UNITED STATES v. HAYMOND, 588 U.S. ___ (2019)


Issues: , ,

139 S.Ct. 2369 (2019)
588 U.S. ___

UNITED STATES, Petitioner
v.
Andre Ralph HAYMOND.

No. 17-1672.
Supreme Court of United States.

Argued February 26, 2019.
Decided June 26, 2019.

ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE TENTH CIRCUIT.

Counsel

Eric J. Feigin for the petitioner.

William D. Lunn for the respondent.

Jeffrey T. Green, Matthew J. Letten, Sidley Austin LLP, Washington, DC, Sarah O’Rourke Schrup, Northwestern Supreme, Court Practicum, Chicago, IL, William D. Lunn, Counsel of Record, Tulsa, OK, for respondent.

Noel J. Francisco, Solicitor General, Brian A. Benczkowski, Assistant Attorney General, Eric J. Feigin, Christopher G. Michel, Assistants to the Solicitor General, William A. Glaser, Attorney, Department of Justice, Washington, DC, for petitioner.

*2371

Syllabus[*]

Respondent Andre Haymond was convicted of possessing child pornography, a crime that carries a prison term of zero to 10 years. After serving a prison sentence of 38 months, and while on supervised release, Mr. Haymond was again found with what appeared to be child pornography. The government sought to revoke his supervised release and secure a new and additional prison sentence. A district judge, acting without a jury, found by a preponderance of the evidence that Mr. Haymond knowingly downloaded and possessed child pornography. Under 18 U.S.C. § 3583(e)(3), the judge could have sentenced him to a prison term of between zero and two additional years. But because possession of child pornography is an enumerated offense under § 3583(k), the judge instead imposed that provision’s 5-year mandatory minimum. On appeal, the Tenth Circuit observed that whereas a jury had convicted Mr. Haymond beyond a reasonable doubt of a crime carrying a prison term of zero to 10 years, this new prison term included a new and higher mandatory minimum resting on facts found only by a judge by a preponderance of the evidence. The Tenth Circuit therefore held that § 3583(k) violated the right to trial by jury guaranteed by the Fifth and Sixth Amendments.

Held: The judgment is vacated, and the case is remanded.

869 F.3d 1153, vacated and remanded.

Justice GORSUCH, joined by Justice GINSBURG, Justice SOTOMAYOR, and Justice KAGAN, concluded that the application of § 3583(k) in this case violated Mr. Haymond’s right to trial by jury. Pp. 2375-2385.

(a) As at the time of the Fifth and Sixth Amendments’ adoption, a judge’s sentencing authority derives from, and is limited by, the jury’s factual findings of criminal conduct. A jury must find beyond a reasonable doubt every fact “`which the law makes essential to [a] punishment'”

*2372

that a judge might later seek to impose. Blakely v. Washington, 542 U.S. 296, 304, 124 S.Ct. 2531, 159 L.Ed.2d 403. Historically, that rule’s application proved straightforward, but recent legislative innovations have raised difficult questions. In Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435, for example, this Court held unconstitutional a sentencing scheme that allowed a judge to increase a defendant’s sentence beyond the statutory maximum based on the judge’s finding of new facts by a preponderance of the evidence. And in Alleyne v. United States, 570 U.S. 99, 133 S.Ct. 2151, 186 L.Ed.2d 314, the Court held that Apprendi‘s principle “applies with equal force to facts increasing the mandatory minimum.” 570 U.S. at 111-112, 133 S.Ct. 2151. The lesson for this case is clear: Based solely on the facts reflected in the jury’s verdict, Mr. Haymond faced a lawful prison term of between zero and 10 years. But just like the facts the judge found at the defendant’s sentencing hearing in Alleyne, the facts the judge found here increased “the legally prescribed range of allowable sentences” in violation of the Fifth and Sixth Amendments. Id., at 115, 133 S.Ct. 2151. Pp. 2375-2379.

(b) The government’s various replies are unpersuasive. First, it stresses that Alleyne arose in a different procedural posture, but this Court has repeatedly rejected efforts to dodge the demands of the Fifth and Sixth Amendments by the simple expedient of relabeling a criminal prosecution. And this Court has already recognized that punishments for revocation of supervised release arise from and are “treat[ed] … as part of the penalty for the initial offense.” Johnson v. United States, 529 U.S. 694, 700, 120 S.Ct. 1795, 146 L.Ed.2d 727. Because a defendant’s final sentence includes any revocation sentence he may receive, § 3583(k)’s 5-year mandatory minimum mirrors the unconstitutional sentencing enhancement in Alleyne. Second, the government suggests that Mr. Haymond’s sentence for violating the terms of his supervised release was actually fully authorized by the jury’s verdict, because his supervised release was from the outset always subject to the possibility of judicial revocation and § 3583(k)’s mandatory prison sentence. But what is true in Apprendi and Alleyne can be no less true here: A mandatory minimum 5-year sentence that comes into play only as a result of additional judicial factual findings by a preponderance of the evidence cannot stand. Finally, the government contends that § 3583(k)’s supervised release revocation procedures are practically identical to historic parole and probation revocation procedures, which have usually been understood to comport with the Fifth and Sixth Amendments. That argument overlooks a critical difference between § 3583(k) and traditional parole and probation practices. Where parole and probation violations traditionally exposed a defendant only to the remaining prison term authorized for his crime of conviction, § 3583(k) exposes a defendant to an additional mandatory minimum prison term beyond that authorized by the jury’s verdict —all based on facts found by a judge by a mere preponderance of the evidence. Pp. 2378-2383.

(c) The Tenth Circuit may address on remand the question whether its remedy— declaring the last two sentences of § 3583(k) “unconstitutional and unenforceable” —sweeps too broadly, including any question concerning whether the government’s argument to that effect was adequately preserved. Pp. 2384-2385.

Justice BREYER agreed that the particular provision at issue, 18 U.S.C. § 3583(k), is unconstitutional. Three features of § 3583(k), considered together, make it less like ordinary supervised-release

*2373

revocation and more like punishment for a new offense, to which the jury right would typically attach. First, § 3583(k) applies only when a defendant commits a discrete set of criminal offenses specified in the statute. Second, § 3583(k) takes away the judge’s discretion to decide whether violation of the conditions of supervised release should result in imprisonment and for how long. Third, § 3583(k) limits the judge’s discretion in a particular manner: by imposing a mandatory minimum term of imprisonment of “not less than 5 years” upon a judge’s finding that a defendant has committed a listed offense. But because the role of the judge in a typical supervised-release proceeding is consistent with traditional parole and because Congress clearly did not intend the supervised release system to differ from parole in this respect, Justice BREYER would not transplant the Apprendi line of cases to the supervised-release context. Pp. 2373-2375.

GORSUCH, J., announced the judgment of the Court and delivered an opinion, in which GINSBURG, SOTOMAYOR, and KAGAN, JJ., joined. BREYER, J., filed an opinion concurring in the judgment. ALITO, J., filed a dissenting opinion, in which ROBERTS, C. J., and THOMAS and KAVANAUGH, JJ., joined.

Justice GORSUCH, announced the judgment of the Court and delivered an opinion, in which Justice GINSBURG, Justice SOTOMAYOR, and Justice KAGAN joined.

Only a jury, acting on proof beyond a reasonable doubt, may take a person’s liberty. That promise stands as one of the Constitution’s most vital protections against arbitrary government. Yet in this case a congressional statute compelled a federal judge to send a man to prison for a minimum of five years without empaneling a jury of his peers or requiring the government to prove his guilt beyond a reasonable doubt. As applied here, we do not hesitate to hold that the statute violates the Fifth and Sixth Amendments.

I

After a jury found Andre Haymond guilty of possessing child pornography in violation of federal law, the question turned to sentencing. The law authorized the district judge to impose a prison term of between zero and 10 years, 18 U.S.C. § 2252(b)(2), and a period of supervised release of between 5 years and life, § 3583(k). Because Mr. Haymond had no criminal history and was working to help support his mother who had suffered a stroke, the judge concluded that Mr. Haymond was “not going to get much out of being in prison” and sentenced him to a prison term of 38 months, followed by 10 years of supervised release.

*2374

After completing his prison sentence, however, Mr. Haymond encountered trouble on supervised release. He sat for multiple polygraph tests in which he denied possessing or viewing child pornography, and each time the test indicated no deception. But when the government conducted an unannounced search of his computers and cellphone, it turned up 59 images that appeared to be child pornography. Based on that discovery, the government sought to revoke Mr. Haymond’s supervised release and secure a new and additional prison sentence.

A hearing followed before a district judge acting without a jury, and under a preponderance of the evidence rather than a reasonable doubt standard. In light of expert testimony regarding the manner in which cellphones can “cache” images without the user’s knowledge, the judge found insufficient evidence to show that Mr. Haymond knowingly possessed 46 of the images. At the same time, the judge found it more likely than not that Mr. Haymond knowingly downloaded and possessed the remaining 13 images.

With that, the question turned once more to sentencing. Under 18 U.S.C. § 3583(e)(3), enacted as part of the Sentencing Reform Act of 1984, a district judge who finds that a defendant has violated the conditions of his supervised release normally may (but is not required to) impose a new prison term up to the maximum period of supervised release authorized by statute for the defendant’s original crime of conviction, subject to certain limits.1  Under that provision, the judge in this case would have been free to sentence Mr. Haymond to between zero and two additional years in prison.

But there was a complication. Under § 3583(k), added to the Act in 2003 and amended in 2006, if a judge finds by a preponderance of the evidence that a defendant on supervised release committed one of several enumerated offenses, including the possession of child pornography, the judge must impose an additional prison term of at least five years and up to life without regard to the length of the prison term authorized for the defendant’s initial crime of conviction.2

*2375

Because Mr. Haymond had committed an offense covered by § 3583(k), the judge felt bound to impose an additional prison term of at least five years. He did so, though, with reservations. It’s one thing, Judge Terence Kern said, for a judge proceeding under a preponderance of the evidence standard to revoke a defendant’s supervised release and order him to serve additional time in prison within the range already authorized by the defendant’s original conviction; after all, the jury’s verdict, reached under the reasonable doubt standard, permitted that much punishment. But the judge found it “`repugnant'” that a statute might impose a new and additional “mandatory five-year” punishment without those traditional protections. Were it not for § 3583(k)’s mandatory minimum, the judge added, he “probably would have sentenced in the range of two years or less.”

On appeal to the Tenth Circuit, Mr. Haymond challenged both the factual support for his new punishment and its constitutionality. On the facts, the court of appeals held that the district court’s findings against Mr. Haymond were clearly erroneous in certain respects. Even so, the court concluded, just enough evidence remained to sustain a finding that Mr. Haymond had knowingly possessed the 13 images at issue, in violation of § 3583(k). That left the question of the statute’s constitutionality, and there the Tenth Circuit concluded that § 3583(k) violated the Fifth and Sixth Amendments. The court explained that a jury had convicted Mr. Haymond beyond a reasonable doubt of a crime carrying a prison term of zero to 10 years. Yet now Mr. Haymond faced a new potential prison term of five years to life. Because this new prison term included a new and higher mandatory minimum resting only on facts found by a judge by a preponderance of the evidence, the court held, the statute violated Mr. Haymond’s right to trial by jury.

