Ramon Hueso v. J.A. Barnhart, 948 F.3d 324

See case details

RAMON HUESO, Petitioner-Appellant, v. J.A. BARNHART, Warden, Respondent-Appellee.

No. 18-6299

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

Decided and Filed: January 9, 2020

20a0012p.06

Before: MOORE, KETHLEDGE, and MURPHY, Circuit Judges.

RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b). Appeal from the United States District Court for the Eastern District of Kentucky at London. No. 6:18-cv-00176—Danny C. Reeves, District Judge.

COUNSEL

ON BRIEF: Cathryn R. Armistead, ARMISTEAD LAW GROUP, PLLC, Nashville, Tennessee, for Appellant. Charles P. Wisdom, Jr., Kyle M. Melloan, UNITED STATES ATTORNEY’S OFFICE, Lexington, Kentucky, for Appellee.

MURPHY, J., delivered the opinion of the court in which KETHLEDGE, J., joined. MOORE, J. (pp. 23–43), delivered a separate dissenting opinion.

OPINION

MURPHY, Circuit Judge. Since the founding, Congress has adjusted and readjusted the important balance between an individual’s interest in correcting a wrongful conviction and society’s interest in stopping perpetual attacks on final criminal judgments. In the Antiterrorism and Effective Death Penalty Act of 1996, Congress adjusted this balance again, this time by placing greater weight on the finality of completed cases. This law generally gives federal prisoners just one chance to overturn a final criminal judgment—by alleging any and all errors in a single motion to vacate under 28 U.S.C. § 2255. The law permits a second § 2255 motion only if prisoners show new evidence of their innocence or a new rule of constitutional law from the Supreme Court. 28 U.S.C. § 2255(h)(1)–(2). Since 1996, therefore, prisoners have not been able to file a second § 2255 motion based on a new rule of statutory law from the Supreme Court.

Unable to invoke new statutory decisions in a second § 2255 motion, prisoners have turned to a different vehicle: a petition for a writ of habeas corpus under 28 U.S.C. § 2241. But they have faced a different obstacle: § 2255 has long barred federal prisoners from seeking habeas relief unless they show that § 2255’s remedy is “inadequate or ineffective to test the legality of [their] detention.” Id. § 2255(e). Courts disagree over when (if ever) § 2255(h)’s limits on second § 2255 motions—when combined with a new statutory decision issued after the denial of a first motion—render § 2255 “inadequate or ineffective” so as to permit a second round of litigation under § 2241. Compare McCarthan v. Dir. of Goodwill Industries-Suncoast, Inc., 851 F.3d 1076 (11th Cir. 2017) (en banc), with Brown v. Caraway, 719 F.3d 583 (7th Cir. 2013). Our court has joined those that have made the broadest inroads into the 1996 limits in § 2255(h). Unlike some courts, for example, we allow new habeas petitions even if a later Supreme Court decision affects only a prisoner’s sentence, not just the prisoner’s conviction. Hill v. Masters, 836 F.3d 591 (6th Cir. 2016).

Ramon Hueso asks us to go further still. He argues that prisoners barred from filing a second § 2255 motion may seek habeas relief under § 2241 based on new decisions from the circuit courts, not just the Supreme Court. Although the Fourth Circuit has blessed an identical request, United States v. Wheeler, 886 F.3d 415, 428–29 (4th Cir. 2018), we must respectfully decline. Among our reasons: Congress allowed prisoners to file a second § 2255 motion only if the Supreme Court adopts a new rule of constitutional law. 28 U.S.C. § 2255(h)(2). We would write this limit out of the statute if we held that new rules from the circuit courts (whether of statutory or constitutional law) could render § 2255 “inadequate or ineffective” and trigger the right to a second round of litigation under § 2241. We thus affirm the denial of Hueso’s habeas petition.

I.

The Supreme Court long ago recognized that “the power to award the writ [of habeas corpus] by any of the courts of the United States, must be given by written law,” not common law. Ex parte Bollman, 8 U.S. 75, 94 (1807). This case concerns the relationship between two of those written laws: 28 U.S.C. §§ 2241 and 2255. The history of these laws—both before and after Congress’s 1996 changes—puts this case’s complicated statutory question in its proper context.

A.

Section 2241, which allows courts to grant “[w]rits of habeas corpus,” dates to the Judiciary Act of 1789. 28 U.S.C. § 2241(a); McCleskey v. Zant, 499 U.S. 467, 477–78 (1991). The Supreme Court initially interpreted this statute, like the common-law writ, not to apply to prisoners who had been convicted by a court of competent jurisdiction. Ex parte Watkins, 28 U.S. 193, 202–03 (1830). But the Court gradually expanded its interpretation of the habeas statute to permit more and more “collateral” attacks on final criminal judgments. See McCleskey, 499 U.S. at 478–79.

This expansion caused two practical problems. For one, courts could issue writs only “within their respective jurisdictions,” so prisoners filed habeas petitions in the court with jurisdiction over the prison detaining them. See Rumsfeld v. Padilla, 542 U.S. 426, 446–47 (2004). This rule channeled the growing number of petitions into the few courts with jurisdiction over prisons, compelling those courts to review cases from faraway locations. United States v. Hayman, 342 U.S. 205, 213–14, 214 n.18 (1952). For another, “res judicata” did not apply to common-law petitions. McCleskey, 499 U.S. at 479. Courts thus read the habeas statute as allowing prisoners to file multiple requests. Id. This reading “stimulated the filing of unmeritorious successive petitions,” which were submitted “with the hope, perhaps, of reaching the ear of a different judge[.]” Louis E. Goodman, Use and Abuse of the Writ of Habeas Corpus, 7 F.R.D. 313, 315 (1948).

In 1948, Congress passed legislation with remedies tailored to each of these two problems. Pub. L. No. 80-773, 62 Stat. 869, 964–68 (1948).

First Remedy: The 1948 law eliminated the need for courts to review distant judgments by creating a new cause of action in 28 U.S.C. § 2255. 62 Stat. at 967–68. Section 2255 afforded prisoners the same rights granted by the habeas statute (now moved to § 2241), but in a “more convenient forum“: the sentencing court, not the court of confinement. Hayman, 342 U.S. at 219. To ensure that prisoners would use this new remedy, § 2255 directed courts not to entertain a habeas petition under § 2241 if a prisoner had not filed (or had unsuccessfully filed) a § 2255 motion. 62 Stat. at 968; see, e.g., Broadus-Bey v. Diamond, 264 F.2d 242, 242–43 (6th Cir. 1959) (per curiam).

That said, § 2255’s ban on habeas filings came with an exception that we have come to call its “saving” or “savings” clause (now in § 2255(e)). This clause clarified that prisoners could not file habeas petitions under § 2241 “unless it also appears that the remedy by motion [under § 2255] is inadequate or ineffective to test the legality of [their] detention.” 62 Stat. at 968. Before 1996, courts read this clause as “allow[ing] resort to § 2241 sparingly.” Prost v. Anderson, 636 F.3d 578, 588 (10th Cir. 2011); cf. Cohen v. United States, 593 F.2d 766, 770–71, 771 n.12 (6th Cir. 1979). For example, a prisoner’s “lack of success” on the merits in a § 2255 proceeding did not show § 2255’s inadequacy. E.g., Litterio v. Parker, 369 F.2d 395, 396 (3d Cir. 1966) (per curiam).

Second Remedy: The 1948 law also limited multiple collateral filings. A (now-superseded) sentence in § 2255 said that a “sentencing court shall not be required to entertain a second or successive [§ 2255] motion for similar relief on behalf of the same prisoner.” 62 Stat. at 967. And a new § 2244 adopted similar language for habeas petitions. Id. at 965–66. The Supreme Court read these two provisions to codify the limits on multiple filings that courts had been developing in common-law fashion before 1948. Sanders v. United States, 373 U.S. 1, 12–14 (1963). By the 1990s, these judicially developed limits—which the Court described as a “qualified application of the doctrine of res judicata“—allowed courts to dismiss successive petitions (those raising the same issues decided in a prior petition) or abusive petitions (those raising new issues that could have been asserted in a prior petition). Schlup v. Delo, 513 U.S. 298, 318–19, 318 n.34 (1995).

Because the 1948 statute allowed courts to exercise discretion when enforcing its limits on multiple filings, courts developed various exceptions to these limits. By the 1990s, prisoners could file a second collateral challenge raising a new claim if they showed “cause” for the delay in asserting the claim and “prejudice” from the alleged error. See McCleskey, 499 U.S. at 493–94; United States v. Flores, 981 F.2d 231, 234–35, 234 n.4 (5th Cir. 1993). They also could file a second collateral challenge if they made “a proper showing of actual innocence.” See Herrera v. Collins, 506 U.S. 390, 404–05 (1993). Changes in the Supreme Court’s interpretation of a criminal statute implicated both of these judicial exceptions. See United States v. Richards, 5 F.3d 1369, 1370–72 (10th Cir. 1993). A new interpretation that was “novel” could prove cause. Reed v. Ross, 468 U.S. 1, 15 (1984). And a new interpretation that limited a statute’s scope could show that a prisoner was innocent of the offense under the statute’s narrowed reach. Bousley v. United States, 523 U.S. 614, 623–24 (1998).

B.

In 1996, Congress replaced these judicially developed limits on multiple § 2255 motions with a categorical limit (now in § 2255(h)) that generally prohibits second § 2255 motions. See Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), Pub. L. No. 104-132, 110 Stat. 1214, 1220. Yet Congress adopted two statutory exceptions to this new statutory limit that codified narrower versions of the prior judicially developed exceptions. A new change-in-law exception allows a second motion based only on a “new rule of constitutional law, made retroactive to cases on collateral review by the Supreme Court, that was previously unavailable.” 28 U.S.C. § 2255(h)(2). And a new actual-innocence exception allows a second motion based only on “newly discovered evidence that, if proven and viewed in light of the evidence as a whole, would be sufficient to establish by clear and convincing evidence that no reasonable factfinder would have found the movant guilty of the offense[.]” Id. § 2255(h)(1).

This case mixes a rule that Congress revised with one it left alone. Start with the revision: Congress did not include changes in statutory interpretation as a ground for a second § 2255 motion. Section 2255(h) instead requires either a new constitutional rule by the Supreme Court or new evidence of innocence. Now for what remained: Congress did not touch § 2255(e)’s text allowing a habeas petition under § 2241 if § 2255’s “remedy by motion is inadequate or ineffective to test the legality of [a prisoner’s] detention.” Putting the two together, prisoners have argued that Congress—in imposing § 2255(h)’s new limits—meant to impliedly keep the old exception for changes in statutory interpretation by rendering § 2255 inadequate under § 2255(e) when those changes occur. Under this view, prisoners may seek relief based on a new statutory interpretation, but they now must do so with a § 2241 petition (not a § 2255 motion) in their court of confinement (not their sentencing court).

Circuit courts fall along a spectrum in addressing this argument. Some reject it outright because Congress did not include a statutory-changes exception in § 2255(h). E.g., Prost, 636 F.3d at 585–86. Many others hold that § 2255(h)’s limits render § 2255 inadequate in “unusual” situations. In re Dorsainvil, 119 F.3d 245, 251 (3d Cir. 1997). These courts permit new habeas petitions if prisoners show both that a Supreme Court decision adopting a new statutory interpretation proves their “actual innocence,” and that they “had no earlier opportunity” to assert that new interpretation in their initial § 2255 motion. Id. at 251–52; Triestman v. United States, 124 F.3d 361, 377 (2d Cir. 1997). Still others allow § 2241 petitions for new statutory cases that affect a prisoner’s sentence, even if the case does not prove the prisoner’s innocence of the offense. Wheeler, 886 F.3d at 427–28; Brown v. Caraway, 719 F.3d 583, 586–88 (7th Cir. 2013).

