Joel WILLIAMS, Petitioner-Appellant, v. Blaine C. LAFLER, Warden, Respondent-Appellee, 494 F. App'x 526

PER CURIAM. Petitioner Joel Williams seeks a writ of habeas corpus or, at a minimum, an evi-dentiary hearing following his conviction of first-degree premeditated murder and possession of a firearm during the commission of a felony. For the reasons that follow, we deny Williams the writ and affirm the district court’s denial of an evidentiary hearing. BACKGROUND This case arises out of the shooting death of Carl Russell Cotton in the parking lot of a party store in Highland Park, Michigan.

Following the shooting, Williams was charged with the first-degree premeditated murder of Cotton and possession of a firearm during the commission of a felony. Prior to trial, defense counsel’s investigators located and interviewed two witnesses to the shooting: Terrell Higginbotham and Tameka Ballard. During the interviews, both witnesses placed Williams at the scene of the crime and indicated that the killing may have been the result of a fight between Williams and the decedent rather than premeditation. Neither Higginbotham nor Ballard, however, were called as witnesses in the trial.

The prosecution presented two witnesses, Jack Albright and Wayne Burton. Though their testimony differed in some respects, both testified that they observed Cotton and Williams arguing just prior to the shooting, and that Williams shot Cotton. At the close of evidence, the jury found Williams guilty as charged. The trial court sentenced Williams to life in prison, plus a mandatory two years’ incarceration for the firearm conviction. Williams appealed his conviction to the Michigan Court of Appeals. He raised a number of claims on his direct appeal, but did not raise ineffective assistance of counsel.

The state appellate court affirmed his conviction in an unpublished opinion. Williams applied for leave to appeal to the Michigan Supreme Court, but was denied. Williams then returned to the trial court and filed a motion for relief from judgment where, for the first time, he raised ineffective assistance of counsel claims pursuant to Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).

Specifically, he alleged that his constitutional rights were violated from ineffective assistance of trial counsel because trial counsel failed to investigate, produce and subpoena witnesses, and to present an adequate defense. He also claimed that his appellate counsel was ineffective for failing to argue trial counsel’s ineffectiveness on direct appeal. The trial court denied his motion. Having exhausted his state court remedies, Williams filed a petition for habeas corpus relief in the district court, as well as a motion for an evidentiary hearing.

He raised a number of issues, including ineffective assistance of appellate counsel based on a failure to argue ineffective assistance of trial counsel. The district court denied Williams’s motion for an evi-dentiary hearing without prejudice and issued an opinion and order denying him a writ of habeas corpus. This appeal followed. There are two issues before the court: (

i) whether Williams should be granted habeas relief based on ineffective assistance of appellate counsel; and (ii) whether the district court abused its discretion in denying Williams’s motion for an evi-dentiary hearing. We address each issue in turn. DISCUSSION I. Habeas relief due to ineffective assistance of appellate counsel Williams asserts that he received ineffective assistance of appellate counsel because appellate counsel did not bring an ineffective assistance of trial counsel claim on direct appeal. We disagree.

The Anti-Terrorism and Effective Death Penalty Act (“AEDPA”) typically governs federal habeas claims. Under AEDPA, this court may grant Williams relief only if the state court’s adjudication on the merits resulted in a decision that was either: (1) “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States” or (2) “based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” 28 U.S.C. § 2254(d)(l)-(2).

However, where the state court did not adjudicate one prong of the Strickland claim on the merits, we review that prong de novo. See Rompilla v. Beard, 545 U.S. 374, 390, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005) (“Because the state courts found the representation adequate, they never reached the issue of prejudice ... and so we examine this element of the Strickland claim de novo.”). In this case, the state court did not reach Strickland’s prejudice inquiry. As such, any review of the prejudice prong is de novo. Williams argues that AEDPA does not provide the standard of review.

