Opinion
No. 103
11-26-2024
The People & c., Respondent, v. David Vaughn, Appellant.
Sam Feldman, for appellant. Melissa Owen, for respondent. The Innocence Project, Inc., The Legal Aid Society, amici curiae.
Sam Feldman, for appellant.
Melissa Owen, for respondent.
The Innocence Project, Inc., The Legal Aid Society, amici curiae.
TROUTMAN, J.
Whether to admit expert testimony concerning the factors that affect the reliability of eyewitness identifications rests within the discretion of the trial court, which balances the probative value of the proffered testimony against "the prospect of trial delay, undue prejudice to the opposing party, confusing the issues or misleading the jury" (People v Primo, 96 N.Y.2d 351, 355 [2001]; see People v McCullough, 27 N.Y.3d 1158, 1161 [2016]; People v Lee, 96 N.Y.2d 157, 160 [2001]). The presence of evidence that corroborates the eyewitness identification is not a determinative factor in deciding an application to admit this sort of expert testimony, nor is it the first part of a rigid two-step process (see McCullough, 27 N.Y.3d at 1161). To the extent our holding in People v LeGrand (8 N.Y.3d 449 [2007]) has been misinterpreted to suggest there is such a two-step process, those cases should not be followed (see e.g. People v Santiago, 17 N.Y.3d 661, 669 [2011]). Here, although the Appellate Division referenced an incorrect framework, we nevertheless hold that Supreme Court did not abuse its discretion in granting defendant's oral, midtrial application only in part and limiting the testimony of defendant's expert to the topic of cross-race effect (217 A.D.3d 781, 783 [2d Dept 2023]). Therefore, we affirm the Appellate Division's order.
I.
On May 9, 2016, two men entered a Brooklyn warehouse during working hours, one of them displaying a badge on a lanyard that he wore over a plain, brown sweatshirt. Inside the warehouse, the men were shown to an office. The man with the badge entered the office, closed the door, and drew a firearm, while the other man remained outside the door. The gunman, a Black man, forced a victim, an Asian man, to surrender $3,000 cash, and the robbers left. That victim and a second victim, who is also Asian, separately called 911 as they exited the warehouse, and for a time they pursued the fleeing robbers through the streets.
Surveillance footage depicts the robbery essentially in its entirety, as well as the robbers' entry into and exit from the warehouse. Although the footage is of fairly good quality, the quality is not so good that the robbers' facial features can be clearly discerned.
Four days later, the first victim was standing at a loading dock when he saw defendant and another man walk by. The victim recognized those men as the robbers, noting that they were wearing the same clothes that the robbers wore during the robbery. In defendant's case, he wore a plain, brown hooded sweatshirt. The victim would later testify at trial that he and defendant locked eyes, and then defendant put his hood up. The victim hailed a passing police car and guided the officers down the path that the robbers took during their earlier flight from the warehouse. Shortly thereafter, the victim spotted defendant and his companion and identified them as the robbers. Defendant was arrested and brought to the precinct, where the sweatshirt was photographed inside-out, vouchered, and placed in an evidence bag. By the time of trial, the sweatshirt, which had been wet from the rain when bagged, was destroyed by mold.
Defendant was charged with, inter alia, robbery in the first degree (Penal Law § 160.15 [4]), and the matter proceeded to a jury trial. After the parties finished selecting a jury, they delivered their opening statements, and the People called their first two witnesses before the court adjourned for a holiday weekend. Testimony resumed five days later. On the morning of the sixth day, the court put on record that defense counsel called chambers late on the day the court adjourned for the holiday weekend and stated that he intended to call an expert to testify about cross-race effect. Defense counsel confirmed this and added that the parties received an email from chambers, approximately two weeks before trial commenced, asking if the defense intended to present expert testimony on cross-race effect. Defense counsel had responded to the court's email by stating that he "may intend" to do so.
Cross-race effect, one of the factors that affect the reliability of eyewitness identifications, is based on scientific studies demonstrating that the likelihood of misidentification is higher when the eyewitness and the defendant are of different races (see People v Boone, 30 N.Y.3d 521, 528 [2017]).
Defense counsel then stated, for the first time, his intention to elicit expert testimony not only on cross-race effect, but also on several other factors that affect the reliability of eyewitness identification, "such as" the reliability of showup identification procedures, the effect of stress, weapon focus, the duration of the encounter, distance and lighting, the use of disguises, and memory decay. Asked by the court for "something in writing," defense counsel acknowledged that he had nothing prepared. The court stated that it was important for defense counsel to identify all the factors on which he intended to present testimony because there may be legal authority establishing whether they are generally accepted in the scientific community, obviating the need for a Frye hearing to determine that issue (see generally Frye v United States, 293 F 1013 [DC Cir 1923]). The court recognized that the general acceptance of cross-race effect had been established in People v Boone (30 N.Y.3d 521, 528-529 [2017]) and asked if defense counsel had case law supporting the general acceptance of the additional factors about which he sought to elicit expert testimony.
Defense counsel did not have any case law to provide but offered to list the factors and supply case law later, prompting the court to admonish him for arriving late: "[T]his is why I asked everybody to be here at 10:30 this morning because I knew this was an outstanding matter and we would be spending some time on it. Now, it's after 11:00." Defense counsel apologized and listed a total of 11 factors for the court's consideration: the eight listed previously plus the influence of private actors, the effect of multiple perpetrators, and the correlation between confidence and accuracy.
The court again emphasized to defense counsel the importance of providing case law and, referring to Boone, opined that the additional factors such as the effect of stress and weapon focus were within the ken of an average juror. Thus, the court indicated that it would likely not allow testimony as to those factors unless defense counsel provided case law to the contrary. The People asked the court to preclude the expert testimony on the ground that defendant's application was untimely, but at that point the court stated that it was "not likely" to deny the application on that ground. The People further argued that the expert testimony should be excluded because the identification was corroborated by the surveillance video and clothing. Defense counsel argued that the jury could not conclude that defendant was the culprit based on the video and clothing alone and that the expert testimony would be informative for a jury, and he asked "at least" for testimony on cross-race effect.
The court chided defense counsel for failing to make a motion in limine:
"This case is a year old. This has never been raised before. We could have had a hearing on this. I could have had both sides brief this issue and, here we are, with a jury in the back, in the middle of trial, and you are... telling me that you want to call this expert witness... [Y]ou were in front of me at least two weeks ago-ten days ago. It could have been brought up before we ever picked a jury. We would have delayed on picking the jury. I would have seen this as a very significant issue."
Testimony resumed, and at the end of the day the court issued a "preliminary ruling" that it was "not likely" to permit testimony on any factor other than cross-race effect. The court stated that it had read several cases and that it had discretion to admit or exclude the testimony because the identification was corroborated by the clothing and video. The court added: "[T]he other issue is... you really needed to raise this issue a while ago, and perhaps it might have been a different result. But getting this information in the middle of the People's case that you're calling a witness... we're doing a lot of catch up here." The court stated that it could have denied defendant's application in its entirety but explained that it was important to afford defendant the "opportunity to put on a defense."
