a crime to falsely claim receipt of military decorations or medals and provides an enhanced penalty if the Congressional Medal of Honor is involved. 18 U. S. C. §§704 (b), (c). Respondent pleaded guilty to a charge of falsely claiming that he had received the Medal of Honor, but reserved his right to appeal his claim that the Act is unconstitutional. The Ninth Circuit reversed, finding the Act invalid under the First Amendment. Held: The judgment is affirmed. Pp. 3−18. 617 F.3d 1198, affirmed. Justice Kennedy, joined by The Chief Justice, Justice Ginsburg, and Justice Sotomayor, concluded that the Act infringes upon speech protected by the First Amendment. Pp. 3–18. (a) The Constitution “demands that content-based restrictions on speech be presumed invalid . . . and that the Government bear the burden of showing their constitutionality.” Ashcroft v. American Civil Liberties Union, 542 U.S. 656, 660. Content-based restrictions on speech have been permitted only for a few historic categories of speech, including incitement, obscenity, defamation, speech integral to criminal conduct, so-called “fighting words,” child pornography, fraud, true threats, and speech presenting some grave and imminent threat the Government has the power to prevent. Absent from these few categories is any general exception for false statements. The Government argues that cases such as Hustler Magazine, Inc., v. Falwell, 485 U.S. 46, 52, support its claim that false statements have no value and hence no First Amendment protection. But all the Government’s quotations derive from cases discussing defamation, fraud, or some other legally cognizable harm associated with a false statement. In those decisions the falsity of the speech at issue was not irrelevant to the Court’s analysis, but neither was it determinative. These prior decisions have not confronted a measure, like the Stolen Valor Act, that targets falsity and nothing more. Even when considering some instances of defamation or fraud, the Court has instructed that falsity alone may not suffice to bring the speech outside the First Amendment; the statement must be a knowing and reckless falsehood. See New York Times v. Sullivan, 376 U.S. 254, 280. Here, the Government seeks to convert a rule that limits liability even in defamation cases where the law permits recovery for tortious wrongs into a rule that expands liability in a different, far greater realm of discourse and expression. The Government’s three examples of false-speech regulation that courts generally have found permissible do not establish a principle that all proscriptions of false statements are exempt from rigorous First Amendment scrutiny. The criminal prohibition of a false statement made to Government officials in communications concerning official matters, 18 U. S. C. §1001, does not lead to the broader proposition that false statements are unprotected when made to any person, at any time, in any context. As for perjury statutes, perjured statements lack First Amendment protection not simply because they are false, but because perjury undermines the function and province of the law and threatens the integrity of judgments. Finally, there are statutes that prohibit falsely representing that one is speaking on behalf of the Government, or prohibit impersonating a Government officer. These examples, to the extent that they implicate fraud or speech integral to criminal conduct, are inapplicable here. While there may exist “some categories of speech that have been historically unprotected,” but that the Court has not yet specifically identified or discussed, United States v. Stevens, 559 U. S. ___, ___, the Government has not demonstrated that false statements should constitute a new category. Pp. 3−10. (b) The Act seeks to control and suppress all false statements on this one subject in almost limitless times and settings without regard to whether the lie was made for the purpose of material gain. Permitting the Government to decree this speech to be a criminal offense would endorse government authority to compile a list of subjects about which false statements are punishable. That governmental power has no clear limiting principle. Pp. 10−11. (c) The Court applies the “most exacting scrutiny” in assessing content-based restrictions on protected speech. Turner Broadcasting System Inc. v. FCC, 512 U.S. 622, 642. The Act does not satisfy that scrutiny. While the Government’s interest in protecting the integrity of the Medal of Honor is beyond question, the First Amendment requires that there be a direct causal link between the restriction imposed and the injury to be prevented. Here, that link has not been shown. The Government points to no evidence supporting its claim that the public’s general perception of military awards is diluted by false claims such as those made by respondent. And it has not shown, and cannot show, why counterspeech, such as the ridicule respondent received online and in the press, would not suffice to achieve its interest. In addition, when the Government seeks to regulate protected speech, the restriction must be the “least restrictive means among available, effective alternatives.” Ashcroft, 542 U. S., at 666. Here, the Government could likely protect the integrity of the military awards system by creating a database of Medal winners accessible and searchable on the Internet, as some private individuals have already done. Pp. 12−18. Justice Breyer, joined by Justice Kagan, concluded that because the Stolen Valor Act, as presently drafted, works disproportionate constitutional harm, it fails intermediate scrutiny, and thus violates the First Amendment. Pp. 1−10. (a) In determining whether a statute violates the First Amendment, the Court has often found it appropriate to examine the fit between statutory ends and means, taking into account the seriousness of the speech-related harm the provision will likely cause, the nature and importance of the provision’s countervailing objectives, the extent to which the statute will tend to achieve those objectives, and whether there are other, less restrictive alternatives. “Intermediate scrutiny” describes this approach. Since false factual statements are less likely than true factual statements to make a valuable contribution to the marketplace of ideas, and the government often has good reason to prohibit such false speech, but its regulation can threaten speech-related harm, such an approach is applied here. Pp. 1−3. (b) The Act should be read as criminalizing only false factual statements made with knowledge of their falsity and with intent that they be taken as true. Although the Court has frequently said or implied that false factual statements enjoy little First Amendment protection, see, e.g., Gertz v. Robert Welch, Inc., 418 U.S. 323, 340, those statements cannot be read to mean “no protection at all.” False factual statements serve useful human objectives in many contexts. Moreover, the threat of criminal prosecution for making a false statement can inhibit the speaker from making true statements, thereby “chilling” a kind of speech that lies at the First Amendment’s heart. See id., at 340−341. And the pervasiveness of false factual statements provides a weapon to a government broadly empowered to prosecute falsity without more. Those who are unpopular may fear that the government will use that weapon selectively against them. Although there are many statutes and common-law doctrines making the utterance of certain kinds of false statements unlawful, they tend to be narrower than the Act, in that they limit the scope of their application in various ways, for example, by requiring proof of specific harm to identifiable victims. The Act lacks any such limiting features. Although it prohibits only knowing and intentional falsehoods about readily verifiable facts within the personal knowledge of the speaker, it otherwise ranges broadly, and that breadth means that it creates a significant risk of First Amendment harm. Pp. 3−8. (c) The Act nonetheless has substantial justification. It seeks to protect the interests of those who have sacrificed their health and life for their country by seeking to preserve intact the country’s recognition of that sacrifice in the form of military honors. P. 8. (d) It may, however, be possible substantially to achieve the Government’s objective in less burdensome ways. The First Amendment risks flowing from the Act’s breadth of coverage could be diminished or eliminated by a more finely tailored statute, for example, a statute that requires a showing that the false statement caused specific harm or is focused on lies more likely to be harmful or on contexts where such lies are likely to cause harm. Pp. 8−10. Kennedy, J., announced the judgment of the Court and delivered an opinion, in which Roberts, C. J., and Ginsburg and Sotomayor, JJ., joined. Breyer, J., filed an opinion concurring in the judgment, in which Kagan, J., joined. Alito, J., filed a dissenting opinion, in which Scalia and Thomas, JJ., joined.

