Young v. Conway

698 F.3d 69, 2012 WL 4876235, 2012 U.S. App. LEXIS 21502

• Court: Court of Appeals for the Second Circuit
• Decided: October 16, 2012
• Docket: Docket 11-830-pr
• Status: Published

Opinion

BARRINGTON D. PARKER, Circuit Judge:

The State of New York appeals from a judgment of the United States District Court for the Western District of New York (Bianchini, M.J.) granting defendant Rudolph Young’s petition for a writ of habeas corpus, vacating his convictions for robbery and burglary, and barring the State of New York from retrying him. 1 Young was convicted at his first trial in August 1993 based on the victim’s in-court identification and her testimony that she had identified him in a lineup held one month after the crime. He was the only member of the lineup whose picture had also been included in a photographic array shown to the victim two days earlier, when she failed to make an identification. After the lineup identification testimony was suppressed as the product of Young’s unconstitutional arrest under the Fourth Amendment, 2 at a second trial held almost six years later, the state trial court nonetheless permitted the victim to identify Young in court as the person who had broken into her home, based on its finding that her in-court identification had a basis independent of the tainted lineup. The New York courts affirmed Young’s convictions on direct appeal. Young filed a petition for habeas corpus arguing, inter alia, that the source of the victim’s in-court identification could not possibly have been independent of the tainted lineup. The district court agreed and granted the petition.

The State now appeals, arguing principally that, because Young had a full and fair opportunity to litigate his Fourth Amendment claim in state court, federal habeas relief is not available. See Stone v. Powell, 428 U.S. 465, 96 S.Ct. 3037, 49 L.Ed.2d 1067 (1976). Because the State failed to raise this non-jurisdictional argument in the district court, we decline in the exercise of our discretion to consider it for the first time on appeal. We agree with the district court that the state courts’ determination that the victim’s in-court identification derived from a source independent of a tainted lineup constituted an unreasonable application of, and was contrary to, clearly established Supreme Court law. See United States v. Crews, 445 U.S. 463, 472-74, 100 S.Ct. 1244, 63 L.Ed.2d 537 (1980); United States v. Wade, 388 U.S. 218, 240-41, 87 S.Ct. 1926, 18 L.Ed.2d 1149 (1967). Finally, although we conclude that the in-court identification substantially and injuriously influenced the jury’s deliberations, and while we share the district court’s “grave doubts whether th[e] circumstantial evidence was ... legally sufficient to convict Young” without it, Young v. Conway, 761 F.Supp.2d 59, 76 (W.D.N.Y.2011), such doubts must be resolved if at all by the state court, not ours. Accordingly, we vacate that portion of the district court’s judgment barring the State from retrying Young, but affirm its vacatur of Young’s convictions.

BACKGROUND

The crimes for which Young was convicted occurred on March 29, 1991. That evening, an intruder entered the home of William and Lisa Sykes carrying an axe and sledgehammer and wearing a blanket draped over his clothes. He wore a scarf around his mouth that covered his lips, nose, ears, and cheeks, leaving only his eyes and the top of his head uncovered. Brandishing the axe over Mr. Sykes’s head, the intruder demanded money and then took some watches from the bedroom. The Sykeses later reported that a pair of binoculars, a red bicycle, a mirror, and a pair of workout gloves from Mrs. Sykes’s car were also missing.

The intruder was in the house for approximately five to seven minutes. After he left, Mr. Sykes immediately called the police. In the police report taken later that evening, which she signed, Mrs. Sykes, who is white, described the intruder as “[a] black man in his twenties, five-ten, medium build.” App. at 131. 3 When the police asked if she could assist in preparing a composite sketch of the intruder, Mrs. Sykes replied that she could not. Id.

Approximately one month later, police showed Mrs. Sykes a photographic array containing six full-color photographs, including one of Young’s entire face. She could not at that time identify Young as the intruder. 4 The next day, Mr. Sykes viewed the same array in his home — with Mrs. Sykes present — but also failed to make an identification.

The next day, Young was arrested and placed in a lineup that Mr. and Mrs. Sykes viewed separately. As noted, the Appellate Division, Fourth Department, subsequently held that there was no probable cause for the arrest. See Young, 202 A.D.2d at 1026, 609 N.Y.S.2d at 726-27. Of the lineup participants, Young was the only person whose picture had been included in the photo array viewed by the Sykes-es. The lineup participants — none of whom wore scarves around their faces or blankets over their bodies — each stepped forward and said three things that the intruder had allegedly said the night of the crime. Mr. Sykes did not identify Young. Instead, he said the voice of a different lineup participant sounded most like the intruder, while the eyes and face of yet another lineup participant most resembled him. Mrs. Sykes, however, identified Young based just on “his eyes and the voice.” App. at 161.

Young was indicted for burglary and two counts of robbery and went to trial in August 1993. At trial, Mrs. Sykes identified Young as “that person that [she] identified at the lineup.” 1993 Trial Tr. (“Trial I”) 57. She later testified that she made this identification based on a “combination” of factors “from seeing him and also the voice.” App. at 73. Due largely to Mrs. Sykes’s in-court identification, which stemmed from the prior lineup, Young was convicted. See 761 F.Supp.2d at 77.

Young’s conviction was reversed on appeal. The Appellate Division concluded that, because the police had obtained Young’s consent to the lineup “by means affected by the primary taint [of his illegal arrest],” and because “the line-up identification flowed directly from the illegal arrest and was not attenuated therefrom,” Mrs. Sykes’s testimony concerning the lineup should not have been admitted at trial. People v. Young, 255 A.D.2d 905, 683 N.Y.S.2d 677, 678 (1998). The court ordered a new trial and provided the prosecution with an opportunity to prove that Mrs. Sykes had “a basis independent of the unlawful arrest and tainted identification procedure” to identify Young in court. Id.

In March 1999 — eight years after the initial incident at the Sykeses’ residence— the trial court held an independent source hearing to determine whether Mrs. Sykes would be permitted to make an in-court identification of Young at a re-trial. At that hearing, Mrs. Sykes described the robbery in detail and testified as to why she remembered the intruder from her observations of him during the crime rather than from the tainted lineup. According to her testimony, she first encountered the intruder standing “a couple of feet” away from her in her well-lit dining room. App. at 117. Mrs. Sykes, who. is “five-eight, fine-nine,” testified that she was “[essentially ... looking at [the intruder] face-to-face,” and therefore estimated his relative height to be 5'10". Id. Young is black, is six feet tall and was almost 34 at the time of the incident. Mrs. Sykes said the intruder had a “[s]carf over his face and a blanket covering all his clothes.” Id. at 131. The scarf “covered up the intruder’s mouth[,] chin,” ears and the “majority of the intruder’s cheeks and jawbone.” Id. at 136-138. Mrs. Sykes did not recall whether the intruder had a beard or mustache, or the length or kind of hair on his head; whether he was muscular or had any “noticeable or distinct physical characteristics,” id. at 138; whether he wore a hat or any jewelry; what type of shoes or pants he wore; whether he was wearing a jacket underneath the blanket; or whether he was wearing gloves.