Since 1996, our court has moved from near one end of this spectrum to the other. Early cases reserved whether § 2255(h)’s limits on successive filings could render § 2255 inadequate even if a new statutory decision proved a prisoner’s innocence. Charles v. Chandler, 180 F.3d 753, 756–58 (6th Cir. 1999) (per curiam). Eventually, though, we found that § 2255(h)’s limits could render § 2255 inadequate if the Supreme Court changed its interpretation of an offense and prisoners lacked a reasonable opportunity to advocate for that interpretation in their initial § 2255 motion. See Wright v. Spaulding, 939 F.3d 695, 702–03 (6th Cir. 2019) (discussing Martin v. Perez, 319 F.3d 799 (6th Cir. 2003)). We cautioned that this rule applied only when prisoners made “a claim of actual innocence,” Bannerman v. Snyder, 325 F.3d 722, 724 (6th Cir. 2003), not when they merely “challenge[d] their sentences,” United States v. Peterman, 249 F.3d 458, 462 (6th Cir. 2001). That actual-innocence limit gave way in 2016. In Hill, we held that a new Supreme Court decision as to sentencing could render § 2255(h)’s limits inadequate (and permit a habeas petition) if the decision showed a sentencing error that was “sufficiently grave to be deemed a miscarriage of justice or a fundamental defect.” 836 F.3d at 595. Yet Hill still requires prisoners to show that their first § 2255 motion did not give them a “reasonable opportunity” to advocate for the interpretation that the Supreme Court later accepted. Wright, 939 F.3d at 705.

C.

In this case, Ramon Hueso asks us to take a step beyond Hill by further contracting § 2255(h)’s limits on second motions and expanding § 2255(e)’s allowance for habeas petitions.

Hueso’s Criminal Case & § 2255 Motion. In October 2009, a district court in Alaska convicted Hueso of conspiring to distribute and possess with intent to distribute illegal drugs, in violation of 21 U.S.C. §§ 846 and 841(a). The default sentencing range for Hueso’s offense fell between 10 years and life imprisonment. 21 U.S.C. § 841(b)(1)(A) (2006). If, however, Hueso had “a prior conviction for a felony drug offense,” his mandatory-minimum sentence jumped from 10 years to 20 years. Id. Federal law defines “felony drug offense” as “an offense that is punishable by imprisonment for more than one year under any law of the United States or of a State or foreign country that prohibits or restricts conduct relating to” certain controlled substances. Id. § 802(44).

At Hueso’s January 2010 sentencing, the government asked for the 20-year mandatory minimum because the State of Washington had twice convicted Hueso of possessing illegal drugs. Wash. Rev. Code § 69.50.4013(1) (2004). Under the state statute, these offenses carried a maximum penalty of five years, which exceeded the one-year requirement in the federal definition of felony drug offense. See id. § 69.50.4013(2) (2004); id. § 9A.20.021(1)(c) (2004). Under the state sentencing guidelines, however, Hueso faced a maximum six-month term. Id. § 9.94A.505(2)(a)(i) (2002); 2002 Wash. Legis. Serv. Ch. 290, § 8 (S.S.H.B. 2338) (West). And he in fact received concurrent 40-day sentences. When the district court sentenced Hueso, Ninth Circuit precedent held that state convictions like Hueso’s were felony drug offenses so long as “the state’s statutory maximum sentence” exceeded one year, even if “the maximum sentence available under the state sentencing guidelines” did not. United States v. Rosales, 516 F.3d 749, 758 (9th Cir. 2008). Hueso did not contest this point, and the district court sentenced him to the longer 20-year mandatory minimum. The Ninth Circuit affirmed. United States v. Hueso, 420 F. App’x 776, 776 (9th Cir. 2011).

Hueso moved to vacate his sentence under § 2255 in the Alaska court that sentenced him. He again did not challenge his 20-year mandatory-minimum sentence. The court denied his motion, and the Ninth Circuit denied a certificate of appealability. United States v. Hueso, No. 11-35855, 2013 U.S. App. LEXIS 16488 (9th Cir. May 15, 2013).

Hueso’s § 2241 Petitions. In 2013, Hueso filed his first habeas petition under § 2241 in the Eastern District of Kentucky, the district in which he is confined. This petition challenged his 20-year sentence on the ground that his state convictions were not “felony drug offenses” and that his mandatory minimum should have been 10 years, not 20 years. Hueso relied on two statutory decisions. The first was Carachuri-Rosendo v. Holder, 560 U.S. 563 (2010). That case asked whether a lawful permanent resident who had pleaded no contest to a state’s misdemeanor drug-possession offense (after having pleaded guilty to an earlier drug-possession offense) had “been convicted of an ‘aggravated felony’” under the immigration laws. Id. at 566–70. Those laws define “aggravated felony,” among other ways, as incorporating the felonies in the Controlled Substances Act (or state equivalents). Id. at 567. That act, in turn, allows the government to treat drug possession as a federal “felony” if a defendant has a prior possession conviction. Id. at 567–68. The government argued that Carachuri-Rosendo’s misdemeanor offense was an “aggravated felony” because he had a prior possession conviction and could have been charged with a federal felony. The Court disagreed, holding that the statute did not cover those whose conduct “hypothetically” “could have received felony treatment under federal law.” Id. at 576.

The second (more relevant) decision came after the Supreme Court remanded a circuit case interpreting “felony drug offense” for reconsideration in light of Carachuri-Rosendo. United States v. Simmons, 649 F.3d 237, 239 (4th Cir. 2011) (en banc). Like the Ninth Circuit, the Fourth Circuit had read “felony drug offense” to cover a state crime with a statutory maximum exceeding one year, even if the defendant could not have been sentenced to more than a year under the sentencing guidelines. Id. at 241. But a divided en banc court relied on Carachuri-Rosendo to overrule its precedent. Id. Simmons held that courts must focus on the maximum penalties that a specific defendant faces (under the sentencing guidelines), not the maximum penalties that any defendant could face (under the statutory maximum). See id. at 243–50.

The district court denied Hueso’s first habeas petition. Because of our then-existing cases interpreting § 2255(e) in 2013, Hueso could not challenge sentencing errors in a habeas petition. Hueso v. Sepanek, No. 13-cv-19, 2013 WL 4017117, at *5 (E.D. Ky. Aug. 6, 2013).

In 2018, after our Hill decision expanded our interpretation of § 2255(e), Hueso filed a second habeas petition in the Eastern District of Kentucky. He again relied on Carachuri-Rosendo and Simmons. The district court again denied relief. It held (procedurally) that Hueso may not file a habeas petition to challenge this alleged sentencing error even after Hill and (substantively) that Hueso’s state convictions still qualified as felony drug offenses under Ninth Circuit precedent.

A new Ninth Circuit case from 2019 has since undercut the substantive portion of the district court’s opinion. United States v. Valencia-Mendoza, 912 F.3d 1215, 1219–24 (9th Cir. 2019). When considering a definition of “felony” like the definition of “felony drug offense,” Valencia-Mendoza agreed with the Fourth Circuit’s Simmons decision and overruled earlier Ninth Circuit cases as irreconcilable with Carachuri-Rosendo. Id. Valencia-Mendoza held that courts “must consider both a crime’s statutory elements and sentencing factors when determining whether an offense is ‘punishable’” by a term exceeding a year. Id. at 1224 (emphasis omitted).

II.

Hueso challenges his 20-year minimum sentence based on the Supreme Court’s Carachuri-Rosendo decision, the Fourth Circuit’s Simmons decision, and the Ninth Circuit’s Valencia-Mendoza decision. Understandably so. Simmons and Valencia-Mendoza both cite Carachuri-Rosendo to hold that statutory definitions like the definition of “felony drug offense” require courts to consider the maximum prison term that a specific defendant may receive under a state’s sentencing guidelines. But Hueso has already litigated one § 2255 motion, and Hueso’s cited cases do not permit him to file a second § 2255 motion because they do not contain new evidence of innocence or new constitutional rules. 28 U.S.C. § 2255(h)(1)–(2). Hueso instead argues that these cases make § 2255’s remedy “inadequate or ineffective” within the meaning of § 2255(e), which, he says, allows him to pursue this sentencing claim in a § 2241 habeas petition.

At the outset, we clarify what we do and do not decide. We do not opine on the merits of Hueso’s claim that his state convictions are not “felony drug offenses.” “[D]iverse viewpoints” exist on this “difficult question.” See United States v. Recinos-Hernandez, 772 F. App’x 115, 117 (5th Cir. 2019) (per curiam); Simmons, 649 F.3d at 250–58 (Agee, J., dissenting).

We likewise do not opine on whether Hueso’s sentencing challenge alleges the type of statutory error that can be asserted on collateral review at all (even in a timely filed first § 2255 motion). The Supreme Court has held that, unlike constitutional or jurisdictional errors, statutory errors do “not provide a basis for collateral attack unless the claimed error constituted ‘a fundamental defect which inherently results in a complete miscarriage of justice.’” United States v. Addonizio, 442 U.S. 178, 185 (1979) (quoting Hill v. United States, 368 U.S. 424, 428 (1962)). The circuit courts have reached different conclusions about the types of sentencing errors that rise to this fundamental-defect level. See Snider v. United States, 908 F.3d 183, 192 n.5 (6th Cir. 2018). All agree that prisoners may challenge statutory errors resulting in a sentence that “was in excess of the maximum authorized by law.” 28 U.S.C. § 2255(a); see, e.g., Snider, 908 F.3d at 189. And we have held that the relevant “law” includes the once-mandatory sentencing guidelines, which “had the force and effect of law.” Hill, 836 F.3d at 599 (citation omitted). We thus allowed a prisoner to challenge an unlawful career-offender enhancement that resulted in a sentence (300 months) exceeding the mandatory guidelines range that would apply without the enhancement (235 to 293 months). Id. at 593, 599. Yet we have since held that this logic does not extend to the now-advisory guidelines. Bullard v. United States, 937 F.3d 654, 658–61 (6th Cir. 2019). Here, Hueso was sentenced under the advisory guidelines. And while he argues that his mandatory-minimum sentence was wrongly increased, his 20-year sentence still fell below what would have been the maximum allowed by law (life imprisonment) even without the challenged enhancement. This type of challenge to a mandatory-minimum enhancement thus may or may not be cognizable in a properly filed first motion under § 2255. We need not decide this issue.

Rather, we decide only that Hueso cannot pursue this type of claim in habeas under § 2241 because his cited cases do not render a § 2255 motion “inadequate or ineffective” within the meaning of § 2255(e). We reach that result for two reasons. His two circuit decisions cannot, as a matter of law, establish § 2255’s inadequacy. And his lone Supreme Court decision issued when his direct appeal was pending, so he could have cited it in the ordinary course.

A.

We have adopted a “reasonable-opportunity standard” for prisoners seeking to prove § 2255’s inadequacy under § 2255(e). Wright, 939 F.3d at 703. Under this standard, prisoners must show “that binding adverse precedent (or some greater obstacle) left [them] with ‘no reasonable opportunity’ to” assert their legal claim on the front end—in their initial § 2255 motion. Id. (quoting In re Davenport, 147 F.3d 605, 610 (7th Cir. 1998)). But we have yet to identify the kind of statutory changes required on the back end—when § 2255(h) bars another motion. Does a new statutory decision from a circuit court suffice to show § 2255’s inadequacy?

Our cases suggest not. They have all involved a new Supreme Court decision. In Hill, we “reiterate[d]” that we “addresse[d] only a narrow subset of § 2241 petitions“—those involving a “a subsequent, retroactive change in statutory interpretation by the Supreme Court[.]” 836 F.3d at 599–600. And our “actual-innocence” cases likewise have all been “based upon Supreme Court decisions announcing new rules of statutory construction unavailable for attack under section 2255.” Hayes v. Holland, 473 F. App’x 501, 501–02 (6th Cir. 2012) (per curiam); see, e.g., Harrington v. Ormond, 900 F.3d 246, 248–50 (6th Cir. 2018); Wooten v. Cauley, 677 F.3d 303, 306 (6th Cir. 2012); Martin, 319 F.3d at 804–05. We now turn the facts of our prior cases into a holding. In addition to whatever else our reasonable-opportunity standard demands, it requires a Supreme Court decision that adopts a new interpretation of a statute after the completion of the initial § 2255 proceedings. This requirement follows both from § 2255’s text and structure and from the backdrop against which Congress enacted § 2255(h)’s limits in 1996.

