William Bracey v. Superintendent Rockview SCI

986 F.3d 274

• Court: Court of Appeals for the Third Circuit
• Decided: January 20, 2021
• Docket: 17-1064
• Status: Published

Opinion

PRECEDENTIAL

UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT ________________

No. 17-1064 ________________

WILLIAM BRACEY, Appellant

v.

SUPERINTENDENT ROCKVIEW SCI; ATTORNEY GEN- ERAL OF THE STATE OF PENNSYLVANIA; DISTRICT ATTORNEY OF DAUPHIN COUNTY ________________

Appeal from the United States District Court for the Middle District of Pennsylvania (D.C. No. 3:11-cv-02329) U.S. District Judge: Hon. Edwin M. Kosik ________________

Argued July 2, 2020 Before: KRAUSE, PHIPPS, Circuit Judges, and BEETLESTONE,* District Judge

(Opinion filed: January 20, 2021)

David R. Fine Travis N. Gery [ARGUED] K&L Gates 17 North Second Street 18th Floor Harrisburg, PA 17101 Counsel for Appellant

Ryan H. Lysaght [ARGUED] Dauphin County Office of District Attorney 101 Market Street Harrisburg, PA 17101 Counsel for Appellee District Attorney of Dauphin County ________________________

OPINION OF THE COURT ________________________

KRAUSE, Circuit Judge.

The Antiterrorism and Effective Death Penalty Act (AEDPA) strictly regulates a defendant’s ability to collaterally

* Honorable Wendy Beetlestone, District Judge, United States District Court for the Eastern District of Pennsylvania, sitting by designation.

2 attack his final judgment of conviction, including by requiring that he exercise “due diligence” along the way. See 28 U.S.C. §§ 2244(b)(2)(B)(i), 2244(d)(1)(D), 2254(e)(2)(A)(ii), 2255(f)(4). This appeal involves § 2244(d)(1)(D), which re- quires a defendant attacking his state conviction to petition the federal courts within one year of “the date on which the factual predicate of the claim . . . could have been discovered through the exercise of due diligence.” Id. § 2244(d)(1)(D).

At the heart of this appeal are two questions about that requirement. First, if a defendant reasonably expects that the prosecution has complied with its obligations under Brady v. Maryland, 373 U.S. 83 (1963), but later discovers that the pros- ecution instead withheld material exculpatory evidence in its possession, does the fact that the withheld evidence could have been found in public records mean the defendant has failed to “exercise . . . due diligence”? Second, what, if anything, is the relevance of our en banc decision in Dennis v. Sec’y, 834 F.3d 263 (3d Cir. 2016), which rejected the notion that a defendant has a duty to search public records for undisclosed Brady ma- terial because, we held, Brady focuses entirely on the prosecu- tion’s affirmative duty of disclosure and permits defendants to expect that government officials will comply with that duty?

Before we can reach those important questions, how- ever, we must first address a threshold issue. This appeal comes to us not from the dismissal of Appellant William Bracey’s underlying habeas petition, but from the denial of his Federal Rule of Civil Procedure 60(b) motion for reconsidera- tion of that dismissal in light of Dennis. So we begin by re- solving whether the appeal of a Rule 60(b) ruling of this kind requires a certificate of appealability (COA) under 28 U.S.C. § 2253(c)(1)(A), and, if so, whether a COA is justified here.

3 We answer both in the affirmative, reaffirming our case law on the COA requirement in this context and concluding that Bracey has shown he is entitled to one.

Returning, then, to the questions at the core of this case, we hold as follows. Case law from the Supreme Court, this Circuit, and other Courts of Appeals establishes that a due dil- igence requirement like the one in § 2244(d)(1)(D) demands a highly fact- and context-specific inquiry, one that depends on the characteristics and reasonable expectations of someone in the petitioner’s shoes. That is where Dennis comes in. Dennis effected a material change in Circuit law with respect to the reasonable expectations of a Brady claimant: While we had previously suggested that defendants had to search for excul- patory evidence themselves, Dennis made clear that a defend- ant can reasonably expect—and is entitled to presume—that the government fulfilled its Brady obligations because the prosecution’s duty to disclose is absolute and in no way hinges on efforts by the defense. By altering the factual predicate and baseline expectations for Brady claims, Dennis correspond- ingly changed what § 2244(d)(1)(D)’s “due diligence” require- ment demands of Brady claimants. Yet in denying Bracey’s Rule 60(b) motion, the District Court did not recognize the ef- fect Dennis had on the relevant decisional law and did not en- gage in the multifactor analysis our case law requires. We therefore will vacate and remand to the District Court for an appropriate consideration of Bracey’s Rule 60(b) motion.1

1 We express our gratitude to David R. Fine and Travis N. Gery of K&L Gates LLP for accepting this matter pro bono, and we commend the superb quality of their briefing and argu- ment in this case. Lawyers who act pro bono fulfill the highest

4 I. FACTUAL BACKGROUND2

Bracey was convicted of murder in 1995. The Com- monwealth’s case relied heavily on the testimony of two coop- erators: Thomas Plummer, Jr., who was an alleged eyewitness to the murder, and Sylvester Bell, who claimed Bracey had confessed to him. At trial, the Government also elicited testi- mony from Plummer and Bell that they had received favorable plea agreements on certain charges in exchange for their testi- mony. Bracey was sentenced to life in prison, and his subse- quent appeal and state habeas petitions were unsuccessful.

In 2010, Bracey learned the Commonwealth had dis- closed only some of the cases that were pending against Plum- mer and Bell.3 As it turned out, just a few months after Bracey was convicted, Plummer pleaded guilty to charges that had not been disclosed to Bracey or the jury. Similarly, the

service that members of the bar can offer to indigent parties and to the legal profession. 2 These facts are drawn from the Superior Court of Pennsylvania’s decision denying Bracey’s third state postcon- viction petition, the Magistrate Judge’s report and recommen- dation in favor of dismissing Bracey’s federal habeas petition, and the District Court’s opinion adopting the Magistrate Judge’s recommendations. 3 According to Bracey, and as set forth in his sister’s affidavit, Bracey first learned of these additional cases when his sister was trying to reinvestigate his case and contacted him about inconsistencies between the docket sheets and the facts elicited at trial.

5 Commonwealth withdrew a “second set of charges” that had been pending against Bell, JA 29, even though, as the Pennsyl- vania Superior Court found, “[t]he Commonwealth did not ad- vise Bracey or the jury of the existence of [that] second set of charges,” JA 30.

Based on this newly discovered information, Bracey pe- titioned for relief under Pennsylvania’s Post Conviction Relief Act, raising claims under Brady and Giglio v. United States, 405 U.S. 150 (1972).4 The Court of Common Pleas dismissed Bracey’s petition as time barred, ruling that the factual basis of the claim could have “been ascertained [earlier] by the exercise of due diligence.” 42 Pa. Cons. Stat. § 9545(b)(1)(ii). The Su- perior Court affirmed, reasoning that if Bracey “had exercised due diligence, he . . . could have reviewed the dockets and de- termined the sentences imposed [on Plummer and Bell’s other charges] at that time” because “[c]riminal dockets are a matter of public record.” JA 30 (internal quotation marks and citation omitted).

Bracey then filed a federal habeas petition in 2011, again asserting Brady claims based on the allegedly withheld material exculpatory evidence. The District Court dismissed the petition as untimely under 28 U.S.C. § 2244(d)(1)(D). It reasoned that—regardless of the prosecution’s “alleged lack of full disclosure, including the specific terms of the plea 4 As the Supreme Court made clear in United States v. Bagley, 473 U.S. 667, 676 (1985), and Giglio, 405 U.S. at 153– 54, Brady’s rule applies equally to evidence that could have been used for impeachment purposes. We will therefore refer to the impeachment material at issue in this case as Brady ma- terial.

6 agreements and possible maximum penalties [for] each of the witnesses,” JA 74—§ 2244(d)(1)(D) obliges a defendant who is aware of witnesses’ favorable plea agreements to continually seek out “the full extent of those plea agreements,” even after the defendant is convicted, JA 75. The District Court con- cluded that “because the full extent of the plea agreements and the sentences received by the witnesses were a matter of public record,” Bracey “could have found the factual predicate of [his Brady] claim through the exercise of due diligence well before October 2010,” id., meaning that he had filed his petition more than one year after the “factual predicate” for his Brady claim “could have been discovered through the exercise of due dili- gence,” 28 U.S.C. § 2244(d)(1)(D). Bracey sought to appeal that ruling in 2012, but in 2013, we denied a COA in view of then-existing case law.

Three years later, another development prompted Bracey to renew his quest for relief: We decided Dennis, 834 F.3d 263. In Dennis, the Commonwealth argued that a defend- ant’s failure to seek out exculpatory materials in the public do- main barred him from bringing a Brady claim. See id. at 289. We squarely rejected that argument, holding that a defendant has no burden to “scavenge for hints of undisclosed Brady ma- terial” even if the material part could be found in public rec- ords. Id. at 290 (quoting Banks v. Dretke, 540 U.S. 668, 695 (2004)). Rather, the prosecution’s “duty to disclose under Brady is absolute—it does not depend on defense counsel’s ac- tions”—and the defense is “entitled to presume that prosecu- tors have ‘discharged their official duties.’” Id. (quoting Banks, 540 U.S. at 696). Thus, “[t]o the extent [our previous case law] ha[d] considered defense counsel’s purported obli- gation to exercise due diligence to excuse the government’s non-disclosure of material exculpatory evidence,” Dennis

7 definitively “reject[ed] that concept as an unwarranted dilution of Brady’s clear mandate.” Id. at 293.

