WALKER v. UNITED STATES (2025)

OPINION

Allen Walker filed a motion seeking habeas relief under 28 U.S.C. § 2255. The government opposed his motion but did not raise the statute of limitations as a defense. The district court, however, denied the motion as untimely. Walker appealed, and we remanded so the district court could determine whether the government's failure to raise the statute of limitations defense amounted to forfeiture or waiver. We cautioned that the district court could not resurrect a waived limitations defense but could potentially consider a forfeited one. On remand, the district court decided that the government forfeited the defense, considered it despite the forfeiture, and again denied Walker's motion as time-barred.

Walker now appeals that forfeiture determination, arguing that because the government waived the statute of limitations defense, the district court should not have considered it. We agree. We reverse and remand so the district court can proceed to the merits of Walker's § 2255 motion.

FACTS AND PROCEDURAL HISTORY

Allen Walker pleaded guilty to conspiracy with intent to distribute and dispense controlled substances in violation of 21 U.S.C. §§ 841 and 846. The district court sentenced him to 96 months in prison on August 20, 2015. Walker did not appeal, so the judgment became final on September 3, 2015. See Fed. R. App. P. 4(b)(1)(A). On that date, the one-year statute of limitations for habeas relief began to run. 28 U.S.C. § 2255(f)(1). That meant Walker had until September 3, 2016, to file a § 2255 motion.

In August 2016, with the deadline less than a month away, Walker sent the district court a letter asking for more time to file his § 2255 motion and for appointment of counsel to help him prepare it. In his letter, Walker explained that he planned to make an ineffective assistance of counsel claim because his trial lawyer never filed a notice of appeal on his behalf—even though Walker signed the notice and instructed his lawyer to file it immediately after he was sentenced. The eleven-page letter carefully detailed the merits of his ineffective-assistance claim along with supporting caselaw. The district court directed the government to respond to Walker's letter. In its response, the government opposed Walker's request for more time. It had obtained an affidavit from Walker's old lawyer, who said Walker had left him a voicemail instructing him not to file the notice of appeal. (Walker disputes this.) Based on the affidavit, the government argued that Walker's underlying ineffective assistance claim was meritless, so granting him an extension and appointing counsel would be pointless. In November 2016, the district court denied Walker's request on jurisdictional grounds but sent Walker a form to use for filing a § 2255 motion. It reasoned that it lacked the authority to adjust deadlines before Walker had even filed the motion that would create a live case or controversy. Walker then mailed another letter to the district court stating that he thought the previous letter he had sent was the § 2255 motion.

On August 17, 2017—almost a year after the statute of limitations expired—Walker filed his § 2255 motion pro se. The district court appointed counsel, who submitted an amended motion. Walker titled the first section of his argument “Timeliness of Filing of § 2255 Motion.” Am. Mot., R. 6, PageID 843. He conceded that the statute of limitations started to run on September 3, 2015, the limitations period ended exactly one year later, he did not file his pro se motion until almost a year after the deadline, and his motion accordingly “was not filed within the one-year statute of limitations under 28 U.S.C. § 2255(f).” Id. To overcome the timeliness obstacle, Walker argued that the district court should either construe his August 2016 letter as a timely filed § 2255 motion or interpret it as having equitably tolled the statute of limitations. The government opposed Walker's amended § 2255 motion, again arguing only the merits of the ineffective assistance claim. It did not make any statute of limitations argument, although it did recount in its procedural history section that Walker received his sentence in August 2015, that Walker requested more time to file his motion in August 2016, and that later, in his follow-up letter to the court, Walker claimed that his August 2016 letter was itself a pro se motion.

Even though the government did not raise a statute of limitations defense, the district court denied Walker's motion as untimely. Walker moved for reconsideration. This time, the government adopted the district court's reasoning, responding that Walker filed his § 2255 motion too late. The district court denied Walker reconsideration.

Walker appealed and argued that (1) the government waived the timeliness argument, so the district court erred by considering it sua sponte, and (2) even if the government merely forfeited the argument, the district court should have given Walker notice and an opportunity to respond before denying his motion on timeliness grounds. We agreed with Walker that the distinction between waiver and forfeiture matters and remanded to the district court to assess whether the government waived or forfeited the statute of limitations defense. We also clarified that in the case of forfeiture, the district court must give Walker the opportunity to present arguments on timeliness. The district court concluded that this was a case of forfeiture, directed Walker to brief the timeliness issue, and ultimately rejected his alternative arguments that his August 2016 letter should be construed as a timely filed § 2255 motion or that his letter tolled the statute of limitations. Walker appealed again, challenging the district court's conclusion that the government merely forfeited timeliness. He did not challenge the district court's resolution of his alternative arguments about the 2016 letter. We granted a certificate of appealability on the narrow waiver-or-forfeiture question.

ANALYSIS

The central question in this appeal is whether the government waived or forfeited the statute of limitations defense when it opposed Walker's § 2255 motion. The district court considered the defense sua sponte and denied Walker's motion on that basis. That's allowed if the government's failure to make the timeliness argument constitutes mere forfeiture. Day v. McDonough, 547 U.S. 198, 202, 126 S.Ct. 1675, 164 L.Ed.2d 376 (2006). But if the government waived the timeliness argument, the district court was “not at liberty” to “bypass, override, or excuse” that waiver. Wood v. Milyard, 566 U.S. 463, 466, 132 S.Ct. 1826, 182 L.Ed.2d 733 (2012). The differences between waiver and forfeiture decide this case, so we begin our analysis by mapping those differences.

I. Comparing Waiver and Forfeiture

Courts and litigants often use the terms forfeiture and waiver interchangeably. See, e.g., Hamer v. Neighborhood Hous. Servs. of Chi., 583 U.S. 17, 20 n.1, 138 S.Ct. 13, 199 L.Ed.2d 249 (2017); Kontrick v. Ryan, 540 U.S. 443, 458 n.13, 124 S.Ct. 906, 157 L.Ed.2d 867 (2004); United States v. Petlechkov, 922 F.3d 762, 767 (6th Cir. 2019); United States v. Montgomery, 998 F.3d 693, 696 (6th Cir. 2021). But they are not the same. Waiver is the “intentional relinquishment or abandonment of a known right.” United States v. Olano, 507 U.S. 725, 733, 113 S.Ct. 1770, 123 L.Ed.2d 508 (1993) (quoting Johnson v. Zerbst, 304 U.S. 458, 464, 58 S.Ct. 1019, 82 L.Ed. 1461 (1938)). It involves a “deliberate decision,” Wood, 566 U.S. at 473, 132 S.Ct. 1826, to “actively abandon[ ]” an issue, Montgomery, 998 F.3d at 697. In contrast, forfeiture is “the failure to make the timely assertion of a right.” Olano, 507 U.S. at 733, 113 S.Ct. 1770. “Waiver is affirmative and intentional, whereas forfeiture is a more passive” act. Berkshire v. Dahl, 928 F.3d 520, 530 (6th Cir. 2019).

