Jesus Ruiz v. United States

990 F.3d 1025

• Court: Court of Appeals for the Seventh Circuit
• Decided: March 10, 2021
• Docket: 18-1114
• Status: Published

Opinion

In the

United States Court of Appeals For the Seventh Circuit ____________________ No. 18-1114 JESUS RUIZ, Petitioner-Appellant, v.

UNITED STATES OF AMERICA, Respondent-Appellee. ____________________

Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:16-cv-2521 — Charles R. Norgle, Judge. ____________________

ARGUED DECEMBER 4, 2020 — DECIDED MARCH 10, 2021 ____________________

Before KANNE, WOOD, and SCUDDER, Circuit Judges. SCUDDER, Circuit Judge. In 1997 a federal jury convicted Je- sus Ruiz of several crimes for his participation in a deadly kid- napping scheme designed to collect drug debts. Ruiz received seven concurrent life sentences plus an additional consecutive term of 45 years’ imprisonment for using a firearm during the underlying crimes of violence in violation of 18 U.S.C. § 924(c). 2 No. 18-1114

Now some 20 years later, Ruiz challenges the validity of his § 924(c) convictions. He contends that the predicate of- fenses underlying these convictions are not “crimes of vio- lence” under the categorical approach required by United States v. Davis, 139 S. Ct. 2319 (2019). Rather than reaching the merits of this claim, however, the district court dismissed Ruiz’s petition on harmless error grounds, concluding that any error in the § 924(c) convictions would have no effect on Ruiz’s seven life sentences. Because we agree that Ruiz is not entitled to relief under 28 U.S.C. § 2255, we affirm. I A Jesus Ruiz worked as an “enforcer” collecting drug debts for a Mexican cartel. Ruiz and his co-conspirators—Luis Al- berto Carreno, Jose de la Paz Sanchez, Miguel Torres, and Sa- lome Varela—collected payments by kidnapping at gunpoint debtors or their family members, holding them hostage, and beating the victims until ransom payments were made. In June 1996 the group committed a spree of four kidnap- pings. Three victims escaped. But a fourth hostage was not so fortunate. Jaime Estrada—a 17-year-old boy and brother of a debtor—was kidnapped by Ruiz and his confederates in Mil- waukee. After the kidnappers drove Estrada to Chicago and held him captive in an apartment, they called his brothers de- manding a $30,000 ransom payment. While waiting for the payment, Torres shot Estrada in the stomach and locked him in a bathroom, leaving him bleeding and vomiting. In the meantime, instead of making the ransom payment, Estrada’s family contacted law enforcement. The FBI inter- vened and orchestrated a controlled ransom delivery No. 18-1114 3

operation. As the FBI moved in on Ruiz, Varela, and Torres, the kidnappers fled the scene and led the FBI on a high-speed chase reaching speeds of nearly 100 miles per hour. At one point during the chase, Varela pointed a gun at a federal agent. The chase ended after an agent struck the conspirators’ car, and Ruiz, Varela, and Torres were apprehended. The next morning, an attendant at a used-car lot on Chi- cago’s west side discovered Estrada alive but gravely wounded. Seventeen days later, he succumbed to his injury. A coroner determined that Estrada had died from his gunshot wound and the 30-hour delay in receiving treatment. B A federal grand jury returned an indictment against Ruiz, Sanchez, Torres, and Varela. In a superseding indictment, Ruiz faced charges of conspiracy to commit racketeering (18 U.S.C. § 1962(d)), conspiracy to commit kidnapping (18 U.S.C. § 1201(c)), kidnapping resulting in death (18 U.S.C. § 1201(a)), assaulting a federal officer (18 U.S.C. § 111), four counts of violating the Hostage Act, including one count re- sulting in death (18 U.S.C. § 1203(a)), and three counts of us- ing a firearm during and in relation to a crime of violence (18 U.S.C. § 924(c)). The indictment listed a different predicate offense for each of the three § 924(c) counts—specifically, the underlying conspiracy to commit kidnapping, kidnapping, and assault on a federal officer charges. A jury convicted Ruiz on all counts. The district court then imposed seven concurrent life sentences, a 10-year concurrent sentence, and—for the three § 924(c) convictions—an addi- tional 45-year consecutive sentence. The district court deter- mined that two counts of conviction carried a mandatory life 4 No. 18-1114

or death sentence. See 18 U.S.C. § 1201(a) (kidnapping, with the district court finding that death resulted); 18 U.S.C. § 1203(a) (hostage taking, with the district court finding that death resulted). Ruiz’s sentencing occurred before the Su- preme Court’s decisions in Apprendi v. New Jersey, 530 U.S. 466 (2000) and Alleyne v. United States, 570 U.S. 99 (2013), so the findings that resulted in the imposition of mandatory life sen- tences were made by the trial judge and not the jury. No as- pect of this appeal, however, presents a question under Ap- prendi or Alleyne. We affirmed Ruiz’s convictions and sentences on appeal. See United States v. Torres, 191 F.3d 799 (7th Cir. 1999). Ruiz was just 18 years old when he committed these crimes. C For the last 20 years, Ruiz has made several attempts to challenge his sentence through 28 U.S.C. § 2255 and § 2241. So far, none has succeeded. As for the appeal before us here, the procedural back- ground began six years ago when the Supreme Court decided Johnson v. United States, 576 U.S. 591 (2015). In Johnson, the Su- preme Court invalidated as unconstitutionally vague the so- called residual clause of the Armed Career Criminal Act, which provided one of the Act’s alternative definitions for a predicate “violent felony.” See 576 U.S. at 606. Ruiz, in turn, sought permission under 28 U.S.C. § 2244(b)(3) to file a new collateral attack, contending that the residual clause of § 924(c)’s definition of “crime of violence” was not only un- constitutionally vague in light of Johnson, but also that his predicate offenses otherwise did not count as crimes of vio- lence under § 924(c)’s elements clause. We granted Ruiz’s No. 18-1114 5