Bracey promptly moved for reconsideration under Rule 5 60(b), contending that, in light of Dennis, the District Court had erred in dismissing his petition under § 2244(d)(1)(D). Specifically, he argued that because “there is no due diligence requir[e]ment under Brady for defendants to discover impeach- ment material and it is stric[t]ly the duty of the prosecutor to provide this information,” § 2244(d)(1)(D) likewise does not require petitioners in his position to undertake efforts to find exculpatory material. JA 125–26. Unmoved, the District Court issued a one-page summary denial, ruling—without ref- erence to, much less discussion of, Dennis—that Bracey’s mo- tion “raise[d] the same arguments” it had already rejected. JA 7. Bracey then requested a COA, which a motions panel of our Court referred to the merits panel.

II. DISCUSSION6

We certified three questions: (A) whether a COA is re- quired in an appeal from the denial of a Rule 60(b) motion

5 Bracey brought his motion under Rule 60(b)(6), which allows a court to “relieve a party or its legal representative from a final judgment, order, or proceeding for . . . any . . . reason that justifies relief.” Fed. R. Civ. P. 60(b)(6). For simplicity, we refer to Bracey’s motion as one brought under Rule 60(b). 6 The District Court had jurisdiction under 28 U.S.C. §§ 2241(a) and 2254(a), and we have jurisdiction under 28 U.S.C. §§ 1291 and 2254(a). “We exercise plenary review over the statute of limitations issue,” Swartz v. Myers, 204 F.3d 417, 419 (3d Cir. 2000), and review the denial of a Rule 60(b)

8 seeking reconsideration of the dismissal of a federal habeas pe- tition on procedural grounds; (B) if a COA is required, whether one should be granted here; and (C) if a COA is granted, whether the District Court abused its discretion in denying Bracey’s Rule 60(b) motion without considering Dennis’s ef- fect on its previous decision dismissing Bracey’s habeas peti- tion. We address each in turn.

A. Whether a COA Is Required

The first question could have been framed as whether Morris v. Horn, 187 F.3d 333 (3d Cir. 1999), remains good law. AEDPA requires a petitioner to obtain a COA before ap- pealing any “final order” in a federal habeas proceeding chal- lenging a state conviction. 28 U.S.C. § 2253(c)(1)(A). Morris held this requirement also applies to appeals, like this one, from the denial of a Rule 60(b) motion seeking reconsideration of the denial of habeas relief, whether the denial of habeas re- lief was on substantive or procedural grounds. 187 F.3d at 340–41. In so holding, we joined a near-consensus of our sister circuits. See United States v. Winkles, 795 F.3d 1134, 1139 (9th Cir. 2015) (collecting cases). Yet Bracey asks us to revisit that holding, arguing Morris was abrogated by two later Su- preme Court decisions: Harbison v. Bell, 556 U.S. 180 (2009), and Gonzalez v. Crosby, 545 U.S. 524 (2005).7 We are unper- suaded.

motion for abuse of discretion, Morris v. Horn, 187 F.3d 333, 341 (3d Cir. 1999). 7 Bracey also points to our statement in Wilson v. Sec’y, 782 F.3d 110 (3d Cir. 2015), that “the vitality of [Morris] is undermined somewhat by . . . Harbison.” Id. at 115. As

9 Harbison, which involved the same provision at issue in Morris and here, asked whether § 2253(c)’s COA requirement applied to an appeal of a district court order resolving a collat- eral issue—namely, a motion to expand the scope of represen- tation for appointed counsel to include a clemency proceeding. 556 U.S. at 182–83. Concluding that § 2253(c) applied only to appeals of “final orders that dispose of the merits of a habeas corpus proceeding,” the Court held that Harbison was not re- quired to obtain a COA before appealing the district court’s ruling. Id. at 183. Harbison did not, however, distinguish be- tween “final orders that dispose of a habeas corpus proceeding” on substantive grounds and ones that “dispose” of the “pro- ceeding” on procedural grounds. See id.

Gonzalez arose in a different context and dealt with a different issue: when a district court, under 28 U.S.C. § 2244(b), may consider a Rule 60(b) motion without the peti- tioner first having obtained “precertification by the Court of Appeals.” 545 U.S. at 530, 538. Section 2244(b) limits the circumstances in which a petitioner may file “a second or suc- cessive” habeas petition, id. at 526, 530, including by requiring him to “move in the appropriate court of appeals for an order authorizing the district court to consider” the petition, 28 U.S.C. § 2244(b)(3)(A). The question in Gonzalez was when a request for relief, “although labeled a Rule 60(b) motion, is in substance a successive habeas petition” that triggers § 2244(b). 545 U.S. at 530–32. To answer that question, the Court distinguished between a “true Rule 60(b) motion” and a

Bracey acknowledges, this was merely dicta, and tentative dicta at that. Our decision today puts to rest any confusion Wilson may have generated.

10 successive habeas petition merely masquerading as a Rule 60(b) motion: The former attacks “some defect in the integrity of the federal habeas proceedings,” such as the application of AEDPA’s statute of limitations or rules on procedural default, so the district court may consider it without a COA, while the latter attacks “the substance of the federal court’s resolution of a claim on the merits” and so requires a COA before the district court may consider it. Id. at 531–32, 534. Because Gonzalez’s Rule 60(b) motion challenged only a statute-of-limitations rul- ing, it qualified as a true Rule 60(b) motion, i.e., not subject to § 2244(b)’s limitations on successive petitions and reviewable by the district court “without precertification.” Id. at 533, 535– 38.

Although Harbison and Gonzalez addressed different issues in different contexts, Bracey links them syllogistically: Because a COA is required only to appeal a final order on “the merits” of a habeas petition, Harbison, 556 U.S. at 183, and because a Rule 60(b) disposition on procedural grounds is not a disposition “on the merits,” Gonzalez, 545 U.S. at 532, a COA is not required to appeal a Rule 60(b) disposition on pro- cedural grounds. Unfortunately for Bracey, that line of reason- ing is less logic than it is word play.

As the Court put it in Gonzalez, a term like “‘on the merits’ has multiple usages.” Id. at 532 n.4. And fatal to Bracey’s syllogism is that those usages were clearly different in Harbison and Gonzalez. Whereas Gonzalez used “the mer- its” to distinguish a “true Rule 60(b) motion” attacking a pro- cedural defect from a disguised successive habeas petition at- tacking the substantive resolution of a habeas claim, see id. at 531–32, Harbison used “the merits” to distinguish “final or- ders” that conclude the habeas proceeding itself from those

11 orders that merely resolve a collateral issue, see 556 U.S. at 183. And while Harbison excluded from § 2253(c)’s COA re- quirement orders that do not conclude habeas proceedings, it made no further distinction among those orders that do con- clude proceedings based on whether the disposition was sub- stantive or procedural in nature.

Neither Harbison nor Gonzalez, therefore, disturbed our holding in Morris—or, for that matter, the identical conclu- sions of most of our sister circuits—that a COA is required when a petitioner appeals the denial of a Rule 60(b) motion seeking reconsideration of a dismissal of a habeas petition, even if that dismissal was on procedural grounds. To the con- trary, Gonzalez went out of its way to observe that “[m]any Courts of Appeals have . . . requir[ed] a habeas petitioner to obtain a COA as a prerequisite to appealing the denial of a Rule 60(b) motion,” 545 U.S. at 535—a “construction of § 2253” that the Court praised as “plausible” and “effective,” with a “sound[] basis in the statute,”8 id. at 535 n.7.

8 Nor can we accept that Harbison and Gonzalez silently overruled so foundational a precedent as Slack v. McDaniel, 529 U.S. 473 (2000). Yet that is what Bracey would have us believe. Whereas Slack held that § 2253(c)’s COA require- ment applies to appeals of the denial of habeas petitions “based on procedural grounds,” 529 U.S. at 484, Bracey reads Harbi- son to require a COA only for denials “on the merits,” meaning on substantive grounds. And even if Bracey’s reading could be limited in a principled way to Rule 60(b) rulings, it would still allow petitioners to end-run the § 2253(c) requirement Slack addressed by challenging procedural defects in the ha- beas proceeding in a Rule 60(b) motion and then appealing

12 Reaffirming Morris today puts us in good company. See Winkles, 795 F.3d at 1141–42; Hamilton v. Sec’y, 793 F.3d 1261, 1265–66 (11th Cir. 2015) (per curiam).9 The upshot is that because Morris remains good law, Bracey requires a COA for us to hear his appeal.

B. Whether a COA Should Issue

We decide whether Bracey is entitled to a COA in two steps. First, we ask whether his underlying claim is “debata- ble” on the merits. Slack v. McDaniel, 529 U.S. 473, 484 (2000). His burden on this point is a light one, id.: He must show only that his claim represents “something more than the

from the denial of that motion without first obtaining a COA. That is neither what Congress has prescribed nor how the Su- preme Court has viewed the COA requirement, and as always, we are loath to conclude the Court has discarded an existing precedent without having said so more clearly. See Shalala v. Ill. Council on Long Term Care, Inc., 529 U.S. 1, 18 (2000). 9 United States v. McRae, 793 F.3d 392 (4th Cir. 2015), is not to the contrary. There, the Fourth Circuit held that a COA was not required to review “a dismissal of a Rule 60(b) motion on jurisdictional grounds” because that ruling was “so far removed from the merits of the underlying habeas petition that it [could not] be said to be” a final order disposing of the merits of the habeas proceeding. Id. at 400. Even assuming McRae is rightly decided, its holding is “narrow[],” id. at 399, and would not extend to the non-jurisdictional statute of limi- tations that formed the basis for the dismissal of Bracey’s peti- tion here, see United States v. Bendolph, 409 F.3d 155, 160 (3d Cir. 2005).

13 absence of frivolity or the existence of mere good faith,” Mil- ler-El v. Cockrell, 537 U.S. 322, 338 (2003) (internal quotation marks and citation omitted). Indeed, he need not even prove that “some jurists would grant the petition,” as “a claim can be debatable even though every jurist of reason might agree, after the COA has been granted and the case has received full con- sideration, that [the] petitioner will not prevail.” Id. Second, we ask whether “jurists of reason would find it debatable whether the district court was correct in its procedural ruling.” Slack, 529 U.S. at 484. On this second point, too, Bracey’s burden is light; he must merely make a “credible showing that the District Court’s denial of his Rule 60(b) motion was erro- neous.” Morris, 187 F.3d at 341. Bracey has made both show- ings.