Trial proceeded. Defendant neither renewed his application nor supplied case law to support it. The court adhered to its preliminary ruling, and defendant's expert testified about cross-race effect.
The jury found defendant guilty, and the court sentenced him as a persistent felony offender to an indeterminate prison term of 20 years to life. The Appellate Division affirmed, concluding that "the People presented sufficient corroborating evidence connecting the defendant to the crime to obviate the need for the additional inquiry required by People v LeGrand," and that "[u]nder the circumstances of this case, the court providently exercised its discretion in limiting the testimony" (217 A.D.3d at 783).
A Judge of this Court granted leave to appeal.
II.
The" 'single greatest cause of wrongful conviction'" is mistaken eyewitness identification (Boone, 30 N.Y.3d at 527). Although eyewitness identifications are often reliable (see Young v Conway, 698 F.3d 69, 80 [2d Cir 2012]; Boone, 30 N.Y.3d at 538 [Garcia, J., concurring]), the prevalence of eyewitness misidentifications, the danger they pose to the "truth-seeking function and integrity of our justice system," and the need to mitigate that danger are so beyond dispute and so well established, in New York State and elsewhere, that we hardly need to discuss the matter further (Boone, 30 N.Y.3d at 528 [majority]). Situational and environmental factors that affect the reliability of eyewitness identifications have been the subject of scientific study for decades, and defendants in criminal proceedings, likewise for decades, have sought to present the testimony of experts in the "scientific data concerning certain factors involved in perception and memory" (People v Mooney, 76 N.Y.2d 827, 831 [1990, Kaye, J., dissenting]; see LeGrand, 8 N.Y.3d at 452). We have repeatedly" 'encouraged'" the trial courts to admit this sort of expert testimony in appropriate cases (Santiago, 17 N.Y.3d at 669, quoting People v Drake, 7 N.Y.3d 28, 31 [2006]).
Questions of the admissibility and scope of expert testimony concerning the factors that affect the reliability of eyewitness identifications in a particular case are addressed to the trial court's sound discretion (see People v Berry, 27 N.Y.3d 10, 19-21 [2016]; People v Abney, 13 N.Y.3d 251, 259-260, 268 [2009]; see also People v Cronin, 60 N.Y.2d 430, 433 [1983]). Courts deciding those questions apply traditional evidentiary principles (see McCullough, 27 N.Y.3d at 1161), which require the courts to weigh the testimony's probative value against its prospect of causing undue prejudice to the opposing party, confusing the issues, misleading the jury, or unduly delaying trial (see People v Perdue, 41 N.Y.3d 245, 252 [2023]; McCullough, 27 N.Y.3d at 1161).
On an application to admit expert testimony of this sort, the trial court may need to determine whether the expert testimony is beyond the ken of the average juror or generally accepted in the scientific community (see Lee, 96 N.Y.2d at 162). Indeed, in Abney, we reversed and ordered a new trial where the trial court abused its discretion in denying an application to present expert testimony on several factors, concluding that the court should have held a Frye hearing to resolve the issue of general acceptance (see Abney, 13 N.Y.3d at 268). While general acceptance may be established at a Frye hearing, a hearing is not necessary in all cases (see People v Wesley, 83 N.Y.2d 417, 426 [1994]). General acceptance may be established through legal precedent (see Matter of Lahey v Kelly, 71 N.Y.2d 135, 144 [1987]; Boone, 30 N.Y.3d at 528 [considering lower court opinions to support the conclusion that cross-race effect is generally accepted in the scientific community]; cf. People v Williams, 35 N.Y.3d 24, 39 [2020] [holding that the court abused its discretion as a matter of law in admitting DNA evidence without a Frye hearing because the cases on which the court relied did not adequately assess whether the evidence was generally accepted in the scientific community]). Where the defendant fails to demonstrate that a topic of the proffered expert testimony is generally accepted in the relevant scientific community, the trial court should exclude or limit the testimony as appropriate (see Berry, 27 N.Y.3d at 19).
Generally, the scientific studies concerning the factors that affect the reliability of eyewitness identifications are beyond the ken of the average juror (see Lee, 96 N.Y.2d at 162). Although that may not be true of every conceivable such factor, we have yet to find testimony of this sort within the ken of the average juror, and we have never upheld the exclusion of this kind of testimony on the ground that it was within the ken of the average juror (cf. Abney, 13 N.Y.3d at 260, 268).
Courts must not decide whether evidence is admissible based solely on the existence or strength of corroborating evidence (see People v Oddone, 22 N.Y.3d 369, 379 [2013]; see also Holmes v South Carolina, 547 U.S. 319, 329-330 [2006]). Nor should courts require adequate corroborating evidence as a prerequisite to weighing other considerations pertinent to admissibility (see McCullough, 27 N.Y.3d at 1161). Rather, courts should be guided by "whether the proffered expert testimony 'would aid a lay jury in reaching a verdict'" (Lee, 96 N.Y.2d at 162; see Abney, 13 N.Y.3d at 266).
Our decision in People v LeGrand does not say otherwise. In that case, we held that the court abused its discretion in excluding testimony concerning three factors that affect the reliability of eyewitness identification-the correlation between confidence and accuracy, the effect of postevent information, and confidence malleability (see 8 N.Y.3d at 452, 458; see also McCullough, 27 N.Y.3d at 1161). In so holding, we observed that the prerequisites for the admission of scientific evidence were satisfied- i.e., the testimony concerning the three factors was "relevant" to assessing the accuracy of the eyewitness identification in that particular case; the testimony was "based on principles that are generally accepted within the relevant scientific community"; the testimony was "proffered by a qualified expert"; and the topic of the testimony was "beyond the ken of the average juror" (LeGrand, 8 N.Y.3d at 452). Although we also considered that that there was "little or no corroborating evidence connecting the defendant to the crime," the "main issue" in LeGrand did not involve corroboration at all, but instead concerned whether the defendant had established at the Frye hearing that the testimony was generally accepted in the scientific community (id.).
To be clear, the considerations enumerated in LeGrand do not displace the traditional principles under which courts weigh the probative value of the testimony against the prospect of delay, prejudice, or confusion (see McCullough, 27 N.Y.3d at 1161). Courts should not apply any two-step process that references corroboration or credibility as gateway factors for the exercise of discretion by the trial court.
III.
Here, the court did not abuse its discretion in granting defendant's application only in part and limiting the expert testimony to the topic of cross-race effect. The record establishes that the court reached its determination by weighing the probative value of the testimony against, among other things, the prospect of unduly delaying trial (see Perdue, 41 N.Y.3d at 252).