But for all the things waiver necessarily is—deliberate, intentional, affirmative—there is one important thing it need not be: explicit. Waiver can be “implied from all the circumstances.” Berghuis v. Thompkins, 560 U.S. 370, 384, 130 S.Ct. 2250, 176 L.Ed.2d 1098 (2010); see also Hemphill v. New York, 595 U.S. 140, 157, 142 S.Ct. 681, 211 L.Ed.2d 534 (2022) (Alito, J., concurring) (“In the prototypical case of implied waiver, the relevant course of conduct signals an intention to relinquish the right at issue.”). There are no “magic words” required for showing intent to waive an argument. D'Ambrosio v. Bagley, 527 F.3d 489, 497 (6th Cir. 2008).

D'Ambrosio holds that even “express” waiver does not have to be explicit. Id. at 495–97. That case addresses the exhaustion requirement of a different habeas provision, which says the state is not “deemed to have waived” the exhaustion requirement unless it “expressly waives” the issue. See 28 U.S.C. § 2254(b)(3). D'Ambrosio explains that even an “express” waiver does not need to be verbal or “expressed in a certain manner.” 527 F.3d at 497. Instead, “[t]he touchstone for determining whether a waiver is express is the clarity of the intent to waive.” Id. There is no “express” requirement here. In fact, § 2255 does not allude to waiver at all in connection with the statute of limitations, let alone add modifiers. And if explicit verbal statements are unnecessary for express waiver, they are undoubtedly unnecessary for ordinary waiver.

So, if a waiver need not be verbal or explicit, how do we determine whether an issue is waived rather than forfeited? There is no clear formula. Instead, it might be easiest to imagine waiver and forfeiture as opposite ends of a continuum. See Montgomery, 998 F.3d at 697. With each additional indication that a party intentionally abandoned an argument rather than merely overlooked it, the party's conduct travels further along that continuum toward waiver. Silence on its own, without other evidence of intent, is not enough for waiver. In the absence of an explicit statement, we look for strong indications of intent. When indications of intent are absent, or when there are affirmative reasons to believe that a party did not intend to abandon an argument, the party's conduct slides the other way, toward forfeiture.

To analyze whether the government's actions in this case add up to forfeiture or to waiver, we first consider a pair of Supreme Court precedents involving, as here, the statute of limitations defense to an untimely habeas petition: Day v. McDonough, 547 U.S. 198, 126 S.Ct. 1675, 164 L.Ed.2d 376 (2006), and Wood v. Milyard, 566 U.S. 463, 132 S.Ct. 1826, 182 L.Ed.2d 733 (2012). In Day, a state prisoner filed an untimely habeas petition. The state miscalculated the number of countable days since the judgment became final and deemed the petition timely. Day, 547 U.S. at 203–04, 126 S.Ct. 1675. The district court dismissed the petition as untimely on its own initiative because there was “no intelligent waiver” by the state, only a “clear computation error.” Id. at 202, 210, 126 S.Ct. 1675. The Supreme Court affirmed but cautioned that it would be “an abuse of discretion to override a State's deliberate waiver of a limitations defense.” Id. at 202, 126 S.Ct. 1675.

That situation arose in Wood. There, the district court directed the state to respond to a prisoner's habeas petition and address potential statute of limitations or exhaustion defenses. Wood, 566 U.S. at 467, 132 S.Ct. 1826. The state's response identified the limitations period's start and end dates and discussed the potential tolling effect of the petitioner's state habeas petition, which indicated the state understood it had a colorable statute of limitations defense. Id. at 469–70, 132 S.Ct. 1826. Then, the state wrote that it did not concede but would “not challenge” timeliness and instead turned to the merits. Id. at 470, 132 S.Ct. 1826. The Supreme Court reasoned that this conduct amounted to waiver because the state evinced “clear and accurate understanding” of the statute of limitations issue but “deliberately steered the District Court away from the question and towards the merits.” Id. at 474, 132 S.Ct. 1826. Hence, the Court concluded that the district court could not dismiss on timeliness grounds. Id.

Day and Wood are instructive starting points. They show us that calculation errors and similar mistakes are hallmarks of forfeiture, whereas party statements directing a court away from an issue the party clearly understands are hallmarks of waiver. Examples of waiver and forfeiture from our cases provide additional insight. For instance, we have held that if a party files a motion or lodges an objection but later withdraws it, it has waived whatever argument it made in the initial motion or objection. See, e.g., United States v. Denkins, 367 F.3d 537, 543–44 (6th Cir. 2004); United States v. Sheppard, 149 F.3d 458, 461 (6th Cir. 1998); United States v. Jackson, 23 F. App'x 254, 255 (6th Cir. 2001). Likewise, we have held that if the district court asserts that some question is a “non-issue” and a party says nothing in response over multiple years of litigation, that silence amounts to a waiver. D'Ambrosio, 527 F.3d at 496. 1 In both situations, we inferred an intent to waive because the party must have known about an issue and yet failed to raise it—either because the party addressed it directly in a motion or objection, or because the district court referred to it repeatedly over years of litigation.

In contrast, when no one raises an issue before the district court, it is less likely that a party ignored it on purpose. For example, when the government simply “neglects to advance” an argument that a plea agreement's terms foreclose habeas relief, it forfeits (rather than waives) that argument on appeal. See Greer v. United States, 938 F.3d 766, 770 (6th Cir. 2019); Hunter v. United States, 160 F.3d 1109, 1113–14 (6th Cir. 1998). In a similar vein, when the government opposes a motion to suppress evidence in the district court but fails to argue that the defendants lack standing to challenge the underlying search, the government forfeits (but does not waive) the standing argument. United States v. Noble, 762 F.3d 509, 526–28 (6th Cir. 2014).

And sometimes, even where an issue is raised before the district court, if a party demonstrates no awareness of a potential argument regarding the issue, we interpret the oversight as unintentional forfeiture rather than deliberate waiver of that argument. As the dissent notes, we see this routinely when criminal defendants fail to challenge trial-procedure rulings or sentencing enhancements despite being on notice of these issues. Montgomery, 998 F.3d at 698. But contrary to the dissent's suggestions, the rule we apply here would not render those omissions waivers. The “mere failure” of a criminal defendant to assert a right or make an argument “at the proper time” does not show the defendant knew they had a colorable argument to make. Dissenting Op. at ––––. Likewise, the fact that the defendant “say[s] nothing in response to the probation officers’ recommendations” at sentencing does not demonstrate the defendant understood the challenges they could make to those recommendations. Id. at ––––.