On the merits, we perform a “threshold inquiry regard- ing” the underlying claim, Pabon v. Mahanoy, 654 F.3d 385, 393 (3d Cir. 2011), “without ‘full consideration of the factual or legal bases adduced in support of th[at] claim[],’” Buck v. Davis, 137 S. Ct. 759, 773 (2017) (quoting Miller-El, 537 U.S. at 336). Under that “limited . . . inquiry,” id. at 774, we con- clude Bracey’s Brady claim is debatable. Although Bracey’s claim that two key government witnesses at his trial received undisclosed benefits in exchange for their testimony has not been factually developed (because it was dismissed as un- timely), it has all the features of a paradigmatic Brady claim. See United States v. Risha, 445 F.3d 298, 299 (3d Cir. 2006) (“[A] failure of the prosecution to disclose impeachment evi- dence, coupled with a duty to disclose, would result in a Brady violation.”). At this stage, Bracey’s claim is sufficiently debat- able to “deserve encouragement to proceed further.” Miller- El, 537 U.S. at 336, 338 (internal quotation marks and citation omitted).

14 And for the reasons we explain below, Bracey has also credibly shown the District Court abused its discretion by denying his Rule 60(b) motion for reconsideration without so much as mentioning Dennis, let alone considering the factors it was required to consider. We therefore conclude that Bracey is entitled to a COA and that, under § 2253(c)(1)(A), we may hear Bracey’s appeal.

C. Whether the Rule 60(b) Denial Was an Abuse of Discretion

We come at last to the heart of the case: whether the Dis- trict Court abused its discretion when it denied Bracey’s Rule 60(b) motion. The standards for evaluating a 60(b) motion are well established. A court may grant equitable relief under Rule 60(b)(6) “in extraordinary circumstances where, without such relief, an extreme and unexpected hardship would occur.” Cox v. Horn, 757 F.3d 113, 120 (3d Cir. 2014) (citation and internal quotation marks omitted). The petitioner “bears the burden of establishing entitlement to such equitable relief,” but the District Court “must consider the full measure of any properly presented facts and circumstances attendant to the [petitioner’s] request.” Id. at 122.

Where, as here, a petitioner seeks Rule 60(b) relief based on an intervening change in the law, we must address three is- sues: First, we ask whether the asserted change is material to the basis on which the district court initially denied habeas re- lief. See Norris v. Brooks, 794 F.3d 401, 404–05 (3d Cir. 2015). If it is, we then evaluate whether the district court ana- lyzed the petitioner’s Rule 60(b) motion in accordance with a multifactor analysis we outlined in Cox, which includes, among other things, a consideration of the effect of the change in decisional law and an assessment of “the merits of [the]

15 petitioner’s underlying . . . claim.” 757 F.3d at 124. Finally, we determine the proper disposition on appeal: If the District Court undertook the requisite multifactor analysis, we review the merits of its ruling for abuse of discretion, id. at 118, but if it did not engage in that analysis or “we cannot determine from what it wrote whether the Court considered [the relevant] fac- tors,” id. at 120, then the District Court per se abused its dis- cretion and we ordinarily remand, because “[t]he grant or de- nial of a Rule 60(b)(6) motion is an equitable matter left, in the first instance, to the discretion of a district court,” id. at 124.

We consider each of these issues in order.

1. Whether Dennis constituted a change in decisional law material to the dismissal of Bracey’s Brady claim on timeliness grounds

The first question we must decide is whether Dennis constituted a material change in decisional law. See Norris, 794 F.3d at 404. Much depends on how we understand the interaction between what AEDPA requires and what Dennis held. If Dennis did not effect a material change in the law gov- erning the timeliness of a habeas petition like Bracey’s, then the District Court did not abuse its discretion in denying his Rule 60(b) motion. See Norris, 794 F.3d at 404–05; Greene v. Superintendent Smithfield SCI, 882 F.3d 443, 451 (3d Cir. 2018). But if, on the other hand, Dennis did shift the legal ground, then the District Court’s summary denial of Bracey’s Rule 60(b) motion without considering Dennis would stand on seriously shaky footing.

The parties offer competing arguments on this point. According to Bracey, although Dennis arose in the context of

16 proving a Brady claim, it also affects the diligence that § 2244(d)(1)(D) expects of a habeas petitioner hoping to assert a Brady claim. For its part, the Commonwealth contends that Dennis has no relevance whatsoever to § 2244(d)(1)(D) and that, absent a material change in the decisional law, we must affirm the District Court.10

Bracey has the better argument. To understand why, we must first explore the relevant decisional principles, including

10 At the outset, the Commonwealth gives two reasons why, in its view, Dennis is inapposite. Neither is persuasive. First, it asserts that “the information in question” in Dennis “was not publicly available.” Appellee’s Br. 21. Be that as it may, see Dennis, 834 F.3d at 289, the entirety of our “due dil- igence” discussion came in response to an argument that “there was no Brady violation because the [relevant evidence] was publicly available,” id., and our rejection of that argument was definitive, id. at 290–91. Second, the Commonwealth argues that the withheld evidence Bracey discovered in 2010, unlike the alibi evidence in Dennis, was not Brady material at all, ei- ther because it went only to impeachment or because “the wit- nesses in question were already impeached during trial by the same fact,” so “[t]he fact that the Commonwealth allegedly of- fered them more for their testimony than came out at trial would have no effect on the case.” Appellee’s Br. 21. But it is hornbook law that impeachment material qualifies as excul- patory under Brady, see Bagley, 473 U.S. at 676; Giglio, 405 U.S. at 154, and that “[t]he mere fact that a witness has been heavily cross-examined or impeached at trial does not preclude a determination that additional impeachment evidence is mate- rial under Brady,” Dennis, 834 F.3d at 300.

17 that a due diligence requirement like the one in § 2244(d)(1)(D) requires a fact- and context-specific inquiry focused on the characteristics and reasonable expectations of each petitioner. Having done so, we are then well positioned to examine Dennis, particularly the change it wrought in our Court’s articulation of a Brady claim and the implications of that change for the reasonable expectations of Brady claimants. Finally, applying Dennis’s lessons to the principles governing AEDPA’s due diligence requirement, we conclude that a peti- tioner in Bracey’s position does not fail to “exercise . . . due diligence,” 28 U.S.C. § 2244(d)(1)(D), by relying on the pros- ecution’s independent and unflagging obligation to disclose material exculpatory evidence in its possession. Dennis there- fore caused a material change in the relevant law, making Bracey’s Rule 60(b) motion worthy of consideration.

a. Section 2244(d)(1)(D) requires a highly fact- and context-specific analysis of what due diligence requires based on each petitioner’s reasonable expecta- tions

In allowing habeas claims to be filed within one year of “the date on which the factual predicate of the claim[s] . . . could have been discovered through the exercise of due dili- gence,” 28 U.S.C. § 2244(d)(1)(D), AEDPA does not impose a one-size-fits-all requirement. Rather, what due diligence re- quires depends on the circumstances of each petitioner: who he is, what facts he knows, what claim he seeks to bring, and what he can reasonably expect in view of his circumstances and the nature of that particular claim. This conclusion flows from the statutory text and is well settled in the case law of this Circuit and others. It also comports with decisions interpreting

18 identically worded due diligence requirements elsewhere in AEDPA, as well as closely related diligence requirements in the habeas context more generally.

We begin, as is customary, with the text. Ross v. Blake, 136 S. Ct. 1850, 1856 (2016). The subject matter of § 2244(d)(1)(D) is the “factual predicate” of the habeas claim, meaning the facts “out of which the point of law arises,” Fact, Black’s Law Dictionary (6th ed. 1990); see also Fact, Black’s Law Dictionary (11th ed. 2019) (noting that by 1899, “predi- cate fact” meant “[a] fact from which a presumption or infer- ence [of the existence of the claim] arises”). And the trigger for that provision’s one-year clock is the date on which those facts “could have been discovered through the exercise of due diligence.” 28 U.S.C. § 2244(d)(1)(D) (emphasis added). That language is critical. In the law, “due diligence” is a relative term; it is “[s]uch a measure of prudence, activity, or assiduity, as is properly to be expected from, and ordinarily exercised by, a reasonable and prudent man under the particular circum- stances; not measured by any absolute standard, but depending on the relative facts of the special case.” Diligence, Black’s Law Dictionary (6th ed. 1990). Put another way, due diligence is the care “expected from a man of ordinary prudence,” which is “measured with reference to the particular circumstances.” Id. So by its plain terms, § 2244(d)(1)(D) requires us to deter- mine when, in “the particular circumstances” at issue here, we should “expect[]” a “reasonable” petitioner to discover the facts giving rise to a habeas claim.

That is precisely what our case law requires. We have long held “that, to satisfy § 2244(d)(1)(D)’s ‘due diligence’ standard, a prisoner must exercise ‘reasonable diligence in the circumstances.’” Wilson v. Beard, 426 F.3d 653, 660 (3d Cir.

19 2005) (quoting Schlueter v. Varner, 384 F.3d 69, 74 (3d Cir. 2004)). That inquiry “is context-specific,” and “[t]he fact that we require a petitioner in one situation to undertake certain ac- tions does not necessitate that we impose the same burden on all petitioners.” Id. at 661.

We have also laid down important markers guiding courts in assessing what due diligence requires. “It is not enough,” we have cautioned, “that [a petitioner] could have learned about [the factual basis for his claim] by happenstance” or “that [he] could have discovered [it] fortuitously.” Id. at 660. Nor must a petitioner “continuously monitor[] [public sources] for [years] . . . on the unlikely chance that he might learn something which would be useful to his case.” Id. at 661 (internal quotation marks and citation omitted). Rather, § 2244(d)(1)(D) requires that we focus on the “reasonabl[e] ex- pect[ations]” of someone “in [the petitioner’s] position,” be- cause a petitioner will have an obligation to investigate only once he has a “reasonable basis . . . to expect that [investiga- tion] would uncover . . . relevant information.” Id. In short, unless “the petitioner should be expected to take actions which would lead him to the information,” id. at 662, his decision not to investigate “[i]s not a failure to exercise due diligence,”11 id. at 661.