; obscenity, see, e.g., Miller v. California, 413 U.S. 15 (1973); defamation, see, e.g., New York Times Co. v. Sullivan, 376 U.S. 254 (1964) (providing substantial protection for speech about public figures); Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974) (imposing some limits on liability for defaming a private figure); speech integral to criminal conduct, see, e.g., Giboney v. Empire Storage & Ice Co., 336 U.S. 490 (1949); so-called “fighting words,” see Chaplinsky v. New Hampshire, 315 U.S. 568 (1942); child pornography, see New York v. Ferber, 458 U.S. 747 (1982); fraud, see Virginia Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 771 (1976); true threats, see Watts v. United States, 394 U.S. 705 (1969) (per curiam); and speech presenting some grave and imminent threat the government has the power to prevent, see Near v. Minnesota ex rel. Olson, 283 U.S. 697, 716 (1931), although a restriction under the last category is most difficult to sustain, see New York Times Co. v. United States, 403 U.S. 713 (1971) (per curiam). These categories have a historical foundation in the Court’s free speech tradition. The vast realm of free speech and thought always protected in our tradition can still thrive, and even be furthered, by adherence to those categories and rules. Absent from those few categories where the law allows content-based regulation of speech is any general exception to the First Amendment for false statements. This comports with the common understanding that some false statements are inevitable if there is to be an open and vigorous expression of views in public and private con-versation, expression the First Amendment seeks to guarantee. See Sullivan, supra, at 271 (“Th[e] erroneous statement is inevitable in free debate”). The Government disagrees with this proposition. It cites language from some of this Court’s precedents to support its contention that false statements have no value and hence no First Amendment protection. See also Brief for Eugene Volokh et al. as Amici Curiae 2–11. These isolated statements in some earlier decisions do not support the Government’s submission that false statements, as a general rule, are beyond constitutional protection. That conclusion would take the quoted language far from its proper context. For instance, the Court has stated “[f]alse statements of fact are particularly valueless [because] they interfere with the truth-seeking function of the marketplace of ideas,” Hustler Magazine, Inc. v. Falwell, 485 U.S. 46, 52 (1988), and that false statements “are not protected by the First Amendment in the same manner as truthful statements,” Brown v. Hartlage, 456 U.S. 45, 60–61 (1982). See also, e.g., Virginia Bd. of Pharmacy, supra, at 771 (“Untruthful speech, commercial or otherwise, has never been protected for its own sake”); Herbert v. Lando, 441 U.S. 153, 171 (1979) (“Spreading false information in and of itself carries no First Amendment credentials”); Gertz, supra, at 340 (“[T]here is no constitutional value in false statements of fact”); Garrison v. Louisiana, 379 U.S. 64, 75 (1964) (“[T]he knowingly false statement and the false statement made with reckless disregard of the truth, do not enjoy constitutional protection”). These quotations all derive from cases discussing def-amation, fraud, or some other legally cognizable harm associated with a false statement, such as an invasion of privacy or the costs of vexatious litigation. See Brief for United States 18–19. In those decisions the falsity of the speech at issue was not irrelevant to our analysis, but neither was it determinative. The Court has never endorsed the categorical rule the Government advances: that false statements receive no First Amendment protection. Our prior decisions have not confronted a measure, like the Stolen Valor Act, that targets falsity and nothing more. Even when considering some instances of defamation and fraud, moreover, the Court has been careful to instruct that falsity alone may not suffice to bring the speech outside the First Amendment. The statement must be a knowing or reckless falsehood. See Sullivan, supra, at 280 (prohibiting recovery of damages for a defamatory falsehood made about a public official unless the statement was made “with knowledge that it was false or with reckless disregard of whether it was false or not”); see also Garrison, supra, at 73 (“[E]ven when the utterance is false, the great principles of the Constitution which secure freedom of expression . . . preclude attaching adverse consequences to any except the knowing or reckless falsehood”); Illinois ex rel. Madigan v. Telemarketing Associates, Inc., 538 U.S. 600, 620 (2003) (“False statement alone does not subject a fundraiser to fraud liability”). The Government thus seeks to use this principle for a new purpose. It seeks to convert a rule that limits liability even in defamation cases where the law permits recovery for tortious wrongs into a rule that expands liability in a different, far greater realm of discourse and expression. That inverts the rationale for the exception. The requirements of a knowing falsehood or reckless disregard for the truth as the condition for recovery in certain defamation cases exists to allow more speech, not less. A rule designed to tolerate certain speech ought not blossom to become a rationale for a rule restricting it. The Government then gives three examples of regulations on false speech that courts generally have found per-missible: first, the criminal prohibition of a false statement made to a Government official, 18 U. S. C. §1001; second, laws punishing perjury; and third, prohibi-tions on the false representation that one is speaking as a Government official or on behalf of the Government, see, e.g., §912; §709. These restrictions, however, do not establish a principle that all proscriptions of false statements are exempt from exacting First Amendment scrutiny. The federal statute prohibiting false statements to Government officials punishes “whoever, in any matter within the jurisdiction of the executive, legislative, or judicial branch of the Government . . . makes any mate-rially false, fictitious, or fraudulent statement or repre-sentation.” §1001. Section 1001’s prohibition on false statements made to Government officials, in communications concerning official matters, does not lead to the broader proposition that false statements are unprotected when made to any person, at any time, in any context. The same point can be made about what the Court has confirmed is the “unquestioned constitutionality of perjury statutes,” both the federal statute, §1623, and its state-law equivalents. United States v. Grayson, 438 U.S. 41, 54 (1978). See also Konigsberg v. State Bar of Cal., 366 U.S. 36, 51, n. 10 (1961). It is not simply because perjured statements are false that they lack First Amendment protection. Perjured testimony “is at war with justice” because it can cause a court to render a “judgment not resting on truth.” In re Michael, 326 U.S. 224, 227 (1945). Perjury undermines the function and province of the law and threatens the integrity of judgments that are the basis of the legal system. See United States v. Dunnigan, 507 U.S. 87, 97 (1993) (“To uphold the integrity of our trial system . . . the constitutionality of perjury statutes is unquestioned”). Unlike speech in other contexts, testi-mony under oath has the formality and gravity necessary to remind the witness that his or her statements will be the basis for official governmental action, action that often affects the rights and liberties of others. Sworn testimony is quite distinct from lies not spoken under oath and sim-ply intended to puff up oneself. Statutes that prohibit falsely representing that one is speaking on behalf of the Government, or that prohibit im-personating a Government officer, also protect the integrity of Government processes, quite apart from merely restricting false speech. Title 18 U. S. C. §912, for ex-ample, prohibits impersonating an officer or employee of the United States. Even if that statute may not require proving an “actual financial or property loss” resulting from the deception, the statute is itself confined to “maintain[ing] the general good repute and dignity of . . . government . . . service itself.” United States v. Lepowitch, 318 U.S. 702, 704 (1943) (internal quotation marks omitted). The same can be said for prohibitions on the unauthorized use of the names of federal agencies such as the Federal Bureau of Investigation in a manner calculated to convey that the communication is approved, see §709, or using words such as “Federal” or “United States” in the collection of private debts in order to convey that the communication has official authorization, see §712. These examples, to the extent that they implicate fraud or speech integral to criminal conduct, are inapplicable here. As our law and tradition show, then, there are instances in which the falsity of speech bears upon whether it is protected. Some false speech may be prohibited even if analogous true speech could not be. This opinion does not imply that any of these targeted prohibitions are somehow vulnerable. But it also rejects the notion that false speech should be in a general category that is presumptively unprotected. Although the First Amendment stands against any “freewheeling authority to declare new categories of speech outside the scope of the First Amendment,” Stevens, 559 U. S., at ___ (slip op., at 9), the Court has acknowledged that perhaps there exist “some categories of speech that have been historically unprotected . . . but have not yet been specifically identified or discussed . . . in our case law.” Ibid. Before exempting a category of speech from the normal prohibition on content-based re-strictions, however, the Court must be presented with “per-suasive evidence that a novel restriction on content is part of a long (if heretofore unrecognized) tradition of proscription,” Brown v. Entertainment Merchants Assn., 564 U. S. ___, ___ (2011) (slip op., at 4). The Government has not demonstrated that false statements generally should constitute a new category of unprotected speech on this basis. III The probable, and adverse, effect of the Act on free- dom of expression illustrates, in a fundamental way, the reasons for the Law’s distrust of content-based speech prohibitions. The Act by its plain terms applies to a false statement made at any time, in any place, to any person. It can be assumed that it would not apply to, say, a theatrical performance. See Milkovich v. Lorain Journal Co., 497 U.S. 1, 20 (1990) (recognizing that some statements nominally purporting to contain false facts in reality “cannot reasonably be interpreted as stating actual facts about an individual” (internal quotation marks and brackets omitted)). Still, the sweeping, quite unprecedented reach of the statute puts it in conflict with the First Amendment. Here the lie was made in a public meeting, but the statute would apply with equal force to personal, whispered conversations within a home. The statute seeks to control and suppress all false statements on this one subject in almost limitless times and settings. And it does so en-tirely without regard to whether the lie was made for the purpose of material gain. See San Francisco Arts & Athletics, Inc. v. United States Olympic Comm., 483 U.S. 522, 539–540 (1987) (prohibiting a nonprofit corporation from exploiting the “commercial magnetism” of the word “Olym-pic” when organizing an athletic competition (internal quotation marks omitted)). Permitting the government to decree this speech to be a criminal offense, whether shouted from the rooftops or made in a barely audible whisper, would endorse government authority to compile a list of subjects about which false statements are punishable. That governmental power has no clear limiting principle. Our constitutional tradition stands against the idea that we need Oceania’s Ministry of Truth. See G. Orwell, Nineteen Eighty-Four (1949) (Centennial ed. 2003). Were this law to be sustained, there could be an endless list of subjects the National Government or the States could single out. Where false claims are made to effect a fraud or secure moneys or other valuable considerations, say offers of employment, it is well established that the Government may restrict speech without affronting the First Amendment. See, e.g., Virginia Bd. of Pharmacy, 425 U. S., at 771 (noting that fraudulent speech generally falls outside the protections of the First Amendment). But the Stolen Valor Act is not so limited in its reach. Were the Court to hold that the interest in truthful discourse alone is sufficient to sustain a ban on speech, absent any evidence that the speech was used to gain