OPINION

ROGERS, Circuit Judge.

These habeas petitions involve the far-reaching effects of an unintentional computer glitch that caused the systematic underrepresentation of African-Americans in the jury pools of Kent County, Michigan. The three petitioners received separate jury trials in Kent County in either 2001 or 2002. At jury selection, the petitioners did not object to the racial composition of their respective jury venires, and were subsequently convicted. A few months later, the Grand Rapids Press published a story detailing how Kent County's jury selection software had a computer glitch that had systematically excluded African-Americans from the jury pool. In light of these revelations, the petitioners each filed motions for post-conviction relief in state court. The state court found that the petitioners had waived these claims by failing to object to the racial composition of the jury venire during voir dire.

The petitioners each filed a § 2254 habeas petition. The Western District of Michigan denied two of the petitions — Carter v. Lafler, No. 10-1247, and Wellborn v. Berghuis, No. 09-1539 — holding that the petitioners had not shown cause to excuse their procedural default. In Ambrose v. Booker, No. 11-1430, the Eastern District of Michigan granted habeas relief. The Ambrose court held that the petitioner could not have known of the computer glitch, presumed prejudice because the glitch caused a structural error, and then granted the petition on the merits. Remand in all three cases is appropriate because the petitioners have shown cause to excuse their defaults, but federal-state comity requires that the district court find actual prejudice before granting relief.

I.

1. Ambrose

On April 19, 2001, a Kent County jury convicted Joseph Ambrose of two counts of armed robbery, one count of carjacking, and one count of felony-firearm possession. Ambrose v. Booker, 781 F.Supp.2d 532, 535 (E.D.Mich.2011). Ambrose sought leave to appeal; however, his appointed counsel withdrew and the Michigan Court of Appeals determined that Ambrose's appeal was frivolous. Id. Ambrose did not appeal to the Michigan Supreme Court.

On July 30, 2002, the Grand Rapids Press reported that a computer glitch had an impact on Kent County's system for selecting jury venires. The glitch was introduced accidentally by the county when it assumed control of the jury selection computer program from a private vendor in April 2001. The problem came to light in 2002, when a local high school teacher, Wayne Bentley, completed a study of minority representation on Kent County juries. Bentley found that the underrepresentation of minorities was statistically significant, and shared his findings with county officials. The county subsequently conducted an internal study that revealed that "nearly 75 percent of the county's 454,000 eligible residents were excluded from potential jury pools since spring 2001" and that "[m]any blacks were excluded from ... jury pools due to a computer glitch that selected a majority of potential candidates from the suburbs." The chief judge of the Kent County Circuit Court, George Buth, stated, "There has been a mistake — a big mistake."

In light of these revelations, Ambrose initiated state post-conviction proceedings. Ambrose requested relief from judgment, claiming he was denied his right to be tried by a jury drawn from a fair cross-section of the community. The trial court denied Ambrose's motion because, among other things, Ambrose failed to object to the venire panel before the jury was empaneled. The Michigan Court of Appeals denied leave to appeal, as did the Michigan Supreme Court. See People v. Ambrose, 474 Mich. 931, 706 N.W.2d 16 (2005) (unpublished table decision).

Ambrose then filed this habeas petition in the United States District Court for the Eastern District of Michigan. The district court referred the matter to a magistrate judge, who held an evidentiary hearing at which four items of evidence were introduced. First, the magistrate judge heard the testimony of Wayne Bentley, the teacher that uncovered the disparate representation. Second, the magistrate judge heard testimony from Terry Holtrop, the case manager for the Kent County Circuit Court. Ambrose, 781 F.Supp.2d at 537-38. Holtrop explained how Bentley's evidence spurred an internal investigation by the County, culminating in the Kent County Jury Management System Report (hereafter "the Report") dated August 1, 2002. The Report described how a transfer of control over a database — from a private vendor to the County in an effort to cut costs — caused the error:

