Nakia William GARRUS, Appellant v. * SECRETARY OF the PENNSYLVANIA DEPARTMENT OF CORRECTIONS; The District Attorney of the County of Philadelphia; The Attorney General for the State of Pennsylvania * Pursuant to Rule 43(c), Fed. R.App. P.
OPINION OF THE COURT
William Garrus (“Garrus”), a Pennsylvania state prisoner, seeks federal habeas relief under 28 U.S.C. § 2254, as amended by the Anti–Terrorism and Effective Death Penalty Act of 1996 (“AEDPA”). Garrus was found guilty in state court of voluntary manslaughter in 2001. At sentencing, the judge increased his sentence beyond the statutory maximum based on 42 Pa. Cons.Stat. § 9714, Pennsylvania's “three strikes” law. In order to do so, the judge made a judicial finding that Garrus had previously been convicted of burglarizing an occupied building, when, in fact, he had only pled guilty to, and been convicted of, second degree burglary (which, under Pennsylvania law, necessarily requires that the burglarized building was unoccupied). In the habeas petition now before us, Garrus argues that this judicial factfinding violated the rule of Apprendi v. New Jersey, requiring that “[o]ther than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt.” 530 U.S. 466, 490, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000). The District Court denied the petition on the basis that the highest state court determination upholding Garrus's sentence was not contrary to or an unreasonable application of clearly established Federal law.
Key to our determination is a single question: whether, pursuant to AEDPA, the state court unreasonably applied Apprendi by allowing Garrus to be sentenced beyond the statutory maximum based on a judicial finding that Garrus burglarized an occupied building, despite his plea to the contrary.1 For the reasons set forth below, we hold that the state court determination upholding Garrus's sentence was objectively unreasonable, and that Garrus is entitled to habeas relief. Accordingly, we will reverse the order of the District Court.
On February 10, 2000, Garrus was at the home of his girlfriend, Toi Bryant, with whom he has a daughter, when Bryant's ex-boyfriend, Charles Goode, showed up at the residence. Shortly after Goode arrived, an argument ensued between him and Bryant involving their child. As the argument escalated, Garrus took his daughter upstairs. He later returned downstairs to the kitchen where Goode and Bryant were arguing. After Garrus and Goode exchanged some heated words, a fight broke out between the two men. Garrus picked up a kitchen knife and stabbed Goode several times in the chest area. Goode broke away from Garrus and ran upstairs to the bathroom. Garrus left the home. When police arrived, both Goode and Bryant identified Garrus as the attacker. Goode later died from his injuries.
Following a jury trial in the Philadelphia County Court of Common Pleas, Garrus was convicted on March 9, 2001, of voluntary manslaughter and possession of an instrument of crime. A conviction for voluntary manslaughter carries a maximum sentence of 20 years of imprisonment. 18 Pa. Cons.Stat. §§ 1103, 2503(c). However, the Commonwealth of Pennsylvania (the “Commonwealth”) notified the trial court that it would seek a sentencing enhancement under Pennsylvania's “three strikes” law, which requires a judge to sentence a defendant to a minimum term of 25 years' imprisonment (and a maximum of up to life imprisonment) if the defendant was previously convicted of two or more separate “crimes of violence.” 42 Pa. Cons.Stat. § 9714. “Crimes of violence” may include first degree burglary (defined as, inter alia, burglary of an occupied structure), but do not include second degree burglary (defined as, inter alia, burglary of an unoccupied structure).2
On April 30, 2001, the trial court held the first of two sentencing hearings. The Commonwealth offered evidence that Garrus had three prior convictions for “crimes of violence:” (1) a January 1995 conviction for first-degree robbery; (2) a February 1995 conviction for first-degree robbery; and (3) a February 1997 plea and conviction for second-degree burglary. The Commonwealth argued that although Garrus had only pled guilty to second degree burglary in 1997, the trial court should consider a police report and witness statements regarding that burglary to find instead that the building Garrus had burglarized in 1997 was occupied. On the basis of the police report and witness statements, the trial court found at the second sentencing hearing on May 8, 2001, that Garrus's prior 1997 conviction for second-degree burglary constituted a crime of violence under § 9714. Commonwealth v. Garrus, June Term 2000 No. 0092, slip op. at 17 (Phila.Cnty.Ct.Com.Pl. Jan. 16, 2002). Additionally, it found that the two 1995 robberies constituted two separate crimes of violence, and that his conviction for voluntary manslaughter was his fourth crime of violence. Id. The trial court sentenced Garrus to a term of 25 to 50 years' imprisonment. Id.
Garrus appealed his conviction and sentence to the Superior Court of Pennsylvania.3 He argued that, by definition, his 1997 second-degree burglary conviction did not constitute a “crime of violence” under Pennsylvania's “three strikes” law. The Superior Court of Pennsylvania adopted the trial court's opinion in its entirety. Commonwealth v. Garrus, No. 2592 EDA 2001 (Pa.Super.Ct. Oct. 18, 2002). The Supreme Court of Pennsylvania granted Garrus's petition seeking allocatur on the sole issue of Garrus's sentence, Commonwealth v. Garrus, 572 Pa. 438, 817 A.2d 455 (Pa.2003), but later dismissed the appeal in September 2003 as having been improvidently granted. Commonwealth v. Garrus, 574 Pa. 655, 832 A.2d 1063 (Pa.2003).
In April 2004, Garrus filed a pro se petition for postconviction relief in the Philadelphia County Court of Common Pleas, pursuant to Pennsylvania's Post Conviction Relief Act (“PCRA”), 42 Pa. Cons.Stat. § 9541 et seq., challenging his sentence. He argued that the trial court, at sentencing, violated his constitutional rights, as articulated in Apprendi, 530 U.S. at 476, by relying on the police report and victim statement to determine that his 1997 burglary conviction was a “crime of violence” under Pennsylvania's “three strikes” law. Garrus also asserted that state law required his two 1995 robbery convictions to be considered as one “crime of violence” under the “three strikes” law. As a result, Garrus claimed that he had, at most, one prior conviction for a “crime of violence,” rendering him ineligible for the 25–year mandatory minimum sentence he had received.
Garrus's PCRA petition was dismissed in February 2005. A month later, the Supreme Court rendered a decision in Shepard v. United States, 544 U.S. 13, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005), in which the Court held that a sentencing court is not permitted to consider police reports in determining whether a prior conviction constituted a “violent crime” under the Armed Career Criminal Act (“ACCA”), 18 U.S.C. § 924, a federal recidivism statute. Id. at 16, 26.
On appeal from the denial of his PCRA petition, Garrus argued that the sentencing court's consideration of the police report and victim statement, with respect to his 1997 burglary conviction, violated his rights under Apprendi and Shepard, and he again argued that his two 1995 robbery convictions constituted one “crime of violence.” The Superior Court of Pennsylvania affirmed the denial of his PCRA petition. Commonwealth v. Garrus, No. 961 EDA 2005 (Pa.Super.Ct. Apr.18, 2006). In its opinion, the Superior Court acknowledged that Apprendi requires that “any fact (other than prior conviction) that increases the maximum penalty for a crime must be charged in an indictment, submitted to a jury, and proven beyond a reasonable doubt.” Id. at 8–9 (quoting Apprendi, 530 U.S. at 476). However, it determined that the prior conviction exception permitted the sentencing judge to find that Garrus had burglarized an occupied building despite his plea to the contrary, and to sentence him beyond the statutory maximum accordingly. The Superior Court also determined that the Supreme Court's decision in Shepard did not affect the validity of the prior conviction exception established in Apprendi. Finally, on the basis that the 1997 burglary constituted a “crime of violence,” the Superior Court declined to reach the issue of whether the two 1995 robberies constituted separate crimes. The Supreme Court of Pennsylvania denied Garrus's petition seeking further appellate review. Commonwealth v. Garrus, 906 A.2d 639 (Pa. Aug. 29, 2006) (table).
On January 16, 2007, Garrus filed a petition for writ of habeas corpus in the United States District Court for the Eastern District of Pennsylvania, pursuant to 28 U.S.C. § 2254. Garrus raised the same challenges to both his 1997 burglary conviction and the two 1995 robbery convictions. The Magistrate Judge denied Garrus's petition, Garrus v. Mazurkiewicz, No. 07–187 (E.D.Pa. May 28, 2008), and concluded that it was precluded from granting federal habeas relief because the sentencing court's consideration of the police report and victim statement in determining whether Garrus had two prior convictions for crimes of violence under Pennsylvania's “three strikes” law was not contrary to, or an unreasonable application of, clearly established Supreme Court precedent. Id. at 38–39.