The record belies the dissent's suggestion that trial delay concerns, specifically those occasioned by a Frye hearing and a full briefing, were not an "actual basis" of the court's ruling (dissenting op at 20).
On the morning of the third day of testimony, for the first time, defendant sought a ruling allowing expert testimony on cross-race effect and 10 additional factors that affect the reliability of eyewitness identifications. Although the court was familiar with the law concerning cross-race effect from our then-recent decision in Boone, the same cannot be said for the rest of the factors. Defendant failed to provide a written application or legal memorandum at the court's request or cite case law to satisfy the prerequisites for the admission of scientific testimony. Even after the court made its "preliminary ruling," defendant had the opportunity to renew his application, supplement it with legal support, or narrow it to some of the more relevant or legally supportable topics (see Abney, 13 N.Y.3d at 259-260, 268). Defendant may have dispelled the court's stated concerns at least with respect to some of the factors if he had responded to the court's direct request for case law supporting his position. For example, our precedent establishes that the correlation between confidence and accuracy is generally accepted in the relevant scientific community and beyond the ken of the average juror, both of which are necessary to demonstrate the testimony's admissibility (see LeGrand, 8 N.Y.3d at 452). Defendant did none of these things.
We cannot conclude that a court abused its discretion in excluding or limiting expert testimony of this sort on such a perfunctory application brought at this juncture of the proceedings (see Santiago, 17 N.Y.3d at 666-667 [reversing where defendant moved before trial and renewed motion after a different Justice was assigned to the case]; Abney, 13 N.Y.3d at 259-260 [reversing where defendant moved before jury selection, the court denied as premature with leave to renew, and defendant renewed motion after the People rested]; LeGrand, 8 N.Y.3d at 453 [reversing where defendant moved prior to retrial]). Indeed, the cases where we have reversed illustrate the importance of a timely motion because admissibility in those cases hinged in part on the outcome of a Frye hearing that the court was able to hold before trial (see LeGrand, 8 N.Y.3d at 458), should have held (see Abney, 13 N.Y.3d at 268), or may have found necessary to hold (see Santiago, 17 N.Y.3d at 672). Of course, the timing of the motion is not dispositive, and indeed is less crucial, if the defendant establishes that a Frye hearing is unnecessary under precedent recognizing that the proposed testimony is generally accepted in the scientific community (see Lahey, 71 N.Y.2d at 144), but defendant made no attempt to do so here.
Contrary to the dissent's assertion, our decision does not incentivize defense attorneys to request Frye hearings if general acceptance can be established without such a hearing (see dissenting op at 19-20). Our decision should incentivize defense attorneys to prepare their motions before trial, provide trial court judges with written submissions supporting their positions, including citation to relevant authority, and address as appropriate any concerns raised by the trial courts concerning those motions.
Even on appeal, defendant makes no attempt to demonstrate that all or even most of the factors concerning which he sought to elicit expert testimony satisfy Frye 's general acceptance standard. Defendant now focuses on three factors that he perceives to be most easily established as the subject of admissible expert testimony-the effect of stress, weapon focus, and the correlation between confidence and accuracy. Our dissenting colleague does the same. By focusing now on the most legally supportable factors and providing support for their admissibility, defendant does what he could and perhaps should have done in the trial court. His failure to have done so in the face of the trial court's stated concerns and direct request for case law is precisely the problem here. Thus, contrary to the dissent, defendant fails even on appeal to obviate the need to conduct a Frye hearing for the trial court to rule on the entirety of the application. Moreover, although our dissenting colleague "would expect" the court first to "hear from counsel on the cause and potential length of any delay" (dissenting op at 19), we presume that this judge was aware of the anticipated length of delay a midtrial Frye hearing would create.
The court expressed that it understood the importance of the issue and that it was sensitive to defendant's right to present a defense. The court also expressed that it would have been amenable to delaying trial for a Frye hearing and a full briefing had defendant made his application prior to jury selection. In the end, the court exercised its discretion by granting defendant's application in part and limiting the expert's testimony to the topic of cross-race effect. We emphasize that the timing of the application and the prospect of trial delay are not dispositive considerations. The court could have delayed the jury trial and held the Frye hearing or granted the application in its entirety if defendant had been able to establish that a Frye hearing was not required. Under the circumstances, however, especially given defense counsel's failure to submit a written application supported by legal authority despite the court's request for that information, we cannot say that there was an abuse of discretion in the court's balancing of the proffered testimony's probative value against the prospect of causing undue prejudice, confusing the issues, misleading the jury, or unduly delaying the proceedings.
Accordingly, the order of the Appellate Division should be affirmed.
RIVERA, J. (dissenting)
Mistaken eyewitness identifications are the leading cause of wrongful convictions. There is a scientific consensus that various factors-including the three raised in this appeal-impact memory in a way that increases the risk of witness misidentification. As the Court has repeatedly recognized, expert testimony regarding factors that decrease the reliability of eyewitness identifications minimizes the grave risk of incarcerating innocent people. Despite acknowledging both that wrongful convictions are a real occurrence and that expert testimony can help jurors understand counterintuitive subjects, the majority decides that judicial economy is more important than a defendant's constitutional right to present a defense (see id. at 12). According to the majority, vague concerns about trial delay alone are sufficient to preclude relevant, scientifically-sound expert testimony on the unreliability of eyewitness identification. That conclusion is unprecedented. It also has no support in this record, where the trial court rejected defendant's expert testimony, not as untimely, but as unnecessary and, accordingly, made neither factual findings pertaining to any possible delay nor a ruling that the probative value of the evidence was outweighed by the danger that its admission would prolong the trial to an unreasonable extent without any corresponding advantage (see People v Davis, 43 N.Y.2d 17, 27 [1977]).
Finally, the majority reaffirms a fundamental error in the application of the corroboration rule by again permitting a court to credit the very eyewitness testimony contested by the proffered expert opinion even though questions concerning the credibility and reliability of such evidence "have been traditionally reserved for the trier of fact" (Holmes v South Carolina, 547 U.S. 319, 330 [2006]). Crediting such testimony before the defendant has an opportunity to test its strength before the factfinder-through cross examination aided by expert opinion-creates a significant risk of wrongful conviction and does not rationally serve the ends of justice. Therefore, I would reverse and order a new trial.
I.
Defendant was tried before a jury for, among other counts, robbery and burglary arising from an armed theft at a warehouse. No forensic evidence linked defendant to the crime and the prosecution's case therefore depended largely on the testimony of two eyewitnesses working at the warehouse who identified defendant as the gunman. The eyewitnesses are Asian men and defendant is Black. The defense was that the eyewitnesses mistakenly identified him.