11 Wilson and Schlueter provide a helpful comparison. In Schlueter, the pertinent facts about a defense attorney’s con- flicts of interest were “common knowledge in the . . . relatively small legal community” where the defendant was tried. 384 F.3d at 74. Under those circumstances, we found it “incon- ceivable” that someone in the petitioner’s position would have needed ten years to discover the conflicts. Id. at 71–72, 74. In

20 Other Courts of Appeals have interpreted § 2244(d)(1)(D) in the same manner. In a decision that we later cited in Schlueter, see 384 F.3d at 74, the Seventh Circuit em- phasized that § 2244(d)(1)(D) requires only “reasonable dili- gence” and that courts “should take into account” the peti- tioner’s particular circumstances. Moore v. Knight, 368 F.3d 936, 938–40 (7th Cir. 2004) (quoting Wims v. United States, 225 F.3d 186, 190 n.4 (2d Cir. 2000)). Where, for instance, the petitioner “had [no] reason to suspect the content of [the trial judge’s] communications to the jury was different than she had represented,” his decision not to investigate that communica- tion until he learned facts giving him “reason to suspect a harmful error” did not constitute a lack of diligence under § 2244(d)(1)(D). Id. at 939–40. The Fifth Circuit, relying on both Moore and Wilson, likewise has emphasized that the proper question is “whether the petitioner should be expected to take actions which would lead him to the information . . . under the circumstances.” Starns v. Andrews, 524 F.3d 612, 618–19 (5th Cir. 2008) (quoting Wilson, 426 F.3d at 662).

Wilson, on the other hand, we found that “[n]o person in Wil- son’s position would reasonably expect that the local news would be a source of information relevant to his case,” and, as such, a failure to investigate the relevant facts did not amount to a lack of due diligence even though those facts had “received widespread attention on local newscasts.” 426 F.3d at 660–61. Distinguishing Schlueter, we explained that “[t]he essential question” under § 2244(d)(1)(D) “is not whether the relevant information was known by a large number of people, but whether the petitioner should be expected to take actions which would lead him to the information.” Id. at 661–62 (emphasis added).

21 Other circuits have had much the same to say. See Ford v. Gonzalez, 683 F.3d 1230, 1235 (9th Cir. 2012) (holding that § 2244(d)(1)(D) requires “‘reasonable diligence in the circum- stances’” and looks to “the petitioner’s particular circum- stances” (quoting Schlueter, 384 F.3d at 74)); Wood v. Spen- cer, 487 F.3d 1, 5 (1st Cir. 2007) (concluding that due diligence requires an assessment of what knowledge of the facts “fairly may be imputed to” a reasonable person in the petitioner’s po- sition).

Moreover, because “identical words used in different parts of the same statute are generally presumed to have the same meaning,” IBP, Inc. v. Alvarez, 546 U.S. 21, 34 (2005); accord Azar v. Allina Health Servs., 139 S. Ct. 1804, 1812 (2019), we find further support in judicial interpretations of other due diligence requirements under AEDPA’s umbrella. One example is 28 U.S.C. § 2255 (the cognate to § 2244(d)(1)(D) for federal judgments), which applies a one- year limitations period to motions to vacate a federal criminal sentence that runs from “the date on which the facts supporting the claim or claims presented could have been discovered through the exercise of due diligence.” 28 U.S.C. § 2255(f)(4). Facing “the question of how to implement the statutory man- date that a petitioner act with due diligence in discovering” the facts underlying a claim, the Supreme Court held that the an- swer depends on when the petitioner “is in a position to realize that he has an interest” in pursuing further action. Johnson v. United States, 544 U.S. 295, 308 (2005). Thus, until some- thing in the petitioner’s case “clearly shows that diligence is in order,” § 2255(f)(4) will not be triggered. Id.; see also Jeffer- son v. United States, 730 F.3d 537, 544–45 (6th Cir. 2013) (holding that § 2255(f)(4) requires courts to focus on what a “reasonable person in [the petitioner’s] circumstances would

22 have [done]” based on baseline expectations about the nature of his claim and the government’s and defense’s respective du- ties).

Another example from which we draw guidance is § 2244(b), which calls for dismissal of any claim in a second or successive habeas petition that was not presented in a prior petition unless “the factual predicate for the claim could not have been discovered previously through the exercise of due diligence.” 28 U.S.C. § 2244(b)(2)(B)(i). For that subsection, too, due diligence requires only “as much as could ‘reasonably’ be expected from someone in [the petitioner’s] circumstances.” In re Wogenstahl, 902 F.3d 621, 629 (6th Cir. 2018) (quoting DiCenzi v. Rose, 452 F.3d 465, 470 (6th Cir. 2000)); see also Solorio v. Muniz, 896 F.3d 914, 920–21 (9th Cir. 2018) (hold- ing that for § 2244(b)(2)(B)(i) the question is whether the peti- tioner “is on notice that new evidence might exist” so that due diligence would require him “to investigate further” (emphasis omitted)).

Finally, we find echoes of these principles in decisions addressing other diligence requirements in the habeas context. Consider Williams v. Taylor, 529 U.S. 420 (2000), which ad- dressed an AEDPA provision making evidentiary hearings pre- sumptively unavailable where the petitioner “has failed to de- velop the factual basis of a claim in State court proceedings.”12 12 One exception to that presumptive rule is when the petitioner shows that the claim relies on “a factual predicate that could not have been previously discovered through the ex- ercise of due diligence,” 28 U.S.C. § 2254(e)(2)(A)(ii), but that exception was conceded not to apply in Williams, see 529 U.S. at 430.

23 Id. at 429 (quoting 28 U.S.C. § 2254(e)(2)). Holding that a pe- titioner has not “failed” to develop a record under § 2254(e)(2) when he acted “diligent[ly],” id. at 431–32, the Supreme Court employed analysis that should, by now, sound familiar: “Dili- gence . . . depends upon whether the [petitioner] made a rea- sonable attempt, in light of the information available at the time, to investigate and pursue claims.” Id. at 435. Thus, if “no evidence . . . would have put a reasonable attorney on no- tice” that an investigation would be fruitful, there is no lack of diligence even if the information in question was available through “public records.”13 Id. at 442–43.

13 Our case law on equitable tolling in the habeas con- text lends even more support. To be entitled to equitable toll- ing, a petitioner must show he has “pursu[ed] his rights dili- gently.” Munchinski v. Wilson, 694 F.3d 308, 329 (3d Cir. 2012) (citation omitted). And as with AEDPA’s statutory dil- igence requirements, see id. at 330–31 (citing Schlueter), the equitable diligence requirement demands “reasonable dili- gence, not maximum feasible diligence.” Id. at 330 (quoting Holland v. Florida, 560 U.S. 631, 653 (2010)). “In other words, the diligence inquiry is fact-specific and depends on the circumstances faced by the particular petitioner,” and “there are no bright line rules as to what conduct is insufficient.” Id. at 331; see id. at 330 (explaining that diligence depends on “the petitioner’s overall level of care and caution in light of his or her particular circumstances” (quoting Doe v. Busby, 661 F.3d 1001, 1013 (9th Cir. 2011))); Ross v. Varano, 712 F.3d 784, 799 (3d Cir. 2013) (requiring courts to “consider[]” diligence “in light of the particular circumstances of the case”); see also, e.g., Baldayaque v. United States, 338 F.3d 145, 153 (2d Cir. 2003) (framing the question as whether the petitioner “act[ed]

24 In light of the statutory text and the broad consensus among the Supreme Court, this Circuit, and other Courts of Appeals, we reaffirm that for purposes of § 2244(d)(1)(D), due diligence depends on each petitioner’s circumstances and the nature of the claim asserted, and it requires that we assess, in light of that context, what a petitioner “would [have] reasona- bly expect[ed]” might result from investigative efforts. Wil- son, 426 F.3d at 661. In this case, then, we ask what a reason- able person in Bracey’s position would have expected about improperly withheld Brady material in the prosecution’s pos- session. Enter Dennis.

as diligently as reasonably could have been expected under the circumstances”); Luna v. Kernan, 784 F.3d 640, 649–50 (9th Cir. 2015) (emphasizing that diligence for purposes of equita- ble tolling depends on the “circumstances” and what “a reason- able person in the petitioner’s shoes” would have understood was appropriate at the time (citation omitted)). Most relevant here, we have explained in that context that “[i]f a petitioner ‘did what he reasonably thought was necessary to preserve his rights . . . based on information he received . . . , then he can hardly be faulted for not acting more “diligently” than he did.’” Munchinski, 694 F.3d at 331 (quoting Holmes v. Spencer, 685 F.3d 51, 65 (1st Cir. 2012)). While we express no opinion on whether Bracey might have succeeded on an equitable toll- ing theory, this case law underscores what we and other courts have said about the statutory tolling provision at issue here.

25 b. Dennis clarified the factual predicate of Brady claims and the reasonable ex- pectations of Brady claimants

Dennis corrected a major misstep in this Circuit’s case law, one that was central to the nature of Brady claims and the relative expectations that the law places on prosecutors and de- fendants with respect to exculpatory information in the govern- ment’s possession. Snapshots of our case law before and after Dennis show why.