Taken together, the recurring theme that emerges from all these cases—Day, Wood, and our circuit precedent—is that to decide whether a party made an affirmative choice to abandon an argument or right, we look for signs that a party knew and understood both the relevant issue and potential relevant arguments about it. See id. at –––– (focusing on whether the record shows that a party “understood” the right (quoting Berghuis, 560 U.S. at 384, 130 S.Ct. 2250)). Silence alone is not enough. But the more a party alludes to the issue or argument, and the more the district court or the party's opponent brings attention to the issue or argument, the likelier it is that the party knew and understood that the issue was live, that it could impact the litigation, and that there was a viable way to respond to it. We can reasonably infer that failure to raise the issue in the face of that high level of knowledge is deliberate, and therefore waiver. 2

II. Application

Now, we turn to this case. Walker argues that the district court used the wrong standard when it decided that the government forfeited instead of waived the statute of limitations defense. He claims that the district court reached the wrong outcome as a result. We start by examining whether the district court employed the wrong legal standard, which is a question we review de novo. Ahearn v. Jackson Hosp. Corp., 351 F.3d 226, 234 (6th Cir. 2003) (citing United States v. Willis, 257 F.3d 636, 642 (6th Cir. 2001)).

A. The district court's standard for evaluating intent.

The district court held that the government did not intend to forego the statute of limitations defense and classified its failure to raise timeliness as forfeiture. But the district court based its intent determination on a misunderstanding of the appropriate standard. It concluded that the government's failure to raise the timeliness issue was unintentional because it never “explicitly relinquished” the defense. Apr. 12, 2021, Order, R. 47, PageID 331 (citation omitted). As we have explained, that is the wrong standard for waiver. A party can intentionally forego an argument through strategic silence and steering the court in a different direction. See, e.g., D'Ambrosio, 527 F.3d at 496. Indeed, Walker alleges implied waiver here and anchors his argument in implicit-waiver caselaw from the Supreme Court and this circuit. The absence of an explicit waiver is not at issue.

The government contends the district court understood that intent to waive can be implied and grappled with this possibility when it made its intent determination. But it's hard to find evidence of that in the court's order. The district court relies on a pair of facts to reach its conclusion on forfeiture: (1) the government “failed to raise the timeliness issue at all and even conceded they did not do so until the Court raised the issue sua sponte,” and (2) “Walker himself admits ․ that the Government initially failed to raise the issue.” Apr. 12, 2021, Order, R. 47, PageID 331–32. These facts just confirm that the government didn't make a timeliness argument, which still begs the question whether doing so was inadvertent or purposeful. And the district court does not appear to recognize the possibility of implicit waiver anywhere in its order. It simply concludes that “[t]hese failures by the Government [to raise the issue] are enough for a finding of forfeiture.” Id. at PageID 332.

In sum, we agree with Walker that the district court erred by requiring explicit relinquishment when it looked for signs of intent.

B. The standard for reviewing the district court's intent determination.

We will not defer to the district court's intent determination because it was tainted by a legal error. See Columbus Educ. Ass'n v. Columbus City Sch. Dist., 623 F.2d 1155, 1159 (6th Cir. 1980). Yet we note that even if the district court had used the right standard, we still might not defer. Generally, whether a party forfeited or waived an argument is a “mixed question of law and fact,” such that we evaluate underlying factual findings for clear error but review de novo whether those facts amount to waiver or forfeiture. United States v. Pembrook, 79 F.4th 720, 727 (6th Cir. 2023) (quoting United States v. Lara, 590 F. App'x 574, 586 (6th Cir. 2014)); United States v. Boudreau, 564 F.3d 431, 435 (6th Cir. 2009). But intent separates forfeiture from waiver. So, does evaluating waiver present a typical fact question about subjective intent or a legal question about whether the pleadings objectively show intent to waive the issue?

Our court has treated intent in both ways. It is frequently a fact question, for example, when an intentional mental state is an element of a crime. See, e.g., United States v. Krimsky, 230 F.3d 855, 860 (6th Cir. 2000). In these scenarios, the district court may have to sift through a complex factual record, scrutinize the credibility of witnesses, review affidavits and other sworn testimony, and more. Because the district court is in a “superior position” to do this, we “normally give[ ] great deference” to its factual determinations. Fairport Int'l Expl., Inc. v. The Captain Lawrence, 245 F.3d 857, 862 (6th Cir. 2001). If intent is a fact question here, then the difference between waiver and forfeiture basically boils down to a factual inquiry. And if that is true, then we may have to review the district court's forfeiture determination under the more deferential clear error standard. It is hard to square this with binding precedent that says we review waiver and forfeiture determinations—which are basically intent determinations—de novo. 3

However, intent is sometimes a legal question, such as when we interpret contracts or wills. See, e.g., Orrand v. Scassa Asphalt, Inc., 794 F.3d 556, 561–62 (6th Cir. 2015); President & Fellows of Harvard Coll. v. Jewett, 11 F.2d 119, 121 (6th Cir. 1925). And there are good reasons to suspect that it is here. The information available to resolve the intent question does not consist of typical evidence of subjective intent (like sworn affidavits or live testimony), but of legal documents. A district court is not obviously in a superior position to a court of appeals to interpret a handful of legal filings. Furthermore, the “evidence” of intent here sheds little light on the attorney's subjective state of mind. At oral argument, the government confirmed that “there's no direct evidence of what the particular prosecutor who responded to Mr. Walker's motion had in her mind at the time.” Oral Arg. Recording at 17:55. Instead, briefings before the district court reveal the attorney's objective intent. In that way, the intent question at the heart of this case resembles those we routinely treat as legal issues. See Orrand, 794 F.3d at 561–62; Jewett, 11 F.2d at 121. And if we approach intent as a question of law for the purposes of waiver and forfeiture, we avoid the tension between two competing standards of review and create harmony with our precedent.

For the standard of review in this case, however, we do not have to categorically resolve this question. Regardless of whether we should view the district court's intent determination as a factual finding or a legal conclusion, we know it was infected by a legal error. At oral argument, the government clarified that if we determine that the district court used an incorrect legal standard to reach its intent finding, we should apply the correct standard to the underlying facts ourselves rather than remanding the case. For that reason, we analyze the “evidence”—the recorded communications with and briefing before the district court—de novo to determine whether the government waived or forfeited its timeliness argument.

C. The government's conduct amounts to waiver.

We turn now to whether the government forfeited or waived the statute of limitations defense under the proper standard. In this case, the government encountered the statute of limitations defense in myriad ways. First, Walker signaled that he was concerned about meeting the deadline in his letter asking for more time to file his motion. After all, why would Walker request an extension if he wasn't at any risk of running out of time? Second, the district court directed the government to respond to Walker's request for more time. Third, the government answered that the court should not give Walker more time because his claim was meritless anyway. There is no way to communicate this message without understanding the statute of limitations issue that lies at the heart of it. Not only did Walker's letter and the district court's instructions place the government on notice, but the government's affirmative response to that notice contemplated that Walker's claim would become time-barred. Fourth, the government knew that the district court denied Walker's request for extra time. Fifth, when Walker finally did file, he titled the first section of his argument “Timeliness of Filing of § 2255 Motion,” and conceded that his motion “was not filed within the one-year statute of limitations.” Am. Mot., R. 6, PageID 843. Walker also requested that the district court “treat his August 8, 2016[,] letter as a § 2255 motion.” Id. at PageID 844. Sixth, the government knew that Walker had been sentenced about two years prior; it said as much in its opposition brief. The government correctly recited the date of Walker's sentencing, explained that Walker requested an extension of time to file his § 2255 motion just under a year later, and noted that the district court denied the request. After marching through this procedural history, the government wrote that Walker “now moves ․ pursuant to 28 U.S.C. § 2255” and cited to Walker's amended motion, which was filed more than two years after he was sentenced. Resp., R. 10, PageID 76. Seventh, the government alluded to a potential dispute on the timeliness issue, observing that Walker believed his letter requesting an extension was itself a pro se § 2255 motion. Despite all this, the government asked the court to take an alternative path—decide the merits of Walker's petition.