Pre-trial, the court asked defense counsel if he intended to move for the presentation of expert testimony regarding cross-racial effects on the reliability of eyewitness identifications. Defense counsel responded that he "may intend" to do so and later clarified that this meant he was "trying to find" an expert who was able to take the case and willing to accept the modest rates paid to expert witnesses who testify on behalf of indigent defendants. A week later, defense counsel confirmed his intent to call an expert. On the following business day, the trial court agreed with the prosecution that his request "can wait" because it "[did] not involve" the prosecution witness testifying that day. During an in-court colloquy the next day, defense counsel specified his intent to present expert testimony on cross-race effects and a host of other factors that diminish the reliability of eyewitness identifications, including those which he presents on appeal: the witness's "stress," "weapon focus," and "confidence relative to... accuracy."
As the majority observes, the trial court described defense counsel's application as having come" 'in the middle of trial'" and" 'in the middle of the People's case'" (majority op at 6), characterizations that the majority endorses (see id. at 2, 13). However, the record shows that the issue of the defense's application came up while the prosecution was presenting the second of its eight witnesses at which point the defense was told the application "can wait." The trial court then permitted defense counsel to flesh out his application the following day after the prosecution finished presenting its third witness. Thus, if the trial court was unfamiliar with the parameters of the application until the middle of the prosecution's case, that was of its own choosing. Either way, the application certainly did not come "midtrial" (majority op at 2, 13). And it certainly preceded the testimonies of the two eyewitnesses to which the expert's testimony pertained.
In response, the trial court expressed doubt about whether "the jurors need[ed] an expert to explain... that it might be stressful during a crime if there is a gunpoint robbery." Defense counsel explained that "an expert would be able to clarify how the human mind works under stress" and that the lay jurors were unaware of "how stress can affect a witness's perception under certain circumstances." Regarding overconfidence, counsel explained that "there are witnesses that will swear and testify that they are 100 percent sure, they have a tremendous amount of confidence as to their identification, but that does not mean that the identification was accurate."
The trial court indicated that it would permit expert testimony on cross-racial identifications, but unless the defense supplied it with "case law," would not permit expert testimony on the effects of stress, weapon focus, and overconfidence because those other subjects were "all within the k[e]n of a juror using their good common sense." The prosecution protested that the court should have barred defendant from calling an expert on cross-racial effects as well, arguing-much like the majority here posits concerning the other factors-that the defense's application was ill-timed. The trial court "chided defense counsel for failing to make a motion in limine," as the majority recounts (majority op at 5) but nonetheless overruled the prosecution's objection and made clear that defense counsels' timing was not the driver of its decision, commenting that it was "not likely to grant [the prosecution's] application" to preclude the expert "based on a timeliness issue." Defense counsel also reminded the trial court that he had indicated pre-trial that he might call an expert witness, and explained that any delay in specifying which expert was attributable to his difficulty locating an expert willing to accept the fee rates set by County Law § 18-b.
The prosecution persisted in its effort to preclude the testimony altogether on the ground that in People v Boone (30 N.Y.3d 521 [2017]), where we first approved of the admission of expert testimony on cross-racial effects, "there was no other corroborating evidence," unlike here, the prosecution continued, where there was "more than one witness who was involved," along with surveillance video. Defense counsel countered that part of the prosecution's corroborating evidence was another eyewitness whose identification raised the same cross-racial concern-i.e., an Asian male identifying a black defendant and that the prosecution's "corroborating evidence would assume that the jury has to take" the corroborating evidence based on an initial presentation without further consideration. The court agreed with the prosecution that its case was "not relying just on the witnesses" because of the surveillance video, "where the jury can have the opportunity to look at the videotape and look at the defendant" and further commented that the surveillance video was "powerful."
Later, the trial court reiterated what it termed its earlier "preliminary ruling", that it would permit expert testimony solely on cross-racial identifications and not "anything else." The trial court also informed the parties that it had independently researched the law in this area. Specifically, the trial court explained that it had "been reading a number of cases... [o]n this identification testimony," and, based on that case law, concluded that "it's really in the discretion of the court with respect to denying it when it is corroborated by other evidence in the case, and there is corroboration for the identification through the clothing of the defendant, as well as a videotape."
The prosecution's case relied primarily on the testimonies of two eyewitnesses, J. and C., each of whom identified defendant in court as one of the men who robbed them-C. for the first time. One victim described the man who entered the office as a six-foot tall black man "in his mid 40s,... had a hat on," and "pulled out a silver gun." The other victim pegged his height at 6 feet, 2 inches with "facial hair, [a] goatee," a "[h]unched-over posture," and described him as wearing a brown hoodie, blue jeans, red headphones, and a silver shield around his neck. Both victims described the man who remained outside the door-the co-defendant-as a shorter black male, who one of them recalled wearing a navy-blue fleece jacket. The in-court descriptions tracked their respective descriptions to the investigating officers after the robbery. Neither victim ever described either of the men as missing a bottom tooth.
The prosecution also presented video from inside the warehouse that captured the robbery committed by two robbers-one holding a gun and wearing what appears to be a brown sweater-over the course of roughly two minutes. But the video does not clearly depict their faces. The prosecution also presented evidence that defendant was wearing a brown sweater when he was arrested and presented a photograph of the sweatshirt turned inside-out.
For his part, defense counsel introduced evidence casting doubt on the eyewitnesses' identifications. During cross-examination J.-who, four days after the robbery, spotted two men who he "had a feeling" were the "two guys" who robbed him and contacted police-confirmed that he "was not 100 percent sure" at that point and that he had made this guess from 30 to 35 feet away. The defense also introduced a meteorologist, who testified that a quarter inch of rain fell nearby at John F. Kennedy International Airport that day, with rainfall peaking between 5 and 6 p.m., when the airport reported fog. Defense counsel also called a police officer who was present for J.'s identification of defendant four days after the robbery, who recalled J. telling the officers he was sure the two men were the perpetrators "because [defendant] still ha[d] the brown sweater on" without mentioning any other distinctive features such as defendant's face or hair.
The defense also introduced a photo of defendant taken before the robbery, which showed that he had a gap in his lower front teeth where one or two teeth were missing, and later had defendant show his missing teeth to the jury. Finally, the defense expert also testified, as limited by the court's ruling, solely on the questionable reliability of cross-racial identifications. He explained that there is an "often replicated or repeated finding" that people are "worse at identifying members of another race or ethnicity when we are talking about stranger identification" and that this cross-racial effect reduces "the likelihood that someone can accurately identify an unfamiliar strange face after a single exposure to that face."
Defense counsel argued during his summation that defendant was "charged with something he did not do... [a]ll because he's wearing what appears to be a brown sweatshirt," and observed that brown is "a very common color," the sweatshirt itself had been destroyed by mold and, since the only available photograph showed it turned inside-out, the jury could not see whether it had any distinctive features like a logo. He also commented on the "terrible" viewing conditions under which J. identified defendant four days after the robbery based on his wearing a similar, plain brown sweatshirt at the time. He urged the jury to skepticize the identifications because "[t]his is a very stressful situation for somebody [who] is getting robbed with a gun." He also reminded them that, although defendant had "a prominent missing bottom tooth" prior to the date of the robbery, the surveillance video showed the robber's teeth as he was talking and no missing tooth was visible, and both victims testified that the man with the gun had all of his teeth. Finally, he noted that C. never identified defendant as the robber before the trial, but instead waited 653 days before pointing to "the only African American man sitting at the Defense table... [w]ith officers behind him."