1. Text and Structure

In recent decades, the Supreme Court has insisted that statutory interpretation depends on statutory text. E.g., Ross v. Blake, 136 S. Ct. 1850, 1856 (2016). Its approach “calls on the judicial interpreter to consider the entire text, in view of its structure and of the physical and logical relation of its many parts.” Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 167 (2012); Star Athletica, L.L.C. v. Varsity Brands, Inc., 137 S. Ct. 1002, 1010 (2017).

Following these instructions, we begin with § 2255(e)’s language:

An application for a writ of habeas corpus in behalf of a prisoner who is authorized to apply for relief by motion pursuant to this section, shall not be entertained if it appears that the applicant has failed to apply for relief, by motion, to the court which sentenced him, or that such court has denied him relief, unless it also appears that the remedy by motion is inadequate or ineffective to test the legality of his detention.

28 U.S.C. § 2255(e). Under this text, prisoners who have had a § 2255 motion “denied” must show that the statute’s remedy (that is, its “means” for seeking relief) is inadequate or ineffective (that is, “insufficient” or “incapable“) to test (that is, “try“) a legal challenge to their detention. See, e.g., Bouvier’s Law Dictionary (1934) (“remedy“); Webster’s New Int’l Dictionary 1254, 1271 (2d. ed. 1934) (“inadequate” and “ineffective“); 11 Oxford English Dictionary 220 (1st ed. 1933) (“test“). Notably, this statutory language does not give courts license to ask whether they consider § 2255 to be inadequate or ineffective in the abstract. Rather, “the inadequacy or ineffectiveness must relate specifically to a procedural deficiency in ‘test(ing) the legality of [a prisoner’s] detention.’” Cain v. Markley, 347 F.2d 408, 410 (7th Cir. 1965). So § 2255(e) asks only whether § 2255’s motion is sufficient to assert a claim on the merits; it does not guarantee “success” on the merits. Litterio, 369 F.2d at 396.

This plain language readily explains why, as a general matter, § 2255(h)’s limits on successive motions do not render § 2255’s remedy inadequate or ineffective under § 2255(e). As all courts agree, the initial § 2255 motion provides a sufficient means to test a prisoner’s claims even though § 2255(h)’s limits bar the prisoner from succeeding on any later-filed claims. See Hill, 836 F.3d at 594; Davenport, 147 F.3d at 608; Dorsainvil, 119 F.3d at 251.

This plain language might also suggest the answer to the question presented here: Should an initial § 2255 motion be considered “inadequate or ineffective” under § 2255(e) if the precedent existing at the time of that motion bars a prisoner from succeeding on a claim? The answer is “no” according to those courts that distinguish between testing a claim and succeeding on a claim. See Prost, 636 F.3d at 590; McCarthan, 851 F.3d at 1085–95. Mistaken precedent may prove that the “circuit law is inadequate,” but it does nothing to prove that “the § 2255 remedial vehicle is inadequate or ineffective to the task of testing the argument.” Prost, 636 F.3d at 590.

Our court, by contrast, has followed a less textual path. We have shaped and reshaped our standards interpreting the phrase “inadequate or ineffective” as if we had the authority to decide this question in common-law fashion. We first required a “showing of ‘actual innocence.’” Martin, 319 F.3d at 804. We then molded that requirement into a multi-part test. Wooten, 677 F.3d at 307–08. We then adopted different tests for different claims. Hill, 836 F.3d at 594–95.

Yet we do not read our cases as jettisoning § 2255’s text altogether. “[S]uch a casual disregard of the rules of statutory interpretation” would put us on a collision course with the Supreme Court’s teaching that “a court’s proper starting point lies in a careful examination of the ordinary meaning and structure of the law itself.” Food Mktg. Inst. v. Argus Leader Media, 139 S. Ct. 2356, 2364 (2019). This view of our cases would also render Congress’s 1996 limits on successive filings superfluous. Before then, courts had been engaging in precisely that sort of common-law development of the limits on multiple filings. Cf. McCleskey, 499 U.S. at 489. But Congress toppled this judicial framework with § 2255(h). Section 2255(e) should not be read as impliedly resurrecting it—in the court of confinement rather than the sentencing court, no less. That view of our cases would undermine both Congress’s 1996 law and its 1948 law.

We instead implement our “reasonable-opportunity standard” in a manner that best reconciles it with § 2255’s text and structure as a whole. When assessed from that view, our standard must require a new Supreme Court decision after the initial § 2255 motion. The Supreme Court has explained that the “[i]nterpretation of a phrase of uncertain reach“—as we have interpreted the phrase “inadequate or ineffective” to be—“is not confined to a single sentence when the text of the whole statute gives instruction as to its meaning.” Star Athletica, 137 S. Ct. at 1010 (quoting Maracich v. Spears, 570 U.S. 48, 65 (2013)). Here, for example, no reasonable interpreter would say that § 2255(f)’s one-year statute of limitations makes § 2255’s remedy “inadequate or ineffective” whenever a prisoner blows that deadline. That reading would obliterate this statute of limitations because prisoners could always bypass the time limit for filing a § 2255 motion by substituting a § 2241 petition in its place. Wooten, 677 F.3d at 307. And “there can be no justification for needlessly rendering [these two provisions] in conflict if they can be interpreted harmoniously.” Scalia & Garner, supra, at 180.

This harmonious-reading rule compels the conclusion that § 2255’s remedy cannot be viewed as “inadequate or ineffective” unless a prisoner identifies a new Supreme Court decision. Without that rule, we would interpret § 2255(e) to abolish two other limits in § 2255. To begin with, § 2255(h)(2) tells courts that new constitutional rules must come from “the Supreme Court” for those rules to authorize another round of § 2255 litigation. But if we held that new statutory rules from the circuit courts sufficed to render § 2255’s remedy “inadequate or ineffective,” prisoners could assert new statutory claims (under § 2241) more easily than they could assert new constitutional claims (under § 2255(h)(2)). That reading generates an “odd” dichotomy. Chazen v. Marske, 938 F.3d 851, 864 (7th Cir. 2019) (Barrett, J., concurring). Indeed, whatever the phrase “inadequate or ineffective” means, it cannot be read to distinguish statutory claims from constitutional claims. So if new statutory rules from the circuit courts could trigger § 2255(e) because prisoners cannot vindicate the new rules under § 2255, new constitutional rules from the circuit courts likewise would have to trigger § 2255(e) because prisoners also could not vindicate them under § 2255. But that interpretation altogether erases § 2255(h)(2)’s requirement of a new Supreme Court decision for successive litigation.

In addition, § 2255(f) postpones the start date for the running of its one-year time limit for first § 2255 motions if prisoners rely on new statutory rights, not just new constitutional rights. See Jamieson v. United States, 692 F.3d 435, 439 (6th Cir. 2012) (per curiam); Prost, 636 F.3d at 585–86. The statute does so by tying the running of this limitations period to “the date on which the right asserted was initially recognized by the Supreme Court, if that right has been newly recognized by the Supreme Court and made retroactively applicable to cases on collateral review[.]” 28 U.S.C. § 2255(f)(3). So a circuit change in interpretation (unlike a Supreme Court change) does not trigger the delayed start date for initial § 2255 motions. Rather, § 2255(f) treats § 2255 motions as untimely when they are based on circuit cases that arise one year after a final judgment (the default start date for the limitations period). If, however, we allowed prisoners to use those same circuit cases to proceed through the habeas backdoor—which has no statute of limitations, see Wooten, 677 F.3d at 306—then the “Supreme Court” limit in § 2255(f)(3) is no limit at all. Under this view, prisoners who could not satisfy that Supreme Court requirement could simply file a § 2241 petition instead.

In sum, we must read our standards in a way that harmonizes § 2255(e) with § 2255(f)(3) and § 2255(h)(2). This view of the section as a whole shows that later circuit decisions do not suffice to prove that an initial § 2255 motion is inadequate or ineffective to test a claim.

2. Statutory History

Apart from § 2255’s text and structure, we must read the current § 2255 against the backdrop of its “statutory history—the statutes repealed or amended by the statute under consideration.” Scalia & Garner, supra, at 256; see, e.g., United States v. Wong Kim Ark, 169 U.S. 649, 653–54 (1898). This statutory history, unlike legislative history, itself reflects the constitutional bicameralism and presentment process and thus is vastly more difficult to manipulate than legislative history. Cf. Scalia & Garner, supra, at 256. For two reasons, this historical look confirms that Congress’s 1996 limits on successive motions in § 2255(h) cannot render § 2255 “inadequate or ineffective” without, at the least, a new Supreme Court precedent.

Reason One: Pre-1996 Limits. As noted, courts had been imposing judicially developed limits on multiple filings before Congress adopted § 2255(h)’s statutory limits. See Schlup, 513 U.S. at 318 & n.34. And, as far as we are aware, no court found that those judicial limits rendered § 2255 inadequate when they precluded new filings—even though § 2255(e)’s language dates back to 1948. Cf. Richards, 5 F.3d at 1370–72. The “cause” exception to those limits is instructive here because it adopted a test similar to our “reasonable-opportunity standard.” Wright, 939 F.3d at 703. This exception allowed prisoners to file another collateral challenge only if a new claim had not been “reasonably available” in prior challenges. Murray v. Carrier, 477 U.S. 478, 488 (1986). If, before 1996, a claim was considered “reasonably available” under these cause rules (and so barred by the judicial limits on successive filings), id., it would make little sense to find today that a prisoner lacked a “reasonable opportunity” to assert that claim under § 2255(e). We should not read Congress’s addition of § 2255(h)—which strengthened the limits on successive motions—as paradoxically relaxing those limits by expanding the meaning of § 2255(e)’s words.

And notably, when prisoners sought to establish “cause” with a new precedent, the Supreme Court’s cases suggested that nothing but a new Supreme Court precedent would do. The new precedent must have departed so dramatically from prior cases that it could be said to have created a new claim that had not been previously “available.” Smith v. Murray, 477 U.S. 527, 537 (1986). The Court “identified several circumstances in which a new rule representing a clear break with the past” might satisfy this test. 1 Brian R. Means, Postconviction Remedies § 24:15, at 891–92 (2019). Each circumstance contemplated a new Supreme Court decision: The Court might “explicitly overrule one of its precedents“; it might “overturn a longstanding and widespread practice” that had been adopted by “a near-unanimous body of lower court authority“; or it might “disapprove of a practice the Court arguably had sanctioned in prior cases.” Id. (discussing Ross, 468 U.S. at 17–18). If, by contrast, a circuit’s decisions precluded a claim—that is, if the claim was “unacceptable to that particular court at that particular time,” Engle v. Isaac, 456 U.S. 107, 130 n.35 (1982) (citation omitted)—the prisoner could not establish “cause” for failing to raise the claim from that fact alone. See Bousley, 523 U.S. at 623–24. Yet if we now held that this same circuit precedent could show § 2255’s inadequacy under § 2255(e) (and allow a prisoner to pursue a habeas claim under § 2241), our conclusion would have rendered meaningless these preexisting “cause” requirements.

Indeed, we would upend more than just the preexisting limits governing successive filings (which have now been superseded by § 2255(h)). The Supreme Court’s “cause” rules continue to apply outside the context of successive petitions. If, for example, a prisoner fails to raise a claim on direct appeal that the prisoner seeks to litigate in a § 2255 motion, courts will consider the claim procedurally defaulted unless the prisoner shows cause and prejudice (or actual innocence). E.g., United States v. Frady, 456 U.S. 152, 166–68 (1982). Because changes in circuit precedent do not generally suffice to show cause, see, e.g., Damon v. United States, 732 F.3d 1, 4–5 (1st Cir. 2013), those same changes should not suffice to show that § 2255 is inadequate.