request. See Ruiz v. United States, No. 16-1193 (7th Cir. Feb. 19, 2016). Ruiz then filed a new § 2255 petition and argued to the district court that his § 924(c) convictions should be vacated because those convictions were based on the residual clause’s unconstitutionally vague definition of “crime of violence,” and, in any event, that the predicate offenses used to support these convictions did not categorically require “the use, at- tempted use, or threatened use of physical force against the person or property of another.” 18 U.S.C. § 924(c)(3)(A). The government maintained that all three of Ruiz’s underlying “crimes of violence” remained valid even after Johnson be- cause they involved as an element the “use, attempted use, or threatened use” of force. Alternatively, the government char- acterized any error as harmless because the validity of the § 924(c) convictions would not affect Ruiz’s multiple life sen- tences. The district court denied Ruiz’s § 2255 motion but, in do- ing so, declined to reach the merits of his claims. The court instead concluded that any error relating to the § 924(c) con- victions was harmless because Ruiz faced seven life sentences, including two mandatory life sentences. Even if Ruiz could show that the reasoning in Johnson required his § 924(c) con- victions to be vacated, the district court explained, it would not change the reality that he remains subject to seven unchal- lenged, valid life sentences. Ruiz appealed, and we granted his certificate of appeala- bility. While his appeal was pending, the Supreme Court de- cided United States v. Davis, holding that the residual clause’s definition of “crime of violence” in § 924(c)(3)(B) is indeed 6 No. 18-1114

void for vagueness under similar reasoning employed in John- son. See 139 S. Ct. 2319, 2336 (2019). II When reviewing the district court’s denial of a § 2255 pe- tition, we review its legal conclusions de novo. See Hrobowski v. United States, 904 F.3d 566, 569 (7th Cir. 2018). Under 28 U.S.C. § 2255(b), if a court “finds that the judgment was rendered without jurisdiction, or that the sentence imposed was not authorized by law,” then “the court shall vacate and set the judgment aside,” and shall discharge the prisoner, re- sentence him, grant a new trial, or correct the sentence “as may appear appropriate.” Mindful of this standard, the parties present their argu- ments from opposite ends of the spectrum. On one end, Ruiz invites us to proceed directly to the merits of his Davis claim and vacate his § 924(c) convictions. On the other end, the gov- ernment asks us to affirm the district court’s harmless error analysis and denial of relief. In the government’s view, be- cause Ruiz advances no challenge to his seven life sentences, any relief on the § 924(c) convictions would not affect the amount of time he spends in prison—the definition of harm- less error, as the government sees it. The question of which route to take—Ruiz’s or the govern- ment’s—is not answered by our case law. Nor has our court had occasion to decide whether two of the potential predicate offenses underlying Ruiz’s § 924(c) convictions—kidnapping resulting in death (18 U.S.C. § 1201(a)) and assault on a fed- eral law enforcement officer (18 U.S.C. § 111)—are crimes of violence under § 924(c)(3)’s elements clause. In the end, we No. 18-1114 7

agree with the district court’s approach, so we decline to ad- dress the more complicated merits questions. A The doctrine of harmless error is the product of judicial reform dating to the early twentieth century. See Chapman v. California, 386 U.S. 18, 48 (1967) (Harlan, J., dissenting) (de- scribing the evolution of the American harmless error rule). Most American appellate courts previously followed the Eng- lish rule, which “held that any error of substance required a reversal of conviction.” Id. (emphasis added). This ap- proach—which applied to constitutional errors and statutory- and common-law violations alike—had the unfortunate effect of devolving the criminal trial into a “game for sowing re- versible error in the record, only to have repeated the same matching of wits when a new trial had been thus obtained.” Kotteakos v. United States, 328 U.S. 750, 759 (1946); see also Chapman, 386 U.S. at 48–49 (Harlan, J., dissenting). Concerned that appellate courts were operating as “im- pregnable citadels of technicality,” Kotteakos, 328 U.S. at 759 (footnote omitted), Congress responded in 1919 by enacting Section 269 of the revised Judicial Code. See Sam Kamin, Harmless Error and the Rights/Remedies Split, 88 Va. L. Rev. 1, 10 (2002). The updated code required appellate courts “to re- verse lower court rulings only where the substantial rights of the parties were adversely affected at trial.” Id. (emphasis added). At the time, though, the reformed harmless error rule applied to only statutory and procedural errors. See id. Errors of constitutional magnitude still warranted reversal of a de- fendant’s conviction. See id. 8 No. 18-1114

More change came in 1967. It was then that the Supreme Court decided Chapman v. California, which upended the di- chotomy between constitutional and non-constitutional er- rors by holding that the doctrine of harmless error applies to “constitutional errors which in the setting of a particular case are so unimportant and insignificant” that they may be deemed inconsequential. 386 U.S. at 22. But some constitutional errors, the Chapman Court recog- nized, remain so intrinsically damaging and basic to our trial system as to never be harmless. See id. at 23 n.8 (listing as ex- amples Payne v. Arkansas, 356 U.S. 560 (1958) (coerced confes- sion); Gideon v. Wainwright, 372 U.S. 335 (1963) (right to coun- sel); and Tumey v. Ohio, 273 U.S. 510 (1927) (impartial judge)). The law has come to call these violations structural errors. See Weaver v. Massachusetts, 137 S. Ct. 1899, 1907 (2017) (“The pur- pose of the structural error doctrine is to ensure insistence on certain basic, constitutional guarantees that should define the framework of any criminal trial.”); see, e.g., Sullivan v. Louisi- ana, 508 U.S. 275 (1993) (failure to give a jury a reasonable- doubt instruction); Batson v. Kentucky, 476 U.S. 79 (1986) (ex- clusion of jurors based on race); Vasquez v. Hillery, 474 U.S. 254 (1986) (exclusion of grand jurors based on race); Waller v. Geor- gia, 467 U.S. 39 (1984) (denial of a public trial); McKaskle v. Wiggins, 465 U.S. 168 (1984) (denial of the right to self-repre- sentation at trial). At bottom, the doctrine of harmless error owes its exist- ence to the concept that a legal error having no consequential effect on a judgment does not necessarily need to be rectified. See 28 U.S.C. § 2111 (“On the hearing of any appeal or writ of certiorari in any case, the court shall give judgment after an examination of the record without regard to errors or defects No. 18-1114 9