We thus pause briefly to address Williams’s standard-of-review arguments. He asserts that the state court never adjudicated his appellate counsel claim on the merits, and that we must therefore review his claims de novo in accord with Cone v. Bell, 556 U.S. 449, 129 S.Ct. 1769, 173 L.Ed.2d 701 (2009). Williams’s argument hinges on a feature of Michigan law known as a Gin-ther hearing, which allows the defendant to develop a testimonial record in support of his claim on direct review. See People v. Snider, 239 Mich.App. 393, 608 N.W.2d 502, 517 (2000); see also People v.

Ginther, 390 Mich. 436, 212 N.W.2d 922, 925 (1973). This mechanism is available when the defendant moves for a new trial or evidentia-ry hearing in the state courts, Snider, 608 N.W.2d at 517, but Williams made no such motions, instead bringing his ineffective assistance of counsel claim in his motion for post-conviction relief from judgment. As Williams’s argument goes, his appellate counsel should have raised the trial counsel issue in a manner that would have afforded Williams the right to supplement the record with a Ginther hearing.

According to Williams, appellate counsel’s failure means that the state court hearing Williams’s motion for post-conviction relief never truly addressed the claim on the merits. Williams draws that proposition from this court’s decision in Brown v.

Smith, in which we held that a state court adjudication is not on the merits, and de novo review thus applies, where the facts that formed “the basis of the claim were not in the record before the [state court], and that court explicitly acknowledged that its review was limited to mistakes apparent on the record.” 551 F.3d 424, 428 (6th Cir.2008) (internal quotation marks and citation omitted.) The Supreme Court’s recent opinion in Cullen v.

Pinholster, — U.S. -, 131 S.Ct. 1388, 1398-1400, 179 L.Ed.2d 557 (2011), which limits the circumstances under which a federal habeas court can consider evidence that was not before the state court, raises some doubt as to whether Brown remains good law. See Welsh v. Lafler, 444 Fed.Appx. 844, 849 (6th Cir.2011) (unpublished opinion). However, we need not determine Pinholster’s effect on Brown in order to resolve this appeal.

Even assuming that the state court did not address Williams’s ineffective assistance of appellate counsel claim on the merits and thus that de novo review applied, we would deny relief for the reasons described below. Analysis Williams alleges that his appellate counsel was ineffective because he did not adequately investigate trial counsel’s performance, which he claims was itself ineffective.

We conclude, however, that Williams’s ineffective assistance of trial counsel claim is without merit, and he therefore has no cause for complaint regarding appellate counsel’s decision not to pursue that argument on appeal. In order to establish ineffective assistance of appellate counsel, Williams must show “that counsel’s performance was deficient ... [and] that the deficient performance prejudiced the defense.” Strickland, 466 U.S. at 687, 104 S.Ct. 2052.

In determining whether counsel’s performance was deficient, we consider “whether, in light of all the circumstances, the identified acts or omissions were outside the wide range of professionally competent assistance,” id. at 690, 104 S.Ct. 2052, or, phrased differently, whether counsel’s performance “fell below an objective standard of reasonableness,” id. at 688, 104 S.Ct. 2052. “Counsel is strongly presumed to have rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment.” Id. at 690, 104 S.Ct. 2052.

Judicial scrutiny of counsel’s performance is therefore “highly deferential.” Id. at 689, 104 S.Ct. 2052. Additionally, tactical choices regarding which issues to pursue on appeal are “properly left to the sound professional judgment of counsel.” United States v. Perry, 908 F.2d 56, 59 (6th Cir.1990). Finally, in order to properly assess Williams’s appellate counsel claim, we must first consider whether trial counsel was ineffective. See, e.g., Howard v.

Bouchard, 405 F.3d 459, 478-79 (6th Cir.2005) (assessing trial counsel’s performance before assessing appellate counsel’s failure to raise trial counsel claim). 1. Trial Counsel A. Strickland’s “Objectively Unreasonable” Prong Williams alleges that his trial counsel’s representation was objectively unreasonable because he failed to exert a reasonable effort to find Higginbotham and Ballard, despite the fact that Michigan law required the prosecuting attorney or investigative law enforcement agency to provide the defendant, upon request, reasonable assistance to locate and serve process on a witness.