The Magistrate Judge also concluded that Garrus's argument that his two 1995 robbery convictions should be considered as one for purposes of the recidivism statute was therefore moot, as Garrus still had at least two prior convictions for crimes of violence—the 1997 burglary conviction and at least one of the 1995 robbery convictions. In any event, the Magistrate Judge reasoned, Garrus's challenge to the 1995 robbery convictions involved an interpretation of state law and could not be considered by federal courts in a habeas petition.
Nevertheless, the Magistrate Judge recommended that the District Court issue a certificate of appealability to this Court, limited to the issue involving the 1997 burglary conviction. The District Court adopted the Magistrate Judge's recommendation, with the exception that it rejected the Magistrate Judge's determination that sufficient grounds existed to issue a certificate of appealability. Garrus v. Johnson, No. 07–187 (E.D.Pa. Aug. 18, 2009). Garrus appealed, and on March 12, 2010, we granted a certificate of appealability to consider the issue of whether the state sentencing court violated Garrus's constitutional rights by labeling the 1997 second-degree burglary conviction a “crime of violence” for purposes of applying Pennsylvania's “three strikes” law.
The District Court had jurisdiction to consider Garrus's habeas petition pursuant to 28 U.S.C. § 2254, as the claim raised in the petition has been properly exhausted in the state courts. See Alston v. Redman, 34 F.3d 1237, 1242 (3d Cir.1994).4 We have jurisdiction to review the District Court's judgment pursuant to 28 U.S.C. §§ 1291 and 2253. We exercise de novo review over a District Court's denial of habeas relief, Vega v. United States, 493 F.3d 310, 314 (3d Cir.2007), and apply the highly deferential AEDPA standard, which, in this case, precludes us from granting habeas relief unless the challenged state court decision was “contrary to, or involved an unreasonable application of, clearly established Federal law [.]” 28 U.S.C. § 2254(d)(1).
Core to a criminal defendant's constitutional rights is the principle that “criminal convictions [must] rest upon a jury determination that the defendant is guilty of every element of the crime with which he is charged, beyond a reasonable doubt.” United States v. Gaudin, 515 U.S. 506, 510, 115 S.Ct. 2310, 132 L.Ed.2d 444 (1995); see also In re Winship, 397 U.S. 358, 364, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970). In Apprendi, the Supreme Court held that this same right applies to a sentencing factor that would increase a defendant's sentence beyond the statutory maximum because, like an element of a separate crime, such a sentencing factor results in a higher sentence than that which could be prescribed for the original crime. 530 U.S. at 476, 490. Thus, there is no “principled basis” for treating such a sentencing factor differently than an element of a crime. Id. at 476. The exception is that a judge can increase the sentence beyond the statutory maximum based on “the fact of a prior conviction” because a prior conviction has already been established through procedural safeguards. Id. at 488–90; Jones v. United States, 526 U.S. 227, 249, 119 S.Ct. 1215, 143 L.Ed.2d 311 (1999) (“[A] prior conviction must itself have been established through procedures satisfying the fair notice, reasonable doubt, and jury trial guarantees.”).
These precepts are “rooted in longstanding common-law practice [.]” Cunningham v. California, 549 U.S. 270, 281, 127 S.Ct. 856, 166 L.Ed.2d 856 (2007). Apprendi recounted their historical context, noting that early on in our history, judges had “very little explicit discretion in sentencing.” 530 U.S. at 479. Any “circumstances mandating a particular punishment” had to be charged to the jury in the indictment; there was “[no] distinction between an ‘element’ of a felony offense and a ‘sentencing factor[.]’ “ Id. at 478, 480. In the 19th century, this idea began to shift, “from statutes providing fixed-term sentences to those providing judges discretion within a permissible range[.]” Id. at 481. Crucially, this shift “has been regularly accompanied by the qualification that [such judicial] discretion was bound by the range of sentencing options prescribed by the legislature.” Id.
By the late 20th century, the Supreme Court began to address state laws that increased a defendant's punishment based on factors found at sentencing, rather than based on factors found at trial. In McMillan v. Pennsylvania, 477 U.S. 79, 106 S.Ct. 2411, 91 L.Ed.2d 67 (1986), the Supreme Court, “for the first time, coined the term ‘sentencing factor’ to refer to a fact that was not found by a jury but that could affect the sentence imposed by the judge.” Apprendi, 530 U.S. at 485 (construing McMillan ). Consistent with longstanding constitutional principles, McMillan held that a “sentencing factor” must at times be found by a jury because “(1) constitutional limits exist to States' authority to define away facts necessary to constitute a criminal offense, and (2) ․ a state scheme that keeps from the jury facts that ‘expos[e] [defendants] to greater or additional punishment,’ may raise serious constitutional concern[s].” Apprendi, 530 U.S. at 486 (quoting McMillan, 477 U.S. at 85–88) (internal citations omitted).
A decade later, the Supreme Court decided Almendarez–Torres v. United States, 523 U.S. 224, 242–47, 118 S.Ct. 1219, 140 L.Ed.2d 350 (1998), a case in which the defendant was convicted for his presence in the United States after being deported, a violation which normally carried a maximum of two years' imprisonment. 523 U.S. at 227. Prior to his initial deportation, Almendarez–Torres was convicted of three aggravated felonies, and according to the sentencing statute, where the original deportation had occurred subsequent to an aggravated felony, the court could impose a maximum sentence of up to twenty years' imprisonment. Id. at 227, 229. Almendarez–Torres argued that he could not be subject to more than two years' imprisonment, because his indictment had not mentioned his earlier aggravated felony convictions. Id. at 227. Importantly, there was no contention that the defendant had not been convicted of the three aggravated felonies, but only that the fact of his prior convictions should have been included in the indictment and charged to the jury. Id. Thus, Almendarez–Torres had earlier been convicted of the facts necessary for the sentencing enhancement, “pursuant to proceedings with substantial procedural safeguards of their own[.]” Apprendi, 530 U.S. at 488. In other words, by judicially finding that Almendarez–Torres's three prior convictions existed, the sentencing judge did not “change a pre-existing definition of a well-established crime, nor ․ ‘evade’ the Constitution, either by ‘presuming’ guilt or ‘restructuring’ the elements of an offense.” Almendarez–Torres, 523 U.S. at 246.
The Supreme Court found that the fact of Almendarez–Torres's prior convictions did not have to be charged to the jury because the constitutional limitations articulated in McMillan did not apply .5 Almendarez–Torres, 523 U.S. at 228, 242–43. As the Supreme Court later explained in Apprendi: “[b]oth the certainty that procedural safeguards attached to any ‘fact’ of prior conviction, and the reality that Almendarez–Torres did not challenge the accuracy of that ‘fact’ in his case, mitigated the due process and Sixth Amendment concerns otherwise implicated in allowing a judge to determine a ‘fact’ increasing punishment beyond the maximum of the statutory range.” 530 U.S. at 488.
A year after Almendarez–Torres, the Supreme Court explained that Almendarez–Torres's recidivism exception was permitted because “a prior conviction must itself have been established through procedures satisfying the fair notice, reasonable doubt, and jury trial guarantees.” Jones, 526 U.S. at 249. In Jones, the trial court gave the defendant a sentence beyond the statutory maximum based on a judicial finding that the carjacking offense he was convicted of involved “serious bodily injury.” Id. at 230–31. The Supreme Court found that the carjacking statute required a jury, rather than a judge, to determine whether the crime involved “serious bodily injury,” citing “grave” constitutional questions that would arise if the statute were to be interpreted otherwise. Id. at 231, 239. Although Jones did not actually reach the constitutional issue, its explanation of Almendarez–Torres and its constitutional discussion were significant, especially given that Apprendi explicitly “confirm[ed] the opinion ․ expressed in Jones [,]” 530 U.S. at 490, that “under the Due Process Clause of the Fifth Amendment and the notice and jury trial guarantees of the Sixth Amendment, any fact (other than prior conviction) that increases the maximum penalty for a crime must be charged in an indictment, submitted to a jury, and proven beyond a reasonable doubt.” 530 U.S. at 476 (quoting Jones, 526 U.S. at 243 n. 6).
Against this historical backdrop, the Supreme Court decided Apprendi, where a defendant pled guilty to unlawful possession of a weapon, a second-degree offense punishable by up to ten years' imprisonment. Despite the defendant's plea and conviction for a second-degree offense, the trial court sentenced him to twelve years' imprisonment, a “punishment identical to that ․ provide[d] for crimes of the first degree,” based on a judicial finding pursuant to a New Jersey hate crime sentencing enhancement, “that the defendant's ‘purpose’ for unlawfully possessing the weapon was ‘to intimidate’ his victim on the basis of a particular characteristic the victim possessed.” 530 U.S. at 468–69, 471, 491. The Apprendi Court began by stating that “[a]t stake in this case are constitutional protections of surpassing importance: the proscription of any deprivation of liberty without ‘due process of law,’ Amdt. 14, and the guarantee that ‘[i]n all criminal prosecutions, the accused shall enjoy the right to a speedy and public trial, by an impartial jury[.]’ “ 530 U.S. at 476–77. It then discussed several centuries of precedent, summarizing that a criminal defendant may not be “expose[d] ․ to a penalty exceeding the maximum he would receive if punished according to the facts reflected in the jury verdict alone.” Id. at 482–83.