Third, the magistrate judge considered statistical analysis submitted by Dr. Paul Stephenson, a statistician who used different methodologies to evaluate the impact of the glitch. Dr. Stephenson first compared the percentage of eligible African-Americans in Kent County and the actual percentage in the venire. He found an absolute disparity of 6.03% and a comparative disparity of 73.1% fewer African-American members than would be expected.1 While Dr. Stephenson believed that these disparities were useful "descriptive statistics," he did not believe they were "viable for inferential purposes." Stephenson next used both a standard deviation and binomial test, and found that although the tests provided "insufficient evidence to demonstrate that the representation of Blacks or African-Americans in [a specific] venire is biased, this is in part due to the size of the venire." Stephenson also noted "while it would be less likely, one also could expect approximately 2% of all venires to contain no ... African-American members." Stephenson then employed a "Chi-square Goodness-of-fit test." Using this test, Dr. Stephenson found that "there is essentially no chance of acquiring the results we obtained if the selection process for potential jurors is unbiased. As a result, there is overwhelming evidence to conclude that the selection process for terms during the first months of 2002 was biased." Dr. Stephenson concluded that "the analysis presented in this report demonstrates that a systematic bias did exist in the selection of individuals summoned for jury duty during the first three months of 2002. This bias would have inevitably led to under representation of ... African-Americans in the terms during this period of time."

Fourth, the magistrate judge considered a report by Dr. Edward Rothman, who analyzed the composition of the Kent County Jury Pool between January 1998 and December 2002. Dr. Rothman used the census figures from the 2000 census, at which time African-Americans comprised 8.24% of the population of Kent County. Dr. Rothman applied these figures to the period April 2001 to August 2002, and concluded that there was a 3.45% absolute disparity between jury-eligible African-Americans and those who appeared on jury venires. Rothman found a comparative disparity of 42%. Based on these four pieces of evidence, the magistrate judge concluded that Petitioner had proven a prima facie case of a fair cross-section claim.

The district court adopted the magistrate judge's report and recommendation. Ambrose, 781 F.Supp.2d at 545-46. First, the district court held that Ambrose had not procedurally defaulted his claim. The court acknowledged that district courts had split on the issue of procedural default with regard to the Kent County computer glitch, but the court concluded that Ambrose had shown good cause to excuse his default because he could not have known of the glitch. Id. at 542-43. Second, the district court found that the number of African-Americans in the jury pool was not "fair and reasonable in relation to the number of such persons in the community." Id. at 543-44. The district court recognized that the absolute disparity of 3.45% fell below the 10% mark established by Swain v. Alabama, 380 U.S. 202, 208-09, 85 S.Ct. 824, 13 L.Ed.2d 759 (1965). However, this 10% mark was not a bright-line rule; to hold otherwise would prevent fair cross-section claims in areas where the minority population was less than 10%. Ambrose, 781 F.Supp.2d at 543-44. Finally, the district court held that the systematic exclusion did not need to be intentional. Id. at 545. The district court granted a conditional writ of habeas corpus and respondents timely appealed.

2. Carter

Gregory Carter's habeas petition was reviewed by both the Eastern and Western Districts of Michigan, which came to opposite conclusions. A jury convicted Carter of armed robbery of a convenience store in Kent County, Michigan. At trial, the prosecution presented "overwhelming" evidence of the robbery, including a videotape which showed Carter robbing the store. Carter v. Lafler, No. 1:09-cv-215, 2010 WL 160814, at *1 (W.D.Mich. Jan. 8, 2010). Petitioner was sentenced to two years' incarceration for a felony firearm charge to be served consecutively with a term of 15 to 40 years' imprisonment for the armed robbery. Id. Carter did not object to the venire panel at trial, and did not challenge the panel's composition at any stage of his direct appeal.

In light of the revelations about the problems with Kent County's jury selection system, Carter initiated state post-conviction proceedings under M.C.R. 6.501. In his motion, Carter challenged the composition of the jury based on flaws in the Kent County jury selection system, and raised an ineffective assistance of counsel claim. The trial court denied the motion, reasoning that Carter had defaulted by failing to object to the venire's composition before the jury was empaneled. Further, the trial court found that Carter's ineffective assistance of counsel claim failed under Michigan's version of Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). Carter applied for leave to appeal, but both the Michigan Court of Appeals and Michigan Supreme Court denied the application "for failure to meet the burden of establishing entitlement to relief under MCR 6.508(D)." People v. Carter, No. 254482 (Mich.App. Aug. 20, 2004); People v. Carter, 472 Mich. 912, 696 N.W.2d 714 (2005) (table opinion).