After initially screaming, Mrs. Sykes said she looked carefully at the intruder, in disbelief that the incident was not a prank, but realized after staring at his eyes that she did not know him. The burglar then walked directly behind Mr. Sykes, brandished the axe over his head, looked directly at Mrs. Sykes, and said, “I will kill him. Give me your wallets.” Id. at 118 (quotation marks omitted).

The three then walked down the hallway to the master bedroom to retrieve Mr. Sykes’s wallet. Mrs. Sykes turned on the hall light and the intruder turned and looked right at her. She continued to watch him as he relieved Mr. Sykes of his money and took the watches. After next demanding that Mrs. Sykes give him her money, he proceeded to rip two telephones out of the walls, instructing Mrs. Sykes not to “look at [his] face.” Id. at 123. The intruder left the house shortly thereafter.

Mrs. Sykes testified that, up until the point the intruder instructed her to avert her eyes, she had continuously looked at his face, primarily his eyes, trying to determine who he was. She said that for many nights after the crime, she woke up seeing the intruder’s eyes in her nightmares. However, she also testified that there was “[n]othing unusual that stood out” about them. Id. at 140. She further acknowledged that she was unable to assist the police in sketching a composite drawing of the intruder’s face and stated that, although she had examined the eyes of each of the faces in the photo array conducted a month after the incident, she was unable to select Young because “a photograph is not a real person” and the pictures in the array did not “look real” to her. Id. at 155.

As to the lineup, Mrs. Sykes testified that she selected Young based solely on his eyes and voice. She contended that, eight years after the incident, she could “completely excise that lineup from [her] mind” and make an in-person identification based solely on the “eyes and voice.” Id. at 168. Based on this testimony, the court found that Mrs. Sykes had demonstrated the ability to make an in-court identification at Young’s subsequent retrial. In so holding, the court stressed that Mrs. Sykes “was in close proximity” to the intruder and “had ample opportunity” to see him. Id. at 195. Although “much of her identification focused on [the intruder’s] eyes,” her certainty helped establish her ability to make an in-court identification. Id.

At Young’s retrial in January 2000, the prosecution offered no physical evidence linking Young to the robbery. Instead, its case-in-chief consisted almost entirely of testimony from the Sykeses and from two other individuals, Taunja Isaac and Nell Kimbrel. The Sykeses’ description of the robbery at trial was largely consistent with Mrs. Sykes’s description at the independent source hearing. She identified Young as the perpetrator. Mr. Sykes did not identify him. The court denied Young’s motion to introduce expert testimony on the ability of a person (1) to make an identification independent of an earlier, tainted lineup identification or (2) generally to make an accurate identification given various factors.

The prosecution’s third primary witness, Isaac, was an acquaintance of Young’s and a convicted felon with a lengthy criminal record. Isaac testified that on April 29th — a month after the Sykes burglary and in connection with Isaac’s arrest for disorderly conduct — she informed the police about a pair of binoculars and three watches that Young had “asked [her] to sell” some time in March or April of 1991. 2000 Trial Tr. (“Trial II”) 131. When asked by a sheriffs investigator to retrieve the items, Isaac was able to recover only the binoculars, which were introduced into evidence. Mr. Sykes identified them as those stolen from his car. A sheriffs investigator testified that he never directed Isaac to show him where or from whom she had gotten them.

Isaac admitted at trial that she “had a very good and close friendship” with Lamont Gordon, another man who “lived around [her] neighborhood” and whom she knew the police to be investigating as a suspect in the Sykes robbery. Id. at 149-150. Gordon was a “[l]ight skinned” African-American male, “about six feet,” of “medium build,” and “between 18 and 20” at the time of the incident. Id. at 159. Isaac further testified that at the time of her arrest she feared she would be charged and sent to jail. She never was.

The fourth primary prosecution witness, Kimbrel, testified that a month after the incident the police found a pair of gloves belonging to the Sykeses in a garbage can at her residence — a house from which as many as 30 people a night would “be in and out” to “smoke cocaine.” Id. at 272-73. Isaac, from whom Kimbrel regularly bought cocaine, and Young were two such people. According to Kimbrel’s testimony, Young had been staying at her house at least once a week and leaving some of his belongings there during the months of March and April 1991, during which time Kimbrel was “constantly [and] continually” using cocaine. Id. at 286. Kimbrel, who had prior felony convictions for larceny and possession of stolen property, could say only that the gloves “had been [in her house] for a while” and that she had never actually seen Young with the gloves. Id. at 261. She testified that “[a]t any point someone could have brought the gloves in [to her house] because [she] had a lot of peoples in the house.” Id. at 274. Kimbrel had discarded the gloves, which were admitted at trial, along with a number of other items upon learning that the police were coming to execute a search warrant.

The defense introduced the testimony of a police officer establishing that Young was 6'0" to 6'1" tall and would have been 34 years old at the time of the incident.

Young was convicted on all three counts and sentenced to two consecutive terms of fifteen years to life imprisonment.

Direct Appeal

The Appellate Division, Fourth Department, affirmed the convictions. People v. Young, 20 A.D.3d 893, 798 N.Y.S.2d 625, 625-26 (2005). Addressing Young’s challenge to Mrs. Sykes’s in-court identification, a majority of the five-justice panel held that the trial court had “properly determined that the People proved by clear and convincing evidence that the victim had an independent basis for her in-court identification” of Young. Id. at 625-26. Two justices dissented on the grounds that

[t]he inability of the victim to assist the police in constructing a composite of the intruder and her inability to select defendant from a photo array prior to the lineup identification procedure strongly suggest that her alleged independent “recollection” of defendant was irrevocably tainted by her having viewed defendant in the lineup and having heard him speak. We therefore must conclude that any in-court identification testimony by the ■ victim would be derived from exploitation of the illegal arrest.

Id. at 627 (Hurlbutt, J.P., and Gorski, J., dissenting) (quotation marks and citations omitted).

A divided panel of the Court of Appeals affirmed. People v. Young, 7 N.Y.3d 40, 46, 817 N.Y.S.2d 576, 850 N.E.2d 623 (2006). The majority declined to “disturb[]” the lower courts’ determination that there was an independent basis for Mrs. Sykes’s identification testimony, an “issue of fact” for which it found “support in the record.” Id. at 44, 817 N.Y.S.2d 576, 850 N.E.2d 623. A dissenting judge concluded that “any in-court identification [was] impermissible as a matter of law.” Id. at 48, 817 N.Y.S.2d 576, 850 N.E.2d 623 (G.B. Smith, J., dissenting). 5

Proceedings Below

Young timely filed a petition for a writ of habeas corpus alleging, inter alia, that the state courts’ conclusion that Mrs. Sykes had an independent source for her in-court identification of Young was an unreasonable application of clearly established federal law. See Wade, 388 U.S. at 241, 87 S.Ct. 1926 (listing six factors courts should consider to determine whether witness has independent basis for in-court identification); see also Crews, 445 U.S. at 472-73 & n. 18, 100 S.Ct. 1244. The State opposed Young’s petition on the merits, but never argued, as it does on appeal, that Stone barred consideration of Young’s claim.