During its summation, the prosecution likewise emphasized the centrality of the victims' identifications and presented weapon focus as buttressing rather than tainting the identification, urging that "the question here isn't how [the victims] remember. How could they forget[?] How could they forget the face of the man who held a gun to them[?] Members of the jury, I submit to you, it doesn't matter... if it was two days, three days or if it was 653 days. That's a face you are not going to forget."
The prosecution further told the jury that "at the end of the day what a lot of things in this case come down to [is] [the victims]" and emphasized that it was "asking [the jury] to rely on their testimony."
The jury found defendant guilty of first-degree robbery and second-degree burglary and the trial court subsequently sentenced him to an aggregate 20-year prison term. The Appellate Division affirmed (217 A.D.3d 781 [2d Dept 2023]) and a Judge of this Court granted defendant leave to appeal (40 N.Y.3d 999).
II.
A.
Misidentifications are the leading cause of wrongful convictions (Boone, 30 N.Y.3d at 527 ["Mistaken eyewitness identifications are 'the single greatest cause of wrongful convictions in this country' ", quoting State v Delgado, 188 N.J. 48, 60, 902 A.2d 888, 895 [2006], citing State v Dubose, 285 Wis.2d 143, 162-163, 699 N.W.2d 582, 592 [2005]; People v Marshall, 26 N.Y.3d 495, 502 [2015] ["Wrongful convictions based on mistaken eyewitness identifications pose a serious danger to defendants and the integrity of our justice system"], citing United States v Wade, 388 U.S. 218, 228 [1967] ["The vagaries of eyewitness identification are well-known; the annals of criminal law are rife with instances of mistaken identification"]; People v Santiago, 17 N.Y.3d 661, 669 [2011] ["(M)istaken eyewitness identifications play a significant role in many wrongful convictions"]). Courts must have a solid basis for precluding expert testimony about the ways in which eyewitness identifications, particularly involving strangers, may be unreliable, for even an eyewitness confident in their identification may be mistaken due to risk factors that impact memory.
As I have previously discussed, the science concerning the many environmental factors that undermine the reliability of eyewitness identifications is now well-settled and undisputable (see People v Perdue, 41 N.Y.3d 245, 257-259 [2023] [Rivera, J., dissenting]). As relevant here, high stress and the presence of a weapon greatly increase the chance of a mistake by an eyewitness. Even as far back as 2001, a survey of psychologists who studied the topic revealed that, when presented with the proposition that "[t]he presence of a weapon impairs an eyewitness's ability to accurately identify the perpetrator's face," 87% found the statement "reliable" and 97% believed it had some "research basis" (Saul M. Kassin et al., On the "General Acceptance" of Eyewitness Testimony Research: A New Survey of the Experts, 56 Am Psych 405, 408, 412 [2001]). When presented with the proposition that "[v]ery high levels of stress impair the accuracy of eyewitness testimony," 60% found it "reliable" and 98% believed there was a "research basis" for it (id. at 405, 408, 412).
The Innocence Project has filed an amicus brief in this appeal further detailing the settled science on this subject (see https://www.nycourts.gov/reporter/webdocs/PeoplevVaughn_AmicusBrief.pdf).
What was once an emerging consensus has solidified. In a recent survey, 78% of experts concurred with the proposition that "[h]igh levels of stress impair the accuracy of an eyewitness's memory for an event" (Travis M. Seale-Calisle et al., New Insights on Expert Opinion About Eyewitness Memory Research, 18 Perspectives on Psych Science 1, 5, 7 [2024]). The same is true of a witness's "confidence at the time of trial" in their possibly stress-affected identification of a defendant as the perpetrator, which "is not a reliable predictor of eyewitness accuracy" (Natl. Research Council, Identifying the Culprit: Assessing Eyewitness Identification 108 [2014], available at http://nap.nationalacadamies. org/download/18891 [last accessed Oct. 28, 2024], citing, inter alia, C. M. Allwood, J. Knutsson, & P. A. Granhag, "Eyewitnesses Under Influence: How Feedback Affects the Realism in Confidence Judgements," Psychology, Crime, and L 12 [1] at 25-38 [2006]). Several meta-analyses confirm that there is no real debate within the scientific community about these issues (see e.g. Angela M. Jones et. al, Comparing the Effectiveness of Henderson Instructions and Expert Testimony: Which Safeguard Improves Jurors' Evaluations of Eyewitness Evidence, 13 J Experimental Criminology 29, 31 [2017]; Kerstin Kocab & Siegfried L. Sporer, The Weapon Focus Effect for Person Identifications and Descriptions: A Meta-Analysis, Advances in Psychology and Law: Vol 1 71-72 [June 2016]; Kenneth A. Deffenbacher et al., A Meta-Analytic Review of the Effects of High Stress on Eyewitness Memory, 28 L & Human Behavior 687, 699 [2004]).
Though sound ethical considerations limit laboratory testing of stress effects, one study of military servicemembers under which each participant was subject to interrogations in high- and low-stress environments and then asked to identify their interrogators after 24 hours, revealed that, for every subject who correctly identified their interrogator after that time, two identified the wrong person (Charles A. Morgan III et al., Accuracy of Eyewitness Memory for Persons Encountered During Exposure to High Intense Stress, 27 Intl. J.L. Psychiatry 265, 273 [2004]). Subjects interrogated under low-stress conditions were much more likely to identify the correct person after 24 hours (see id.). Similarly, contemporary studies have shown that "the presence of a weapon reduced both identification accuracy and feature accuracy" and accuracy is further reduced "in threatening scenarios than in non-threatening ones" (Natl. Research Council, Identifying the Culprit at 93, citing J.M Fawcett et al.," Of Guns and Geese: A Meta-Analytic Review of the 'Weapon Focus' Literature," 19 Psychology, Crime and L 35-36 [2013]).