Before Dennis, our Brady decisions were “inconsistent and could easily confuse,” with some suggesting that the de- fendant himself had an “obligation to exercise due diligence” in collecting material exculpatory evidence and that his failure to do so would “excuse the government’s non-disclosure of [such] evidence.” Dennis, 834 F.3d at 291–93. In United States v. Starusko, 729 F.2d 256 (3d Cir. 1984), for instance, we stated that the government bore no obligation under Brady “to furnish a defendant with information which [the defendant] already has or, with any reasonable diligence, [could] obtain himself.” Id. at 262. And in Grant v. Lockett, 709 F.3d 224 (3d Cir. 2013), we rejected a petitioner’s Brady claim after de- termining that his counsel could have discovered the withheld evidence through “reasonable diligence.” Id. at 231. Before Dennis, therefore, the government had access to a potent argu- ment in every Brady case: “that because defense counsel could or should have discovered the [Brady evidence] with due dili- gence, the prosecution was not required to disclose it.” Dennis, 834 F.3d at 291; see id. at 291 n.20.

That changed with Dennis. There, we confronted the question whether the government’s duty to disclose could be excused where the evidence in question was assertedly

26 available in public records. See id. at 289–91. We answered that question in the negative, and in strong terms: There is no “affirmative due diligence duty of defense counsel as part of Brady” and “no support [for] the notion that defendants must scavenge for hints of undisclosed Brady material.” Id. at 290 (quoting Banks, 540 U.S. at 695). Rather, “the duty to disclose under Brady is absolute—it does not depend on defense coun- sel’s actions.” Id. Consequently, the defense “is entitled to presume that prosecutors have ‘discharged their official du- ties’” by sharing all material exculpatory information in their possession, id. (quoting Banks, 540 U.S. at 696), and the de- fense’s diligence in seeking out exculpatory material on its own “plays no role in the Brady analysis,” id. at 291.

Our en banc opinion in Dennis thus decisively rejected the line of cases embracing a due diligence obligation and re- turned us to first principles as to both the factual predicate of a Brady claim and the reasonable expectations of a defendant in the Brady context. Whereas we had previously suggested that the factual trigger for a Brady violation was the prosecution’s failure to disclose material exculpatory evidence not otherwise accessible to the defendant, Dennis embraced an “absolute” prosecutorial duty to disclose “not depend[ent] on” the de- fense’s efforts. 834 F.3d at 290. And whereas we had previ- ously suggested that a defendant bore an independent obliga- tion to seek out Brady material, Dennis held that the defense may “rely on the prosecutor’s duty to turn over exculpatory ev- idence,” with any inquiry into the defendant’s ability to dis- cover that evidence being “beside the point.” Id. at 291. Thus, while it had the effect of bringing our case law back in line with clearly established Supreme Court precedent, id., Dennis

27 reflected a significant change in our own jurisprudence, with important consequences for this appeal.14

14 According to the Dissent, Dennis effected no change in our understanding of Brady because it had long been “con- trolling circuit law” that a prosecutor’s Brady obligations in- cluded the production of impeachment evidence otherwise publicly available. Dissent at 5. But the case the Dissent cites for this proposition, Wilson v. Beard, 589 F.3d 651 (3d Cir. 2009), was recognized by the majority in Dennis as the single outlier where “we got it right,” Dennis, 834 F.3d at 292, in con- trast to our many prior cases embracing the “due diligence ex- ception to Brady,” id.; see, e.g., United States v. Starusko, 729 F.2d 256, 262 (3d Cir. 1984); United States v. Perdomo, 929 F.2d 967, 973 (3d Cir. 1991). Those cases “ha[d] made clear that Brady does not compel the government ‘to furnish a de- fendant with information which . . . with any reasonable dili- gence, he can obtain himself.’” United States v. Pelullo, 399 F.3d 197, 213 (3d Cir. 2005) (citation omitted). And where “[our] cases conflict, the earlier is the controlling authority and the latter is ineffective as precedent[].” United States v. Rivera, 365 F.3d 213, 213 (3d Cir. 2004). So it was not Wilson, but “th[at] earlier case law [that] control[led]”—unless and until that case law was reconsidered en banc. Holland v. N.J. Dep’t of Corr., 246 F.3d 267, 278 n.8 (3d Cir. 2001).

That is precisely why our cases post-dating Wilson— without so much as a mention of that case—continued to hold “[a] Brady claim lacked merit” where “trial counsel could . . . have accessed [a prosecution witness’s] criminal history through [public] records.” Grant v. Lockett, 709 F.3d 224, 230–31 (3d Cir. 2013). It’s why the District Court held in 2012

28 c. Dennis wrought a material change in the decisional law with respect to the reasonable expectations of a petitioner in Bracey’s position

Dennis did not involve § 2244(d)(1)(D) directly and did not alter that provision’s requirement that a petitioner exercise “reasonable diligence in the circumstances,” Wilson, 426 F.3d at 660 (quoting Schlueter, 384 F.3d at 74). The question before us is, rather, whether Dennis materially altered what a peti- tioner in Bracey’s “position would [have] reasonably

that Bracey failed to exercise due diligence “because the full extent of the plea agreements and the sentences received by the witnesses were a matter of public record,” JA 75, and why we, in 2013, denied a COA. It’s why the question whether we should continue to recognize a “reasonable diligence” excep- tion to Brady was a major point of contention between the Den- nis majority and dissent. Compare 834 F.3d at 289–93 (hold- ing Grant, Perdomo, and Starusko overruled “[t]o the extent [they] considered defense counsel’s purported obligation to ex- ercise due diligence” an exception to Brady) with id. at 362–64 (observing “[t]he reasonable diligence ‘branch of the Brady doctrine’ [wa]s evident . . . in our own precedents” and was the approach of “our court, [and] ten out of the twelve regional courts of appeals”) (Fisher, J., dissenting). And it’s why an en banc was required to overturn those precedents and bring us in line with Supreme Court precedent by holding that “the con- cept of ‘due diligence’ plays no role in the Brady analysis.” Id. at 291; see 3d Cir. I.O.P. 9.1 (“[N]o subsequent panel overrules the holding in a precedential opinion of a previous panel. Court en banc consideration is required to do so.”).

29 expect[ed],” id. at 661, about the possibility of undisclosed Brady material, thereby shifting the ground on which Bracey’s habeas petition was dismissed. That, it did.

This conclusion flows from both lines of doctrine dis- cussed above. As we have explained, § 2244(d)(1)(D)’s due diligence requirement—like other identically worded require- ments in AEDPA—obligates a petitioner to investigate the fac- tual basis for a potential claim only once he has a “reasonable . . . expectation” that an investigation would produce relevant information. Wilson, 426 F.3d at 661–62. Dennis, in turn, clar- ified the reasonable expectations of Brady claimants by hold- ing that the prosecution’s “duty to disclose under Brady is ab- solute” and that the defense “is entitled to presume that prose- cutors have ‘discharged their official duties’” under Brady. 834 F.3d at 290 (citation omitted). Put differently, § 2244(d)(1)(D) asks whether a “person in [the petitioner’s] position would reasonably expect” that independent investiga- tion would yield evidence of a Brady violation, Wilson, 426 F.3d at 661, and Dennis answers that, absent evidence to the contrary, a petitioner would reasonably expect—and, indeed, “is entitled to presume,” 834 F.3d at 290—the exact opposite: that there is no Brady violation to be discovered.

We find overwhelming support for this conclusion in the decisions of our sister circuits that have adopted the view of Brady we embraced in Dennis. When those circuits have assessed a statutory due diligence requirement in the context of a Brady claim, they have recognized what we do today: A pe- titioner’s failure to search for Brady material of which he is unaware and which he is entitled to presume is non-existent

30 does not fall short of the diligence required by § 2244(d)(1)(D). See Wilson, 426 F.3d at 661–62; Dennis, 834 F.3d at 290.

Take Douglas v. Workman, 560 F.3d 1156 (10th Cir. 2009), which also involved a Brady claim alleged to be un- timely under § 2244(d)(1)(D). See id. at 1181. As in this case, the government in Douglas argued that the petitioner had not “exercise[d] . . . due diligence” with respect to the underpin- nings of his Brady claim because he “could have uncovered” the Brady material had he investigated further. Id. The Tenth Circuit rejected that argument. Previewing what we would say in Dennis, the court reasoned it was “appropriate for [the peti- tioner] to assume that his prosecutors would not stoop to im- proper litigation conduct to advance prospects for gaining a conviction,” which in turn meant that he had not “fail[ed] to exercise due diligence” under § 2244(d)(1)(D). Id. (quoting Banks, 540 U.S. at 694). Accepting the government’s contrary argument, the court explained, would subvert the expectation on which Brady is built, namely that it is “incumbent on the State to set the record straight.” Id. (quoting Banks, 540 U.S. at 676).

Other Courts of Appeals are in agreement. In Jefferson, 730 F.3d 537, which involved the substantively identical due diligence requirement in § 2255(f)(4), the Sixth Circuit held that because the government’s obligation to disclose Brady material “exist[s] regardless of whether the defendant[] . . . ask[s]” or searches for the information,15 the petitioner was

15 In comparable contexts, some courts have cited de- fendants’ “repeated[] request[s]” for Brady materials as suffi- cient evidence of diligence. E.g., Juniper v. Zook, 876 F.3d 551, 564 (4th Cir. 2017). But to be clear, such requests are not

31 entitled to “rely on” government compliance and “assume that [the] prosecutors would not stoop” to violating Brady to secure a conviction. Id. at 545–46 (quoting Banks, 540 U.S. at 694). Without a specific “basis for believing [the prosecution] had failed to comply with Brady,” the court explained, a petitioner need not independently search for Brady material. Id. at 545 (quoting Strickler v. Greene, 527 U.S. 263, 287 (1999)); see also In re Wogenstahl, 902 F.3d at 629 (holding, in a case in- volving § 2244(b)(2)(B)(i), that “[t]he prosecution has a con- stitutional obligation under Brady to provide material exculpa- tory and impeachment evidence, . . . and the defendant is not required to request continuously Brady information in order to show due diligence”).

The Ninth Circuit’s case law falls neatly in step. Facing a Brady claim in the context of § 2244(b)(2)(B)(i), it held that although a petitioner “on notice” of a possible Brady violation must “exercise due diligence in investigating” that possibility, one without any “indication . . . that the alleged exculpatory evidence existed” would have “no reason to investigate” fur- ther and therefore would not have failed to exercise due dili- gence. Solorio, 896 F.3d at 920. Absent facts indicating that the default assumption should be overridden, the petitioner is not “on inquiry notice to investigate further.” Id. at 920–21.