After “[reviewing the Wade factors,” the district court concluded that “all of them are on Petitioner’s side of the scale, and none of them are on the government’s.” 761 F.Supp.2d at 75. Accordingly, the district court found the New York courts’ determination that Mrs. Sykes’s in-court identification stemmed from an “independent source” to be an unreasonable application of Wade. Id. at 76. Left with “no doubt ... that the [trial court’s] error [in admitting the identification testimony] influenced the jury’s deliberations in a way that was substantial and injurious,” the court ordered Young’s convictions vacated. Id. at 77 (quotation marks and citations omitted). Finally, because it believed “the prosecution would not be able to secure a conviction based upon legally sufficient evidence so as to satisfy due process concerns,” the court also ordered “the extraordinary remedy of precluding the prosecution from retrying Young on the charges stemming from the Sykes home invasion.” Id. at 83. This appeal followed.

We review a district court’s grant of habeas relief de novo, and the underlying findings of fact for clear error. See Ramchair v. Conway, 601 F.3d 66, 72 (2d Cir.2010).

DISCUSSION

The State presents scant argument on appeal challenging the district court’s conclusion that the state court unreasonably applied Wade. Its principal contention is that, because Young had a full and fair opportunity to litigate his “Fourth Amendment claim” in state court, federal habeas corpus provides him no remedy. See Stone, 428 U.S. at 494, 96 S.Ct. 3037. Because the State failed to raise this non-jurisdictional argument below, we decline to consider it. See infra Part II. We further hold that the district court properly concluded that the state courts’ determination, that Mrs. Sykes’s testimony was properly admitted as having a source independent of the tainted lineup, constituted an unreasonable application of, and was contrary to, clearly established Supreme Court law, see Crews, 445 U.S. at 472-74, 100 S.Ct. 1244; Wade, 388 U.S. at 241, 87 S.Ct. 1926. See infra Part I. Finally, we hold, that the in-court identification substantially and injuriously influenced the jury’s deliberations, warranting vacatur of Young’s convictions. See infra Part III.

I.

Whether an in-court identification has a source independent of an earlier tainted identification is a mixed question of law and fact. See Wade, 388 U.S. at 241-42, 87 S.Ct. 1926. “On federal habeas review, mixed questions of law and fact translate to ‘mixed constitutional questions (ie., application of constitutional law to fact).’ ” Overton v. Newton, 295 F.3d 270, 277 (2d Cir.2002) (quoting Williams v. Taylor, 529 U.S. 362, 400, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000) (O’Connor, J., concurring)). Under the Anti-Terrorism and Effective Death Penalty Act of 1996 (AEDPA), such questions “are subject to the standard set forth in 28 U.S.C. § 2254(d)(1), which requires the habeas court to determine whether the state court’s decision ‘was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.’ ” Id. (quoting 28 U.S.C. § 2254(d)(1)).

The exclusionary rule applies not only to the “direct products” of unconstitutional invasions of defendants’ Fourth Amendment rights, but also to the indirect or derivative “fruits” of those invasions. Crews, 445 U.S. at 470, 100 S.Ct. 1244 (quotation marks and citations omitted) (citing Wong Sun v. United States, 371 U.S. 471, 484, 83 S.Ct. 407, 9 L.Ed.2d 441 (1963)). Whether such fruit is “of the poisonous tree” — in which case it must be excluded at trial — depends on “whether the chain of causation proceeding from the unlawful conduct has become so attenuated or has been interrupted by some intervening circumstance so as to remove the taint imposed upon that evidence by the original illegality.” Id. at 471, 83 S.Ct. 407 (quotation marks and citations omitted). In the case of unconstitutional arrests, any of three distinct elements of in-court identifications may “ha[ve] been come at by exploitation of the violation of the defendant’s Fourth Amendment rights”: (1) the victim’s presence at trial to identify the defendant; (2) the defendant’s presence at trial so that he can be identified; and (3) the victim’s “knowledge of and [ ] ability to reconstruct the prior criminal occurrence and to identify the defendant from her observations of him at the time of the crime.” Id. (quotation marks and citations omitted).

As to the third element, “given the vagaries of human memory and the inherent suggestibility of many identification procedures,” “intervening photographic and lineup identifications — both of which are conceded to be suppressible fruits of the Fourth Amendment violation — [may] affect the reliability of the in-court identification and render it inadmissible as well.” Id. at 472, 83 S.Ct. 407 (citing Patrick M. Wall, Eye-Witness Identification in Criminal Cases 40-64 (1965); Frederic D. Woocher, Note, Did Your Eyes Deceive You? Expert Psychological Testimony on the Unreliability of Eyewitness Identification, 29 Stan. L.Rev. 969, 974-89 (1977)).

[I]t is a matter of common experience that, once a witness has picked out the accused at the line-up, he is not likely to go back on his word later on, so that in practice the issue of identity may (in the absence of other relevant evidence) for all practical purposes be determined there and then, before the trial.

Wade, 388 U.S. at 229, 87 S.Ct. 1926 (quotation marks and citations omitted) (citing Glanville Williams & H.A. Hammelmann, Identification Parades, Part I, Crim. L.Rev. 479, 482 (1963)). Indeed, social science research indicates that false identification rates increase, and accuracy on the whole decreases, when there are multiple identification procedures. E.g., Ryan D. Godfrey & Steven E. Clark, Repeated Eyewitness Identification Procedures: Memory, Decision Making, and Probative Value, 34 Law & Hum. Behav. 241, 241, 256 (2010) (explaining this effect as the result either of “misplaced familiarity due to the memory of the suspect” from earlier identification or of “heightened expectations and suggestiveness”).

Whether an in-court identification is nevertheless admissible depends on whether it has an origin “independent” of the tainted lineup, Wade, 388 U.S. at 242, 87 S.Ct. 1926 — whether it “rested on an independent recollection of [the victim’s] initial encounter with the assailant, uninfluenced by the pretrial identifications,” Crews, 445 U.S. at 473, 100 S.Ct. 1244. In answering that question, courts consider the following factors: (1) the victim’s prior opportunity to observe the alleged criminal act; (2) any discrepancy between any prelineup description and the defendant’s actual description; (3) any identification pri- or to the lineup of another person; (4) a photographic identification of the defendant prior to the illegal lineup; (5) the victim’s failure to identify the defendant on a prior occasion; and (6) the lapse of time between the alleged act and the lineup identification. 6 Wade, 388 U.S. at 241, 87 S.Ct. 1926; see also Crews, 445 U.S. at 473 nn. 18-19, 100 S.Ct. 1244 (citing with approval and applying Wade factors “in the context of [a] Fourth Amendment violation” to determine whether in-court identification had origins independent of earlier identification tainted by arrest lacking probable cause (quotation marks omitted)). It “is also relevant” to consider facts “disclosed concerning the conduct of the lineup.” Wade, 388 U.S. at 241, 87 S.Ct. 1926. The State bears the burden of proving, by clear and convincing evidence, that the in-court identification was “based upon observations of the suspect other than the [tainted] lineup identification.” Id. at 240, 87 S.Ct. 1926.