By way of remedy, the court held the last two sentences of § 3583(k), which mandate a 5-year minimum prison term, “unconstitutional and unenforceable.” 869 F.3d 1153, 1168 (C.A.10 2017). The court then vacated Mr. Haymond’s revocation sentence and remanded the case to the district court for resentencing without regard to those provisions. In effect, the court of appeals left the district court free to issue a new sentence under the preexisting statute governing most every other supervised release violation, § 3583(e). Following the Tenth Circuit’s directions, the district court proceeded to resentence Mr. Haymond to time served, as he had already been detained by that point for approximately 28 months. We granted review to consider the Tenth Circuit’s constitutional holding. 586 U. S. ___, 139 S.Ct. 398, 202 L.Ed.2d 309 (2018).

II

Together with the right to vote, those who wrote our Constitution considered the right to trial by jury “the heart and lungs, the mainspring and the center wheel” of our liberties, without which “the body must die; the watch must run down; the government must become arbitrary.” Letter from Clarendon to W. Pym (Jan. 27, 1766), in 1 Papers of John Adams 169 (R. Taylor ed. 1977). Just as the right to vote sought to preserve the people’s authority over their government’s executive and legislative functions, the right to a jury trial sought to preserve the people’s authority over its judicial functions. J. Adams, Diary Entry (Feb. 12, 1771), in 2 Diary and Autobiography of John Adams 3 (L. Butterfield ed. 1961); see also 2 J. Story, Commentaries on the Constitution § 1779, pp. 540-541 (4th ed. 1873).

*2376

Toward that end, the Framers adopted the Sixth Amendment’s promise that “[i]n all criminal prosecutions the accused shall enjoy the right to a speedy and public trial, by an impartial jury.” In the Fifth Amendment, they added that no one may be deprived of liberty without “due process of law.” Together, these pillars of the Bill of Rights ensure that the government must prove to a jury every criminal charge beyond a reasonable doubt, an ancient rule that has “extend[ed] down centuries.” Apprendi v. New Jersey, 530 U.S. 466, 477, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000).

But when does a “criminal prosecution” arise implicating the right to trial by jury beyond a reasonable doubt? At the founding, a “prosecution” of an individual simply referred to “the manner of [his] formal accusation.” 4 W. Blackstone, Commentaries on the Laws of England 298 (1769) (Blackstone); see also N. Webster, An American Dictionary of the English Language (1st ed. 1828) (defining “prosecution” as “the process of exhibiting formal charges against an offender before a legal tribunal”). And the concept of a “crime” was a broad one linked to punishment, amounting to those “acts to which the law affixes … punishment,” or, stated differently, those “element[s] in the wrong upon which the punishment is based.” 1 J. Bishop, Criminal Procedure §§ 80, 84, pp. 51-53 (2d ed. 1872) (Bishop); see also J. Archbold, Pleading and Evidence in Criminal Cases *106 (5th Am. ed. 1846) (Archbold) (discussing a crime as including any fact that “annexes a higher degree of punishment”); Blakely v. Washington, 542 U.S. 296, 309, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004); Apprendi, 530 U.S. at 481, 120 S.Ct. 2348.

Consistent with these understandings, juries in our constitutional order exercise supervisory authority over the judicial function by limiting the judge’s power to punish. A judge’s authority to issue a sentence derives from, and is limited by, the jury’s factual findings of criminal conduct. In the early Republic, if an indictment or “accusation … lack[ed] any particular fact which the laws ma[d]e essential to the punishment,” it was treated as “no accusation” at all. 1 Bishop § 87, at 55; see also 2 M. Hale, Pleas of the Crown *170 (1736); Archbold *106. And the “truth of every accusation” that was brought against a person had to “be confirmed by the unanimous suffrage of twelve of his equals and neighbours.” 4 Blackstone 343. Because the Constitution’s guarantees cannot mean less today than they did the day they were adopted, it remains the case today that a jury must find beyond a reasonable doubt every fact “`which the law makes essential to [a] punishment'” that a judge might later seek to impose. Blakely, 542 U.S. at 304, 124 S.Ct. 2531 (quoting 1 Bishop § 87, at 55).

For much of our history, the application of this rule of jury supervision proved pretty straightforward. At common law, crimes tended to carry with them specific sanctions, and “once the facts of the offense were determined by the jury, the judge was meant simply to impose the prescribed sentence.” Alleyne v. United States, 570 U.S. 99, 108, 133 S.Ct. 2151, 186 L.Ed.2d 314 (2013) (plurality opinion) (internal quotation marks and brackets omitted). Even when judges did enjoy discretion to adjust a sentence based on judge-found aggravating or mitigating facts, they could not “`swell the penalty above what the law ha[d] provided for the acts charged'” and found by the jury. Apprendi, 530 U.S. at 519, 120 S.Ct. 2348 (THOMAS, J., concurring) (quoting 1 Bishop § 85, at 54); see also 1 J. Bishop, Criminal Law §§ 933-934(1), p. 690 (9th ed. 1923) (“[T]he court determines in each

*2377

case what within the limits of the law shall be the punishment” (emphasis added)). In time, of course, legislatures adopted new laws allowing judges or parole boards to suspend part (parole) or all (probation) of a defendant’s prescribed prison term and afford him a period of conditional liberty as an “act of grace,” subject to revocation. Escoe v. Zerbst, 295 U.S. 490, 492, 55 S.Ct. 818, 79 L.Ed. 1566 (1935); see Anderson v. Corall, 263 U.S. 193, 196-197, 44 S.Ct. 43, 68 L.Ed. 247 (1923). But here, too, the prison sentence a judge or parole board could impose for a parole or probation violation normally could not exceed the remaining balance of the term of imprisonment already authorized by the jury’s verdict. So even these developments did not usually implicate the historic concerns of the Fifth and Sixth Amendments. See Blakely, 542 U.S. at 309, 124 S.Ct. 2531; Apprendi, 530 U.S. at 498, 120 S.Ct. 2348 (Scalia, J., concurring); 4 Atty. Gen.’s Survey of Release Proc. 22 (1939); 2 id., at 333.

More recent legislative innovations have raised harder questions. In Apprendi, for example, a jury convicted the defendant of a gun crime that carried a maximum prison sentence of 10 years. But then a judge sought to impose a longer sentence pursuant to a statute that authorized him to do so if he found, by a preponderance of the evidence, that the defendant had committed the crime with racial bias. Apprendi held this scheme unconstitutional. “[A]ny fact that increases the penalty for a crime beyond the prescribed statutory maximum,” this Court explained, “must be submitted to a jury, and proved beyond a reasonable doubt” or admitted by the defendant. 530 U.S. at 490, 120 S.Ct. 2348. Nor may a State evade this traditional restraint on the judicial power by simply calling the process of finding new facts and imposing a new punishment a judicial “sentencing enhancement.” Id., at 495, 120 S.Ct. 2348. “[T]he relevant inquiry is one not of form, but of effect—does the required [judicial] finding expose the defendant to a greater punishment than that authorized by the jury’s guilty verdict?” Id., at 494, 120 S.Ct. 2348.

While “trial practices ca[n] change in the course of centuries and still remain true to the principles that emerged from the Framers'” design, id., at 483, 120 S.Ct. 2348, in the years since Apprendi this Court has not hesitated to strike down other innovations that fail to respect the jury’s supervisory function. See, e.g., Ring v. Arizona, 536 U.S. 584, 122 S.Ct. 2428, 153 L.Ed.2d 556 (2002) (imposition of death penalty based on judicial factfinding); Blakely, 542 U.S. at 303, 124 S.Ct. 2531 (mandatory state sentencing guidelines); Cunningham v. California, 549 U.S. 270, 127 S.Ct. 856, 166 L.Ed.2d 856 (2007) (same); United States v. Booker, 543 U.S. 220, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005) (mandatory federal sentencing guidelines); Southern Union Co. v. United States, 567 U.S. 343, 132 S.Ct. 2344, 183 L.Ed.2d 318 (2012) (imposition of criminal fines based on judicial factfinding).3

Still, these decisions left an important gap. In Apprendi, this Court recognized that “`[i]t is unconstitutional for a legislature to remove from the jury the assessment of facts that increase the prescribed range of penalties.'” 530 U.S. at 490, 120

*2378

S.Ct. 2348. But by definition, a range of punishments includes not only a maximum but a minimum. And logically it would seem to follow that any facts necessary to increase a person’s minimum punishment (the “floor”) should be found by the jury no less than facts necessary to increase his maximum punishment (the “ceiling”). Before Apprendi, however, this Court had held that facts elevating the minimum punishment need not be proven to a jury beyond a reasonable doubt. McMillan v. Pennsylvania, 477 U.S. 79, 106 S.Ct. 2411, 91 L.Ed.2d 67 (1986); see also Harris v. United States, 536 U.S. 545, 122 S.Ct. 2406, 153 L.Ed.2d 524 (2002) (adhering to McMillan).

Eventually, the Court confronted this anomaly in Alleyne. There, a jury convicted the defendant of a crime that ordinarily carried a sentence of five years to life in prison. But a separate statutory “sentencing enhancement” increased the mandatory minimum to seven years if the defendant “brandished” the gun. At sentencing, a judge found by a preponderance of the evidence that the defendant had indeed brandished a gun and imposed the mandatory minimum 7-year prison term.

This Court reversed. Finding no basis in the original understanding of the Fifth and Sixth Amendments for McMillan and Harris, the Court expressly overruled those decisions and held that “the principle applied in Apprendi applies with equal force to facts increasing the mandatory minimum” as it does to facts increasing the statutory maximum penalty. Alleyne, 570 U.S. at 112, 133 S.Ct. 2151. Nor did it matter to Alleyne‘s analysis that, even without the mandatory minimum, the trial judge would have been free to impose a 7-year sentence because it fell within the statutory sentencing range authorized by the jury’s findings. Both the “floor” and “ceiling” of a sentencing range “define the legally prescribed penalty.” Ibid. And under our Constitution, when “a finding of fact alters the legally prescribed punishment so as to aggravate it” that finding must be made by a jury of the defendant’s peers beyond a reasonable doubt. Id., at 114, 133 S.Ct. 2151. Along the way, the Court observed that there can be little doubt that “[e]levating the low end of a sentencing range heightens the loss of liberty associated with the crime: The defendant’s expected punishment has increased as a result of the narrowed range and the prosecution is empowered, by invoking the mandatory minimum, to require the judge to impose a higher punishment than he might wish.” Id., at 113, 133 S.Ct. 2151 (internal quotation marks omitted).

By now, the lesson for our case is clear. Based on the facts reflected in the jury’s verdict, Mr. Haymond faced a lawful prison term of between zero and 10 years under § 2252(b)(2). But then a judge— acting without a jury and based only on a preponderance of the evidence—found that Mr. Haymond had engaged in additional conduct in violation of the terms of his supervised release. Under § 3583(k), that judicial factfinding triggered a new punishment in the form of a prison term of at least five years and up to life. So just like the facts the judge found at the defendant’s sentencing hearing in Alleyne, the facts the judge found here increased “the legally prescribed range of allowable sentences” in violation of the Fifth and Sixth Amendments. Id., at 115, 133 S.Ct. 2151. In this case, that meant Mr. Haymond faced a minimum of five years in prison instead of as little as none. Nor did the absence of a jury’s finding beyond a reasonable doubt only infringe the rights of the accused; it also divested the “`people at large'”—the men and women who make up a jury of a defendant’s peers—of their constitutional authority to set the metes

*2379

and bounds of judicially administered criminal punishments. Blakely, 542 U.S. at 306, 124 S.Ct. 2531 (quoting Letter XV by the Federal Farmer (Jan. 18, 1788), in 2 The Complete Anti-Federalist 315, 320 (H. Storing ed. 1981)).4

III

In reply, the government and the dissent offer many and sometimes competing arguments, but we find none persuasive.