So too with the Fifth Circuit. In Starns, 524 F.3d 612, which like this case involved § 2244(d)(1)(D), the court held that a petitioner “d[oes] not fail to act with due diligence in not

necessary to establish a Brady claim: Under Brady, “the duty to disclose . . . evidence is applicable even though there has been no request by the accused.” Strickler, 527 U.S. at 280 (citing United States v. Agurs, 427 U.S. 97, 107 (1976)).

32 investigating further [into alleged Brady material]” because he has every right to “assum[e] the state could be taken at its word.” Id. at 619. And recently, returning to this issue in the context of § 2244(b)(2)(B)(i), the Fifth Circuit emphasized that where the petitioner has “no reason . . . to suspect that docu- ments were being withheld,” due diligence did not require him to search for such documents.16 In re Will, 970 F.3d 536, 542–

16 Here, too, we find confirmation in closely related doc- trines. In Juniper, for instance, the court faced a Brady claim in the context of a habeas petitioner’s quest for an evidentiary hearing, which implicated the judicial interpretation of § 2254(e)(2) requiring the petitioner to “ha[ve] diligently pur- sued” his claims in state court. 876 F.3d at 563 (citation omit- ted); see Williams, 529 U.S. at 432, 435. In assessing that re- quirement against Brady’s backdrop, Juniper held that the pe- titioner had exercised appropriate diligence simply by “re- quest[ing] that the Commonwealth turn over” the exculpatory evidence. 876 F.3d at 564. But see supra n.15. The court’s conclusion flows from what Dennis recognized: that unless some fact tips him off otherwise, a petitioner may reasonably trust that the government is acting according to its constitu- tional obligations. See 834 F.3d at 290.

Also illuminating are cases in which a habeas petitioner must show “cause” to excuse a procedural default, which typi- cally requires diligence on the part of the petitioner. In Banks, for example, where the state failed to disclose impeachment material, the Supreme Court held that the petitioner “had cause for failing to investigate” because “it was . . . appropriate” to presume that prosecutors would fulfill their disclosure obliga- tions. 540 U.S. at 693–94. The Court therefore rejected the notion “that the prosecution can lie and conceal and the

33 43 & nn.22–24 (5th Cir. 2020) (citing Banks, 540 U.S. at 695, and Strickler, 527 U.S. at 289).

These decisions confirm what our case law dictates. Once Brady is understood to impose an affirmative disclosure obligation on the government, one in which criminal defend- ants are entitled to place their faith, a defendant’s lack of inde- pendent investigation does not equate to a lack of due dili- gence, at least not without facts giving him a reasonable basis to suspect a Brady violation. To the contrary, in the typical case it is “unreasonable to expect the [petitioner]” to harbor suspicions that the government is defying its obligations,

prisoner still has the burden to discover the evidence . . . so long as the potential existence of a prosecutorial misconduct claim might have been detected.” Id. at 696 (internal quotation marks and citation omitted). Likewise, in Strickler, the Court explained that a defendant usually “ha[s] no basis for believing the Commonwealth ha[s] failed to comply” with its Brady ob- ligations and thus cannot be expected to conduct his own in- vestigation based on “[m]ere speculation that some exculpa- tory material may have been withheld.” 527 U.S. at 286–87. That is true even when the defendant “could have obtained” the material from public sources. Id. at 284–85. “In the context of a Brady claim,” in other words, “a defendant cannot conduct [a] ‘reasonable and diligent investigation’ [necessary] . . . to preclude a finding of procedural default when the evidence is in the hands of the State.” Id. at 287–88 (citation omitted). Given the close nexus between Strickler’s reasoning and the issues we confront in this case, it is little surprise that Dennis extensively relied on Strickler in reaching its conclusion. See Dennis, 834 F.3d at 291 (citing Strickler, 527 U.S. at 286–89).

34 Wilson, 426 F.3d at 661, because such an expectation would be “fundamentally at odds with Brady itself,” Jefferson, 730 F.3d at 546.17 We hold, therefore, that a habeas petitioner’s Brady claim is timely under § 2244(d)(1)(D) so long as it is filed within one year of the date on which the petitioner has reason to believe that the prosecution may have violated its duty of disclosure.

Crucially, the defendant’s “reasonabl[e] expect[ation]” that the government will comply with Brady, see Wilson, 426 F.3d at 661, does not evaporate upon conviction or after trial. Rather, unless and until there are reasons to think otherwise, that reasonable expectation continues past trial, into postcon- viction proceedings and beyond. See, e.g., Banks, 540 U.S. at 693; Strickler, 527 U.S. at 286–87. Moreover, we must be at- tentive to each petitioner’s “circumstances,” Wilson, 426 F.3d at 660 (quoting Schlueter, 384 F.3d at 74), and in particular “take into account that prisoners are limited by their physical

17 Likewise, the Pennsylvania Supreme Court recently reached the same conclusion with regard to Pennsylvania’s Post Conviction Relief Act, see 42 Pa. Cons. Stat. § 9545(b)(1)(ii), holding that the due diligence inquiry “calls for a circumstance-dependent analysis of the petitioner’s knowledge, not that of the public at large,” and therefore that courts cannot presume petitioners know facts that “appear in the public record,” Commonwealth v. Small, 238 A.3d 1267, 1283, 1286 (Pa. 2020); see also Commonwealth v. Burton, 158 A.3d 618, 638 (Pa. 2017) (holding that the public records pre- sumption “does not apply to pro se prisoner petitioners” be- cause “prisoners’ access to public records is distinctly compro- mised”).

35 confinement,” Moore, 368 F.3d at 940. See also Ross, 712 F.3d at 802 (explaining that due diligence does not “expect Herculean efforts on the part of a lay person who is a convicted and incarcerated prisoner”). Thus, even after a defendant goes to trial, is convicted, and is sentenced to serve a term of impris- onment, due diligence does not require him to shed his reason- able reliance on the government’s compliance with Brady.

We add two clarifications about the scope of today’s holding. First, we agree with the Fifth Circuit that a due dili- gence requirement like the one in § 2244(d)(1)(D) “cannot be collapsed” into “the merits of [the] Brady [claim]” itself. John- son v. Dretke, 442 F.3d 901, 909 (5th Cir. 2006) (interpreting § 2244(b)(2)(B)(i)). But our holding does nothing of the sort. To be clear, “where the record demonstrates that the defendant or defense counsel was aware of the potential Brady material but failed to pursue investigation of that ultimate claim,” id. at 910, nothing in Dennis or any other decision of this Circuit, including today’s, stands in the way of any of the consequences that AEDPA attaches to a lack of due diligence, see Gage v. Chappell, 793 F.3d 1159, 1166 (9th Cir. 2015). The baseline expectations that Dennis established for Brady claimants in the context of AEDPA’s due diligence requirements hold true only where the petitioner has no reasonable basis in fact to be “aware of the potential Brady material.” Johnson, 442 F.3d at 910.

Second, our holding today is a narrow one. Due dili- gence requirements like the one in § 2244(d)(1)(D) are highly context-specific, and we address them here only in the context of a Brady claim. Cf., e.g., Ross, 712 F.3d at 799–800, 802–04 (applying the due diligence prong of the equitable tolling test to ineffective assistance of counsel claims allegedly barred by

36 § 2244(d)’s limitations period). And as far as Brady claims go, due diligence requirements remain substantial: If there is a rea- sonable basis for a petitioner to believe additional investigation will yield undisclosed Brady material, that petitioner must in- vestigate or else risk the statutory consequences. See, e.g., Solorio, 896 F.3d at 921 (concluding that petitioner was “on inquiry notice” more than a year before he filed his petition and therefore that he did not exercise due diligence under § 2244(b)(2)(B)(i)). And any petitioner who clears one of AEDPA’s due diligence hurdles will still face other meaning- ful procedural bars—not to mention AEDPA’s daunting defer- ential standard on the merits.

As for this appeal, we conclude that Dennis changed the relevant decisional law on which the dismissal of Bracey’s un- derlying habeas petition rested. The District Court dismissed Bracey’s habeas petition as untimely because the allegedly un- disclosed Brady material was a matter of public record, mean- ing Bracey had failed to “exercise . . . due diligence,” 28 U.S.C. § 2244(d)(1)(D). Neither that holding nor our denial of Bracey’s request for a COA was unreasonable under our then- existing circuit law. But that basis for dismissal became not just “debatable,” Slack, 529 U.S. at 484, but untenable after Dennis. At that point, it should have been apparent that the District Court had misapprehended the “factual predicate,” 28 U.S.C. § 2244(d)(1)(D), of Bracey’s Brady claim—namely, that the prosecution had withheld Brady material, not that the prosecution had withheld Brady material the defense could not otherwise access. It should have been equally apparent that the Court had misapprehended what a “person in [Bracey’s] posi- tion would [have] reasonably expect[ed],” Wilson, 426 F.3d at 661, about the possibility of wrongfully withheld infor- mation—namely, that without any “reason for believing [the

37 prosecution] had failed to comply,” Bracey “was entitled to rely on the prosecutor[’s] fulfilling [his] Brady obligation.” Dennis, 834 F.3d at 291. Together, these changes brought on by Dennis had a material impact on the rationale for the District Court’s original dismissal of Bracey’s habeas petition: the un- timeliness of that petition under § 2244(d)(1)(D) for lack of due diligence.