Apprendi explained that Almendarez–Torres was “at best an exceptional departure from ․ historic practice” and a “limited” and “narrow” exception based on “unique facts[.]” Id. at 487, 488 n. 14, 489–90. Twice in its opinion, the Apprendi Court emphasized that Almendarez–Torres was different because “Almendarez–Torres had admitted the three earlier convictions for aggravated felonies[,]” and those prior convictions “had been entered pursuant to proceedings with substantial procedural safeguards of their own [.]” Id. at 488; see also id. (reiterating the same points). Thus, “no question concerning the right to a jury trial or the standard of proof that would apply to a contested issue of fact was before the Court.” Id.
Having explained the historical principles and the Almendarez–Torres exception, the Apprendi Court summarized the law as follows, in the now-famous Apprendi rule:
“[O]ur reexamination of our cases in this area, and of the history upon which they rely, confirms the opinion that we expressed in Jones. Other than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt. With that exception, we endorse the statement of the rule set forth in the concurring opinions in that case: ‘[I]t is unconstitutional for a legislature to remove from the jury the assessment of facts that increase the prescribed range of penalties to which a criminal defendant is exposed. It is equally clear that such facts must be established by proof beyond a reasonable doubt.’ “
Id. at 490 (quoting Jones, 526 U.S. at 252–53 (Stevens, J., concurring); citing id. at 253 (Scalia, J., concurring)). Applying this newly articulated rule, the Apprendi court ruled that it was unconstitutional for the sentencing judge to increase Apprendi's sentence beyond the statutory maximum based on a judicial finding that the crime to which Apprendi pled guilty constituted a hate crime. In doing so, it compared Almendarez–Torres, reiterating that: “there is a vast difference between accepting the validity of a prior judgment of conviction entered in a proceeding in which the defendant had the right to a jury trial and the right to require the prosecutor to prove guilt beyond a reasonable doubt, and allowing the judge to find the required fact under a lesser standard of proof.” Apprendi, 530 U.S. at 496.
Based on these constitutional principles, Garrus requests habeas relief under AEDPA. Specifically, he claims that the trial court violated his Sixth Amendment and Due Process rights by finding that his prior 1997 conviction constituted a “crime of violence” and argues that the state appellate court decisions affirming that finding were “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States” in Apprendi. See 28 U.S.C. § 2254(d)(1).
AEDPA “ ‘imposes a highly deferential standard for evaluating state-court rulings' and ‘demands that state-court decisions be given the benefit of the doubt.’ “ Felkner v. Jackson, ––– U.S. ––––, ––––, 131 S.Ct. 1305, 1307, 179 L.Ed.2d 374 (2011) (quoting Renico v. Lett, ––– U.S. ––––, ––––, 130 S.Ct. 1855, 1862, 176 L.Ed.2d 678 (2010)). We must use habeas corpus as a “ ‘guard against extreme malfunctions in the state criminal justice systems,’ not a substitute for ordinary error correction through appeal.” Harrington v. Richter, ––– U.S. ––––, ––––, 131 S.Ct. 770, 786, 178 L.Ed.2d 624 (2011). Thus, we may grant habeas relief under AEDPA only if the state court's adjudication of a claim “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States[,]” or “resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.”6 28 U.S.C. § 2254(d).
Because the state court identified and relied on the correct governing legal rule established by Apprendi, we must apply AEDPA's “unreasonable application” standard.7 See Williams v. Taylor, 529 U.S. 362, 407–08, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000). Under AEDPA's “unreasonable application” standard, we “may not issue the writ [if we merely] conclude [ ] ․ that the relevant state-court decision applied clearly established federal law erroneously or incorrectly.” Id. at 411. “Rather, that application must be ‘objectively unreasonable.’ “ Renico, 130 S.Ct. at 1862 (quoting Williams, 529 U.S. at 409). In other words, we must consider whether the state court's “determination can[ ] be reconciled with any reasonable application of the controlling standard[.]” Panetti v. Quarterman, 551 U.S. 930, 953, 127 S.Ct. 2842, 168 L.Ed.2d 662 (2007) (emphasis added). Garrus must show that “the state court's ruling ․ was so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.” Harrington, 131 S.Ct. at 786–87.
This case turns on the meaning of the “unreasonable application” standard. Garrus argues that the state court unreasonably applied Apprendi and its prior conviction exception by permitting the trial court to determine at sentencing that Garrus's prior 1997 conviction for burglarizing an unoccupied building (second-degree burglary) actually involved burglarizing an occupied building. In response, the Commonwealth argues that Garrus cannot meet his burden of proving that the state court's application of Apprendi was objectively unreasonable because neither Apprendi nor any case since has explicitly defined the scope of “the fact of a prior conviction” exception. However, our determination is not dependent on a fully delineated interpretation or application of “the fact of a prior conviction” exception, only a reasonable one. Panetti, 551 U.S. at 953 (“That the standard is stated in general terms does not mean the application was reasonable.”). The AEDPA standard is not so “myopic” nor “constrained” that it requires the full scope of all clearly established laws to be precisely defined. Jamison v. Klem, 544 F.3d 266, 273 (3d Cir.2008); see also id. (noting that if a Supreme Court decision must precisely resolve a given issue under AEDPA, only the “contrary to” standard would exist, and the “unreasonable application” standard would be rendered meaningless). Rather, AEDPA requires that if a clearly established rule or principle is stated, however general, state courts must adhere to a reasonable application of that rule.8 Panetti, 551 U.S. at 953.
Apprendi clearly requires that any fact “other than the fact of a prior conviction” must be submitted to a jury if it will increase the penalty beyond the statutory maximum. 530 U.S. at 490. Under AEDPA there is undoubtedly a broad spectrum of reasonable applications of this rule, but a fact that actually defies and contradicts a prior conviction falls squarely outside the spectrum of reasonable interpretations. Cf. Wilson v. Knowles, 638 F.3d 1213, 1215 (9th Cir.2011) (“Courts may reasonably disagree about some of the precise boundaries of the exception․ But that does not mean that Apprendi is amorphous․ The judge's fact-finding seven years after the 1993 conviction extended beyond any reasonable interpretation of the prior conviction exception.”). We can find no reasonable basis for applying Apprendi to allow judicial finding of a fact that is inconsistent with and contradicts the facts established through constitutional procedures. The very reason for the exception is to recognize facts that have already “been established through procedures satisfying the fair notice, reasonable doubt, and jury trial guarantees.” Jones, 526 U.S. at 249. The phrase “fact of a prior conviction” might be reasonably construed to allow judicial factfinding of a fact underlying a prior conviction or a fact consistent with a prior conviction. However, absent case law authorizing some type of paradoxical interpretation, “fact of a prior conviction” cannot reasonably be interpreted to allow judicial factfinding of a fact contradicting a prior conviction or a fact inconsistent with a prior conviction.9
No existing precedent so much as hints that a paradoxical interpretation of Apprendi might be reasonable.10 Thus, under Apprendi, it was objectively unreasonable for the state court to allow Garrus to be sentenced beyond the statutory maximum based on a mere allegation that actually defied and contradicted his prior conviction. The Commonwealth proffers several cases, which it argues could have supported the state court's decision to apply the prior conviction exception in Garrus's case, but its attempts fall flat.
Despite the Commonwealth's attempts to paint it as such, Almendarez–Torres is not a panacea allowing recidivism-related judicial factfinding, or supporting the state court's determination that Apprendi's prior conviction exception permitted the sentencing judge to find that Garrus had burglarized an occupied building despite his plea to the contrary. Apprendi cited Almendarez–Torres as the “limited” and “narrow” exception to the Apprendi rule, 530 U.S. at 488 n. 14, 489–90, and explained that the constitutionality of that exception rested on (1) “the certainty that procedural safeguards attached to any ‘fact’ of prior conviction,” and (2) “the reality that Almendarez–Torres did not challenge the accuracy of that ‘fact’ in his case [.]” Id. at 488. In fact, Apprendi emphasized these factors at least twice, and it is important to note that neither factor is present in this case. Regarding the first factor, the only “fact” judicially found by the sentencing court in Almendarez–Torres was the actual fact that the defendant was convicted, and that “fact” had previously been established through a conviction process applying procedural safeguards. In contrast, Garrus was only accused of—not convicted of—the violent crime of burglarizing an occupied building (which would have been first-degree burglary). More importantly, Garrus only pled guilty and was convicted of burglarizing an unoccupied building (second-degree burglary). Thus, unlike the Almendarez–Torres defendant, whose sentence was enhanced based on actual prior convictions, Garrus's sentence was enhanced beyond the statutory maximum without a basis in any jury determination or plea admission. Because the “fact” of which Garrus was accused was never found by a jury nor was it an element of the crime for which he pled guilty, no procedural safeguards attached to the finding of this “fact.”
