Summary
holding that "errors in applying advisory guidelines are procedural," and opining that a new rule that erroneously sentencing a defendant under current Guidelines rather than less punitive Guidelines in effect at the time of offense is a violation of the Ex Post Facto Clause would therefore not be applied retroactively under Teague
Summary of this case from United States v. AllenOpinion
No. 11–1245.
2013-07-31
Bernard HAWKINS, Petitioner–Appellant, v. UNITED STATES of America, Respondent–Appellee.
Julius L. Loeser, James R. Thompson, Jr., Thomas G. Weber, Sean G. Wieber (argued), Attorneys, Winston & Strawn LLP, Chicago, IL, for Petitioner–Appellant. Gary Bell, David E. Hollar (argued), Attorneys, Office of the United States Attorney, Hammond, IN, for Respondent–Appellee.
Julius L. Loeser, James R. Thompson, Jr., Thomas G. Weber, Sean G. Wieber (argued), Attorneys, Winston & Strawn LLP, Chicago, IL, for Petitioner–Appellant. Gary Bell, David E. Hollar (argued), Attorneys, Office of the United States Attorney, Hammond, IN, for Respondent–Appellee.
Before POSNER, KANNE, and ROVNER, Circuit Judges.
POSNER, Circuit Judge.
This opinion, a supplement to the panel opinion reported at 706 F.3d 820, explains why a majority of the panel does not believe that rehearing is warranted by the Supreme Court's decision in Peugh v. United States, –––U.S. ––––, 133 S.Ct. 2072, 186 L.Ed.2d 84 (2013), rendered after the panel opinion.
The panel had held that an error in calculating a defendant's guidelines sentencing range does not justify postconviction relief unless the defendant had, as in Narvaez v. United States, 674 F.3d 621 (7th Cir.2011), been sentenced in the pre- Booker era, when the guidelines were mandatory rather than merely advisory. Peugh holds that a sentence violates the Constitution's ex post facto clause if in calculating the defendant's advisory guidelines range (as the judge is required to do even though he can if he wants sentence the defendant outside that range) the judge had calculated the range in effect when he sentenced the defendant, rather than when the defendant committed the crime for which he's being sentenced, if the earlier range was lower (less punitive). 133 S.Ct. at 2088. The arguable significance of Peugh for the present case is that the Court held that an error in calculating a merely advisory guidelines range nevertheless invalidated the sentence.
The issue in this case differs from that in Peugh in several respects, however. One is that Peugh involved constitutional error—a violation of the ex post facto clause. Our case involves no claim of constitutional error—no claim for example that Hawkins's sentence exceeded the statutory maximum (as the sentence in Narvaez did, if mandatory guidelines are treated as equivalents of statutes, as the panel opinion in this case suggested they can be, 706 F.3d at 822). There is just a claim that the sentencing judge miscalculated the advisory guidelines range and might have given a lower sentence had he not miscalculated it. Peugh tells us only that the advisory nature of the guidelines in the present era, the Booker era, which allows the sentencing judge broad discretion, nevertheless does not excuse constitutional violations arising from the judge's miscalculating the applicable guideline.
Another difference between Peugh and the present case is that the two cases are governed by different legal standards. The standard governing ex post facto challenges, involved in Peugh, is not the same as the standard for postconviction relief for nonconstitutional errors, applicable to our case. Peugh holds that the ex post facto clause is violated when “a change in law creates a ‘significant risk’ of a higher sentence.” 133 S.Ct. at 2088. In contrast, postconviction relief requires a showing that “the error ‘had substantial and injurious effect or influence in determining the jury's verdict.’ Under this standard, habeas [corpus] petitioners ... are not entitled to habeas [corpus] relief based on trial error unless they can establish that it resulted in ‘actual prejudice.’ ” Brecht v. Abrahamson, 507 U.S. 619, 637, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993) (citations omitted). Postconviction review is therefore proper when for example the judge imposes a sentence that he had no authority to impose, as in Narvaez, since the consequence for the defendant in such a case is “actual prejudice”—an “injurious effect” on the judgment. But it doesn't follow that postconviction relief is proper just because the judge, though he could lawfully have imposed the sentence that he did impose, might have imposed a lighter sentence had he calculated the applicable guidelines sentencing range correctly.