Studies likewise have continued to reinforce that consensus that witness confidence in identifications is both a poor predictor of the identification's accuracy yet a dangerously persuasive feature for lay jurors. "[P]ost-identification feedback," meaning positive reinforcement from others concerning the identification, is largely to blame (Gary L. Wells and Laura Smalarz, Lives Destroyed by Distorted Recollections of Fluency, Attention, View, and Confidence: A Sin of Bias in Eyewitness Identification, 11 J Applied Research, Memory & Cognition, 461, 461 [2022]). For example, in one study, subjects were shown a simulated crime, followed by a lineup from which many of the subjects identified a suspect despite the perpetrator's absence from the lineup (see id.). Subjects who were then told "good, you identified the suspect" later "showed a shocking degree of distortion in their recall of the ease with which they were able to pick the person out of the lineup, how good their view was, how much attention they paid during witnessing, and the confidence they had at the time of the identification" (id.). Science confirms the tendency of lay jurors to assign weight to this inflated confidence, thereby insidiously distorting "the truth-seeking function and integrity of our criminal justice system" (Boone, 30 N.Y.3d at 528). Studies show that subjects participating in simulations as jurors can "significantly discriminate between accurate and mistaken eyewitnesses when the witness had not received confirming postidentification feedback" (Laura Smalarz & Gary L. Wells, Post-Identification Feedback to Eyewitnesses Impairs Evaluators' Abilities to Discriminate Between Accurate and Mistaken Testimony, 38 L & Human Behavior 192, 199 [2014]). Once the eyewitness is told that they identified the correct person and conveys the identification to the jurors, however, "the ability of [the jurors] to discriminate between accurate and mistaken testimony was totally eliminated" (id. at 200).
B.
In light of this decades-old consensus, the Court has long approved the propriety of expert testimony regarding the reliability of eyewitness identifications. Notably, at the time of defendant's trial, there was ample precedent-including from this Court-based on these studies, supporting the admission of the expert testimony regarding the specific factors impairing reliability that the trial court barred defendant from presenting here.
The trial court's preclusion of the expert testimony here is due to its consideration of alleged corroborating evidence as dispositive. As I discuss more fully below, that was error (see infra, § III). Courts presented with applications to admit expert testimony engage in a two-step inquiry to determine its admissibility. At the first step, the court must evaluate" 'whether the expert testimony 'would help to clarify an issue calling for professional or technical knowledge, possessed by the expert and beyond the ken of the typical juror' '" (People v Nicholson, 26 N.Y.3d 813, 828 [2016], quoting People v Williams, 20 N.Y.3d 579, 583-584 [2013], quoting De Long v County of Erie, 60 N.Y.2d 296, 307 [1983]). If so, the court must then assess whether other experts in the relevant field have accepted the scientific methods and techniques the evidence's offeror has proffered. "[I]n recognition that expert testimony... may involve novel scientific theories and techniques, a trial court may need to determine whether the proffered expert testimony is generally accepted by the relevant scientific community" (People v Lee, 96 N.Y.2d 157, 162 [2001]). Unlike many other jurisdictions, this Court has held onto the framework for assessing general acceptance first announced over a century ago in Frye v United States (293 F 1013, 1014 [DC Cir 1923]), the central inquiry of which is" 'whether the accepted techniques, when properly performed, generate results accepted as reliable within the scientific community generally'" (Parker v Mobil Oil Corp., 7 N.Y.3d 434, 446 [2006], quoting People v Wesley, 83 N.Y.2d 417, 422 [1994]). Of course, courts need not conduct a full Frye hearing every time a party moves to introduce expert testimony. "Judicial precedent may also support a conclusion with respect to the general reliability of a disputed scientific technique short of a hearing" (People v Williams, 35 N.Y.3d 24, 38 [2020]; see also Wesley, 83 N.Y.2d at 436 [Kaye, Ch. J., concurring] ["Once a scientific procedure has been proved reliable, a Frye inquiry need not be conducted each time such evidence is offered" and (c)ourts thereafter may take judicial notice of reliability of the general procedure"]).
The majority acknowledges that parties and trial courts may consider previous Frye hearings and determinations in other cases (see majority op at 7-8). Nevertheless, the majority ignores the abundance of precedent establishing the admissibility of expert testimony covering the factors at issue here. In People v Lee, the Court first recognized that these scientifically-recognized factors that impair the reliability of eyewitness identifications and the studies that revealed and confirmed them are "beyond the ken of the typical juror" (96 N.Y.2d 157, 162 [2001]). The Court specifically pointed to several of the factors on which defendant sought to present expert testimony here in People v Young (7 N.Y.3d 40 [2006]). There, the Court observed that the defendant moved to present expert testimony "that, where a crime is committed with a weapon, the victim's tendency to focus on the weapon may interfere with his or her observation of the criminal; that a stressful experience is more likely to be a memorable one up to a moderate level of stress, but that events generating higher levels of stress are remembered less well; and that there is only a very weak relationship between a witness's confidence in the accuracy of his or her recollection and its true accuracy-i.e., that people very confident in their recollections are often wrong" (id. at 43).
The majority's observations that "we have yet to find testimony of this sort within the ken of the average juror" and "have never upheld the exclusion of this kind of testimony on the ground that it was within the ken of an average juror" (majority op at 8 n 2) is squarely at odds with the trial court's contrary determination here. Perhaps without realizing it, the majority highlights the trial court's grave error in excluding the expert testimony when it concedes that the three factors at issue in this appeal are "legally supportable" (majority op at 12). In doing so, the majority bolsters a conclusion antithetical to the one it reaches today.
The Young Court ultimately held that, under the circumstances present in that case, the trial court there did not abuse its discretion when it denied the defendant's application (see id. at 46). However, the Court acknowledged that the proffered factors "were relevant to [the] identification of [the] defendant" and "could have been valuable to a juror... in assessing [the identifier's] testimony" because "the expert [could] tell the jury something significant that jurors would not ordinarily be expected to know already" regarding the factors that impact the accuracy of identification which were not "so obvious or well known that ordinary jurors would not benefit from hearing them" (id. at 45).
As concerns the second step, "[c]ourts across the country now accept that certain factors can greatly affect the accuracy of eyewitness identification testimony" and have admitted expert testimony on a variety of them including, as relevant here, "the weak correlation between confidence and accuracy regarding an identification;... the diminishing reliability of an identification caused by the witness's focus on a weapon;... and... the negative impact of high stress at the time of the observation on the ability to retain an accurate perception of events" (Jones v United States, 262 A.3d 1114, 1125-1126 [DC 2021] [footnotes omitted] [cataloging cases]). So too have courts in New York, including this one, incorporated scientific consensus regarding these factors into our jurisprudence. As far back as a decade ago, the Court had already repeatedly held that the perils of witness confidence and post-identification feedback have gained sufficient acceptance in the scientific community, such that expert testimony concerning them is admissible (see Santiago, 17 N.Y.3d at 672; People v Abney, 13 N.Y.3d 251, 268 [2009]). And lower courts have long held the same regarding many other factors including, among other things, weapon focus and event stress (see e.g. People v Abney, 31 Misc.3d 1231 [A] at *29-35 [Sup Ct, New York County 2011]; People v Norstrand, 35 Misc.3d 367, 373 [Sup Ct, Monroe County 2011]). The scientific consensus has only grown stronger since then (see e.g. Seale-Calisle et al., New Insights on Expert Opinion About Eyewitness Memory Research, 18 Perspectives on Psych Science at 1, 5, 7 Natl. Research Council, Identifying the Culprit; Jones et. al, Comparing the Effectiveness of Henderson Instructions and Expert Testimony, 13 J Experimental Criminology at 29, 31; Kocab & Sporer, The Weapon Focus Effect, Advances in Psychology and Law: Vol 1 71-72; Fawcett et al., "Of Guns and Geese," Psychology, Crime and L 19 [1] at 35-36).