a material advantage, it would give government a broad censorial power unprecedented in this Court’s cases or in our constitutional tradition. The mere potential for the exercise of that power casts a chill, a chill the First Amendment cannot permit if free speech, thought, and discourse are to remain a foundation of our freedom. IV The previous discussion suffices to show that the Act conflicts with free speech principles. But even when examined within its own narrow sphere of operation, the Act cannot survive. In assessing content-based restrictions on protected speech, the Court has not adopted a free-wheeling approach, see Stevens, 559 U. S., at ___ (slip op., at 7) (“The First Amendment’s guarantee of free speech does not extend only to categories of speech that survive an ad hoc balancing of relative social costs and benefits”), but rather has applied the “most exacting scrutiny.” Turner Broadcasting System, Inc. v. FCC, 512 U.S. 622, 642 (1994). Although the objectives the Government seeks to further by the statute are not without significance, the Court must, and now does, find the Act does not satisfy exacting scrutiny. The Government is correct when it states military medals “serve the important public function of recognizing and expressing gratitude for acts of heroism and sacrifice in military service,” and also “ ‘foste[r] morale, mission accomplishment and esprit de corps’ among service members.” Brief for United States 37, 38. General George Washington observed that an award for valor would “cherish a virtuous ambition in . . . soldiers, as well as foster and encourage every species of military merit.” General Orders of George Washington Issued at Newburgh on the Hudson, 1782–1783 (Aug. 7, 1782), p. 30 (E. Boynton ed. 1883). Time has not diminished this idea. In periods of war and peace alike public recognition of valor and noble sacrifice by men and women in uniform reinforces the pride and national resolve that the military relies upon to fulfill its mission. These interests are related to the integrity of the military honors system in general, and the Congressional Medal of Honor in particular. Although millions have served with brave resolve, the Medal, which is the highest military award for valor against an enemy force, has been given just 3,476 times. Established in 1861, the Medal is reserved for those who have distinguished themselves “conspicuously by gallantry and intrepidity at the risk of his life above and beyond the call of duty.” 10 U. S. C. §§3741 (Army), 6241 (Navy and Marine Corps), 8741 (Air Force), 14 U. S. C. §491 (Coast Guard). The stories of those who earned the Medal inspire and fascinate, from Dakota Meyer who in 2009 drove five times into the midst of a Taliban ambush to save 36 lives, see Curtis, President Obama Awards Medal of Honor to Dakota Meyer, The White House Blog (Sept. 15, 2011) (all Internet materials as visited June 25, 2012, and available in Clerk of Court’s case file); to Desmond Doss who served as an army medic on Okinawa and on June 5, 1945, rescued 75 fellow soldiers, and who, after being wounded, gave up his own place on a stretcher so others could be taken to safety, see America’s Heroes 88–90 (J. Willbanks ed. 2011); to William Carney who sustained multiple gunshot wounds to the head, chest, legs, and arm, and yet carried the flag to ensure it did not touch the ground during the Union army’s assault on Fort Wagner in July 1863, id., at 44–45. The rare acts of courage the Medal celebrates led President Truman to say he would “rather have that medal round my neck than . . . be president of the United States.” Truman Gives No. 1 Army Medal to 15 Heroes, Washington Post, Oct. 13, 1945, p. 5. The Government’s interest in protecting the integrity of the Medal of Honor is beyond question. But to recite the Government’s compelling interests is not to end the matter. The First Amendment requires that the Government’s chosen restriction on the speech at issue be “actually necessary” to achieve its interest. En-tertainment Merchants Assn., 564 U. S., at ___ (slip op., at 12). There must be a direct causal link between the restriction imposed and the injury to be prevented. See ibid. The link between the Government’s interest in protecting the integrity of the military honors system and the Act’s restriction on the false claims of liars like respondent has not been shown. Although appearing to concede that “an isolated misrepresentation by itself would not tarnish the meaning of military honors,” the Government asserts it is “common sense that false representations have the tendency to dilute the value and meaning of military awards,” Brief for United States 49, 54. It must be acknowledged that when a pretender claims the Medal to be his own, the lie might harm the Government by demeaning the high purpose of the award, diminishing the honor it confirms, and creating the appearance that the Medal is awarded more often than is true. Furthermore, the lie may offend the true holders of the Medal. From one perspective it in-sults their bravery and high principles when falsehood puts them in the unworthy company of a pretender. Yet these interests do not satisfy the Government’s heavy burden when it seeks to regulate protected speech. See United States v. Playboy Entertainment Group, Inc., 529 U.S. 803, 818 (2000). The Government points to no evidence to support its claim that the public’s general perception of military awards is diluted by false claims such as those made by Alvarez. Cf. Entertainment Merchants Assn., supra, at ___–___ (slip op., at 12–13) (analyzing and rejecting the findings of research psychologists demonstrating the causal link between violent video games and harmful effects on children). As one of the Government’s amici notes “there is nothing that charlatans such as Xavier Alvarez can do to stain [the Medal winners’] honor.” Brief for Veterans of Foreign Wars of the United States et al. as Amici Curiae 1. This general proposition is sound, even if true holders of the Medal might experience anger and frustration. The lack of a causal link between the Government’s stated interest and the Act is not the only way in which the Act is not actually necessary to achieve the Government’s stated interest. The Government has not shown, and cannot show, why counterspeech would not suffice to achieve its interest. The facts of this case indicate that the dynamics of free speech, of counterspeech, of refutation, can overcome the lie. Respondent lied at a public meeting. Even before the FBI began investigating him for his false statements “Alvarez was perceived as a phony,” 617 F. 3d, at 1211. Once the lie was made public, he was ridiculed online, see Brief for Respondent 3, his actions were reported in the press, see Ortega, Alvarez Again Denies Claim, Ontario, CA, Inland Valley Daily Bulletin (Sept. 27, 2007), and a fellow board member called for his resignation, see, e.g., Bigham, Water District Rep Requests Alvarez Resign in Wake of False Medal Claim, San Bernardino Cty., CA, The Sun (May 21, 2008). There is good reason to believe that a similar fate would befall other false claimants. See Brief for Reporters Committee for Freedom of the Press et al. as Amici Curiae 30–33 (listing numerous examples of public exposure of false claimants). Indeed, the outrage and contempt expressed for respondent’s lies can serve to reawaken and reinforce the public’s respect for the Medal, its recipients, and its high purpose. The acclaim that recipients of the Congressional Medal of Honor receive also casts doubt on the proposition that the public will be misled by the claims of charlatans or become cynical of those whose heroic deeds earned them the Medal by right. See, e.g., Well Done, Washington Post, Feb. 5, 1943, p. 8 (reporting on Pres-ident Roosevelt’s awarding the Congressional Medal of Honor to Maj. Gen. Alexander Vandegrift); Devroy, Medal of Honor Given to 2 Killed in Somalia, Washington Post, May 24, 1994, p. A6 (reporting on President Clinton’s awarding the Congressional Medal of Honor to two special forces soldiers killed during operations in Somalia). The remedy for speech that is false is speech that is true. This is the ordinary course in a free society. The response to the unreasoned is the rational; to the uninformed, the enlightened; to the straight-out lie, the simple truth. See Whitney v. California, 274 U.S. 357, 377 (1927) (Brandeis, J., concurring) (“If there be time to expose through discussion the falsehood and fallacies, to avert the evil by the processes of education, the remedy to be ap-plied is more speech, not enforced silence”). The theory of our Constitution is “that the best test of truth is the power of the thought to get itself accepted in the competition of the market,” Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting). The First Amendment itself ensures the right to respond to speech we do not like, and for good reason. Freedom of speech and thought flows not from the beneficence of the state but from the inalienable rights of the person. And suppression of speech by the government can make exposure of falsity more difficult, not less so. Society has the right and civic duty to engage in open, dynamic, rational discourse. These ends are not well served when the government seeks to orchestrate public discussion through content-based mandates. Expressing its concern that counterspeech is insuf- ficient, the Government responds that because “some military records have been lost . . . some claims [are] un-verifiable,” Brief for United States 50. This proves little, however; for without verifiable records, successful criminal prosecution under the Act would be more difficult in any event. So, in cases where public refutation will not serve the Government’s interest, the Act will not either. In addition, the Government claims that “many [false claims] will remain unchallenged.” Id., at 55. The Government provides no support for the contention. And in any event, in order to show that public refutation is not an adequate alternative, the Government must demonstrate that unchallenged claims undermine the public’s perception of the military and the integrity of its awards system. This showing has not been made. It is a fair assumption that any true holders of the Medal who had heard of Alvarez’s false claims would have been fully vindicated by the community’s expression of outrage, showing as it did the Nation’s high regard for the Medal. The same can be said for the Government’s interest. The American people do not need the assistance of a government prosecution to express their high regard for the special place that military heroes hold in our tradi-tion. Only a weak society needs government protection or intervention before it pursues its resolve to preserve the truth. Truth needs neither handcuffs nor a badge for its vindication. In addition, when the Government seeks to regulate protected speech, the restriction must be the “least restrictive means among available, effective alternatives.” Ashcroft, 542 U. S., at 666. There is, however, at least one less speech-restrictive means by which the Government could likely protect the integrity of the military awards system. A Government-created database could list Congressional Medal of Honor winners. Were a database accessible through the Internet, it would be easy to verify and expose false claims. It appears some private individuals have already created databases similar to this, see Brief for Respondent 25, and at least one data- base of past winners is online and fully searchable, see Congressional Medal of Honor Society, Full Archive, http://www.cmohs.org/recipient-archive.php. The Solicitor General responds that although Congress and the Department of Defense investigated the feasibility of establishing a database in 2008, the Government “concluded that such a database would be impracticable and insuf-ficiently comprehensive.” Brief for United States 55. Without more explanation, it is difficult to assess the Gov-ernment’s claim, especially when at least one database of Congressional Medal of Honor winners already exists. The Government may have responses to some of these criticisms, but there has been no clear showing of the necessity of the statute, the necessity required by exacting scrutiny. * * * The Nation well knows that one of the costs of the First Amendment is that it protects the speech we detest as well as the speech we embrace. Though few might find respondent’s statements anything but contemptible, his right to make those statements is protected by the Constitution’s guarantee of freedom of speech and expression. The Stolen Valor Act infringes upon speech protected by the First Amendment. The judgment of the Court of Appeals is affirmed. It is so ordered.