Another reason to doubt the applicability of Peugh to our case is that the Supreme Court didn't say that the rule it was announcing was to have retroactive effect, and thus be subject to invocation even by persons whose sentences became final before June 10 of this year, when the Supreme Court handed down Peugh. Yet unless it is retroactive, Peugh can't help our petitioner, whose sentence became final seven years ago. A decision is retroactive if it is a “substantive” decision but (with immaterial exceptions) not if it's a “procedural” one. “New substantive rules generally apply retroactively. This includes decisions that narrow the scope of a criminal statute by interpreting its terms.... 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.... New rules of procedure, on the other hand, generally do not apply retroactively. They do not produce a class of persons convicted of conduct the law does not make criminal, but merely raise the possibility that someone convicted with use of the invalidated procedure might have been acquitted otherwise.” Schriro v. Summerlin, 542 U.S. 348, 351–52, 124 S.Ct. 2519, 159 L.Ed.2d 442 (2004) (emphasis in original; citations omitted). So since Peugh says that “failing to calculate the correct Guidelines range constitutes procedural error,” 133 S.Ct. at 2083, the implication is that the rule announced in Peugh won't be applied retroactively—and the statement in Schriro that a procedural error “merely raise[s] the possibility that someone convicted with use of the invalidated procedure might have been acquitted otherwise” is an exact description of the present case.
Now that we know that errors in applying the advisory guidelines are procedural, cases reinterpreting the advisory guidelines—including Chambers v. United States, 555 U.S. 122, 129 S.Ct. 687, 172 L.Ed.2d 484 (2009), on which our petitioner relies for his argument that his prior offense wasn't a “violent felony”—don't have retroactive application either, though there may be exceptions (none applicable to this case): Chambers may still be considered “substantive” when the pre-Chambers understanding of “violent felony” resulted in a sentence that exceeded either the statutory maximum under the Armed Career Criminal Act, Welch v. United States, 604 F.3d 408, 412 and n. 5 (7th Cir.2010), or (as in Narvaez ) the guidelines range when the guidelines were still mandatory. See also Brown v. Caraway, 719 F.3d 583, 586–88 (7th Cir.2013).
Our panel opinion does not deny that the district judge had committed an error that would be corrigible on direct review. But we found the social interest in a belated correction of the error outweighed by the social interest in the finality of judicial decisions, including sentences. About finality all that the panel dissent said was that “were we writing on a blank slate, we might argue against the majority's elevation of finality over fairness.” 706 F.3d at 828. That's vague. As the panel opinion pointed out, “an erroneous computation of an advisory guidelines sentence is reversible (unless harmless) on direct appeal; it doesn't follow that it's reversible years later in a postconviction proceeding.” Id. at 824. The panel dissent evinced no recognition of the importance of finality to an effective judicial system, or of the difficulty of balancing “fairness” (meaning what exactly?) against finality. Finality is an institutional value and it is tempting to subordinate such a value to the equities of the individual case. But there are dangers, especially if so vague a term as “fairness” is to be the touchstone.
Judicial systems that ignore the importance of finality invite unreasonable delay in the disposition of cases. A case in point is Brazil, which like a number of other countries, such as India, has an extraordinary problem of judicial delay—unbelievable really from a U.S. perspective. In India there are “30 million cases pending in various courts and an average time span of 15 years to get the dispute resolved through court system.” Vandana Ajay Kumar, “Judicial Delay in India: Causes & Remedies,” 4 Journal of Law, Policy & Globalization 16 (2012), www. iiste. org/ Journals/ index. php/ JLPG/ article/ view/ 2069 (visited July 24, 2013). In Brazil more than 70 percent of the cases pending in its courts in 2010 had been filed in earlier years—often many years earlier. See CNJ (Conselho Nacional de Justiça [National Council of Justice] ), Justice in Numbers 2010 6 (2010). Since 1988, “the backlog of cases in judicial dockets has multiplied by a factor of ten, and as a consequence trial delays have more than doubled.... Some cases have been pending since the 1940s.... Conservative statistics estimate the number of lawsuits awaiting final decision to be more than 50 million. Between 1995 and 1999, 32.2 million processes [cases] entered the Brazilian courts. However, only 22.6 million of these were decided during the same period. This leaves a deficit of almost 10 million processes left unjudged.” Augusto Zimmermann, “How Brazilian Judges Undermine the Rule of Law: A Critical Appraisal,” 11 International Trade & Business L.Rev. 179, 192 (2008).