We turn now to an analysis of Mrs. Sykes’s in-court identification, considering whether the Wade factors could reasonably support a finding by clear and convincing evidence that it had an independent basis. In the course of assessing those factors, we reference an extensive body of scientific literature presented to us by amicus curiae the Innocence Project in support of affirmance. 7 That literature in dicates that certain circumstances surrounding a crime — including the perpetrator’s wearing a disguise, the presence of a weapon, the stress of the situation, the cross-racial nature of the crime, the passage of time between observation and identification, and the witness’s exposure to defendant through multiple identification procedures' — may impair the ability of a witness, such as Mrs. Sykes, to accurately process what she observed. Many of these factors are counterintuitive and, therefore, not coterminous with “common sense.” We note that the research presented to us by the Innocence Project — to which the State has offered no response 8 —has been reviewed, replicated, and retested, and is generally accepted in the research community. See State v. Henderson, 208 N.J. 208, 283, 27 A.3d 872 (2011) (reviewing comprehensive 86-page report of court-appointed Special Master based on seven experts’ testimony and hundreds of scientific studies, including many of those cited by the Innocence Project and by us below; referring to such research as “the gold standard in terms of the applicability of social science research to the law” (quotation marks and citations omitted)). As a result, as the New Jersey Supreme Court has pointed out, “social scientists, forensic experts, law enforcement agencies, law reform groups, legislatures and courts” routinely rely upon the research during legal proceedings regarding eyewitness testimony. Report of the Special Master at 73, Henderson, 208 N.J. 208, 27 A.3d 872. 9

Indeed, this Court has previously approved lower courts’ drawing upon this body of literature in appropriate cases. In United States v. Luis, for example, we endorsed a “flexible approach” to trial judges’ use of “a specific eyewitness charge in order to ameliorate the concerns expressed in ... [Wade ] relating to the dangers inherent in eyewitness testimony that may lead to misidentification.” 835 F.2d 37, 41 (2d Cir.1987); see also United States v. Sema, 799 F.2d 842, 850 (2d Cir.1986) (noting our “full[ ] aware[ness] of the dangers of testimony based purely on eyewitness identification,” dangers on which we have “often commented”), abrogated on other grounds by, United States v. DiNapoli, 8 F.3d 909, 914 n. 5 (2d Cir.1993). In United States v. Veal, we affirmed the district court’s exclusion of expert testimony on eyewitness identification in favor of “a very detailed instruction to the jury regarding the evaluation of eyewitness identification evidence,” observing that district courts “may properly address the dangers of unreliable eyewitness identification testimony by giving a jury charge appropriate to the circum stances of the case.” No. 98-1539, .1999 WL 446783, at *1 (2d Cir. June 16, 1999) (citing Luis, 835 F.2d at 41). Trial courts should continue to draw upon this literature to the extent they deem it helpful in fashioning such charges. That said, we caution that much eyewitness identification testimony is reliable and is, and should be, routinely accepted by juries. See, e.g., Luis, 835 F.2d at 42 (finding no abuse of discretion in district court’s refusal to give defense counsel’s requested eyewitness identification charge in part because “[t]he[ ] circumstances provide strong indicia that the identification of [defendant] as a participant in the crime for which he was charged was reliable”).

The first and, we think, most important Wade factor concerns the victim’s prior opportunity to observe the alleged criminal act. Here, Mrs. Sykes testified that she observed the perpetrator for five to seven minutes in good lighting. However, she was afforded no meaningful opportunity to perceive him during that time given that his body was covered with a blanket and the entirety of his face below his eyes — including his lips, nose, ears, and cheeks — was covered by a scarf. And as for those eyes, “nothing unusual ... stood out ” about them. App. at 140 (emphasis added). Nor did the intruder have any other “noticeable or distinct physical characteristics.” Id. at 138. Indeed, following the incident, Mrs. Sykes could neither assist in a composite sketch nor recall any details of the perpetrator’s mouth, ears, forehead, facial hair, or hairstyle. Since then, she has been unable to articulate anything but the barest of details about him — his height, gender, and race — instead repeatedly emphasizing his wholly unremarkable and decontextualized eyes. In Raheem v. Kelly, 257 F.3d 122, 138 (2d Cir.2001), this Court found a witness’s testimony at a hearing insufficient to establish an independent basis for his in-court identification of an alleged shooter, where the witness described the shooter’s eyes as “weird” and “different” because of the shooter’s “glare” and the manner in which he “fixed his eyes”; there was “something about it that stood out.” Id. at 139. Here, Mrs. Sykes offered an even flimsier description and admitted that nothing unusual stood out about the perpetrator’s eyes. Thus we are not persuaded that her observation of them could serve as an independent basis for her in-court identification.

In assessing this first Wade factor, we find illuminating the social science research, presented by the Innocence Project, addressing the effect of disguises, weapons, stress, and cross-racial identifications on those identifications’ accuracy. First, as the amicus curiae observes, even “subtle disguises can ... impair identification accuracy.” Brian L. Cutler & Margaret Bull Kovera, Evaluating Eyewitness Identification 43 (2010); see also Brian L. Cutler et al., The Reliability of Eyewitness Identification: The Role of System and Estimator Variables, 11 Law & Hum. Behav. 233, 240, 244-45 (1987) (reporting results of experiment showing that when “perpetrator” wore a hat, only 27% of participants’ identifications were accurate, versus 45% when “perpetrator” did not wear a hat). Here, the perpetrator of the Sykes robbery wore no minimal disguise; as noted above, his body and face were almost entirely covered. That disguise deprived Mrs. Sykes of any meaningful opportunity to observe him; social science research suggests it likely also impaired her ability accurately to perceive the little she was able to observe.

Second, the scientific literature indicates that the presence of a weapon during a crime “will draw central attention, thus decreasing the ability of the eyewitness to adequately encode and later recall peripheral details.” Nancy Mehrkens Steblay, A Metcu-Analytic Review of the Weapon Focus Effect, 16 Law & Hum. Behav. 413,414 (1992). For example, an analysis of 19 weapon-focus studies involving 2082 identifications found that, on average, identification accuracy decreased approximately 10% when a weapon was present. Id. at 415-17. Our own Court has recognized that “it is human nature for a person toward whom a [weapon] is being pointed to focus h[er] attention more on the [weapon] than on the face of the person pointing it.” Raheem, 257 F.3d at 138. Here, the perpetrator entered the Sykeses’ home wielding an axe and a sledgehammer. Mrs. Sykes “observed” him while he brandished the axe over the head of her wheelchair-bound husband, threatening to kill him.

Third, high levels of stress have been shown to induce a defensive mental state that can result in a diminished ability accurately to process and recall events, leading to inaccurate identifications. Kenneth A. Deffenbacher et al., A Meta-Analytic Review of the Effects of High Stress on Eyewitness Memory, 28 Law & Hum. Behav. 687, 687, 699-700 (2004); see also Charles A. Morgan III et al., Accuracy of Eyewitness Memot'y for Persons Encountered During Exposure to Highly Intense Stress, 27 Int’l J.L. & Psychiatry 265 (2004). For example, a review of 16 studies involving 1727 participants found that accurate identifications decreased 22.2% under high stress conditions. Deffenbacher et al., A Meta-Analytic Review of the Effects of High Stress on Eyewitness Memory at 692, 694 (reporting that overall proportion of correct identifications for high stress conditions was 0.42 versus 0.54 for low stress conditions). Here, as if the stress of the situation were not already clear, Mrs. Sykes testified that she was “scared” and “petrified.” App. at 142.