566.US.2011_11-139

Ordinarily, the Government must assess a deficiency against a tax- payer within “3 years after the return was filed,” 26 U. S. C. §6501(a), but that period is extended to 6 years when a taxpayer “omits from gross income an amount properly includible therein which is in excess of 25 percent of the amount of gross income stated in the return,” §6501(e)(1)(A). Respondent taxpayers overstated the basis of certain property that they had sold. As a result, their returns understated the gross income they received from the sale by an amount in excess of 25%. The Commissioner asserted the deficiency outside the 3-year limitations period but within the 6-year period. The Fourth Circuit concluded that the taxpayers’ overstatements of basis, and resulting understatements of gross income, did not trigger the extended limitations period. Held: The judgment is affirmed. 634 F.3d 249, affirmed. Justice Breyer delivered the opinion of the Court, except as to Part IV–C, concluding that §6501(e)(1)(A) does not apply to an overstatement of basis. Pp. 2–8. (a) In Colony, Inc. v. Commissioner, 357 U.S. 28, the Court interpreted a provision of the Internal Revenue Code of 1939 containing language materially indistinguishable from the language at issue here, holding that taxpayer misstatements that overstate the basis in property do not fall within the statute’s scope. The Court recognized that such an overstatement wrongly understates a taxpayer’s income, but concluded that the phrase “omits . . . an amount” limited the statute’s scope to situations in which specific receipts are left out of the computation of gross income. The Court also noted that while the statute’s language was not “unambiguous,” id., at 33, the statutory history showed that Congress intended to restrict the extended limitations period to situations that did not include overstatements of basis. Finally, the Court found its conclusion “in harmony with the unambiguous language of §6501(e)(1)(A),” id., at 37, the provision enacted as part of the Internal Revenue Code of 1954 and applicable here. Pp. 2–4. (b) Colony determines the outcome of this case. The operative language of the 1939 provision and the provision at issue is identical. It would be difficult to give the same language here a different interpretation without overruling Colony, a course of action stare decisis counsels against. John R. Sand & Gravel Co. v. United States, 552 U.S. 130, 139. The Government suggests that differences in other nearby parts of the 1954 Code favor a different interpretation than the one adopted in Colony. However, its arguments are too fragile to bear the significant weight it seeks to place upon them. Pp. 4–7. (c) The Court also rejects the Government’s argument that a recently promulgated Treasury Regulation interpreting the statute’s operative language in its favor should be granted deference under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837. See National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U.S. 967, 982. Colony has already interpreted the statute, and there is no longer any different construction that is consistent with Colony and available for adoption by the agency. Pp. 7–8. Breyer, J. delivered the opinion of the Court, except as to Part IV–C. Roberts, C. J., and Thomas and Alito, JJ., joined that opinion in full, and Scalia, J., joined except for Part IV–C. Scalia, J., filed an opinion concurring in part and concurring in the judgment. Kennedy, J., filed a dissenting opinion, in which Ginsburg, Sotomayor, and Kagan, JJ., joined.

except as to Part IV–C. Ordinarily, the Government must assess a deficiency against a taxpayer within “3 years after the return was filed.” 26 U. S. C. §6501(a) (2000 ed.). The 3-year period is extended to 6 years, however, when a taxpayer “omits from gross income an amount properly includible therein which is in excess of 25 percent of the amount of gross income stated in the return.” §6501(e)(1)(A) (emphasis added). The question before us is whether this latter provision applies (and extends the ordinary 3-year limitations period) when the taxpayer overstates his basis in property that he has sold, thereby understating the gain that he received from its sale. Following Colony, Inc. v. Commissioner, 357 U.S. 28 (1958), we hold that the provision does not apply to an overstatement of basis. Hence the 6-year period does not apply. I For present purposes the relevant underlying circumstances are not in dispute. We consequently assume that (1) the respondent taxpayers filed their relevant tax returns in April 2000; (2) the returns overstated the basis of certain property that the taxpayers had sold; (3) as a result the returns understated the gross income that the taxpayers received from the sale of the property; and (4) the understatement exceeded the statute’s 25% threshold. We also take as undisputed that the Commissioner asserted the relevant deficiency within the extended 6-year limitations period, but outside the default 3-year period. Thus, unless the 6-year statute of limitations applies, the Government’s efforts to assert a tax deficiency came too late. Our conclusion—that the extended limitations period does not apply—follows directly from this Court’s earlier decision in Colony. II In Colony this Court interpreted a provision of the In- ternal Revenue Code of 1939, the operative language of which is identical to the language now before us. The Commissioner there had determined “that the taxpayer had understated the gross profits on the sales of certain lots of land for residential purposes as a result of having overstated the ‘basis’ of such lots by erroneously including in their cost certain unallowable items of development expense.” Id., at 30. The Commissioner’s assessment came after the ordinary 3-year limitations period had run. And, it was consequently timely only if the taxpayer, in the words of the 1939 Code, had “omit[ted] from gross income an amount properly includible therein which is in excess of 25 per cen- tum of the amount of gross income stated in the return . . . .” 26 U. S. C. §275(c) (1940 ed.). The Code provision ap- plicable to this case, adopted in 1954, contains materially indistinguishable language. See §6501(e)(1)(A) (2000 ed.) (same, but replacing “per centum” with “percent”). See also Appendix, infra. In Colony this Court held that taxpayer misstatements, overstating the basis in property, do not fall within the scope of the statute. But the Court recognized the Commissioner’s contrary argument for inclusion. 357 U. S., at 32. Then as now, the Code itself defined “gross income” in this context as the difference between gross revenue (often the amount the taxpayer received upon selling the prop- erty) and basis (often the amount the taxpayer paid for the property). Compare 26 U. S. C. §§22, 111 (1940 ed.) with §§61(a)(3), 1001(a) (2000 ed.). And, the Commissioner pointed out, an overstatement of basis can diminish the “amount” of the gain just as leaving the item entirely off the return might do. 357 U. S., at 32. Either way, the error wrongly understates the taxpayer’s income. But, the Court added, the Commissioner’s argument did not fully account for the provision’s language, in particular the word “omit.” The key phrase says “omits . . . an amount.” The word “omits” (unlike, say, “reduces” or “un- derstates”) means “ ‘[t]o leave out or unmentioned; not to insert, include, or name.’ ” Ibid. (quoting Webster’s New International Dictionary (2d ed. 1939)). Thus, taken literally, “omit” limits the statute’s scope to situations in which specific receipts or accruals of income are left out of the computation of gross income; to inflate the basis, however, is not to “omit” a specific item, not even of profit. While finding this latter interpretation of the language the “more plausibl[e],” the Court also noted that the language was not “unambiguous.” Colony, 357 U. S., at 33. It then examined various congressional Reports discussing the relevant statutory language. It found in those Reports “persuasive indications that Congress merely had in mind failures to report particular income receipts and accruals, and did not intend the [extended] limitation to apply whenever gross income was understated . . . .” Id., at 35. This “history,” the Court said, “shows . . . that the Congress intended an exception to the usual three-year statute of limitations only in the restricted type of situation already described,” a situation that did not include overstatements of basis. Id., at 36. The Court wrote that Congress, in enacting the provision, “manifested no broader purpose than to give the Commissioner an additional two [now three] years to investigate tax returns in cases where, because of a taxpayer’s omission to report some taxable item, the Commissioner is at a special disadvantage . . . [because] the return on its face provides no clue to the existence of the omitted item. . . . [W]hen, as here [i.e., where the overstatement of basis is at issue], the understatement of a tax arises from an error in reporting an item disclosed on the face of the return the Commissioner is at no such disadvantage . . . whether the error be one affecting ‘gross income’ or one, such as overstated deductions, affecting other parts of the return.” Ibid. (emphasis added). Finally, the Court noted that Congress had recently enacted the Internal Revenue Code of 1954. And the Court observed that “the conclusion we reach is in har- mony with the unambiguous language of §6501(e)(1)(A),” id., at 37, i.e., the provision relevant in this present case. III In our view, Colony determines the outcome in this case. The provision before us is a 1954 reenactment of the 1939 provision that Colony interpreted. The operative language is identical. It would be difficult, perhaps impossible, to give the same language here a different interpretation without effectively overruling Colony, a course of action that basic principles of