Reason Two: Choice-of-Law Problems. The problems that Congress sought to fix in 1948 confirm the need for a new Supreme Court decision. By placing a prisoner’s collateral challenge in the same court that sentenced the prisoner, § 2255 ended “the unseemly spectacle of federal district courts trying the regularity of proceedings had in courts of coordinate jurisdiction.” John J. Parker, Limiting the Abuse of Habeas Corpus, 8 F.R.D. 171, 172–73 (1949). The rule that only a later Supreme Court decision may trigger § 2255(e)’s authorization for new habeas filings at least lessens the potential friction between the sentencing court (which hears the § 2255 motion) and the court of confinement (which hears the § 2241 petition). Any new decision from the Supreme Court binds both courts. So when considering a § 2241 petition relying on that new decision, the court of confinement need not “grade” the opinions of the sentencing circuit. Its marching orders will have come from the Supreme Court.

If we held that circuit decisions could prove § 2255’s inadequacy, by contrast, the statute would retain some of the “difficulties that had arisen in administering the habeas corpus jurisdiction of the federal courts” before 1948. Hayman, 342 U.S. at 219. Prisoners often may be sentenced within one circuit (whose law would apply to a § 2255 motion), but confined in another (whose law would apply to a § 2241 petition). What happens if the circuit precedent in the court of confinement is more favorable to the prisoner’s claim than the circuit precedent in the sentencing court? Even before 1996, prisoners suggested that the sentencing court’s less favorable caselaw rendered a § 2255 motion inadequate and allowed resort to habeas in the court of confinement. If true, some courts reasoned, “a prisoner would have the right in every instance to retest the legality of his detention” “because of a possible application of different legal principles by the court confronted with a habeas corpus petition and the court which has already ruled on a section 2255 motion[.]” Cain, 347 F.2d at 410. That is, a rule allowing circuit decisions to trigger § 2255(e) would give prisoners the right to “shop around the country in the hope of receiving a more favorable ruling.” Early v. Lamanna, No. 98-3892, 1999 WL 435156, at *5 (6th Cir. June 17, 1999); 3 Charles A. Wright & Sarah N. Welling, Federal Practice & Procedure § 623, at 648 & n.26 (4th ed. 2011).

At the least, this rule would trigger a difficult “choice-of-law question.” Chazen, 938 F.3d at 864 (Barrett, J., concurring). When a court of confinement hears a claim in a § 2241 petition, should it apply its precedent or the circuit precedent from the sentencing court? On the one hand, a rule following the sentencing court’s law “ensures that the law that prevails in the judicial circuit of any federal prisoner’s conviction, or a substantially similar law, is the law that will be applied to the prisoner’s § 2241 petition seeking vacation of a conviction.” Id. at 865 (citation omitted). It also eliminates the arbitrary nature of tying the law that governs a § 2241 petition to the location in which the prisoner happens to be detained. Id. On the other hand, a rule following the court of confinement’s law comports with the background norm that each court should apply its own precedent on the meaning of federal law. E.g., AER Advisors, Inc. v. Fidelity Brokerage Servs., LLC, 921 F.3d 282, 288 n.5 (1st Cir. 2019). As then-Judge Ginsburg said, “because there is ultimately a single proper interpretation of federal law, the attempt to ascertain and apply diverse circuit interpretations simultaneously is inherently self-contradictory.” In re Korean Air Lines Disaster of Sept. 1, 1983, 829 F.2d 1171, 1175 (D.C. Cir. 1987). And while Congress could order a departure from that default principle, it did not do so in 1948. It fixed the problem of one circuit assessing another’s work by changing the venue, not the choice-of-law rules.

This case features all of these problems. Hueso was sentenced in Alaska, so Ninth Circuit precedent governed his direct appeal and § 2255 proceedings. He also relies heavily on the Fourth Circuit’s Simmons decision, even though he was not convicted there and is not confined there. Yet he is now imprisoned in Kentucky, so Sixth Circuit precedent would normally govern. But our cases point in opposite directions on the merits of Hueso’s sentencing claim. Compare United States v. Rockymore, 909 F.3d 167, 170–71 (6th Cir. 2018), with United States v. Montgomery, 893 F.3d 935, 940–41 (6th Cir. 2018). And what if we were to disagree with Simmons and Valencia-Mendoza? Reasonable minds might conclude that the Supreme Court’s reading of “an unrelated, civil statutory scheme[]” in Carachuri-Rosendo “offers [no] clear direction” in this context. Simmons, 649 F.3d at 250 (Duncan, J., dissenting); see id. at 250–58 (Agee, J., dissenting). A requirement that the new interpretation spring from the Supreme Court, with its power to impose a uniform interpretation among the circuits, avoids these inter-circuit feuds—the same type of feuds that Congress sought to eliminate in 1948.

3. Competing Views

Neither the Fourth Circuit’s decision in Wheeler nor the dissent in this case persuades us to take a different view. The Fourth Circuit found “no need to read [§ 2255(e)’s] savings clause as dependent only on a change in Supreme Court law” because its text (unlike § 2255(h)(2)’s text) does not identify a new Supreme Court case as a prerequisite for § 2255 to be considered “inadequate or ineffective.” Wheeler, 886 F.3d at 428–29. True enough. But we still must decide what § 2255(e)’s general “inadequate or ineffective” language means. Even if our precedent requires us to overlook that this general phrase is connected to the words “to test,” we still should interpret the phrase in the way that best comports with the rest of § 2255. See Star Athletica, 137 S. Ct. at 1010. And § 2255(h)(2) requires a new Supreme Court decision before triggering the right to a second or successive § 2255 motion. We would read this limit right out of the statute if we held that the limit also rendered § 2255 “inadequate or ineffective” under § 2255(e) so as to permit the filing of a § 2241 petition instead.

The Fourth Circuit, by comparison, thought that § 2255(h)(2)’s requirement of a new Supreme Court decision “cut[] the other way.” Wheeler, 886 F.3d at 428. It said: “Congress could have made savings clause relief dependent only on changes in Supreme Court constitutional law by using the identical language in § 2255(e), but it did not.” Id. at 428–29. The court thus replaced the harmonious-reading canon with a “discordant-reading” canon. Under its logic, Congress also could have made savings clause relief dependent only on a federal prisoner’s compliance with § 2255(f)’s time limits. Does its failure to do so mean that courts may hold that those time limits render § 2255 “inadequate or ineffective” under § 2255(e)? We think not. Courts should read § 2255(e)’s general “inadequate or ineffective” phrase in a manner that harmonizes it with the rest of § 2255, not in a manner that creates conflicts in the statutory scheme. For centuries, courts have recognized that “[i]f any section [of a law] be intricate, obscure, or doubtful, the proper mode of discovering its true meaning is by comparing it with the other sections, and finding out the sense of one clause by the words or obvious intent of the other.” Scalia & Garner, supra, at 167 (quoting 1 Edward Coke, The First Part of the Institutes of the Laws of England, or a Commentary upon Littleton § 728, at 381a (1628; 14th ed. 1791)). The Fourth Circuit failed to do so.

The Fourth Circuit also said that § 2255(e) applies to those who have already been “denied” relief under § 2255, so Congress meant for it to cover “those prisoners filing a successive § 2255 motion.” Wheeler, 886 F.3d at 429. We do not dispute that § 2255(e)’s language covers those who have had one § 2255 motion denied if they show that § 2255’s remedy is inadequate or ineffective. But we fail to see how that fact favors the Fourth Circuit’s conclusion that they may rely on new circuit cases over our conclusion that they must rely on new Supreme Court cases.

To its credit, the dissent in this case does confront our concern that interpreting § 2255(e)’s “inadequate or ineffective” language to include statutory changes from the circuit courts would treat the limits in § 2255(f)(3) and (h)(2) as a nullity. The dissent says those limits would bar most untimely or successive § 2255 motions (and so would not be superfluous) because only a narrow subset of circuit changes would trigger § 2255(e): those showing statutory errors that are “sufficiently grave to be deemed a miscarriage of justice or a fundamental defect.” Dissent at 30 (quoting Hill, 836 F.3d at 595). This response confirms the textual problem with the dissent’s view. As noted, the Supreme Court has long held that a statutory error “does not provide a basis for collateral attack“—whether under § 2241, § 2254, or § 2255—“unless the claimed error constituted ‘a fundamental defect which inherently results in a complete miscarriage of justice.’” Addonizio, 442 U.S. at 186 (quoting Hill, 368 U.S. at 428); see Reed v. Farley, 512 U.S. 339, 353–55 (1994). Well before Congress’s 1996 amendments, a federal prisoner needed to satisfy this fundamental-defect requirement when raising statutory errors in collateral attacks. After those amendments, a prisoner still must satisfy this fundamental-defect element even if the prisoner files a timely first-in-time § 2255 motion. If satisfying that preexisting mandate also automatically overcomes Congress’s second-or-successive limits in § 2255(h) or its time limits in § 2255(f), those limits do no additional work. We refuse to read the limits out of the statute in this way.

Again to its credit, the dissent recognizes the incongruity of allowing a prisoner to raise a successive statutory claim based on a new circuit decision under § 2241 when the prisoner must rely on a new Supreme Court decision to raise a successive constitutional claim under § 2255(h)(2). Dissent at 31–32. The dissent reconciles this oddity not by requiring new Supreme Court decisions for both statutory and constitutional claims (as we do), but by suggesting that even new constitutional circuit decisions can render § 2255 “inadequate or ineffective” under § 2255(e). We are not aware of any other opinion that has suggested that a new constitutional decision by a circuit court could render § 2255 inadequate or ineffective and trigger another round of litigation under § 2241. Yet, as the dissent rightly notes, that is the logical implication of Wheeler‘s holding, despite § 2255(h)(2)‘s plain text requiring “a new rule of constitutional law, made retroactive to cases on collateral review by the Supreme Court,” for successive litigation under § 2255. For this reason, too, we think that our reading best implements § 2255‘s text and structure.

In sum, Hueso must identify a new Supreme Court decision to show that § 2255‘s remedy is inadequate or ineffective. He thus may not rely on the Ninth Circuit‘s decision in Valencia-Mendoza or the Fourth Circuit‘s decision in Simmons as the basis for his habeas claim.

B.

That leaves Hueso with one Supreme Court precedent—Carachuri-Rosendo—in support of his view that § 2255‘s remedy is inadequate to test his sentencing claim. His reliance on this case fails for a timing reason: The Supreme Court decided Carachuri-Rosendo in June 2010. See id. at 563. Carachuri-Rosendo thus issued well before Hueso filed his § 2255 motion in June 2011. Indeed, Hueso could have cited Carachuri-Rosendo on the direct appeal in his criminal case, which was decided in March 2011. See Hueso, 420 F. App‘x at 776. Because Hueso faced no “obstacle” to citing this Supreme Court decision in support of his sentencing claim in earlier proceedings, he had a “reasonable opportunity” to rely on the decision. Wright, 939 F.3d at 703 (citation omitted). This decision thus does not establish his initial § 2255 motion‘s inadequacy.

In response, Hueso argues that he could not have relied on Carachuri-Rosendo on direct appeal or in his initial § 2255 motion because of the then-existing Ninth Circuit precedent that foreclosed his claim. Yet the “new legal arguments,” see Wright, 939 F.3d at 705, that Carachuri-Rosendo generated included the argument that later won in the Ninth Circuit—that this new decision overruled prior circuit precedent interpreting terms like “felony drug offense,” see Valencia-Mendoza, 912 F.3d at 1219. Indeed, if Hueso had timely raised his claim, the Ninth Circuit may well have found that its earlier decisions conflicted with this “higher intervening authority” in his case, not in Valencia-Mendoza‘s case. Id.

* * *

Disagreements among the circuit courts often create the potential for the unequal treatment of similarly situated parties. Here, Gerald Wheeler could pursue in the Fourth Circuit the habeas claim that Ramon Hueso may not raise in the Sixth. A federal prisoner‘s ability to seek habeas relief under § 2241 should not depend on where the executive branch opts to confine him. But a circuit conflict would exist over the meaning of § 2255(e) no matter how we resolved Hueso‘s case. And we must follow the approach that stays true to the text and structure of § 2255, not one that implements our own personal views of justice and fairness. We affirm.