Looking objectively at the record, these facts show that the government understood the timeline for Walker's motion and knew not only that timeliness was at issue in this case, but specifically that it could make a viable statute of limitations argument. Yet the government chose not to pursue it. This knowledge distinguishes the government's situation from the commonplace reality that sometimes parties unintentionally forfeit an argument because they do not recognize an issue or discover a plausible argument on point until later in the litigation. 4

There is more. We can discern strategic reasons why the government might have chosen this course of action. For example, it may have decided against a timeliness argument to avoid responding to the potentially thorny question of whether Walker's August 2016 letter either stopped the clock or was itself a timely filed motion, especially given the leniency afforded to pro se litigants. See United States v. Asakevich, 810 F.3d 418, 423–24 (6th Cir. 2016); see also Castro v. United States, 540 U.S. 375, 381–83, 124 S.Ct. 786, 157 L.Ed.2d 778 (2003). After all, the eleven-page letter contained a lengthy and articulate discussion of the facts and law supporting Walker's ineffective-assistance claim. Relatedly, the government may have decided, based on Walker's lawyer's affidavit, that it had an ironclad merits argument worth pursuing instead. While it is not necessary for us to identify a strategic reason behind the government's waiver, the failure to raise timeliness here is not a mystifying oversight explained only by inadvertence, further supporting our conclusion.

It is true that the government did not say outright that it was abandoning the timeliness issue. However, the government did not have to “steer[ ]” the district court away from the statute of limitations issue with an explicit statement to convey its intent to waive. See Wood, 566 U.S. at 474, 132 S.Ct. 1826. Given the procedural context, its redirection toward the merits was enough, because waiver can be implied with “silence, coupled with an understanding of [the issue] and a course of conduct indicating waiver.” Berghuis, 560 U.S. at 384, 130 S.Ct. 2250. The dissent acknowledges, as it must, that a “course of conduct” can imply an intent to waive. Dissenting Op. at ––––. Yet it still faults Walker for not providing direct evidence of the government attorney's subjective state of mind. See id. at ––––. Our cases impose no such requirement. See D'Ambrosio, 527 F.3d at 496. And demanding the kind of proof that the dissent seeks would obviate altogether the doctrine of implied waiver, which rests on the premise that courts can discern a party's intent to waive through its actions rather than its words. Tellingly, the dissent does not suggest we remand for testimony from the government's attorney. No party does, which is unsurprising because inserting that kind of requirement into our claims-processing rules would risk turning a forfeiture-versus-waiver dispute into a mini trial. 5

The government instead argues that its actions, as reflected in the pleadings, are better described as forfeiture. It highlights a pair of cases where a party was made aware of an issue before the district court and did not address it, but we still treated the issue as merely forfeited. See United States v. Mabee, 765 F.3d 666 (6th Cir. 2014); Cradler v. United States, 891 F.3d 659 (6th Cir. 2018). But both Mabee and Cradler emphasized that waiver is “explicit[ ],” and neither contended with the possibility of implied waiver. 765 F.3d at 671, 673; 891 F.3d at 665 n.1. And their facts differ from our case's in important ways. In Mabee, we held that a criminal defendant had not intentionally relinquished his right to challenge a sentencing enhancement in part because he contested key facts underlying the enhancement, and his statements about whether the enhancement applied were “conflicting” and “ambiguous at best.” Mabee, 765 F.3d at 672. Not so here, where the government's brief accurately recounted the timeline of Walker's motion and the context indicated the government knew it had a colorable limitations argument, but the government nonetheless made no statements, “ambiguous” or otherwise, addressing timeliness. And while the government's initial silence regarding timeliness did not constitute waiver in Cradler, throughout that litigation, the petitioner had represented that his motion was timely based on his interpretation of Supreme Court caselaw. 891 F.3d at 665. So there was less evidence that the government was aware of its timeliness argument than there is here, where Walker openly conceded that he filed his motion after the deadline.

Because the government waived rather than forfeited the statute of limitations defense, it was improper for the district court to dismiss Walker's § 2255 motion sua sponte on timeliness grounds. See Day, 547 U.S. at 202, 126 S.Ct. 1675.

CONCLUSION

We reverse the district court's dismissal of Walker's § 2255 motion and remand for proceedings consistent with this opinion.

DISSENT

The dividing line between a “forfeiture” and a “waiver” has befuddled courts for decades. And my colleagues in this case thoughtfully place the government's failure to raise a statute-of-limitations defense in its initial response on the “waiver” side of this line. But I do not think the government's conduct rose to the high level that the Supreme Court has suggested a “waiver” requires in this context. I thus respectfully dissent.

Allen Walker sought relief from his federal conviction by filing a postconviction motion under 28 U.S.C. § 2255. His conviction had become final in September 2015. But he did not file his § 2255 motion until August 2017—almost a year beyond the one-year statute of limitations. See 28 U.S.C. § 2255(f)(1). Walker conceded that he had “not filed” his motion in time. Mot., R.6, PageID 843. But he asked the court to treat an earlier (and timely) letter as his § 2255 motion. Id., PageID 844. He also suggested that “equitable tolling” could save the motion. Id., PageID 844 n.3. In response, the government argued the merits of Walker's claims and said nothing about the statute of limitations. Resp., R.10, PageID 75–79. The district court nevertheless dismissed Walker's motion as untimely. Order, R.12, PageID 91. The government only later adopted the court's reasoning in response to Walker's motion for reconsideration. Resp., R.16, PageID 161.

On these facts, should we describe the government's conduct as a “forfeiture” or a “waiver” of the statute-of-limitations defense? Courts have used those verbs “interchangeably” for decades. Freytag v. Comm'r of Internal Revenue, 501 U.S. 868, 894 n.2, 111 S.Ct. 2631, 115 L.Ed.2d 764 (1991) (Scalia, J., concurring in part and concurring in the judgment). Yet the terms have different roots. And while I find it debatable whether the government forfeited its statute-of-limitations defense, I do not think its initial silence (even when coupled with Walker's express statute-of-limitations arguments) amounted to an implied waiver. If applied broadly, such a waiver rule would all but end our routine practice of reviewing arguments that criminal defendants make for the first time on appeal.