which do not affect the substantial rights of the parties.”). Our cases make clear that the doctrine likewise applies to ordinary sentencing errors. See, e.g., United States v. Clark, 906 F.3d 667, 671 (7th Cir. 2018) (“In a criminal-sentencing case, a finding of harmless error ‘removes the pointless step of returning to the district court when we are convinced that the sentence the judge imposes will be identical to the one we remanded.’” (quoting United States v. Abbas, 560 F.3d 660, 667 (7th Cir. 2009))). At its core, harmless error review is an equitable doc- trine allowing courts to decline to afford relief when an error does not affect an existing judgment. See Chapman, 386 U.S. at 22–24. Two Supreme Court cases define and govern our modern harmless error doctrine. The first case, Chapman, remains the leading decision for reviewing constitutional errors on direct appeal and places the burden on the government to show that the error “was harmless beyond a reasonable doubt.” Id. at 24. The second case, Brecht v. Abrahamson, applies to constitu- tional infirmities identified and advanced by state prisoners on collateral review in federal habeas corpus proceedings. See 507 U.S. 619 (1993). Under Brecht, the state prisoner bears the burden of demonstrating that the error “had [a] substantial and injurious effect or influence in determining the jury’s ver- dict.” Id. at 637–38. To date the Supreme Court has not addressed whether Chapman, Brecht, or a third standard applies to federal prison- ers seeking post-conviction relief under 28 U.S.C. § 2255. Nei- ther has our court taken a position on the issue and indeed “our caselaw gestures in conflicting directions.” Daniels v. United States, 939 F.3d 898, 903 (7th Cir. 2019). In Lanier v. United States, for example, we applied a Chapman-like 10 No. 18-1114

harmless error standard to the petitioner’s § 2255 motion. See 220 F.3d 833, 839 (7th Cir. 2000). More recently in Sorich v. United States, however, we applied the Brecht standard to a § 2255 motion on the joint agreement of the parties. See 709 F.3d 670, 674 (7th Cir. 2013). All of these principles apply here. And against this legal framing of the harmless error doctrine, we turn to what all of this means for Ruiz’s § 924(c) convictions and his 45-year con- secutive sentence. B In the midst of the uncertainty surrounding harmless error review in the context of § 2255 petitions, we need not plant our feet firmly on the correct standard to apply in Ruiz’s case. Under either measure—and with the extraordinary fact pat- tern before us—we cannot say that any error underlying Ruiz’s § 924(c) convictions could be considered anything other than harmless. See Daniels, 939 F.3d at 903 (declining to “resolve [the] tension” in our case law because the error was “harmless under any standard”). Recall that in addition to the 45-year consecutive sentence, Ruiz received seven concurrent life terms. For two of his con- victions—the hostage taking of Jaime Estrada (18 U.S.C. § 1203(a)) and the related kidnapping (18 U.S.C. § 1201(a))— the district court determined, based on the jury’s verdict and factual findings at sentencing, that death resulted and that it was required to impose a punishment of life imprisonment or death. Ruiz does not challenge the validity of those convic- tions or, for that matter, any of his seven life sentences. With the reality of Ruiz’s stark situation before us, it is dif- ficult to see how any relief—even a complete vacatur of the No. 18-1114 11

§ 924(c) convictions and their accompanying sentences— would reduce the time that Ruiz must serve in prison. Stated otherwise, he cannot show any prejudice befalling him from any erroneous § 924(c) convictions. Nor has Ruiz established that he will suffer any concrete, non-speculative collateral consequences if we decline to reach the merits of his Davis claim, let alone any consequences affecting his “custody” for purposes of habeas relief. See 28 U.S.C. § 2255. Ruiz begs to differ and presses us to presume that his § 924(c) convictions and his 45-year consecutive sentence carry collateral consequences. Relying on Sibron v. New York, 392 U.S 40 (1968), Ruiz emphasizes that “the obvious fact of life [is] that most criminal convictions do in fact entail adverse collateral legal consequences.” Id. at 55. At that level of gener- ality, Ruiz is right. Indeed, the Supreme Court has been “will- ing to presume that a wrongful criminal conviction has con- tinuing collateral consequences (or, what is effectively the same, to count collateral consequences that are remote and unlikely to occur).” See Spencer v. Kemna, 523 U.S. 1, 8 (1998) (citing Sibron, 392 U.S. at 55–56). But recognizing that general presumption does not establish the more specific precept that a criminal conviction can never be harmless—that circum- stances may exist where collateral consequences are exceed- ingly remote and highly unlikely to ever manifest themselves. Nor, of course, does Ruiz’s general observation account for the reality that his pursuit of habeas relief depends on identi- fying a collateral consequence that rises to the level of impact- ing his ongoing “custody,” as required by § 2255. Perhaps the closest Ruiz gets to identifying a realistic col- lateral consequence is his mentioning the $300 special assess- ment that the district court ordered him to pay for the § 924(c) 12 No. 18-1114