Carter filed a habeas petition in the United States District Court for the Eastern District of Michigan. The district court held that there was cause and prejudice to excuse Carter's procedural default. Carter v. Lafler, No. 06-cv-10552, 2009 WL 649889, at *3 (E.D.Mich. Mar. 10, 2009). The district court recognized that Carter had procedurally defaulted by failing to object at trial to the composition of the venire panel. The default was excused, however, because the district court found that Carter could not have known about the systematic exclusion caused by the computer glitch at the time the jury was sworn. Id. Having excused the default, the court determined that an evidentiary hearing was necessary to determine whether Carter's jury was empaneled during the period impacted by the computer glitch. Id. at *4-5. Because Carter was imprisoned in the Western District of Michigan, and because the trial had occurred in the Western District, the court transferred the action "for an evidentiary hearing and determination on the Sixth Amendment claim." Id. at *5.

Upon receiving the case, the Western District of Michigan assigned the case to a magistrate judge. The magistrate judge recommended that the district court deny the habeas petition as procedurally defaulted without an evidentiary hearing. Carter v. Lafler, No. 1:09-cv-215, 2010 WL 160814, at *4-13 (W.D.Mich. Jan. 8, 2010). The district court adopted this recommendation because the court determined that Carter had not shown good cause for his procedural default. Id. at *3. Although Carter was unaware of the computer glitch at the time of his trial, the district court held that the racial composition of the jury venire gave him notice of his claim. Id. Carter timely appeals.

3. Wellborn

Carl Wellborn was convicted of sexually abusing his granddaughters by a Kent County Jury in March 2001. Wellborn v. Berghuis, No. 1:05-cv-346, 2009 WL 891708, at *8-14 (W.D.Mich. Mar. 31, 2009) (report and recommendation). On direct appeal, Wellborn claimed that he was denied a jury drawn from a fair cross-section of the community based on the Kent County computer glitch. The Michigan Court of Appeals found that Wellborn had defaulted his fair cross-section claim because he failed to raise it before the jury was sworn. Id. at *14. The Michigan Court of Appeals affirmed Wellborn's convictions, and the Michigan Supreme Court denied leave to appeal. Id.

Wellborn raised a variety of claims in his habeas petition, but only appeals his fair cross-section claim. The district court found that this claim was procedurally defaulted, and held that there was no cause to excuse the default. Id. at *3. Because the petitioner observed the juror array, the court reasoned, he was on notice of a potential underrepresentation. Id. The court found that the petitioner must show evidence that the underrepresentation was intentionally concealed, a showing Wellborn did not make. Id. at *3-4. The court acknowledged that the Eastern District had addressed the computer glitch and excused procedural default, but found Wellborn's case distinguishable on two grounds: Wellborn was white and there was overwhelming evidence of Wellborn's guilt. Id. at *3-4. Wellborn filed this timely appeal.

II.

As an initial matter, the parties disagree on the standard of review. The Antiterrorism and Effective Death Penalty Act ("AEDPA") requires federal courts to grant considerable deference to state courts and presume the correctness of state court fact finding. 28 U.S.C. § 2254(d), (e)(1). However, this deferential standard applies only to claims that state courts have adjudicated on the merits. Respondents admit that "the State court ... refus[ed] to examine the merits of Petitioner's claim." Even without this concession, it is evident that the state courts rejected petitioners' fair cross-section claims on procedural grounds, based on the failure to object to the jury panel at trial. For this reason, AEDPA deference does not apply and the court reviews legal conclusions de novo and findings of fact for clear error. See, e.g., Dyer v. Bowlen, 465 F.3d 280, 283-84 (6th Cir.2006).