I. Forfeiture

How does a party “forfeit” a right? The Supreme Court has defined a forfeiture as “the failure to make the timely assertion of a right[.]” United States v. Olano, 507 U.S. 725, 733, 113 S.Ct. 1770, 123 L.Ed.2d 508 (1993). The doctrine grew out of the common-law “procedural principle” that an appellate court generally will not review a claim of error if a litigant failed to raise the alleged error in a lower court. Yakus v. United States, 321 U.S. 414, 444–45, 64 S.Ct. 660, 88 L.Ed. 834 (1944) (citing cases); see Robert J. Martineau, Considering New Issues on Appeal, 40 Vand. L. Rev. 1023, 1026–28 (1987); Lester B. Orfield, The Scope of Appeal in Criminal Cases, 84 U. Pa. L. Rev. 825, 840 (1936). So a forfeiture generally occurs when a litigant tries to enforce a right in a manner that runs afoul of the claims-processing rules of the “tribunal having jurisdiction” over the right. Yakus, 321 U.S. at 444, 64 S.Ct. 660.

The doctrine has developed two characteristics that matter today. As for the first, courts have long had discretion to overlook a party's failure to follow the procedural rules for raising a right and to consider the party's arguments despite this forfeiture. See, e.g., United States v. Atkinson, 297 U.S. 157, 160, 56 S.Ct. 391, 80 L.Ed. 555 (1936). In the criminal context, for example, the “plain error” test in Federal Rule of Criminal Procedure 52(b) codified that traditional practice. See Olano, 507 U.S. at 731–32, 113 S.Ct. 1770. As for the second, courts have not traditionally examined the reasons for a party's failure to follow the proper procedure. A forfeiture thus could arise from the “bungling of improperly trained lawyers” as much as from a party's intentional disregard of the rules. Orfield, supra, at 845. Today, then, forfeitures regularly result from a lawyer's legal or factual mistakes. See Day v. McDonough, 547 U.S. 198, 211, 126 S.Ct. 1675, 164 L.Ed.2d 376 (2006); Bannister v. Knox Cnty. Bd. of Educ., 49 F.4th 1000, 1011–12 (6th Cir. 2022).

In Day, the Supreme Court clarified how these forfeiture rules apply to the government's right to raise a statute-of-limitations defense in the habeas context. In the Antiterrorism and Effective Death Penalty Act (AEDPA), Congress adopted a one-year limitations period both for a state prisoner's habeas petition under 28 U.S.C. § 2254 and for a federal prisoner's postconviction motion under 28 U.S.C. § 2255. See 28 U.S.C. §§ 2244(d), 2255(f). The claims-processing rules for § 2254 petitions—in particular, Rule 5(d) of the Rules Governing Section 2254 Cases (“State Habeas Rules”)—require a State's answer to a petition to include any statute-of-limitations defense. State Habeas Rule 12 also allows district courts to follow the Federal Rules of Civil Procedure that do not conflict with the habeas laws and rules. And the Civil Rules likewise require defendants to plead statute-of-limitations defenses in an answer. See Day, 547 U.S. at 207–08, 126 S.Ct. 1675.

The State in Day forfeited its statute-of-limitations defense because it did not raise that defense in the way these rules required. In its answer to a state prisoner's § 2254 petition, the State instead conceded that the prisoner had filed a timely claim. See Day, 547 U.S. at 203, 126 S.Ct. 1675. The State's answer also revealed, though, that its concession had rested on a miscalculation and that the prisoner had, in fact, filed an untimely petition. Id. A magistrate judge noticed the State's error and recommended that the district court dismiss the petition as untimely. See id. at 204, 126 S.Ct. 1675. The district court did so, excusing the State's forfeiture. Id. Affirming this result, the Supreme Court held that district courts in the habeas context have discretion to forgive a State's forfeiture and to invoke the statute of limitations on their own initiative. See id. at 205–11, 126 S.Ct. 1675.

Unlike the state prisoner in Day, Walker sought relief from a federal conviction under § 2255. The district court ordered the federal government to respond to Walker's motion “in accordance with” Rule 5 of the Rules Governing Section 2255 Proceedings (“Federal Habeas Rules”). Order, R.2, PageID 37. Unlike its counterpart in the State Habeas Rules, Federal Habeas Rule 5 does not expressly mention any requirement that the government's response include a statute-of-limitations defense. And the government's response here disagreed with Walker on the merits of his constitutional claims but said nothing about AEDPA's statute of limitations. Resp., R.10, PageID 75–79. Did the government's failure to raise the statute-of-limitations defense in its response (like the State's failure in Day) forfeit that defense for the entire § 2255 proceedings?

The answer is not obvious to me. Recall that a forfeiture requires a party to “fail[ ] to make the timely assertion of a right” in the litigation. Olano, 507 U.S. at 733, 113 S.Ct. 1770. So courts cannot decide whether a party has forfeited a right without identifying the procedural rules for when the party must invoke it. Consider the following example: a party files a motion to dismiss under Civil Rule 12(b)(6) on the merits without making any statute-of-limitations argument because the party believes that the merits raise legal questions but the statute-of-limitations defense raises fact-bound questions. Nobody would say that the party has forfeited the right to assert a statute-of-limitations defense later in the litigation by failing to include it in this motion. Why? The relevant rules did not require the party to raise that defense in a motion to dismiss. Cf. Fed. R. Civ. P. 12(h).

We thus must ask when § 2255’s rules required the government to identify its statute-of-limitations defense here. That statutory section says nothing on this topic. See 28 U.S.C. § 2255(b). As mentioned, moreover, Federal Habeas Rule 5(b) does not require the government to identify a statute-of-limitations defense in its answer. Cf. Wood v. Milyard, 566 U.S. 463, 470, 132 S.Ct. 1826, 182 L.Ed.2d 733 (2012). Rather, it says: “The answer must address the allegations in the motion. In addition, it must state whether the moving party has used any other federal remedies, including any prior post-conviction motions under these rules or any previous rules, and whether the moving party received an evidentiary hearing.” Federal Habeas Rule 5(b). And while Federal Habeas Rule 12 says that a district court “may” follow any nonconflicting Civil or Criminal Rules when considering a § 2255 motion, the rule does not tell the district court that it must do so. A response to a § 2255 motion thus might not resemble a traditional “answer” in a civil suit. Cf. United States v. Boniface, 601 F.2d 390, 392–93 (9th Cir. 1979). Perhaps courts should extend to this § 2255 context the ordinary civil rule that defendants must include statutes-of-limitations defenses in their answers. See Wood, 566 U.S. at 470, 132 S.Ct. 1826. But I see nothing in the Federal Habeas Rules that unambiguously compels this approach. Nor does existing precedent resolve whether the government forfeits a statute-of-limitations defense if it raises the defense only later rather than in its initial response.