2. Whether the District Court properly ap- plied the totality-of-the-circumstances analysis required by our precedent

Having decided that there was a material change in de- cisional law, we next address whether the District Court properly concluded that Bracey’s motion, which was based on that intervening precedent, failed to meet the high threshold for 60(b) relief. “[W]e have not foreclosed the possibility that a change in controlling precedent, even standing alone, might give reason for 60(b)(6) relief,” but we have also observed that “intervening changes in the law rarely justify [such] relief” without more. Cox, 757 F.3d at 121, 124. The ultimate ques- tion is whether the petitioner has shown “extraordinary circum- stances where, without such relief, an extreme and unexpected hardship would occur.” Id. at 115.18

18 The Dissent posits that where a 60(b)(6) motion is predicated on a change in decisional law, the only qualifying decisions are those that would “apply retroactively.” Dissent at 5–6. Though the Dissent doesn’t say as much, the effect of such a rule would be to require the change in decisional law to fit one of Teague v. Lane’s “two exceptions to [the] rule of nonretroactivity for cases on collateral review.” 489 U.S. 288, 307, 310 (1989). This novel proposition—starkly absent from

38 the briefing and devoid of any support in Supreme Court or our Court’s precedent—confuses the test for reopening a final judgment so a court can consider a petition at all, with the test for assessing the merits of that petition once under considera- tion. These are distinct inquiries, and we deal today only with the first.

For “a change in law []—when accompanied by appro- priate equitable circumstances—[to] support Rule 60(b)(6) re- lief,” Satterfield v. Dist. Att’y Phila., 872 F.3d 152, 161 (3d Cir. 2017), the change need only be “relevant” to the court’s basis for “finding that [defendant’s] petition was untimely un- der AEDPA,” id. at 159; see also Norris, 794 F.3d at 405 (ob- serving that a change in decisional law that is “relevant to the movant’s position” can be sufficient in the presence of other Cox factors to “support a Rule 60(b) motion”). Thus, in Cox itself, we explained that, although the intervening precedent in that case “did not announce a new constitutional rule or right for criminal defendants, but rather an equitable rule prescribing and expanding the opportunity for review of their Sixth Amendment claims,” 757 F.3d at 124, that change in decisional law was sufficient to require remand for “explicit consideration of the [other Cox factors],” id. at 126—particularly where, as here, no court had ever considered the merits of the underlying claim, id. at 124–25.

Teague, on the other hand, relates to the retroactive ap- plication of substantive or procedural law to the validity of the petitioner’s underlying conviction, limiting retroactivity to cases in which the substantive conduct was “beyond the power of the criminal law-making authority to proscribe” or which announce a “watershed rule[] of criminal procedure.” 489 U.S.

39 In Cox, we set out a number of factors the district court should consider in answering that question. Id. at 122–26. One is the effect of the change in decisional law on the district court’s prior ruling, which carries particular weight where, as here, that change concerns a “constitutional rule or right for criminal defendants.” Id. at 124. Another key factor is “the merits of a petitioner’s underlying . . . claim.” Id. As we have explained, the district court “need not provide a remedy under 60(b)(6) for claims of dubious merit,” id. at 125, but because relief may be warranted if “there is merit in the defense or claim,” Lasky v. Cont’l Prods. Corp., 804 F.2d 250, 256 n.10 (3d Cir. 1986), it is especially important for the court to “assess the merits” in cases where, as here, “the merits of the [under- lying] claim were never considered prior to judgment,” Cox, 757 F.3d at 124 (citing Lasky, 804 F.2d at 256 n.10). Other factors include “[p]rinciples of finality and comity,” id., at 125; the “movant’s diligence in pursuing review,” id. at 126; and, where appropriate, “the imperative of correcting a

at 311. That retroactivity inquiry has no bearing whatsoever on the materiality of a jurisprudential change under 60(b)(6) and would portend a sea change in that rule if it did. Indeed, given that the Supreme Court has never recognized “a new rule of criminal procedure capable of meeting” Teague’s standard, Ramos v. Louisiana, 140 S. Ct. 1390, 1407 (2020), has denied retroactivity for so fundamental a rule as Batson v. Kentucky, 476 U.S. 79 (1986), see Teague, 489 U.S. at 296, and has cau- tioned that it is “unlikely that any [watershed] rules ha[ve] yet to emerge,” Whorton v. Bockting, 549 U.S. 406, 417 (2007) (second alteration in original), the Dissent’s approach would essentially write 60(b)(6) out of existence where changes in decisional law are concerned.

40 fundamentally unjust incarceration,” Satterfield v. Dist. Att’y Phila., 872 F.3d 152, 162 (3d Cir. 2017) (quoting Murray v. Carrier, 477 U.S. 478, 495 (1986)). The analysis required by Cox is thus both “flexible” and “multifactor.” 757 F.3d at 122.

Here, however, the District Court failed to conduct any analysis at all. Instead, in response to Bracey’s 60(b) motion, it issued a one-page order that summarily rejected relief with- out citing Cox, addressing any of the Cox factors, or even ac- knowledging the material change in law discussed in the mo- tion. Thus, at the very least, “we cannot determine from what it wrote whether the [District] Court considered [those] fac- tors.” Id. at 120. Under Cox, that was a per se abuse of discre- tion. Id. at 124.

3. The proper disposition on appeal

As a general matter, where a district court fails to “em- ploy the full, case-specific analysis we require when faced with a 60(b)(6) motion,” id. at 120, “[t]he approach in Cox calls for a remand to the District Court to determine if [extraordinary circumstances] exist[],” United States v. Doe, 810 F.3d 132, 152 (3d Cir. 2015), because “[t]he grant or denial of a Rule 60(b)(6) motion is an equitable matter left, in the first instance, to the discretion of a district court,” Cox, 757 F.3d at 124. On the other hand, we can affirm on any basis supported by the record, TD Bank N.A. v. Hill, 928 F.3d 259, 270 (3d Cir. 2019), so a remand may not be necessary if we can determine from the record on appeal that a petitioner cannot show extraordi- nary circumstances as a matter of law.

Here, because the District Court not only failed to “con- sider the full measure of [the] properly presented facts and cir- cumstances attendant to [Bracey’s] request,” Cox, 757 F.3d at

41 122, but also precluded development of the record concerning those facts and circumstances, remand is required. In his pro se request for 60(b) relief, Bracey not only cited to Cox and its “critical factors in [the] rule 60(b) analysis,” JA 126; he also discussed the effect of our intervening decision in Dennis on the District Court’s dismissal of his petition; the constitutional nature of that change in law establishing that “the prosecutor has a duty to seek justice by providing [Giglio] material,” JA 125; the factual support for and the merits of his Brady claim; his state and federal habeas petitions, which recounted those facts in even more detail; the procedural history of his state and federal petitions, which reflected that no court had ever re- quired the State to answer his habeas petition, much less ad- dressed the merits; and “the need for an evidentiary hearing” to develop his habeas claim, JA 127. As a pro se pleading, moreover, Bracey’s motion, while cogent in its own right, must also be liberally construed. See Erickson v. Pardus, 551 U.S. 89, 94 (2007).

Yet the District Court’s order denying Bracey’s motion did not even mention Dennis, much less discuss its significance for Bracey’s Brady claim or whether, as a claim implicating a basic “constitutional . . . right for criminal defendants,” Cox, 757 F.3d at 124, that “change in controlling precedent, even standing alone, might give reason for 60(b)(6) relief,” id. at 121. Nor did—or even could—the District Court “assess the merits” of Bracey’s underlying Brady claim, id. at 124, be- cause it denied his request for an evidentiary hearing on his 60(b) motion, stating only—and erroneously—that the motion “raise[d] the same arguments” the Court had previously re- jected. JA 7.

42 In sum, this is not a situation where the petitioner failed to “properly present[]” any “facts or circumstances,” id. at 122, that could possibly justify 60(b)(6) relief. Nor is it a situation where the district court misapplied the Cox factors, but the rec- ord, in any event, forecloses a finding of extraordinary circum- stances as a matter of law.19 What we confront instead is a

19 In the Dissent’s view Bracey did not “properly pre- sent[]” a claim for 60(b)(6) relief because the “facts and cir- cumstances” he raised in his motion “are not extraordinary” and “do not justify re-opening” his case. Dissent at 6 (empha- sis omitted). That the Dissent, instead of remanding, would take upon itself to decide the merits of this 60(b)(6) motion— where we have no evidentiary record before us, no reasoned opinion to review, and no briefing on appeal concerning any Cox factor other than change in decisional law—is puzzling to say the least. See Cox, 757 F.3d at 124 (explaining that where a district court “base[s] its decision solely on the reasoning” that a change in decisional law is per se inadequate for 60(b)(6) relief “and fail[s] to consider how, if at all, the . . . other fac- tor[s] highlighted by the parties would figure into its 60(b)(6) analysis,” the appellate court “will remand to give it the oppor- tunity to conduct that equitable evaluation”); Satterfield, 872 F.3d at 162 (requiring the district court, in the first instance, to “evaluate the nature of the change [in decisional law] along with all of the equitable circumstances and clearly articulate the reasoning underlying its ultimate determination”); id. (ob- serving that where, “[a]s best we can tell, [the District Court] incorrectly focused on whether [the change in decisional law], in isolation, was sufficient to serve as an extraordinary circum- stance,” the appellate court “will vacate the order of the District Court . . . and remand to it to carry out another analysis”).

43 district court that abused its discretion by failing to apply the Cox factors at all, and—to the extent additional record devel- opment would have been helpful to assess those factors—by denying the petitioner’s request for an evidentiary hearing. In that situation, we must remand not only because “[t]he grant or denial of a Rule 60(b)(6) motion is an equitable matter left, in the first instance, to the discretion of a district court,” id. at 124, but also because the record, as such, does not allow for mean- ingful appellate review.