convictions. See 18 U.S.C. § 3013(a)(2)(A). Though this posi- tion would have merit on direct appeal, it falls short here be- cause § 2255 serves as a remedy to contest a prisoner’s cus- tody—not the imposition of fines or other special assess- ments. See Ryan v. United States, 688 F.3d 845, 849 (7th Cir. 2012) (“A collateral attack under § 2241, § 2254, or § 2255 con- tests only custody, however, and not fines or special assess- ments.”). Ruiz also suggests that if he were convicted of another § 924(c) offense someday in the future, his prior firearm con- victions would trigger increased penalties. But this contention rests on a big “if”: that scenario could come to pass only if Ruiz somehow managed to possess a gun while incarcerated. In further effort to establish a collateral consequence, Ruiz observes that Congress and the Bureau of Prisons generally differentiate among prisoners based on their offense of con- viction, and it is within the government’s prerogative to enact measures unique to individuals like Ruiz with § 924(c) con- victions. See, e.g., 18 U.S.C. § 3632(d)(4)(D)(xxii) (rendering prisoners ineligible for time credit if convicted of a § 924(c) offense); 28 C.F.R. § 550.55(b)(5)(ii) (listing convictions that make inmates ineligible for early release, including firearm offenses). We remain unpersuaded by these examples, how- ever, because Ruiz still faces other valid convictions carrying life sentences that prevent him from taking advantage of time credit for good behavior. See 18 U.S.C. § 3624(b)(1) (prisoners with life sentences ineligible for good-time credit). At bottom, Ruiz points to no traditional collateral consequences—like the loss of the right to vote, participate on a jury, or own a fire- arm—that would not also result from his unchallenged con- victions and life sentences. And though the range of potential No. 18-1114 13

adverse collateral consequences remains broad, Ruiz cannot show that any of them rise to the level of “custody” in the face of seven life sentences he does not even challenge. On another front, Ruiz urges us to consider how future legislative or judicial developments could eventually under- mine his non-§ 924(c) convictions or life sentences, such that it would be preferable to litigate his Davis challenge now. But in addition to not identifying any potential and likely collat- eral consequences, neither has Ruiz forecasted any foreseea- ble changes in the law that call into question his seven life sen- tences. All Ruiz has done, and understandably so, is offer views about how the law may someday change to afford him relief for all seven life sentences. And even if the law were to change—either through legislative action or by Supreme Court decisions—Ruiz would face yet another hurdle of counting on Congress or the Court to apply the law retroac- tively. In the end, Ruiz has not put forth circumstances enabling us to conclude that any error with his § 924(c) convictions is anything other than harmless. C In a similar vein to harmless error review, the government invites us by analogy to consider the reasoning supporting the concurrent sentence doctrine. This discretionary doctrine allows courts to “pretermit decision about convictions pro- ducing concurrent sentences, when the extra convictions do not have cumulative effects.” Ryan, 688 F.3d at 849. Put an- other way, the doctrine “allows appellate courts to decline to review a conviction carrying a concurrent sentence when one ‘concurrent’ conviction has been found valid.” Cheeks v. Gaetz, 14 No. 18-1114

571 F.3d 680, 689 (7th Cir. 2009) (quoting United States v. Kim- berlin, 675 F.2d 866, 867 (7th Cir. 1982)). The extent to which the doctrine may apply “depends on the degree of prejudice that may be attributed to the chal- lenged conviction.” Id. at 689 (quoting Cramer v. Fahner, 683 F.2d 1376, 1380 (7th Cir. 1982)). Where no prejudice re- sults from foregoing review of the challenged conviction, a court may properly exercise its discretion in declining to reach the merits of the conviction. See Hill v. Werlinger, 695 F.3d 644, 649 n.1 (7th Cir. 2012) (explaining that the doc- trine requires (1) an equal or longer sentence on an unchal- lenged or affirmed conviction and (2) no adverse collateral consequences to the prisoner by declining to review the chal- lenged conviction). A recent application of the concurrent sentence doctrine came in Ryan v. United States, 688 F.3d 845 (7th Cir. 2012). In Ryan, a federal jury convicted the former Illinois governor of several crimes based on his involvement in a corruption scan- dal. Among his many crimes of conviction was one RICO vi- olation, which resulted in a sentence of 78 months. See id. at 848. Ryan’s remaining sentences—60-month sentences on seven mail-fraud convictions, 60-month sentences on three false-statement counts, and 36-month sentences on four tax counts—ran concurrently with each other and with the 78- month RICO sentence. See id. On post-conviction review un- der § 2255, Ryan challenged the validity of his RICO convic- tion and the seven mail-fraud convictions that served as the underlying predicate acts necessary to sustain his RICO con- viction. See 18 U.S.C. § 1962(d). Rather than review the valid- ity of each of the seven mail-fraud convictions, we instead considered only whether there were enough valid mail-fraud No. 18-1114 15

convictions to uphold the RICO conviction. See Ryan, 688 F.3d at 848. “An attempt to decide on collateral review whether each of the seven mail-fraud convictions was valid,” we ex- plained, would be “unnecessary” and “would smack of an ad- visory opinion.” Id. at 849, 852. Ruiz’s situation, of course, does not fit within the concur- rent sentence doctrine because his § 924(c) convictions yielded consecutive sentences to be served in addition to his seven life sentences. Even so, the same considerations of futil- ity, speculation, and preservation of judicial resources that underpinned our discretion in Ryan to not review all seven mail-fraud convictions rings true here too. Reviewing the va- lidity of Ruiz’s § 924(c) convictions in the face of seven re- maining and valid life sentences is similarly unnecessary, as our review would lead to no practical or concrete sentencing relief for Ruiz. Make no mistake. We are not adopting a “consecutive sen- tence doctrine” analogous to the concurrent sentence doc- trine. In ordinary cases, such a doctrine could not operate. In- deed, in most circumstances involving consecutive sentences, a prisoner would suffer tangible prejudice if an invalid con- viction remained on his record because he would be required to serve a longer actual prison term. But based on the excep- tional circumstances presented here—Ruiz’s unchallenged seven life sentences and a 45-year consecutive sentence—the government’s analogy to the concurrent sentence doctrine re- inforces our harmless error analysis: Ruiz has not demon- strated that he will suffer actual prejudice by our foregoing review of his Davis claim. 16 No. 18-1114