1. Cause for Procedural Default

There is no dispute that petitioners defaulted their claims by failing to object to the jury venire at trial, in violation of Michigan's contemporaneous objection rule. See, e.g., People v. Dixon, 217 Mich.App. 400, 552 N.W.2d 663, 667 (1996). However, petitioners have shown "cause and prejudice" sufficient to excuse the default under Coleman v. Thompson, 501 U.S. 722, 750, 111 S.Ct. 2546, 115 L.Ed.2d 640 (1991). The cause inquiry "ordinarily turn[s] on whether ... some objective factor external to the defense impeded counsel's efforts to comply with the State's procedural rule," and is satisfied by "a showing that the factual or legal basis for a claim was not reasonably available to counsel." Murray v. Carrier, 477 U.S. 478, 488, 106 S.Ct. 2639, 91 L.Ed.2d 397 (1986). As discussed below, petitioners have shown cause because the factual basis for the claim — the computer glitch — was not reasonably available to counsel, and petitioners could not have known that minorities were underrepresented in the jury pool by looking at the venire panel. Moreover, the Supreme Court has found cause to excuse a procedural default in a factually similar case. See Amadeo v. Zant, 486 U.S. 214, 108 S.Ct. 1771, 100 L.Ed.2d 249 (1988).

It is undisputed that petitioners lacked actual knowledge of the glitch, and the record demonstrates that they could not have reasonably known of the error at the time the jury was sworn. The glitch was a mistyped parameter in the software, buried in a mountain of computer code, that was only discovered after a broad statistical analysis led to an extensive internal investigation. This is not the type of error that would be reasonably discoverable by a defendant at jury selection. The difficulty of detecting this problem is underscored by the fact that the glitch persisted for nearly two years without detection by defendants, judges, or Kent County officials.

Further, looking at the venire panel did not provide petitioners with notice of their claims. Petitioners' claims arise from the right "to be tried by an impartial jury drawn from sources reflecting a cross section of the community." Berghuis v. Smith, ___ U.S. ___, 130 S.Ct. 1382, 1387, 176 L.Ed.2d 249 (2010) (emphasis added). A petitioner raising this claim is challenging the pool from which the jury is drawn, and not necessarily the venire panel directly before him. Accordingly, the composition of one panel does not indicate whether a fair cross-section claim exists.

The irrelevance of the composition of a single venire panel is underscored by the fact that a petitioner may bring a claim even if minorities are included in his panel. The Sixth Amendment guarantees only the opportunity for a representative jury, not a representative jury itself. United States v. Biaggi, 909 F.2d 662, 678 (2d Cir.1990). The focus, therefore, is on the procedure for selecting juries, and not the outcome of that process. As the First Circuit eloquently put it:

Barber v. Ponte, 772 F.2d 982, 993 (1st Cir.1985), vacated on other grounds, 772 F.2d 996 (1st Cir.1985) (en banc). It may be true that a venire panel's composition may put a petitioner on notice of a procedural irregularity in some instances, for example in areas where the minority population is so high that the probability of an all-white panel is minuscule. However, where the minority population is comparatively small, the actual composition of the venire panel does not provide reasonable notice of the existence or non-existence of a fair cross-section claim. As the district court persuasively stated in Ambrose:

Ambrose v. Booker, No. 06-13361-BC, 2011 WL 1806426, at *2 (E.D.Mich. May 11, 2011). Here, each petitioner saw only one venire panel, which did not alert any of the petitioners to the systematic underrepresentation. Even the judges and Kent County officials who were confronted with venire panels daily did not notice the underrepresentation. To say that petitioners should have looked at a single venire panel and noticed something that the officials failed to notice for sixteen months rings hollow.2

Even if the composition of one venire panel were relevant, the absence of minorities on a panel would not have alerted petitioners that a fair cross-section claim existed. Due to the small size of the African-American population in Kent County, it is entirely possible that a defendant could receive an all-white juror array even if there were no underrepresentation of African-Americans in the jury pool. Dr. Stephenson concluded as much after conducting a standard deviation test and a binomial test on the Kent County jury pool. Stephenson Report, at 3. Because of Kent County's small African-American population (8.25%) and the small size of venire panels, Dr. Stephenson concluded that one "could expect approximately 2% of all venires to contain no Black or African-American members." Id. at 4. Even if the computer had been glitch-free, petitioners could have received a panel without a single African-American panel member. Given the small minority population in Kent County, if petitioners had faced an all-white jury panel, they could have believed that there was no underrepresentation problem.