Fourth, social science research indicates that people are significantly more prone to identification errors when trying to identify someone of a different race, a phenomenon known as “own-race bias.” “There is a considerable consistency across [scientific] studies, indicating that memory for own-race faces is superior to memory for other-race faces.” Robert K. Bothwell et al., Cross-Racial Identification, 15 Personality & Soc. Psychol. Bull. 19, 19, 23 (1989) (conducting meta-analysis of 14 studies finding that own-race bias effect “occurs for both Black and White subjects in 79% of the samples”). Studies have thus found a “tendency for people to exhibit better memory for faces of [members of their own race] than for faces of [members of another race].” Tara Anthony et al., Cross-Racial Facial Identification: A Social Cognitive Integration, 18 Personality & Soc. Psychol. Bull. 296, 299 (1992). Studies suggest that own-race bias is especially pronounced where, as here, the person making the identification is Caucasian and the person being identified is African-American. Henry F. Fradella, Why Judges Should Admit Expert Testimony on the Unreliability of Eyewitness Testimony, 2 Fed. Cts. L.Rev. 1, 14 (2007).

Before proceeding to the second Wade factor, we also consider the flip-side of the first one: whether Mrs. Sykes’s limited opportunity to observe the perpetrator at the crime scene influenced her subsequent perceived ability to identify Young, and whether those subsequent identifications in turn altered her perceived ability to draw on her limited observation of Young at the crime scene. After all, as we have seen, the in-court identification followed three instances in which Mrs. Sykes was exposed to the defendant subsequent to the crime: she viewed his picture in the photo array and his person both in the unconstitutional lineup and also during the first trial. Here, the Innocence Project has directed us to scientific research indicating that such prior identifications may taint subsequent in-court identifications due to a phenomenon known as the “mugshot exposure effect,” or “unconscious transference,” whereby a witness selects a person in a later identification procedure based on a sense of familiarity deriving from her exposure to him during a prior one. This phenomenon occurs because “the witness [is] unable to partition his or her memory in such a way as to know that the suspect’s increased familiarity is due to the exposure [in the photo array], rather than the suspect’s presence at the time of the crime.” Godfrey & Clark, at 242; see also People v. Santiago, 17 N.Y.3d 661, 673, 934 N.Y.S.2d 746, 958 N.E.2d 874 (2011) (recognizing unconscious transference); Henderson, 208 N.J. at 255 (“[Successive views of the same person can make it difficult to know whether the later identification stems from a memory of the original event or a memory of the earlier identification procedure.”). This phenomenon is especially pronounced where, as here, the witness initially makes no identification from a photo array, but then selects someone whose picture was included in the photo array during a later identification procedure. Godfrey & Clark, at 247. For example, an analysis of 17 experiments showed that while only 15% of witnesses made an incorrect identification when the suspects in the lineup were viewed for the first time in the lineup, 37% of the witnesses made an incorrect identification when they had seen a suspect in a prior mugshot. Kenneth A. Deffenbacher et ah, Mugshot Exposure Effects: Retroactive Interference, Mugshot Commitment, Source Confusion, and Unconscious Transference, 30 Law & Hum. Behav. 287, 299 (2006).

In addition, where, as here, a victim identifies the defendant during an identification procedure prior to her in-court identification, her memory can be tainted by the “mugshot commitment effect”: having identified that person as the perpetrator, she becomes attached to her prior identification. As a result, she is more likely to identify him again in a subsequent identification procedure, even if he is innocent. See Charles A. Goodsell et ah, Effects of Mugshot Commitment on Lineup Performance in Young and Older Adults, 23 Applied Cognitive Psychol. 788, 789 (2009). In one study, 72% of persons who made an inaccurate identification from a mugshot book later made the same mistaken identification in a lineup. Id. at 795. This phenomenon may be observed even when the actual culprit is present in the second identification procedure, and the previously selected innocent person is absent. For example, in one experiment, 60% of participants indicated that the actual culprit was not present in the lineup, while only 12% correctly identified him from the lineup. Id. at 798. This research demonstrates that “[m]ugshot choosers will select their prior mugshot choice if given the opportunity and will reject a lineup that does not contain it” even when the opportunity to select the actual culprit is available. Id. Moreover, where a witness has been primed with information supporting an erroneous identification — such as repeated exposure to Mr. Young’s face in photographs, the illegal lineup, and the 1993 trial — research suggests she is often more confident in her erroneous selection. See Elizabeth Loftus, Semantic Integration of Verbal Information into a Visual Memory, 4 J. Experimental Psych.: Hum. Learning And Memory 19-31 (1978).

Here, our analysis of Mrs. Sykes’s ability to identify Young independent of the tainted lineup is also informed by the possibility that her in-court identification was due to her prior choice of Young during the illegal lineup, which in turn may likely have been due to her prior exposure to his picture in the photo array. Combined with Mrs. Sykes’s minimal opportunity to observe the perpetrator during the commission of a stressful, violence-threatening, cross-racial crime, Mrs. Sykes’s multiple exposures to and identifications of Young thereafter strongly suggests that there could be no independent basis for Mrs. Sykes’s identification. Nevertheless, we examine the remaining Wade factors.

* * *

The second Wade factor is the existence of any discrepancy between any pre-lineup description and the defendant’s actual description. Immediately after the robbery, Mrs. Sykes described the perpetrator as “[a] black man in his twenties, five-ten, medium build,” App. at 131; Young, by contrast, is a six-foot tall African-American who, at the time of the robbery, was almost 34 years old. The two-inch height differential between Mrs. Sykes’s estimated and Young’s actual height is significant for at least two reasons: first, the difference between 5'10" and six feet is the difference between a man of average height and a tall man. See Cynthia L. Ogden, et al., Centers for Disease Control, Mean Body Weight, Height, and Body Mass Index, United States, 1960-2002 (2004) at 15, available at http://www.cdc. gov/nchs/data/ad/ad347.pdf. Second, and more important, Mrs. Sykes estimated the perpetrator’s height in relative terms — in relation to her own height of 5'8" or 5'9"— based on her ability essentially to stand “face-to-face” with him. If Young was the perpetrator, there would have been a three-to four-inch height differential between them, rendering a description of their encounter as “face-to-face” unlikely. Even leaving aside the divergence of Mrs. Sykes’s description from Young’s actual appearance, Mrs. Sykes’s post-incident description does not “instill any confidence as to the reliability of [her] identificationf ] of [Young] as the [robber] independently of the [tainted lineup], for though [she] provided general information as to the [robber’s] age, height, and weight, [she] provided virtually no detail about his face.” Raheem, 257 F.3d at 138. Therefore, this factor also weighs against the State.

The third, fourth, and fifth Wade factors probe whether the witness, prior to the lineup, identified another person; whether the witness identified the defendant by photograph prior to the lineup; and whether the witness failed to identify the defendant on a prior occasion. Here, Mrs. Sykes has offered differing testimony as to whether, from the photo array conducted one month after the crime, she selected someone other than Young as the perpetrator or failed to identify anyone at all. Either way, her inability to identify Young goes to the third through fifth Wade factors and, for each, counsels against concluding that Mrs. Sykes had an independent basis for her in-court identification of Young. In Crews, by contrast — where the in-court identification was found to have an independent basis — the “the victim failed to identify anyone other than respondent, ... [and] twice selected respondent without hesitation in nonsuggestive pretrial identification procedures.” 445 U.S. at 473 n. 18,100 S.Ct. 1244.