A

The government begins by pointing out that Alleyne arose in a different procedural posture. There, the trial judge applied a “sentencing enhancement” based on his own factual findings at the defendant’s initial sentencing hearing; meanwhile, Mr. Haymond received his new punishment from a judge at a hearing to consider the revocation of his term of supervised release. This procedural distinction makes all the difference, we are told, because the Sixth Amendment’s jury trial promise applies only to “criminal prosecutions,” which end with the issuance of a sentence and do not extend to “postjudgment sentence-administration proceedings.” Brief for United States 24; see also post, at 2393-2395 (ALITO, J., dissenting) (echoing this argument).

But we have been down this road before. Our precedents, Apprendi, Blakely, and Alleyne included, have repeatedly rejected efforts to dodge the demands of the Fifth and Sixth Amendments by the simple expedient of relabeling a criminal prosecution a “sentencing enhancement.” Calling part of a criminal prosecution a “sentence modification” imposed at a “postjudgment sentence-administration proceeding” can fare no better. As this Court has repeatedly explained, any “increase in a defendant’s authorized punishment contingent on the finding of a fact” requires a jury and proof beyond a reasonable doubt “no matter” what the government chooses to call the exercise. Ring, 536 U.S. at 602, 122 S.Ct. 2428.

To be sure, and as the government and dissent emphasize, founding-era prosecutions traditionally ended at final judgment. But at that time, generally, “questions of guilt and punishment both were resolved in a single proceeding” subject to the Fifth and Sixth Amendment’s demands. Douglass, Confronting Death: Sixth Amendment Rights at Capital Sentencing, 105 Colum. L. Rev. 1967, 2011 (2005); see also supra, at 7. Over time, procedures changed as legislatures sometimes bifurcated criminal prosecutions into separate trial and penalty phases. But none of these developments licensed judges to sentence individuals to punishments beyond the legal limits fixed by the facts found in the jury’s verdict. See ibid. To the contrary, we recognized in Apprendi and Alleyne, a “criminal prosecution” continues and the defendant remains an “accused” with all the rights provided by the Sixth Amendment, until a final sentence is imposed. See Apprendi, 530 U.S. at 481-482, 120 S.Ct. 2348.

Today, we merely acknowledge that an accused’s final sentence includes any supervised release sentence he may receive. Nor in saying that do we say anything new: This Court has already recognized that supervised release punishments

*2380

arise from and are “treat[ed] … as part of the penalty for the initial offense.” Johnson v. United States, 529 U.S. 694, 700, 120 S.Ct. 1795, 146 L.Ed.2d 727 (2000). The defendant receives a term of supervised release thanks to his initial offense, and whether that release is later revoked or sustained, it constitutes a part of the final sentence for his crime. As at the initial sentencing hearing, that does not mean a jury must find every fact in a revocation hearing that may affect the judge’s exercise of discretion within the range of punishments authorized by the jury’s verdict. But it does mean that a jury must find any facts that trigger a new mandatory minimum prison term.5

This logic respects not only our precedents, but the original meaning of the jury trial right they seek to protect. The Constitution seeks to safeguard the people’s control over the business of judicial punishments by ensuring that any accusation triggering a new and additional punishment is proven to the satisfaction of a jury beyond a reasonable doubt. By contrast, the view the government and dissent espouse would demote the jury from its historic role as “circuitbreaker in the State’s machinery of justice,” Blakely, 542 U.S. at 306, 124 S.Ct. 2531, to “`low-level gatekeeping,'” Booker, 543 U.S. at 230, 125 S.Ct. 738. If the government and dissent were correct, Congress could require anyone convicted of even a modest crime to serve a sentence of supervised release for the rest of his life. At that point, a judge could try and convict him of any violation of the terms of his release under a preponderance of the evidence standard, and then sentence him to pretty much anything. At oral argument, the government even conceded that, under its theory, a defendant on supervised release would have no Sixth Amendment right to a jury trial when charged with an infraction carrying the death penalty. We continue to doubt whether even Apprendi‘s fiercest critics “would advocate” such an “absurd result.” Blakely, 542 U.S. at 306, 124 S.Ct. 2531.6

B

Where it previously suggested that Mr. Haymond’s supervised release revocation proceeding was entirely divorced from his criminal prosecution, the government next turns around and suggests that Mr. Haymond’s sentence for violating the terms of his supervised release was actually fully authorized by the jury’s verdict. See also post, at 2389-2390 (ALITO, J., dissenting) (proposing a similar theory). After all, the government observes, on the strength of the jury’s findings the judge was entitled to impose as punishment a term of supervised release; and, in turn, that term of supervised release was from the outset always subject to the possibility of judicial

*2381

revocation and § 3583(k)’s mandatory prison sentence. Presto: Sixth Amendment problem solved.

But we have been down this road too. In Apprendi and Alleyne, the jury’s verdict triggered a statute that authorized a judge at sentencing to increase the defendant’s term of imprisonment based on judge-found facts. This Court had no difficulty rejecting that scheme as an impermissible evasion of the historic rule that a jury must find all of the facts necessary to authorize a judicial punishment. See Alleyne, 570 U.S. at 117, 133 S.Ct. 2151; Apprendi, 530 U.S. at 483, 120 S.Ct. 2348. And what was true there can be no less true here: A mandatory minimum 5-year sentence that comes into play only as a result of additional judicial factual findings by a preponderance of the evidence cannot stand. This Court’s observation that “postrevocation sanctions” are “treat[ed] … as part of the penalty for the initial offense,” Johnson, 529 U.S. at 700, 120 S.Ct. 1795, only highlights the constitutional infirmity of § 3583(k): Treating Mr. Haymond’s 5-year mandatory minimum prison term as part of his sentence for his original offense makes clear that it mirrors the unconstitutional sentencing enhancement in Alleyne. See supra, at 2379-2380.

Notice, too, that following the government down this road would lead to the same destination as the last: If the government were right, a jury’s conviction on one crime would (again) permit perpetual supervised release and allow the government to evade the need for another jury trial on any other offense the defendant might commit, no matter how grave the punishment. And if there’s any doubt about the incentives such a rule would create, consider this case. Instead of seeking a revocation of supervised release, the government could have chosen to prosecute Mr. Haymond under a statute mandating a term of imprisonment of 10 to 20 years for repeat child-pornography offenders. 18 U.S.C. § 2252(b)(2). But why bother with an old-fashioned jury trial for a new crime when a quick-and-easy “supervised release revocation hearing” before a judge carries a penalty of five years to life? This displacement of the jury’s traditional supervisory role, under cover of a welter of new labels, exemplifies the “Framers’ fears that the jury right could be lost not only by gross denial, but by erosion.” Apprendi, 530 U.S. at 483, 120 S.Ct. 2348 (internal quotation marks omitted).

C

Pivoting once more, the government and the dissent seem to accept for argument’s sake that “postjudgment sentence-administration proceedings” can implicate the Fifth and Sixth Amendments. See post, at 2376-2379. But, they contend, § 3583(k)’s supervised release revocation procedures are practically identical to historic parole and probation revocation procedures. See, e.g., Gagnon v. Scarpelli, 411 U.S. 778, 93 S.Ct. 1756, 36 L.Ed.2d 656 (1973); Morrissey v. Brewer, 408 U.S. 471, 92 S.Ct. 2593, 33 L.Ed.2d 484 (1972). And, because those other procedures have usually been understood to comport with the Fifth and Sixth Amendments, they submit, § 3583(k)’s procedures must do so as well.

But this argument, too, rests on a faulty premise, overlooking a critical difference between § 3583(k) and traditional parole and probation practices. Before the Sentencing Reform Act of 1984, a federal criminal defendant could serve as little as a third of his assigned prison term before becoming eligible for release on parole. See 18 U.S.C. § 4205(a) (1982 ed.). Or he might avoid prison altogether in favor of probation. See § 3561 (1982 ed.). If the defendant violated the terms of his parole or probation, a judge could send him to

*2382

prison. But either way and as we’ve seen, a judge generally could sentence the defendant to serve only the remaining prison term authorized by statute for his original crime of conviction. See supra, at 2376-2377; Morrissey, 408 U.S. at 477, 92 S.Ct. 2593 (“The essence of parole is release from prison, before the completion of sentence” (emphasis added)). Thus, a judge could not imprison a defendant for any longer than the jury’s factual findings allowed —a result entirely harmonious with the Fifth and Sixth Amendments. See Apprendi, 530 U.S. at 498, 120 S.Ct. 2348 (Scalia, J., concurring); Blakely, 542 U.S. at 309, 124 S.Ct. 2531.

All that changed beginning in 1984. That year, Congress overhauled federal sentencing procedures to make prison terms more determinate and abolish the practice of parole. Now, when a defendant is sentenced to prison he generally must serve the great bulk of his assigned term. In parole’s place, Congress established the system of supervised release. But “[u]nlike parole,” supervised release wasn’t introduced to replace a portion of the defendant’s prison term, only to encourage rehabilitation after the completion of his prison term. United States Sentencing Commission, Guidelines Manual ch. 7, pt. A(2)(b) (Nov. 2012); see Doherty, Indeterminate Sentencing Returns: The Invention of Supervised Release, 88 N. Y. U. L. Rev. 958, 1024 (2013).

In this case, that structural difference bears constitutional consequences. Where parole and probation violations generally exposed a defendant only to the remaining prison term authorized for his crime of conviction, as found by a unanimous jury under the reasonable doubt standard, supervised release violations subject to § 3583(k) can, at least as applied in cases like ours, expose a defendant to an additional mandatory minimum prison term well beyond that authorized by the jury’s verdict—all based on facts found by a judge by a mere preponderance of the evidence. In fact, § 3583(k) differs in this critical respect not only from parole and probation; it also represents a break from the supervised release practices that Congress authorized in § 3583(e)(3) and that govern most federal criminal proceedings today. Unlike all those procedures, § 3583(k) alone requires a substantial increase in the minimum sentence to which a defendant may be exposed based only on judge-found facts under a preponderance standard. And, as we explained in Alleyne and reaffirm today, that offends the Fifth and Sixth Amendments’ ancient protections.7

D

The dissent suggests an analogy between revocation under § 3583(k) and prison disciplinary procedures that do not normally require the involvement of a jury. Post, at 2396-2397. But the analogy is a strained one: While the Sixth Amendment surely does not require a jury to find every fact that the government relies on to adjust the terms of a prisoner’s confinement (say, by reducing some of his privileges as a sanction for violating the prison rules), that does not mean the government can send a free man back to prison for years based on judge-found facts.

*2383

Again, practice in the early Republic confirms this. At that time, a term of imprisonment may have been understood as encompassing a degree of summary discipline for alleged infractions of prison regulations without the involvement of a jury. See F. Gray, Prison Discipline in America 22-23, 48-49 (1848). But that does not mean any sanction, no matter how serious, would have been considered part and parcel of the original punishment. On the contrary, the few courts that grappled with this issue seem to have recognized that “infamous” punishments, such as a substantial additional term in prison, might implicate the right to trial by jury. See, e.g., Gross v. Rice, 71 Me. 241, 246-252 (1880); In re Edwards, 43 N.J.L. 555, 557-558 (1881).