Moreover, the Supreme Court has found cause to excuse a procedural default in a factually similar case. In Amadeo v. Zant, a district attorney conspired with the jury commissioners of a Georgia county to systematically exclude African-Americans and women from the jury pool. 486 U.S. 214, 108 S.Ct. 1771, 100 L.Ed.2d 249 (1988). The wrongdoers purposefully underrepresented minorities, but were careful to keep representation within a presumptively valid statistical range (under 10%). The petitioner did not object at trial to the racial composition of the jury venire. Several months later, the petitioner learned of a memorandum that detailed a scheme "to underrepresent black people and women on the master jury lists from which ... juries were drawn." Id. at 217, 108 S.Ct. 1771. The memorandum was hidden in a public record — intermingled with irrelevant data — and only discovered in unrelated litigation. The Supreme Court excused the procedural default, reasoning that the memorandum was not reasonably available to petitioner's lawyers. Id. at 224, 108 S.Ct. 1771. The court based its decision on the fact that petitioner's lawyers did not deliberately bypass a jury challenge. Id. at 226-28, 108 S.Ct. 1771. Similarly, in this case, petitioners could not have discovered the computer glitch prior to trial and did not deliberately bypass a jury challenge. Further, the memorandum in Amadeo was filed in publicly accessible records whereas here the error was contained in a single line of code in the jury selection software. It would make little sense to excuse a failure to discover a publicly filed memorandum, but require petitioners to scour non-public programming codes of jury selection software. Accordingly, under the reasoning of Amadeo, petitioners' defaults are excused.

Respondents attempt to distinguish Amadeo, but are ultimately unsuccessful. Respondents first argue that, unlike in these cases, the officials in Amadeo deliberately reduced the minority representation in the juror pool. Although this may be true, the focus of the cause-and-prejudice test is not on the intent of the officials but rather on petitioners' knowledge of the underrepresentation. For example, if a government official intended to conceal an intentional underrepresentation, there would be no good cause to excuse procedural default if the petitioner discovered the scheme. Therefore, a government official's lack of intent is irrelevant.

Respondents do not persuasively explain how the disparity in this case was any more detectable than the one in Amadeo. Respondents argue that the scheme in Amadeo was specifically tailored to avoid detection by keeping the underrepresentation within statistically presumptive levels. They contend that systematic underrepresentation is discoverable unless a scheme exists that was specifically tailored to avoid detection. It is true that the government officials in Amadeo endeavored to conceal the underrepresentation by keeping it to less than 10% absolute disparity. Amadeo, 486 U.S. at 218, 108 S.Ct. 1771. However, in these cases, the absolute disparity was only 3.45%; the underrepresentation would be at least as difficult to detect.

Policy considerations also support this reasoning. First, finding good cause to excuse the default does not undermine the purpose of the contemporaneous objection rule. The purpose of preservation requirements, such as the objection rule, is to allow the trial court to correct errors and ensure that a trial is fair in the first instance. See United States v. Young, 470 U.S. 1, 15, 105 S.Ct. 1038, 84 L.Ed.2d 1 (1985). In the ordinary case, a contemporaneous objection allows a trial court to fix an error at the outset of trial, saving valuable time and resources. See United States v. Lowry, No. 91-6169, 1992 WL 92746, at *4 (6th Cir. Apr. 22, 1992). However, in this case, even if petitioners had objected prior to the empanelment of the jury, the trial court could not have remedied the error. Had the trial court granted the hypothetical contemporaneous objection, it would have dismissed the jury panel and asked the Kent County Commissioners to draw a new panel. This panel would have come from the same jury pool, affected by the same yet-to-be discovered computer glitch. The underrepresentation would have persisted even if petitioners had objected at trial.

Second, the procedural default rule was not intended to bar defaults such as petitioners'. The Supreme Court has made clear that the test is intended to prevent defense counsel from defaulting in state court for strategic gain. As the Court stated, "defense counsel may not make a tactical decision to forgo a procedural opportunity — for instance, an opportunity to object at trial or to raise an issue on appeal — and then, when he discovers that the tactic has been unsuccessful, pursue an alternative strategy in federal court." Reed v. Ross, 468 U.S. 1, 14, 104 S.Ct. 2901, 82 L.Ed.2d 1 (1984) (internal citations omitted). In this case, respondents do not — nor could they — contend that petitioners tactically declined to raise a pre-trial challenge to their venire panels. Petitioners' claims are premised on the fact that they were unaware of the defects in the challenged jury selection procedure. Respondents would be hard-pressed to argue that petitioners tactically relied on the computer glitch of which they did not know.