As to the second factor, Garrus makes clear in his reply brief that, unlike Almendarez–Torres, he did not concede or admit the fact at issue. In fact, at the sentencing hearing, he questioned the judicial finding that his 1997 conviction was a “crime of violence.” (Sentencing Hr'g Tr. 28, May 8, 2001, see App. at 177a.) A defendant waives his Apprendi rights only if he “either stipulates to the relevant facts or consents to judicial factfinding.” Blakely v. Washington, 542 U.S. 296, 310, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004) (citing Apprendi, 530 U.S. at 488; Duncan v. Louisiana, 391 U.S. 145, 158, 88 S.Ct. 1444, 20 L.Ed.2d 491 (1968)). Thus, neither of the two factors that allowed Almendarez–Torres to stand as a narrow exception to Apprendi apply in this case. As Apprendi explained, “there is a vast difference between accepting the validity of a prior judgment of conviction entered in a proceeding in which the defendant had the right to a jury trial and the right to require the prosecutor to prove guilt beyond a reasonable doubt,” as occurred in Almendarez–Torres, and “allowing the judge to find the required fact under a lesser standard of proof[,]” as occurred in both Apprendi and here in Garrus's case. See 530 U.S. at 496. Therefore, it was objectively unreasonable for the state court to find that Almendarez–Torres authorized the application of Apprendi's prior conviction exception in this case.
Had Garrus been convicted of first-degree burglary in 1997 (based on a jury finding or plea establishing that he burglarized a building that was occupied and adapted for overnight accommodation), we do not doubt that under Almendarez–Torres, the court sentencing him for his later manslaughter conviction could have taken note of the earlier first-degree burglary conviction and held that it qualified as a crime of violence under Pennsylvania's “three strikes” law. However, Garrus was convicted of second-degree burglary.
Given the highly deferential AEDPA standard, it is even possible that if Pennsylvania merely had a generic burglary statute that did not differentiate between burglarizing an occupied or unoccupied building, a reasonable application of the “fact of a prior conviction” exception might have permitted the judge to look at facts underlying the prior burglary conviction to determine whether Garrus was convicted of burglarizing an occupied building.11 See, e.g., United States v. Santiago, 268 F.3d 151, 153, 157 (2d Cir.2001) (finding that Apprendi allows a sentencing judge to find “not only the mere fact of previous convictions but [certain] other related issues as well”);12 cf. Taylor v. United States, 495 U.S. 575, 602, 110 S.Ct. 2143, 109 L.Ed.2d 607 (1990) (holding that under a federal “three strikes” law, the sentencing judge may only find facts that were necessary to the prior conviction). However, the sentencing judge in this case looked at the previous conviction for burglarizing an unoccupied building, and nonetheless found facts supporting exactly the opposite proposition: that the building had been occupied. Because Garrus was convicted of burglarizing an unoccupied building, the finding that the building was occupied was simply not a “fact of a prior conviction.” The finding that the building was occupied could be viewed as a fact of a prior accusation of first-degree burglary, or as a fact contradicting a prior conviction for second-degree burglary. Either way, it cannot be reasonably interpreted as a “fact of a prior conviction.”
We likewise disagree with the Commonwealth's convoluted argument that Shepard somehow supports the state court's determination because it was decided on a statutory basis rather than a constitutional one. As a threshold matter, we note that Garrus contends that Shepard supports him rather than the Commonwealth. However, because Shepard's holding was limited to a federal statute that is inapplicable here and its constitutional discussion was mere dicta, Garrus cannot rely on it for relief. See Carey, 549 U.S. at 74 (“ ‘[C]learly established Federal law’ in § 2254(d)(1) ‘refers to the holdings, as opposed to the dicta, of th[e] [Supreme] Court's decisions as of the time of the relevant state-court decision.’ “ (quoting Williams, 529 U.S. at 412)).13 Nevertheless, we will address the Commonwealth's argument that Shepard somehow made the state court's application of Apprendi reasonable. Cf. Price v. Vincent, 538 U.S. 634, 643 and n. 2, 123 S.Ct. 1848, 155 L.Ed.2d 877 (2003) (even cases that are not binding might be relevant to the consideration of whether a state court decision is objectively unreasonable).
Shepard involved a sentencing enhancement pursuant to 18 U.S.C. § 924, which applies where a defendant has been previously convicted of violent felonies, including burglaries of buildings. 544 U.S. at 15–16. Shepard had previously been convicted of burglary four times, but he was convicted of burglarizing structures under Massachusetts law, and thus it was unclear from his convictions whether he had burglarized buildings or some other types of structures such as vehicles. Id. at 16–17; see also id. at 31 (O'Connor, J., dissenting). The Supreme Court held, based on principles of statutory interpretation, that the sentencing court was not permitted to look at police reports in order to determine whether the structures the defendant had burglarized were buildings. Id. at 16–22. It ruled that the sentencing court was “limited to the terms of the charging document, the terms of a plea agreement or [the terms of the] transcript of [the plea] colloquy ․ or to some comparable judicial record of this information.” Id. at 26.
The Commonwealth suggests that the state court could have gleaned support from Shepard by finding that because Shepard's holding was not constitutionally based, the use of police reports for judicial factfinding was constitutionally permissible. However such non sequitur reasoning is objectively unreasonable. The Shepard Court did nothing more than employ the classic doctrine of constitutional avoidance by deciding the case on a statutory basis. It is entirely unreasonable to read constitutional avoidance as indicative that a practice is constitutionally permissible. In fact, if anything can be read into the Shepard Court's discussion of the Constitution in dicta, it is that the majority of the Court expressed grave constitutional concerns: four justices cited concerns that looking at police reports underlying a prior conviction posed “serious risks of unconstitutionality” under Apprendi, and Justice Thomas would have gone further and found the entire “fact of a prior conviction” exception to be unconstitutional. See id. at 25 (plurality); id. at 26–28 (Thomas, J., concurring in part and concurring in the judgment).
Even assuming for purposes of this analysis that the Supreme Court expressed constitutional uncertainty about Apprendi's application in Shepard, it does not follow that Apprendi's application here is unclear. In fact, nothing in Shepard supports the Commonwealth's argument that “the fact of a prior conviction” might be reasonably construed to mean a fact contradicting a prior conviction because Shepard simply did not involve a fact contradicting a prior conviction. Crucially, Shepard had four times been convicted of burglarizing structures, and the police reports did not contradict this fact. Rather, the police reports supported the conviction for burglary of structures, and merely explained specifically that the types of structures involved were buildings. As the plurality noted, the factual allegations in the Shepard police reports might have been permitted by Apprendi as “a fact about a prior conviction[.]” Shepard, 544 U.S. at 25 (plurality) (emphasis added). The Shepard dissent's argument also highlights this distinction by stating that sentencing judges should be able to refer to “internally consistent parts of the record from the earlier conviction.” Id. at 31 (O'Connor, J., dissenting). In contrast, the police report alleging that Garrus burglarized an occupied building cannot reasonably be interpreted as a fact consistent with or a fact about Garrus's conviction for burglarizing an unoccupied building.
Rather than supporting the Commonwealth's argument, Shepard demonstrates that even judicial factfinding that is consistent with a prior conviction might be unconstitutional. By no means does it support the proposition that Apprendi's “fact of a prior conviction” exception might be reasonably interpreted to allow judicial factfinding that contradicts a prior conviction.
Finally, the Commonwealth strains to argue that some Circuit court cases show that the state court's application of Apprendi was reasonable. See Kessee v. Mendoza–Powers, 574 F.3d 675 (9th Cir.2009); United States v. Smith, 474 F.3d 888, 892 (6th Cir.2007), abrogation on other grounds recognized by United States v. Johnson, 640 F.3d 195, 205 (6th Cir.2011); Boyd v. Newland, 467 F.3d 1139 (9th Cir.2006); United States v. Hollingsworth, 414 F.3d 621 (6th Cir.2005); United States v. Williams, 410 F.3d 397 (7th Cir.2005); United States v. Camacho–Ibarquen, 410 F.3d 1307 (11th Cir.2005); United States v. Kempis–Bonola, 287 F.3d 699 (8th Cir.2002); Santiago, 268 F.3d at 151. We disagree; none of these cases demonstrates that it was reasonable for the state court to apply Apprendi to sentence a defendant beyond the statutory maximum based on a fact contradicting a prior conviction.14 In short, none of these cases involves a judicial finding of fact that contradicted the defendant's prior conviction.15 In fact, at least one of the cases did not even involve judicial factfinding that increased the penalty beyond the prescribed statutory maximum. See Smith, 474 F.3d at 892.