Besides, parties also may eliminate any forfeiture concerns under the Civil Rules by moving for leave to amend their answers to include new defenses. Courts “should freely give leave when justice so requires.” Fed. R. Civ. P. 15(a)(2); see Day, 547 U.S. at 208–09, 126 S.Ct. 1675. So the district court in this case could have notified the government of its statute-of-limitations concerns before dismissing Walker's motion as untimely. Indeed, every Justice in Day (including the dissenters) blessed this alternative to a sua sponte dismissal. See 547 U.S. at 209 & n.9, 126 S.Ct. 1675; id. at 216–17 & n.2, 126 S.Ct. 1675 (Scalia, J., dissenting). If the district court had followed the alternative here, the government could have sought to amend its response to invoke the statute of limitations. We have every reason to believe that the government would have done so. After all, it has steadfastly sought to enforce the statute of limitations ever since it responded to Walker's motion for reconsideration. I also see no reason why the court could not have “freely given” the government leave to amend. Fed. R. Civ. P. 15(a)(2). And this step would have barred any claim that the government had forfeited the statute-of-limitations defense. If this case had included this modest additional process, I doubt anyone would say that the government's initial omission amounted to a waiver.

All told, Day’s holding that district courts may shortcut the amendment process by immediately dismissing an untimely suit may well have caused more problems than it solved—as the several rounds of litigation in this case attest. That said, the government has conceded that it forfeited any statute-of-limitations defense. I would accept this concession here. Still, the concession makes no difference to the question presented. If the government forfeited the statute-of-limitations defense, the district court still had the discretion (rather than the duty) to consider the defense under Day. See 547 U.S. at 209, 126 S.Ct. 1675; see also Perrone v. United States, 889 F.3d 898, 909 (7th Cir. 2018). Admittedly, the forfeiture could open up an additional appellate avenue for Walker if he could show that the court abused its discretion by relying on the statute of limitations. But we limited our certificate of appealability to the single issue whether the government's conduct amounted to a forfeiture or a waiver. I thus would not reach any subsidiary issue about whether the court abused its discretion in this way.

II. Waiver

How does a party “waive” a right? It depends. Many sources—from the U.S. Constitution, to federal statutes, to state common law—create “rights.” The Supreme Court has generally recognized that parties may give up most of these rights. See United States v. Mezzanatto, 513 U.S. 196, 201, 115 S.Ct. 797, 130 L.Ed.2d 697 (1995); Shutte v. Thompson, 82 U.S. 151, 159, 15 Wall. 151, 21 L.Ed. 123 (1873). And if a party has waived (rather than forfeited) a right, courts lack discretion to enforce it anyway. See Wood, 566 U.S. at 466, 132 S.Ct. 1826; see also Olano, 507 U.S. at 732–33, 113 S.Ct. 1770. But the Court has added that “[w]hat suffices for waiver depends on the nature of the right at issue.” New York v. Hill, 528 U.S. 110, 114, 120 S.Ct. 659, 145 L.Ed.2d 560 (2000). That is, while forfeiture turns on court rules for enforcing a right, waiver turns on the right itself.

Start with the waiver of constitutional rights. When describing what the Constitution demands for a party to waive a right found in that document, the Supreme Court has long defined “waiver” as the “intentional relinquishment or abandonment of a known right.” Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 682, 119 S.Ct. 2219, 144 L.Ed.2d 605 (1999) (quoting Johnson v. Zerbst, 304 U.S. 458, 464, 58 S.Ct. 1019, 82 L.Ed. 1461 (1938)); see Brewer v. Williams, 430 U.S. 387, 404, 97 S.Ct. 1232, 51 L.Ed.2d 424 (1977). This test has two parts. See Berghuis v. Thompkins, 560 U.S. 370, 382, 130 S.Ct. 2250, 176 L.Ed.2d 1098 (2010). The “known right” part shows that parties must have “full awareness” of the “right” and the “consequences” of giving it up. Zerbst, 304 U.S. at 464, 58 S.Ct. 1019; Moran v. Burbine, 475 U.S. 412, 421, 106 S.Ct. 1135, 89 L.Ed.2d 410 (1986). The “intentional relinquishment” part shows that parties must make a “voluntary” decision to abandon the right free from any “intimidation, coercion, or deception.” Zerbst, 304 U.S. at 464, 58 S.Ct. 1019; Moran, 475 U.S. at 421, 106 S.Ct. 1135.

Despite adopting a uniform constitutional test, the Court has applied this test in a way that requires different things for different rights. In the test's most demanding form (governing the right to counsel or to a jury trial), the Court has required parties to personally and expressly relinquish a right. See Gonzalez v. United States, 553 U.S. 242, 248–49, 128 S.Ct. 1765, 170 L.Ed.2d 616 (2008); Schneckloth v. Bustamonte, 412 U.S. 218, 235–38, 93 S.Ct. 2041, 36 L.Ed.2d 854 (1973). The Court has also held, however, that a party's lawyer can give up other rights on the party's behalf—presumably as long as the lawyer acts knowingly and voluntarily. See Gonzalez, 553 U.S. at 249–53, 128 S.Ct. 1765. And courts have held that a party can impliedly waive other rights (such as the right to represent oneself) by acting in a way that conflicts with the exercise of the right (such as by hiring a lawyer). See Carson v. United States, 88 F.4th 633, 646 (6th Cir. 2023); United States v. Ortiz, 82 F.3d 1066, 1071 (D.C. Cir. 1996).

But this case concerns a statutory right rather than a constitutional one. Waiver in the statutory context does not depend on any constitutional test. Rather, the Supreme Court has asked an ordinary question of statutory interpretation to decide both whether parties may waive statutory rights and what conduct they must undertake to do so. See Hill, 528 U.S. at 116, 120 S.Ct. 659; see also Zedner v. United States, 547 U.S. 489, 500–03, 126 S.Ct. 1976, 164 L.Ed.2d 749 (2006); Crosby v. United States, 506 U.S. 255, 258–62, 113 S.Ct. 748, 122 L.Ed.2d 25 (1993). The Court starts with the recognition that Congress passes laws “against a background of common-law adjudicatory principles.” Astoria Fed. Sav. & Loan Ass'n v. Solimino, 501 U.S. 104, 108, 111 S.Ct. 2166, 115 L.Ed.2d 96 (1991). So when a statute says nothing about waiver, the Court presumes that parties may waive statutory rights. Mezzanatto, 513 U.S. at 201, 115 S.Ct. 797. That said, the text of some statutes can rebut this presumption. The Court, for example, has read Title VII to bar employees from entering a “prospective waiver” of that law's antidiscrimination protections. Alexander v. Gardner-Denver Co., 415 U.S. 36, 51, 94 S.Ct. 1011, 39 L.Ed.2d 147 (1974); cf. Zedner, 547 U.S. at 500–03, 126 S.Ct. 1976. And when a statute lists the ways in which a party may waive a right, the Court sometimes concludes that Congress meant to “occupy the field” and bar waivers in other ways. Mezzanatto, 513 U.S. at 201, 115 S.Ct. 797.

