Marvel DANIEL, Petitioner-Appellee, v. Cindi S. CURTIN, Respondent-Appellant, 499 F. App'x 400

OPINION KAREN NELSON MOORE, Circuit Judge. The state appeals a judgment from the United States District Court for the Eastern District of Michigan granting Marvel Daniel’s (“Daniel”) petition for a writ of habeas corpus on the grounds of ineffective assistance of counsel.

In addition to arguing that the district court erred in granting the writ, the state contends that the district court improperly held an evi-dentiary hearing, that all evidence presented during that hearing should be omitted from this court’s consideration, and that the deferential standard of review set out in the Anti-Terrorism and Death Penalty Act of 1996 (“AEDPA”) — rather than de novo review — should apply because Daniel’s claim was adjudicated on the merits in state court.

Because in light of intervening Supreme Court precedent the district court improperly granted Daniel’s petition for a writ of habeas corpus, we REVERSE the judgment of the district court and REMAND for further proceedings consistent with this opinion. I. BACKGROUND AND PROCEDURAL HISTORY On July 12, 2003, Derrick Mitchell (“Mitchell”) was shot while driving his automobile in east Detroit. Police officers received a call alerting them to the shooting at 9:30 p.m. and found Mitchell dead at the scene when they arrived approximately five minutes later. R. 11-12 (Trial Tr., 7/20/04, at 55:3, 56:24-57:5).

Mitchell was a “known drug dealer” in the neighborhood and notoriously “hid his drugs in stereo equipment in his car.” People v. Daniel, No. 257658, 2005 WL 3440436, at *1 (Mich.Ct.App. Dec. 15, 2005) (unpublished opinion). That night at the scene, police took written statements from two eyewitnesses, Shawn O’Neil (“O’Neil”) and Dollie Muex (“Muex”), who were sitting on the front porches of nearby houses when the shooting took place. R. 34-4 (Exh. 2). Neither O’Neil nor Muex knew Mitchell.

Both witnesses reported hearing a single gunshot and then seeing Mitchell’s car collide into the back of a parked vehicle before crashing into a nearby tree. Id. at 2-A, 2-B. Both witnesses reported that after the vehicle crashed into the tree, two or three young boys ran toward the car, took something out of it, and then ran in the same direction they had come, away from the vehicle. Id. O’Neil stated that he did not see who shot Mitchell nor did he “notice . any other vehicles around the vehicle that crashed.” Id. at 2-A.

Muex reported seeing a “white [2002 mini] van ... with black interior, tinted windows” alongside Mitchell’s vehicle immediately before the shooting. Id. at 2-B. Muex also stated that the passenger window of the van appeared to be open. Id. Officers Shawn Stallard (“Stallard”) and Steven Compton (“Compton”) completed preliminary police reports documenting the incident. Id. at 2-C, 2-D.

Stallard wrote in his report that a “[white] newer model mini van [with] gold rims” was reported at the scene, and that he spoke with Muex, who reported hearing a single gun shot and witnessing the minivan traveling “next to/slightly behind” the victim’s vehicle. Id. at 2-C. Compton wrote in his report that the suspect was driving a “white” “2002 model” “minivan w[ith] gold rims,” and that Muex confirmed that the shots were fired from the “pass[enger] side front window of the sus[pect’s] vehficle.]” Id. at 2-D.

In November 2008, Daniel was arrested in connection with the shooting and charged with first-degree murder. R. 11-2 (State Ct. Docket at 1). During a two-day bench trial in July of 2004, the prosecution presented four key witnesses to the shooting: Darius Scott (“Scott”), Prentice Graham (“Graham”), Shantrice Riley (“Riley”), and Simon Melton (“Melton”). None of these witnesses provided statements to the police on the night in question. R. Ills (Trial Tr., 7/26/04, at 82:5-7). Daniel presented an alibi witness in his defense.

The Michigan Court of Appeals summarized the testimony given at trial as follows: One witness testified that defendant was driving his van at the time of the shooting. Two other witnesses also testified that defendant was driving his van at the time of the shooting, but they equivocated on this point at trial, stating that they assumed defendant was driving because they recognized the van as belonging to him.

One of the prosecution’s witnesses, Darius Scott, testified that, shortly after the shooting, defendant contacted him by telephone and simply stated, “yep, yep, yep, yep.” Before that time, Scott and defendant had discussed the shooting death of another drug dealer, Shawn, who sold drugs with defendant out of a house on Novara Street. Scott testified that he believed Mitchell was involved in Shawn’s shooting. ... Defendant offered an alibi defense. His girlfriend testified that he was at her home from 2:00 p.m. until approximately midnight on the date Mitchell was shot.

During this timeframe, she had the keys to defendant’s gray conversion van, and the van remained parked nearby. She testified that defendant’s van could not have been near Novara Street at the time of the shooting. Daniel, 2005 WL 8440486, at *1. The trial judge found sufficient evidence of Daniel’s involvement in the shooting, but reasonable doubt as to whether the shooting was premeditated. R. 11-15 (Trial Tr., 7/26/04, at 99:3-15). As a result, Daniel was convicted of second-degree murder and sentenced to fifty-to-seventy-five years in prison.