---

DISSENT

---

KAREN NELSON MOORE, Circuit Judge, dissenting. In 2009, Ramon Hueso was convicted in the District of Alaska of one count of unlawfully and knowingly conspiring with another to distribute and possess with intent to distribute 50 grams or more of a mixture or substance containing actual methamphetamine. See Hueso v. Barnhart, No. 6:18-176, 2018 WL 6172513, at *2 (E.D. Ky. Nov. 26, 2018). Hueso had been convicted twice before of drug offenses in the State of Washington, so he was subject to a mandatory-minimum sentence of 20 years. Id. Had Hueso been convicted in Alaska of the same federal offense and with the same criminal record in 2019, however, his mandatory-minimum sentence would have been 10 years. This discrepancy is the result of the Ninth Circuit changing its law on how to determine whether a prior offense is a “felony drug offense.” See United States v. Valencia-Mendoza, 912 F.3d 1215, 1224 (9th Cir. 2019). Hueso, who is now incarcerated within this circuit, seeks relief under 28 U.S.C. § 2241. If he were collaterally attacking his sentence for the first time, he would almost certainly prevail based on the intervening change in law. The only reason he will remain in prison for another decade is because he has already sought habeas relief once, and did not prevail the first time. He should not be foreclosed from seeking such relief, and therefore I dissent.

I.

A jury convicted Hueso of one count of unlawfully and knowingly conspiring with another to distribute and possess with intent to distribute 50 grams or more of a mixture or substance containing actual methamphetamine, in violation of 21 U.S.C. §§ 846 and 841(b)(1)(A). Hueso, 2018 WL 6172513, at *2. Because the government gave notice under 21 U.S.C. § 851 that Hueso had two prior felony drug convictions, he was subject to the enhanced penalties of § 841(b)(1)(A): namely, the then-applicable 20-year mandatory minimum. Id.

Whether one of Hueso‘s prior drug convictions was properly categorized as a “felony drug offense” is at the heart of this case. “A ‘felony drug offense’ is defined as ‘an offense that is punishable by imprisonment for more than one year under any law of . . . a State . . . that prohibits or restricts conduct relating to narcotic drugs.‘” United States v. Rosales, 516 F.3d 749, 758 (9th Cir. 2008) (quoting 21 U.S.C. § 802(44)). The statutory maximum under Washington law for both of Hueso‘s prior drug convictions was five years. Hueso, 2018 WL 6172513, at *5. Under Ninth Circuit law at the time of Hueso‘s federal conviction, “[t]o determine whether a state ‘felony drug offense’ is punishable by more than one year,” a court would “look to the state‘s statutory maximum sentence and not the maximum sentence available under the state sentencing guidelines.” Rosales, 516 F.3d at 758. Thus, Hueso‘s prior convictions qualified as felony drug offenses, and he was subject to the enhanced federal mandatory minimum.

But Hueso was not actually subject to imprisonment for a year or longer for either of his prior Washington convictions. R. 1-2 (Pet. App‘x at 2) (Page ID #30); see Appellant Br. at 17–18. The maximum sentence under Washington‘s sentencing guidelines to which Hueso was subject was six months. R. 1-2 (Pet. App‘x at 2) (Page ID #30); see Appellant Br. at 17–18. At the time of Hueso‘s federal conviction, Rosales was the law of the Ninth Circuit, and so the statutory maximum—as opposed to the maximum under the sentencing guidelines—controlled. The law of the Ninth Circuit has since changed. In 2019, the Ninth Circuit held in United States v. Valencia-Mendoza that “courts must consider both a crime‘s statutory elements and sentencing factors when determining whether an offense is ‘punishable’ by a certain term of imprisonment.” 912 F.3d 1215, 1224 (9th Cir. 2019). Except in certain circumstances that existed neither in Valencia-Mendoza nor in this case, Washington‘s sentencing guidelines are binding, and therefore it is the prescribed guidelines range—not the statutory maximum—that must be considered when determining whether a prior conviction was a felony offense. This is the basis of Hueso‘s § 2241 petition.

A challenge to the validity of a federal conviction or sentence is generally brought as a motion pursuant to § 2255, whereas a petition concerning the manner or execution of a sentence is appropriate under § 2241. Hill v. Masters, 836 F.3d 591, 594 (6th Cir. 2016). A prisoner‘s ability to access § 2255 relief is limited, however, when the movant is filing a “second or successive” motion. Such a motion is barred unless it “(1) contains newly discovered evidence that ‘would be sufficient to establish by clear and convincing evidence that no reasonable factfinder would have found the movant guilty,’ or (2) is based on ‘a new rule of constitutional law, made retroactive to cases on collateral review by the Supreme Court, that was previously unavailable.‘” Id. (quoting 28 U.S.C. § 2255(h)). Yet § 2255 has an escape valve of sorts, which is known as the savings clause. This clause allows some prisoners to file their claims under § 2241, free from § 2255‘s restrictions on second or successive motions. 28 U.S.C. § 2255(e).

The savings clause says:

An application for a writ of habeas corpus in behalf of a prisoner who is authorized to apply for relief by motion pursuant to [§ 2255], shall not be entertained if it appears that the applicant has failed to apply for relief, by motion, to the court which sentenced him, or that such court has denied him relief, unless it also appears that the remedy by motion is inadequate or ineffective to test the legality of his detention.

Id.

Until 2016, this court allowed prisoners to use the savings clause to challenge their convictions only when they identified retroactively applicable changes in statutory law that established their innocence. See United States v. Peterman, 249 F.3d 458, 462 (6th Cir. 2001). That changed, however, with Hill. Hill established that, in some limited circumstances, a prisoner can use the savings clause to bring a § 2241 petition to challenge a sentence, even though the prisoner is not asserting that they should not have been convicted of the underlying crime. In Hill, the prisoner was sentenced under the pre-United States v. Booker, 543 U.S. 220 (2005), mandatory guidelines and received a sentencing enhancement as a career offender. Hill, 836 F.3d at 592–94. After Hill‘s conviction, direct appeal, and first § 2255 motion, the Supreme Court issued Descamps v. United States, 570 U.S. 254 (2013), which directed the lower courts to use the categorical approach when the statute at issue consists of a single set of elements that define the crime “more broadly than the generic offense.” Descamps, 570 U.S. at 261. The Fourth Circuit—the jurisdiction in which Hill was convicted and sentenced—applied Descamps to Hill‘s crime of conviction and determined that it was not a violent felony. Hill, 836 F.3d at 595. Thus, if Hill had been convicted and sentenced post-Descamps, he would not have been subject to the mandatory career-offender sentencing enhancement. But because Descamps dealt with statutory interpretation rather than constitutional rights, Hill could not have brought a second motion under § 2255. This court held that Hill had access to relief through the savings clause and was able to file a § 2241 petition. The circumstances in which a prisoner could proceed as Hill did require (1) a case of statutory interpretation, (2) that is retroactively applicable and “could not have been invoked in [the] original petition,” and (3) an error “sufficiently grave to be ‘deemed a miscarriage of justice,’ or, put another way, a ‘fundamental defect’ corrigible in a habeas corpus proceeding.” Id. at 598 (quoting Brown v. Caraway, 719 F.3d 583, 586 (7th Cir. 2013)). The question before us here is whether the “case of statutory interpretation” may come from a circuit court, rather than the Supreme Court.

II.

Before delving into this question, the majority states at the outset of its opinion that it will not opine on “whether Hueso‘s sentencing challenge alleges the type of statutory error that can be asserted on collateral review at all,” given that “his 20-year sentence . . . fell below what would have been the maximum allowed by law (life imprisonment) even without the challenged enhancement.” Maj. Op. at 10. Indeed, there is no need for further opinion on whether such a challenge is cognizable, because we have already held that it is. In Hill, we considered “whether § 2241 may also support a challenge to a misapplied enhancement resulting in a sentence that does not exceed the statutory maximum,” and held that it could. 836 F.3d at 596–97. Specifically, we held that because Hill was sentenced under the mandatory sentencing guidelines pre-Booker, he could collaterally attack his sentence even though it was lower than the statutory maximum. 836 F.3d at 599 (“Serving a sentence imposed under mandatory guidelines (subsequently lowered by retroactive Supreme Court precedent) shares similarities with serving a sentence imposed above the statutory maximum. Both sentences are beyond what is called for by law, and both raise a fundamental fairness issue.“) (citation omitted). Like Hill, Hueso was sentenced pursuant to a mandatory framework that did not allow the district court to exercise discretion in varying downward, resulting in a sentence that is “beyond what is called for by law“; in Hill‘s case, this framework was a mandatory sentencing range, and in Hueso‘s case, it was a mandatory minimum. Although the sentencing courts in these cases “might have imposed [] sentence[s] equally as harsh as th[ose] mandated by [law],” to deny a claim based on this “frail conjecture” would constitute an “arbitrary disregard of the petitioner‘s right to liberty.” Hicks v. Oklahoma, 447 U.S. 343, 346 (1980).

On this point, the majority cites Bullard v. United States, 937 F.3d 654, 659 (6th Cir. 2019), perhaps in an attempt to find some ambiguity in Hill‘s holding. Bullard, however, only further highlights the problem Hill identified with forcing prisoners to bear sentences that are no longer mandatory. In Bullard, the prisoner argued that “the district court misclassified him as a career offender, which resulted in a higher recommended sentence.” 937 F.3d at 657. Yet Bullard was sentenced pursuant to the post-Booker, advisory guidelines, unlike Hill, whom we permitted to mount a collateral attack on his sentence because it was rendered pursuant to the pre-Booker, mandatory guidelines. The panel made clear in Bullard that the advisory nature of the guidelines was central to its decision: “[A] misapplication-of-an-advisory-guidelines-range claim is . . . not cognizable under § 2255.” Id. at 660 (quoting Snider v. United States, 908 F.3d 183, 191 (6th Cir. 2018), cert. denied, 139 S. Ct. 1573 (2019)). The problem the panel had with Bullard‘s claim was not that his resulting sentence would still be within the statutory maximum. See id. at 658. Rather, the panel rejected his collateral attack on the guidelines range because that range was not binding on the sentencing court.

Here, by contrast, the district court was required to sentence Hueso to no less than 20 years in prison—a doubling of his mandatory minimum, constraining the district court‘s discretion, as it made very clear at sentencing. See AK R. 84 (Sent‘g Hr‘g Tr. at 11–12, 14) (“And so, you know, tough as it is, you know, it‘s not—I didn‘t write law, I just have to impose the law. . . . None of us wrote the law, . . . [a]nd somewhere along the line, the legislature, I guess, in its war on drugs decided that if you‘re going to do it once and then do it again, not learn from your mistakes, then they‘re going to treat you seriously. . . . [a]nd so that‘s why we—that‘s why the sentence, although tough, is deemed reasonable by those who set . . . the sentence.“). Reading Bullard consistently with our precedent in Hill, it is clear that we permit collateral attacks to mandatory sentences when the sentence has become no longer mandatory, even if the resulting sentence is within the statutory maximum. See United States v. Wheeler, 886 F.3d 415, 434 (4th Cir. 2018). To the extent the majority implies that collateral attacks may be unavailable to any prisoners whose sentences fall within their statutory ranges, our binding precedent in Hill states otherwise.

III.

Turning to the merits, the majority raises textual, doctrinal, and prudential reasons for rejecting Hueso‘s § 2241 petition, none of which are convincing. I consider each in turn.

A.

The majority‘s professed adherence to plain-language review falls short, as it ignores the simple fact that the savings clause in § 2255(e) does not contain the textual limitations placed on second or successive motions in § 2255(h). Whereas the savings clause permits recourse to § 2241 when a remedy by § 2255 motion is “inadequate or ineffective,” § 2255(h) identifies only two specific scenarios in which successive § 2255 motions are permitted. As we recognized in Hill, this represents an unmistakable difference between these two provisions: “A § 2241 petition is not subject to the general rule against second or successive motions in the absence of newly discovered evidence or a new rule of constitutional law.” 836 F.3d at 594. And as the Supreme Court has emphasized, such differences must mean something. “Where Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.” Russello v. United States, 464 U.S. 16, 23 (1983) (alteration and internal quotation marks omitted). In this vein, “Congress could have made savings clause relief dependent only on changes in Supreme Court constitutional law by using the identical language in [the savings clause], but it did not.” Wheeler, 886 F.3d at 428–29. I would presume that the clear textual differences between § 2255(h) and § 2255(e) are not the result of legislative inadvertence.