D Consider the alternative pressed by Ruiz. His position would require us to confront complex legal questions yet to be addressed by our court. This complexity is not imaginary. Take, for example, Ruiz’s conviction for kidnapping Estrada in violation of 18 U.S.C. § 1201(a). Ruiz and the government dispute whether the jury convicted Ruiz of “simple” kidnapping or kidnapping resulting in death. This dispute is largely driven by conflicting indicators in the so-called Shepard documents. See Shepard v. United States, 544 U.S. 13, 16 (2005) (explaining that courts, when applying the modified categorical ap- proach, may examine a limited class of documents, such as the indictment, jury instructions, and other trial court rec- ords); see also Haynes v. United States, 936 F.3d 683, 687 (7th Cir. 2019) (applying Shepard in the context of § 924(c) convic- tions). The indictment seems to have charged Ruiz with kid- napping resulting in death, but the jury instructions sug- gested that he was convicted of simple kidnapping. The dif- ference matters because simple kidnapping is not a crime of violence, see United States v. Jenkins, 849 F.3d 390, 394 (7th Cir. 2017), cert. granted, judgment vacated, 138 S. Ct. 1980 (2018), reinstated sub. nom., United States v. Jackson, 932 F.3d 556, 557 (7th Cir. 2019), but kidnapping resulting in death might be. Taking the next step in the analysis, if we were to agree with the government that Ruiz was convicted of kidnapping resulting in death, we would then have to decide whether that offense qualifies as a “crime of violence” under the elements clause in 18 U.S.C. § 924(c)(3). At least two circuits have an- swered that question yes, but both did so over dissenting opinions. See In re Hall, 979 F.3d 339 (5th Cir. 2020); United No. 18-1114 17

States v. Ross, 969 F.3d 829 (8th Cir. 2020). Though kidnapping resulting in death “sure sounds like a ‘crime of violence,’” making that determination is by no means simple. Ross, 969 F.3d at 845 (Stras, J., concurring in the judgment in part and dissenting in part). Indeed, because we have already held that simple kidnapping is not a crime of violence, our inquiry would focus on whether the statutory language, “if the death of any person results,” necessarily involves the use of force as understood in § 924(c)(3)(A). And if we were apt to disagree with the Fifth and Eighth Circuits, we would create a circuit split. To be sure, we have benefitted from outstanding represen- tation provided by all counsel on appeal, and if we were to address the merits questions, we would be aided by their thorough briefing and effective advocacy. Yet, given the unique circumstances before us, this appeal does not require us to answer these difficult questions. III As for the views of our dissenting colleague, we agree with nearly all of them. Our disagreement is limited in that we stop short of recognizing a nearly per se rule that an unlawful con- viction always constitutes a prejudicial error as a matter of law—regardless of the sentence. We do not see such an ironclad requirement as consistent with habeas relief concentrating on the ongoing lawfulness of a petitioner’s custody. Nor, in our respectful view, could such an unyielding rule be reconciled with the concurrent sentence doctrine. It matters not that the concurrent sentence doctrine arose as a discretionary principle, for adoption of a categorical legal rule that a wrongful conviction is always remediable 18 No. 18-1114

and never harmless would prohibit the exercise of discretion in all cases. Our decision is also narrow. In almost all situations, the combination of a constitutionally infirm conviction and con- secutive sentences will be prejudicial to a defendant. Over- whelmingly, unlawful convictions carry with them conse- quences, such that harmless error will have no place in a proper analysis. This case, however, presents the exceedingly rare occasion in which the opposite is true. Absent some extraordinary and unexpected change in the law with retroactive application, Ruiz’s seven life sentences will remain in place. Vacating Ruiz’s § 924(c) convictions (assuming his Davis claim has merit) does nothing to change that unfortunate reality for Ruiz. On top of that, Ruiz would face the nearly insurmount- able challenge of persuading a court to reduce not just one or two, but all seven of his life sentences to secure any prospect of tangible relief. For these reasons, we AFFIRM. No. 18-1114 19

WOOD, Circuit Judge, dissenting. In this proceeding under 28 U.S.C. § 2255, Jesus Ruiz has argued compellingly that he stands convicted of multiple nonexistent crimes, for which he has been sentenced to a term of 45 years. That sentence is to run consecutively to his life sentences on other counts. The government concedes that one conviction underlying the 45- year sentence was erroneous. Ordinarily this would cry out for relief. But because of the life sentences, the majority sees no point in recognizing or correcting this error. It believes, to put it formally, that Ruiz has not, and never can, suffer any prejudice from the extra 45 years, and so no action is required. Both for formal reasons and for practical reasons, I would hold that a conviction for a noncrime is always prejudicial er- ror as a matter of law, regardless of the sentence and how it relates to other convictions and sentences from the same or other proceedings. Furthermore, my crystal ball is not as clear as the majority’s. Future legal developments whose likeli- hood, while perhaps not high, is real, may at a stroke sweep away all seven life sentences and make that 45-year sentence of immediate concern. Its existence on Ruiz’s record cannot be brushed away as harmless error. And now is the time to take action. If Ruiz were to try to bring a second motion under section 2255 in the wake of a pertinent change, there is no guarantee he could meet the exacting criteria of 28 U.S.C. § 2255(h) to pursue it. I would reverse the district court’s de- nial of relief and reach the merits of Ruiz’s claims under United States v. Davis, 139 S. Ct. 2319 (2019). I therefore re- spectfully dissent. 20 No. 18-1114

