UNITED STATES v. MCLAURIN (2026)

ORDER

A member of the Court requested a poll for en banc review, but the poll failed to produce a majority of judges in regular active service and not disqualified who voted in favor of rehearing en banc. Chief Judge Diaz and Judges Wilkinson, King, Gregory, Agee, Wynn, Thacker, Harris, Richardson, Quattlebaum, Heytens, Benjamin, and Berner voted to deny rehearing en banc. Judges Niemeyer and Rushing voted to grant rehearing en banc.

Judge Wynn wrote a statement respecting the denial of rehearing en banc. Judge Richardson, with whom Chief Judge Diaz, and Judges Wilkinson, Niemeyer, King, Agee, Quattlebaum, and Rushing joined, wrote a statement respecting the denial of rehearing en banc.

Entered at the direction of Judge Gregory.

Judge Richardson's well-written opinion attached to this Order does not oppose the denial of en banc review—indeed, it could not. He, along with nearly all the judges who joined his separate writing, voted against rehearing the case en banc. Yet they collectively identify a serious and compelling concern, one that lies well within this Court's power to correct by granting such review. That inconsistency—articulating a forceful basis for en banc review while simultaneously voting to deny it—is one I find difficult to reconcile.

A doctrinal error has ensnared our district courts. Because of it, we routinely vacate substantively valid criminal judgments. The culprit is our misunderstanding of the right of a defendant to be present at sentencing. The right is indisputable. But our Court has confused its foundation. This error can—and should—be fixed with a modest doctrinal change. Fixing our precedent would not only correct the law but would return some modicum of order to sentencing and promote judicial economy.¹

The error is our conclusion that the oral pronouncement is the operative sentence. Explaining why that's wrong requires some background.

Rule 43 of the Federal Rules of Criminal Procedure states a simple rule: “the defendant must be present at ․ sentencing.” This rule protects a defendant's procedural right—rooted in the Due Process Clause—to be present at certain proceedings. See Lewis v. United States, 146 U.S. 370, 372, 13 S.Ct. 136, 36 L.Ed. 1011 (1892); United States v. Gagnon, 470 U.S. 522, 526, 105 S.Ct. 1482, 84 L.Ed.2d 486 (1985). Our Court has reasonably understood this procedural right to require an oral sentence pronouncement in the defendant's presence.

But this raises the question: What happens when a district court later imposes a different written sentence from the one it orally pronounced? After all—like in other areas of the law—a district court's spoken word is not its last. In the context of criminal sentencing, district courts are required to finally impose a defendant's sentence in a written “judgment of conviction.” See Fed. R. Crim. P. 32(k)(1). The right to be present for the oral pronouncement of one's sentence would ring hollow if a judge could change the sentence in the final judgment. That's why our Court has long recognized that a district court errs when the written judgment includes a punishment that is different than what was orally pronounced. See, e.g., Rakes v. United States, 309 F.2d 686, 687–88 (4th Cir. 1962).

But it is important to consider why a discrepancy between a written judgment and an oral pronouncement matters. It's because the written judgment is the final word on a defendant's sentence. Put differently, the written sentencing judgment—not the oral pronouncement—binds. As I explain, this understanding of the primacy of the written judgment finds support in relevant principles of law and history.

Unfortunately, our Court has long suggested the opposite: that the oral pronouncement of a defendant's sentence “controls” over the written judgment. See, e.g., United States v. Morse, 344 F.2d 27, 29 n.1 (4th Cir. 1965). Other courts of appeals have suggested the same. See United States v. Daddino, 5 F.3d 262, 266 n.5 (7th Cir. 1993) (collecting cases). We recently revived this erroneous idea in United States v. Rogers. 961 F.3d 291, 296 (4th Cir. 2020) (“[I]f a conflict arises between the orally pronounced sentence and the written judgment, then the oral sentence controls.”). And the next year in United States v. Singletary, we began to read the idea for all its worth. See 984 F.3d 341, 345 (4th Cir. 2021). In Singletary, our Court counterintuitively held that conditions of supervised release “appearing for the first time in a written judgment in fact have not been ‘imposed’ on the defendant.” Id. We referred to the conditions included in the written judgment as “nullities.” Id. at 344. And, as a remedy, we vacated the defendant's entire sentence and required the district court to resentence the defendant. Id. at 346.

The approach we took in Singletary has led to a flood of appeals by defendants seeking to vacate their otherwise-valid sentences because of minor discrepancies between the oral pronouncement of a sentence and the written judgment. In case after case, we have slowly expanded the reach of our so-called Rogers-Singletary doctrine. That expansion continues in this case, in which the panel majority allowed a defendant to challenge part of his sentence more than a decade after it became final. For more than a decade, everyone—including the courts, prosecutors, probation office, and even the defendant—understood the defendant's sentence to include those supervised release conditions listed in the written judgment. No longer. That result is troubling and presents concerns about the finality of a whole swath of otherwise-valid judgments.² But, given the conflicting state of the doctrine, the majority opinion in this case is not without support—especially given the idea that “discretionary conditions appearing for the first time in a written judgment in fact have not been ‘imposed’ on the defendant.” See Singletary, 984 F.3d at 345 (emphasis omitted). If we mean what we say (and I hope we don't), the result in this case might follow since a defendant cannot be convicted for a violation of a condition of supervised release that was never genuinely part of his sentence. Our error continues to spread.