See Mich.Code of Crim. Proc. § 767.40a(5). Trial counsel, he argues, thus failed to use a significant resource at his disposal. Williams further alleges that this was not a strategic decision because “[a] purportedly strategic decision is not objectively reasonable when the attorney has failed to investigate his options and make a reasonable choice between them.” Towns v. Smith, 395 F.3d 251, 258 (6th Cir.2005) (internal quotation marks and citation omitted).

Williams alleges that this conduct was “particularly unreasonable” given that, as he describes it, the two witnesses would have provided Williams with a substantial argument that the prosecutor had not proved premeditation beyond a reasonable doubt. In his pre-trial interview, Higginbotham indicated that Cotton threatened to kill Williams, that Higginbotham had to hold Cotton back from Williams, that Williams appeared scared of Cotton, and that Cotton pushed Williams to the ground just before the shooting.

In her pre-trial interview, Ballard indicated that Cotton was pushing to get at Williams and acting “wild and out of control,” that Cotton threatened to kill Williams, and that Williams appeared scared of Cotton. Williams argues that, because lack of premeditation was the defense’s primary strategy, it was objectively unreasonable not to locate (and then call) these witnesses. As proof that lack of premeditation was, indeed, the primary strategy, Williams asserts that trial counsel: (

i) previewed two defenses in his opening statement, mistaken identity and lack of premeditation; (ii) referred to a lack of premeditation as “the main point” at the trial’s opening; and (iii) moved for a directed verdict on the premeditation issue at the close of the prosecution’s case. Williams’s characterization of trial counsel’s dominant strategy is overzealous. Though trial counsel previewed a premeditation defense, he also previewed, as the State points out, a misidentification defense, i.e., that Williams was not the shooter. Williams describes the misidentification defense as a “last resort.” We disagree: Trial counsel (

i) argued that the state had charged an innocent person; (ii) attacked the credibility of witnesses who identified Williams as the shooter; (iii) argued that the decedent made a dying declaration identifying someone else as the shooter; (iv) argued that there was no significant physical evidence connecting Williams to the crime scene; (

v) argued that physical evidence such as blood splatter and gunshot residue was inconsistent with Williams being the shooter; (vi) asserted near the end of his closing argument that there were “serious questions of identification” in the ease; and (vii) never once acknowledged that Williams was at the scene of the crime or had shot the victim. It appears to us that trial counsel was attempting to inject reasonable doubt into the proceedings based on either misidenti-fication or a lack of premeditation.

It is perfectly reasonable for trial counsel to select witnesses and evidence that preserve each possible avenue of defense, and to reject what would seriously undermine any single strategy. See, e.g., Collier v. Lafler, 419 Fed.Appx. 555, 560 (6th Cir.2011) (trial counsel “was not deficient for failing to present evidence that ... would have been detrimental to his defense” and the decision constituted “reasonable trial strategy”).

Here, where it appears that trial counsel was attempting to preserve the possibility of both misidentification and a lack of premeditation, it also seems clear that calling witnesses who put Williams at the crime scene would have been detrimental to his misidentification defense. His failure to do so therefore is not objectively unreasonable. The same can be said for trial counsel’s alleged failure to locate Higginbotham and Ballard.

Williams asserts that trial counsel was remiss in his efforts to find the witnesses because, at one point, he said that his investigator had been searching for them, and at another time, he stated that he would not be calling Higginbotham and Ballard because “we’re unable to.” Lafler, by contrast, asserts that “there is no indication” that trial counsel did not call the witnesses because he could not find them; to the contrary, he interviewed them, transcribed the interviews, and then simply chose a different strategy at trial. Neither party’s argument is unassailable.