Moreover, as noted above, Mrs. Sykes’s failure to identify Young in the photo array, followed by her “successful” identification of him at the lineup, not only suggests that the robbery itself may have provided an inadequate basis for her in-court identification; but in light of the fact that Young was the only lineup participant whose picture was also included in the photo array, it also suggests that the photo array itself may have been the only basis for the very lineup identification suppressed as the fruit of Young’s unlawful arrest.

The sixth and final Wade factor is “the lapse of time between the alleged act and the lineup identification.” Wade, 388 U.S. at 241, 87 S.Ct. 1926. As the amicus curiae observes, research indicates that the passage of time both degrades correct memories and heightens confidence in incorrect ones. See Kenneth A. Deffenbacher et al., Forgetting the Once-Seen Face: Estimating the Strength of an Eyewitness’s Memory Representation, 14 J. Experimental Psychol.: Applied 139, 147-48 (2008). The Supreme Court has stated that a delay of seven months between a crime and a pre-trial identification is “a seriously negative factor” weighing against independent reliability “in most cases.” Neil, 409 U.S. at 201, 93 S.Ct. 375. According to studies, even a one-week delay can cause the “typical eyewitness viewing a perpetrator’s face that [is] not highly distinctive ... to have no more than a 50% chance of being correct in his or her lineup identification.” Deffenbacher et al., Forgetting, at 147.

Here, over one month passed between the robbery and Mrs. Sykes’s identification of Young at the suppressed lineup. Early in this intervening one month — a not inconsiderable period of time — Mrs. Sykes failed to identify him (or worse, identified someone else) while being shown his photograph only one week after the alleged act. Leaving the photo array aside for the moment, the month-long delay between the robbery and confrontation weighs in favor of the accused when the circumstances of the crime and the prior description given by the witness indicate nothing distinctive about the alleged perpetrator or about the witness’s ability to perceive anything distinctive about him. There is no basis in the record to conclude that the reliability of Mrs. Sykes’s identification somehow strengthened over the course of this month, or that she remembered details she had forgotten during the first photo array. The same is true of the year that passed between the robbery and her identification of Young at his first trial and the eight years separating the robbery from the independent source hearing. See Raheem, 257 F.3d at 139 (concluding that length of time between crime and confrontation did not suggest reliability where witnesses made wrong identifications of gunman from photo array conducted less than one week after event, three weeks passed between crime and their selection of defendant at improper lineup, and more than five years passed between crime and in-court identification, during which time this Court “[saw] nothing to suggest that those witnesses’ identifications of [defendant] became more reliable”).

Finally, we note that using the suppressed lineup as the second end-point makes sense only when the lineup is “non-suggestive.” Crews, 445 U.S. at 473 n. 18, 100 S.Ct. 1244; see also Wade, 388 U.S. at 241, 87 S.Ct. 1926 (directing courts to consider facts disclosed “concerning the conduct of the lineup”). But here, as we observed earlier, Mrs. Sykes may have been influenced by her exposure to Young’s photograph in the photo array. For all of these reasons, the sixth factor weighs significantly in favor of the petitioner.

Because all six Wade factors weigh against a finding that Mrs. Sykes’s in-court identification could derive from a source other than the tainted lineup, the State failed to meet its burden to prove an independent basis by clear and convincing evidence. The Court of Appeals’ determination otherwise was based on its mistaken impression that the independent source inquiry was an “issue of fact” to be resolved by the trial court and upheld on appeal so long as there was “support in the record.” Young, 7 N.Y.3d at 44, 817 N.Y.S.2d 576, 850 N.E.2d 623. Thus, not only did the Court of Appeals apply the wrong legal standard, but also its conclusion constituted an unreasonable application of the correct standard (Crews and Wade) to the facts of this particular case. Bell, 535 U.S. at 694, 122 S.Ct. 1843. Its conclusion was therefore both contrary to, and an unreasonable application of, Supreme Court law. Id.

II.

The State’s primary argument on appeal is that we are barred by Stone from considering Young’s claim and, notwithstanding the State’s failure to raise the issue below, we should consider it on appeal. We hold that, because the Stone rule is non-jurisdictional, it is waivable by the State, and we decline in the exercise of our discretion to consider it on appeal.

In Stone, the Supreme Court held that “where the State has provided an opportunity for full and fair litigation of a Fourth Amendment claim, a state prisoner may not be granted federal habeas corpus relief on the ground that evidence obtained in an unconstitutional search or seizure was introduced at his trial.” 428 U.S. at 494, 96 S.Ct. 3037. The Court grounded its holding in the fact that the exclusionary rule is merely “a judicially created means of effectuating the rights secured by the Fourth Amendment” — not a constitutional right personal to the defendant. Id. at 482, 96 S.Ct. 3037. Because “in the case of a typical Fourth Amendment claim, asserted on collateral attack, a convicted defendant is usually asking society to redetermine an issue that has no bearing on the basic justice of his incarceration,” habeas relief is unavailable unless the defendant was denied an opportunity for full and fair litigation in state court. Id. at 491 n. 31, 96 S.Ct. 3037.

However, the Supreme Court has made it clear that Stone’s limitation on federal habeas relief is not jurisdictional. Withrow v. Williams, 507 U.S. 680, 686, 113 S.Ct. 1745, 123 L.Ed.2d 407 (1993) (collecting cases discussing Stone as prudential and equitable rather than jurisdictional in nature). It is well-settled that non-jurisdictional arguments and defenses may be waived, and that we have the discretion — but are not required — to consider them on appeal. E.g., Gonzalez v. Thaler, — U.S. -, 132 S.Ct. 641, 648, 181 L.Ed.2d 619 (2012); Bowles v. Russell, 551 U.S. 205, 216-17, 127 S.Ct. 2360, 168 L.Ed.2d 96 (2007) (Souter, J., dissenting); Sharkey v. Quarantillo, 541 F.3d 75, 88 (2d Cir.2008).

The State nonetheless contends that Stone is jurisdictional because it “made [habeas] relief categorically unavailable by collateral review.” Resp.’s Reply Br. 3. There is no merit to this contention; Stone itself refutes it. In responding to Justice Brennan’s dissent that Stone “la[id] the groundwork for a ‘drastic withdrawal of federal habeas jurisdiction,’” the Stone majority explained that it held

only that a federal court need not apply the exclusionary rule on habeas review of a Fourth Amendment claim absent a showing that the state prisoner was denied an opportunity for a full and fair litigation of that claim at trial and on direct review. Our decision does not mean that the federal court lacks jurisdiction over such a claim.

