Supervised Release Revocations After Haymond: Lingering Questions

Supervised release revocation is supposed to be different from a new criminal prosecution. A defendant on supervised release does not receive a full jury trial. The government generally proves violations by a preponderance of the evidence, not beyond a reasonable doubt. The judge decides whether a violation occurred and what sanction follows.

But United States v. Haymond complicated that framework.

In Haymond, the Supreme Court addressed 18 U.S.C. § 3583(k), a provision that required a mandatory minimum prison term when certain defendants on supervised release committed specified offenses. The Court held that applying § 3583(k)'s mandatory minimum based on judge-found facts violated the Fifth and Sixth Amendments. The constitutional problem was that a judge, using a preponderance standard, found facts that triggered a mandatory five-year prison term rather than requiring a jury to find those facts beyond a reasonable doubt.

The decision did not eliminate supervised release revocation. It did not hold that every revocation requires a jury. And it did not broadly import Apprendi and Alleyne into every supervised-release proceeding. But it raised a difficult constitutional question that remains important today: when does revocation stop looking like punishment for the original offense and start looking like punishment for a new crime?

That question remains unsettled in important ways.

Key Takeaways

• Haymond was important, but narrow. The controlling opinion is generally treated as Justice Breyer's concurrence, not Justice Gorsuch's broader plurality.
• Ordinary revocations under § 3583(e)(3) usually remain outside the Sixth Amendment jury-trial rule. Courts have mostly refused to extend Haymond to mandatory revocation under § 3583(g).
• Robinson adds another important limit: the exclusionary rule generally does not apply in supervised-release revocation hearings.
• Carpenter reinforces that ordinary § 3583(e)(3) revocations are not "criminal prosecutions" requiring jury trials.
• Doka remains one of the leading post-Haymond cases preserving § 3583(e)(3)'s constitutionality.
• The strongest future challenges are likely to involve revocation schemes that impose mandatory imprisonment based on judge-found new criminal conduct.

I. The Statutory Framework: What § 3583 Does

Section 3583 governs federal supervised release after imprisonment. It authorizes courts to impose a term of supervised release after a felony or misdemeanor conviction and sets maximum supervised-release terms depending on the class of offense. It also identifies mandatory conditions, such as not committing another crime, not unlawfully possessing controlled substances, complying with drug testing, making restitution when required, cooperating in DNA collection, and complying with sex-offender registration requirements when applicable.

The statute also gives courts authority to modify, reduce, enlarge, terminate, or revoke supervised release. The most important revocation provisions for post-Haymond litigation are § 3583(e)(3), § 3583(g), and § 3583(k).

Section 3583(e)(3) is the ordinary revocation provision. It allows a court to revoke supervised release and impose imprisonment if the court finds, by a preponderance of the evidence, that the defendant violated a condition of supervised release. Section 3583(g), by contrast, requires revocation and imprisonment for certain violations, including possession of a controlled substance, possession of a firearm, refusal to comply with drug testing, and repeated positive drug tests. But § 3583(g) still caps imprisonment by reference to § 3583(e)(3).

That statutory difference matters. Most lower courts have treated § 3583(e)(3) and § 3583(g) as different from § 3583(k), the provision invalidated in Haymond.

II. What Haymond Actually Held

The defendant in Haymond was on supervised release after a federal child-pornography conviction. While on supervision, he was alleged to have possessed additional child pornography. The district judge found the violation by a preponderance of the evidence. Under ordinary revocation rules, the judge could have imposed a prison term within the limits of § 3583(e)(3). But because § 3583(k) applied, the judge was required to impose at least five years in prison.

That mandatory minimum was the constitutional problem.

The Supreme Court concluded that § 3583(k), as applied, crossed the constitutional line because it imposed a new mandatory minimum based on judge-found facts. The plurality reasoned that this resembled the kind of factfinding forbidden by Apprendi and Alleyne: facts that increase a defendant's legally required punishment generally must be found by a jury beyond a reasonable doubt.