Indeed, we have already held that § 2241 covers some territory that § 2255 does not. Compare Wooten v. Cauley, 677 F.3d 303, 307 (6th Cir. 2012) (holding, in part, that a petitioner may use a § 2241 petition based on “the existence of a new interpretation of statutory law“), with 28 U.S.C. § 2255(h) (no exception for new interpretations of statutory law). See also United States v. Barrett, 178 F.3d 34, 51 (1st Cir. 1999) (“[T]he § 2255 savings clause . . . must mean something.“). The majority expends great effort in suggesting alternative routes that previous panels of this court could have taken in construing the text of the savings clause,¹ but fails to acknowledge that our interpretation of “inadequate or ineffective“—as meaning something other than what the two exceptions in § 2255(h) provide—is well established.

Accepting the premise that the savings clause stands for something, the majority contends further that if we allowed a prisoner to access the savings clause based on an intervening circuit court decision, we would “abolish” the requirement of § 2255(h)(2) that a successive motion be based on a new rule of constitutional law from the Supreme Court. Why? Because “prisoners could assert new statutory claims (under § 2241) more easily than they could assert new constitutional claims (under § 2255(h)(2)).” Maj. Op. at 14.

The majority has it backwards. Throughout the opinion, it fails to acknowledge that the savings clause of § 2255(e)—the only way for a federal prisoner to seek successive review of his sentence based on a change in statutory law—is extremely narrow and does not create an “easier” path for relief compared to § 2255(h). A brief overview of 28 U.S.C. § 2255‘s structure explains why. In the first subsection, § 2255(a), the statute provides four circumstances in which a prisoner may attack a sentence: 1) when “the sentence was imposed in violation of the Constitution or laws of the United States,” 2) when the sentencing court “was without jurisdiction to impose such sentence,” 3) when “the sentence was in excess of the maximum authorized by law,” and 4) when the sentence “is otherwise subject to collateral attack.” 28 U.S.C. § 2255(a). In construing this statute, the Supreme Court has long drawn a distinction between those § 2255 claims that assert “constitutional or jurisdictional” errors and those that do not. Snider v. United States, 908 F.3d 183, 189 (6th Cir. 2018) (citing Davis v. United States, 417 U.S. 333, 346 (1974)), cert. denied, 139 S. Ct. 1573 (2019). Claims of constitutional error need not demonstrate a “complete miscarriage of justice” to be cognizable under § 2255. Id.

Thus, even if a prisoner is filing his second, third, or fourth § 2255 motion, he may overcome the statute‘s general bar on such iterative motions if he invokes “a new rule of constitutional law” made retroactive by the Supreme Court. 28 U.S.C. § 2255(h)(2).

Claims in the latter category—such as those raising statutory claims, like Hueso‘s—are subject to a more onerous standard. “[N]onconstitutional claims that could not have been asserted on direct appeal can be raised on collateral review only if the alleged error constituted ‘a fundamental defect which inherently results in a complete miscarriage of justice.‘” Stone v. Powell, 428 U.S. 465, 477 n.10 (1976) (quoting Davis, 417 U.S. at 346). The same is true of second or successive statutory claims brought via the savings clause, under which a prisoner seeking to challenge his sentence must demonstrate an “error sufficiently grave to be deemed a miscarriage of justice or a fundamental defect.” Hill, 836 F.3d at 595 (citing Brown v. Caraway, 719 F.3d 583, 586 (7th Cir. 2013), and Williams v. Warden, Fed. Bureau of Prisons, 713 F.3d 1332, 1343 (11th Cir. 2013)). Yet the majority repeatedly ignores just how narrow § 2255(e)‘s supplementary path to relief is. Arguing that Hueso‘s interpretation of the savings clause would “write th[e] [§ 2255(h)(2)] limit out of the statute,” Maj. Op. at 2, the majority fails to acknowledge our well-established precedent that “the § 2255 remedy is not considered inadequate or ineffective simply because § 2255 relief has already been denied, or because the petitioner is procedurally barred from pursuing relief under § 2255, or because the petitioner has been denied permission to file a second or successive motion to vacate.” Charles v. Chandler, 180 F.3d 753, 756 (6th Cir. 1999) (citations omitted). Thus, when the majority asks, “Does a new statutory decision from a circuit court suffice to show § 2255‘s inadequacy?“, Maj. Op. at 11, it is asking the wrong question. I agree with the majority that the answer to its question is no, as does Hueso.² Only when a prisoner, relying on a new, retroactive decision, can demonstrate an “error sufficiently grave to be deemed a miscarriage of justice or a fundamental defect” does the savings clause become available. Hill, 836 F.3d at 595. Thus, allowing a prisoner to raise an intervening circuit court decision via § 2241 does not allow him to utilize statutory law “more easily” than constitutional law, given this onerous requirement imposed on statutory claims.

The majority suggests that this interpretation would allow a showing of fundamental defect to “automatically overcome[] Congress‘s second-or-successive limits in § 2255(h) or its time limits in § 2255(f).” Maj. Op. at 20. This cannot be the case, the majority continues, because a federal prisoner filing his first § 2255 motion based on a statutory error must show fundamental defect, and there must be some difference between what is required to file a first and a second or successive motion. Again, the majority tells only half of the story. Even when a federal prisoner can demonstrate that a statutory error in his conviction or sentence constituted fundamental defect, savings-clause relief will be unavailable unless that prisoner can demonstrate “a case of statutory interpretation . . . that is retroactive and could not have been invoked in the initial § 2255 motion.” Hill, 836 F.3d at 595. Only by ignoring the other requirements of accessing savings-clause relief, as clarified in Hill, can the majority characterize the interpretation of § 2255 offered here as “read[ing] the limits out of the statute.” Maj. Op. at 20.

Still, one might wonder why an intervening circuit court construction of statutory law—if creating a “fundamental defect” in a petitioner‘s sentence—could afford relief under § 2241, while an intervening circuit court construction of constitutional law could not. I agree with the majority that such a reading would fail to harmonize § 2255 with § 2241. Instead, I would permit intervening circuit court decisions of constitutional law—again, only if triggering a “fundamental defect” in a petitioner‘s sentence—to make savings clause relief available. This reading would not render § 2255(h)(2) moot in any sense. After all, in the post-AEDPA landscape, a prisoner can file successive § 2255 motions so long as he demonstrates a new Supreme Court rule of constitutional law, even without a showing of error constituting a “complete miscarriage of justice.” See Davis, 417 U.S. at 346; 28 U.S.C. § 2255(h)(2). See also United States v. Mikalajunas, 186 F.3d 490, 495–96 (4th Cir. 1999) (“[T]he scope of review of non-constitutional error is more limited than that of constitutional error; a nonconstitutional error does not provide a basis for collateral attack unless it involves ‘a fundamental defect which inherently results in a complete miscarriage of justice,’ or is ‘inconsistent with the rudimentary demands of fair procedure.‘“) (citations omitted). Yet if the petitioner seeks to rely on a circuit court decision of constitutional law, he would need to make a showing of “fundamental defect” under § 2241, as discussed above. This reading would avoid 1) creating an “odd state of affairs” in which a prisoner with a second or successive statutory claim could secure relief based on an intervening circuit court case, while a prisoner with a second or successive constitutional claim could not, Chazen v. Marske, 938 F.3d 851, 864 (7th Cir. 2019) (Barrett, J., concurring), and 2) rendering § 2255(h)(2) superfluous, because the only instances in which intervening circuit court decisions of constitutional law would suffice for savings-clause purposes would be those involving “error[s] sufficiently grave to be deemed a miscarriage of justice or a fundamental defect.” Hill, 836 F.3d at 595. Not only is there “no need to read the savings clause as dependent only on a change in Supreme Court law,” Wheeler, 886 F.3d at 428, but, in order to harmonize § 2255(h) with § 2241, there is an obligation to permit savings-clause relief for statutory and constitutional claims alike.

IV.

The majority next attempts to justify its holding by reference to our precedent, beginning with a discussion of our past rejection of § 2241 petitions that did not raise claims of actual innocence. The relevance of this discussion is unclear, as this case is about something different: Whether an intervening change in circuit law may render a § 2255 motion “inadequate or ineffective.” Historical overview of this question confirms that we have never limited use of the savings clause to situations in which Supreme Court law has changed. In fact, in analyzing whether a prisoner may access the savings clause, we have explicitly considered whether the prisoner has demonstrated an intervening change in circuit law.

A.

In none of the cases cited by the majority have we limited use of the savings clause to situations involving intervening Supreme Court decision that postdate a prisoner‘s initial § 2255 motion. The majority describes our recent decision in Wright v. Spaulding, 939 F.3d 695 (6th Cir. 2019), as stating that our earlier decision in Martin v. Perez, 319 F.3d 799 (6th Cir. 2003), included this limitation. See Maj. Op. at 6. In Wright, a panel of this circuit stated that Martin limited successive habeas petitions to situations in which a prisoner‘s argument “had been blocked by binding precedent until a new Supreme Court case cleared the path.” 939 F.3d at 702 (citing Martin, 319 F.3d at 805). In Martin, however, we never stated that it “mattered,” Wright, 939 F.3d at 702 (discussing Martin), whether the intervening change in law came from the Supreme Court or a circuit court. Nothing in Martin addresses the issue we face here: Whether the distinction between Supreme Court and circuit court caselaw makes a difference.3 Nor did we limit the savings clause in this way in United States v. Peterman, 249 F.3d 458 (6th Cir. 2001), another case cited by the majority in its historical overview. Peterman rejected three prisoners’ § 2241 petitions “because [they] ha[d] not shown an intervening change in the law that establishe[d] their actual innocence.” Id. at 462 (emphasis added).4 Absent in this formulation is any reference to—let alone a requirement of—an intervening change in Supreme Court law. Nor is there any such requirement in Bannerman v. Snyder, 325 F.3d 722 (6th Cir. 2003), another case cited by the majority in its doctrinal review.

Finally, our decision in Hill did not limit all savings-clause claims to situations involving intervening changes in Supreme Court law. In Hill, we did state that a petitioner could access the savings clause “when a subsequent, retroactive change in statutory interpretation by the Supreme Court reveals that a previous conviction is not a predicate offense for a career-offender enhancement.” 836 F.3d at 600. Yet “we reiterate[d] that our decision addresses only a narrow subset of § 2241 petitions,” id. at 599, and here Hueso brings a different type of claim from the one brought by Hill. In fact, we stated more generally in Hill:

When seeking to petition under § 2241 based on a misapplied sentence, the petitioner must show (1) a case of statutory interpretation, (2) that is retroactive and could not have been invoked in the initial § 2255 motion, and (3) that the misapplied sentence presents an error sufficiently grave to be deemed a miscarriage of justice or a fundamental defect.

Id. at 595 (emphasis added). This enunciation of our standard clearly does not limit all uses of § 2241 to cases involving “a Supreme Court case of statutory interpretation“; had we intended such a limit, we would have said so.

The majority‘s reliance on other precedent interpreting “cause” for failing to raise arguments earlier is similarly unavailing. The majority argues that before Congress codified limits on successive § 2255 motions in AEDPA, “the Supreme Court‘s cases suggested that nothing but a new Supreme Court precedent” would establish “cause” for a new motion. Maj. Op. at 16. Specifically, the majority states:

If . . . a circuit‘s decisions precluded a claim—that is, if the claim was “unacceptable to that particular court at that particular time,” Engle v. Isaac, 456 U.S. 107, 130 n.5 (1982) (citation omitted)—the prisoner could not establish “cause” for failing to raise the claim from this fact alone. See Bousley, 523 U.S. at 623–24.