Trial counsel stated that he had “searched and searched for ... weeks,” “high and low, day and night,” and could not locate Higginbotham because Higginbotham had, apparently, gone into hiding in order to avoid child support payments. Trial counsel further stated that he did not believe he would be able to locate Ballard even with a continuance from the court. It appears that trial counsel was, indeed, diligent in his efforts to locate Ballard and Higginbotham, and succeeded in locating and interviewing them as part of trial preparation.

It is possible that, had he been able to procure the witnesses for trial, he would have called them. But it is also possible that, due to trial strategy, he would not have called them, and we decline to speculate what trial counsel might have attempted to demonstrate through examination. In light of all this, we cannot say that trial counsel’s “failure” to procure Higginbotham and Ballard overcomes the strong presumption that he “made all significant decisions in the exercise of reasonable professional judgment.” Strickland, 466 U.S. at 690, 104 S.Ct. 2052.

Nor can we say that the state court’s decision on this matter entitles Williams to habeas relief.

To be sure, the state court addressed this issue somewhat summarily, stating only that it “[did] not find any professional errors committed by defense counsel” because “counsel’s decision regarding who to present before a jury ... is purely a strategic choice.” The court was thus unwilling to “second-guess counsel’s decisions.” Despite its succinct approach, we decline to disturb its decision: we have not found, and Williams has not offered, case law that would render the state court’s finding untenable, and we believe the record supports the state court’s conclusion. See, e.g., Roush v.

Burt, BIB Fed.Appx. 754, 761 (6th Cir.2008) (finding that there was no reason to deem counsel ineffective for failing to call a witness to the stand where the record indicated that counsel had investigated what that person would provide as a witness); cf. Hawkins v.

Coyle, 547 F.3d 540, 546-547 (6th Cir.2008) (instructing that where the state court disposes of a federal constitutional claim with “little-to-no articulated analysis,” this court conducts a “careful and independent review of the record and applicable law, but [we] cannot reverse unless the state court’s decision is contrary to or an unreasonable application of federal law”) (internal quotations marks omitted). B.

Strickland’s Prejudice Prong Because we find that trial counsel’s performance falls within the “wide range of professionally competent assistance,” Strickland, 466 U.S. at 690, 104 S.Ct. 2052, we need not address Strickland’s prejudice prong. However, we note for the sake of thoroughness that, even if Williams showed trial counsel’s performance to be objectively unreasonable, he could not show that it was prejudicial.

Counsel’s objectively unreasonable performance prejudices the defendant where “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Strickland, 466 U.S. at 694, 104 S.Ct. 2052. “A reasonable probability is a probability sufficient to undermine confidence in the outcome.” Id. Williams argues that trial counsel’s unreasonable performance prejudiced Williams because: (

i) Higginbotham’s and Ballard’s testimony contradicted certain elements of the prosecution’s witnesses’ testimony concerning premeditation; and (ii) the jury asked about the definition of premeditation during deliberations, suggesting it may have had reasonable doubt had it heard testimony suggesting that the killing was not premeditated. Williams cites no cases to support his argument. As we have observed, trial counsel attempted to inject reasonable doubt into the proceedings based on either misidentification or a lack of premeditation.

It is possible that, hearing testimony indicating a lack of premeditation, the jury would have discounted the prosecution’s evidence tending to show premeditation, and would have, for whatever reason, found Williams innocent. But we see at least three reasons why it is a stretch to call this possibility “a reasonable probability.” Strickland, 466 U.S. at 694, 104 S.Ct. 2052. First, nothing suggests that the jury would have deemed Higginbotham and Ballard more credible than the prosecution’s witnesses.

Second, their testimony would (presumably) have shown simply that Williams and the decedent were fighting — this does not negate premeditation, particularly when the record indicates that Williams was armed prior to the altercation. Third, the testimony would have seriously undermined one of the defense’s two arguments — i.e., that the state had charged the wrong man — by placing an armed Williams at the crime scene and in a heated fight with the victim. In light of the foregoing, the testimony might not only have had no effect on the verdict, but might have done Williams more harm than good.