Upon a thorough and circumspect examination of the potential arguments or theories that “supported or ․ could have supported, the state court's decision[,]” we believe that no “fair-minded jurist could disagree that those arguments or theories are inconsistent with the holding in” Apprendi.16 Harrington, 131 S.Ct. at 786; see also Panetti, 551 U.S. at 953 (The “record ․ cannot, under any reasonable interpretation of the controlling legal standard, support [the state's] legal ruling.”). In fact, no fairminded jurist could disagree that the state court's paradoxical interpretation of the prior conviction exception renders an absurd result under Apprendi, allowing a judge that applies the “three strikes” statute to make the very factual finding that an earlier judge would have been prohibited from making. As we noted above, Apprendi pled guilty to unlawful possession of a weapon, a second-degree offense punishable by a maximum of ten years' imprisonment, but the judge increased the sentence beyond the statutory maximum based on the finding of an additional factor, effectively imposing punishment on the defendant for a first-degree offense. Apprendi, 530 U.S. at 491. Thus, under Apprendi, it is clear that a judge applying the initial sentence for second-degree burglary (under Pennsylvania law) cannot enhance the sentence beyond the statutory maximum based on the finding that the defendant burglarized an occupied building, effectively imposing punishment on the defendant for a first-degree offense. The Commonwealth today would have us believe that Apprendi may be reasonably applied to allow a later judge sentencing a defendant to look back at the defendant's initial conviction and make the factual finding that the initial sentencing judge was prohibited from making. Apprendi cannot be reasonably construed or applied to allow this absurd result.
Accordingly, we conclude that it was objectively unreasonable for the state court to apply Apprendi's prior conviction exception to allow judicial factfinding based on an allegation that actually defied and contradicted the elements of the prior conviction. The rule of Apprendi and the prior conviction exception simply cannot be reconciled with the state court decision. In sum, Garrus has shown that “the state court's ruling ․ was so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.” Harrington, 131 S.Ct. at 786–87.
We recognize the rarity of today's determination; seldom does a state court's decision fail to withstand AEDPA's highly deferential standard. However, our conclusion that the state court unreasonably applied clearly established law by relying on a narrow and inapplicable exception is not unprecedented. See Abdul–Kabir v. Quarterman, 550 U.S. 233, 258, 127 S.Ct. 1654, 167 L.Ed.2d 585 (2007) (finding that the state court unreasonably applied clearly established law by relying on an inapplicable “narrow holding” and “ignoring the fundamental principles established by [the Supreme Court's] most relevant precedents”); Williams, 529 U.S. at 397 (finding that the state court unreasonably applied clearly established law by relying on an “inapplicable” exception to the law clearly established by Strickland ). AEDPA severely constrained our ability to review state court decisions, but it did not render our review void or meaningless. See, e.g., Harrington, 131 S.Ct. at 786 (“ § 2254(d) stops short of imposing a complete bar on federal court relitigation of claims already rejected in state proceedings.”); Jamison, 544 F.3d at 273 (the “unreasonable application” standard is not meaningless). We must use our habeas authority only to “guard against extreme malfunctions in the state criminal justice systems,” Harrington, 131 S.Ct. at 786, and today, we exercise our authority to do so.
We conclude that the District Court erred in determining that § 2254(d)(1) prevented it from granting Garrus habeas relief. The state court unreasonably applied Apprendi by allowing Garrus to be sentenced beyond the statutory maximum based on a judicial finding that he burglarized an occupied building when he was actually convicted of burglarizing an unoccupied building.
Having determined that AEDPA does not bar federal habeas relief, we must also consider whether the state court determination had a “substantial and injurious effect” on Garrus's sentence. Fry v. Pliler, 551 U.S. 112, 116–20, 127 S.Ct. 2321, 168 L.Ed.2d 16 (2007); see also Horn, 536 U.S. at 272 (“none of our post-AEDPA cases have suggested that a writ of habeas corpus should automatically issue if a prisoner satisfies the AEDPA standard”). There is no doubt that the state court's determination relied on its application of Apprendi to uphold the trial court's decision to increase the length of Garrus's sentence, and that Garrus has therefore shown a “substantial and injurious effect.” We conclude that the District Court erred in failing to grant habeas relief under these circumstances. We will reverse the order of the District Court, and remand with directions that a writ of habeas corpus be conditionally granted, providing that petitioner be resentenced or released from custody within 120 days, unless, within that time, the Commonwealth of Pennsylvania determines that petitioner qualifies for the sentencing enhancement based on his two 1995 robbery convictions.
Judge Greenaway has succinctly explained how the Court has gone astray in this case, and I join his dissent in its entirety. I write separately because I think the reason why the majority has erred illuminates how it has done so.
Since its enactment in 1996, no law has so vexed the United States Courts of Appeals as the Antiterrorism and Effective Death Penalty Act (AEDPA), 28 U.S.C. § 2254. In twelve years, the Supreme Court has granted certiorari in ninety-four cases arising under AEDPA, forty-six of which involved questions of federal court deference to decisions of state courts.1 Thirty-four of those cases (approximately seventy-four percent) have been reversed because the court of appeals failed to afford sufficient deference to the state court.2 Remarkably, twenty-two of those cases—almost fifty percent—were reversed without dissent.3
The Supreme Court decisions catalogued in the margin echo a common mistake: the failure to adhere to AEDPA's extraordinarily deferential standard of review. Our Court makes that same mistake today.
The majority's fundamental error is manifest in the question we presented to the parties: “whether the district court erred in considering the defendant's 1997 conviction for second-degree burglary in determining whether the defendant should be sentenced under Pennsylvania's ‘three strikes' law, 42 Pa. Cons.Stat. § 9714.” This was the wrong question to ask in an AEDPA case that originated in state court. As the Commonwealth of Pennsylvania correctly argued, “[u]nder the federal habeas statute, ‘the only question that matters [is] whether [the] state court decision is contrary to, or involved an unreasonable application of, clearly established federal law.’ “ Appellee's Supp. Br. 1 (emphasis added) (quoting Lockyer v. Andrade, 538 U.S. 63, 71, 123 S.Ct. 1166, 155 L.Ed.2d 144 (2003)). Indeed, “a state prisoner must show that the state court's ruling on the claim being presented in federal court was so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.” Harrington v. Richter, ––– U.S. ––––, –––– – ––––, 131 S.Ct. 770, 786–87, 178 L.Ed.2d 624 (2011). Even petitioner Garrus acknowledged—albeit in somewhat apologetic fashion—that the question we presented invited argument under the wrong legal standard:
The question presented is somewhat ambiguous because the district court did not “consider” the 1997 burglary conviction for purposes of “determining whether the defendant should be sentenced under Pennsylvania's ‘three strikes' law.” Instead, applying 28 U.S.C. § 2254(d)(1), the district court analyzed whether the Pennsylvania Superior Court unreasonably applied clearly established U.S. Supreme Court precedent when it upheld the state sentencing court's decision to rely on police reports and witness statements to make findings of fact about the 1997 burglary conviction, only to rely on those findings to impose an enhanced sentence under the Three Strikes Law.
Appellant's Br. 1 n.1.
Instead of posing a question that circumvented AEDPA's deferential standard of review, we should have asked whether any Supreme Court decision directly prohibits a state court judge from looking to the facts underlying a defendant's prior conviction when considering a recidivism enhancement. The answer to that question is “no.”
Central to the state court's decision was the issue of who—judge or jury—should find facts regarding prior convictions when assessing a defendant's history of recidivism. Because recidivism “ ‘does not relate to the commission of the offense, but goes to the punishment only, [it] therefore ․ may be subsequently decided’ “ by the sentencing judge. Almendarez–Torres v. United States, 523 U.S. 224, 244, 118 S.Ct. 1219, 140 L.Ed.2d 350 (1998) (quoting Graham v. West Virginia, 224 U.S. 616, 629, 32 S.Ct. 583, 56 L.Ed. 917 (1912)). Garrus did not even attempt to explain why his case does not fall squarely within the recidivism exception announced in Almendarez–Torres. Instead, he relied entirely on Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000). But as then-Judge Sotomayor wrote for the Court of Appeals for the Second Circuit, in a case not subject to AEDPA's deferential standard of review, “recidivism has long been considered a distinct issue,” and “[n]othing in Apprendi itself—as it involved a hate crime statute enhancing sentences based on the motivation underlying the crime—calls that distinction into question.” United States v. Santiago, 268 F.3d 151, 156 (2d Cir.2001). Judge Sotomayor then clearly explained why Apprendi is of no help to habeas petitioners such as Garrus:
Almendarez–Torres explains why recidivism requires special treatment, and absent an explicit Supreme Court ruling to the contrary, we decline to institute a policy that runs counter to the principles set forth in that opinion. In short, we read Apprendi as leaving to the judge ․ the task of finding not only the mere fact of previous convictions but other related issues as well. Judges frequently must make factual determinations for sentencing, so it is hardly anomalous to require that they also determine the “who, what, when, and where” of a prior conviction.