III.
Here, the majority upholds the trial court's rejection of counsel's request to present expert testimony concerning these scientifically-proven phenomena, not because the majority questions their general acceptance in the scientific community, but because, the majority says, defendant failed to bring the trial court's attention to this general acceptance. The majority reaches its conclusion that the trial court acted within its discretion to obviate "the prospect of unduly delaying the trial" (majority op at 10). That reasoning is flawed as a matter of fact and law.
First, the record establishes that the trial court limited the expert's testimony not due to some vague concern about trial delay but because the court determined on the merits that the testimony was unnecessary. The Appellate Division similarly addressed the merits of defendant's proffered expert testimony and made no mention of a timeliness bar as justification for the trial court's decision (see 217 A.D.3d at 783). Indeed, an amorphous claim of trial delay alone is insufficient to deny a defendant their constitutional right to present a defense.
To portray the trial court's ruling as delay-driven, the majority relies mostly on the trial court's comment that defendant" 'really needed to raise this issue a while ago, and perhaps it might have been a different result'" (majority op at 6). But that remark is nothing more than speculation that if defense counsel had more time, he might have persuaded the trial court to permit expert testimony on the three factors presented on appeal. Contrary to the majority, the remark does not suggest timeliness grounds for the trial court's exclusion of the testimony. Rather, the trial court made this comment after explaining that it was denying defense counsel's request after "reading a number of cases" and concluding that "it's really within the discretion of the [c]ourt with respect to denying [expert evidence] when [the identification] is corroborated by other evidence in the case."
The majority accurately observes that we have never concluded that a court "abused its discretion in excluding or limiting expert testimony of this sort on" what it calls "such a perfunctory application brought at this juncture of the proceedings" (majority op at 11). The majority also points to other cases featuring what it terms "a properly timed motion" as "illustrat[ing] the importance of" such, but the decisions it references-unlike the one on appeal here-did not hinge on either the punctiliousness or timing of the defendants' motions at issue in them, but on whether the trial courts there erred in rejecting those motions on their merits (id. at 12, citing Santiago, 17 N.Y.3d at 672; Abney, 17 N.Y.3d at 672; LeGrand, 8 N.Y.3d at 458). The majority's mere correlation of certain fact patterns with certain outcomes is insufficient to alter or displace the legal principles that actually drove those outcomes in those cases.
Second, even if unreasonable trial delay could justify rejecting expert testimony on a matter as significant as eyewitness reliability in a case where the perpetrator's identity is disputed, the record here does not establish any judicial findings of a lengthy delay that would support limiting the expert testimony as done here. Undue delay is a valid basis for excluding relevant evidence when the evidence's" 'probative value is outweighed by the danger that its admission would prolong the trial to an unreasonable extent without any corresponding advantage'" (Davis, 43 N.Y.2d at 27, quoting Jerome Prince, Richardson on Evidence § 147 [10th ed 1973]). Here, there was no discussion about how much longer either side would need if such testimony was permitted. One would expect that if the trial court was going to deny the testimony on a procedural rather than substantive ground, that the court would first hear from counsel on the cause and potential length of any delay, and whether any delay was unreasonable. Indeed, given the trial court's approval of the expert's testimony on cross-racial effects, it is unclear how permitting this same expert to explain, during the course of his same appearance, the impact of the other relevant effects would raise "the prospect of unduly delaying the trial" (majority op at 10). Moreover, the testimony provided an obvious "corresponding advantage" since it would have educated the jurors on matters counterintuitive to their common understanding of the way human memory works-information, in other words, outside the ken of the jurors and essential to the truth-seeking function of the trial (Davis, 43 N.Y.2d at 27 [internal quotation marks omitted]). Today's decision penalizing defendants for a trial court's unfamiliarity with the law incentivizes defense attorneys to, out of an abundance of caution, request Frye hearings in every case involving an eyewitness identification, even when precedent supports admission of the expert testimony. Irrespective of the point at which they make such requests, the result will be unnecessarily prolonged criminal proceedings and wastes of valuable time and resources-byproducts the majority presumably would find objectionable (see majority op at 11-12 & n 4).
The majority "presume[s] that this judge was aware of the anticipated length of delay a midtrial Frye hearing would create" (majority op at 12-13). That presumption is in tension with the majority's refusal to "presume" that trial courts are familiar with the law regarding the admissibility of expert testimony on factors affecting the reliability of eyewitness identifications. The majority's resort to simply assuming away this weakness in its reasoning as a basis for denying defendant the chance to present this valuable expert testimony is troubling.
The record belies the majority's suggestion that" 'undue prejudice to the opposing party, confusing the issues or misleading the jury'" played a role in the trial court's ruling, in addition to the prospect of delay (majority op at 1-2, quoting People v Primo, 96 N.Y.2d 351, 355 [2001]; see also id. at 8, 13). But more to the point, it is difficult to imagine how, in a case that hinged so heavily on eyewitness identifications, admission of relevant, Frye tested expert testimony concerning factors that undermine the reliability of such identifications would inject prejudice, confuse the issues, or mislead the jury-certainly no more so than expert testimony on cross-racial effect, which the trial court rightly admitted.
The actual basis for the trial court's ruling-that the expert testimony was unnecessary-was wrong on the law. Just before the defense's expert testified about the effects of cross-racial identifications, the trial court informed counsel that it had "been reading a number of cases... [o]n this identification testimony" which convinced it that it had ruled correctly when defense counsel first confirmed his intent to present expert testimony about the other factors as well. But the trial court misunderstood the law pronounced in a "number of cases... [o]n this identification testimony" that it had "been reading." As the majority states, "our precedent establishes that the correlation between confidence and accuracy is generally accepted in the relevant scientific community and beyond the ken of the average juror" (majority op at 11). Nevertheless, the majority unpersuasively holds that the trial court acted within its discretion when its ruling overlooked a constellation of precedents of which it could have taken judicial notice and admitted the testimony (see Williams, 35 N.Y.3d at 38).
Strikingly, the majority faults defendant for failing to furnish the trial court with case law supporting his position (see majority op at 11). But, the trial court based its conclusion on its own research and reading of the cases. And of course, trial judges are expected to remain current on "[j]udicial precedent" that "may... support a conclusion with respect to the general reliability of a disputed scientific technique" (Williams, 35 N.Y.3d at 38). Indeed, as a general matter, "[t]rial judges are presumed to know the law and to apply it in making their decisions" (Walton v Arizona, 497 U.S. 639, 653 [1990], overruled on other grounds, Ring v Arizona, 536 U.S. 584 [2002]; see also Bell v Cone, 543 U.S. 447, 455 [2005] [observing that, on federal habeas review, a court should not "presume so lightly that a state court failed to apply its own law"]). Thus, defendant is no more at fault than a lawyer who presents cases that fail to persuade the judge. Had the trial court here known the law, its "concerns at least with respect to some of the factors" (majority op at 11), including those raised on this appeal, would have been dispelled.