Its absence does not undermine our confidence in the trial’s outcome. 2. Appellate Counsel Williams claims that his appellate counsel was ineffective because he did not adequately investigate trial counsel’s performance, which he claims was itself ineffective. If trial counsel performed adequately, however, this court’s inquiry is “at an end; by definition, appellate counsel cannot be ineffective for a failure to raise an issue that lacks merit.” Greer v. Mitchell, 264 F.3d 663, 676 (6th Cir.2001).

Having concluded that trial counsel performed reasonably, and that even if he did not, Williams cannot show prejudice, we need not delve into appellate counsel’s performance. In order to afford Williams a more thorough analysis of his claim, however, we briefly address the merits. Appellate counsel is constitutionally ineffective “if his or her conduct falls below an objective standard for reasonableness and causes the petitioner prejudice.” Williams v. Anderson, 460 F.3d 789, 800 (6th Cir.2006). “Failure of appellate counsel to raise an issue can amount to constitutionally ineffective assistance.” McFarland v.

Yukins, 356 F.3d 688, 710 (6th Cir.2004). However, it is clear that appellate counsel is not required to raise every non-frivolous issue on appeal. See id.; see also Jones v. Barnes, 463 U.S. 745, 751, 103 S.Ct. 3308, 77 L.Ed.2d 987 (1983). Strategic and tactical choices regarding which issues to pursue on appeal are “properly left to the sound professional judgment of counsel.” Perry, 908 F.2d at 59. In fact, “the hallmark of effective appellate advocacy” is the “process of ‘winnowing out weaker arguments on appeal and focusing on’ those more likely to prevail.” Smith v.

Murray, 477 U.S. 527, 536, 106 S.Ct. 2661, 91 L.Ed.2d 434 (1986) (quoting Barnes, 463 U.S. at 751-52, 103 S.Ct. 3308.) Here, Williams’s appellate counsel raised five arguments on appeal, including challenges to pre-trial identification procedures, evidentiary rulings, and the sufficiency of the evidence. Williams argues that counsel was nevertheless ineffective because he did not raise the ineffectiveness of Williams’s trial counsel.

But, as discussed above, trial counsel’s behavior was not objectively unreasonable, and Williams has not shown that there is a “reasonable probability” that, had trial counsel called Higginbotham and Ballard, he would have been acquitted. Strickland, 466 U.S. at 694, 104 S.Ct. 2052. Because Williams has not shown that his trial counsel was constitutionally ineffective, he cannot assert a meritorious claim that appellate counsel was constitutionally ineffective for declining to so argue.

We also reject one of Williams’s peripheral theories regarding appellate counsel’s alleged insufficiency: appellate counsel “neither knew nor researched controlling legal authority and thus did not review all the relevant facts before choosing which issues to appeal.” Williams cites a letter appellate counsel wrote to Williams stating that he would not raise ineffective assistance of trial counsel claims “because I cannot support them either in Law or in the transcripts.

Many of your suggestions involves [sic] trial strategy which are not appealable issues.” This letter, Williams argues, shows that appellate counsel did not know he could supplement the record via a Ginther hearing and thus rendered objectively unreasonable performance. We disagree. The letter shows only what it says: that appellate counsel did not believe he could support an ineffective assistance of trial counsel claim based on either the law or the record.

Given that this court considers “strategic and tactical choices” regarding which issues to pursue on appeal “properly left to the sound professional judgment of counsel,” Perry, 908 F.2d at 59, Williams does not meet his burden merely by pointing to these two sentences in a letter. II.

Evidentiary hearing Williams asserts that, even if this court finds his Strickland claims unpersuasive on the record, he is nevertheless entitled to an evidentiary hearing in federal court regarding whether appellate counsel reasonably investigated trial counsel’s representation and whether trial counsel used reasonable efforts to find Higginbotham and Ballard. The district court denied his request for such a hearing. We review a district court’s denial of an evidentiary hearing for an abuse of discretion. Cornwell v. Bradshaw, 559 F.3d 398, 410 (6th Cir.2009).