Third, were we to hold that petitioners cannot demonstrate cause, we would be instituting a de facto requirement that defense attorneys in Kent County automatically raise fair cross-section objections. A prudent lawyer in such a situation would recognize that a failure to raise a fair cross-section claim would bar a later assertion of the right should a computer glitch — or even a scheme such as the one in Amadeo — come to light. Such a result would run counter to the lawyer's duty to avoid frivolous litigation, as well as the procedural default doctrine's concern with judicial efficiency. Instead, finding cause to excuse petitioners' failure to object would leave open a critical safety valve to deal with real fair-cross-section claims.

Respondents rely on a number of our cases in this context, but these cases are factually or analytically distinguishable. First, respondents rely on United States v. Boulding, a case involving two federal defendants who failed to object to the racial diversity of the jury panel until after the jurors were sworn. 412 Fed.Appx. 798, 800 (6th Cir.2011). Defendants' counsel explained that he had not objected during jury selection because he did not want to offend or alienate the jury. Id. at 802. We held that "defendants' counsel had ample opportunity to make an objection, and could have done so in a sidebar or requested a recess if concerned about the objection's effect on the jury." Id. at 802. This case is distinguishable from the instant case in at least two ways. First, the default in Boulding was precisely the type of strategic default contemplated by the court in Reed. Counsel wanted to avoid offending the jury, so he withheld his objection until after the jury was empaneled. Here, respondents cannot show that petitioners strategically withheld their objections. Second, the Boulding defendant knew of the potential objection but chose to wait to object until it was untimely. Petitioners, by contrast, did not know of the underrepresentation of minorities in the jury pool until months after their trials. This again goes to the motive of the failure to object: in Boulding it was strategic, here it was unwitting.

Notwithstanding respondents' arguments to the contrary, United States v. Blair, 214 F.3d 690, 699 (6th Cir.2000), is also distinguishable. In Blair, the defendant "failed to recognize or chose to ignore a potential challenge to the jury selection plan" because this court had not yet decided that the plan violated the Equal Protection Clause. Blair held that a defendant must raise a claim if the factual basis is known, even if no court has ruled on the legal implications of those specific facts. Here, in contrast, petitioners did not know of the factual basis of the claim.

Finally, respondents argue that because other defendants are able to raise fair cross-section claims by looking at the venire panel, petitioners should have too. However, the cases on which respondents rely do not involve a similarly subtle absolute disparity or a hidden cause of underrepresentation. See United States v. Buchanan, 213 F.3d 302, 309 (6th Cir.2000) (overturned on other grounds); United States v. Jackman, 46 F.3d 1240, 1243 (2d Cir.1995). In the ordinary case, fair cross-section claims should be raised at jury selection. However, where the underrepresentation is as obscure as the one in this case — due to a small minority population and a small absolute disparity — a failure to object must be excused.

2. Prejudice for Procedural Default

Having shown cause, petitioners must show actual prejudice to excuse their default, even if the error is structural. The Supreme Court has held that a petitioner must show "actual prejudice" to excuse a default. In Francis v. Henderson, 425 U.S. 536, 96 S.Ct. 1708, 48 L.Ed.2d 149 (1976), a defendant was convicted in state court of murder. He first raised a challenge to the racial composition of his grand jury in his § 2254 petition. The Court held that a petitioner must show both cause and "actual prejudice" to excuse the procedural default. The Court stated, "[t]he presumption of prejudice which supports the existence of the right is not inconsistent with a holding that actual prejudice must be shown in order to obtain relief from a statutorily provided waiver for failure to assert it in a timely manner." Id. at 542 n. 6, 96 S.Ct. 1708 (quoting Davis v. United States, 411 U.S. 233, 245, 93 S.Ct. 1577, 36 L.Ed.2d 216 (1973)) (the "Henderson-Davis rule").