Daniel moved for a new trial, an eviden-tiary hearing, and a court-appointed investigator, and argued that he was denied effective assistance of counsel, among other claims. R. 11-2 (State Ct. Docket at 6). After holding an evidentiary hearing in which trial counsel, Daniel, and an additional alibi witness testified, the trial judge concluded that Daniel’s counsel was not ineffective and denied Daniel’s motion for a court-appointed investigator. R. 11-17 (Ginther Hr’g, 4/25/05, at 67:12-18). The Michigan Court of Appeals affirmed. Daniel, 2005 WL 3440436, at *1, *6.

Daniel petitioned for rehearing, arguing that he was impeded from presenting evidence to substantiate his claim by the trial court’s refusal to appoint an investigator, but the Michigan Court of Appeals summarily denied his petition. R. 11-18 (Mieh.Ct.App.Order). The Michigan Supreme Court denied leave to appeal. R. 11-19 (Mich.Sup.Ct.Order). Daniel filed a petition for a writ of habe-as corpus in federal district court raising the same claims presented in state court. R. 1 (Habeas Pet.). The district court granted an evidentiary hearing on two issues, R. 12 (Dist. Ct.

Order, 11/4/08, at 1), and then granted Daniel’s request to expand the hearing’s scope to address whether counsel was ineffective in failing to investigate eyewitnesses O’Neil and Muex, R. 21 (Dist. Ct. Order, 9/11/09, at 1). During the evidentiary hearing, O’Neil expanded on his prior statements to the police and testified that he witnessed a white minivan drive alongside Mitchell’s vehicle prior to the shooting. R. 28 (Evid. Hr’g Tr., 2/22/10, at 23:24-27:5). O’Neil also testified that a picture of Daniel’s van did not resemble the van he saw on the night of the shooting. Id. at 28:25-29:1.

An affidavit from Muex was introduced into evidence largely affirming her prior statement to the police. Id. at 39:21-40:9; see also 34-7 (Muex Aff.). The district court concluded that trial counsel’s failure to investigate O’Neil’s and Muex’s testimony amounted to ineffective assistance of counsel that prejudiced Daniel and, thus, granted Daniel’s petition for habeas relief. R. 36 (Dist. Ct. Op. at 13-17).

In doing so, the district court applied de novo review because it determined that the Michigan Court of Appeals “ignored the gravamen of petitioner’s claim,” and “failed to provide funds to petitioner to hire an investigator.” Id. at 11. The district court also stated, however, that it would have reached the same result “[e]ven under the higher AEDPA review standard.” Id. at 17 n. 3. The state timely appeals. II. ANALYSIS The district court’s opinion in this matter was issued prior to important Supreme Court precedent governing the habeas corpus legal landscape.

Thus, the Supreme Court’s decisions in Harrington v. Richter, -U.S.-, 131 S.Ct. 770, 178 L.Ed.2d 624 (2011), and Cullen v. Pinholster, — U.S.-, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011), require that we scrutinize two key aspects of the district court’s decision: the application of de novo review and the consideration of evidence introduced during the federal evidentiary hearing. We address each issue before considering the merits of Daniel’s habeas petition. A. Standard of Review “We review the district court’s legal conclusions in habeas proceedings de novo and its findings of fact for clear error.” Akins v.

Easterling, 648 F.3d 380, 385 (6th Cir.2011) (internal quotation marks omitted). Because Daniel’s habeas petition was filed after AEDPA’s effective date, we apply AEDPA in our review of his petition. Id.

Under AEDPA, to uphold the district court’s grant of habeas relief we must conclude that the Michigan court’s decision “with respect to any claim that was adjudicated on the merits in State court proceedings” was (1) “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court” 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). “A state court’s decision would be considered contrary to established law if it is diametrically different from or opposite in character or nature to federal law as determined by the Supreme Court.” Akins, 648 F.3d at 385 (internal quotation marks and alterations omitted). “If the state court identifies the correct governing legal principle from the Supreme Court’s decisions, habeas relief is available under the unreasonable application clause if the state court unreasonably applies that principle to the facts of the prisoner’s case or unreasonably extends or unreasonably refuses to extend a legal principle from the Supreme Court precedent to a new context.” Id. (internal quotation marks and alterations omitted).

In order to warrant habeas relief, “the state court’s application of clearly established law must be objectively unreasonable” and not just “erroneous[ ] or incorrect ].” Id. at 385-86 (internal quotation marks and alterations omitted). “A decision adjudicated on the merits in a state court and based on a factual determination will not be overturned on factual grounds unless objectively unreasonable in light of the evidence presented in the state-court proceeding.” Id. at 386 (internal quotation marks omitted). “The state court’s factual determinations are ‘presumed to be correct,’ and the petitioner has ‘the burden of rebutting the presumption of the correctness by clear and convincing evidence.’ ” Id. (quoting 28 U.S.C. § 2254(e)(1)).

This deferential standard of review under AEDPA is limited to claims “adjudicated on the merits in State court proceedings.” 28 U.S.C. § 2254(d). The district court in this case, citing our decision in Maples v. Stegall, 340 F.3d 433, 436-37 (6th Cir.2003), concluded that.de novo review was appropriate because Daniel’s ineffective-assistance-of-counsel claim was not decided on the merits by the state court. R. 36 (Dist. Ct. Op. at 11).