If this were not a case governed by AEDPA and the question presented was whether what befell Garrus violated the rule of Apprendi, the majority's decision might well be correct. It is no secret that Almendarez–Torres is one of the most tenuous precedents of the Supreme Court; three of the five sitting justices who participated in Apprendi have openly criticized Almendarez–Torres.4 Nevertheless, I am convinced that even critics of Almendarez–Torres would be hard-pressed to hold that the state court in this case violated the rule of Apprendi when it found facts pursuant to a state recidivist statute.
The existence of a circuit split demonstrates that it is wrong to conclude that “ ‘fairminded jurists could [not] disagree’ on the correctness of the state court's decision” in this case. Harrington v. Richter, –––U.S. ––––, ––––, 131 S.Ct. 770, 786, 178 L.Ed.2d 624 (2011) (quoting Yarborough v. Alvarado, 541 U.S. 652, 664, 124 S.Ct. 2140, 158 L.Ed.2d 938 (2004)). The Courts of Appeals for the Second and Eighth Circuits have found no constitutional violation in circumstances similar to those presented in this appeal. See Portalatin v. Graham, 624 F.3d 69 (2d Cir.2010) (en banc); United States v. Davis, 260 F.3d 965 (8th Cir.2001).
In Portalatin, the Second Circuit consolidated the habeas petitions of three prisoners who had been sentenced under New York's persistent felony offender (PFO) statute. The petitioners argued that the statute was unconstitutional because it required a sentencing judge to assess the nature, rather than the mere fact of, predicate felony convictions before imposing a PFO sentence. They urged the court to construe the Almendarez–Torres exception to Apprendi narrowly so as to prevent a sentencing judge from considering the facts underlying a defendant's prior conviction when forming an opinion about his criminal history. Finding “no clear holding of the Supreme Court to command such a result,” the Second Circuit explained that, “ ‘[g]iven the lack of holdings from th[e] [Supreme Court]’ construing the recidivism exception as narrowly as petitioners urge, ‘it cannot be said that the state court unreasonably applied clearly established federal law.’ “ Id. at 92 (quoting Carey v. Musladin, 549 U.S. 70, 77, 127 S.Ct. 649, 166 L.Ed.2d 482 (2006)). Commenting on the “lack of guidance as to the precise scope of the recidivism exception,” the court concluded:
It might well be constitutionally significant whether a sentencing judge is required to find, for example, that a defendant's criminal history is “especially violent” before imposing a sentence, or whether, as in New York, a sentencing judge simply must find that the nature of his criminal history justifies “extended incarceration and life-time supervision.” ․ The Supreme Court may answer that question at some future time. But, if our Court cannot divine a clear answer from the Court's existing holdings, AEDPA prevents us from faulting a state court for selecting one reasonable conclusion over another.
Id. at 93; see also United States v. Snype, 441 F.3d 119, 148 (2d Cir.2006) (noting that while the continued viability of Almendarez–Torres has been questioned, the Supreme Court has not reversed that decision).
The Eighth Circuit reached a similar result in Davis. There, the appellant had been sentenced to a term of life imprisonment as a repeat offender following a federal conviction for attempted armed bank robbery, an offense that carried a statutory maximum of twenty-five years. He argued that the question of whether his past robbery convictions were qualifying felonies for purposes of the recidivist statute should have been decided by the jury, not by the sentencing judge. The court disagreed, finding that the “fact of prior conviction includes not only the fact that a prior conviction exists, but also a determination of whether a conviction is one of the enumerated types qualifying for the sentence enhancement.” Davis, 260 F.3d at 969. Conceding that Apprendi casts doubt on the viability of Almendarez–Torres, the Eighth Circuit nonetheless concluded that its role was “to apply Supreme Court precedent as it stands, and not as it may develop.” Id.
Adverse to the Second Circuit's decision in Portalatin and the Eighth Circuit's decision in Davis stands the opinion of the Ninth Circuit in Wilson v. Knowles, 638 F.3d 1213 (9th Cir.2011). After a 1993 car accident, Wilson pleaded no contest to gross vehicular manslaughter while driving under the influence of alcohol and to proximately causing bodily injury while driving under the influence of alcohol. When sentencing Wilson for a separate 2000 conviction for driving under the influence with a prior felony, a California judge relied on the information and preliminary hearing transcripts from Wilson's 1993 convictions to find numerous additional facts about the earlier offenses, including that Wilson personally inflicted bodily injury, that the bodily injury was great, and that his victim was not an accomplice. These factual findings led the judge to conclude that Wilson's 1993 convictions qualified as two predicate offenses for purposes of California's Three Strikes Law, and Wilson received an enhanced sentence of 25–years to life.
The Ninth Circuit held that the state court's factfinding was an unreasonable application of Apprendi. See id. at 1215 (“It would be unreasonable to read Apprendi as allowing a sentencing judge to find the kinds of disputed facts at issue here․ The judge's fact-finding seven years after the 1993 conviction extended beyond any reasonable interpretation of the prior conviction exception.”). Chief Judge Kozinski dissented, stating: “The Supreme Court hasn't straightened all this out․ AEDPA deference can be a bitter pill to swallow. In some habeas cases, we must reject what appear to be valid constitutional claims because petitioner's rights have not yet been clearly established by the Supreme Court.” Id. at 1217 (Kozinski, C.J., dissenting) (citations omitted).
Although it is possible that Wilson was correctly decided while Portalatin, Davis, and Chief Judge Kozinski's dissent in Wilson were all in error, it is immaterial for AEDPA purposes which line of analysis is correct. The mere fact of a difference of opinion among courts of appeals leads ineluctably to the conclusion that a state court cannot run afoul of AEDPA regardless of which of these two paths it chooses.
Stated differently, I am loath to conclude that the United States Court of Appeals judges who decided Portalatin and Davis are not “fairminded.” Yarborough, 541 U.S. at 664. Indeed, the intersection of criminal procedure and constitutional law often presents difficult questions upon which reasonable minds can differ.5 And while it is true that no Supreme Court decision explicitly authorized the state judge to delve into Garrus's prior conviction, the fact that no decision of the Court prohibited her from doing so is dispositive under AEDPA. Unless and until the Supreme Court overrules or modifies Almendarez–Torres, state courts cannot properly be held to have violated the Constitution when they find facts at sentencing to determine whether a recidivist statute applies. For that reason, I respectfully dissent.
Under the Anti–Terrorism and Effective Death Penalty Act of 1996 (“AEDPA”), Congress vested the Supreme Court—and only the Supreme Court—with the authority to determine clearly established law for purposes of analyzing an inmate's federal challenge to a state court judgment. Today, the majority feigns fidelity to this enduring tenet of AEDPA jurisprudence. In the process, the majority assumes the role of legislator, rewriting AEDPA to add this Court, along with the Supreme Court, as the two judicial bodies capable of delineating clearly established law. Armed with this newly created authority, the majority wades through murky Supreme Court precedent and emerges with what it purports to be an unassailable legal principle: that the prior-conviction exception enshrined in Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000), forbids a sentencing court from finding facts inconsistent with a prior conviction when applying a state recidivism statute.
Unlike the majority, I cannot discern any principle in Apprendi or other authoritative Supreme Court precedent that would render this interpretation of the prior-conviction exception objectively unreasonable, the deferential standard by which we must judge the state court's determination. Indeed, the majority identifies not one Supreme Court case expounding on the contours of the amorphous prior-conviction exception. This lack of clarity coupled with the inherent tension between Apprendi and Almendarez–Torres v. United States, 523 U.S. 224, 118 S.Ct. 1219, 140 L.Ed.2d 350 (1998)—a controlling case the majority brushes aside as having limited relevance—belies the majority's triumphant conclusion that the state court's determination was objectively unreasonable.
I do not disagree that the majority's interpretation of the prior-conviction exception is a reasonable and even favorable one, preventing an otherwise seemingly inequitable result. Of course, our task is not to simply choose the preferred reading of an ambiguous legal phrase. Absent guidance from the Supreme Court, we are left to determine whether the state court's interpretation was objectively unreasonable. Given that this area of law is, at best, in a state of flux, AEDPA “demands that [the] state-court decision[ ] be given the benefit of the doubt.” Renico v. Lett, ––– U.S. ––––, ––––, 130 S.Ct. 1855, 1862, 176 L.Ed.2d 678 (2010) (internal quotation marks and citation omitted). For this reason, I am compelled to respectfully dissent.