Id. This statement overreaches. First, neither the Supreme Court in Engle v. Isaac nor the Ninth Circuit in the omitted citation above (a dissenting opinion by Judge Poole in Myers v. State of Wash., 646 F.2d 355, 364 (9th Cir. 1981)) were discussing the preclusive effect of a circuit court‘s decisions. The Supreme Court in Engle was explaining that a habeas petitioner under § 2254 “may not bypass the state courts simply because he thinks they will be unsympathetic to the claim,” because “[e]ven a state court that has previously rejected a constitutional argument may decide, upon reflection, that the contention is valid.” Engle, 456 U.S. at 130. Federal circuit courts, by contrast, are unable to alter course upon their own reflection. See, e.g., 6th Cir. R. 32.1 (“Published panel opinions are binding on later panels. A published opinion is overruled only by the court en banc.“). Similarly, Judge Poole in Myers was describing his unwillingness to allow § 2254 petitioners to bypass the state courts: “[U]nless the objection was truly so far ahead of its time that its conception would have been visionary, I think that the state trial court is entitled to an opportunity to consider it before it may provide the basis for reversal on federal habeas.” Myers v. State of Wash., 646 F.2d 355, 364 (9th Cir. 1981) (Poole, J., dissenting), cert. granted, judgment vacated sub nom. Washington v. Myers, 456 U.S. 921 (1982). Neither of these opinions dealt with adverse circuit precedent, and neither is binding on this court.

Second, although the Supreme Court in Bousley v. United States did reject a § 2255 petition despite settled, adverse circuit precedent foreclosing the prisoner from raising an argument earlier, it did so in the context of procedural default on a direct appeal, not in the context of a successive § 2255 motion. See 523 U.S. 614, 623 (1998). Its treatment of adverse circuit precedent is not binding on our court‘s interpretation of the savings clause, as Judge Posner explained in In re Davenport, 147 F.3d 605 (7th Cir. 1998). In that case, notwithstanding the Bousley decision a month prior, the Seventh Circuit permitted a prisoner to file a successive § 2255 motion because settled adverse circuit precedent would have rendered an earlier argument futile. Id. at 611 (“The trial judge, bound by our pre-Bailey cases, would not listen to [the prisoner]; stare decisis would make us unwilling (in all likelihood) to listen to him.“).⁵

In fact, published caselaw in this circuit has supported consideration of intervening circuit law as a basis for accessing the savings clause. We have held that when a prisoner seeks access to § 2241 via the savings clause because a change in law meant the prisoner was actually innocent, that change in law did not have to come from the Supreme Court; a change in circuit law was sufficient. See Phillips v. United States, 734 F.3d 573, 582 (6th Cir. 2013) (denying a petition based on actual innocence because “[n]either this Circuit nor the Supreme Court has issued any ‘new decisions interpreting . . .’ the criminal offense” at issue); see also Haque v. Warden, 665 F. App‘x 390, 396–97 (6th Cir. 2016) (considering an intervening circuit case but declining to grant a § 2241 petition because the case did not create a “new rule of statutory interpretation“). There is no meaningful distinction between the claims in these cases and Hueso‘s claim. Thus, the majority‘s statement that our savings-clause cases have “all involved a new Supreme Court decision,” Maj. Op. at 11, ignores the fact that we have explicitly considered new circuit court decisions in the savings-clause analysis.

B.

Because the majority believes that the savings clause does not permit petitions like Hueso‘s, it does not reach the questions of whether Valencia-Mendoza—the Ninth Circuit case on which Hueso relies for his savings-clause argument—1) “could not have been invoked in the initial § 2255 motion” and 2) was a retroactive decision. Hill, 836 F.3d at 595. I would reach both of these questions and hold that Hueso has satisfied this portion of the Hill test.

1.

On the question of whether earlier invocation of the decision was possible, the majority is correct that Carachuri-Rosendo v. Holder, 560 U.S. 563 (2010), on which Valencia-Mendoza was based in part, was issued before Hueso filed his § 2255 motion—indeed, it was issued during the pendency of his direct appeal. But it is the later-decided Valencia-Mendoza, not Carachuri-Rosendo, that abrogated “binding adverse precedent” and “clear[ed] a path” for Hueso‘s habeas petition. Wright v. Spaulding, 939 F.3d 695, 703, 706 (6th Cir. 2019). After all, Carachuri-Rosendo did not deal with the definition of a felony offense, or with federal drug laws at all. It was an immigration case, interpreting a different statutory term in a different context.

Moreover, the fact that Carachuri-Rosendo was a building block for the Ninth Circuit‘s decision in Valencia-Mendoza does not undermine Hueso‘s argument that he could not have raised the Ninth Circuit case earlier. On this point, the doctrinal chronology leading up to this case is important, and similar to that of Hill, in which the petitioner was permitted to use the savings clause after an intervening decision despite the existence of supportive, Supreme Court precedent at the time he filed his first § 2255 motion. At first blush, Hill may appear dissimilar; after all, the relevant new cases that Hill relied upon in his § 2241 petition—Descamps and United States v. Royal, 731 F.3d 333 (4th Cir. 2013)—were decided after Hill filed his direct appeal and § 2255 motion, so Hill clearly could not have invoked them earlier. Yet the Descamps decision was based on Supreme Court caselaw that predated Hill‘s conviction. See Descamps, 570 U.S. at 260 (citing Taylor v. United States, 495 U.S. 575 (1990)). The Court said in Descamps that its “caselaw . . . all but resolves this case” and that the Descamps decision was merely applying that caselaw in “the only way we have ever allowed.” Id. at 260, 263. Thus, at least some of the building blocks of Descamps were available to Hill in his direct appeal and earlier § 2255 motion, yet he was not foreclosed from filing a § 2241 petition based on Descamps. If we bar Hueso‘s petition because at least one of the building blocks of Valencia-Mendoza was available to him at the time of his direct appeal, then Hill‘s petition ought to have been barred as well.

There is a further complication, however. A panel of this circuit recently opined that “[a]t no point did the Hill panel consider and consciously decide whether Hill‘s claim had previously been unavailable. . . . There was no ‘application of the judicial mind’ to that question, so there was no ‘decision’ about it.” Wright, 939 F.3d at 704.⁶ In other words, Wright says that the foregoing discussion of Hill—specifically related to Hill‘s ability to rely on Descamps despite an earlier opportunity to raise its foundational cases as a basis for relief—is not the holding of Hill. I am not so convinced. Although the government in Hill conceded that the petitioner could not have invoked his claim any earlier, see Corrected Br. for the Respondent Warden at 38, 43, Hill v. Masters, 836 F.3d 591 (6th Cir. 2016) (No. 15-5188), and the panel in Hill did not offer independent analysis on this issue, see Hill, 836 F.3d at 595–96, the fact remains that the parties could not stipulate their way around § 2255(e)‘s requirement that in order to file a § 2241 petition, a § 2255 motion must have been deemed “inadequate or ineffective.” If Descamps had been published a year before Hill filed his § 2241 petition and the government had erroneously conceded that Hill could not have raised it earlier, we surely would not have granted Hill relief based on the parties’ flawed stipulation of law.

In any event, whether or not Hill‘s acceptance of the parties’ agreement that Hill could not have raised his claim earlier qualifies as a holding, the fact that we permitted this claim despite existing Supreme Court caselaw that Hill could have theoretically raised earlier is highly significant in Hueso‘s case, which raises precisely the same issue. Here, existing Supreme Court caselaw—namely Carachuri-Rosendo—supported Hueso‘s argument that his prior convictions did not qualify as felony drug offenses, but adverse circuit precedent—United States v. Rios-Beltran, 361 F.3d 1204 (9th Cir. 2004), and United States v. Murillo, 422 F.3d 1152 (9th Cir. 2005)—directly foreclosed this argument when he first litigated a § 2255 motion. As in Hill, it was not until the circuit court explicitly abrogated its old rule that Hueso could reasonably be expected to raise the claim. In Wright, by contrast, we concluded that the petitioner—who argued that the Supreme Court‘s decision in Mathis constituted intervening law, which he could not have raised earlier—did not “need Mathis to clear a path through erroneous Fourth Circuit precedent.” 939 F.3d at 706. We concluded that Wright identified only one binding circuit case on point that, in fact, was consistent with Mathis—far from the “binding adverse precedent” Wright needed to show that he had “no reasonable opportunity” to make his winning argument earlier. Id. at 703, 706 n.8. In this case, on the other hand, the now-overruled Ninth Circuit cases cited by Hueso unambiguously foreclosed the “new legal arguments” he raises in his § 2241 petition. Id. at 705. See, e.g., Rios-Beltran, 361 F.3d at 1208 (“The actual sentence imposed on an individual for a prior conviction . . . is not the relevant inquiry. We look to the maximum penalty allowed by law in determining whether a prior conviction constitutes an aggravated felony under state law for purposes of § 2L1.2.“). This case is closer to Hill, in which controlling circuit precedent foreclosed relief.

2.

Because the majority rejects Hueso‘s argument that Valencia-Mendoza‘s alteration of the “felony offense” definition allows him to use the savings clause in § 2255(e), it also does not reach the question of whether Valencia-Mendoza applies retroactively. See Hill, 836 F.3d at 595. “New substantive rules generally apply retroactively,” Schriro v. Summerlin, 542 U.S. 348, 351 (2004), whereas “rules of criminal procedure” generally do not, Teague v. Lane, 489 U.S. 288, 310 (1989). A new rule is substantive “if it alters the range of conduct or the class of persons that the law punishes.” Schriro, 542 U.S. at 353. “Such rules apply retroactively because they ‘necessarily carry a significant risk that a defendant stands convicted of “an act that the law does not make criminal“’ or faces a punishment that the law cannot impose upon him.” Id. at 352 (quoting Bousley v. United States, 523 U.S. 614, 620 (1998) (citation omitted)).

The question here is whether Valencia-Mendoza, in applying Carachuri-Rosendo to the case before it, has retroactive application. I would hold that it does, following the Fourth Circuit‘s analysis in Miller v. United States, 735 F.3d 141, 145 (4th Cir. 2013). In Miller, the Fourth Circuit confronted the same issue we face here, examining whether a prior circuit case—United States v. Simmons, 649 F.3d 237 (4th Cir. 2011) (en banc)—applied retroactively. As the Ninth Circuit did in Valencia-Mendoza, the Fourth Circuit had applied Carachuri-Rosendo in Simmons to determine whether a petitioner‘s earlier state conviction was for an offense punishable by imprisonment for a term exceeding one year. Like Hueso, Simmons had received a doubling of his mandatory-minimum sentence—from five to ten years, in Simmons‘s case—under 21 U.S.C. § 841(b), because existing precedent required the sentencing court to determine whether Simmons‘s prior conviction was punishable by a prison term exceeding one year by referencing “the maximum aggravated sentence that could be imposed for that crime upon a defendant with the worst possible criminal history.” Id. at 241 (quoting United States v. Harp, 406 F.3d 242, 246 (4th Cir. 2005)). For the same reasons discussed in Valencia-Mendoza, the en banc court in Simmons reversed circuit precedent and held that under Carachuri-Rosendo, a prior conviction under North Carolina law was punishable by more than one year of imprisonment only if the defendant‘s conviction, based on individualized offense characteristics and criminal history, permitted such a sentence. Id. at 244.

Subsequently, the Fourth Circuit held in Miller that a prisoner could benefit retroactively from the Simmons decision because Simmons “applied Carachuri to create a new substantive rule,” even if the Supreme Court‘s decision in Carachuri-Rosendo itself did not create a new substantive rule. Miller, 735 F.3d at 145. Specifically, the Miller court held that ”Simmons requires the court to look at how much prison time the defendant was exposed to given his own criminal history at the time he was sentenced and any aggravating factors that were actually alleged against him,” and in doing so “altered ‘the class of persons that the law punishes.‘” Id. at 146 (quoting Schriro, 542 U.S. at 353). Unlike a procedural rule, the new rule in Simmons did not “merely raise the possibility that someone convicted with use of [an] invalidated procedure might have been acquitted otherwise.” Id. at 145 (quoting Schriro, 542 U.S. at 352).