In support of this conclusion, the district court found that (1) the Michigan Court of Appeals “ignored the gravamen of petitioner’s claim” discussing only the reasonableness of counsel’s decision not to have the two eyewitnesses testify as opposed to counsel’s decision not to investigate those witnesses; and (2) the state court impeded Daniel from presenting evidence in support of his claim by denying Daniel’s request for funding for an investigator. Id.

Since the district court issued its decision, the Supreme Court has clarified the meaning of “adjudicated on the merits” for the purposes of AEDPA’s deferential review. In Harrington v. Richter, the Supreme Court held that “[w]hen a federal claim has been presented to a state court and the state court has denied relief, it may be presumed that the state court adjudicated the claim on the merits in the absence of any indication or state-law procedural principles to the contrary.” 131 S.Ct. at 784-85. In essence, the Court clarified that “§ 2254(

d) does not require a state court to give reasons before its decision can be deemed to have been ‘adjudicated on the merits.’ ” Id. at 785. We have analyzed Richter’s exact import in light of prior Supreme Court precedent and concluded that it is only “when there is no explanation as to either Strickland prong” that “a habeas court must afford both prongs of AEDPA deference after determining what arguments or theories could have supported the state court’s decision.” Rayner v. Mills, 685 F.3d 631, 637 (6th Cir.2012) (internal quotation marks and alterations omitted).

This means that when a state court discusses only one Strickland prong, we continue to follow the Supreme Court’s decision in Wiggins v. Smith, 539 U.S. 510, 534, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003), and review de novo the prong not reached by the state court. Id. at 637-38; see also Davis v. Lafler, 658 F.3d 525, 537 (6th Cir.2011) (en banc). There is no dispute that Daniel raised the same ineffective-assistance-of-counsel claim that he now advances in federal court before the Michigan Court of Appeals. See R. 11-18 (Daniel Mich. Ct.App. Br. at 13).

It is also clear that the Michigan court expressed no view on the merits of the first prong of Daniel’s Strickland claim: whether counsel’s failure to investigate the two eye witnesses constitutes deficient performance. Nevertheless, it does appear that the Michigan Court of Appeals addressed the prejudice prong of the Strickland inquiry insofar as it concluded that “[d]efendant has not met his burden of proving that the outcome of his case would have been different if his counsel had ... called other witnesses to testify.” Daniel 2005 WL 3440436, at *3.

Accordingly, we review the deficient-performance prong of Daniel’s Strickland claim de novo, but, contrary to the holding of the district court, still afford the prejudice prong AEDPA deference. B. Federal Evidentiary Hearing The district court conducted an evi-dentiary hearing on Daniel’s ineffective-assistance-of-counsel claim, at least in part because the state court refused to provide funds for Daniel to hire an investigator, which Daniel argues impeded him from presenting evidence to the Michigan state courts.

The state argues that the district court erred in granting Daniel a federal evidentiary hearing and that we may not consider the evidence introduced during that hearing on appeal. We agree with the state in light of recent Supreme Court precedent. The Supreme Court held in Cullen v. Pinholster that claims under § 2254(d)(1) of AEDPA must be assessed solely on the record before the state court. The Court explained:

Section 2254(d)(1) refers, in the past tense, to a state-court adjudication that ‘resulted in’ a decision that was contrary to, or ‘involved’ an unreasonable application of, established law. This backward-looking language requires an examination of the state-court decision at the time it was made. It follows that the record under review is limited to the record in existence at that same time i.e., the record before the state court. 131 S.Ct. at 1398.

Accordingly, we are limited to the state-court record and we may not consider the new evidence presented during the evidentiary hearing before the district court, at least insofar as it is relevant to the prejudice prong of Daniel’s Strickland claim. See Campbell v. Bradshaw, 674 F.3d 578, 586 (6th Cir.2012) (“In other words, ‘[i]f a. claim has been adjudicated on the merits by a state court, a federal habeas petitioner must overcome the limitation of § 2254(d)(1) on the record that was before that state court.’ ” (quoting Pinholster, 131 S.Ct. at 1400)).

We also note that Daniel has not presented any reason why the strictures of Pinholster would not apply in his particular case. The majority in Pinholster did recognize that “state prisoners may sometimes submit new evidence in federal court” because it is axiomatic that “not all federal habeas claims by state prisoners fall within the scope of § 2254(d)”; some claims will not have been previously “adjudicated on the merits in [the] State court proceedings.” Pinholster, 131 S.Ct. at 1401 (internal quotation marks omitted).

In dissent, Justice Sotomayor suggested that “[t]here may be situations in which new evidence supporting a claim adjudicated on the merits gives rise to an altogether different claim” and that “the majority [did] not intend to suggest that review is limited to the state-court record when a petitioner’s inability to develop the facts supporting his claim was the fault of the state court itself.” Id. at 1417 n. 5 (Soto-mayor, J., dissenting). The majority seemed to agree with this proposition, at least in the abstract.

See id. at 1401 n. 10 (“Though we do not decide where to draw the line between new claims and claims adjudicated on the merits, Justice Sotoma-yor’s hypothetical involving new evidence of withheld exculpatory witness statements may well present a new claim.” (internal citations omitted)). Daniel makes no argument that his case falls into this exception, and we therefore have no occasion to consider its existence or scope.