The reasons for this delay are various, but one reason is that the Brazilian judicial system has only a weak concept of finality. Apart from allowing interlocutory appeals promiscuously, Brazil's judicial culture permits reopening (which is what our case involves) promiscuously. The costs in judicial overload are very great.
We are not Brazil or India. But we have to worry about delay in our federal judicial system as well, because of the difficulty of filling federal judicial vacancies and the increasing complexity of federal cases, which increases the time required for deciding them. Such costs must be kept in mind in deciding how generously to allow postconviction retraction of sentences. Fairness to victims of errors in guidelines calculation that might or might not have lengthened a sentence (or shortened it for that matter, thus conferring a windfall on the defendant) must be balanced against the harm to victims of judicial delay brought about by judges' neglect of the social interest in judicial finality. Furthermore, as the Supreme Court explained in Teague v. Lane, 489 U.S. 288, 310, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989) (citations omitted), “the application of new rules to cases on collateral review may be more intrusive than the enjoining of criminal prosecutions, for it continually forces the States to marshal resources in order to keep in prison defendants whose trials and appeals conformed to then-existing constitutional standards” (emphasis in original; citation omitted). In our case, it's the federal government rather than the states that have to marshal those resources. And the procedural error is in sentencing rather than at trial. But the point is the same—it's costly for government to have to defend sentences and resentence defendants long after the original sentences were imposed. ROVNER, Circuit Judge, dissenting from the denial of rehearing.
Since July 25, 2003, Bernard Hawkins has been sitting in a Federal Correctional Institution, where he is scheduled to remain for approximately twelve-and-a-half years. It is uncontroverted that the district court erred when it calculated his sentence using the career offender enhancement, and had the court not erred, his calculated sentencing range would have been approximately ten times less—somewhere in the range of 15–21 months. Yet despite the known and conceded error, we are told that for the sake of principles of finality, Hawkins must remain in prison for the entire 151–month sentence. My dissent to the panel opinion elucidated the reasons why I believe this was the wrong result. In the interim, the Supreme Court issued a decision in Peugh v. United States, ––– U.S. ––––, 133 S.Ct. 2072, 186 L.Ed.2d 84 (2013), addressing the question of how appellate courts should view the effect of errors that sentencing courts make when they select the incorrect United States Sentencing Guideline as a starting point. In light of that decision, and for the reasons articulated in the dissent to the panel opinion, I believe it is our duty to reconsider Mr. Hawkins' case and therefore I respectfully dissent from the denial of rehearing.
Before the Supreme Court issued its decision in Peugh, we knew with indisputable certainty the following legal facts: (1) The district court erred when it labeled Mr. Hawkins a career offender. He was not. Chambers v. United States, 555 U.S. 122, 130, 129 S.Ct. 687, 172 L.Ed.2d 484 (2009).(2) This is the type of error that can be corrected retroactively. Narvaez v. United States, 674 F.3d 621, 625 (7th Cir.2011).(3) Such an error constitutes a miscarriage of justice and could be corrected on post-conviction review, at least when the Guidelines were mandatory. Id. at 629.
The one and only question for which we lacked a definitive answer was whether the holding in Narvaez could be applied in post-conviction cases after Booker—that is, when the Guidelines were no longer mandatory. In the panel opinion, this was the critical (and only) distinction between Narvaez and this case. Hawkins v. United States, 706 F.3d 820, 824–25 (7th Cir.2013).