What’s more, a tradition of summary process in prison, where administrators face the “formidable task” of controlling a large group of potentially unruly prisoners, does not necessarily support the use of such summary process outside the prison walls. O’Lone v. Estate of Shabazz, 482 U.S. 342, 353, 107 S.Ct. 2400, 96 L.Ed.2d 282 (1987); cf. Morrissey, 408 U.S. at 482, 92 S.Ct. 2593. We have long held that prison regulations that impinge on the constitutional rights inmates would enjoy out-side of prison must be “reasonably related to legitimate penological interests” in managing the prison. Turner v. Safley, 482 U.S. 78, 89, 107 S.Ct. 2254, 96 L.Ed.2d 64 (1987). That approach, we have said, ensures that corrections officials can “`anticipate security problems'” and address “`the intractable problems of prison administration.'” O’Lone, 482 U.S. at 349, 107 S.Ct. 2400; see also Dahne v. Richey, 587 U. S. ___, ___, 139 S.Ct. 1531, 1532 ___ L.Ed.2d ___ (2019) (ALITO, J., dissenting from denial of certiorari) (“To maintain order, prison authorities may insist on compliance with rules that would not be permitted in the outside world”). Whether or not the Turner test applies to prisoners’ jury trial rights, we certainly have never extended it to the jury rights of persons out in the world who retain the core attributes of liberty. Cf. Griffin v. Wisconsin, 483 U.S. 868, 874, n. 2, 107 S.Ct. 3164, 97 L.Ed.2d 709 (1987) (reserving question whether Turner applies to probation). Even the government has not asked us to do so today.8

E

Finally, much of the dissent is consumed by what it calls the “potentially revolutionary” consequences of our opinion. Post, at 2386-2387; see also post, at 2394, 2399 (calling our opinion “inexcusable,” “unpardonabl[e],” and “dangerous”); post, at 2388 (our opinion threatens to bring “the whole concept of supervised release … crashing down”); post, at 2391 (under our opinion, “the whole system of supervised release would be like a 40-ton truck speeding down a steep mountain road with no brakes”). But what agitates the dissent so much is an issue not presented here: whether all supervised release proceedings comport with Apprendi. As we have emphasized, our decision is limited to § 3583(k)—an unusual provision enacted little more than a decade ago—and the Alleyne problem raised by its 5-year mandatory minimum term of imprisonment. See n. 7, supra. Section § 3583(e), which governs supervised release revocation proceedings generally, does not contain any

*2384

similar mandatory minimum triggered by judge-found facts.

Besides, even if our opinion could be read to cast doubts on § 3583(e) and its consistency with Apprendi, the practical consequences of a holding to that effect would not come close to fulfilling the dissent’s apocalyptic prophecy. In most cases (including this one), combining a defendant’s initial and post-revocation sentences issued under § 3583(e) will not yield a term of imprisonment that exceeds the statutory maximum term of imprisonment the jury has authorized for the original crime of conviction. That’s because “courts rarely sentence defendants to the statutory maxima,” United States v. Caso, 723 F.3d 215, 224-225 (C.A.D.C. 2013) (citing Sentencing Commission data indicating that only about 1% of defendants receive the maximum), and revocation penalties under § 3583(e)(3) are only a small fraction of those available under § 3583(k). So even if § 3583(e)(3) turns out to raise Sixth Amendment issues in a small set of cases, it hardly follows that “as a practical matter supervised-release revocation proceedings cannot be held” or that “the whole idea of supervised release must fall.” Post, at 2388. Indeed, the vast majority of supervised release revocation proceedings under subsection (e)(3) would likely be unaffected.

In the end, the dissent is left only to echo an age-old criticism: Jury trials are inconvenient for the government. Yet like much else in our Constitution, the jury system isn’t designed to promote efficiency but to protect liberty. In what now seems a prescient passage, Blackstone warned that the true threat to trial by jury would come less from “open attacks,” which “none will be so hardy as to make,” as from subtle “machinations, which may sap and undermine i[t] by introducing new and arbitrary methods.” 4 Blackstone 343. This Court has repeatedly sought to guard the historic role of the jury against such incursions. For “however convenient these may appear at first, (as doubtless all arbitrary powers, well executed, are the most convenient) yet let it be again remembered, that delays, and little inconveniences in the forms of justice, are the price that all free nations must pay for their liberty in more substantial matters.” Id., at 344.9

IV

Having concluded that the application of § 3583(k)’s mandatory minimum in this

*2385

case violated Mr. Haymond’s right to trial by jury, we face the question of remedy. Recall that the Tenth Circuit declared the last two sentences of § 3583(k) “unconstitutional and unenforceable.” Those two sentences provide in relevant part that “[i]f a defendant required to register under [SORNA]” commits certain specified offenses, “the court shall revoke the term of supervised release and require the defendant to serve a term of imprisonment [of] not … less than 5 years.”

Before us, the government suggests that the Tenth Circuit erred in declaring those two sentences “unenforceable.” That remedy, the government says, sweeps too broadly. In the government’s view, any constitutional infirmity can be cured simply by requiring juries acting under the reasonable doubt standard, rather than judges proceeding under the preponderance of the evidence standard, to find the facts necessary to trigger § 3583(k)’s mandatory minimum. This remedy would be consistent with the statute’s terms, the government assures us, because “the court” authorized to revoke a term of supervised release in § 3583(k) can and should be construed as embracing not only judges but also juries. And, the government insists, that means we should direct the court of appeals to send this case back to the district court so a jury may be empaneled to decide whether Mr. Haymond violated § 3583(k). Unsurprisingly, Mr. Haymond contests all of this vigorously.

We decline to tangle with the parties’ competing remedial arguments today. The Tenth Circuit did not address these arguments; it appears the government did not even discuss the possibility of empaneling a jury in its brief to that court; and this Court normally proceeds as a “court of review, not of first view,” Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7, 125 S.Ct. 2113, 161 L.Ed.2d 1020 (2005). Given all this, we believe the wiser course lies in returning the case to the court of appeals for it to have the opportunity to address the government’s remedial argument in the first instance, including any question concerning whether that argument was adequately preserved in this case.

*

The judgment of the court of appeals is vacated, and the case is remanded for further proceedings.

It is so ordered.

Justice BREYER, concurring in the judgment.

I agree with much of the dissent, in particular that the role of the judge in a supervised-release proceeding is consistent with traditional parole. See post, at 2390-2391 (opinion of ALITO, J.). As 18 U.S.C. § 3583 makes clear, Congress did not intend the system of supervised release to differ from parole in this respect. And in light of the potentially destabilizing consequences, I would not transplant the Apprendi line of cases to the supervised-release context. See post, at 2388-2389; cf. Alleyne v. United States, 570 U.S. 99, 122, 133 S.Ct. 2151, 186 L.Ed.2d 314 (2013) (BREYER, J., concurring in part and concurring in judgment); United States v. Booker, 543 U.S. 220, 327, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005) (BREYER, J., dissenting in part); Blakely v. Washington, 542 U.S. 296, 329-330, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004) (BREYER, J., dissenting); Harris v. United States, 536 U.S. 545, 569-570, 122 S.Ct. 2406, 153 L.Ed.2d 524 (2002) (BREYER, J., concurring in part and concurring in judgment); Apprendi v. New Jersey, 530 U.S. 466, 555, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000) (BREYER, J., dissenting).

*2386

Nevertheless, I agree with the plurality that this specific provision of the supervised-release statute, § 3583(k), is unconstitutional. Revocation of supervised release is typically understood as “part of the penalty for the initial offense.” Johnson v. United States, 529 U.S. 694, 700, 120 S.Ct. 1795, 146 L.Ed.2d 727 (2000). The consequences that flow from violation of the conditions of supervised release are first and foremost considered sanctions for the defendant’s “breach of trust”—his “failure to follow the court-imposed conditions” that followed his initial conviction— not “for the particular conduct triggering the revocation as if that conduct were being sentenced as new federal criminal conduct.” United States Sentencing Commission, Guidelines Manual ch. 7, pt. A, intro. 3(b) (Nov. 2018); see post, at 2392-2393. Consistent with that view, the consequences for violation of conditions of supervised release under § 3583(e), which governs most revocations, are limited by the severity of the original crime of conviction, not the conduct that results in revocation. See § 3583(e)(3) (specifying that a defendant may as a consequence of revocation serve no “more than 5 years in prison if the offense that resulted in the term of supervised release is a class A felony, [no] more than 3 years in prison if … a class B felony,” and so on).

Section 3583(k) is difficult to reconcile with this understanding of supervised release. In particular, three aspects of this provision, considered in combination, lead me to think it is less like ordinary revocation and more like punishment for a new offense, to which the jury right would typically attach. First, § 3583(k) applies only when a defendant commits a discrete set of federal criminal offenses specified in the statute. Second, § 3583(k) takes away the judge’s discretion to decide whether violation of a condition of supervised release should result in imprisonment and for how long. Third, § 3583(k) limits the judge’s discretion in a particular manner: by imposing a mandatory minimum term of imprisonment of “not less than 5 years” upon a judge’s finding that a defendant has “commit[ted] any” listed “criminal offense.”

Taken together, these features of § 3583(k) more closely resemble the punishment of new criminal offenses, but without granting a defendant the rights, including the jury right, that attend a new criminal prosecution. And in an ordinary criminal prosecution, a jury must find facts that trigger a mandatory minimum prison term. Alleyne, 570 U.S. at 103, 133 S.Ct. 2151.

Accordingly, I would hold that § 3583(k) is unconstitutional and remand for the Court of Appeals to address the question of remedy. Because this is the course adopted by the plurality, I concur in the judgment.

Justice ALITO, with whom THE CHIEF JUSTICE, Justice THOMAS, and Justice KAVANAUGH join, dissenting.

I do not think that there is a constitutional basis for today’s holding, which is set out in Justice BREYER’s opinion, but it is narrow and has saved our jurisprudence from the consequences of the plurality opinion, which is not based on the original meaning of the Sixth Amendment, is irreconcilable with precedent, and sports rhetoric with potentially revolutionary implications. The plurality opinion appears to have been carefully crafted for the purpose of laying the groundwork for later decisions of much broader scope.

I

A

What do I mean by this? Many passages in the opinion suggest that the entire system

*2387

of supervised release, which has been an integral part of the federal criminal justice system for the past 35 years, is fundamentally flawed in ways that cannot be fixed. Under the Sentencing Reform Act of 1984 (SRA), whenever a federal court sentences a criminal defendant to a term of imprisonment, the court may include in the sentence a term of supervised release, and under some circumstances supervised release is mandatory. 18 U.S.C. § 3583. When a court imposes a term of supervised release, the order must specify the conditions with which the defendant is required to comply, § 3583(d), and a judge may revoke supervised release and send a defendant back to prison if the judge finds by a preponderance of the evidence that the defendant violated one of those conditions, § 3583(e)(3).

Many statements and passages in the plurality opinion strongly suggest that the Sixth Amendment right to a jury trial applies to any supervised-release revocation proceeding. Take the opinion’s opening line: “Only a jury, acting on proof beyond a reasonable doubt, may take a person’s liberty.” Ante, at 2373. In a supervised-release revocation proceeding, a judge, based on the preponderance of the evidence, may make a finding that “take[s] a person’s liberty,” ibid., in the sense that the defendant is sent back to prison. Later, after noting that the Sixth Amendment applies to a “criminal prosecution,” the plurality gives that term a broad definition that appears to encompass any supervised-release revocation proceeding. The plurality defines a “crime” as any “`ac[t] to which the law affixes … punishment,'” and says that a “prosecution” is “`the process of exhibiting formal charges against an offender before a legal tribunal.'” Ante, at 2376. These definitions explain what the terms in question mean in general use, but they were not formulated for the purpose of specifying what “criminal prosecution” means in the specific context of the Sixth Amendment. The plurality, however, uses them for precisely that purpose, and in so doing boldly suggests that every supervised-release revocation proceeding is a criminal prosecution. See ante, at 2379 (“[A] `criminal prosecution’ continues and the defendant remains an `accused’ with all the rights provided by the Sixth Amendment, until a final sentence is imposed…. [A]n accused’s final sentence includes any supervised release sentence he may receive”).