While it is true that the state court did not provide funding for an independent investigator as he requested, Daniel was afforded the benefit of a Ginther hearing during which he could have introduced evidence relevant to his ineffective-assistance-of-counsel claims. Indeed, at the Ginther hearing the trial judge stated: “I’m not precluding [Daniel] or [defense counsel] from calling any witness that [they] want to call. I’ll make that abundantly clear for the record.” R. 11-17 (Ginther Hr’g, 4/25/05, at 48:2-4).

Accordingly, we conclude that Daniel’s case falls squarely within the Supreme Court’s holding in Pinholster that claims decided on the merits in state court and assessed under § 2254(d)(1) must be confined to the record before the state court. As a result, we may not consider the additional evidence presented during the federal evidentiary hearing. See Campbell, 674 F.3d at 590 n. 3 (“We similarly cannot review the testimony ... [given] at the federal evidentiary hearing because it was not part of the state court record.”). C.

Ineffective-Assistance-of-Counsel Claim The Sixth Amendment guarantees criminal defendants “the right to the effective assistance of counsel.” Strickland v. Washington, 466 U.S. 668, 686, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984) (internal quotation marks omitted).

To state a claim for ineffective assistance of counsel, a defendant must demonstrate: (1) “that counsel’s representation fell below an objective standard of reasonableness”; and (2) “that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Id. at 688, 694, 104 S.Ct. 2052. “Our review of counsel’s performance is highly deferential and counsel is strongly presumed to have rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment.” Johnson v.

Bell, 525 F.3d 466, 487 (6th Cir.2008) (internal quotation marks omitted), cert. denied, — U.S. -, 129 S.Ct. 1668, 173 L.Ed.2d 1039 (2009). 1. Performance In Wiggins, the Supreme Court held that counsel’s failure to conduct an adequate investigation before deciding on a defense strategy constituted ineffective assistance of counsel under Strickland. 539 U.S. at 533-34, 123 S.Ct. 2527.

The Court explained that “strategic choices made after less than complete investigation are reasonable only to the extent that reasonable professional judgments support the limitations on investigation” and that consequently “[a] decision not to investigate ... must be directly assessed for reasonableness in all the circumstances.” Id. at 533, 123 S.Ct. 2527 (internal quotation marks omitted).

The Court then concluded that the state court’s application of Strickland was “objectively unreasonable” because the state court failed to assess whether counsel’s decision to cease further mitigation investigation upon the acquisition of two distinct types of records “actually demonstrated reasonable professional judgment.” Id. at 527, 123 S.Ct. 2527.

The Court explained that, instead, “[t]he state court merely assumed that the investigation was adequate” when in fact the record revealed that “counsel chose to abandon their investigation at an unreasonable juncture, making a fully informed decision with respect to sentencing strategy impossible.” Id. at 527-28, 123 S.Ct. 2527. We conclude that Wiggins is directly applicable to the case before us. Daniel’s counsel failed to investigate — in fact failed even to attempt to contact — the only two unbiased eyewitnesses to the murder, O’Neil and Muex. See R. 11-17 (Ginther Hr’g Tr., 4/25/05, at 16:17-19).

Instead, counsel relied solely on these witnesses’ statements to the police to conclude that they would not be helpful to Daniel’s defense. Id. at 16:18-19. At the Ginther hearing before the state court, Daniel’s counsel justified this decision by explaining that he was concerned that Muex and O’Neil might corroborate the testimony from the prosecution’s witnesses because their statements were not “necessarily ... inconsistent with what the prosecution witnesses were talking about.” Id. at 13:23-14:5.

In particular, he expressed concern about the witnesses corroborating the removal of the object from Mitchell’s vehicle immediately following the shooting, as this could support a finding of felony first-degree murder. Id. at 15:3-12. While the reasons stated by Daniel’s counsel might explain why an attorney ultimately would decide not to present the testimony of a witness at trial, they do not explain why counsel would choose not to conduct even minimal investigation into the potential utility of a witness. See Porter v.

McCollum, 558 U.S. 30, 130 S.Ct. 447, 453, 175 L.Ed.2d 398 (2009) (concluding that counsel’s performance was constitutionally deficient where counsel “did not even take the first step of interviewing witnesses or requesting records”); Towns v.

Smith, 395 F.3d 251, 260 (6th Cir.2005) (stating that while under some circumstances it may have been “objectively reasonable” to decline to call a particular witness, “it was objectively unreasonable for counsel to make that decision without first investigating [the potential witness], or at least making a reasoned professional judgment that such investigation was unnecessary.”)- Indeed, applying Wiggins we have held on numerous occasions that a decision made by counsel without the necessary investigation constitutes deficient performance. See, e.g., Couch v.

Booker, 632 F.3d 241, 246 (6th Cir.2011) (“While the point of the Sixth Amendment is not to allow Monday-morning quarterbacking of defense counsel’s strategic decisions, a lawyer cannot make a protected strategic decision without investigating the potential bases for it.”); English v. Romanowski, 602 F.3d 714, 728 (6th Cir.2010) (“Although defense counsel’s decision not to call Ceruti was not an unreasonable trial strategy, we hold that counsel’s failure to adequately investigate that decision before trial was deficient performance sufficient to satisfy the first prong of the Strickland test.”); Bigelow v.

