February 20, 2026
Ellingburg v. United States, 607 U.S. ____ (2025) (Kavanaugh, J.)
Response by Lula Hagos
Geo. Wash. L. Rev. On the Docket (Oct. Term 2025)
Slip Opinion | SCOTUSblog
Note: A version of this Response originally appeared on the Sentencing Matters Substack.
Ellingburg and Restitution’s Constitutional Reckoning
For years, criminal restitution has occupied a doctrinal gray zone. On the one hand, it is imposed in criminal court against defendants, embedded in a criminal judgment, enforced by the government, and backed by the coercive machinery of the criminal legal system. On the other hand, restitution is often described as something other than punishment. The dominant narrative casts restitution as a victim-centered, compensatory remedy akin to civil relief rather than a penal sanction. That narrative has helped restitution expand both quietly and rapidly while attracting less sustained scrutiny than other monetary sanctions that courts readily acknowledge are punitive.
Over time, that insulation has become increasingly hard to defend. Restitution is frequently the most consequential financial component of a criminal sentence and, because of the government’s extensive power to enforce it, it can also be the most durable. In its Fiscal Year 2024 Annual Report, the U.S. Sentencing Commission reported that courts ordered approximately $13.5 billion in restitution that year––the highest annual amount in more than two decades.1 At the same time, restitution determinations often proceed with relatively thin procedural protections for a sanction that can shadow a person for decades.
Against that backdrop, on January 20, 2026, the Supreme Court issued a seminal opinion in Ellingburg v. United States.2 In a unanimous decision, the Court held that restitution under the Mandatory Victims Restitution Act3 (“MVRA”) is “plainly criminal punishment” for purposes of the Constitution’s Ex Post Facto Clause,4 reversing a decision by the U.S. Court of Appeals for the Eighth Circuit that MVRA restitution was a civil remedy and not subject to ex post facto limits.5 The holding is narrow in what it decides, but it brings long-needed clarity to restitution’s constitutional status and provides a new baseline for the disputes that follow.
At the time of Holsey Ellingburg Jr.’s offense, federal law capped restitution enforcement at twenty years from entry of judgment. Under that regime, his restitution obligation would have ended in 2016. Congress later enacted the MVRA and extended enforcement to twenty years after release from imprisonment. Applied retroactively, that change extended Ellingburg’s restitution exposure to 2042. By that point, the amount he owed had also grown substantially through interest and collection over time,6 increasing from roughly $7,567 to more than $13,000.
Those facts presented a straightforward ex post facto question. The Constitution bars the government from retroactively increasing punishment by changing the legal consequences of an offense after the fact. Whether this bar applies depends on a threshold question about restitution’s nature. If restitution is punishment, retroactively extending its enforcement increases the penalty and is impermissible. If restitution is a civil remedy, retroactive enforcement is constitutionally permissible.
Writing for the majority, Justice Kavanaugh answered that question unequivocally. The Court held that restitution under the MVRA is “plainly criminal punishment” for purposes of the Ex Post Facto Clause.7 The Court reasoned that the statute’s text, design, and operation make restitution part of the criminal sentence, rather than a civil remedy.
The Court emphasized that the MVRA labels restitution a “penalty” tied to a criminal offense and integrates it into the sentencing process after conviction, alongside incarceration and fines.8 The Court also pointed to the MVRA’s placement in Title 18’s sentencing framework and its procedural integration with other criminal penalties. Just as importantly, restitution under the MVRA is pursued and enforced by the government, not by victims in the way civil plaintiffs would control a compensatory claim. The MVRA’s enforcement mechanisms, the Court explained, also carry a host of coercive consequences that plainly resemble punishment. Taken together, the statute’s features make it “abundantly clear” that MVRA restitution is punishment.9
The Court also leaned on its own prior descriptions of MVRA restitution as serving punitive aims, in cases like Paroline v. United States10 and Pasquantino v. United States.11 Those cases did not squarely address the ex post facto question but repeatedly characterized MVRA restitution as more than mere compensation.12 Ellingburg now makes the conclusion explicit.
Ellingburg thus resolves a deeply entrenched circuit split over restitution’s nature in criminal court. But perhaps its most important contribution is that it does away with a legal fiction—that restitution is a civil remedy free of constitutional scrutiny and the Constitution’s limits on criminal penalties—that has shaped restitution doctrine for years. In Ellingburg, the Court finally rejected that fiction, aligning doctrine with restitution’s institutional reality under the MVRA.
Justice Thomas, joined by Justice Gorsuch, concurred but wrote separately to emphasize an originalist foundation for ex post facto doctrine.13 Relying on the Court’s 1798 seriatim opinions in Calder v. Bull,14 he argued that modern jurisprudence has placed too much weight on civil and criminal labels. In his view, the relevant question instead turns on whether a law imposes “a coercive penalty for a public wrong,” regardless of whether Congress brands it as civil or criminal.15 A “public wrong” is an injury to the sovereign in its sovereign capacity, brought by the United States rather than a private party.16 Applied here, Justice Thomas reasoned that because Ellingburg was prosecuted by the United States, rather than the bank that was robbed, the ex post facto constraint applies. The concurrence signals a willingness, at least by two Justices, to revise the civil-criminal boundary more broadly for ex post facto purposes.