Later statements are even more explicit. Quoting Blakely v. Washington, 542 U.S. 296, 304, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004), out of context, the plurality states that “a jury must find beyond a reasonable doubt every fact which the law makes essential to a punishment that a judge might later seek to impose.” Ante, at 2370 (internal quotation marks and alteration omitted). If sending a defendant found to have violated supervised release back to prison is “punishment,” then the thrust of the plurality’s statement is that any factual finding needed to bring that about must be made by a jury, not by a judge, as is currently done.

Also telling is the plurality’s response to the Government’s argument that Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000), Blakely, and Alleyne v. United States, 570 U.S. 99, 133 S.Ct. 2151, 186 L.Ed.2d 314 (2013), apply only to a defendant’s sentencing proceeding and not to a supervised-release revocation proceeding, which the Government describes as a “postjudgment sentence-administration proceedin[g].” Brief for United States 24. Rejecting this argument, the plurality huffs that “the demands of the Fifth and Sixth Amendments” cannot be “dodge[d]” “by the simple expedient of relabeling a criminal prosecution a … `sentence modification’ imposed at a `postjudgment

*2388

sentence administration proceeding.'” Ante, at 2379. The meaning of this statement is unmistakable and cannot have been inadvertent: A supervised-release revocation proceeding is a criminal prosecution and is therefore governed by the Sixth Amendment (and the Fifth Amendment to boot). And there is more. See ante, at 2380 (“any accusation triggering a new and additional punishment [must be] proven to the satisfaction of a jury beyond a reasonable doubt”); ante, at 2380 (“a jury must find all of the facts necessary to authorize a judicial punishment”).

Finally, while the plurality appears to say that the Sixth Amendment does not apply to parole revocation proceedings, see ante, at 2381-2382,10 the plurality characterizes supervised release as “critical[ly] differen[t],” ante, at 2381-2382. This is so, the plurality explains, because parole relieved a prisoner from serving part of the prison sentence originally imposed, whereas a term of supervised release is added to the term of imprisonment specified by the sentencing judge. As I will explain, this difference is purely formal and should have no constitutional consequences. But for now the important point is the plain implication of what the plurality says: Parole was constitutional, but supervised release… well, that is an entirely different animal.

The intimation in all these statements is clear enough: All supervised-release revocation proceedings must be conducted in compliance with the Sixth Amendment— which means that the defendant is entitled to a jury trial, which means that as a practical matter supervised-release revocation proceedings cannot be held. In 2018, federal district courts completed 1809 criminal jury trials. Admin. Office of U. S. Courts, Judicial Business of the United States Courts (2018) (Table T-1). During that same year, they adjudicated 16,946 revocations of supervised release, ibid.(Table E-7A), and there is simply no way that the federal courts could empanel enough juries to adjudicate all those proceedings, let alone try all those proceedings in accordance with the Sixth Amendment’s Confrontation Clause. So, if every supervised-release revocation proceeding is a criminal prosecution, as the plurality suggests, the whole concept of supervised release will come crashing down.11

Where the plurality is headed is demonstrated —ironically—by its insistence that it is not going all the way—for now. The plurality writes: “[O]ur opinion,” ante, at 2383, 2383-2384, does “not pass judgment one way or the other on § 3583(e)’s consistency with Apprendi,” ante, at 2382, n. 7. Section 3583(e) sets out the procedure to be followed in all supervised-release revocation proceedings, so if that provision is not consistent with Apprendi, the whole idea of supervised release must fall. The strategy of the plurality opinion is only thinly veiled. It provides the framework to be used in ending supervised release. It provides no clear ground for limiting the rationale of the opinion so that it does not lead to that result. And then it says: We are not doing that today.

*2389

B

Is it possible to read the plurality opinion more narrowly? Can it be understood to condemn only one narrow statutory provision, namely, § 3583(k), which required the judge to send respondent Haymond back to prison for at least five years once the judge found that he had violated a condition of his supervised release by again possessing child pornography? On this reading, the only Sixth Amendment defect would be the mandatory minimum period of additional confinement that the statute imposes. There would be no problem if the judge had been free to choose the term, if any, of additional confinement. Does the plurality mean to go no further than this?

There are passages in the opinion that hint at this narrower interpretation. The plurality analogizes the mandatory minimum term of additional confinement required by § 3583(k) to the mandatory minimum term of initial imprisonment found to violate the Sixth Amendment in Alleyne, see ante, at 2377-2379. But the previously quoted statements pointing to a broader understanding remain, and the plurality does nothing to disavow that reading. To the contrary, the plurality doubles down, assuring us that this broader understanding would not be too disruptive. See ante, at 2383-2384.

A narrower interpretation of the plurality opinion is also contradicted by another important statement in the opinion. The plurality says that the maximum “lawful prison term” “reflected in the jury’s verdict” in respondent’s case was “10 years.” Ante, at 2378. This statement is full of meaning because if 10 years is the maximum amount of time that respondent could lawfully be required to spend in prison on the basis of the jury’s verdict, there is a serious constitutional defect in the very design of the supervised-release system. That is so because the concept of supervised release is based on a fundamentally different conception of the maximum term of confinement authorized by a guilty verdict.

To understand this, it is important to understand the relationship between the system of supervised release and the old federal parole system it replaced. By abolishing parole and substituting supervised release, the SRA sought to retain the chief benefit of parole, i.e., providing a transition period of monitoring to ensure that a prisoner who leaves prison has been sufficiently reformed so that he is able to lead a law-abiding life. At the same time, the SRA aimed to promote truth in sentencing and thus to eliminate a much-derided feature of the old parole system. See United States Sentencing Commission, Guidelines Manual ch. 1, pt. A (Nov. 2018) (USSG). Under the parole system, a defendant who was convicted of a serious crime and given what seemed to be a stiff sentence could be and not infrequently was set free after serving only a fraction of the sentence originally pronounced. A prisoner was generally eligible for parole after serving only one-third of his sentence, and a sentence of life was treated as a sentence of 30 years.12  Therefore, a defendant sentenced to imprisonment for life could be out on the streets after only 10 years.

The SRA changed this, and now a defendant must serve the full term of imprisonment imposed at sentencing minus only a small deduction for good behavior in prison. USSG ch. 1, pt. A.1(3); 18 U.S.C. § 3624(b); Barber v. Thomas, 560 U.S. 474, 481-482, 130 S.Ct. 2499, 177 L.Ed.2d 1 (2010). But to provide the same sort of transition period as was furnished under parole, a sentencing court may, and in

*2390

some cases must, add a period of supervised release. See § 3583. The replacement of parole with supervised release changed the form of federal sentences but not their substance. Here is an example: A pre-SRA sentence of nine years’ imprisonment meant three years of certain confinement and six years of possible confinement depending on the defendant’s conduct in the outside world after release from prison. At least for present purposes, such a sentence is the substantive equivalent of a post-SRA sentence of three years’ imprisonment followed by six years of supervised release. In both situations, the period of certain confinement (three years) and the maximum term of possible confinement (nine years) are the same. If anything, the defendant in the post-SRA case is treated more favorably because he is guaranteed release from prison after three years; his release at that point is not dependent on a decision by a parole board.

As this example shows, the concept of supervised release rests on the idea that a defendant sentenced to x years of imprisonment followed by y years of supervised release is really sentenced to a maximum punishment of x + y years of confinement, with the proviso that any time beyond x years will be excused if the defendant abides by the terms of supervised release. And on this understanding, the maximum term reflected in the jury’s verdict in respondent’s case was not 10 years, as the plurality claims, but 10 years plus the maximum period of supervised release that the statute authorized.13

None of this matters in respondent’s case because the sum of his original sentence (38 months) and the additional time imposed for violating supervised release (60 months) is less than 120 months, but adoption of the rule toward which the plurality opinion seems to point would make a big difference in many cases. Under that rule, a term of supervised release could never be ordered for a defendant who is sentenced to the statutory maximum term of imprisonment, and only a short period of supervised release could be ordered for a defendant sentenced to a term of imprisonment that is close to the statutory maximum. Moreover, in many cases, a judge, before beginning a supervised-release revocation proceeding, would have to anticipate the period of additional confinement that the judge would find appropriate if a particular violation or set of violations was shown. For example, suppose that the statutory maximum term of certain confinement authorized by the offense of conviction is 10 years and that a prisoner is sentenced to and serves eight years. Suppose that the term of supervised release imposed at the time of sentencing is five years. Before starting a supervised-release revocation proceeding in this hypothetical case, the judge would have to decide whether to rule out the possibility of sending the defendant back to prison for more than two years. Unless the judge was willing

*2391

to do this—without knowing all the facts—the judge would have to convene a jury. It would be strange to put judges in that predicament.

The plurality appreciates the implication of its understanding of the maximum term of imprisonment authorized by a jury verdict in the post-SRA era. In footnote 4, the plurality says that it need not decide whether its interpretation of the Sixth Amendment leads to the results I have just outlined. See ante, at 2378-2379, n. 4. But here again, while formally reserving decision on this question, the opinion provides no theory that might permit what the SRA contemplates.

In short, under the plurality opinion, the whole system of supervised release would be like a 40-ton truck speeding down a steep mountain road with no brakes.

II

This should not have been a difficult or complicated case. I start with the proposition that the old federal parole system did not implicate the Sixth Amendment’s jury trial right. A parole revocation proceeding was not a “criminal prosecution” within the meaning of the Sixth Amendment, and revocation did not result in a new sentence. See, e.g., United States v. Williams, 558 F.2d 224, 226 (C.A.5 1977); Hyser v. Reed, 318 F.2d 225, 237 (C.A.D.C. 1963). When a prisoner was paroled, the Executive was simply exercising the authority conferred by law to grant the defendant a conditional release from serving part of the sentence imposed after a guilty verdict. Mistretta v. United States, 488 U.S. 361, 364-365, 109 S.Ct. 647, 102 L.Ed.2d 714 (1989).

Supervised release, for reasons already explained, is not fundamentally different and therefore should not be treated any differently for Sixth Amendment purposes. When a jury finds a federal defendant guilty of violating a particular criminal statute, the maximum period of confinement authorized is the maximum term of imprisonment plus the maximum term of supervised release. If a prisoner does not end up spending this full period in confinement, that is because service of part of the period is excused due to satisfactory conduct during the period of supervised release. Any other reading exalts form over substance in a way that has enormous consequences that cannot be justified on constitutional grounds.

Once this is understood, it follows that the procedures that must be followed at a supervised-release revocation proceeding are the same that had to be followed at a parole revocation proceeding, and these were settled long ago. At a parole revocation hearing, the fundamental requisites of due process had to be observed, but a parolee did not have a right to a jury trial. See, e.g., United States v. Carlton, 442 F.3d 802, 807 (C.A.2 2006); United States v. Huerta-Pimental, 445 F.3d 1220, 1225 (C.A.9 2006). Neither the Confrontation Clause nor the formal rules of evidence had to be followed. See, e.g., Morrissey v. Brewer, 408 U.S. 471, 488-489, 92 S.Ct. 2593, 33 L.Ed.2d 484 (1972); Gagnon v. Scarpelli, 411 U.S. 778, 782, n. 5, 93 S.Ct. 1756, 36 L.Ed.2d 656 (1973). Due process did not require proof beyond a reasonable doubt as is necessary at trial, see, e.g., DeWitt v. Ventetoulo, 6 F.3d 32, 36-37 (C.A.1 1993); Whitehead v. United States Parole Comm’n, 755 F.2d 1536, 1537 (C.A.11 1985); Mack v. McCune, 551 F.2d 251, 254 (C.A.10 1977); and the Double Jeopardy Clause did not apply, see, e.g., Kell v. United States Parole Comm’n, 26 F.3d 1016, 1020 (C.A.10 1994) (citing cases).