Haviland, 576 F.3d 284, 289 (6th Cir.2009) (concluding that counsel’s failure to seek corroboration for an alibi defense constituted deficient performance under Strickland); Poindexter v. Booker, 301 Fed.Appx. 522, 528-29 (6th Cir.2008) (unpublished opinion) (concluding that counsel’s failure to investigate two alibi witnesses constituted deficient performance under Strickland).

In this case, it would have been clear to even the most unseasoned counsel that Muex and O’Neil were potentially important witnesses by virtue of the fact that they were the only disinterested witnesses to the shooting who provided statements to police on the night in question. See Wiggins, 539 U.S. at 525, 123 S.Ct. 2527 (“[A]ny reasonably competent attorney would have realized that pursuing these leads was necessary to making an informed choice among possible defenses.”).

In light of this fact, it is difficult to fathom why counsel would not make even the most minimal efforts to contact and interview these individuals to probe their utility to the defense. See id. at 533, 123 S.Ct. 2527 (while Strickland “does not require counsel to investigate every conceivable line of ... evidence” no matter the prospect of its utility, decisions as to how to allocate investigatory resources must be reasonable). Indeed, it is difficult to understand how counsel could have made a reasoned professional judgment not to utilize Muex and O’Neil without at least speaking to them once. See id.

Moreover, contrary to Daniel’s trial counsel’s assertions, the statements that O’Neil and Muex gave to the police on the night of the shooting were not so obviously unhelpful to the defense to warrant no further investigation; O’Neil’s and Muex’s statements did conflict in significant respects with the testimony presented by the prosecution’s witnesses at trial. Muex stated to the police on the night of the shooting that the van she saw was a white minivan with gold rims and that the shots were fired from the van while it was moving. See R. 34-4 (Exh. 2 at 2-B).

This contradicted testimony from the prosecution’s witnesses that the shooter was in a gray conversion van and that the shooter exited the van during the course of the shooting. See R. 11-12 (Trial Tr., 7/20/04, at 33:11-14) (Scott Test.); id. at 70:16-23 (Graham Test.); R. 11-13 (Test. Tr., 7/20/04, at 75:18-77:5) (Graham Test.); id. at 117:7-13 (Riley Test.).

Because there was no physical evidence linking Daniel to the scene of the crime and, thus, the prosecution’s case hinged on these witnesses’ ability to identify Daniel at the shooting by virtue of the presence of his van, Muex’s contradiction of the van description was of great potential utility to the defense. Both Muex and O’Neil also testified that two or three young men approached the victim’s car on foot after the shooting occurred. See R. 34-4 (Exh. 2 at 2-A, 2-B).

This directly contradicted the testimony of at least two of the prosecution’s witnesses, who said that the shooter emerged from the van during the course of the shooting and removed the item from the victim’s car. See R. 11-15 (Trial Tr., 7/20/04, at 7:10-12, 8:1-12, 9:10-12) (Melton Test.); R. 11-13 (Trial Tr., 7/20/04, at 120:10-15) (Riley Test.).

Moreover, this contradiction exposes the weakness of counsel’s justification for not pursuing Muex and O’Neil as potential defense witnesses because Muex’s and O’Neil’s statements actually undermined the prosecution witnesses’ testimony that might support a finding of felony murder. Muex’s and O’Neil’s statements suggest that the removal of the box from the victim’s vehicle was unrelated to the shooting — or at least not perpetrated by the shooter.

Indeed, if the opposite were true, or if it was at least clear that these witnesses bolstered the prosecution’s case, it would seem strange that the prosecution had not elicited their assistance. At the very least, defense counsel should have clarified whether these witnesses could offer any potentially damaging testimony to bolster the prosecution’s case in order to be prepared to rebut it, particularly given the minimal effort involved. See Rompilla v.

Beard, 545 U.S. 374, 385-86, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005) (concluding counsel’s failure to review file containing defendants’ criminal history in order to counter the prosecution’s case constituted deficient performance). The minimal effort required to investigate Muex’s and O’Neil’s utility to the defense also distinguishes this case from the Supreme Court’s decision in Richter.

In Richter, the Supreme Court considered a claim of ineffective assistance of counsel premised on' counsel’s failure to elicit the assistance of various experts to provide analysis of the types and patterns of blood found at the crime scene. In that case, Richter was convicted of the murder of Patrick Klein (“Klein”) in large part based on testimony from Joshua Johnson (“Johnson”), who had called 911 and was present at the scene of the crime, also ■with gun shoot wounds, when the police arrived. Richter, 131 S.Ct. at 781.

Richter’s defense centered on refuting Johnson’s version of the events by contending that another man, Christian Bran-seombe (“Branscombe”), had shot Klein in self defense after Branscombe was attacked by Klein and Johnson. Id. at 782. Richter contended that if counsel had secured a blood expert on the matter, the “expert could have testified that the blood in Johnson’s doorway could not have come from Johnson and could have come from Klein, thus suggesting that Richter’s version of the shooting was correct and Johnson’s a fabrication.” Id. at 789.

The Supreme Court concluded that counsel’s failure to secure the experts did not constitute deficient performance because counsel was reasonable in determining that the risks that the blood evidence might undermine Richter’s version of the events justified not pursuing that forensic evidence. Id. at 789-90. In so doing, the Court recognized that “[w]hen defense counsel does not have a solid case, the best strategy can be to say that there is too much doubt about the State’s' theory for a jury to convict.” Id. at 791.