This circuit has accordingly declined to presume prejudice for the purposes of procedural default when considering structural error claims, although we have not yet addressed the issue in the context of a fair cross-section claim. For example, in Keith v. Mitchell, 455 F.3d 662, 673 (6th Cir. 2006), a defendant failed to object to the exclusion of three scrupled jurors3 during his murder trial. Defendant relied on Gray v. Mississippi, 481 U.S. 648, 107 S.Ct. 2045, 95 L.Ed.2d 622 (1987), in which the Supreme Court stated that the exclusion of a scrupled juror is not subject to the harmless error analysis. The Keith defendant sought to extend Gray's holding to "obviate any requirement for him to demonstrate prejudice resulting from his procedural default." Keith, 455 F.3d at 674. This court declined in part because Gray had not been extended by the Supreme Court or any federal court of appeals to the procedural default context. More importantly, we reasoned that extending Gray to the procedural default context would require us to ignore the fundamental differences between direct and collateral review in the federalist system. Id. at 675. We concluded that the argument for the "extension of Gray is weakened, not because the issue is any less important, but because considerations of federalism and respect for the state trial process demand that it be so." Id. The procedural default doctrine is "grounded in concerns of comity and federalism"; therefore, "[w]hen a federal court vacates the judgment of a state trial court, a showing of actual prejudice is required to insure that defaulted claims are considered only when they will have made a difference." Id.

This court has expressly found the Henderson-Davis rule to apply to a jury selection claim, albeit in dicta on a direct review of a federal appeal. In United States v. Ovalle, 136 F.3d 1092 (6th Cir. 1998), four defendants raised challenges to the Eastern District of Michigan's jury selection plan on direct appeal. Two of the defendants raised timely objections to the plan; therefore, we determined that all defendants were entitled to relief because the state court had the opportunity to address the problem. Id. at 1108-09. However, we noted that the non-objecting defendants would not have otherwise been entitled to relief because they had failed to show cause and actual prejudice under the Henderson-Davis rule, even though the error was structural. Petitioner attempts to distinguish this case by arguing that it was a federal case, applying Federal Rule of Criminal Procedure 12. However, in Henderson, the Supreme Court explicitly rejected the notion that the federal courts could strictly impose procedural requirements and then review a state's requirements in a different fashion. The Court stated, in pertinent part:

Henderson, 425 U.S. at 541, 96 S.Ct. 1708. Ovalle reflects this court's "actual prejudice" requirement.

Other circuits have required a petitioner to show "actual prejudice," even where the petitioner claimed a structural error had occurred. In Purvis v. Crosby, 451 F.3d 734 (11th Cir.2006), the Eleventh Circuit refused to presume prejudice where an ineffective assistance of counsel claim arose from a failure to object to a structural error. The court reasoned that allowing the petitioner to dress up a procedural default as an ineffective counsel claim would allow the defendant to circumvent Henderson and Davis. Id. at 742-43. Similarly, the Seventh Circuit declined to presume prejudice for the purposes of procedural default notwithstanding a petitioner's claims that alleged perjured testimony and an alleged Brady violation constituted structural error. Ward v. Hinsley, 377 F.3d 719, 725 (7th Cir.2004). The Seventh Circuit reasoned that "[t]he procedural default doctrine is grounded in concerns of comity and federalism. These concerns are in no way diminished if the federal claim raised before the federal habeas court is one of structural error." Id. at 726 (internal citations and quotation marks omitted).

Comity requires us to avoid overturning defaulted claims absent a showing of actual prejudice. Given the Supreme Court's express language, and the procedural default rule's roots in comity and federalism, a petitioner must show that he was actually prejudiced regardless of the nature of the underlying constitutional claim. The purpose of the procedural default rule is to protect state court procedures from undue federal intervention. Federal courts are accordingly closely circumscribed in their review of defaulted claims. Although the nature of the procedural right is important, federal courts should not reverse state court decisions unless a petitioner can show that the outcome would have been different.