The Supreme Court in Peugh, however, rejected just this distinction, instructing that the advisory nature of the Guidelines and the presence of discretion do not alleviate the infirmities that arise when a sentencing court chooses the improper Guideline range as a starting point. Peugh, 133 S.Ct. at 2086. It is true that Peugh involved a sentence that violated the Ex–Post Facto Clause of the Constitution, but the reasoning of Peugh broadly addressed the exact question we are faced with here—whether the harm caused by an error in sentencing is somehow mitigated when the Guidelines are merely advisory as opposed to mandatory. The government's position in Peugh mirrors that of the panel opinion in this case—that an advisory system mitigates the harm because such an error merely creates a risk and not a guarantee of injury. Now that the Guidelines are advisory, the argument goes, they lack sufficient legal force and effect to attain the status of a law, as they arguably had before.
The Supreme Court handily rejected these arguments, reasoning that the Guidelines are much more than an advisory set of guideposts that allow a district court to sentence as if starting with a blank slate. First, the Peugh court reasoned, district courts are required to consult the Guidelines and must first correctly calculate the Guidelines sentence before beginning any analysis about deviations. Peugh, 133 S.Ct. at 2079, 2083, 2087. The wagon wheels of the sentencing machinery are not rolling off into an open field of grass. This strict requirement—to begin all sentencing proceedings by correctly calculating the applicable sentencing range ( see id.; Gall v. United States, 552 U.S. 38, 49, 128 S.Ct. 586, 169 L.Ed.2d 445 (2007))—aligns the wheels of the sentencing machinery into the deep grooves created by the Guidelines. If a court does choose to deviate, it must consider the extent of the deviation and ensure that the justification is sufficiently compelling to support the degree of variance. Peugh, 133 S.Ct. at 2087 (citing Gall, 552 U.S. at 50, 128 S.Ct. 586.). Furthermore, a court of appeals is permitted to presume that a sentence that comports with the Guidelines is reasonable. Id. The reviewing court also, in considering whether the district court's sentence was reasonable, weighs the extent of any departure from the Guidelines in determining whether the district court abused its discretion. Id. The effect of the Guidelines, therefore, is strong and anchoring. (And it was particularly strong in this case, see Hawkins, 706 F.3d at 831 (Rovner, J. dissenting) (describing the district court's verging-on-improper allegiance to the Guidelines).).
Moreover, as the Peugh court points out, empirical evidence supports the view that the Sentencing Guidelines greatly influence the sentences imposed by judges. Peugh, 133 S.Ct. at 2084. Even under the new advisory Guidelines system, district courts have, in the vast majority of cases, imposed either within-Guidelines sentences or sentences that depart from the Guidelines based on the government's motion. Id. In less than 20% of cases since 2007 have district courts imposed above or below-Guidelines sentences absent a government motion. Id. Moreover, the data indicates that when the Guidelines range moves up or down, the sentences move with it. Id. Thus the initial miscalculation of the sentencing Guidelines sets the wheels of sentencing into the tracks in which they will stay in the vast majority of cases.
Nowhere in Peugh's lengthy discussion about the magnetism of the advisory system does the Court argue that the pull can only cause sufficient harm if the miscalculationcauses or results from a constitutional violation. This is not to say that there are not legitimate differences between constitutional violations and other errors. There are indeed. This is why most sentencing errors are not cognizable on collateral review in the first place. Welch v. United States, 604 F.3d 408, 412 (7th Cir.2010). But this court and others have held that in exceptional cases “the claimed error of law was a fundamental defect which inherently results in a complete miscarriage of justice and [where] it presents exceptional circumstances where the need for the remedy afforded by the writ of habeas corpus is apparent.” Davis v. United States, 417 U.S. 333, 346, 94 S.Ct. 2298, 41 L.Ed.2d 109 (1974) (quoting Hill v. United States, 368 U.S. 424, 428, 82 S.Ct. 468, 7 L.Ed.2d 417 (1962)); see also Narvaez, 674 F.3d at 628.