stare decisis wisely counsel us not to take. John R. Sand & Gravel Co. v. United States, 552 U.S. 130, 139 (2008) (“[S]tare decisis in respect to statu- tory interpretation has special force, for Congress remains free to alter what we have done” (internal quotation marks omitted)); Patterson v. McLean Credit Union, 491 U.S. 164, 172–173 (1989). The Government, in an effort to convince us to interpret the operative language before us differently, points to differences in other nearby parts of the 1954 Code. It suggests that these differences counsel in favor of a different interpretation than the one adopted in Colony. For example, the Government points to a new provision, §6501(e)(1)(A)(i), which says: “In the case of a trade or business, the term ‘gross income’ means the total of the amounts received or accrued from the sale of goods or services (if such amounts are required to be shown on the return) prior to the diminution by the cost of such sales or services.” If the section’s basic phrase “omi[ssion] from gross income” does not apply to overstatements of basis (which is what Colony held), then what need would there be for clause (i), which leads to the same result in a specific subset of cases? And why, the Government adds, does a later paragraph, referring to gifts and estates, speak of a taxpayer who “omits . . . items includible in [the] gross estate”? See §6501(e)(2) (emphasis added). By speaking of “items” there does it not imply that omission of an “amount” cov- ers more than omission of individual items—indeed that it includes overstatements of basis, which, after all, di- minish the amount of the profit that should have been re- ported as gross income? In our view, these points are too fragile to bear the sig- nificant argumentative weight the Government seeks to place upon them. For example, at least one plausible reason why Congress might have added clause (i) has nothing to do with any desire to change the meaning of the general rule. Rather when Congress wrote the 1954 Code (prior to Colony), it did not yet know how the Court would interpret the provision’s operative language. At least one lower court had decided that the provision did not apply to overstatements about the cost of goods that a business later sold. See Uptegrove Lumber Co. v. Commissioner, 204 F.2d 570 (CA3 1953). But see Reis v. Commissioner, 142 F.2d 900, 902–903 (CA6 1944). And Congress could well have wanted to ensure that, come what may in the Supreme Court, Uptegrove’s interpretation would remain the law where a “trade or business” was at issue. Nor does our interpretation leave clause (i) without work to do. TRW Inc. v. Andrews, 534 U.S. 19, 31 (2001) (noting canon that statutes should be read to avoid making any provision “superfluous, void, or insignificant” (internal quotation marks omitted)). That provision also explains how to calculate the denominator for purposes of determining whether a conceded omission amounts to 25% of “gross income.” For example, it tells us that a merchant who fails to include $10,000 of revenue from sold goods has not met the 25% test if total revenue is more than $40,000, regardless of the cost paid by the merchant to acquire those goods. But without clause (i), the general statutory definition of “gross income” requires subtracting the cost from the sales price. See 26 U. S. C. §§61(a)(3), 1012. Under such a definition of “gross income,” the cal- culation would take (1) total revenue from sales, $40,000, minus (2) “the cost of such sales,” say, $25,000. The $10,000 of revenue would thus amount to 67% of the “gross income” of $15,000. And the clause does this work in respect to omissions from gross income irrespective of our interpretation regarding overstatements of basis. The Government’s argument about subsection (e)(2)’s use of the word “item” instead of “amount” is yet weaker. The Court in Colony addressed a similar argument about the word “amount.” It wrote: “The Commissioner states that the draftsman’s use of the word ‘amount’ (instead of, for example, ‘item’) suggests a concentration on the quantitative aspect of the error—that is whether or not gross income was understated by as much as 25%.” 357 U. S., at 32. But the Court, while recognizing the Commissioner’s logic, rejected the argument (and the significance of the word “amount”) as insufficient to prove the Commissioner’s conclusion. And the addition of the word “item” in a different subsection similarly fails to exert an interpretive force sufficiently strong to affect our conclusion. The word’s appearance in subsection (e)(2), we concede, is new. But to rely in the case before us on this solitary word change in a different subsection is like hoping that a new batboy will change the outcome of the World Series. IV A Finally, the Government points to Treasury Regulation §301.6501(e)–1, which was promulgated in final form in December 2010. See 26 CFR §301.6501(e)–1 (2011). The regulation, as relevant here, departs from Colony and interprets the operative language of the statute in the Government’s favor. The regulation says that “an un- derstated amount of gross income resulting from an overstatement of unrecovered cost or other basis constitutes an omission from gross income.” §301.6501(e)–1(a)(1)(iii). In the Government’s view this new regulation in effect overturns Colony’s interpretation of this statute. The Government points out that the Treasury Regulation constitutes “an agency’s construction of a statute which it administers.” Chevron, U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842 (1984). See also Mayo Foundation for Medical Ed. and Research v. United States, 562 U. S. ___ (2011) (applying Chevron in the tax context). The Court has written that a “court’s prior judicial construction of a statute trumps an agency construction otherwise entitled to Chevron deference only if the prior court decision holds that its construction follows from the unambiguous terms of the statute . . . .” National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U.S. 967, 982 (2005) (emphasis added). And, as the Government notes, in Colony itself the Court wrote that “it cannot be said that the language is unambiguous.” 357 U. S., at 33. Hence, the Government concludes, Colony cannot govern the outcome in this case. The question, rather, is whether the agency’s construction is a “permissible construction of the statute.” Chevron, supra, at 843. And, since the Government argues that the regulation embodies a reasonable, hence permissible, construction of the statute, the Government believes it must win. B We do not accept this argument. In our view, Colony has already interpreted the statute, and there is no longer any different construction that is consistent with Colony and available for adoption by the agency. C The fatal flaw in the Government’s contrary argument is that it overlooks the reason why Brand X held that a “prior judicial construction,” unless reflecting an “unambiguous” statute, does not trump a different agency construction of that statute. 545 U. S., at 982. The Court reveals that reason when it points out that “it is for agencies, not courts, to fill statutory gaps.” Ibid. The fact that a statute is unambiguous means that there is “no gap for the agency to fill” and thus “no room for agency discre- tion.” Id., at 982–983. In so stating, the Court sought to encapsulate what earlier opinions, including Chevron, made clear. Those opinions identify the underlying interpretive problem as that of deciding whether, or when, a particular statute in effect delegates to an agency the power to fill a gap, thereby implicitly taking from a court the power to void a reasonable gap-filling interpretation. Thus, in Chevron the Court said that, when “Congress has explicitly left a gap for the agency to fill, there is an express delegation of authority to the agency to elucidate a specific provision of the statute by regulation. . . . Sometimes the legislative delegation to an agency on a particular question is implicit rather than explicit. [But in either instance], a court may not substitute its own construction of a statutory provision for a reasonable interpretation made by the administrator of an agency.” 467 U. S., at 843–844. See also United States v. Mead Corp., 533 U.S. 218, 229 (2001); Smiley v. Citibank (South Dakota), N. A., 517 U.S. 735, 741 (1996); INS v. Cardoza-Fonseca, 480 U.S. 421, 448 (1987); Morton v. Ruiz, 415 U.S. 199, 231 (1974). Chevron and later cases find in unambiguous language a clear sign that Congress did not delegate gap-filling authority to an agency; and they find in ambiguous language at least a presumptive indication that Congress did delegate that gap-filling authority. Thus, in Chevron the Court wrote that a statute’s silence or ambiguity as to a particular issue means that Congress has not “directly addressed the precise question at issue” (thus likely delegating gap-filling power to the agency). 467 U. S., at 843. In Mead the Court, describing Chevron, explained: “Congress . . . may not have expressly delegated authority or responsibility to implement a particular provision or fill a particular gap. Yet it can still be apparent from the agency’s generally conferred authority and other statutory circumstances that Congress would expect the agency to be able to speak with the force of law when it addresses ambiguity in the statute or fills a space in the enacted law, even one about which Congress did not actually have an intent as to a particular result.” 533 U. S., at 229 (internal quotation marks omitted). Chevron added that “[i]f a court, employing traditional tools of statutory construction, ascertains that Congress had an intention on the precise question at issue, that intention is the law and must be given effect.” 467 U. S., at 843, n. 9 (emphasis added). As the Government points out, the Court in Colony stated that the statutory language at issue is not “unambiguous.” 