Ellingburg’s holding has significant and immediate implications for restitution beyond the ex post facto doctrine. Perhaps the most consequential issue concerns the Sixth Amendment jury trial right. Under Apprendi v. New Jersey,17 any fact other than a prior conviction that increases the maximum authorized punishment must be found by a jury beyond a reasonable doubt. For years, defendants raising Sixth Amendment challenges to restitution have been turned away on two overlapping grounds: 1) restitution is compensatory rather than punitive, and 2) the MVRA does not specify a numeric “statutory maximum” because it requires the “full amount” of a victim’s losses.18
Ellingburg largely eliminates the force of the first rationale by holding that restitution is “plainly” criminal punishment. The remaining obstacle is the “no statutory maximum” argument, but the doctrinal tension is now harder to avoid because restitution’s amount is determined by contested findings about loss and causation that can dramatically increase a defendant’s punishment exposure.
If Apprendi were applied to restitution, the procedural implications would be significant. The restitution amount would need to be charged in the indictment, submitted to the jury, and proven beyond a reasonable doubt by the government, unless a defendant admitted the amount. That process would align restitution more closely with fines and forfeiture—monetary sanctions the Court has treated as punishment in other constitutional contexts.
The Sixth Amendment issue is actively percolating in the courts.19 Ellingburg’s classification of restitution as punishment gives these cases a clearer analytical starting point, even if courts continue to wrestle with the “no statutory maximum” question.
Another potential downstream consequence of the Ellingburg decision involves the Eighth Amendment’s Excessive Fines Clause, which applies to financial penalties imposed as punishment. Restitution has often avoided excessive fines analysis because it has been framed as victim compensation rather than punishment. After Ellingburg, categorical insulation becomes harder to maintain.
Courts may still resist treating ordinary restitution orders as excessive where they track concrete, easily quantifiable losses. But harder cases, such as those involving extraordinary amounts, tenuous causation determinations, or speculative valuations, present a viable opportunity for the defense to make arguments about proportionality. The Court’s framework in United States v. Bajakajian20 (involving forfeiture) remains the touchstone for evaluating punitive monetary sanctions under the Excessive Fines Clause, and Ellingburg now strengthens the predicate for bringing restitution challenges within that conversation.
Ellingburg does not revolutionize restitution overnight. But it marks a pivotal shift in restitution jurisprudence. The Court’s reclassification of restitution as punishment collapses doctrine that allowed it to be imposed and enforced as a criminal sanction while being treated as civil for constitutional purposes.
The doctrinal consequences will take time to develop. Still, after Ellingburg, courts evaluating restitution claims begin from a dramatically different baseline, one that is closer to restitution’s institutional and practical reality and more attentive to the constitutional implications of punitive monetary sanctions.
Lula Hagos is an Associate Professor of Law and the Director of the Criminal Defense and Justice Clinic, which she launched in Fall 2021. Professor Hagos’s scholarly interests explore issues that lie at the intersection of criminal law and procedure, race, and access to justice.
Recommended Citation
Lula Hagos, Ellingburg and Restitution’s Constitutional Reckoning, Geo. Wash. L. Rev. On the Docket (Feb. 20, 2026), https://www.gwlr.org/ellingburg-restitution-reckoning.
References
[1] U.S. Sent’g Comm’n, 2024 Annual Report 16 (2024), https://www.ussc.gov/sites/default/files/pdf/research-and-publications/annual-reports-and-sourcebooks/2024/2024-Annual-Report.pdf [https://perma.cc/5K6M-Q773].
[2] No. 24–482, 2026 WL 135982 (U.S. Jan. 20, 2026).
[3] 18 U.S.C. § 3663A.
[4] Ellingburg, 2026 WL 135982, at *2.
[5] 113 F.4th 839, 842 (8th Cir. 2024) (per curiam).
[6] See 18 U.S.C. § 3612.
[7] Ellingburg, 2026 WL 135982, at *2.
[8] See id.
[9] Id. at *3.
[10] 572 U.S. 434 (2014).
[11] 544 U.S. 349 (2005).
[12] See, e.g., Paroline, 572 U.S. at 456; Pasquantino, 544 U.S. at 365.
[13] Ellingburg, 2026 WL 135982, at *4 (Thomas, J., concurring).
[14] 3 U.S. (3 Dall.) 386 (1798).
[15] Ellingburg, 2026 WL 135982, at *4 (Thomas, J., concurring).
[16] See id. at *8–9.
[17] 530 U.S. 466 (2000).
[18] 18 U.S.C. § 3664(f)(1)(A).
[19] See, e.g., Petition for a Writ of Certiorari at i, Stroud v. United States, No. 22-11208 (U.S. June 18, 2025), cert. denied, 2026 WL 79762 (U.S. Jan. 12, 2026).
[20] 524 U.S. 321 (1998).