While there are arguably some similarities between Richter and the present case, Richter is sufficiently distinguishable so as to make it inapplicable here. It is true that in Richter defense counsel made a decision not to pursue evidence because of its potential to harm the defense and that counsel in this instance provided a similar justification for not investigating Muex and O’Neil. However, in Richter the foregone investigation would have required a substantial commitment of resources, as commission of a blood expert would have been a costly undertaking.

The Court reasoned that counsel was justified in deciding to allocate precious and limited defense resources elsewhere, and to focus instead on “draw[ing] attention to weaknesses in [the state’s experts’] conclusions.” Id. In this instance, however, the effort required for counsel to conduct at least a preliminary investigation of Muex and O’Neil was minimal. Counsel was provided police reports with the witnesses’ contact information and, therefore, counsel could have at least attempted to contact the witnesses by telephone to explore their testimony and determine their utility to the defense.

Even if counsel suspected that these witnesses might be harmful to the defense, a simple phone call to confirm the validity of these suspicions would have had no impact on counsel’s ability to pursue other avenues of defense. Based on the foregoing, we conclude that counsel’s failure to investigate the only two unbiased witnesses to the murder constitutes deficient performance under both Wiggins and Strickland. See Wiggins, 539 U.S. at 527-28, 123 S.Ct. 2527 (concluding state-court conclusion that “merely assumed that the investigation was adequate ... was objectively unreasonable”). 2.

Prejudice The fact that Daniel’s counsel’s performance fell below objective standards of reasonableness is not, however, the end of our inquiry. We must consider whether Daniel was prejudiced by his counsel’s error under AEDPA’s deferential standard of review. Before doing so, it is important to emphasize the proper bounds within which such analysis must operate.

The purpose of federal habeas review under AEDPA is to “guard against extreme malfunctions in the state criminal justice systems” and thus we are not to utilize it as “a substitute for ordinary error correction through appeal.” Richter, 131 S Ct. at 786 (internal quotation marks omitted). For this reason, when reviewing the reasonableness of a state court’s application of clearly established federal law, we are to “presum[e] that state courts know and follow the law,” Woodford v.

Visciotti, 537 U.S. 19, 24, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002), and to uphold “[the] state court’s determination that a claim lacks merit ... so long as ‘fairmind-ed jurists could disagree’ on the correctness of the state court’s decision,” Richter, 131 S.Ct. at 786. We must be careful to distinguish “an unreasonable application of federal law ... from an incorrect application of federal law” as AEDPA authorizes us only to correct the former, not the later. Id. at 785 (internal quotation marks omitted).

In short, “even a strong case for relief does not mean the state court’s contrary conclusion was unreasonable.” Id. at 786. Furthermore, we look only to the Supreme Court’s prior decisions in determining whether the state court unreasonably applied clearly established federal law because “circuit precedent does not constitute ‘clearly established Federal law, as determined by the Supreme Court’” and “[i]t therefore cannot form the basis for habeas relief under AEDPA.” Parker v. Matthews, — U.S. -, 132 S.Ct. 2148, 2155, 183 L.Ed.2d 32 (2012) (internal quotation marks omitted); see also Renico v.

Lett, 559 U.S. 766, 130 S.Ct. 1855, 1865-66, 176 L.Ed.2d 678 (2010). In assessing whether a petitioner has satisfied the prejudice prong of the Strickland analysis, the Supreme Court has emphasized that “the Strickland inquiry requires ... probing and fact-specific analysis.” Sears v. Upton, — U.S. -, 130 S.Ct. 3259, 3266, 177 L.Ed.2d 1025 (2010).

The Court has concluded that a state court’s prejudice determination was “unreasonable insofar as it failed to evaluate the totality of the available ... evidence” in determining whether, absent the error, there “existed a reasonable probability that the result ... would have been different.” Williams v. Taylor, 529 U.S. 362, 397-98, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000); see also Lafler v. Cooper, — U.S. -, 132 S.Ct. 1376, 1391, 182 L.Ed.2d 398 (2012) (holding petitioner satisfied the prejudice prong of the Strickland analysis).

Where “a verdict or conclusion [is] only weakly supported by the record [it] is more likely to have been affected by errors than one with overwhelming support in the record” because the failure of counsel to pursue a defense capable of instilling “reasonable doubt respecting guilt” can make the difference between conviction and acquittal. Strickland, 466 U.S. at 695-96, 104 S.Ct. 2052. The case against Daniel admittedly is weak in several important respects. First, the prosecution presented no physical evidence linking Daniel to the scene of the crime or to the commission of the shooting.

Instead, the prosecution built its case solely on the testimony of witnesses linking Daniel to the scene of the crime. The prosecution’s four key witnesses equivocated in their testimony placing Daniel at the scene of the crime, and several of the witnesses provided internally inconsistent testimony. Scott testified that he saw Daniel’s van — a brown conversion van — at the scene of the murder, R. 11-12 (Trial Tr., 7/20/04, at 12:15-18, 14:1-3), but Scott later retracted this testimony and suggested instead that the van might have been gray, id. at 33:11-22, 37:5-10.