357 U. S., at 33. But the Court decided that case nearly 30 years before it decided Chevron. There is no reason to believe that the linguistic ambiguity noted by Colony reflects a post-Chevron conclusion that Congress had delegated gap-filling power to the agency. At the same time, there is every reason to believe that the Court thought that Congress had “directly spoken to the question at hand,” and thus left “[no] gap for the agency to fill.” Chevron, supra, at 842–843. For one thing, the Court said that the taxpayer had the better side of the textual argument. Colony, 357 U. S., at 33. For another, its examination of legislative history led it to believe that Congress had decided the question definitively, leaving no room for the agency to reach a contrary result. It found in that history “persuasive indications” that Congress intended overstatements of basis to fall outside the statute’s scope, and it said that it was satisfied that Congress “intended an exception . . . only in the restricted type of situation” it had already described. Id., at 35–36. Further, it thought that the Commissioner’s inter- pretation (the interpretation once again advanced here) would “create a patent incongruity in the tax law.” Id., at 36–37. And it reached this conclusion despite the fact that, in the years leading up to Colony, the Commissioner had consistently advocated the opposite in the circuit courts. See, e.g., Uptegrove, 204 F.2d 570; Reis, 142 F.2d 900; Goodenow v. Commisioner, 238 F.2d 20 (CA8 1956); American Liberty Oil Co. v. Commissioner, 1 T.C. 386 (1942). Cf. Slaff v. Commisioner, 220 F.2d 65 (CA9 1955); Davis v. Hightower, 230 F.2d 549 (CA5 1956). Thus, the Court was aware it was rejecting the expert opinion of the Commissioner of Internal Revenue. And finally, after completing its analysis, Colony found its interpretation of the 1939 Code “in harmony with the [now] unambiguous language” of the 1954 Code, which at a minimum suggests that the Court saw nothing in the 1954 Code as inconsistent with its conclusion. 357 U. S., at 37. It may be that judges today would use other methods to determine whether Congress left a gap to fill. But that is beside the point. The question is whether the Court in Colony concluded that the statute left such a gap. And, in our view, the opinion (written by Justice Harlan for the Court) makes clear that it did not. Given principles of stare decisis, we must follow that interpretation. And there being no gap to fill, the Government’s gap-filling regulation cannot change Colony’s interpretation of the statute. We agree with the taxpayer that overstatements of basis, and the resulting understatement of gross income, do not trigger the extended limitations period of §6501(e)(1)(A). The Court of Appeals reached the same conclusion. See 634 F.3d 249 (CA4 2011). And its judgment is affirmed. It is so ordered. APPENDIX We reproduce the applicable sections of the two relevant versions of the U. S. Code below. Section 6501 was amended and reorganized in 2010. See Hiring Incentives to Restore Employment Act, §513, 124Stat. 111. But the parties agree that the amendments do not affect this case. We therefore have referred to, and reproduce here, the section as it appears in the 2000 edition of the U. S. Code. Title 26 U. S. C. §275 (1940 ed.) “Period of limitation upon assessment and collection. . . . . . “(a) General rule. “The amount of income taxes imposed by this chapter shall be assessed within three years after the return was filed, and no proceeding in court without assessment for the collection of such taxes shall be begun after the expiration of such period. . . . . . “(c) Omission from gross income. “If the taxpayer omits from gross income an amount properly includible therein which is in excess of 25 per centum of the amount of gross income stated in the return, the tax may be assessed, or a proceeding in court for the collection of such tax may be begun without assessment, at any time within 5 years after the return was filed.” Title 26 U. S. C. §6501 (2000 ed.) “Limitations on assessment and collection. “(a) General rule “Except as otherwise provided in this section, the amount of any tax imposed by this title shall be assessed within 3 years after the return was filed (whether or not such return was filed on or after the date prescribed) or, if the tax is payable by stamp, at any time after such tax became due and before the expiration of 3 years after the date on which any part of such tax was paid, and no proceeding in court without assessment for the collection of such tax shall be begun after the expiration of such period. . . . . . . . . “(e) Substantial omission of items “Except as otherwise provided in subsection (c)— “(1) Income taxes “In the case of any tax imposed by subtitle A— “(A) General rule “If the taxpayer omits from gross income an amount properly includible therein which is in excess of 25 percent of the amount of gross income stated in the return, the tax may be assessed, or a proceeding in court for the collection of such tax may be begun without assessment, at any time within 6 years after the return was filed. For purposes of this subparagraph— “(i) In the case of a trade or business, the term ‘gross income’ means the total of the amounts received or accrued from the sale of goods or services (if such amounts are required to be shown on the return) prior to diminution by the cost of such sales or services; and “(ii) In determining the amount omitted from gross income, there shall not be taken into account any amount which is omitted from gross income stated in the return if such amount is disclosed in the return, or in a statement attached to the return, in a manner adequate to apprise the Secretary of the nature and amount of such item. . . . . . “(2) Estate and gift taxes “In the case of a return of estate tax under chapter 11 or a return of gift tax under chapter 12, if the taxpayer omits from the gross estate or from the total amount of the gifts made during the period for which the return was filed items includible in such gross estate or such total gifts, as the case may be, as exceed in amount 25 percent of the gross estate stated in the return or the total amount of gifts stated in the return, the tax may be assessed, or a proceeding in court for the collection of such tax may be begun without assessment, at any time within 6 years after the return was filed. . . .”

565.US.2011_10-1259

The Government obtained a search warrant permitting it to install a Global-Positioning-System (GPS) tracking device on a vehicle registered to respondent Jones’s wife. The warrant authorized installation in the District of Columbia and within 10 days, but agents installed the device on the 11th day and in Maryland. The Government then tracked the vehicle’s movements for 28 days. It subsequently secured an indictment of Jones and others on drug trafficking conspiracy charges. The District Court suppressed the GPS data obtained while the vehicle was parked at Jones’s residence, but held the remaining data admissible because Jones had no reasonable expectation of privacy when the vehicle was on public streets. Jones was convicted. The D. C. Circuit reversed, concluding that admission of the evidence obtained by warrantless use of the GPS device violated the Fourth Amendment. Held: The Government’s attachment of the GPS device to the vehicle, and its use of that device to monitor the vehicle’s movements, constitutes a search under the Fourth Amendment. Pp. 3–12. (a) The Fourth Amendment protects the “right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures.” Here, the Government’s physical intrusion on an “effect” for the purpose of obtaining information constitutes a “search.” This type of encroachment on an area enumerated in the Amendment would have been considered a search within the meaning of the Amendment at the time it was adopted. Pp. 3–4. (b) This conclusion is consistent with this Court’s Fourth Amendment jurisprudence, which until the latter half of the 20th century was tied to common-law trespass. Later cases, which have deviated from that exclusively property-based approach, have applied the analysis of Justice Harlan’s concurrence in Katz v. United States, 389 U.S. 347, which said that the Fourth Amendment protects a person’s “reasonable expectation of privacy,” id., at 360. Here, the Court need not address the Government’s contention that Jones had no “reasonable expectation of privacy,” because Jones’s Fourth Amendment rights do not rise or fall with the Katz formulation. At bottom, the Court must “assur[e] preservation of that degree of privacy against government that existed when the Fourth Amendment was adopted.” Kyllo v. United States, 533 U.S. 27, 34. Katz did not repudiate the understanding that the Fourth Amendment embodies a particular concern for government trespass upon the areas it enumerates. The Katz reasonable-expectation-of-privacy test has been added to, but not substituted for, the common-law trespassory test. See Alderman v. United States, 394 U.S. 165, 176; Soldal v. Cook County, 506 U.S. 56, 64. United States v. Knotts, 460 U.S. 276, and United States v. Karo, 468 U. S. 705—post-Katz cases rejecting Fourth Amendment challenges to “beepers,” electronic tracking devices representing another form of electronic monitoring—do not foreclose the conclusion that a search occurred here. New York v. Class, 475 U.S. 106, and Oliver v. United States, 466 U.S. 170, also do not support the Government’s position. Pp. 4–12. (c) The Government’s alternative argument—that if the attachment and use of the device was a search, it was a reasonable one—is forfeited because it was not raised below. P. 12. 615 F.3d 544, affirmed. Scalia, J., delivered the opinion of the Court, in which Roberts, C. J., and Kennedy, Thomas, and Sotomayor, JJ., joined. Sotomayor, J., filed a concurring opinion. Alito, J., filed an opinion concurring in the judgment, in which Ginsburg, Breyer, and Kagan, JJ., joined.