The Peugh court had no reason to address whether its theory (that advisory Guidelines do not mitigate the harm) also applied to non-constitutional miscarriages of justice that are cognizable on post-conviction review, but it would be hard to imagine what the reasoning might be for distinguishing them. In Peugh's case, the harm occurred when the district judge chose the current Sentencing Guidelines rather than the lower Guidelines in effect at the time Peugh committed the crime. In Hawkins, the error occurred when the district court judge calculated Hawkins' sentence as if he were a career offender, when he was not. In both cases, the question is the same—does the advisory nature of the Guidelines mitigate the harm caused by an initial improper calculation of the Guidelines? The nature of the harm is relevant in determining whether the harm can be addressed on post-conviction review in the first place. But after that, it is hard to imagine how the advisory system has a different effect on ex post facto violations where there is a “significant risk” of a higher sentence ( Peugh, 133 S.Ct. at 2088) than on errors of law which inherently result in a complete miscarriage of justice. ( Davis, 417 U.S. at 346, 94 S.Ct. 2298,Narvaez, 674 F.3d at 628).
The Peugh court rejected the idea that a Guidelines error causes only a potential prejudice, stating “that a district court may ultimately sentence a given defendant outside the Guidelines range does not deprive the Guidelines of force as the framework for sentencing.” Peugh, 133 S.Ct. at 2076. The Guidelines exert a strong controlling influence notwithstanding the fact that they are no longer mandatory. Id. at 2084, 2086–88.
The majority finds itself dissatisfied with the dissent's brief discussion about the limitations of post-conviction review. But in Narvaez we said that when a court labels a defendant as a career offender when he is not, (1) the court has erred, ( Narvaez, 674 F.3d at 625); (2) the error constitutes a miscarriage of justice ( id. at 629); and (3) that error is therefore cognizable on post-conviction review ( id. at 625). Narvaez resolved any question as to whether a sentencing error could be redressable on post conviction relief. This ended the matter and thus our discussion was initially succinct.
The majority reasons that because the Guidelines were binding before Booker, we should view Narvaez's sentence, which exceeded the Guidelines, as a sentence that exceeded the statutory maximum. But the Guidelines, even before Booker, were not the equivalent of statutes. Departures, for example, allowed courts to sentence above and below the Guidelines but did not permit changes beyond the actual statutory range. See, e.g., United States v. Mancari, 463 F.3d 590, 597 (7th Cir.2006).
A more detailed explanation would have noted that although it is true that, in general, sentencing errors are not cognizable in § 2255 proceedings, the Supreme Court has never set forth a per se rule that a sentencing error could never rise to the level of a miscarriage of justice. Such relief is reserved for exceptional cases, but is available. In Davis v. United States, the Supreme Court found a non-constitutional, nonjurisdictional error to be a miscarriage of justice where a subsequent change in law rendered the defendant's conviction and sentence unlawful, and specifically rejected the notion that only claims of constitutional dimension are cognizable under § 2255. Davis, 417 U.S. at 345, 346–47, 94 S.Ct. 2298. And as I noted in my dissent, this is not a run-of-the-mill sentencing error. Few Guidelines sentences have as profound an effect as the career offender label. And when the magnitude of error creates a complete miscarriage of justice, that error begins to look much like a constitutional error. After all, when we label an error as a complete miscarriage of justice, are we not saying that the defendant has not received any of the justice that he is due? This is a profound error indeed and, for these purposes, should be treated in the same manner as a constitutional error.
Turning to the dissent's discussion of retroactivity, for the sake of economy, I will rely on the long string cite of cases I provided in the dissent, all of which held that Chambers (and its closely related ancestor, Begay ) apply retroactively on collateral review. Hawkins, 706 F.3d at 828 (Rovner, J., dissenting). Peugh does indeed state that “failure to calculate the correct Guidelines range constitutes procedural error,” ( Peugh, 133 S.Ct. at 2080) and ordinarily procedural errors cannot be corrected retroactively. But it seems to me that Peugh's reference to “procedural error” refers to errors made in the procedure required in post- Booker sentencing—that the first procedural step any sentencing court must take is to correctly calculate the proper Guidelines range. Gall v. United States, 552 U.S. 38, 51, 128 S.Ct. 586, 169 L.Ed.2d 445 (2007); Peugh, 133 S.Ct. at 2080. This does not mean that the subsequent substantive error—creating a class of persons convicted as career offenders when they are not—is not redressable retroactively. I see no reason why Peugh's brief reference to procedural error alters this and other courts' assessments that the rule announced in Chambers was substantive and applies retroactively.