Although Scott’s most definitive basis for connecting Daniel to the crime was his testimony that Daniel called him shortly after the murder and said “yep, yep, yep, yep” and then hung up the phone, id. at 17:12-18:8, the prosecution did not present phone records corroborating that Scott actually received this call. Two other witnesses, Graham and Melton, reported seeing Daniel’s gray van at the scene of the crime and seeing Daniel driving the van earlier in the day, but both equivocated on the point of whether they were certain they saw Daniel driving the van when the shooting occurred.

R. 11-13 (73:7-11, 100:1-11, 103:21-104:9, 107:3-10) (Graham Test.); R. 11-15 (5:16-21, 14:13-15:14, 32:13-21) (Melton Test.). Riley testified that she saw a “big,” “gray” van at the scene of the crime, R. 11-13 (117:7-13), but stated that she could not identify the passenger or the driver, id. at 123:9. Thus, while the prosecution undeniably presented evidence connecting Daniel to the crime, that evidence was far from air tight and certainly left potential for reasonable doubt.

This inconclusive testimony was further weakened by the fact that each of the prosecution’s witnesses was biased in some way, either by virtue of friendship with the victim and/or by offering statements to the police only after having been brought in based on charges against them. Scott offered information to the police only after he had been taken into custody on a gun charge. R. 11-12 (Trial Tr., 7/20/04, at 36:1-2, 37:20-22).

Graham was a friend of Mitchell’s, but not of Daniel’s, id. at 67:3-7; 69:12-17; R. 11-13 (Trial Tr., 7/20/04, at 91:5-12), and provided the police with a statement only after he was arrested for a parole violation, id. at 103:12-16. Riley also knew Mitchell well, id. at 124:22-23, and the prosecutor admitted at closing arguments that Riley may have stretched her testimony as a result of her relationship with Mitchell, R. 11-15 (Trial Tr., 7/26/04, at 71:5-7). Melton was Mitchell’s friend, id. at 3:7-13, and did not provide a statement until he was picked up by police, id. at 20:21-21:19; 25:6-11.

In spite of these weaknesses and inconsistencies, however, the fact-finder in Daniel’s bench trial “found” the prosecution’s witnesses to be “by and large, ... credible.” Id. at 98:9-13. At the Ginther hearing, the judge further explained that he found Scott’s testimony as to Daniel calling Scott immediately after the murder and stating “yep, yep, yep” to be “the most damning evidence against” Daniel. R. 17 (Ginther Hr’g, at 60:22-25).

While' the statements by Muex and O’Neil did in some ways contradict testimony presented by the prosecution’s witnesses, these contradictions were not so startling as to undermine fundamentally the prosecution’s case. Although Muex and O’Neil contended that the van in question was white, not gray, this contradiction is of limited utility because the shooting occurred at dusk. As light continually faded from the street, shadows could have altered the apparent color of the van, causing witnesses to assume that the van was white when in fact it was gray, or vice versa.

Indeed, the fact-finder was not bothered by the fact that the prosecution’s key witness shifted from testifying that the van was brown to testifying that the van was gray, perhaps for this very reason. Had Muex and O’Neil testified that the van was an obviously different color, such as red, or that the shooter was driving a small compact car, their testimony might have had a greater impact on the weight of evidence presented.

However, given their statements with an inconsistency possibly attributable to the setting sun, we cannot say that the state court was unreasonable in determining that there was not a reasonable probability that the statements’ inclusion would have impacted the outcome of the trial. The second difference between Muex’s and O’Neil’s statements and the trial testimony was that Muex and O’Neil said that several boys, and not the shooter, removed the box from the victim’s vehicle after the murder. This is admittedly a material difference.

However, the judge in the bench trial found Daniel guilty of second-degree murder, not first-degree felony murder, and therefore made no factual finding as to who removed the box from the victim’s car and whether this was connected to the shooting. R. 15 (Trial Tr., 7/26/04, at 99:7-15). The practical result is that this factual inconsistency is therefore largely irrelevant and could have served only to undermine further the credibility of the prosecution’s witnesses.

Given that there were already a large number of inconsistencies in the prosecution’s witnesses’ testimony — inconsistencies that the trial judge acknowledged, but nevertheless concluded did not undermine their credibility — it was reasonable for the state court to determine that further contradictions would have been cumulative. As a result, we cannot conclude that the state court was unreasonable in determining that there was no reasonable probability that introduction of this evidence would have impacted the outcome of the trial.

We cannot deny that, were we evaluating prejudice outside of AEDPA’s confines, we likely would reach a different result. Counsel’s failure even to attempt to contact the only two unbiased witnesses to the murder is, in our view, deplorable. The evidence against Daniel was weak because the prosecution’s case was premised entirely on testimony from biased witnesses who the trial court recognized “have some contradictions in their testimony.” R. 11-15 (Trial Tr., 7/26/04, at 98:10-11).

Muex and O’Neil could have weakened the prosecution’s case by providing contradictory testimony with impartiality that the prosecution’s material witnesses lacked. However, our task today is not to consider ab initio whether Darnel’s counsel’s performance was prejudicial under Strickland. We are limited to deciding whether the state court’s Strickland determination was an unreasonable application of clearly established Federal law. Because we cannot conclude that the state court’s decision was so unreasonable, we must affirm the decision of the Michigan Court of Appeals. III.