The right at issue here arises from AEDPA. As noted, that law created a one-year statute of limitations for federal prisoners to seek postconviction relief under § 2255. See id. § 2255(f). The law says nothing about whether (and if so how) the government may waive this defense. The law's silence thus triggers the “background presumption” that parties may waive their rights, including their statute-of-limitations defenses. Mezzanatto, 513 U.S. at 203, 115 S.Ct. 797; cf. United States v. Kwai Fun Wong, 575 U.S. 402, 409–10, 135 S.Ct. 1625, 191 L.Ed.2d 533 (2015). Indeed, the Supreme Court has twice recognized that a State can waive the analogous statute of limitations that AEDPA created for habeas petitions filed under § 2254. See Wood, 566 U.S. at 474, 132 S.Ct. 1826; Day, 547 U.S. at 202, 211, 126 S.Ct. 1675.

That conclusion leads to the next question: what does AEDPA require for a waiver? The Supreme Court in Wood incorporated its constitutional test into the statute, asking whether the government has engaged in an “intentional relinquishment or abandonment of [its] known” statute-of-limitations defense. 566 U.S. at 474, 132 S.Ct. 1826 (citation omitted). That decision may well make sense if we presume that a statute silent on waiver adopts background rules. Traditional sources defined waiver in a similar way: “the voluntary abandonment or surrender, by a capable person, of a right known by him to exist, with the intent that such right shall be surrendered and such person forever deprived of its benefit.” Renzo D. Bowers, Treatise on the Law of Waiver § 1, at 19 (1914). But again, the Court has applied this constitutional test with different levels of vigor depending on the right at stake, sometimes requiring express waivers and other times allowing implied ones.

So we must next consider what version of this test applies. I agree with my colleagues that we should not read AEDPA (like, say, a criminal defendant's right to a jury trial) to require an express waiver. For one thing, courts have long held that waivers can be implied—not just express. Traditionally, a waiver might arise from “acts ․ manifesting an intent and purpose not to claim the supposed advantage; or by a course of acts and conduct, or by so neglecting and failing to act, as to induce the belief that it was his intention and purpose to waive.” Id. § 1, at 19; see Dunkel Oil Corp. v. Indep. Oil & Gas Co., 70 F.2d 967, 969 (7th Cir. 1934). For another thing, AEDPA elsewhere included a provision that does require express waivers. When defending against a § 2254 petition, a State may raise the defense that the petitioner failed to exhaust a claim in state court. See 28 U.S.C. § 2254(b)(1)(A). AEDPA adds: “A State shall not be deemed to have waived the exhaustion requirement or be estopped from reliance upon the requirement unless the State, through counsel, expressly waives the requirement.” Id. § 2254(b)(3) (emphasis added). But Congress included no similar provision for the statute of limitations. And we normally assume that Congress acts intentionally with its inclusion of language in one place and its exclusion elsewhere. See Duncan v. Walker, 533 U.S. 167, 173, 121 S.Ct. 2120, 150 L.Ed.2d 251 (2001). Implied waivers thus can count.

Yet what type of conduct amounts to an “implied” waiver? Because Wood incorporated the constitutional test, the conduct presumably still must satisfy the two constitutional elements for an effective waiver: the government must voluntarily and knowingly relinquish the statute of limitations. See Berghuis, 560 U.S. at 382–83, 130 S.Ct. 2250. To voluntarily give up a right in an implied way, a party must engage in “conduct clearly and reasonably exemplifying” a choice to abandon it. Bowers, supra, § 1, at 20. And, as we have often said when finding that criminal defendants did not waive an objection, the “mere failure” to assert the right at the proper time does not reveal such an implied choice. United States v. Fowler, 819 F.3d 298, 306 (6th Cir. 2016); see United States v. Russell, 26 F.4th 371, 375 (6th Cir. 2022); United States v. Holland, 799 F. App'x 380, 384 (6th Cir. 2020); United States v. Mabee, 765 F.3d 666, 671 (6th Cir. 2014); United States v. Bowen, 194 F. App'x 393, 405 (6th Cir. 2006); United States v. Koeberlein, 161 F.3d 946, 949 (6th Cir. 1998); United States v. Branham, 97 F.3d 835, 842 (6th Cir. 1996); see also Berkshire v. Dahl, 928 F.3d 520, 530 (6th Cir. 2019). Rather, a party's conduct must affirmatively contradict any intent to exercise the right. Defendants, for example, cannot both represent themselves and retain lawyers. So if a “defendant appears at trial with counsel,” it is safe to say that the defendant has relinquished the right to self-representation. Ortiz, 82 F.3d at 1071.

To knowingly give up a right in an implied way, the record must next show that the party “understood” the right. Berghuis, 560 U.S. at 384, 130 S.Ct. 2250. Yet if a failure to enforce a right flows out of a legal or factual mistake, the party has not knowingly waived it. Day confirms this point. The government there arguably entered an express waiver because it told the court that the petitioner had timely filed for relief. 547 U.S. at 203, 126 S.Ct. 1675. But this statement was not a valid waiver because it rested on an “inadvertent error” about how to calculate the limitations period. Id. at 211, 126 S.Ct. 1675.

Precedent applying the right to remain silent from Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966), shows what an implied waiver looks like. Even if a defendant does not sign an express waiver, the Supreme Court has held, the defendant's conduct can reveal “an implied waiver of the right to remain silent.” Berghuis, 560 U.S. at 384, 130 S.Ct. 2250. Defendants typically will know of this right's scope because the Miranda warnings exist to alert them of it. Id. at 385–86, 130 S.Ct. 2250. And after receiving these warnings, defendants sometimes start talking to the police. This decision to speak (if voluntary) qualifies as a “course of conduct” that contradicts an intent to invoke a right not to speak. Id. at 386, 130 S.Ct. 2250 (quoting North Carolina v. Butler, 441 U.S. 369, 373, 99 S.Ct. 1755, 60 L.Ed.2d 286 (1979)).

Here, however, Walker has not made out such an implied waiver. As the party asserting waiver, he has the burden of proving it by a preponderance of the evidence. See id. at 384, 130 S.Ct. 2250; Nationwide Mut. Ins. v. Home Ins., 330 F.3d 843, 846 (6th Cir. 2003). He has met neither element. To start, the government did not engage in a “course of conduct” that affirmatively contradicted its reliance on the statute-of-limitations defense. Berghuis, 560 U.S. at 386, 130 S.Ct. 2250 (quoting Butler, 441 U.S. at 373, 99 S.Ct. 1755). At most, it failed to raise this defense at the time the court's rules required parties to raise it even after Walker had flagged the issue by making arguments as to why his motion was timely. But the “mere failure” to raise a defense at the proper time has never sufficed to show an implied waiver. Fowler, 819 F.3d at 306. To the contrary, that conduct is the very definition of “forfeiture”: “the failure to make the timely assertion of a right[.]” Olano, 507 U.S. at 733, 113 S.Ct. 1770. And the government has steadfastly sought to enforce this defense ever since the district court alerted the government to it. The government's course of conduct, then, looks nothing like a defendant's decision to speak and thus to relinquish the right not to speak. Berghuis, 560 U.S. at 386, 130 S.Ct. 2250.