And even if the Peugh holding could not be applied retroactively (although I think it can), the outcome here does not require a retroactive application of Peugh's holding which addresses the specific ex post facto violation in that case. As the dissent initially pointed out, Hawkins had all the legal pieces of the puzzle necessary to warrant a remand for resentencing—an error constituting a miscarriage of justice redressable on postconviction review. See Hawkins, 706 F.3d at 826–27, 832 (Rovner, J. dissenting). It is the widely applicable rationale of Peugh as opposed to the specific holding in cases of ex post facto violations that I am suggesting should inform a decision in Hawkins—that is, that advisory Guidelines do not mitigate the harm caused by errant sentencing calculations that have extraordinary effect.
And what of the importance of finality—an issue the original dissent addressed in its discussion of Rozeier and Meirovitz, and its attempts to limit the applicability of the holding in the name of finality? See Hawkins, 706 F.3d at 828, 832 (Rovner, J. dissenting). As a philosophical matter, I believe that fairness is the lifeblood of our system of justice, and more specifically, justice requires the ability to rectify substantial uncontroverted judicial errors that cause significant injury. This is why in our anthropomorphization of Justice, she is wearing a blindfold, and not running shoes. If allowing fairness to prevail in limited situations involving grave miscarriages of justice subverts finality, then I suppose I agree with my dissenting brother in the Eleventh Circuit who, in a similar case, decried the “elevation of form over substance; of finality over fairness.” Rozier v. United States, 701 F.3d 681, 690 (11th Cir.2012) (Hill, J. dissenting). “Due process,” Judge Hill wrote,
is the defining virtue of our system of criminal justice. But we should ask ourselves why. Is it because it achieves finality? Or is it because we believe that, more often than not, we will reach a correct result where certain process is due the criminal defendant. The goal is a correct result—not simply the provision of process. To be sure, we do not guarantee a correct result. But where all know the result is error, to adhere to the process as though it were the end goal is unfair in the purest sense of the word.
Id. at 690–91.
It simply cannot be that the judicial system is incapable of balancing what the majority calls the “vague” notion of fairness against finality. Rectifying errors does indeed cause some amount of delay (although as I will address in a moment, in this case the delay would be minimal). But if finality were our only or even the more important institutional goal, we would not permit any postconviction relief at all. In fact, we could do away with direct appeals altogether. After all, all of these lines of finality are drawn based on theories involving the balance between correcting unjust error and finality. The question is where to draw the line—that is, how to balance finality in cases, for example, where an uncontroverted error has increased a defendant's sentence tenfold and where the error could not have been raised at trial or on direct review.
As my brother in the Eighth Circuit pointed out in a case nearly identical to this one (and quoting another dissent from the Eleventh Circuit), denying relief for the sake of finality:
does not build confidence in our court system because this looks to the world like a court refusing to acknowledge or make amends for its own mistake. Second, to the extent that there have been administrative costs and delay in considering [the petitioner's] request for relief, they have already been incurred, and we need only grant him that relief to end his very expensive incarceration. Third, because the only issue before us is a purely legal one, there is no evidence we must consult. Thus spoilation [sic] is not a concern. And finally, [the petitioner's] case presents no comity concerns insofar as he seeks to correct a sentence imposed in federal court and not by the state.
Sun Bear v. United States, 644 F.3d 700, 712 (8th Cir.2011) (Melloy, J., dissenting) (quoting Gilbert v. United States, 640 F.3d 1293, 1334 (11th Cir.2011) (Martin, J., dissenting)).