CONCLUSION Based on the foregoing, we REVERSE the judgment of the district court granting Daniel’s petition for a writ of habeas corpus, and REMAND for further proceedings consistent with this opinion. . The district court denied habeas relief on Daniel’s remaining claims. Daniel does not appeal this determination. . We acknowledge that the Michigan court’s Strickland analysis focused solely on counsel’s failure to call Muex and O’Neil as witnesses during trial, which is slightly different from the Strickland claim premised on counsel’s failure to investigate Muex and O’Neil as potential defense witnesses.

See Daniel, 2005 WL 3440436, at *3. Nevertheless, because the prejudice inquiries for both Strickland claims in this instance are identical — both hinged on the reasonable probability that Muex’s and O’Neil's testimony would have impacted the outcome of the trial based on the information contained in the police reports memorializing Muex’s and O’Neil’s statements on the night of the shooting — it is appropriate to conclude that Michigan court’s prejudice determination is applicable to both Strickland claims. .

Because the new evidence is relevant to only our prejudice determination, we decline to consider whether such evidence could be considered in the course of our de novo review of the performance prong of the Strickland inquiry. . Admittedly, Wiggins is different insofar as it involved counsel's obligations during the sentencing phase of a capital trial. We do recognize that the ABA imposes heightened professional standards on counsel in the sentencing phase of a capital case. See Bobby v.

Van Hook, 558 U.S. 4, 130 S.Ct. 13, 17, 175 L.Ed.2d 255 (2009) (discussing 2003 ABA Guidelines for capital defense counsel); see also id. at 16, 130 S.Ct. 13 (The ABA guidelines are "useful ... only to the extent they describe the professional norms prevailing when the representation took place."); Rom-pilla v. Beard, 545 U.S. 374, 387, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005) (discussing 1982 ABA Standards for Criminal Justice, which are not capital-sentencing specific, as relevant to determining reasonableness of counsel's conduct during capital-sentencing phase).

However, the Sixth Amendment right to counsel is at core concerned with a defendant receiving "a fair trial.” Strickland, 466 U.S. at 686, 104 S.Ct. 2052 (explaining that the right to effective assistance of counsel extends to “[a] capital sentencing proceeding” because such a proceeding "is sufficiently like a trial in its adversarial format and in the existence of standards for its decisions.”). Therefore, despite their capital-sentencing contexts, Strickland and Wiggins remain applicable to counsel’s conduct during the merits portion of a criminal trial. See, e.g., Ramonez v.

Berghuis, 490 F.3d 482, 488 (6th Cir.2007) ("[T]he state court ignored the central teaching of Strickland, as reaffirmed by Wiggins, that the investigation leading to the choice of a so-called trial strategy must itself have been reasonably conducted lest the ‘strategic’ choice erected upon it rest on a rotten foundation.” (citations omitted)). . The Michigan Court of Appeals correctly articulated the standard for prejudice under Strickland as requiring Daniel to demonstrate "a reasonable probability that the result of the proceeding would have been different” but for counsel’s errors.

Daniel, 2005 WL 3440436, at *1. Later in its analysis, however the Michigan court incorrectly restated this standard in concluding that Daniel "has not met his burden of proving that the outcome of his case would have been different if his counsel had ... called other witnesses to testify.” Id. at *3; see also id. at *3 n. 2 ("Defendant does not make any argument, however, to support that ... the outcome of his trial would have been different.”).

Although neither party has raised this issue before this court, this statement by the Michigan Court of Appeals is an incorrect statement of law because Strickland requires only that a petitioner demonstrate that "it is reasonably likely the result would have been different” not "that counsel’s actions more likely than not altered the outcome.” Richter, 131 S.Ct. at 792 (internal quotation marks omitted).

The Supreme Court in the past has remanded cases for determination whether a petitioner was prejudiced by counsel’s errors under the correct Strickland standard where the state court articulated the proper prejudice standard, but then deviated from it in application of Strickland to the case before it. See Sears, 130 S.Ct. at 3265-66 (concluding that the state court failed to apply the proper prejudice inquiry as the Supreme Court has "never limited the prejudice inquiry under Strickland to cases in which there was only ‘little or no mitigation evidence’ presented”).

The Supreme Court has recognized also, however, that "the difference between Strickland’s prejudice standard and a more-probable-than-not standard is slight and matters ‘only in the rarest case.’ ” Richter, 131 S.Ct. at 792. Accordingly, and because the issue has not been raised by the parties, we conclude that despite the Michigan court’s imprecise language, the court was considering Daniel’s Strickland claim under the proper standard that it first articulated in its opinion. See Holland v.

Jackson, 542 U.S. 649, 654-55, 124 S.Ct. 2736, 159 L.Ed.2d 683 (2004) ("As we explained in [Woodford v.] Visciotti[, 537 U.S. 19, 24, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002) ], § 2254(

d) requires that state-court decisions be given the benefit of the doubt' " because “[rjeadiness to attribute error is inconsistent with the presumption that state courts know and follow the law.” (internal quotation marks omitted)). . Moreover, not even Muex's and O’Neil’s statements were perfectly consistent. Muex's statement indicates that three boys approached the victim's vehicle after the shooting, R. 34-4 (Exh. 2-B, Muex Stmt.), while O'Neil’s statement indicated that there were only two boys who approached the vehicle, id. (Exh. 4, O’Neil Stmt.).