I A A federal defendant has a due process right to be tried and convicted only for a crime that actually exists. “[C]onviction and punishment … for an act that the law does not make crim- inal … ‘inherently results in a complete miscarriage of justice’ … .” Davis v. United States, 417 U.S. 333, 346 (1974). A convic- tion for a “nonexistent offense” thus reflects a “fundamental … defect” in a criminal judgment and must be set aside. In re Davenport, 147 F.3d 605, 611 (7th Cir. 1998). As the majority notes, Ruiz was indicted on numerous counts associated with his kidnapping of four victims, one of whom died. Pertinent here, the indictment also charged him with three counts of using a firearm during and in relation to a crime of violence, in violation of 18 U.S.C. § 924(c). He was convicted on all counts, and he received seven concurrent life sentences for the racketeering, kidnapping, and hostage of- fenses; he also received an additional consecutive sentence of 45 years for the three firearms offenses. Two of the life sen- tences were mandatory under the governing statutes, 18 U.S.C. §§ 1201(a) and 1203(a). At the time of the offenses, Ruiz was 18 years old. In this appeal, Ruiz is not directly challenging any of the life sentences. He argues only that his three section 924(c) con- victions are based on conduct that is no longer criminal after the Supreme Court’s 2019 Davis decision, 139 S. Ct. 2319, and so the sentences associated with them must be set aside. Each of those three convictions matters: the 45-year term repre- sented a five-year consecutive sentence on Count 9, a 20-year consecutive sentence on Count 10, and another 20-year No. 18-1114 21

consecutive sentence on Count 11. So setting aside even one of the counts of conviction would have a concrete impact. The government concedes that at least one of Ruiz’s firearms con- victions can no longer stand: Count 9, which was based on the predicate offense of “conspiracy to kidnap.” That conviction is no longer valid because conspiracy to kidnap is not a crime of violence under section 924(c)(3)(A). See D’Antoni v. United States, 916 F.3d 658, 665 (7th Cir. 2019). Despite the government’s concession, as well as the strength of Ruiz’s arguments that his other section 924(c) con- victions are also invalid after Davis, the majority declines to reach the merits because it sees no prejudice to Ruiz stemming from this “fundamental” error. Any error was harmless, the majority reasons, because even if we were to invalidate one or more of Ruiz’s section 924(c) convictions, it believes that there is nothing that would change the fact that he is subject to seven concurrent life sentences. As a result, it believes, our de- cision would have no practical effect. I have no quarrel with the proposition that harmless-error analysis is required for a section 2255 motion. See Fed. R. Civ. P. 61 (“At every stage of the proceeding, the court must disre- gard all errors and defects that do not affect any party’s sub- stantial rights.”); Rules Governing Section 2255 Proceedings for the United States District Courts, Rule 12 (applicability of the Federal Rules of Civil Procedure). My dispute is with the application of that rule. The harmless-error inquiry involves two variables: a de- fect and an outcome. Courts engaged in harmless-error anal- ysis generally ask the following counterfactual question: whether, absent the identified defect in a judicial proceeding, the outcome of the proceeding would be different. If so, then 22 No. 18-1114

the defendant was harmed by the error and relief is appropri- ate. (Although certain structural defects are deemed always to be harmful, for present purposes I do not take issue with the majority’s implicit conclusion, ante at 8, that no such al- leged error occurred here.) In Ruiz’s case, the relevant defect appears in the court’s jury instruction, which made it possible for Ruiz to be con- victed based on facts that the law does not criminalize. The relevant outcome is Ruiz’s conviction. See California v. Roy, 519 U.S. 2, 5 (1996) (holding that where there is “an error in the instruction that defined the crime,” the proper inquiry is “whether the error had substantial and injurious effect or in- fluence in determining the jury’s verdict” (emphasis added) (internal quotation marks omitted)). When viewed from this perspective, we are safe in concluding that but for the error in the judicial proceedings (i.e., the defective instruction), the outcome would have been different (i.e., Ruiz would not have been convicted on one or more of the section 924(c) counts). Under the harmless-error test, a conviction for a noncrime is, by definition, harmful. The majority goes off track by analyzing the wrong varia- bles: Instead of asking whether Ruiz’s conviction would stand had his jury been instructed to find the elements of a valid crime, the majority jumps to the question whether the practi- cal length of Ruiz’s overall sentence would be any different if his firearms convictions were invalidated. Put another way, the majority views Ruiz’s conviction as the error, not the harm. This approach finds no support in the harmless-error ju- risprudence. The principal cases upon which the majority re- lies actually reinforce the notion that the appropriate object No. 18-1114 23

when assessing harm is the conviction, not the resulting sen- tence. See Chapman v. California, 386 U.S. 18, 26 (1967) (finding no harmless error where a defect in the trial proceedings “did not contribute to petitioners’ convictions” (emphasis added)); Brecht v. Abrahamson, 507 U.S. 619, 623 (1993) (assessing whether the “error had substantial and injurious effect or in- fluence in determining the jury’s verdict” (emphasis added) (internal quotation marks and citations omitted)). Even the historical backdrop that the majority cites suggests why harmless error is inapposite here: when the harm in question is a conviction for conduct that the law criminalized at the time of conviction but that has since been recognized as not falling under any criminal prohibition, a court’s vacation of that conviction does not create an incentive for litigants to en- gage in a “game for sowing reversible error in the record.” Kotteakos v. United States, 328 U.S. 750, 759 (1946). No defend- ant has that much foresight.