For the past 35 years, it has been understood that the same rules apply at a supervised-release revocation proceeding. There

*2392

is no good reason to depart from that understanding.

III

The plurality tries to suggest a reason by sprinkling its opinion with quotations from venerable sources, but all are far afield. (John Adams was not writing about the Sixth Amendment when he made a diary entry in 1771 or when he wrote to William Pym in 1766. See ante, at 2375-2376.) And the plurality makes no real effort to show that the Sixth Amendment was originally understood to require a jury trial in a proceeding like a supervised-release revocation proceeding. Of course, nothing like supervised release—or for that matter, parole—existed when the Sixth Amendment was ratified, so I will not attempt to make the affirmative case that the Sixth Amendment was specifically understood not to apply to such proceedings. But there is a strong case for the proposition that the terms of the Sixth Amendment and the original understanding of the scope of the jury trial right do not require the plurality’s interpretation. And our prior precedents emphatically refute that interpretation.

The Sixth Amendment limits the scope of the jury trial right in three significant ways: It provides “who may assert the right (`the accused’); when the right may be asserted (`[i]n all criminal prosecutions’); and what the right guarantees” (“the right to a … trial, by an impartial jury”). Rothgery v. Gillespie County, 554 U.S. 191, 214, 128 S.Ct. 2578, 171 L.Ed.2d 366 (2008) (ALITO, J., concurring). The plurality can reach its conclusion only by ignoring these limitations.

A

I begin with who may assert the jury trial right. The text of the Sixth Amendment makes clear that this is “a right of the `accused’ and only the `accused.'” A. Amar, The Bill of Rights 111 (1998). The “accused” is an individual “[c]harged with a crime, by a legal process.” N. Webster, An American Dictionary of the English Language (1828); see also 2 J. Bouvier, Law Dictionary 50 (10th ed. 1860) (Bouvier Law Dictionary) (“One who is charged with a crime or misdemeanor”).

“At the founding, `accused’ described a status preceding `convicted.'” Betterman v. Montana, 578 U. S. ___, ___, 136 S.Ct. 1609, 1614, 194 L.Ed.2d 723 (2016). Blackstone, for example, spoke of “the accused” in outlining the beginning of a criminal prosecution, see 4 W. Blackstone, Commentaries on the Laws of England 313 (1769), and spoke of “the offender” and “the criminal” after conviction, see id., at 370, 371, 373, 378, 379. See also id., at 279 (referring to “the party accused before he is condemned”). And “[t]his understanding of the Sixth Amendment language—`accused’ as distinct from `convicted’ …— endures today.” Betterman, 578 U. S., at ___, 136 S.Ct., at 1614 (citing Black’s Law Dictionary 26 (10th ed. 2014) (defining “accused” as “a person who has been arrested and brought before a magistrate or who has been formally charged” (emphasis added))).

Despite the plurality’s suggestion otherwise, see ante, at 2379-2380, respondent was no longer the “accused” while he served his term of supervised release. To be sure, he was formerly the accused—at the time when he was duly indicted and tried for possession of child pornography. But after a jury convicted him and authorized the judge to sentence him to terms of imprisonment and supervised release, respondent was transformed into the convicted. And his status as such remained the same while he served his sentences, including during the proceeding to determine whether he had adhered to the conditions

*2393

attached to the term of supervised release that was permitted by law and thus implicitly authorized by the jury’s verdict.

This is especially so given that respondent’s reimprisonment was not primarily a punishment for new criminal conduct. The principal reason for assigning a penalty to a supervised-release violation is not that the violative act is a crime (indeed, under other provisions in § 3583, the act need not even be criminal); rather, it is that the violative act is a breach of trust. USSG ch. 7, pt. A, intro. 3(b) (recommended reimprisonment terms are designed to “sanction primarily the defendant’s breach of trust,” not “new criminal conduct”). In other words, it makes little sense to treat respondent as the accused—i.e., one charged with a crime—when he has been charged not with a crime, but with violating the terms of a jury-authorized sentence that flowed from his original conviction. The plurality’s extension of the jury trial right to respondent’s supervised-release revocation proceeding thus flounders from the start for the simple reason that respondent cannot easily be viewed as an “accused” in the conventional sense of the term.

B

It is similarly awkward to characterize a supervised-release revocation proceeding as part of the defendant’s “criminal prosecution.” A supervised-release revocation proceeding is not part of the criminal prosecution that landed a defendant in prison in the first place because “[a] `criminal prosecution’ … ends when sentence has been pronounced on the convicted or a verdict of `Not guilty’ has cleared the defendant of the charge.” F. Heller, Sixth Amendment to the Constitution of the United States 54 (1951). This follows from the early understanding that a “prosecution” concludes when a court enters final judgment. See, e.g., Webster, An American Dictionary of the English Language (defining a prosecution as the “process of exhibiting formal charges against an offender before a legal tribunal, and pursuing them to final judgment” (emphasis added)); The Universal English Dictionary 465 (J. Craig ed. 1869) (“[T]he institution of legal proceedings against a person; the process of exhibiting formal charges against an offender before a legal tribunal, and pursuing them to final judgment” (emphasis added)); H. Holthouse, New Law Dictionary 344 (1847) (defining prosecution as “the means adopted to bring a supposed offender to justice and punishment by due course of law”); Bouvier Law Dictionary 396 (“The means adopted to bring a supposed offender to justice and punishment by due course of law”).

Our precedents reflect this understanding by defining the end of criminal prosecutions to be the entry of final judgment and imposition of sentence. In the Sixth Amendment context, for example, the Court has explained that “[c]riminal proceedings generally unfold in three discrete phases”: a prearrest phase, a charging phase that extends through trial, and a sentencing phase. Betterman, 578 U. S., at ___, 136 S.Ct., at 1611. As the Court described the final phase, the criminal proceeding ends “[a]fter conviction, [when] the court imposes sentence.” Ibid.; see also id., at ___, 136 S.Ct., at 1614 (“And `trial’ meant a discrete episode after which judgment (i.e., sentencing) would follow”). That description echoed the Court’s earlier characterization of the process, beginning to end: “criminal indictment, trial by jury, and judgment by court.” Apprendi, 530 U.S. at 478, 120 S.Ct. 2348; see also ibid., n. 4 (citing Blackstone to explain that “`judgment’ by the court” was “the stage approximating in modern terms the imposition of sentence” (emphasis added)). And even outside the Sixth Amendment context,

*2394

we have said that “[t]he general rule is that finality in the context of a criminal prosecution is defined by a judgment of conviction and the imposition of sentence.” Fort Wayne Books, Inc. v. Indiana, 489 U.S. 46, 54, 109 S.Ct. 916, 103 L.Ed.2d 34 (1989).

In fact, two prior precedents—which the plurality effectively ignores—drew this exact line in stating that parole- and probation-revocation proceedings are not part of a criminal prosecution. Unless the plurality is willing to own up to attempting to overrule these precedents, its failure to engage with them is inexcusable.

The first is Morrissey, 408 U.S. at 472, 92 S.Ct. 2593, a landmark case in which the Court held that due process requires a State to afford a parolee “some opportunity to be heard” before revoking parole. In considering that question, the Court “beg[an] with the proposition that the revocation of parole is not part of a criminal prosecution and thus the full panoply of rights due a defendant in such a proceeding does not apply in parole revocations.” Id., at 480, 92 S.Ct. 2593. The Court made clear that “[p]arole arises after the end of the criminal prosecution, including imposition of sentence.” Ibid. (emphasis added).

The second is Gagnon, 411 U.S. 778, 93 S.Ct. 1756, where the Court considered whether a probationer has a right to appointed counsel prior to the revocation of probation. There, the Court reasoned that “[p]robation revocation, like parole revocation, is not a stage of a criminal prosecution.” Id., at 782, 93 S.Ct. 1756. Thus, in both contexts, the Court emphasized that parole- and probation-revocation proceedings are not part of a criminal prosecution. And that understanding carried significant consequences: It denied parolees and probationers the “full panoply of rights” to which a defendant is entitled in a criminal prosecution. Morrissey, 408 U.S. at 480, 92 S.Ct. 2593.

Supervised-release revocation proceedings are not part of the defendant’s criminal prosecution for the same reasons. As we said in United States v. Johnson, 529 U.S. 53, 59, 120 S.Ct. 1114, 146 L.Ed.2d 39 (2000), which the plurality all but ignores, “[s]upervised release has no statutory function until confinement ends,” which itself has no function until the criminal prosecution has ended. It follows, then, that “the revocation of [supervised release] is not part of a criminal prosecution.” Morrissey, 408 U.S. at 480, 92 S.Ct. 2593.

The fact that Morrissey and Gagnon involved parole and probation, not supervised release, does not matter for present purposes. Cf. ante, at 2376-2377, 2384-2382. These cases did not turn on any features of parole or probation that might distinguish them from supervised release. Rather, those decisions recognized an obvious fact: The administration of a sentence occurs after a court imposes that sentence —i.e., after the criminal prosecution has ended. That fact is equally true here. No matter what penalties flow from the revocation of parole, probation, or supervised release, the related proceedings are not part of the criminal prosecution.

In recognition of this, the courts of appeals for the past 35 years have overwhelmingly declined to apply the Sixth Amendment in supervised-release revocation proceedings, and they have done so precisely on the ground that these proceedings are not part of criminal prosecutions. This is true as to the jury trial right;14  the Speedy Trial Clause;15  the Confrontation

*2395

Clause;16  and the right to counsel.17  As then-Judge Gorsuch succinctly put it not too long ago, “settled precedent” dictates that Sixth Amendment rights “d[o] not apply to supervised release revocation proceedings and the due process guarantees associated with these proceedings are `minimal.'” United States v. Henry, 852 F.3d 1204, 1206-1207 (C.A.10 2017) (quoting Morrissey, 408 U.S. at 485, 489, 92 S.Ct. 2593). And even the court below agreed: “Revocation of supervised release is not part of a criminal prosecution, so defendants accused of a violation of the conditions of supervised release have no right to a jury determination of the facts constituting that violation.” 869 F.3d 1153, 1163 (C.A.10 2017).

Attempting to claim that a criminal prosecution actually extends through any period of supervised release, the plurality appears to arrive at an unintended destination. The plurality says (while mischaracterizing Apprendi and Alleyne, see infra, at 2395) that “a `criminal prosecution’ continues and the defendant remains an `accused’ with all the rights provided by the Sixth Amendment, until a final sentence is imposed.” Ante, at 2387. That is exactly right. And the Court’s precedents emphatically say that a sentence is “imposed” at final judgment, supra, at 2392-2393, not again and again every time a convicted criminal wakes up to serve a day of supervised release and violates a condition of his release. That postjudgment conduct during the administration of supervised release, and any proceedings to adjudicate violations of the release conditions, necessarily occurs “after the end of the criminal prosecution, including imposition of sentence.” Morrissey, 408 U.S. at 480, 92 S.Ct. 2593 (emphasis added).