We decide whether the attachment of a Global-Positioning-System (GPS) tracking device to an individual’s vehicle, and subsequent use of that device to monitor the vehicle’s movements on public streets, constitutes a search or seizure within the meaning of the Fourth Amendment. I In 2004 respondent Antoine Jones, owner and operator of a nightclub in the District of Columbia, came under suspicion of trafficking in narcotics and was made the target of an investigation by a joint FBI and Metropolitan Police Department task force. Officers employed various investigative techniques, including visual surveillance of the nightclub, installation of a camera focused on the front door of the club, and a pen register and wiretap covering Jones’s cellular phone. Based in part on information gathered from these sources, in 2005 the Government applied to the United States District Court for the District of Columbia for a warrant authorizing the use of an electronic tracking device on the Jeep Grand Cherokee registered to Jones’s wife. A warrant issued, authorizing installation of the de- vice in the District of Columbia and within 10 days. On the 11th day, and not in the District of Columbia but in Maryland, [ 1 ] agents installed a GPS tracking device on the undercarriage of the Jeep while it was parked in a public parking lot. Over the next 28 days, the Government used the device to track the vehicle’s movements, and once had to replace the device’s battery when the vehicle was parked in a different public lot in Maryland. By means of signals from multiple satellites, the device established the vehicle’s location within 50 to 100 feet, and communicated that location by cellular phone to a Government computer. It relayed more than 2,000 pages of data over the 4-week period. The Government ultimately obtained a multiple-count indictment charging Jones and several alleged co-conspirators with, as relevant here, conspiracy to distribute and possess with intent to distribute five kilograms or more of cocaine and 50 grams or more of cocaine base, in violation of 21 U. S. C. §§841 and 846. Before trial, Jones filed a motion to suppress evidence obtained through the GPS device. The District Court granted the motion only in part, suppressing the data obtained while the vehicle was parked in the garage adjoining Jones’s residence. 451 F. Supp. 2d 71, 88 (2006). It held the remaining data admissible, because “ ‘[a] person traveling in an automobile on public thoroughfares has no reasonable expectation of privacy in his movements from one place to another.’ ” Ibid. (quoting United States v. Knotts, 460 U.S. 276, 281 (1983) ). Jones’s trial in October 2006 produced a hung jury on the conspiracy count. In March 2007, a grand jury returned another indictment, charging Jones and others with the same conspir- acy. The Government introduced at trial the same GPS-derived locational data admitted in the first trial, which connected Jones to the alleged conspirators’ stash house that contained $850,000 in cash, 97 kilograms of cocaine, and 1 kilogram of cocaine base. The jury returned a guilty verdict, and the District Court sentenced Jones to life imprisonment. The United States Court of Appeals for the District of Columbia Circuit reversed the conviction because of admission of the evidence obtained by warrantless use of the GPS device which, it said, violated the Fourth Amend- ment. United States v. Maynard, 615 F.3d 544 (2010). The D. C. Circuit denied the Government’s petition for rehearing en banc, with four judges dissenting. 625 F.3d 766 (2010). We granted certiorari, 564 U. S. ___ (2011). II A The Fourth Amendment provides in relevant part that “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated.” It is beyond dispute that a vehicle is an “effect” as that term is used in the Amendment. United States v. Chadwick, 433 U.S. 1, 12 (1977) . We hold that the Government’s installation of a GPS device on a target’s vehicle, [ 2 ] and its use of that device to monitor the vehicle’s movements, constitutes a “search.” It is important to be clear about what occurred in this case: The Government physically occupied private property for the purpose of obtaining information. We have no doubt that such a physical intrusion would have been considered a “search” within the meaning of the Fourth Amendment when it was adopted. Entick v. Carrington, 95 Eng. Rep. 807 (C. P. 1765), is a “case we have described as a ‘monument of English freedom’ ‘undoubtedly familiar’ to ‘every American statesman’ at the time the Constitution was adopted, and considered to be ‘the true and ultimate expression of constitutional law’ ” with regard to search and seizure. Brower v. County of Inyo, 489 U.S. 593, 596 (1989) (quoting Boyd v. United States, 116 U.S. 616, 626 (1886) ). In that case, Lord Camden expressed in plain terms the significance of property rights in search-and-seizure analysis: “[O]ur law holds the property of every man so sacred, that no man can set his foot upon his neighbour’s close without his leave; if he does he is a trespasser, though he does no damage at all; if he will tread upon his neighbour’s ground, he must justify it by law.” Entick, supra, at 817. The text of the Fourth Amendment reflects its close connection to property, since otherwise it would have referred simply to “the right of the people to be secure against unreasonable searches and seizures”; the phrase “in their persons, houses, papers, and effects” would have been superfluous. Consistent with this understanding, our Fourth Amendment jurisprudence was tied to common-law trespass, at least until the latter half of the 20th century. Kyllo v. United States, 533 U.S. 27, 31 (2001) ; Kerr, The Fourth Amendment and New Technologies: Constitutional Myths and the Case for Caution, 102 Mich. L. Rev. 801, 816 (2004). Thus, in Olmstead v. United States, 277 U.S. 438 (1928) , we held that wiretaps attached to telephone wires on the public streets did not constitute a Fourth Amendment search because “[t]here was no entry of the houses or offices of the defendants,” id., at 464. Our later cases, of course, have deviated from that exclusively property-based approach. In Katz v. United States, 389 U.S. 347, 351 (1967) , we said that “the Fourth Amendment protects people, not places,” and found a violation in attachment of an eavesdropping device to a public telephone booth. Our later cases have applied the analysis of Justice Harlan’s concurrence in that case, which said that a violation occurs when government officers violate a person’s “reasonable expectation of privacy,” id., at 360. See, e.g., Bond v. United States, 529 U.S. 334 (2000) ; California v. Ciraolo, 476 U.S. 207 (1986) ; Smith v. Maryland, 442 U.S. 735 (1979) . The Government contends that the Harlan standard shows that no search occurred here, since Jones had no “reasonable expectation of privacy” in the area of the Jeep accessed by Government agents (its underbody) and in the locations of the Jeep on the public roads, which were visible to all. But we need not address the Government’s contentions, because Jones’s Fourth Amendment rights do not rise or fall with the Katz formulation. At bottom, we must “assur[e] preservation of that degree of privacy against government that existed when the Fourth Amendment was adopted.” Kyllo, supra, at 34. As explained, for most of our history the Fourth Amendment was understood to embody a particular concern for government trespass upon the areas (“persons, houses, papers, and effects”) it enumerates. [ 3 ] Katz did not repudiate that understanding. Less than two years later the Court upheld defendants’ contention that the Government could not introduce against them conversations between other people obtained by warrantless placement of electronic surveillance devices in their homes. The opinion rejected the dissent’s contention that there was no Fourth Amendment violation “unless the conversational privacy of the homeowner himself is invaded.” [ 4 ] Alderman v. United States, 394 U.S. 165, 176 (1969) . “[W]e [do not] believe that Katz, by holding that the Fourth Amendment protects persons and their private conversations, was intended to withdraw any of the protection which the Amendment extends to the home . . . .” Id., at 180. More recently, in Soldal v. Cook County, 506 U.S. 56 (1992) , the Court unanimously rejected the argument that although a “seizure” had occurred “in a ‘technical’ sense” when a trailer home was forcibly removed, id., at 62, no Fourth Amendment violation occurred because law enforcement had not “invade[d] the [individuals’] privacy,” id., at 60. Katz, the Court explained, established that “property rights are not the sole measure of Fourth Amendment violations,” but did not “snuf[f] out the previously recognized protection for property.” 506 U. S., at 64. As Justice Brennan explained in his concurrence in Knotts, Katz did not erode the principle “that, when the Government does engage in physical intrusion of a constitutionally protected area in order to obtain information, that intrusion may constitute a violation of the Fourth Amendment.” 460 U. S., at 286 (opinion concurring in judgment). We have embodied that preservation of past rights in our very definition of “reasonable expectation of privacy” which we have said to be an expectation “that has a source outside of the Fourth Amendment, either by reference to concepts of real or personal property law or to understandings that are recognized and permitted by society.” Minnesota v. Carter, 525 U.S. 83, 88 (1998) (internal quotation marks omitted). Katz did not narrow the Fourth Amendment’s scope. [ 5 ] The Government contends that several of our post-Katz cases foreclose the conclusion that what occurred here constituted a search. It relies principally on two cases in which we rejected Fourth Amendment challenges to “beepers,” electronic tracking devices that represent another form of electronic monitoring. The first case, Knotts, upheld against Fourth Amendment challenge the use of a “beeper” that had been placed in a container of chloroform, allowing law enforcement to monitor the location of the container. 460 U. S., at 278. We said that there had been no infringement of Knotts’ reasonable expectation of privacy since the information obtained—the location of the automobile carrying the container on public roads, and the location of the off-loaded container in open fields near Knotts’ cabin—had been voluntarily conveyed to the public. [ 6 ] Id., at 281–282. But as we have discussed, the Katz reasonable-expectation-of-privacy test has been added to, not substituted for, the common-law trespassory test. The holding in Knotts addressed only the former, since the latter was not at issue. The beeper had been placed in the container before it came into Knotts’ possession, with the consent of the then-owner. 460 U. S., at 278. Knotts did not challenge that installation, and we specifically de- clined to consider its effect on the Fourth Amendment analysis. Id., at 279, n. Knotts would be relevant, perhaps, if the Government were making the argument that what would otherwise be an unconstitutional search is not such where it produces only public information. The Government does not make that argument, and we know of no case that would support it. The second “beeper” case, United States v. Karo, 468 U.S. 705 (1984) , does not suggest a different conclusion. There we addressed the question left open by Knotts, whether the installation of a beeper in a container amounted to a search or seizure. 