B Although the harmless-error test looks only to whether a defect in judicial proceedings makes the difference between conviction or acquittal, in one limited context courts do ask whether an invalid conviction will have an effect on the length of a defendant’s sentence. That has come to be called the con- current-sentence doctrine. See, e.g., Ryan v. United States, 688 F.3d 845, 849 (7th Cir. 2012). The concurrent-sentence doctrine has some similarities to harmless-error analysis, but the two rules differ in both source and scope. First, the concurrent-sentence doctrine is a discre- tionary, judicially created, tool, see Steffes v. Pollard, 663 F.3d 276, 280 (2011), whereas the harmless-error test is mandatory, 24 No. 18-1114

see Fed. R. Civ. P. 61; In re Rafdo Enters., Inc., 297 F.2d 505, 507 (7th Cir. 1962). This means that the concurrent-sentence doc- trine is not just a particular application of harmless error. Were that the case, the concurrent-sentence doctrine would be mandatory, not discretionary. Even if the two ideas were thought to overlap, the insur- mountable problem for the majority is that (as it concedes) the concurrent-sentence doctrine does not apply here. Ruiz’s sen- tences are consecutive, not concurrent: the district court sen- tenced him to life plus 45 years. His section 924(c) sentences have not begun to run and will not start unless and until his life sentences are shortened or removed. There is no “consec- utive-to-a-life-sentence doctrine.” To extend the concurrent- sentence doctrine to this setting would be to call into question the purpose of sentencing defendants to consecutive sen- tences beyond life terms—a common and longstanding judi- cial practice. The majority thus errs, in my estimation, by looking to the concurrent-sentence doctrine to justify its replacement of the appropriate object of harmless-error inquiry (the conviction) with an inappropriate one (the overall length of the sentence). The fact that the overall length of a sentence is relevant to the concurrent-sentence doctrine does not mean that it is relevant to harmless-error analysis. Contrary to the position the major- ity takes today, it is always the case that where there is a chal- lenge to a conviction for a noncrime, the defendant’s “sub- stantial rights” have been affected, and so there is harmful er- ror. No. 18-1114 25

II Even accepting the faulty premise that harmless-error analysis permits us to look past the inherent prejudice Ruiz suffers from a defective conviction, the majority’s analysis still falls flat. The majority concludes that Ruiz’s defective firearms con- victions are harmless because he is unable to show that he suf- fers any collateral consequences from those convictions that he does not otherwise suffer from his life sentences. A glaring problem with the majority’s reasoning is that it ignores the clear command of Sibron v. New York, 392 U.S. 40 (1968). Sibron rejects “all inquiry into the actual existence of specific collateral consequences” and establishes a presump- tion that there are collateral consequences associated with each conviction in a criminal judgment. Id. at 55. Critically, the Supreme Court observed that it saw “no relevance in the fact that Sibron [was] a multiple offender” because it is “im- possible … to say at what point the number of convictions on a man’s record renders his reputation irredeemable.” Id. at 56. Of particular relevance here, the Sibron Court also noted the possibility that future legal reforms might invalidate Sibron’s remaining convictions, leaving only the challenged conviction to stand. The Court explained: We cannot foretell what opportunities might present themselves in the future for the removal of other con- victions from an individual’s record. The question of the validity of a criminal conviction can arise in many contexts and the sooner the issue is fully litigated the better for all concerned. … And it is far better to elimi- nate the source of a potential legal disability than to 26 No. 18-1114

require the citizen to suffer the possibly unjustified consequences of the disability itself for an indefinite period of time before he can secure adjudication of the State’s right to impose it on the basis of some past ac- tion. Id. at 56–57 (citation omitted). Although Sibron dealt with mootness, not harmless error, its reasoning applies with equal force here. For all intents and purposes, Sibron creates a categorical rule that criminal con- victions carry collateral consequences—full stop. The major- ity’s willingness to overlook a sentence of 45 years is incon- sistent with Sibron. The majority also ignores Sibron’s lan- guage about the possibility of future legal reforms. Sibron ef- fectively tells courts not to assume the impossibility of legal reforms that might invalidate a defendant’s other convictions or modify his sentence. Under Sibron, the mere chance (no matter how remote) that future reforms might invalidate or decrease the length of Ruiz’s life sentences means that Ruiz’s section 924(c) convictions have practical consequences dis- tinct from those of his life sentences. With one eye on history and the other on current develop- ments, it is not hard to imagine future legal reforms that would alter Ruiz’s remaining convictions or sentences. And these possibilities are far from remote. At the time Ruiz com- mitted the conduct for which he was convicted and sentenced to mandatory life without parole, he was only a few months past his 18th birthday. In Miller v. Alabama, 567 U.S. 460 (2012), the Supreme Court held that mandatory life imprisonment without parole for those under the age of 18 at the time of their crimes violates the Eighth Amendment. The Court justi- fied this rule on the theory that offenders under the age of 18 No. 18-1114 27

have “diminished culpability” because of their “lack of ma- turity,” “underdeveloped sense of responsibility,” and char- acter that is less “well formed,” as compared with that of adults. Id. at 471. Miller upended many convictions for which the defendants had received life sentences. For those who had also been sentenced to consecutive terms of years, the latter sentences sprang into relevance with their post-Miller pro- ceedings. Miller is part of a long line of cases recognizing that the Constitution demands different penal treatment for young of- fenders—a group that up until now has been defined as those under the age of 18. See, e.g., Graham v. Florida, 560 U.S. 48 (2010); Roper v. Simmons, 543 U.S. 551 (2005). As the Miller Court noted, these decisions are grounded in science. Courts have paid heed to “developments in psychology and brain science [that] show fundamental differences between juvenile and adult minds” including the “parts of the brain involved in behavior control.” Miller, 567 U.S. at 471–72 (quoting Gra- ham, 560 U.S. at 68). For now, they are using the age of 18 as the relevant cut-off point, largely because of the scientific community’s assessments regarding the length of the devel- opmental period in the human brain. But science does not stand still, and there is no reason to think that it will do so going forward. The scientific commu- nity’s views on the development of the brain evolve all the time. One of the medical authorities on which the Supreme Court has relied most heavily on questions of neurological de- velopment is the American Association on Intellectual and Developmental Disabilities (AAIDD). Since Atkins v. Virginia, 536 U.S. 304 (2002), nearly every Supreme Court case concern- ing intellectual and developmental disabilities has drawn 28 No. 18-1114