But Haymond was fractured. Justice Gorsuch wrote a broad plurality opinion. Justice Breyer supplied the fifth vote, but his concurrence was narrower. Because of that, lower courts generally treat Justice Breyer's concurrence as the controlling opinion. The Seventh Circuit in Carpenter collected circuit authority recognizing Justice Breyer's concurrence as the controlling rule.

Under Justice Breyer's approach, § 3583(k) was constitutionally problematic because it had three distinctive features: (1) it applied only when a defendant committed a discrete set of federal criminal offenses; (2) it removed judicial discretion; and (3) it imposed a mandatory minimum term of imprisonment. Those features made § 3583(k) look less like ordinary revocation and more like punishment for a new offense.

III. Why Lower Courts Have Read Haymond Narrowly

Since Haymond, most federal courts have limited the decision to its unusual statutory context. The dominant view is that ordinary supervised-release revocation remains governed by § 3583(e)(3), where the judge may revoke release and impose imprisonment within statutory caps tied to the original offense.

The Seventh Circuit's 2024 decision in United States v. Carpenter, 104 F.4th 655 (7th Cir. 2024) is a useful example. Carpenter argued that his supervised-release revocation under § 3583(e)(3) triggered the Sixth Amendment jury-trial guarantee. The Seventh Circuit rejected that argument, explaining that supervised-release revocations are not "criminal prosecutions" within the meaning of the Sixth Amendment.

The court emphasized that revocation proceedings address conditional liberty and the administration of an existing sentence. They do not impose punishment for a new offense in the same way a new criminal prosecution does. That distinction allowed the court to affirm the denial of Carpenter's motion for a jury trial.

The Second Circuit reached a similar result in United States v. Doka, 955 F.3d 290 (2d Cir. 2020). There, the district court revoked Doka's supervised release based on findings by a preponderance of the evidence that he committed second-degree assault, possessed controlled substances with intent to distribute, and used controlled substances. The Second Circuit upheld the revocation and sentence, distinguishing § 3583(e)(3) from § 3583(k) because § 3583(e)(3) does not impose a mandatory minimum, does not apply only to a discrete set of federal offenses, and does not eliminate judicial discretion. Other courts have relied on Doka for the proposition that ordinary revocation under § 3583(e)(3) remains constitutional after Haymond.

The result is a clear post-Haymond pattern: § 3583(k) is constitutionally suspect because of its mandatory-minimum structure, but § 3583(e)(3) generally remains valid.

IV. Supervised Release as Part of the Original Sentence

The traditional theory is that supervised release revocation punishes a defendant's breach of trust, not the new conduct itself. Under that theory, revocation is part of the original sentence. The defendant has already been convicted, and the court is administering the consequences of that original judgment.

That theory explains why revocation proceedings normally do not include the full protections of a criminal trial. A judge may find a violation by a preponderance of the evidence. The Federal Rules of Evidence do not apply with full force. The defendant receives due-process protections, but not the full trial package.

Haymond did not reject that traditional framework. Instead, it identified a situation where the statutory structure looked different. Section 3583(k) did not merely authorize a judge to respond to breach of trust. It required a severe mandatory prison term based on judge-found conduct that itself resembled a new federal offense.

That is why the case remains important. It shows that Congress cannot avoid the jury-trial right simply by labeling a punishment "revocation" if the statute operates like punishment for a new crime.

V. The Lingering Problem With Mandatory Revocation Under § 3583(g)

The hardest post-Haymond question involves mandatory revocation under § 3583(g).

Section 3583(g) requires revocation and imprisonment for certain violations, including possession of a controlled substance, possession of a firearm, refusal to comply with drug testing, and repeated positive drug tests. But unlike § 3583(k), it does not impose a new five-year mandatory minimum term. Instead, it requires imprisonment within the maximum term authorized by § 3583(e)(3).

Defendants have argued that § 3583(g), like § 3583(k), imposes imprisonment based on judge-found facts without a jury. So far, courts have mostly rejected that argument.