In arguing for presuming prejudice on habeas in this case, petitioners cite a number of our cases that are distinguishable. Quintero v. Bell, 368 F.3d 892 (6th Cir. 2004), is distinguishable because, in that case, counsel's error was so egregious that prejudice could not be reasonably disputed. Trial counsel had failed to object to the inclusion of seven jurors that had served on juries that had convicted petitioner's co-conspirators. Id. at 893. We held that failing to object was an "abandonment of `meaningful adversarial testing' throughout the proceeding," which made "`the adversary process presumptively unreliable.'" Id. (quoting United States v. Cronic, 466 U.S. 648, 659, 104 S.Ct. 2039, 80 L.Ed.2d 657 (1984)). The same cannot be said about the failure to raise a fair cross-section claim. In Quintero, the disputed jurors had already convicted petitioner's co-conspirators. A fair cross-section claim does not present the same danger because there is not the same inevitability of result.

Petitioners also rely on language in Johnson v. Sherry, 586 F.3d 439, 444 (6th Cir.2009), a case in which we addressed a defaulted claim arising from a courtroom closure. We suggested a "strong likelihood" that if the performance was deficient, it would be deemed prejudicial, reasoning in part that the right to a public trial is a structural guarantee. The tentative and conditional nature of this language — in a case that, unlike this one, involved Strickland prejudice regarding a court closure claim — is not sufficient to overcome our reading of the clear import of Supreme Court and Sixth Circuit precedent to the jury selection cases we have before us.

Finally, petitioners parse language from three Supreme Court cases, but cannot show how these cases require a different result. In United States v. Frady, 456 U.S. 152, 170, 102 S.Ct. 1584, 71 L.Ed.2d 816 (1982), for instance, the Court's holding was that the petitioner had to show actual prejudice to overcome his default in not objecting to a jury instruction, and that he had not done so. The Court rejected Frady's argument that prejudice was not required for the jury instruction claim, without saying anything about whether presumed prejudice for the jury instruction claim would meet the "actual prejudice" requirement for overcoming procedural default. Second, petitioners claim that the Court in Amadeo v. Zant declined to impose an "actual prejudice" requirement. However, the Amadeo court declined to reach the issue of prejudice because it was conceded by the government below. 486 U.S. at 228 n. 6, 108 S.Ct. 1771. Third, petitioners argue that Justice Thomas's dissent from the denial of a writ of certiorari in Bell v. Quintero, ___ U.S. ___, 125 S.Ct. 2240, 161 L.Ed.2d 506 (2005), somehow demonstrates Henderson is no longer valid. Justice Thomas noted that "[t]he Court of Appeals' [s] holding also rests on a confusion — the idea that the presence of a structural error, by itself, is necessarily related to counsel's deficient performance and warrants a presumption of prejudice." Id. at 2242. Because the writ was not granted, petitioner infers that the majority rejected Justice Thomas's opinion. However, there are a multitude of reasons that certiorari may not have been granted. The fact that Justice Thomas reiterated the Henderson-Davis rule in his dissent does not mean the majority took the other position.

Because the district courts here did not address actual prejudice, a remand is necessary. We are then left with the question of the proper standard on remand. We are guided in part by the Eleventh Circuit's analysis of a similar question in Hollis v. Davis, 941 F.2d 1471, 1480 (11th Cir.1991). In that case, the petitioner claimed that his counsel was ineffective for failing to object to Alabama's systematic exclusion of African-American jurors from grand and petit juries. To excuse this default, the Hollis court required that petitioner show actual prejudice, which involved determining whether there was a reasonable probability that "a properly selected jury [would] have been less likely to convict." Id. at 1482. The Eleventh Circuit's analysis is persuasive. The most important aspect to the inquiry is the strength of the case against the defendant.4 As the Eleventh Circuit reasoned, "a transcript could show a case against [petitioner] so strong, and defense so weak, that a court would consider it highly improbable that an unbiased jury could acquit." Id. at 1483 (internal quotation marks omitted). In that circumstance, actual prejudice would not be shown.

Although the instant petitions do not involve a Strickland claim, this standard is appropriate because it balances the competing demands of constitutionally protected equal protection interests and comity toward the state courts. We recognize that the application of the actual prejudice standard in cases such as these presents a particularly challenging charge to the district courts below to answer the question, "what would have happened?" The law nonetheless requires that the question be answered — with a careful look at the transcripts involved, and with judgment that takes into account a fair balance of the competing interests of comity toward the final judgments of the state's criminal processes and the protection of constitutional equal protection interests.

III.

The district court orders in all three cases are reversed, and the cases are remanded for further proceedings consistent with this opinion.