468 U. S., at 713. As in Knotts, at the time the beeper was installed the container belonged to a third party, and it did not come into possession of the defendant until later. 468 U. S., at 708. Thus, the specific question we considered was whether the installation “with the consent of the original owner constitute[d] a search or seizure . . . when the container is delivered to a buyer having no knowledge of the presence of the beeper.” Id., at 707 (emphasis added). We held not. The Government, we said, came into physical contact with the container only before it belonged to the defendant Karo; and the transfer of the container with the unmonitored beeper inside did not convey any information and thus did not invade Karo’s privacy. See id., at 712. That conclusion is perfectly consistent with the one we reach here. Karo accepted the container as it came to him, beeper and all, and was therefore not entitled to object to the beeper’s presence, even though it was used to monitor the container’s location. Cf. On Lee v. United States, 343 U.S. 747 –752 (1952) (no search or seizure where an informant, who was wearing a concealed microphone, was invited into the defendant’s business). Jones, who possessed the Jeep at the time the Government trespassorily inserted the information-gathering device, is on much different footing. The Government also points to our exposition in New York v. Class, 475 U.S. 106 (1986) , that “[t]he exterior of a car . . . is thrust into the public eye, and thus to examine it does not constitute a ‘search.’ ” Id., at 114. That statement is of marginal relevance here since, as the Government acknowledges, “the officers in this case did more than conduct a visual inspection of respondent’s vehicle,” Brief for United States 41 (emphasis added). By attaching the device to the Jeep, officers encroached on a protected area. In Class itself we suggested that this would make a difference, for we concluded that an officer’s momentary reaching into the interior of a vehicle did constitute a search. [ 7 ] 475 U. S., at 114–115. Finally, the Government’s position gains little support from our conclusion in Oliver v. United States, 466 U.S. 170 (1984) , that officers’ information-gathering intrusion on an “open field” did not constitute a Fourth Amendment search even though it was a trespass at common law, id., at 183. Quite simply, an open field, unlike the curtilage of a home, see United States v. Dunn, 480 U.S. 294, 300 (1987) , is not one of those protected areas enumerated in the Fourth Amendment. Oliver, supra, at 176–177. See also Hester v. United States, 265 U.S. 57, 59 (1924) . The Government’s physical intrusion on such an area—unlike its intrusion on the “effect” at issue here—is of no Fourth Amendment significance. [ 8 ] B The concurrence begins by accusing us of applying “18th-century tort law.” Post, at 1. That is a distortion. What we apply is an 18th-century guarantee against un- reasonable searches, which we believe must provide at a minimum the degree of protection it afforded when it was adopted. The concurrence does not share that belief. It would apply exclusively Katz’s reasonable-expectation-of-privacy test, even when that eliminates rights that previously existed. The concurrence faults our approach for “present[ing] particularly vexing problems” in cases that do not involve physical contact, such as those that involve the transmission of electronic signals. Post, at 9. We entirely fail to understand that point. For unlike the concurrence, which would make Katz the exclusive test, we do not make trespass the exclusive test. Situations involving merely the transmission of electronic signals without trespass would remain subject to Katz analysis. In fact, it is the concurrence’s insistence on the exclusivity of the Katz test that needlessly leads us into “particularly vexing problems” in the present case. This Court has to date not deviated from the understanding that mere visual observation does not constitute a search. See Kyllo, 533 U. S., at 31–32. We accordingly held in Knotts that “[a] person traveling in an automobile on public thoroughfares has no reasonable expectation of privacy in his movements from one place to another.” 460 U. S., at 281. Thus, even assuming that the concurrence is correct to say that “[t]raditional surveillance” of Jones for a 4-week period “would have required a large team of agents, multiple vehicles, and perhaps aerial assistance,” post, at 12, our cases suggest that such visual observation is constitutionally permissible. It may be that achieving the same result through electronic means, without an accompany- ing trespass, is an unconstitutional invasion of privacy, but the present case does not require us to answer that question. And answering it affirmatively leads us needlessly into additional thorny problems. The concurrence posits that “relatively short-term monitoring of a person’s movements on public streets” is okay, but that “the use of longer term GPS monitoring in investigations of most offenses” is no good. Post, at 13 (emphasis added). That introduces yet another novelty into our jurisprudence. There is no precedent for the proposition that whether a search has occurred depends on the nature of the crime being investigated. And even accepting that novelty, it remains unexplained why a 4-week investigation is “surely” too long and why a drug-trafficking conspiracy involving sub- stantial amounts of cash and narcotics is not an “extra- ordinary offens[e]” which may permit longer observation. See post, at 13–14. What of a 2-day monitoring of a suspected purveyor of stolen electronics? Or of a 6-month monitoring of a suspected terrorist? We may have to grapple with these “vexing problems” in some future case where a classic trespassory search is not involved and resort must be had to Katz analysis; but there is no reason for rushing forward to resolve them here. III The Government argues in the alternative that even if the attachment and use of the device was a search, it was reasonable—and thus lawful—under the Fourth Amendment because “officers had reasonable suspicion, and in- deed probable cause, to believe that [Jones] was a leader in a large-scale cocaine distribution conspiracy.” Brief for United States 50–51. We have no occasion to consider this argument. The Government did not raise it below, and the D. C. Circuit therefore did not address it. See 625 F. 3d, at 767 (Ginsburg, Tatel, and Griffith, JJ., concurring in denial of rehearing en banc). We consider the argument forfeited. See Sprietsma v. Mercury Marine, 537 U.S. 51 , n. 4 (2002). * * * The judgment of the Court of Appeals for the D. C. Circuit is affirmed. It is so ordered. Notes 1 In this litigation, the Government has conceded noncompliance with the warrant and has argued only that a warrant was not required. United States v. Maynard, 615 F.3d 544, 566, n. (CADC 2010). 2 As we have noted, the Jeep was registered to Jones’s wife. The Government acknowledged, however, that Jones was “the exclusive driver.” Id., at 555, n. (internal quotation marks omitted). If Jones was not the owner he had at least the property rights of a bailee. The Court of Appeals concluded that the vehicle’s registration did not affect his ability to make a Fourth Amendment objection, ibid., and the Government has not challenged that determination here. We therefore do not consider the Fourth Amendment significance of Jones’s status. 3 Justice Alito’s concurrence (hereinafter concurrence) doubts the wisdom of our approach because “it is almost impossible to think of late-18th-century situations that are analogous to what took place in this case.” Post, at 3 (opinion concurring in judgment). But in fact it posits a situation that is not far afield—a constable’s concealing himself in the target’s coach in order to track its movements. Ibid. There is no doubt that the information gained by that trespassory activity would be the product of an unlawful search—whether that information consisted of the conversations occurring in the coach, or of the destinations to which the coach traveled. In any case, it is quite irrelevant whether there was an 18th-century analog. Whatever new methods of investigation may be devised, our task, at a minimum, is to decide whether the action in question would have constituted a “search” within the original meaning of the Fourth Amendment. Where, as here, the Government obtains information by physically intruding on a constitutionally protected area, such a search has undoubtedly occurred. 4 Thus, the concurrence’s attempt to recast Alderman as meaning that individuals have a “legitimate expectation of privacy in all conversations that [take] place under their roof,” post, at 6–7, is foreclosed by the Court’s opinion. The Court took as a given that the homeowner’s “conversational privacy” had not been violated. 5 The concurrence notes that post-Katz we have explained that “ ‘an actual trespass is neither necessary nor sufficient to establish a constitutional violation.’ ” Post, at 6 (quoting United States v. Karo, 468 U.S. 705, 713 (1984) ). That is undoubtedly true, and undoubtedly irrelevant. Karo was considering whether a seizure occurred, and as the concurrence explains, a seizure of property occurs, not when there is a trespass, but “when there is some meaningful interference with an individual’s possessory interests in that property.” Post, at 2 (internal quotation marks omitted). Likewise with a search. Trespass alone does not qualify, but there must be conjoined with that what was present here: an attempt to find something or to obtain information. Related to this, and similarly irrelevant, is the concurrence’s point that, if analyzed separately, neither the installation of the device nor its use would constitute a Fourth Amendment search. See ibid. Of course not. A trespass on “houses” or “effects,” or a Katz invasion of privacy, is not alone a search unless it is done to obtain information; and the obtaining of information is not alone a search unless it is achieved by such a trespass or invasion of privacy. 6 Knotts noted the “limited use which the government made of the signals from this particular beeper,” 460 U. S., at 284; and reserved the question whether “different constitutional principles may be applicable” to “dragnet-type law enforcement practices” of the type that GPS tracking made possible here, ibid. 7 The Government also points to Cardwell v. Lewis, 417 U.S. 583 (1974) , in which the Court rejected the claim that the inspection of an impounded vehicle’s tire tread and the collection of paint scrapings from its exterior violated the Fourth Amendment. Whether the plural-ity said so because no search occurred or because the search was rea-sonable is unclear. Compare id., at 591 (opinion of Blackmun, J.) (“[W]e fail to comprehend what expectation of privacy was infringed”), with id., at 592 (“Under circumstances such as these, where probable cause exists, a warrantless examination of the exterior of a car is not unreasonable . . . ”). 8 Thus, our theory is not that the Fourth Amendment is concerned with “any technical trespass that led to the gathering of evidence.” Post, at 3 (Alito, J., concurring in judgment) (emphasis added). The Fourth Amendment protects against trespassory searches only with regard to those items (“persons, houses, papers, and effects”) that it enumerates. The trespass that occurred in Oliver may properly be understood as a “search,” but not one “in the constitutional sense.” 466 U. S., at 170, 183.