428 U.S. at 494 n. 37, 96 S.Ct. 3037 (quoting id. at 517, 96 S.Ct. 3037 (Brennan, J., dissenting)) (emphasis added). 10 Indeed, the Supreme Court has held, in the highly analogous circumstances in which a state fails to assert procedural default, exhaustion, or non-retroactivity under Teague v. Lane, 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989), in opposition to habeas petitions — none of which is jurisdictional— that the courts of appeal have the discretion, but are by no means required, to address those defenses for the first time on appeal. 11

Here, despite four years and numerous opportunities to do so, the State never raised Stone and the record is bereft of any reason as to why it failed to do so. See Boardman v. Estelle, 957 F.2d 1523, 1537 (9th Cir.1992) (“The Supreme Court has enforced strict procedural forfeitures on habeas petitioners in the interests of efficient and final adjudication. Why should not the state be similarly held to a pedestrian rule of appellate procedure? Concerns of federalism and respect for a state’s criminal judgments are marginal here because the state brought the problem on itself.”). Moreover, indulging that argument would require a remand to afford Young the opportunity to argue either why Stone does not apply or why he did not have a full and fair opportunity to litigate his claim. On a petition that has been pending for over five years, that would constitute an unjustifiable waste of scarce judicial resources, undermining the comity and federalism concerns that also underlie Stone. See Withrow, 507 U.S. at 686-87, 113 S.Ct. 1745 (citing Stone, 428 U.S. at 491 n. 31, 96 S.Ct. 3037); cf. Agard v. Portuondo, 159 F.3d 98, 100 (2d Cir.1998) (“Teague itself is driven in part by ... comity[,] [b]ut comity also calls for representatives of states not to agree to federal courtsf] expending substantial time in addressing the merits of a case, only to argue belatedly that the merits should not have been reached.”), rev’d on other grounds, 529 U.S. 61, 120 S.Ct. 1119, 146 L.Ed.2d 47 (2000).

What is more, it is difficult to imagine a case further afield from the “typical Fourth Amendment claim, asserted on collateral attack” than this one, where the issue Young is “asking society to redetermine” has everything to do with “the basic justice of his incarceration.” Stone, 428 U.S. at 491 n. 31, 96 S.Ct. 3037. At issue is the admission of evidence that, in this case, lacks the typical indicia of reliability that ordinarily weigh against re-litigating a Fourth Amendment claim on collateral review. See Stone, 428 U.S. at 490, 96 S.Ct. 3037 (observing that “ordinarily the evidence [sought to be suppressed] ... establishes beyond virtually any shadow of a doubt that the defendant is guilty” (quotation marks and citations omitted)). It also requires vacatur of his convictions. See infra, Part III. For all of these reasons, we exercise our discretion not to consider the State’s argument that Stone bars habeas relief here.

III.

We turn now to the question of whether Young was prejudiced by the improper admission of Mrs. Sykes’s identification testimony. When, as here, there is no state court holding to which AEDPA deference applies, a federal court in habeas must determine whether a state court’s error in admitting identification testimony had a “substantial and injurious effect or influence in determining the jury’s verdict.” Brecht v. Abrahamson, 507 U.S. 619, 623, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993); see also Fry v. Pliler, 551 U.S. 112, 121, 127 S.Ct. 2321, 168 L.Ed.2d 16 (2007). “The principal factors to be considered ... are the importance of the witness’s wrongly admitted testimony and the overall strength of the prosecution’s case.” Raheem, 257 F.3d at 142 (quotation marks and citations omitted). In assessing importance, we consider whether the testimony bore on an issue that was critical to the jury’s decision; whether it was material to the establishment of the critical fact or whether it was instead corroborated and cumulative; and whether the wrongly admitted evidence was emphasized in arguments to the jury. Id. (citing Wray v. Johnson, 202 F.3d 515, 526 (2d Cir.2000)).

Here, Mrs. Sykes’s identification testimony clearly bore on an essential and critical issue: the identity of the robber. Her testimony was also crucial to the prosecution’s case. The only other evidence tying Young to the robbery was Isaac’s testimony that the items found in her home, matching the description of items stolen from the Sykeses’ house, were given to her by Young to sell. Leaving aside the question of whether Isaac’s testimony may have been influenced by a desire to protect Gordon — a close acquaintance whom she knew also to be a suspect in the Sykes robbery (and who better matched Mrs. Sykes’s physical description of the perpetrator) — Isaac provided no insight into how Young acquired those items to begin with. Meanwhile, Kimbrel could not link the gloves found at her residence — from which as many as 30 people came and went on a nightly basis — to Young, and the state offered no physical evidence that would have identified Young as the robber.

On these facts, we have little difficulty concluding that the admission of the unreliable in-court identification had a “substantial and injurious” influence on the jury’s deliberations. Here, we again find illuminating research presented by the In nocenee Project, indicating that identification evidence is “comparable to or more impactive than physical evidence ... and even sometimes [than] confession evidence.” Melissa Boyce et al., Belief of Eyewitness Identification Evidence, in 2 Handbook of Eyewitness Psychology: Memory for People 501, 505 (R.C.L. Lindsay et al. eds., 2007). Moreover, “[t]he existence of eyewitness identification evidence increases the perceived strength of the other evidence presented.” Boyce, at 505.

Studies also suggest that jurors tend to overestimate “the likely accuracy of eyewitness evidence,” John C. Brigham & Robert K. Bothwell, The Ability of Prospective Jurors to Estimate the Accuracy of Eyewitness Identifications, 7 Law & Hum. Behav. 19, 28 (1983), perhaps because they “rely heavily on eyewitness factors that are not good indicators of accuracy,” Tanja Rapus Benton et al., Has Eyewitness Testimony Research Penetrated the American Legal System?: A Synthesis of Case History, Juror Knowledge, and Expert Testimony, in 2 Handbook of Eyewitness Psychology: Memory for People 453, 484 (R.C.L. Lindsay et al. eds., 2007). Social scientists theorize that jurors do this, as the Innocence Project explains, because many of the scientific principles underlying the reliability of eyewitness testimony are counter-intuitive or do not comport with common sense. Michael R. Leippe, The Case for Expert Testimony About Eyewitness Memory, 1 Psychol. Pub. Pol’y & L. 909, 921 (1995). Whatever the cause, the effect is that jurors frequently cannot accurately discriminate between correct and mistaken eyewitnesses and rely on the testimony of mistaken eyewitnesses. Id. at 925.

The jurors may also erroneously have relied on certainty as an indicator of accuracy. 12 “[M]ock-juror studies have found that confidence has a major influence on mock-jurors’ assessments of witness credibility and verdicts.” Neil Brewer & Gary L. Wells, The Confidence-Accuracy Relationship in Eyewitness Identification: Effects of Lineup Instructions, Foil Similarity, and Target-Absent Base Rates, 12 J. Experimental Psychol.: Applied 11, 11 (2006). Yet scientific research suggests that “eyewitness confidence is a poor postdictor of accuracy.” Steven M. Smith et al., Postdictors of Eyewitness Errors: Can False Identifications Be Diagnosed?, 85 J. Applied Psychol. 542, 548 (2000). Because eyewitnesses sincerely believe their testimony and are often unaware of the factors that may have contaminated their memories, they are more likely to be certain about their testimony. See United States v. Bartlett, 567 F.3d 901, 906 (7th Cir.2009) (explaining that the “problem with eyewitness testimony is that witnesses who think they are identifying the wrongdoer — who are credible because they believe every word they utter on the stand — may be mistaken”). And because jurors confound certainty and accuracy, cross-examination is less likely to be effective in discrediting eyewitnesses. Jules Epstein, The Great Engine that Couldn’t: Science, Mistaken Identifications, and the Limits of Cross-Examination, 36 Stetson L.Rev. 727, 772 (2007); Henderson, 208 N.J. at 234-37, 27 A.3d 872; see also People v. LeGrand, 8 N.Y.3d 449, 458, 835 N.Y.S.2d 523, 867 N.E.2d 374 (2007) (noting that scientific research relating to correlation between confidence and accuracy, effect of post-event information on accuracy, and confidence malleability is “generally accepted by social scientists and psychologists working in the field”). This research only reinforces our independent determination that the improper admission of Mrs. Sykes’s uncorroborated identification testimony at Young’s trial had a substantial and injurious effect on the jury’s verdict.