The Sixth Circuit's decision in United States v. Robinson, 63 F.4th 530 (6th Cir. 2023) is central. Robinson was on supervised release after serving a reduced sentence for firearm and drug offenses. During a traffic stop, police found drugs and a firearm. The government conceded a Fourth Amendment violation but argued that the exclusionary rule did not apply in the revocation proceeding. The district court revoked supervised release and imposed a 28-month prison sentence.

On appeal, the Sixth Circuit held that the exclusionary rule does not apply to supervised-release revocation hearings. The court reasoned that extending the exclusionary rule into that context would provide minimal additional deterrence while imposing significant costs on the rehabilitative and supervisory purposes of supervised release. The court also rejected Robinson's jury-trial challenge to § 3583(g), distinguishing § 3583(g) from § 3583(k).

That holding matters for two reasons. First, it confirms that many constitutional protections available in criminal trials do not automatically carry over into revocation proceedings. Second, it reinforces the post-Haymond distinction between § 3583(k) and § 3583(g): § 3583(g) requires revocation for certain violations, but it does not impose the same new mandatory minimum that made § 3583(k) unconstitutional.

VI. The Exclusionary Rule After Robinson

Robinson also adds an important Fourth Amendment dimension to post-Haymond revocation practice.

The exclusionary rule generally prevents the government from using unlawfully obtained evidence in a criminal trial. But revocation hearings are different. Relying on the Supreme Court's approach in Pennsylvania Board of Probation and Parole v. Scott, 524 U.S. 357 (1998), the Sixth Circuit concluded that the deterrent benefits of exclusion are weaker in the revocation context, while the costs are substantial.

For defendants, this is a serious practical limitation. Evidence that might be suppressed in a criminal prosecution may still be used to revoke supervised release. That means a defendant may win or have a strong suppression argument in the criminal case but still face revocation based on the same conduct.

This is one of the overlooked consequences of the current supervised-release framework. Revocation proceedings can produce real imprisonment while using fewer procedural protections than a criminal trial.

VII. Standard of Proof After Haymond

Haymond also prompted renewed scrutiny of the preponderance standard used in revocation proceedings.

Ordinarily, a supervised-release violation need only be proven by a preponderance of the evidence. That remains the rule under § 3583(e)(3). Courts have continued to uphold that standard after Haymond, emphasizing that ordinary revocation does not impose a new mandatory minimum based on a new criminal offense.

The key distinction is not merely the standard of proof. The key distinction is what the judge-found fact does. If the fact only permits discretionary revocation within the statutory limits of the original sentence, courts usually uphold the process. But if the fact triggers a new mandatory punishment that resembles a penalty for a new offense, Haymond becomes more relevant.

That is why defense counsel should not argue Haymond too broadly. The better argument is targeted: the court should examine whether the revocation statute changes the legally required punishment based on judge-found new criminal conduct.

VIII. New Criminal Conduct and the Breach-of-Trust Fiction

A major unresolved issue is the treatment of new criminal conduct.

Many revocations are based on conduct that could also be prosecuted as a new offense: firearm possession, drug possession, assault, fraud, false statements, or other criminal acts. Courts often say the revocation punishment is for breach of trust, not the new crime. But when the new criminal conduct drives the revocation sentence, the distinction can become thin.

That does not mean every new-crime violation requires a jury. Current law says otherwise. Carpenter rejected a jury-trial claim even where some alleged violations involved serious criminal conduct, including arson-related allegations. The Seventh Circuit still treated the proceeding as a revocation of conditional liberty rather than a new criminal prosecution.

But the tension remains. If the revocation system can impose imprisonment based on conduct that looks like a new crime, while avoiding indictment, jury trial, proof beyond a reasonable doubt, and sometimes suppression remedies, then supervised release occupies a constitutionally sensitive space.

That is the unresolved legacy of Haymond.

IX. Practical Impact for Defense Counsel

For defense counsel, Haymond arguments should be precise.