significantly from the medical conclusions set forth in the AAIDD’s treatise, INTELLECTUAL DISABILITY: DEFINITION, CLASSIFICATION, AND SYSTEMS OF SUPPORTS (11th ed. 2010). See Moore v. Texas 137 S. Ct. 1039, 1048–53 (2017); Brumfield v. Cain, 576 U.S. 305, 308, 315, 319, 320 (2015) (citing the 10th edition); Hall v. Florida, 572 U.S. 701, 713 (2014); Atkins v. Virginia, 536 U.S. 304, 308 n.3, 317 n.22 (2002) (citing the 9th edition). Just this year, the AAIDD released the 12th edition of its treatise. See INTELLECTUAL DISABILITY: DEFINITION, DIAGNOSIS, CLASSIFICATION, AND SYSTEMS OF SUPPORTS (12th ed. 2021). In it, the Association defines the end of the human intellectual developmental period as “the age of 22”—not 18. See id. at 1, 13, & 32. See also Frequently Asked Questions on Intellectual Dis- ability, AAIDD https://www.aaidd.org/intellectual-disabil- ity/definition/faqs-on-intellectual-disability. Interestingly enough, this harmonizes the judgment of the scientific com- munity with federal law, which since 2000 has recognized 22 as the age at which neurological development ends. See 42 U.S.C. § 15002(8) (definitions for programs for individuals with developmental disabilities). Given the heavy emphasis the Supreme Court has placed on scientific evidence in this corner of its jurisprudence, the scientific community’s evolving views on the neurological de- velopmental period may prove to have wide ranging effects on the law. It is not at all fanciful to think that, at some point in the not-too-distant future, the Court might revise the Miller line of cases and push the relevant age at which the Eighth Amendment prohibits mandatory life sentences without pa- role to 22. Critically, the rule set forth in Miller is retroactive on collateral review. See Montgomery v. Louisiana, 136 S. Ct. 718, 732 (2016). Thus, if the Court, at any point during Ruiz’s imprisonment, were to revise the rule in Miller and raise the No. 18-1114 29

Eighth Amendment age-line to 22, Ruiz’s life sentences would become eligible for immediate resentencing. At this point, his consecutive 45-year sentence stemming from his section 924(c) convictions would take on immense practical signifi- cance. See United States v. Cephus, 684 F.3d 703, 710 (7th Cir. 2012). It is impossible to assess how likely such a change is, but the probability is certainly well above zero. (Many people would not have predicted the original Miller decision until it was issued.) That is all that matters for Ruiz’s case. Another possible change would be statutory: many people did not pre- dict the Fair Sentencing Act, 124 Stat. 2372, and the later First Step Act, 132 Stat. 5194, which taken together retroactively lowered many drug sentences. Such a legislative change could also affect Ruiz’s life sentences (and would be particu- larly difficult to present in a successive section 2255 motion). Beyond all this, we should remember the Court’s admonition in Sibron: courts are ill-equipped to “foretell what opportuni- ties might present themselves in the future for the removal of other convictions from an individual’s record” and so should err on the side of “eliminat[ing] the source of a potential legal disability” when the issue is squarely presented. Sibron, 392 U.S. at 56–57. That is precisely the case here.

III Finally, the majority suggests that we should not address Ruiz’s claims on the merits because they involve complicated issues of first impression. But Article III courts have a duty to decide cases before them, no matter how novel or complicated the issues may be. See Cohens v. Virginia, 19 U.S. (6 Wheat.) 264, 404 (1821) (“The judiciary cannot, as the legislature may, 30 No. 18-1114

avoid a measure because it approaches the confines of the constitution. We cannot pass it by because it is doubtful. With whatever doubts, with whatever difficulties, a case may be at- tended, we must decide it, if it be brought before us.”). In this case, neither harmless error nor the concurrent-sentence doc- trine permits us to avoid our “unflagging obligation” to de- cide the case before us. Colo. River Water Conservation Dist. v. United States, 424 U.S. 800, 817 (1976). This case is not a good candidate for avoidance in any event: the majority greatly exaggerates the difficulty of resolv- ing Ruiz’s merits claims. The government already has con- ceded that conspiracy to commit kidnapping is not a crime of violence, and so Ruiz’s Count 9 conviction is gone. Next, as the majority acknowledges, for Count 10 the jury expressly found only the predicate elements of simple kidnapping be- yond a reasonable doubt in its verdict, and we already have held that simple kidnapping is not a crime of violence. United States v. Jenkins, 849 F.3d 390, 395 (7th Cir. 2017), vacated, 138 S. Ct. 1980 (2018), reinstated sub nom. United States v. Jackson, 932 F.3d 556, 558 (7th Cir. 2019). That might well take care of Count 10. What’s left is Ruiz’s Count 11 conviction. To resolve that, we would need to apply the well-known modified cate- gorical approach to the elements of the predicate offense, see Shepard v. United States, 544 U.S. 13 (2005), a task we regularly undertake. These circumstances hardly merit avoidance. In my view, Ruiz’s life sentences do not excuse us from the duty to examine his firearms convictions under Davis, nor do those life sentences render any error in the firearms counts harmless, and so I respectfully dissent.