CONCLUSION

For the reasons just stated, the judgment of the district court granting the writ and ordering Young’s convictions vacated is affirmed. Because Young is currently serving a term of 25 years to life on an unrelated charge, that judgment does not require his release. Although the State did not address (in any fashion) the district court’s “extraordinary remedy of precluding the prosecution from retrying Young,” Young, 761 F.Supp.2d at 83, we view the district court’s imposition of that stricture as premature, see DiSimone v. Phillips, 518 F.3d 124, 126-28 (2d Cir.2008) (vacating order barring retrial because petitioner had not argued that retrial would be barred by “double jeopardy, or would necessarily involve constitutionally insufficient evidence”). If the State seeks to retry Young (without, of course, the eyewitness identification of Mrs. Sykes), Young is free to argue in state court that the re-prosecution is barred.

1 . The parties consented below to disposition of the matter by a magistrate judge pursuant to 28 U.S.C. § 636(c)(1).

2 . The Appellate Division, Fourth Department, determined that Young was arrested without probable cause. See People v. Young, 202 A.D.2d 1024, 609 N.Y.S.2d 725, 726-27 (1994).

3 . At a hearing held eight years later, Mrs. Sykes staled that she also told the police the intruder’s skin was light black.

4 . At a hearing held five days after the photo array, Mrs. Sykes stated that she selected someone other than Young. However, at a hearing held eight years later, she asserted that she made no identification from the photo array.

5 . The dissenting judge further chided the majority for ''miss[ing] the opportunity to hold that ..., as a matter of law, where eyewitness identification is attenuated and possibly tainted, and corroborating evidence is weak, courts should allow expert testimony concerning eyewitness identification.” Id. at 50, 817 N.Y.S.2d 576, 850 N.E.2d 623.

6 . By contrast, in determining whether a witness’s pre-trial identification has reliability independent of unduly suggestive identification procedures, courts examine the following factors: "[ (1) ] the opportunity of the witness to view the criminal at the time of the crime, [ (2) ] the witness’ degree of attention, [ (3) ] the accuracy of the witness' prior description of the criminal, [ (4) ] the level of certainty demonstrated by the witness at the confrontation, and [ (5) ] the length of time between the crime and the confrontation.” Neil v. Biggers, 409 U.S. 188, 199-200, 93 S.Ct. 375, 34 L.Ed.2d 401 (1972).

7 . The Innocence Project is an organization dedicated primarily to providing pro bono legal and related investigative services to indi gent prisoners whose actual innocence may be established through post-conviction evidence. It pioneered the post-conviction DNA model that has led to the exoneration of 289 innocent persons to date, the vast majority of whom were originally convicted based, at least in part, on the testimony of eyewitnesses who turned out to be mistaken.

8 . In a footnote in its reply brief, the State argues that, “[b]ecause [Young] did not present any expert testimony at the independent source hearing on the reliability of identification procedures, he cannot rely upon any here.” Resp.’s Reply Br. 8 n. 2 (citing Cullen v. Pinholster, - U.S. -, 131 S.Ct. 1388, 1398, 179 L.Ed.2d 557 (2011) (holding that habeas review “is limited to the record that was before the state court that adjudicated the claim on the merits”)). Our conclusion that Mrs. Sykes's in-court identification lacked an independent source is reinforced, but not compelled or controlled by, the literature we discuss below.

9 . See also State v. Guilbert, 306 Conn. 218, 234 & n. 8, 49 A.3d 705, 720 & n. 8 (2012) (collecting cases reflecting a general acceptance across circuits of scientific studies questioning the reliability of eyewitness identifications).

10 . Three of the four courts of appeal to have considered the issue appear to agree. See United States v. Ishmael, 343 F.3d 741, 742-43 (5th Cir.2003) (exercising discretion to apply Stone as procedural bar rather than treating it as mandatory bar); Tart v. Massachusetts, 949 F.2d 490, 497 n. 6 (1st Cir.1991) (reaching merits of Fourth Amendment claim in habeas proceeding in part because Commonwealth had not asserted Stone); Davis v. Blackburn, 803 F.2d 1371, 1372-73 (5th Cir.1986) (per curiam) (divided panel) (holding that in appropriate cases "a federal court is not foreclosed from sua sponte applying the principles of Stone ” noting that Stone is a "prudential” not a "jurisdictional” rule); Wallace v. Duckworth, 778 F.2d 1215, 1220 n. 1 (7th Cir.1985) (reasoning that because "respondents never raised any Stone [ ] argument, and since the rule of Stone [ ] is not a jurisdictional rule, we need not raise the issue sua sponte." (citation omitted)); but see Woolery v. Arave, 8 F.3d 1325, 1328 (9th Cir.1993) ("reading] Stone as a categorical limitation on the applicability of fourth amendment exclusionary rules in habeas corpus proceedings”).

11 . See, e.g., Day v. McDonough, 547 U.S. 198, 209, 126 S.Ct. 1675, 164 L.Ed.2d 376 (2006) (holding that where state fails to raise AEDPA statute of limitations defense, "district courts are permitted, but not obliged, to consider, sua sponte, the timeliness of a state prisoner’s habeas petition”); Trest v. Cain, 522 U.S. 87, 89, 118 S.Ct. 478, 139 L.Ed.2d 444 (1997) (holding that "[a] court of appeals is not 'required' to raise the issue of procedural default sua sponte ” given that "procedural default ... is not a jurisdictional matter”); Granberry v. Greer, 481 U.S. 129, 131, 107 S.Ct. 1671, 95 L.Ed.2d 119 (1987) (pre-AEDPA, holding that states’ failure to raise exhaustion defenses permit courts of appeal "to exercise discretion in each case to decide whether the administration of justice would be better served by insisting on exhaustion or by reaching the merits”); Danforth v. Minnesota, 552 U.S. 264, 289, 128 S.Ct. 1029, 169 L.Ed.2d 859 (2008) (holding that state can waive Teague non-retroactivity argument by not asserting it in a timely manner); Caspari v. Bohlen, 510 U.S. 383, 389, 114 S.Ct. 948, 127 L.Ed.2d 236 (1994) (holding that consequences of waiver are that "a federal court may, but need not, decline to apply Teague if the State does not argue it”) (citing Schiro v. Farley, 510 U.S. 222, 228-29, 114 S.Ct. 783, 127 L.Ed.2d 47 (1994)); Schiro, 510 U.S. at 228, 114 S.Ct. 783 (explaining that "[t]he Teague bar to the retroactive application of new rules is not ... jurisdictional”).

12 . We note that, in concluding that Mrs. Sykes's in-court identification had an independent basis, the trial court itself expressly relied on the fact that Mrs. Sykes “seemed most certain of her ability to identify [Young].” App. at 195.