We will therefore remand for the District Court to “take the first pass at weighing the equitable factors,” Satterfield, 872

In any event, Bracey did “properly present” a claim, suf- ficient to merit the District Court’s consideration under Cox, because he raised with the District Court a material change in decisional law and other equitable factors, including the im- portance of the right at issue, the merits of his claim, and the need for an evidentiary hearing. See supra at 41–42. And even if he had done no more than argue Dennis’s material change in the law, the District Court was still obligated to consider the other Cox factors because we have squarely rejected “any cat- egorical rule that a change in decisional law [alone] is never an adequate basis for Rule 60(b)(6) relief.” Cox, 757 F.3d at 121; see also Satterfield, 872 F.3d at 162 (“Cox . . . requires a dis- trict court to consider the full panoply of equitable circum- stances before reaching its decision.”). Thus, the District Court abused its discretion under our well-established prece- dent by failing to engage in the necessary “case-dependent analysis, fully in line with Rule 60(b)(6)’s equitable moor- ings.” Cox, 757 F.3d at 124. And whether we agree with that precedent or not, it is the controlling law of our Circuit.

44 F.3d at 162, with additional briefing and factfinding as neces- sary, Cox, 757 F.3d at 126, including any “factors the parties care to brief that we have not just discussed,” Doe, 810 F.3d at 153.

III. CONCLUSION

For the foregoing reasons, we will vacate and remand to the District Court for an appropriate consideration of Bracey’s Rule 60(b) motion.

45 Bracey v. Superintendent Rockview SCI, No. 17-1064

PHIPPS, Circuit Judge, concurring in part and dissenting in part.

I agree with the two preliminary holdings in today’s opinion, viz., that a habeas petitioner must receive a certificate of appealability to appeal an adverse ruling on a Rule 60 motion and that such a certificate should issue in this case. But I disagree with the Majority’s decision to vacate the District Court’s order denying William Bracey’s Rule 60(b)(6) motion for two reasons. First, I do not believe that Dennis v. Secretary, 834 F.3d 263 (3d Cir. 2016) (en banc), constitutes a change in decisional law relevant to Bracey’s habeas petition. Second, even if it does, I disagree that the District Court abused its discretion in denying Bracey’s Rule 60(b)(6) motion. See Giordano v. McCartney, 385 F.2d 154, 155 (3d Cir. 1967) (“A motion for relief under Rule 60(b) is directed to the sound discretion of the trial court and its exercise of that discretion will not be disturbed unless there was a clear abuse.” (citations omitted)).

In May 1995, Bracey was tried and convicted of first degree murder. Fifteen years after that trial, on October 13, 2010, Bracey learned from public court filings that two witnesses who testified against him had additional charges pending against them at the time of his trial. The court records that Bracey discovered also revealed that after the trial, those two witnesses received favorable treatment on those additional charges due to their cooperation with the prosecution at Bracey’s trial.

1 In 2011, Bracey petitioned for a writ of habeas corpus based on his discovery of that information – which the prosecution never disclosed to him. 1 His petition challenged his conviction on the ground that he was unable to impeach those two witnesses at trial with that newly discovered information.2

In August 2012, the District Court denied Bracey’s habeas petition as untimely. Bracey advocated for an accrual date of October 13, 2010 – the date he discovered the additional impeachment information. The District Court rejected Bracey’s argument because that later accrual date depends, not on the date of Bracey’s actual discovery of the facts, but on “the date on which the factual predicate of the claim or claims presented could have been discovered through the exercise of

1 The two witnesses entered plea agreements on the additional charges, but the transcripts do not specify when they entered those agreements. Unless those plea agreements were in place when those witnesses testified against Bracey, the prosecution would not have been obligated to disclose them: post-trial favorable treatment of a witness is not within the scope of Brady disclosures. See Bell v. Bell, 512 F.3d 223, 234 (6th Cir. 2008) (en banc) (rejecting “a new definition of Brady material that includes possible post-trial witness favorable treatment— something never previously considered by any court to be within Brady’s ambit”). 2 Even without that information at trial, Bracey’s counsel still impeached those witnesses on several grounds – including their motives to cooperate with the prosecution in light of other charges pending against them.

2 due diligence.” 28 U.S.C. § 2244(d)(1)(D).3 And in reviewing Bracey’s diligence, the District Court determined that the impeachment information, which was available from public court records, “could have been discovered through the exercise of due diligence well before October of 2010.” Bracey v. Lamas, No. 11-cv-02329, Mem. Op. at 7 (M.D. Pa. Aug. 8, 2012) (JA73a). With that finding, that Bracey did not exercise the diligence needed for a later accrual date, see 28 U.S.C. § 2244(d)(1)(D), the District Court concluded that Bracey’s habeas petition was untimely.

Bracey now seeks to re-open that final judgment denying his habeas petition through a Rule 60(b)(6) motion. He relies on this Court’s 2016 decision in Dennis, which held that Brady obligates the prosecution to disclose publicly available exculpatory information in its actual or constructive possession. 834 F.3d at 291; see generally Brady v. Maryland, 373 U.S. 83 (1963). And according to Bracey, that holding means that reasonable diligence under the habeas statute, see 28 U.S.C. § 2244(d)(1)(D), does not require a habeas petitioner to search public records for charges that were both known to the prosecution and pending against prosecution witnesses. Without an obligation to look for that information, Bracey argues that – due to Dennis – his habeas petition was timely and should be re-opened.

3 See also id. § 2244(d)(1)(A) (setting a default accrual date for a habeas petition as the date on which the judgment in the state- court criminal proceeding became final); Schlueter v. Varner, 384 F.3d 69, 74 (3d Cir. 2004) (“Section 2244(d)(1)(D) provides a petitioner with a later accrual date than section 2244(d)(1)(A) only ‘if vital facts could not have been known.’” (quoting Owens v. Boyd, 235 F.3d 356, 359 (7th Cir. 2000))).

3 The Majority embraces that approach, which extinguishes a habeas petitioner’s diligence obligations with respect to suppressed Brady information. I initially part ways with the Majority for a more basic reason: Bracey’s argument was available to him before Dennis. The relevant point for which Bracey cites Dennis – the prosecution’s obligation under Brady to produce publicly available criminal history information within its actual or constructive possession – had been previously articulated in this Circuit. For that reason, the change in law brought about by Dennis is not relevant to Bracey’s habeas petition.

To explain, at the time of Bracey’s 2011 habeas petition, this Circuit had already made clear that a prosecutor’s disclosure obligations included exculpatory information from publicly available court records within the prosecutor’s actual or constructive possession. See Wilson v. Beard, 589 F.3d 651, 664 (3d Cir. 2009) (“‘[T]he fact that a criminal record is a public document cannot absolve the prosecutor of her responsibility to provide that record to defense counsel.’”); Hollman v. Wilson, 158 F.3d 177, 180 (3d Cir. 1998) (“[E]vidence of a government witness’s prior criminal history is evidence which must be produced to the defense.”). While a criminal defendant’s diligence obligations under Brady may have been uncertain in other respects before Dennis, there was no doubt at the time of Bracey’s initial habeas petition in 2011 that a prosecutor had to produce exculpatory information related to a prosecution witness’s criminal history:

We have clearly held that the prosecution bears the burden of disclosing to the defense a prosecution witness’s criminal record, whether

4 or not an explicit request has been made by defense counsel.

Wilson, 589 F.3d at 663 (emphasis added). Dennis broadened that rule, but the key legal principle that Bracey relies on – a prosecutor’s Brady obligations to produce criminal record impeachment information, even if otherwise publicly available – was controlling circuit law at the time of Bracey’s habeas petition.

Thus, Dennis did not change circuit law relevant to Bracey’s diligence obligations under the habeas statute. And without doing so, Dennis does not provide a basis to re-open the judgment denying Bracey’s habeas petition on timeliness grounds.

But even if Dennis were a material change in circuit law, the District Court did not abuse its discretion in denying Bracey’s Rule 60(b)(6) motion. Bracey did not meet his burden of proving the extraordinary circumstances needed to apply Dennis retroactively and re-open the District Court’s final judgment.

New decisional rules do not automatically apply retroactively to closed civil cases. See Harper v. Va. Dep’t of Taxation, 509 U.S. 86, 97 (1993) (“When this Court applies a rule of federal law to the parties before it, that rule is the controlling interpretation of federal law and must be given full retroactive effect in all cases still open on direct review . . . .”). And Bracey’s habeas case was closed when Dennis was decided. To apply Dennis retroactively to his closed case, Bracey invoked Rule 60(b)(6), which operates as a catch-all to re-open a final judgment for “any other reason that justifies

5 relief.” Fed. R. Civ. P. 60(b)(6). In this Circuit, Rule 60(b)(6) allows for the possibility that a new decisional rule can apply retroactively to a prior final judgment, as if that new rule were controlling at the time of the judgment. But that possibility actualizes only in “extraordinary circumstances,” Cox v. Horn, 757 F.3d 113, 115 (3d Cir. 2014), when the change in law is “accompanied by appropriate equitable circumstances.” Satterfield v. Dist. Att’y Phila., 872 F.3d 152, 161 (3d Cir. 2017). Those extraordinary circumstances needed to reopen a final judgment under Rule 60(b) “rarely occur in the habeas context,” and “‘[t]his very strict interpretation of Rule 60(b) is essential if the finality of judgments is to be preserved.’” Gonzalez v. Crosby, 545 U.S. 524, 535 (2005) (parenthetically quoting Liljeberg v. Health Servs. Acquisition Corp., 486 U.S. 847, 873 (1988) (Rehnquist, C.J., dissenting)).

Here, Bracey has not established such extraordinary circumstances – as was his burden to do. See Cox, 757 F.3d at 122. In his Rule 60(b)(6) motion, Bracey emphasized that he did not know of the impeachment information until 2010. He also argued that Dennis eliminated his diligence obligations due to the prosecutor’s duty to disclose impeachment materials. And he advanced one equitable consideration for re- opening his judgment – that he filed his Rule 60(b)(6) motion within 60 days of the Dennis decision. Based on those “properly presented facts and circumstances attendant to [his] request,” Cox, 757 F.3d at 44 (emphasis added), Bracey did not meet his burden of proof. Those identified facts and circumstances are not extraordinary, and they do not justify re- opening of a final judgment.

6 Thus, even treating Dennis as a change in relevant decisional law, the District Court did not abuse its discretion by denying Bracey’s Rule 60(b)(6) motion.