Analysis As noted above, Pinholster has limited “ § 2254(d)(1) review to the state-court records” where the state court ruled on the merits. 131 S.Ct. at 1398-99. Even assuming that the district court could have granted Williams an evidentiary hearing, it did not abuse its discretion in refusing to do so. Under the provisions of AEDPA which allow evidentiary hearings notwithstanding Pinholster, a petitioner is not permitted to develop new facts in federal court unless he or she, in relevant part, was “diligent in developing” the claim in state court. Williams v.

Taylor, 529 U.S. 420, 437, 120 S.Ct. 1479, 146 L.Ed.2d 435 (2000); see also 28 U.S.C. § 2254(e)(2). Phrased differently, a district court may not hold an evidentiary hearing on a habe-as claim if the petitioner has “failed” to develop the factual basis for the claim in state court. McFarland, 356 F.3d at 712. “ ‘Failed’ within the meaning of [AEDPA] refers to ‘a lack of diligence, or some greater fault, attributable to the prisoner or the prisoner’s counsel.’ ” Getsy v. Mitchell, 495 F.3d 295, 310 (6th Cir.2007) (quoting Williams, 529 U.S. at 432, 120 S.Ct. 1479).

A petitioner “who was diligent in seeking a state evidentiary hearing, but whose requests were denied in the state courts, did not ‘[fail] to develop the factual basis of the claim.’ ” McFarland, 356 F.3d at 712. Nevertheless, an evidentiary hearing is discretionary, and even if allowed by AEDPA, the district court must consider whether the hearing could enable the petitioner to prove factual allegations which, if true, would entitle him or her to federal habeas relief. Schriro v. Landrigan, 550 U.S. 465, 474, 127 S.Ct. 1933, 167 L.Ed.2d 836 (2007).

Thus, “if the record refutes the applicant’s factual allegations or otherwise precludes habeas relief, a district court is not required to hold an evidentiary hearing.” Id. In this case, Williams sought an evidentiary hearing on the appellate counsel claim during the state post-conviction proceedings. He therefore did not “fail” to develop its factual basis below. See McFarland, 356 F.3d at 712. The question remains, however, whether the district court abused its discretion in denying Williams an evidentiary hearing. We conclude that it did not.

The record precludes habeas relief because it reveals neither ineffective assistance of trial nor appellate counsel, and the extra-record evidence to which Williams makes reference would not enable him “to prove factual allegations which, if true, would entitle him ... to federal habeas relief.” Landrigan, 550 U.S. at 474, 127 S.Ct. 1933. The new evidence Williams would offer appears to include a showing that: (

i) appellate counsel did not adequately investigate trial counsel’s performance; and (ii) trial counsel did not “exert a reasonable effort” to find Higginbotham and Ballard in order to call them as witnesses at the trial. This might include affidavits and testimony from counsel, and from Higginbotham and Ballard regarding how they would have testified. However, as discussed above, even if Higginbotham and Ballard indicated that they would have testified to a lack of premeditation, that would not entitle Williams to habeas relief.

And if it came to light that trial counsel’s decision not to call the witnesses was due to error rather than strategy, Williams still could not show prejudice: as we have said, the lack of testimony was not prejudicial, and whether its absence was the result of error or strategy is irrelevant to that determination. CONCLUSION For the foregoing reasons, we DENY Williams’s petition for habeas relief and AFFIRM the district court’s judgment denying Williams an evidentiary hearing. .

The Michigan Court of Appeals summarized the underlying facts of the case in the opinion it issued affirming Williams’s conviction and sentence: The witnesses testified that, prior to the shooting, defendant argued with Cotton and Cotton told defendant not to show his gun or not to shoot him. Defendant and Cotton were standing approximately six to eight feet apart when defendant began shooting. After the first shot, Cotton fell down to the ground and defendant fired a couple of more shots at Cotton.