I. AEDPA Imposes a High Threshold
While the majority outlines the basic AEDPA framework governing our inquiry, AEDPA's prominent role in this case requires further elaboration.
As amended by AEDPA, 28 U.S.C. § 2254 provides, in pertinent part: “[A] circuit judge ․ shall entertain an application for a writ of habeas corpus in behalf of a person in custody pursuant to the judgment of a State court only on the ground that he is in custody in violation of the Constitution or laws or treaties of the United States.” 28 U.S.C. § 2254(a). “[W]ith respect to any claim that was adjudicated on the merits in State court proceedings,” the writ shall not issue unless the adjudication of the claim “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States” or was “based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” Id. § 2254(d)(1)-(2).
A state court decision is “contrary to” Supreme Court precedent “if the state court applies a rule that contradicts the governing law set forth in [Supreme Court] cases,” or “if the state court confronts a set of facts that are materially indistinguishable from a decision of th[e] Court and nevertheless arrives at a result different from [the Court's] precedent.” Williams v. Taylor, 529 U.S. 362, 405–06, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000). A state court decision involves an unreasonable application of federal law “if the state court identifies the correct governing legal rule from [the Supreme] Court's cases but unreasonably applies it to the facts of the particular state prisoner's case.” Id. at 407. To qualify as an “unreasonable application” of clearly established federal law, the state court's application must be objectively unreasonable. Id. at 409. That the state court decision was incorrect or erroneous is insufficient if the error was nonetheless objectively reasonable. Id. at 410; see also Renico, 130 S.Ct. at 1862 (“[A]n unreasonable application of federal law is different from an incorrect application of federal law.” (quoting Williams, 529 U.S. at 410)). “Indeed, ‘a federal habeas court may not issue the writ simply because that court concludes in its independent judgment that the relevant state-court decision applied clearly established federal law erroneously or incorrectly.’ “ Renico, 130 S.Ct. at 1862 (quoting Williams, 529 U.S. at 411).
As a result, the “contrary” and “unreasonable application” thresholds are “highly deferential ․ for evaluating state-court rulings, which demand [ ] that state-court decisions be given the benefit of the doubt.” Cullen v. Pinholster, ––– U.S. ––––, ––––, 131 S.Ct. 1388, 1398, 179 L.Ed.2d 557 (2011) (quoting Woodford v. Visciotti, 537 U.S. 19, 24, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002) (per curiam)). These standards are “ ‘difficult to meet,’ because the purpose of AEDPA is to ensure that federal habeas relief functions as a ‘guard against extreme malfunctions in the state criminal justice systems,’ and not as a means of error correction.” Greene v. Fisher, –––U.S. ––––, ––––, 132 S.Ct. 38, 43, 181 L.Ed.2d 336 (2011) (quoting Harrington v. Richter, ––– U.S. ––––, ––––, 131 S.Ct. 770, 786, 178 L.Ed.2d 624 (2011)).
I agree with the majority that the focus of our inquiry is on the “unreasonable” prong because the state court identified controlling Supreme Court precedent. See Williams, 529 U.S. at 406 (“[A] run-of-the-mill state-court decision applying the correct legal rule from our cases to the facts of a prisoner's case would not fit comfortably within § 2254(d)(1)'s ‘contrary to’ clause.”). As such, we can grant Garrus's petition only if the sentencing court's application of Pennsylvania's recidivism statute transcended the bounds of this undeniably deferential framework.
II. Majority's Discussion of Supreme Court Precedent Falls Short
It is undeniable that the majority's interpretation of Apprendi's prior-conviction exception cannot stand in the absence of clearly established law. See Howes v. Fields, ––– U.S. ––––, ––––, 132 S.Ct. 1181, 1187, 182 L.Ed.2d 17 (2012) (rejecting a circuit court's determination that Supreme Court precedent clearly established a categorical rule regarding custodial interrogations of inmates). Recognizing its obligation, the majority spends pages examining Supreme Court precedent. Yet nowhere in that examination does the majority grapple with precedent that undermines its position. This principled avoidance is the only means by which the majority can reach the conclusion that Supreme Court precedent clearly establishes that it would be objectively unreasonable to interpret the phrase “the fact of a prior conviction” as the state court did in this case.
I do not disagree with the majority that Apprendi is at the forefront of our inquiry. I do disagree, however, with the majority's perception that Apprendi is the be-all and end-all of our inquiry. The majority's singular focus is apparent at the beginning of its opinion, stating that the “[k]ey” to resolving this case involves answering “whether, pursuant to AEDPA, the state court unreasonably applied Apprendi.' “ (Majority Op. at 3.) However, as the majority acknowledges, Garrus was sentenced under 42 Pa. Cons.Stat. § 9714, Pennsylvania's recidivism statute. Despite this admission, Almendarez–Torres—a critical Supreme Court case examining a sentencing court's power to engage in factfinding when applying a recidivism statute—plays second fiddle to Apprendi in the majority's opinion. And when the majority does address Almendarez–Torres, it sidesteps the very logic in Almendarez–Torres that casts even the slightest doubt on the notion that the majority's interpretation of the prior-conviction exception is clearly established.
Almendarez–Torres involved a federal statute that permitted an enhanced penalty for any person unlawfully in the United States who had previously been deported after being convicted of an aggravated felony. 523 U.S. at 226. The defendant pled guilty to being in the United States, without permission, after being deported. Id. At sentencing, the defendant argued that the enhancement should not apply because his three previous convictions for aggravated felonies, which the defendant did not contest, were not charged in the current indictment. Id. at 227. The district court rejected this argument, found the existence of the prior convictions, and applied the sentence enhancement. Id. The Fifth Circuit affirmed. Id. at 227–28.
The Supreme Court held that the district court properly treated the defendant's prior convictions as a sentencing factor and not a separate offense. Id. at 235. The Court repeatedly stressed that its analysis was predicated on the unique role recidivism plays at sentencing. See id. (“At the outset, we note that the relevant statutory subject matter is recidivism. That subject matter—prior commission of a serious crime—is as typical a sentencing factor as one might imagine.”); id. (“With recidivism as the subject matter in mind ․”); id. at 241 (“[The statute at issue] involves one of the most frequently found factors that affects sentencing—recidivism.”); id. at 243 (“[T]he sentencing factor at issue here—recidivism—is a traditional, if not the most traditional, basis for a sentencing court's increasing an offender's sentence.”). The Court noted that, historically, recidivism operated outside of the general framework requiring elements of an offense be submitted to a jury because “recidivism ‘does not relate to the commission of the offense, but goes to the punishment only, and therefore ․ may be subsequently decided.’ “ Id. (quoting Graham v. West Virginia, 224 U.S. 616, 624, 32 S.Ct. 583, 56 L.Ed. 917 (1912)).
Two years after Almendarez–Torres, the Supreme Court rendered a decision in what we have described as its “seminal case” on sentencing—Apprendi. Reinhold v. Rozum, 604 F.3d 149, 152 (3d Cir.2010). In Apprendi, the defendant received an enhanced sentence of twelve years for weapon offenses punishable by a statutory maximum of ten years, based on a determination by the sentencing court, not the jury, that the crime was racially motivated. 530 U.S. at 470–71. Rejecting this approach, the Court pronounced that the Sixth Amendment required that “[o]ther than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt.” Id. at 490. Central to the Court's determination was the fact that the hate crime enhancement required a finding that the defendant's purpose was to intimidate on the basis of race—an inquiry into the defendant's mens rea, an element of the offense. Id. at 492–93.
The Court attempted to dispel any apparent tension between the rule announced in Apprendi and the Court's decision in Almendarez–Torres. “Whereas recidivism ‘does not relate to the commission of the offense’ itself, New Jersey's biased purpose inquiry goes precisely to what happened in the ‘commission of the offense.’ “ Id. at 496 (quoting Almendarez–Torres, 523 U.S. at 230). But the Court was cognizant that “it is arguable that Almendarez–Torres was incorrectly decided” based on a “logical application” of the Court's analysis in Apprendi. Id. at 489. In the end, the Court reasoned that Almendarez–Torres “represents at best an exceptional departure” from Apprendi. Id . at 487. While the Apprendi Court provided less than unbridled support for Almendarez–Torres, it is clear that Almendarez–Torres had continuing vitality after Apprendi.