C

The plurality attempts to pass off its reasoning as nothing more than the logical outgrowth of the Apprendi line of cases, but that is untrue. The plurality invokes these cases to support the idea that the Sixth Amendment cannot be evaded by “[r]elabeling” of a criminal prosecution as a “`sentence modification'” imposed at a “`postjudgment sentence-administration proceeding.'” Ante, at 2392-2393; see also ibid. (claiming that Apprendi “recognized” how long a criminal prosecution continues). But nothing like that was involved in Apprendi or later related cases. Instead, the Court in those cases rejected what it saw as attempts to place the label “sentencing enhancement” on what, in its view, were essentially elements of charged offenses. See, e.g., Blakely, 542 U.S. at 306, 124 S.Ct. 2531 (rejecting the idea that “the jury need only find whatever facts the legislature chooses to label elements of the crime, and that those it labels sentencing factors—no matter how much they may increase the punishment—may be found by the judge”). All of the cases in the Apprendi line involved actual sentencing

*2396

proceedings, and thus there was never any question whether they arose in a “criminal prosecution.” That is not this case.

The plurality insists that it is simply applying Apprendi‘s understanding of the jury trial right when it says that “a jury must find beyond a reasonable doubt every fact which the law makes essential to a punishment that a judge might later seek to impose.” Ante, at 2387 (internal quotation marks and alteration omitted). But that is wrong.

1

Since Apprendi itself, the Court has time and again endeavored to draw its understanding of the jury trial right from historical practices that existed at the founding and soon afterward. See Apprendi, 530 U.S. at 495, 120 S.Ct. 2348 (looking to the “historical pedigree of the jury”); Alleyne, 570 U.S. at 111, 133 S.Ct. 2151 (emphasizing that Apprendi looked to “common-law and early American practice”). As Justices GINSBURG and SOTOMAYOR recently explained, courts applying Apprendi must “examine the historical record, because `the scope of the constitutional jury right must be informed by the historical role of the jury at common law.'” Southern Union Co. v. United States, 567 U.S. 343, 353, 132 S.Ct. 2344, 183 L.Ed.2d 318 (2012) (quoting Oregon v. Ice, 555 U.S. 160, 170, 129 S.Ct. 711, 172 L.Ed.2d 517 (2009)); see also id., at 167-168, 129 S.Ct. 711 (“Our application of Apprendi‘s rule must honor the `longstanding common-law practice’ in which the rule is rooted” (quoting Cunningham v. California, 549 U.S. 270, 281, 127 S.Ct. 856, 166 L.Ed.2d 856 (2007))). Thus, where “[t]he historical record demonstrates that the jury played no role” in a particular context, Ice, 555 U.S. at 168, 129 S.Ct. 711, there is “no encroachment … by the judge upon facts historically found by the jury,” id., at 169, 129 S.Ct. 711, and Apprendi does not govern.

In this case, the plurality can muster no support for the proposition that the jury trial right was extended to anything like a supervised-release or parole revocation proceeding at the time of the adoption of the Sixth Amendment. Supervised release was not instituted until 1984, and parole was unknown until the 19th century, so close historic analogues are lacking. But the nearest practices that can be found do not support the plurality.

Prior to and at the time of the adoption of the Sixth Amendment, convicted criminals were often released on bonds and recognizances that made their continued liberty contingent on good behavior. See L. Friedman, Crime and Punishment in American History 38-39 (1993); A. Hirsch, The Rise of the Penitentiary 7 (1992) (“Since courts in the eighteenth-century frequently demanded that offenders provide monetary sureties for future good behavior, convicts stayed put until they scraped together the requisite funds”). If a prisoner released on such a bond did not exhibit good behavior, the courts had discretion to forfeit the bond (a loss of property) or to turn the individual over to the sheriff (a loss of liberty) until new conditions could be arranged. See Friedman, supra, at 39. There is no evidence that there was a right to a jury trial at such proceedings, and the plurality does not even attempt to prove otherwise.

Corporal punishment of prisoners is also inconsistent with the plurality’s suggestion that a convicted criminal has the right to a jury trial before a punishment is imposed for legally proscribed conduct. See ante, at 2376. Well into the 19th century, prisoners were whipped for misbehavior. See Friedman, supra, at 37, 77, n. *; M. Kann, Punishment, Prisons, and Patriarchy 120, 182 (2005). Virginia law, for example, provided

*2397

that a prisoner could be punished “by stripes” if he were guilty of “profanity, indecent behavior, idleness, neglect or willful mismanagement of work, insubordination, an assault not amounting to felony, or a violation of any of the rules prescribed by the governor.” Va. Code, Tit. 56, ch. 213, § 22 (1849). Massachusetts law gave the warden “all necessary means” “to suppress insurrection, enforce obedience, and maintain order in the prison,” provided however “that no convict shall be punished… by more than ten stripes” without meeting certain conditions. Mass. Gen. Laws, ch. CXVIII, § 21 (1828). And even at the turn of the century, courts entertained imposition of reasonable corporal punishment provided that it was authorized by lawfully adopted rule or regulation. See, e.g., State v. Nipper, 166 N.C. 272, 277-280, 81 S.E. 164, 167-168 (1914); Davis v. State, 81 Miss. 56, 33 So. 286 (1902); Werner v. State, 44 Ark. 122, 131-132 (1884); Cornell v. State, 74 Tenn. 624, 624-631 (1881). There is no suggestion in these authorities that a jury finding of a violation was needed.18

Later, when parole and probation were introduced, courts, with the assistance of parole and probation officials, supervised the conditional release of parolees and probationers, and juries played no part in this process. See 4 Atty. Gen.’s Survey of Release Proc. 1 (1939) (Parole Survey); 2 id., at 2 (Probation Survey).

The well-settled revocation power wielded by courts and other officials brings this point home. A violation of the conditions permitted not only the defendant’s reimprisonment, see Parole Survey 4; Probation Survey 2, but several other penalties as well. In the parole context, these penalties most often included the forfeiture of good time credits—a reduction in prison time based on good behavior—that the parolees had accrued prior to their release on parole, as well as the forfeiture of any time served for the duration of their parole. Parole Survey 249-253; see also Friedman, supra, at 159 (stating in the context of 19th century good time laws that “[t]o forfeit `good time’ was a terrible penalty”). Many States also conditioned the future availability of parole on mandatory minimum terms of reimprisonment, and others even rendered certain parole violators ineligible for future parole. Parole Survey 255-258. And in the probation context, several courts refused to give credit for time spent on probation. Probation Survey 334-335, and n. 52. Thus, courts and parole boards could not only revoke conditional liberty but they could also subject violators to longer periods of

*2398

imprisonment and erase the fact that the violators had served a substantial portion of their lives on the streets under strict conditions.

From each of the foregoing examples, a clear historical fact emerges: American juries have simply played “no role” in the administration of previously imposed sentences. Ice, 555 U.S. at 168, 129 S.Ct. 711. As a result, it is impossible to say with a straight face that the “application of Apprendi‘s rule” to supervised-release revocation proceedings “honor[s] the `longstanding common-law practice’ in which the rule is rooted.” Id., at 167-168, 129 S.Ct. 711 (quoting Cunningham, 549 U.S. at 281, 127 S.Ct. 856).

2

The plurality’s extension of the jury trial right to the administration of previously imposed sentences also sidelines what has until now been the core feature of the Apprendi line of cases—a meaningful connection to the trial for the charged offense. “The touchstone for determining whether a fact must be found by a jury beyond a reasonable doubt is whether the fact constitutes an `element’ or `ingredient’ of the charged offense.” Alleyne, 570 U.S. at 107, 133 S.Ct. 2151 (plurality opinion); see also Southern Union Co., 567 U.S. at 349, 132 S.Ct. 2344 (“Apprendi‘s `core concern’ is to reserve to the jury `the determination of facts that warrant punishment for a specific statutory offense'” (quoting Ice, 555 U.S. at 170, 129 S.Ct. 711)); Ice, 555 U.S. at 168, 129 S.Ct. 711 (noting the jury’s historic role as a “bulwark” between the government and the accused “at the trial for an alleged offense” (emphasis added)). The Court’s rationale has been that “the core crime and the fact triggering [an increased maximum or] mandatory minimum sentence together constitute a new, aggravated crime, each element of which must be submitted to the jury.” Alleyne, 570 U.S. at 113, 133 S.Ct. 2151. And this rationale, of course, is key to the Apprendi line of cases, because the Sixth Amendment protects only the rights of “the accused,” that is, those charged with a particular crime. See supra, at 2391-2393.

In Apprendi itself, the Court emphasized the relevance of the charged offense when distinguishing Almendarez-Torres v. United States, 523 U.S. 224, 118 S.Ct. 1219, 140 L.Ed.2d 350 (1998). The Court explained that the “reasons supporting [a recidivism] exception” in Almendarez-Torres did not apply in Apprendi because, “[w]hereas recidivism `does not relate to the commission of the offense’ itself, New Jersey’s biased purpose inquiry goes precisely to what happened in the `commission of the offense.'” Apprendi, 530 U.S. at 496, 120 S.Ct. 2348 (quoting Almendarez-Torres, 523 U.S. at 230, 244, 118 S.Ct. 1219).

Here, the factual basis for revoking respondent’s supervised release did not “g[o] precisely to what happened in the `commission of the offense'”; it did not even “relate to the commission of the offense.” Apprendi, 530 U.S. at 496, 120 S.Ct. 2348. It had virtually nothing to do with the child-pornography offense that led to respondent’s conviction, incarceration, and supervised release. The same would be true of a defendant convicted of burglary, arson, or any other crime: His failure to attend an employment class or to pass a drug test while on supervised release would have nothing to do with how he carried out those offenses. And it would be impossible for “the core crime” and a postjudgment fact affecting respondent’s sentence to be submitted “together” as one “new, aggravated crime” for proof to a jury. Alleyne, 570 U.S. at 113, 133 S.Ct. 2151. Thus, no reasonable person would describe such postjudgment facts that go

*2399

only to the administration of a previously imposed sentence as “ingredients” or “elements” of the charged offense. Insofar as the charged statutory offense has been part and parcel of “Apprendi‘s core concern,” that concern “is inapplicable to the issue at hand,” and thus, “so too is the Sixth Amendment’s restriction on judge-found facts.” Ice, 555 U.S. at 170, 129 S.Ct. 711.

It is telling that the plurality never brings itself to acknowledge this clear departure from the Apprendi line of cases. For nearly two decades now, the Court has insisted that these cases turn on “a specific statutory offense,” and its “ingredients” and “elements.” Yet today we learn that— at least as far as the plurality is concerned —none of that really mattered.

3

The plurality also errs by failing to distinguish between the unconditional liberty interests with which Apprendi is concerned and the conditional liberty interests at issue in cases like this one. Cf. ante, at 2373 (“Only a jury, acting on proof beyond a reasonable doubt, may take a person’s liberty”). When a person is indicted and faces the threat of prison and supervised release, his unconditional liberty hangs in the balance. See Apprendi, 530 U.S. at 476, 120 S.Ct. 2348 (“At stake in this case are constitutional protections of surpassing importance: the proscription of any deprivation of liberty without `due process of law,’ Amdt. 14 …”); id., at 484, 120 S.Ct. 2348 (“If a defendant faces punishment beyond that provided by statute when an offense is committed under certain circumstances but not others, it is obvious that both the loss of liberty and the stigma attaching to the conviction are heightened”); id., at 495, 120 S.Ct. 2348 (“The degree of criminal culpability the legislature chooses to associate with particular, factually distinct conduct has significant implications both for a defendant’s very liberty, and for the heightened stigma associated with an offense the legislature has selected as worthy of greater punishment”).