Simmons did for the prisoner there precisely what Valencia-Mendoza does for Hueso here: It removed him from the “class of persons” on whom the law imposed double the mandatory-minimum sentence to which he would have otherwise been subject. A decision that “narrow[s] the scope of a criminal statute,” Schriro, 542 U.S. at 352, by eliminating certain individuals, like Hueso, from the ambit of a mandatory-minimum sentencing enhancement qualifies as a new substantive rule, whether or not the resulting sentence could have theoretically been imposed by the sentencing court. As a case in point, Valencia-Mendoza afforded relief to a prisoner who had incorrectly received a sentencing enhancement under U.S.S.G. § 2L1.2, despite the fact that his resulting prison sentence—twenty-four months—was still within the statutory maximum set by 8 U.S.C. § 1326. The new rule from Valencia-Mendoza thus significantly shrunk the class of persons who are to be punished as prior felony offenders, and should have retroactive application.

V.

Finally, the majority raises prudential concerns with an interpretation of the savings clause that would allow a prisoner like Hueso to seek resentencing based on an update in the law of the circuit where he was sentenced. First, the majority argues that its rule “at least lessens the potential friction between the sentencing court . . . and the court of confinement.” Maj. Op. at 17. A rule taking its “marching orders” from the Supreme Court, the majority says, avoids a system in which courts of confinement “‘grade’ the opinions of the sentencing circuit.” Id. This misunderstands the dynamic of claims like Hueso‘s. For us to grant Hueso relief under § 2241 would not involve any complex interpretation or “grading” of Ninth Circuit precedent. The Ninth Circuit has unambiguously overruled its own precedent; we are simply tasked with applying this controlling precedent to Hueso‘s straightforward claim, and sending him back to the Ninth Circuit for resentencing—which is undisputedly how the Ninth Circuit would treat his claim if he were able to file a § 2255 motion there. Even when the relied-upon intervening circuit law is less clear than a case like Valencia-Mendoza, there will be no greater intercircuit friction than in cases involving a prisoner asking the court of confinement to interpret intervening Supreme Court law as striking down binding law of the sentencing circuit. Yet we have allowed such claims—which may involve far more scrutiny by our circuit on other circuits—for years.⁷ See Martin v. Perez, 319 F.3d 799, 805 (6th Cir. 2003). On this point, the majority cannot make up its mind. On the one hand, it favors a system in which our court applies Supreme Court decisions to the law of the sentencing circuit, because “[a]ny new decision from the Supreme Court binds both courts.” Maj. Op. at 17. Yet when it comes to the Supreme Court‘s decision in Carachuri-Rosendo, the majority says this case “offers [no] clear direction” in this context, id. at 18 (quoting Simmons, 649 F.3d at 250 (Duncan, J., dissenting)), contrary to its initial statement that Supreme Court decisions provide the certainty necessary to opine on other circuits’ law. At bottom, applying intervening circuit court decisions set forth in § 2241 petitions would be no more difficult or uncomfortable than what this circuit has done for years.

Second, the majority frets over a rule allowing forum-shopping prisoners—who, in the majority‘s hypothetical, have the capacity to effect their own transfers to circuits with more favorable law—to benefit when “the circuit precedent in the court of confinement is more favorable to the prisoner‘s claim than the circuit precedent in the sentencing court.” Maj. Op. at 17. Yet Hueso has not proposed such a rule, nor would I. Such a rule “would base the choice of law decision on the fortuitous placement of a prisoner by the Bureau of Prisons, not the more rational factor of the place of conviction.” Chaney v. O‘Brien, No. 07-cv-121, 2007 WL 1189641, at *3 (W.D. Va. Apr. 23, 2007) (citing Zuniga-Hernandez v. Gilkey, 242 F. Supp. 2d 549, 554 (S.D. Ill. 2001)), aff‘d, 241 F. App‘x 977 (4th Cir. 2007). Ninth Circuit law, the law of the circuit where Hueso was sentenced, is clearly the law to be applied to Hueso‘s § 2241 petition. Sixth Circuit law is immaterial to assessing the petition. Nothing about a rule permitting prisoners to raise intervening circuit court decisions would result in the type of confusion the majority imagines.

***

For all space the majority devotes to these exaggerated concerns, it is silent on the Kafkaesque implications of its holding. To be clear: The majority today withholds relief from Hueso—whose legal arguments have now been undisputedly accepted by the Ninth Circuit, where he would be resentenced under a new rule that would cut his mandatory minimum in half and could result in his immediate release from prison—because, nearly a decade ago, he did not argue to the Ninth Circuit that its standing interpretation of the law was incorrect. See Appellee Br. at 16 (“[I]t appears Hueso could have been the prisoner to raise Carachuri-Rosendo to the Ninth Circuit and—based on the language in Valencie-Mendoza [sic]—he would have succeeded.“); Maj. Op. at 22 (“[I]f Hueso had timely raised his claim, the Ninth Circuit may well have found that its earlier decisions conflicted with this ‘higher intervening authority’ in his case, not in Valencia-Mendoza‘s case.“). This suggestion is surprising given the majority‘s recognition of “society‘s interest in stopping perpetual attacks on final criminal judgments.” Maj. Op. at 1. The rule created today not only incentivizes but requires prisoners to raise arguments to courts that are squarely foreclosed by binding precedent, to the detriment of judicial efficiency. See In re Davenport, 147 F.3d at 610 (“It would just clog the judicial pipes to require defendants, on pain of forfeiting all right to benefit from future changes in the law, to include challenges to settled law in their briefs on appeal and in postconviction filings.“). Judicial economy aside, it is unclear how the interests of justice are furthered when we reject a meritorious claim from a prisoner who actually did raise winning arguments years before they became the law of the circuit where he was sentenced, see Petition for Writ of Habeas Corpus at 16, Hueso v. Sepanek, No. 13-cv-19, 2013 WL 4017117, (E.D. Ky. Aug. 6, 2013), but not in the circuit where he was sentenced. In the world the majority creates, relief inures to Valencia-Mendoza and not to Hueso, prisoners sentenced in the same circuit, purely by chance. All of this the majority fashions based on an erroneous interpretation of text and precedent.

Each day that Ramon Hueso sits in prison longer than the law requires, our invocations of justice and fairness ring a little hollower. In the death-penalty context, Justice Blackmun once expressed a concern that “[t]he more the Court constrains the federal courts’ power to reach the constitutional claims of those sentenced to death, the more the Court undermines the very legitimacy of capital punishment itself.” Sawyer v. Whitley, 505 U.S. 333, 359–60 (1992) (Blackmun, J. concurring). I fear the “crusade to erect petty procedural barriers” against undisputedly meritorious claims, manifest in the majority opinion and other recent decisions by this court, works a similar effect on this circuit‘s habeas jurisprudence. Coleman v. Thompson, 501 U.S. 722, 758–59 (1991) (Blackmun, J., dissenting). On this point, Chief Judge Gregory‘s dissenting words in the Fourth Circuit‘s now-vacated opinion in United States v. Surratt are apt:

I do not doubt that the majority is sympathetic to [the defendant]. In the end, I suppose we just have fundamentally different views on the role of habeas corpus, as well as the role of the judiciary in granting the writ. I see it as our solemn responsibility to guard against a morbid encroachment upon that which is so precious our Framers ensured its continued vitality in our Constitution. . . . Our abdication of this responsibility begs the question: quis custodiet ipsos custodies? Who will guard the guards themselves?

United States v. Surratt, 797 F.3d 240, 276 (4th Cir. 2015) (Gregory, J., dissenting), pet. for reh‘g en banc granted, (4th Cir. Dec. 2, 2015).⁸ The majority opinion provides no answers to Chief Judge Gregory‘s question.

For the foregoing reasons, I dissent.

Notes

1

Throughout its opinion, the majority suggests that, were it not bound by this court‘s precedent, it would constrict the availability of habeas corpus even more radically than it does today. For example, through a gossamer distinction between “relief” and “remedy,” see Leah M. Litman, Legal Innocence and Federal Habeas, 104 VA. L. REV. 417, 488 & n.336 (2018), the majority would apparently adopt the interpretation of § 2255(e) offered in Prost v. Anderson, 636 F.3d 578 (10th Cir. 2011) (Gorsuch, J.), which would bar § 2241 relief for prisoners convicted of acts that had become no longer criminal, prisoners whose sentences exceeded the statutory maximum, and prisoners whose convictions or sentences violated their constitutional rights.

2

The majority erects and tears down the same straw man in its invocation of § 2255(f)‘s one-year statute of limitations. “[N]o reasonable interpreter,” the majority states, “would say that § 2255(f)‘s one-year statute of limitations makes § 2255‘s remedy ‘inadequate or ineffective’ whenever a prisoner blows that deadline.” Maj. Op. at 14. Again, I agree. Time and again we have held that more is required of a prisoner seeking access to the savings clause. A showing of a missed deadline would not satisfy Hill‘s requirement of “error sufficiently grave to be deemed a miscarriage of justice or a fundamental defect.” Hill, 836 F.3d at 595.

3

As Wright‘s lesson on the definition of a holding should illuminate, see 939 F.3d at 700–02, there is a difference between a) holding that, in an individual case, a new Supreme Court decision has reopened the door to collateral relief and b) holding that only new Supreme Court decisions can reopen this door.

4

We have, of course, eliminated the actual-innocence requirement since deciding Peterman, as the majority acknowledges. Maj. Op. at 6.

5

Nor has the Bousley decision stopped the federal government from repeatedly arguing to the Supreme Court and to this court that the savings clause provides relief when controlling circuit precedent foreclosed a legal claim at the time of sentencing. See, e.g., U.S. Br. in Opp. at 9, 11–13, Dority v. Roy, No. 10-8286 (May 16, 2011); Corrected Br. for the Respondent Warden at 45, Hill v. Masters, 836 F.3d 591 (6th Cir. 2016) (No. 15-5188).

6

As an initial matter, Wright‘s discussion of Hill is itself dicta per the framework Wright sets forth for determining which parts of judicial opinions are binding: It is not “clear that the [Wright] court intended to rest the judgment” on its conclusion that Hill lacked a holding on the petitioner‘s prior opportunity to invoke a new case. 939 F.3d at 701. The opinion did conclude that ”Hill held nothing inconsistent with the reasonable-opportunity standard,” id. at 703, but it did not “actively appl[y] the conclusion to the case in front of it,” id. at 701. There is no further discussion of Hill or what its lack of discussion on “reasonable opportunity” meant for Wright in his habeas petition.

7

Relatedly, the majority misstates the effect of Congress‘s enactment of 28 U.S.C. § 2255. See Maj. Op. at 4 (“The 1948 law eliminated the need for courts to review distant judgments by creating a new cause of action in 28 U.S.C. § 2255.“). Even under the majority‘s preferred construction of the savings clause, however, there would be some need for such review of distant judgments. To characterize the enactment of § 2255 as “eliminat[ing]” any need for collateral review reads the language of § 2255(e)—which, since 1948, has explicitly contemplated such a need—into thin air.

8

The Fourth Circuit voted to rehear Surratt en banc, thereby vacating the panel decision. See 4th Cir. Local Rule 35(c). Before the en banc court could issue an opinion, however, the President commuted Surratt‘s sentence from life imprisonment to 200 months’ imprisonment. See Commutations Granted by President Barack Obama (2009–2017), United States Department of Justice, https://www.justice.gov/pardon/obama-commutations. A majority of the Fourth Circuit‘s judges concluded that this commutation rendered Surratt‘s appeal moot. United States v. Surratt, 855 F.3d 218, 219 (4th Cir.), cert. denied, 138 S. Ct. 554 (2017).

Case Details

Case Name: Ramon Hueso v. J.A. Barnhart

Court Name: Court of Appeals for the Sixth Circuit

Date Published: Jan 9, 2020

Citations: 948 F.3d 324; 18-6299

Docket Number: 18-6299

Court Abbreviation: 6th Cir.