Moreover, we have to ask ourselves how much it would impose on finality to correct the type of error we lament today. We have rectified sentencing errors before by ordering limited remands to address only the narrow issue that warrants reconsideration. In Paladino, we created a system in which the court of appeals could order a limited remand to permit a sentencing judge to determine whether she would have re-imposed her original sentence, knowing that the Guidelines were merely discretionary. United States v. Paladino, 401 F.3d 471, 484 (7th Cir.2005). We used a similar mechanism to allow a district court judge to consider whether he would have imposed the same sentence knowing that he had discretion to depart from the 100:1 crack cocaine to powder cocaine disparity. United States v. Taylor, 520 F.3d 746, 747–48 (7th Cir.2008). And in United States v. Redmond, 667 F.3d 863, 876 (7th Cir.2012), we remanded where the district court suggested that the defendant's status as a career offender was a significant factor in its sentence, and it was not clear that the court recognized its complete discretion to deviate from the Guidelines' career-offender calculation. Limited remands allow a sentencing court to correct errors with minimal use of time and resources. In this situation, Hawkins' case could be remanded back to the district court judge for a recalculation using the Guidelines applicable to a defendant who had not been classified as a career offender. All other sentencing factors and § 3553 factors would remain the same. The marginal cost to the court system would be small (in this case) compared to the incredible injury that would befall a non-career-offender sentenced at career-offender levels (such as Hawkins).
This court is no stranger to this balancing of fairness and finality. As we noted in Paladino,
the entry of an illegal sentence is a serious error routinely corrected on plain-error review. To tell a defendant we know your sentence would have been 60 months shorter had the district judge known the guidelines were merely advisory, because he's told us it would have been—but that is your tough luck and you'll just have to stew in prison for 60 additional months because of an acknowledged violation of the Constitution—would undermine the fairness, the integrity, and the public repute of the federal judicial process.
Paladino, 401 F.3d at 483 (quoting United States v. Pawlinski, 374 F.3d 536, 541 (7th Cir.2004)).
Of course the violation in that case was of a constitutional nature and came to the court of appeals on direct review. And these distinctions make a difference. See, e.g., Hawkins, 706 F.3d at 829–30 (Rovner, J., dissenting). This is precisely why this court has held, and I wholeheartedly agree, that ordinary sentencing errors are not cognizable on post-conviction (§ 2255) review. But when the magnitude of the error creates a complete miscarriage of justice, the error begins to approximate that of a constitutional error. How can we tell a prisoner that he has received no justice but yet considerations of finality prevent any remedy? Surely the solution to the backlogs in Brazil and India that the majority describes cannot be to offer no justice at all—and that is precisely what happens when we allow a “complete miscarriage of justice” to stand uncorrected.
As Judge Hill of the Eighth Circuit stated in another case involving an errantly imposed career offender enhancement:
I reluctantly conclude that our court is determined to deny relief to every confined habeas petitioner whose sentence has been unlawfully enhanced under either the career offender guideline or the armed career criminal statute. We have repeatedly held that procedural rules deprive us of the ability to correct an enhancement that we all agree is error because it was not authorized by law when imposed. To the petitioner, who is serving five, ten or even fifteen years more than he would be in the absence of the error, we say, “Sorry. We know your enhancement was error, but there is nothing we can do. Our hands are tied by procedural rules. We cannot fix this.”
We do this in the name of “finality.” We say we are protecting the integrity of the Great Writ; we cannot go about correcting old mistakes or no conviction or sentence will ever be final.
Of course, finality is desirable. There was a time when there appearedto be no finality in our habeas procedures. The rules we adopted to introduce some finality into the habeas process were long overdue.
But finality must not be our highest goal. The Great Writ is enshrined in our Constitution because we believe that no one may be deprived of life, liberty, or property by the government in violation of law. If a petitioner can show that he is illegally incarcerated, he is entitled to release. Fairness requires it. Justice is the ultimate goal in the grant of the Writ.
Rozier, 701 F.3d at 689–90 (Hill, J., dissenting).
The correct result here, however, does not rely on my personal view that justice and fairness must prevail in our legal system, nor the similar positions of other judges within and outside this circuit. It relies only on a logical legal path based on the precedent of this court and the Supreme Court. That path is simple: The district court erred in finding that Hawkins was a career criminal. Such an error constitutes a miscarriage of justice that can be remedied via petition for relief under § 2255, and, regardless of their advisory nature, the Sentencing Guidelines are influential enough that errors in their calculation cause harm.
The Supreme Court's reasoning in Peugh—which is consistent with the tenets of fairness that are the quintessence of our system of justice—calls for us to rehear this case.