But convictions have consequences. “[G]iven a valid conviction, the criminal defendant [may be] constitutionally deprived of his liberty.” Meachum v. Fano, 427 U.S. 215, 224, 96 S.Ct. 2532, 49 L.Ed.2d 451 (1976). To this end, “[s]upervised release is `a form of postconfinement monitoring’ that permits a defendant a kind of conditional liberty by allowing him to serve part of his sentence outside of prison.” Mont v. United States, 587 U. S. ___, ___-___, 139 S.Ct. 1826, 1833, ___ L.Ed.2d ___ (2019) (quoting Johnson, 529 U.S. at 697, 120 S.Ct. 1795). Convicts like respondent on supervised release thus enjoy only conditional liberty. He most certainly was not “a free man.” Ante, at 2382-2383. This means, then, that “[r]evocation” of supervised release “deprives an individual, not of the absolute liberty to which every citizen is entitled, but only of… conditional liberty.” Morrissey, 408 U.S. at 480, 92 S.Ct. 2593. It is perhaps for that reason that the decisions of this Court that mention “conditional liberty” speak only of general due process rights, not other constitutional protections that unaccused and unconvicted individuals enjoy. See, e.g., Connecticut Bd. of Pardons v. Dumschat, 452 U.S. 458, 101 S.Ct. 2460, 69 L.Ed.2d 158 (1981); Vitek v. Jones, 445 U.S. 480, 100 S.Ct. 1254, 63 L.Ed.2d 552 (1980); Wolff v. McDonnell, 418 U.S 539, 94 S.Ct. 2963, 41 L.Ed.2d 935 (1974); Morrissey, 408 U.S. 471, 92 S.Ct. 2593.

* * *

Today’s decision is based in part on an opinion that is unpardonably vague and suggestive in dangerous ways. It is not grounded on any plausible interpretation

*2400

of the original meaning of the Sixth Amendment, and it is contradicted by precedents that are unceremoniously overruled. It represents one particular view about crime and punishment that is ascendant in some quarters today but is not required by the Constitution. If the Court eventually takes the trip that this opinion proposes, the consequences will be far reaching and unfortunate.

For these reasons, I respectfully dissent.

Footnotes

[*] The syllabus constitutes no part of the opinion of the Court but has been prepared by the Reporter of Decisions for the convenience of the reader. See United States v. Detroit Timber & Lumber Co., 200 U.S. 321, 337, 26 S.Ct. 282, 50 L.Ed. 499.

1 Section 3583(e)(3) states in pertinent part: “The court may, after considering the factors set forth in section 3553(a)(1), … (3) revoke a term of supervised release, and require the defendant to serve in prison all or part of the term of supervised release authorized by statute for the offense that resulted in such term of supervised release without credit for time previously served on postrelease supervision, if the court … finds by a preponderance of the evidence that the defendant violated a condition of supervised release, except that a defendant whose term is revoked under this paragraph may not be required to serve on any such revocation more than 5 years in prison if the offense that resulted in the term of supervised release is a class A felony, more than 3 years in prison if such offense is a class B felony, more than 2 years in prison if such offense is a class C or D felony, or more than one year in any other case ….”

2 Section 3583(k) provides: “Notwithstanding subsection (b), the authorized term of supervised release for any offense under section 1201 involving a minor victim, and for any offense under section 1591, 1594(c), 2241, 2242, 2243, 2244, 2245, 2250, 2251, 2251A, 2252, 2252A, 2260, 2421, 2422, 2423, or 2425, is any term of years not less than 5, or life. If a defendant required to register under the Sex Offender Registration and Notification Act (SORNA) commits any criminal offense under chapter 109A, 110, or 117, or section 1201 or 1591, for which imprisonment for a term longer than 1 year can be imposed, the court shall revoke the term of supervised release and require the defendant to serve a term of imprisonment under subsection (e)(3) without regard to the exception contained therein. Such term shall be not less than 5 years.”

3 The Court has recognized two narrow exceptions to Apprendi‘s general rule, neither of which is implicated here: Prosecutors need not prove to a jury the fact of a defendant’s prior conviction, Almendarez-Torres v. United States, 523 U.S. 224, 118 S.Ct. 1219, 140 L.Ed.2d 350 (1998), or facts that affect whether a defendant with multiple sentences serves them concurrently or consecutively, Oregon v. Ice, 555 U.S. 160, 129 S.Ct. 711, 172 L.Ed.2d 517 (2009).

4 Because we hold that this mandatory minimum rendered Mr. Haymond’s sentence unconstitutional in violation of Alleyne v. United States, 570 U.S. 99, 133 S.Ct. 2151, 186 L.Ed.2d 314 (2013), we need not address the constitutionality of the statute’s effect on his maximum sentence under Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000).

5 The dissent asserts that “a sentence is `imposed’ at final judgment, not again and again every time a convicted criminal … violates a condition of his release.” Post, at 2395 (opinion of ALITO, J.) (citation omitted). But saying it does not make it so. As Johnson recognized, when a defendant is penalized for violating the terms of his supervised release, what the court is really doing is adjusting the defendant’s sentence for his original crime. Even the dissent recognizes that the sword of Damocles hangs over a defendant “every time [he] wakes up to serve a day of supervised release.” Post, at 2395.

6 But perhaps we underestimate their fervor. While not openly embracing that result, the dissent fails to articulate any meaningful limiting principle to avoid it. If, as the dissent suggests, a term of supervised release is interchangeable with whatever sanction is prescribed for a violation, why stop at life in prison? The dissent replies that we might discover some relevant limitation in the Eighth Amendment, which does not mention jury trials, but is unwilling to find that limitation in the Sixth Amendment, which does. Post, at 2390, n. 4.

7 Just as we have no occasion to decide whether § 3583(k) implicates Apprendi by raising the ceiling of permissible punishments beyond those authorized by the jury’s verdict, see n. 4, supra, we do not pass judgment one way or the other on § 3583(e)’s consistency with Apprendi. Nor do we express a view on the mandatory revocation provision for certain drug and gun violations in § 3583(g), which requires courts to impose “a term of imprisonment” of unspecified length.

8 Contrary to the dissent’s characterization, we do not suggest that any prison discipline that is “too harsh” triggers the right to a jury trial. Post, at 2396-2397, n. 9 (emphasis deleted). Instead, we distinguish between altering a prisoner’s conditions of confinement, which generally does not require a jury trial, and sentencing a free man to substantial additional time in prison, which generally does.

9 Justice BREYER agrees that a jury was required here for three reasons “considered in combination.” Post, at 2386 (opinion concurring in judgment). Two of the reasons seem to amount to the same thing—a worry that § 3583(k) imposes a new mandatory minimum sentence without a jury. And for the reasons we’ve already given, we can agree that this is indeed a problem under Alleyne. But Justice BREYER’s remaining reason is another story. He stresses that § 3583(k)’s mandatory minimum applies only to a “discrete set of federal criminal offenses.” Post, at 2386. But why should that matter? Whether the Sixth Amendment is violated in “discrete” instances or vast numbers, our duty to enforce the Constitution remains the same. Besides, any attempt to draw lines based on when an erosion of the jury trial right goes “too far” would prove inherently subjective and depend on judges’ intuitions about the proper role of the juries that are supposed to supervise them. As we have previously explained, “[w]hether the Sixth Amendment incorporates [such a] manipulable standard rather than Apprendi‘s bright-line rule depends on the plausibility of the claim that the Framers would have left definition of the scope of jury power up to judges’ intuitive sense of how far is too far.” Blakely v. Washington, 542 U.S. 296, 308, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004). And we continue to think that claim is “not plausible at all, because the very reason the Framers put a jury-trial guarantee in the Constitution” was to ensure the jury trial right would limit the power of judges and not be ground down to nothing through a balancing of interests by judges themselves. Ibid.

10. But even on this point, the plurality hedges, saying that “historic parole and probation revocation procedures … have usually been understood to comport with the Fifth and Sixth Amendments.” Ante, at 2381-2382 (emphasis added).

11. The plurality casts this argument as “echo[ing] an age-old criticism: Jury trials are inconvenient for the government.” Ante, at 2384. Not at all. My only point is to say that if a questionable interpretation of the Sixth Amendment, see infra, at 2391-2399, would potentially lead to absurd results, that is an additional reason to suspect that something has gone awry.

12. See O’Hara, Parole, 79 Geo. L. J. 1162, 1164-1165 (1991).

13. In respondent’s case that was life. See § 3583(k). Anything approaching that maximum would have been very harsh, but the judge in respondent’s case did not impose such a term, and there are statutory restraints on the imposition of excessive additional terms. In determining the additional period to be ordered as a result of a supervised-release violation, a judge is required to take into account almost all of the factors that must be considered at sentencing. See § 3583(e). The Sentencing Guidelines provide recommended terms for particular violations. See USSG ch. 1, pt. B; id., ch. 7. And the additional terms imposed in such cases are subject to review on appeal. See, e.g., United States v. Wheeler, 814 F.3d 856 (C.A.7 2016); United States v. Cordova, 461 F.3d 1184 (C.A.10 2006). If the Constitution restricts the length of additional imprisonment that may be imposed based on a violation of supervised release, the relevant provision is the Eighth Amendment, not the Sixth. Cf. Lewis v. United States, 518 U.S. 322, 116 S.Ct. 2163, 135 L.Ed.2d 590 (1996).

14. See, e.g., United States v. Carlton, 442 F.3d 802, 806-810 (C.A.2 2006); United States v. Dees, 467 F.3d 847, 854-855 (C.A.3 2006); United States v. Ward, 770 F.3d 1090, 1096-1099 (C.A.4 2014); United States v. Hinson, 429 F.3d 114, 117-119 (C.A.5 2005); United States v. McIntosh, 630 F.3d 699, 703 (C.A.7 2011); United States v. Gavilanes-Ocaranza, 772 F.3d 624, 628-629 (C.A.9 2014); Cordova, 461 F.3d at 1186.

15. See, e.g., Gavilanes-Ocaranza, 772 F.3d at 628; United States v. House, 501 F.3d 928, 931 (C.A.8 2007).

16. See, e.g., United States v. Rondeau, 430 F.3d 44, 47-48 (C.A.1 2005); United States v. Kelley, 446 F.3d 688, 690-692 (C.A.7 2006); United States v. Hall, 419 F.3d 980, 985-986 (C.A.9 2005); United States v. Ray, 530 F.3d 666, 667-668 (C.A.8 2008); United States v. Ojudun, 915 F.3d 875, 888 (C.A.2 2019).

17. See, e.g., United States v. Boultinghouse, 784 F.3d 1163, 1171 (C.A.7 2015); United States v. Owen, 854 F.3d 536, 541 (C.A.8 2017); United States v. Spangle, 626 F.3d 488, 494 (C.A.9 2010).

18. The plurality offers only a few tepid responses. First, the plurality appears to concede that a jury trial is unnecessary where penalties for postjudgment conduct are not too harsh. Ante, at 2382-2383. I suspect that the prisoners who endured corporal punishment would have challenged the plurality’s suggestion that their punishment was not that harsh. But in any event, a too-harsh standard —something that would appear to be more at home in an Eighth Amendment analysis —is hardly a principled way of determining whether a jury trial is constitutionally required. Second, the plurality suggests that my reasoning amounts to an extension of Turner v. Safley, 482 U.S. 78, 107 S.Ct. 2254, 96 L.Ed.2d 64 (1987), because Turner addresses only the relaxation of a prisoner’s constitutional rights and the Court has “never extended it to the jury rights of persons out in the world who retain the core attributes of liberty.” Ante, at 2383. But a convicted criminal on supervised release does not “retain the core attributes of liberty,” ibid., and Turner is not implicated here because, as I have shown, the Sixth Amendment does not apply and thus the criminal has no jury trial right that Turner might relax. And once again, more notable than the plurality’s lack of real answers is its inability to point to any affirmative evidence that the jury ever played a historical role in the administration of previously imposed sentences.