After shooting Cotton at least three times, defendant stepped closer to Cotton and shot him several more times until he emptied the gun. When there were no more bullets in the gun, defendant hit Cotton in the face with the gun.... [Defendant did not instantly or impulsively attack Cotton, or inflict the deadly wounds within only seconds....

Cotton received five gunshot wounds to his head and body, including one wound above the right eye and one wound to the middle of [sic] upper abdomen.... [T]here was no evidence of a struggle, an attack by Cotton or any other occurrence that would prompt any unthinking use of the firearm. People v. Williams, Mich.Ct.App. No. 257142, April 13, 2006, at 5, 2006 WL 954167. .

As stated above, Brown concerned whether an ineffective assistance of counsel claim had been "adjudicated on the merits” where the facts that formed "the basis of the claim were not in the record before the [state court], and that court explicitly acknowledged that its review was limited to mistakes apparent on the record.” 551 F.3d at 428-29 (internal quotation omitted). In deciding the question, we relied on Brady v.

Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), jurisprudence: "[I]n the context of a Brady claim, when the petitioner’s habeas claim involves Brady material that was uncovered only during the federal habeas proceedings, AEDPA deference does not apply to an earlier, state-court Brady adjudication involving a different mix of allegedly improperly withheld evidence.” Brown, 551 F.3d at 429 (citing Joseph v. Coyle, 469 F.3d 441, 469 (6th Cir.2006)).

The Brown court instructed that "the same principle applies generally whenever new, substantial evidence supporting a habe-as claim comes to light during the proceedings in federal district court.” Id. In other words, under Brown, if a hearing in federal court brings new and significant evidence to light, then federal courts are not bound by AEDPA because, as the thinking goes, there was no adjudication on the merits to trigger AEDPA.

But if a court is bound by AEDPA — because the state court ruled on the merits — then it is also bound by the clear rule in Pinholster: Habeas corpus review under § 2254(d)(1) is limited to the record that was before the state court, and federal courts thus may not grant evidentiary hearings in order to supplement the facts found below. See Pinholster, 131 S.Ct. at 1398. The tension between Brown and Pinholster is obvious: Brown effectively uses federal evidentiary hearings to determine whether AEDPA applies, while Pinholster instructs that, under AEDPA § 2254(d)(1), evi-dentiary hearings are prohibited.

Yet, it may be possible for the two principles to work harmoniously in tandem, to wit, in a case where the state court did not, in fact, adjudicate the merits. In such a case, arguably, our decision in Brown would allow an evidentiary hearing and de novo review, and Pinholster would present no obstacle because it applies only where "the state court ... adjudicated the claim on the merits.” Pinholster, 131 S.Ct. at 1398. (And, of course, AEDPA’s evi-dentiary hearing provision, which Pinholster does not affect, is still in play.

See id. at 1400-01 (discussing the continued vitality of § 2254(e)(2).)) In any event, whether Pinhol-ster and Brown can actually coexist as a practical matter is a question better addressed by a case in which it affects the outcome. We thus decline to take up the issue. . In Greer, this court articulated several considerations to weigh when assessing whether an attorney on direct appeal has performed with reasonable competence. Those considerations are not dispositive in this case, and many of them are irrelevant here. 264 F.3d at 679. .

As Pinholster instructs, § 2254(e)(2), which specifically addresses evidentiary hearings, “continues to have force where § 2254(d)(1) does not bar federal habeas relief. For example, not all federal habeas claims by state prisoners fall within the scope of § 2254(d), which applies only to claims 'adjudicated on the merits in State court proceedings.' " Thus, § 2254(e)(2) still allows but “restricts the discretion of federal habeas courts to consider new evidence when deciding claims that were not adjudicated on the merits in state court.” Pinholster, 131 S.Ct. at 1401.