A broad argument that every revocation requires a jury is unlikely to succeed under current law. Courts have repeatedly rejected that position. The stronger argument asks whether the specific statute or revocation theory has the same features that troubled Justice Breyer in Haymond.

Counsel should ask:

1. Does the statute apply only to a discrete set of criminal offenses?
2. Does it remove judicial discretion?
3. Does it impose a mandatory minimum or otherwise increase the legally required punishment?
4. Does the revocation sentence operate more like punishment for new conduct than sanctioning breach of trust?
5. Was the alleged violation also charged or chargeable as a separate crime?
6. Was the judge required to impose imprisonment based on facts not found by a jury beyond a reasonable doubt?
7. Is the government relying on evidence that would be suppressible in a criminal case?
8. Does the revocation theory effectively allow the government to punish a new offense without the procedural protections of a criminal prosecution?

These questions should be preserved clearly in the district court. Many appellate challenges fail because the issue is reviewed only for plain error or because circuit precedent already forecloses relief. Preservation matters because the law is still developing, and future Supreme Court review would likely require a clean record.

X. What Haymond Does Not Do

It is equally important to understand what Haymond does not do.

• It does not invalidate § 3583(e)(3).
• It does not require jury trials for ordinary supervised-release revocations.
• It does not prohibit judges from finding violations by a preponderance of the evidence in ordinary revocation proceedings.
• It does not automatically invalidate § 3583(g).
• It does not apply the exclusionary rule to all revocation hearings.
• It does not hold that every revocation based on new criminal conduct is a new prosecution.

That is why lower courts have been able to contain Haymond. The decision matters, but mainly where the revocation statute creates a mandatory punishment that resembles punishment for a new offense.

XI. The Continuing Tension

The deeper tension is conceptual.

Federal courts often describe supervised release revocation as punishment for "breach of trust." That framing helps explain why revocation does not require the full protections of a criminal trial. But some revocation statutes and revocation practices focus heavily on new criminal conduct. When the new conduct triggers mandatory imprisonment, the breach-of-trust theory becomes harder to maintain.

Robinson, Carpenter, and Doka show how lower courts are managing that tension. Robinson limits both the exclusionary rule and the jury-trial argument under § 3583(g). Carpenter rejects the claim that § 3583(e)(3) revocations are criminal prosecutions requiring jury trials. Doka confirms that § 3583(e)(3) remains constitutional because it lacks the three features that made § 3583(k) unconstitutional.

Together, those cases show that Haymond has been contained — but not erased.

Conclusion

Nearly seven years after Haymond, lower courts have largely confined the decision to § 3583(k). Ordinary revocations under § 3583(e)(3) remain judge-decided proceedings. Mandatory revocation under § 3583(g) has mostly survived constitutional challenge. Justice Breyer's concurrence continues to define the narrow controlling rule.

But Haymond still matters.

It gives defendants a constitutional framework for challenging revocation schemes that impose serious mandatory imprisonment based on judge-found new criminal conduct. It also reminds courts that supervised release is not a constitutional blank check. When revocation begins to look like a new prosecution, the Fifth and Sixth Amendments remain in the background — and sometimes they move to the center.

References

• United States v. Haymond, 588 U.S. 634, 139 S. Ct. 2369 (2019).
• 18 U.S.C. § 3583.
• United States v. Robinson, 63 F.4th 530 (6th Cir. 2023).
• United States v. Carpenter, 104 F.4th 655 (7th Cir. 2024).
• United States v. Doka, 955 F.3d 290 (2d Cir. 2020).
• United States v. Garner, 969 F.3d 550 (5th Cir. 2020).
• United States v. Oliver, 41 F.4th 1093 (9th Cir. 2022).
• United States v. Austin, 501 F. Supp. 3d 266 (S.D.N.Y. 2020).
• Pennsylvania Board of Probation and Parole v. Scott, 524 U.S. 357 (1998).
• Johnson v. United States, 529 U.S. 694 (2000).