To fix our Rogers-Singletary mess, our Court must recognize the centrality of the written judgment. This minor conceptual shift offers meaningful theoretical and practical benefits.

Start with the basics. A written judgment is just that: a judgment. See Judgment, Black's Law Dictionary (12th ed. 2024) (“A court or other tribunal's final determination of the rights and obligations of the parties in a case.”). The power to enter judgments is core to Article III's judicial power. In every other context that comes to mind, we treat a court's written judgment as the court's last word.³ The Federal Rules of Criminal Procedure also recognize the importance of the written judgment: Rule 32(k) requires a district court to list a defendant's sentence in “the judgment of conviction.” And it is the written judgment that officials rely on—in the courts, the prison system, and the probation system—to understand and enforce a defendant's sentence. That common-sense reality confirms that the written judgment is the only sentence truly “imposed.” So it is the only sentence that “controls” in every relevant sense.⁴

Indeed, the Supreme Court has emphasized the centrality of the written judgment: “Courts reduce their opinions and verdicts to judgments precisely to define the rights and liabilities of the parties․ This Court, like all federal appellate courts, does not review lower courts' opinions, but their judgments.” Jennings v. Stephens, 574 U.S. 271, 277, 135 S.Ct. 793, 190 L.Ed.2d 662 (2015). That is why the Supreme Court made clear long ago that “[t]he only sentence known to the law is the sentence or judgment entered upon the records of the court.” Hill v. United States ex rel. Wampler, 298 U.S. 460, 464, 56 S.Ct. 760, 80 L.Ed. 1283 (1936).⁵

Given this background, recognizing the centrality of the written judgment is most consistent with background principles of law and common sense. But the relevant history supports this understanding, too. “In England, the defendant's sentence after oral announcement was entered in the record along with the fact that the defendant was present for the oral announcement.” United States v. Griffin, No. 21-50294, 2022 WL 17175592, at *6 (5th Cir. Nov. 23, 2022) (Oldham, J., dissenting). Judges understood the written record to be binding, as shown by the care they took to change the written record in the case of any later changes to the sentence. Id. “The same was true at common law in the United States.” Id. Both legal principle and historical practice, then, point the same direction.

To recap, the key to reforming the Rogers-Singletary line of cases is to recognize the centrality of the written judgment. To be sure, this change will not (and should not) affect the core of a Rogers-Singletary claim. A defendant can still successfully challenge his sentence if a material portion of the sentence was not orally pronounced. But an error does not change whether an unannounced portion of a sentence was imposed; it only changes whether that aspect of the sentence was correctly imposed (i.e., imposed with the defendant present).

Clarifying this theoretical point would clear the way to solve what I see as a substantial remedial problem, too. After all, a defendant's right to have his sentence orally pronounced is a procedural right.⁶ And the proper remedy for the denial of a procedural right is a procedural remedy. We ought not demand that each defendant be fully resentenced. Rather, the remedy should be more limited: A district court should have the option to either conform the written judgment to the oral pronouncement or conduct a new sentencing hearing.⁷ See Kemp, 88 F.4th at 551–53 (Quattlebaum, J., concurring). Either way, this remedy would address the concerns presented by Rule 43 and the Due Process Clause by ensuring that the ultimate sentence imposed by the court (in its written judgment) was the same as that orally pronounced in the presence of the defendant.

In fact, our current approach—vacating the entire sentence—is inconsistent with our stated view that the oral pronouncement controls. If the oral pronouncement of a defendant's sentence genuinely controlled a defendant's sentence, then the proper remedy would be to require district courts to amend the written judgment to match the oral pronouncement. Some courts use this remedy. See, e.g., Griffin, 2022 WL 17175592, at *5 (Oldham, J., dissenting) (describing the Fifth Circuit's occasional use of this approach). Indeed, we used to take this approach but have since abandoned it. Compare Morse, 344 F.2d at 30–31, with Singletary, 984 F.3d at 346. This internal inconsistency is another reason to start recognizing the centrality of the written judgment—and the procedural nature of the Rule 43 right.

Recognizing the centrality of the written judgment is legally correct—which is reason enough to do so. But there are at least two practical benefits, too: increased finality and judicial economy. These benefits warrant taking the extraordinary step of going en banc for a course correction.

First, finality. In a series of decisions—including this case—our Court has allowed defendants to indirectly challenge their initial sentence well after the appellate process has concluded. That outcome deeply undermines the values of predictability and finality that are central to the rule of law. Like other procedural rights, a challenge to a Rule 43 violation can be forfeited by a defendant. So once a defendant's sentence becomes final, there is a written judgment that controls the defendant's sentence moving forward. This approach would obviate the concerns presented in this case—where a defendant avoids the consequences of his sentence years after it is imposed.