Next, the record lacks the type of evidence that would convince me that the government knowingly gave up the statute-of-limitations defense. Cf. United States v. Petlechkov, 922 F.3d 762, 768 (6th Cir. 2019). Nothing discloses why the government failed to raise the statute of limitations at the proper time. Did the government lawyer recognize that it had to raise the affirmative defense in it its response? Did she (mistakenly) believe that the letter that Walker had previously filed could count as his § 2255 petition? Did she have a correct view of the requirements for equitable tolling? We lack the answers to any of these questions. For all we know, the response's failure to include the defense arose from the “bungling” of an overburdened lawyer—not from an intentional choice to forgo the defense. Orfield, supra, at 845. But such “an inadvertent error” does not show a waiver. Day, 547 U.S. at 211, 126 S.Ct. 1675. And I agree with my colleagues that we should not remand for an evidentiary hearing about the government lawyer's knowledge on these questions. Rather, since Walker bears the burden of proof, I would hold this lack of clarity against him. Because the record fails to establish a knowing waiver, there was no knowing waiver.

If anything, Walker's argument will likely hurt far more criminal defendants than it will help. In the criminal context, the difference between forfeiture and waiver is the difference between no review and plain-error review. See Olano, 507 U.S. at 732–33, 113 S.Ct. 1770. And although the government occasionally raises new arguments on appeal, see, e.g., Russell, 26 F.4th at 374–75, defendants much more often do so. Walker's waiver logic could bar review in many of these cases. Probation officers, for example, routinely recommend sentencing enhancements under the Sentencing Guidelines. See Holland, 799 F. App'x at 382. Just as Walker's § 2255 motion alerted the government to his belief that the statute of limitations did not bar review, the probation officers in these cases alert defendants to their belief that the facts justify the enhancements. See id. Like the government here, however, defendants often say nothing in response to the probation officers’ recommendations. See id. at 384. Traditionally, we have held that this chain of events showed only a forfeiture, so we have reviewed the new arguments for plain error. See id. Should we now treat the chain of events as a waiver? Must we scour the record to speculate whether it reveals a strategic reason for defense counsel's failure to object? I would not go down this road. But that fact neutrally leads me to conclude that the government committed only a forfeiture here.

The main case on which Walker relies—D'Ambrosio v. Bagley, 527 F.3d 489 (6th Cir. 2008)—does not change things. There, we held that the State had “expressly waive[d]” the argument that a habeas petitioner had not exhausted a claim in state court under AEDPA's express-waiver provision. 28 U.S.C. § 2254(b)(3); see D'Ambrosio, 527 F.3d at 494–97. The district court in that case had noted its understanding from talking with counsel that the State “would not argue that the ․ claim was unexhausted,” and the court expressed its belief that the petitioner's claim was exhausted because he could no longer seek relief in state court. D'Ambrosio, 527 F.3d at 496. The State never corrected the district court's understanding that it would not raise an exhaustion defense. Id. Rather, it agreed that the petitioner could not seek state-court relief and asserted a procedural-default defense. Id. Yet habeas “exhaustion” exists whenever a petitioner can “no longer” seek “state-court remedies,” “regardless of the reason for their unavailability.” Woodford v. Ngo, 548 U.S. 81, 92–93, 126 S.Ct. 2378, 165 L.Ed.2d 368 (2006). So the State's express concession that the petitioner could not seek state-court relief was identical to an express concession that the claim was exhausted. And its reliance on the procedural-default defense affirmatively contradicted any exhaustion defense because the same claim could not be both procedurally defaulted and unexhausted.

The facts of Walker's case look nothing like the facts in D'Ambrosio. Unlike the district court in D'Ambrosio, the district court here immediately noted its understanding that the claim was untimely. And unlike the State in D'Ambrosio, the government never expressly said that Walker's claim was timely. Nor did it raise other defenses that contradicted its intent to rely on a statute-of-limitations defense. At most, the government simply failed to assert the statute of limitations at the proper time. Before today, that conduct would have counted as a classic forfeiture.

In short, I would affirm. I thus respectfully dissent.

FOOTNOTES

1.   Recall that D'Ambrosio is a case about express waiver. D'Ambrosio, 527 F.3d at 495. Crucially, in D'Ambrosio, we decided that years of silence constituted not just an ordinary waiver, but an unambiguous and clear waiver. Id. at 496.

2.   The dissent questions whether, to enforce the statute of limitations against a habeas petition under 28 U.S.C. § 2255, the government needs to raise it at all. See Dissenting Op. at –––– – –––– (discussing Federal Habeas Rule 5(b)). Even if it does, according to the dissent, the district court here could have “notified the government of its statute-of-limitations concerns” and then given the government leave to amend its response, thereby allowing the government to avoid forfeiting the defense. Id. We would not be so quick to invite judges to raise affirmative defenses for the government. In any event, we need not resolve that issue here. As the dissent recognizes, the government concedes that it forfeited the statute of limitations defense. The government's only argument on appeal is that it did not waive the defense. Indeed, by affirmatively conceding that it at least forfeited the statute of limitations defense, the government has waived any argument that it didn't need to raise the defense in the first place.

3.   It is also hard to reconcile this with another reality: Generally, in waiver and forfeiture cases, neither the Supreme Court nor our court remands to the district court to take evidence or make factual findings about the subjective intent of the parties, opting instead to answer the intent question from a clean slate on appeal based on the filings below. See, e.g., Wood, 566 U.S. at 474, 132 S.Ct. 1826; Boudreau, 564 F.3d at 435; In re Anheuser-Busch Beer Labeling Mktg. & Sales Pracs. Litig., 644 F. App'x 515, 527–29 (6th Cir. 2016).

4.   The dissent characterizes the government's actions here as a “mere failure to raise [the] defense at the proper time.” Dissenting Op. at –––– (quotation marks omitted). That is a strawman. As we have explained, the government kept mum on the statute of limitations in the face of multiple indicia—from both Walker and the district court—that timeliness was at issue and that the government might have a colorable argument on it. See Berghuis, 560 U.S. at 384, 130 S.Ct. 2250. The dissent's analogy to caselaw on Miranda rights reinforces our point. The district court and Walker repeatedly warned the government of its right to challenge timeliness, and the government indicated that it understood a possible timeliness defense. Yet the government chose instead to address the merits, “contradict[ing] an intent to invoke” the statute of limitations. Dissenting Op. at –––– (citing Berghuis, 560 U.S. at 386, 130 S.Ct. 2250).

5.   Walker briefly argues in the alternative that the district court erred by considering the government's forfeited statute of limitations defense without first deciding whether there were any extraordinary circumstances that excused the forfeiture. See Hutcherson v. Lauderdale County, 326 F.3d 747, 757 (6th Cir. 2003) (noting that “[c]ourts generally lack the ability to raise an affirmative defense sua sponte” outside “special circumstances”) (citation omitted). His argument falls outside the scope of the certificate of appealability we granted. We need not address the argument or decide whether we can even consider it because we hold that the government waived the timeliness defense.

BLOOMEKATZ, Circuit Judge.

BLOOMEKATZ, delivered the opinion of the court in which MOORE, J., concurred. MURPHY, J. (pp –––– – ––––), delivered a separate dissenting opinion.