The majority devotes no attention to the integral role recidivism played in Almendarez–Torres, instead trying to minimize Almendarez–Torres's, significance entirely. Quoting Apprendi, the majority refers to Almendarez–Torres as having “limited” and “narrow” application. (Majority Op. at 15 (quoting Apprendi, 530 U.S. at 488 n. 14, 489).) Put in its proper light, however, this characterization in no way subverts the importance of Almendarez–Torres to Garrus's case.
Apprendi did not involve the application of a recidivism statute. Instead, Apprendi addressed a hate crime enhancement that inquired as to the defendant's mens rea, which the Court iterated was “as close as one might hope to come to a core criminal offense ‘element.’ “ 530 U.S. at 493. By contrast, Almendarez–Torres expressly excluded recidivism as an exception to the nascent legal principle enshrined three years later in Apprendi. Almendarez–Torres, 523 U.S. at 244 (“[T]o hold that the Constitution requires that recidivism be deemed an ‘element’ of petitioner's offense would mark an abrupt departure from a longstanding tradition of treating recidivism as going to the punishment only.” (internal quotation marks and citation omitted)). As a result, it is unsurprising that Apprendi ensured that Almendarez–Torres did not apply outside the recidivism context. However, recidivism is squarely at issue in this case, belying the majority's attempt to minimize Almendarez–Torres's, relevance.
As much as the majority would like to forget Almendarez–Torres, there is little debate that Almendarez–Torres remains binding precedent. After Apprendi cast doubt on the continuing vitality of Almendarez–Torres, we addressed whether Almendarez–Torres's viability had been short-lived. Because the Apprendi Court expressly declined to overrule Almendarez–Torres, we unsurprisingly applied Almendarez–Torres and held that predicate offenses under a recidivism statute need not be charged in an indictment and submitted to the jury. See United States v. Weaver, 267 F.3d 231, 250 (3d Cir.2001) (“Recently, in Apprendi [ ], the Court upheld the validity of Almendarez–Torres ․”). We have since reaffirmed the viability of Almendarez–Torres and have not retreated from this determination. United States v. Arrelucea–Zamudio, 581 F.3d 142, 157 n. 15 (3d Cir.2009) ( “Although several Supreme Court decisions have cast doubt on the statute's continuing constitutional viability post-Apprendi, ․ we are bound by Almendarez–Torres ․ .”); United States v. Ordaz, 398 F.3d 236, 241 (3d Cir.2005) (“The holding in Almendarez–Torres remains binding law․”).1 The Supreme Court continues to recognize the separate treatment afforded prior convictions in Almendarez–Torres. See United States v. O'Brien, ––– U.S. ––––, ––––, 130 S.Ct. 2169, 2174, 176 L.Ed.2d 979 (2010).
III. State Court's Determination Was Objectively Reasonable
Against this proper legal backdrop, there can be no doubt that, for purposes of sentencing under Pennsylvania's recidivism statute, the state court's determination that Garrus's second-degree burglary conviction could be treated as a first-degree offense was not objectively unreasonable. Almendarez–Torres established that a sentencing court applying a recidivism statute that carries with it a sentence above the statutory maximum is permitted to engage in judicial factfinding regarding the defendant's prior convictions. In Apprendi, the Court clarified that such judicial factfinding was impermissible, except that relating to “the fact of a prior conviction.” 530 U.S. at 490. In other words, the exact undefined judicial factfinding permitted by Almendarez–Torres.
Apprendi provides no guidance regarding the scope or meaning of the prior-conviction exception. For good reason, Apprendi's true import is not the establishment of the exception to the rule but the rule itself—that a judge cannot find facts regarding an element of the offense where the defendant faces a sentence above the statutory maximum. Apprendi did not need to expound on the scope of the prior-conviction exception because Apprendi was not, contrary to Almendarez–Torres, a recidivism case. The prior-conviction exception remains as amorphous and undefined as the day Apprendi was decided. See Kessee v. Mendoza–Powers, 574 F.3d 675, 676 (9th Cir.2009) ( “What is the scope of the ‘prior conviction’ exception to the general rule that a sentencing judge may not make factual findings that increase the statutory maximum criminal penalty? The Supreme Court has not yet answered that question.”).
I do not deny that the majority's reading of the prior-conviction exception—prohibiting a sentencing court from applying a recidivism statute and finding a fact inconsistent with a prior conviction—is a reasonable interpretation.2 Under AEDPA, that the majority's reading was reasonable does not ipso facto render all other readings patently unreasonable. See id. (noting that, under AEDPA, just because an appellate court's own interpretation of the prior-conviction exception is reasonable does not mean that the state court's reading was unreasonable). “Because AEDPA authorizes federal courts to grant relief only when state courts act unreasonably, it follows that ‘the more general the rule’ at issue—and thus the greater the potential for reasoned disagreement among fair-minded judges—‘the more leeway state courts have in reaching outcomes in case-by-case determinations.” Renico, 130 S.Ct. at 1864 (quoting Yarborough v. Alvarado, 541 U.S. 652, 664, 124 S.Ct. 2140, 158 L.Ed.2d 938 (2004)). In my view, the prior-conviction exception is sufficiently general and undefined, such that we must, under AEDPA, defer to the state court. Given the logic of Almendarez–Torres and the fact that the Supreme Court has yet to clarify the scope of the prior-conviction exception, I simply cannot subscribe to the majority's finding that the state court's decision here was objectively unreasonable.
A few additional points bear mentioning. First, the majority opinion suffers from an internal inconsistency. Relying as it does on Apprendi, a case founded on the Sixth Amendment right to a jury trial, one would expect the majority to require that a jury determine whether or not Garrus's prior conviction for burglary qualifies as a crime of violence. Yet the majority does the opposite, requiring that the judge determine that Garrus's prior conviction was necessarily not a crime of violence.
Second, the majority notes that “[n]o existing precedent so much as hints that [the state court's] paradoxical interpretation of Apprendi might be reasonable.” (Majority Op. at 20.) This point suffers from two flaws. One could easily reverse the logic of this statement and assert that no Supreme Court precedent suggests that the majority's interpretation is the only reasonable interpretation. Under AEDPA, that is sufficient to uphold the state court's determination.
More importantly, the proposition is simply inaccurate. The majority makes clear that it is troubled by the fact that Garrus pled guilty to second degree burglary only to have a judge find facts years later inconsistent with that prior conviction and without affording Garrus procedural safeguards. In United States v. Santiago, 268 F.3d 151 (2d Cir.2001), then-Judge Sotomayor, writing for the court, addressed whether the requirement under the federal recidivism statute, the Armed Career Criminal Act, that the defendant have three predicate convictions “committed on occasions different from one another,” is a factual issue that must be submitted to a jury under Apprendi. Id. at 152. In resolving the issue in the negative, Judge Sotomayor stressed that courts make determinations all the time in the context of recidivism for which the Apprendi procedural safeguards do not attach. See id. at 156 (“The determination of ‘the fact of a prior conviction’ implicitly entails many subsidiary findings, not the least of which is that the defendant being sentenced is the same defendant who previously was convicted of those prior offenses, a fact that could be quite controversial indeed. Determination of this question would not necessarily come with the ‘procedural safeguards' noted in Apprendi.”). Judge Sotomayor went on to conclude:
In short, we read Apprendi as leaving to the judge, consistent with due process, the task of finding not only the mere fact of previous convictions but other related issues as well. Judges frequently must make factual determinations for sentencing, so it is hardly anomalous to require that they also determine the ‘who, what, when, and where’ of a prior conviction.
Id. Santiago refutes the majority's unyielding interpretation of the phrase “the fact of a prior conviction,” which is even more meaningful considering that Santiago was not a case decided under the highly deferential AEDPA framework.
Finally, there is little doubt that Almendarez–Torres's continuing vitality is in jeopardy. See Shepard v. United States, 544 U.S. 13, 27–28, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005) (Thomas, J., concurring) (“Almendarez–Torres ․ has been eroded by this Court's subsequent Sixth Amendment jurisprudence, and a majority of the Court now recognizes that Almendarez–Torres was wrongly decided․ [I]n an appropriate case, this Court should consider Almendarez–Torres ' continuing viability.”). However, we must apply Almendarez–Torres until the Supreme Court holds otherwise. See Hohn v. United States, 524 U.S. 236, 252–53, 118 S.Ct. 1969, 141 L.Ed.2d 242 (1998) (“Our decisions remain binding precedent until we see fit to reconsider them, regardless of whether subsequent cases have raised doubts about their continuing vitality.”); Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 484, 109 S.Ct. 1917, 104 L.Ed.2d 526 (1989) (“If a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to this Court the prerogative of overruling its own decisions.”).
For these reasons, I respectfully dissent.
FISHER, Circuit Judge, joined by McKEE, Chief Judge, and SLOVITER, SCIRICA, RENDELL, AMBRO, FUENTES, SMITH, and JORDAN, Circuit Judges.