Second, judicial economy. Our district courts face persistent resource constraints. Our current Rogers-Singletary remedy—requiring that the most minor scrivener's error (or stray remark at sentencing) requires a full resentencing—burdens our district courts in exchange for no perceptible benefits. A more carefully prescribed remedy for a Rule 43 violation would thus reduce the burden on the courts while protecting defendants' rights.⁸

* * *

In sum, our caselaw transforms a defendant's right to be present at sentencing into a tangled web of inconsistencies that calls into question heaps of otherwise-valid judgments. Rather than continuing to follow our earlier (unwise) suggestion that the oral sentencing pronouncement controls over the written judgment, we should recognize the centrality of the written judgment. The written judgment is the sentence. It is time we said so.

FOOTNOTES

1.   I am not the first member of our Court to express concerns in this area. See, e.g., United States v. McLaurin, 168 F.4th 693, 712–13 (4th Cir. 2026) (Niemeyer, J., dissenting); United States v. Lassiter, 96 F.4th 629, 640–42 (4th Cir. 2024) (Agee, J., concurring in part and concurring in the judgment); United States v. Kemp, 88 F.4th 539, 547–53 (4th Cir. 2023) (Quattlebaum, J., concurring); United States v. Tostado, No. 23-4423, ––– F.4th ––––, ––––, 2026 WL 881728, at *7–9 (4th Cir. Mar. 31, 2026) (Rushing, J., dissenting). And I venture a guess that our district court colleagues have their own concerns.

2.   Despite the troubling result reached here, I am concerned that this case is not a good vehicle for en banc review. My purpose in writing this statement is not to quarrel with the outcome in this case but to raise broader concerns with the state of our caselaw.

3.   For example, that is why interlocutory orders “merge” into a court's final judgment. Hain Celestial Group, Inc. v. Palmquist, ––– U.S. ––––, 146 S. Ct. 724, 732, ––– L.Ed.2d –––– (2026). That is also why the Federal Rules distinguish between the time that a “court announces a ․ sentence” and the “entry of the judgment.” Fed. R. App. P. 4(b)(2). Consider another illustration from the civil context: If a district court announced a summary judgment ruling orally but handed down a different ruling in its written decision, it would take a whole lot of gumption for an attorney to argue that the oral pronouncement controlled over the written judgment.

4.   If the oral pronouncement of a defendant's sentence genuinely controlled, then we would be free to ignore the written judgment. A discrepancy between the written judgment and oral pronouncement would be of no concern. We would expect judges, prison officials, and probation officers to ignore the written judgment and instead pull the transcript of an oral hearing from PACER. But our Court doesn't (explicitly) require that. This is but one way that our cases suggesting that “the oral sentence controls” are internally inconsistent. Rogers, 961 F.3d at 296; see United States v. Kemp, 88 F.4th 539, 547–53 (4th Cir. 2023) (Quattlebaum, J., concurring).

5.   Our Court did not wrestle with this background principle (or the Supreme Court's direction) when it first suggested in a footnote that the “oral sentence ․ is controlling.” Morse, 344 F.2d at 29 n.1.

6.   That the Rule 43 right is a procedural right is made clear by the fact that in Rogers cases we ask only whether the imposed sentence (the written judgment) was announced in the defendant's presence. Other rules protect a defendant's substantive right to have a valid sentence imposed. Should there be any doubt, remember that the right is protected by the Due Process Clause and the Federal Rules of Criminal Procedure. See Griffin, 2022 WL 17175592, at *6 (Oldham, J., dissenting).

7.   Often, it is unclear in Rogers cases whether the district court created an error by inadvertently adding a provision to the written judgment or instead simply forgot to announce a provision that it intended to impose. Giving the district court the opportunity to decide the proper remedy allows it to more precisely fix its error. For example, if the court forgot to announce an important condition of supervised release, it could hold a new hearing to announce the intended sentence. But if the court inadvertently added a new (or inconsequential) condition, it could simply amend its written judgment to fix the problem.

8.   It is unclear whether the current scheme helps defendants in the mine run of cases. The current remedy of a full resentencing for any discrepancy—along with our procedural-reasonableness caselaw, see United States v. Shields, 779 F. Supp. 3d 774, 779–82 (S.D.W. Va. 2025)—encourages district courts to map out the criminal judgment in advance of the sentencing hearing. To avoid the risk of having sentences vacated on appeal, district courts are incentivized to rely on that pre-selected sentence, instead of meaningfully engaging with the defendant. United States v. Tostado, No. 23-4423, ––– F.4th ––––, ––––, 2026 WL 881728, at *5 (4th Cir. Mar. 31, 2026) (Rushing, J., dissenting). It is far from clear that this scheme benefits defendants, who are forced to convince district court judges to change their minds. On the other hand, allowing a district court to conform the written judgment to the less severe sentence pronounced at the sentencing hearing will necessarily work to a defendant's benefit by removing unannounced conditions from the written judgment.