Finally, the majority is wrong inasmuch as it approves of the trial court's reliance on the surveillance footage and photo of the brown sweatshirt as sufficiently corroborative to preclude testimony about stress, weapon focus, and the pitfalls of witness confidence (see majority op at 7-10). For starters, the majority fails to justify the obvious incongruity in the trial court's ruling which regarded the prosecution's corroborating evidence sufficient to disallow testimony on those three subjects, but insufficient to preclude testimony on cross-racial effects. In addition, that evidence was weak because the surveillance footage failed to capture any of defendant's distinguishing identifying features and the ubiquity of brown sweatshirts similar to the one defendant wore when police arrested him.
IV.
The majority also attempts to recast the role that the prosecution's evidence corroborating the eyewitness identification at issue plays in the admissibility of expert testimony regarding the factors affecting the reliability of such identifications. As an initial matter, since the majority's holding is on different grounds from that of the trial court and Appellate Division-namely, the trial court's purportedly-overriding interest in avoiding trial delay (majority op at 2), its commentary on corroboration is "not necessary to the decision" here and therefore should be construed as obiter dicta (McCormick v Pennsylvania Cent. R. Co., 49 NY 303, 309 [1872]; see also United States v Rubin, 609 F.2d 51, 69 n 2 [2d Cir 1979] [Friendly J., concurring] ["A judge's power to bind is limited to the issue that is before (them); (they) cannot transmute dictum into decision by waving a wand and uttering the word 'hold' "]).
See also Hon. Pierre N. Leval, Madison Lecture: Judging Under the Constitution: Dicta about Dicta, 81 NYU L Rev. 1249, 1257 (2006) "To identify dictum, it is useful to turn the questioned proposition around to assert its opposite, or to assert whatever alternative proposition the court rejected in its favor. If the insertion of the rejected proposition into the court's reasoning, in place of the one adopted, would not require a change in either the court's judgment or the reasoning that supports it, then the proposition is dictum. It is superfluous. It had no functional role in compelling the judgment."
Regardless, as I discuss below, the majority's "clarification" fails to address a fundamental problem of the "corroborating evidence" rule. Under existing case law, New York courts may consider whether eyewitness evidence controverted by the defendant corroborates the prosecution's case. Where a court must make a factual or credibility determination solely within the province of the trier of fact, such evidence can never be judicially deemed reliable for purposes of precluding expert testimony on eyewitness identification. In other words, eyewitness testimony cannot be self-corroborating when the defendant challenges the testimony based on the scientifically-identified risk factors that impact witness memory. Therefore, the trial court should not have applied the rule here.
Amici The Innocence Project and the Legal Aid Society recommend abandonment of the corroboration rule altogether on the grounds that it contravenes traditional evidence principles and abridges the constitutional right of criminal defendants to present a complete defense. I note that defendant does not adopt this position and therefore the continued propriety of the rule is not at issue on this appeal.
A.
Applying the traditional Frye standard, the Court has sanctioned the introduction of expert testimony on a vast array of subjects, including, for example: the effect of drugs and alcohol "on a person's ability to act purposefully" (see People v Cronin, 60 N.Y.2d 430, 433 [1983]); the effects of rape trauma syndrome on complaining witnesses (see People v Taylor, 75 N.Y.2d 277, 289 [1990]); the adequacy of building security measures to deter crime (see Price ex rel. Price v New York City Hous. Auth., 92 N.Y.2d 553, 558-559 [1998]); child sexual abuse syndrome (see People v Carroll, 95 N.Y.2d 375, 387 [2000]); and whether a fire was deliberately set (see People v Rivers, 18 N.Y.3d 222, 228 [2011]). Not so for testimony regarding the reliability of eyewitness identifications, however. For that, the Court has special rules developed over the past two decades.
After initial reticence on the subject (see e.g. People v Mooney, 76 N.Y.2d 827, 828 [1990]; see also id. at 832 [Kaye, Ch. J., dissenting]), the Court first decided in Lee that expert testimony on the subject is proper under the Frye standard (see 96 N.Y.2d at 160-162). But rather than treat this like other expert testimony subject to traditional relevancy principles, the Court concluded that a trial court's discretion in admitting it must be guided by "other relevant factors," including most pertinently "the existence of corroborating evidence" (id. at 163). The Court drew on this reasoning five years later in Young, which involved a victim of a home invasion who identified defendant as her attacker even though the attacker wore a scarf on their face (see 7 N.Y.3d at 42). The Young Court acknowledged that expert testimony on the reliability of eyewitness identifications-specifically the effects of race differences and witness confidence-"[c]ertainly... could have been valuable to a juror" because it was "counter-intuitive" (id. at 45). Nevertheless, the Court agreed that this relevant testimony was inadmissible given "[t]he corroborating evidence" that "significantly diminishe[d] the importance of the proffered expert testimony"-specifically, testimony from defendant's "acquaintances" that the defendant gave them some gloves and binoculars-both items stolen from the victim-a month after the break-in (see id. at 43).
Then in People v LeGrand, the Court set out the test still applied today:
"[W]here the case turns on the accuracy of eyewitness identifications and there is little or no corroborating evidence connecting the defendant to the crime, it is an abuse of discretion for a trial court to exclude expert testimony on the reliability of eyewitness identifications if that testimony is (1) relevant to the witness's identification of defendant, (2) based on principles that are generally accepted within the relevant scientific community, (3) proffered by a qualified expert and (4) on a topic beyond the ken of the average juror" (8 N.Y.3d 449, 452 [2007]).
Almost ten years later, the Court in People v McCullough, announced in a 4-3 decision that the lower courts had misunderstood LeGrand and the corroboration rule as follows:
"To the extent LeGrand has been understood to require courts to apply a strict two-part test that initially evaluates the strength of the corroborating evidence, it should instead be read as enumerating factors for trial courts to consider in determining whether expert testimony on eye-witness identification" 'would aid a lay jury in reaching a verdict'" (27 N.Y.3d 1158, 1161 [2016], quoting Lee, 96 N.Y.2d at 162).
The clarification apparently didn't stick and so the Court now tries again. Courts determining the admissibility of the type of expert testimony at issue here, the majority says, must "weigh the [evidence]'s probative value against its prospect of causing undue prejudice to the opposing party, confusing the issues, misleading the jury, or unduly delaying trial" (majority op at 8, citing, inter alia, McCullough, 27 N.Y.3d at 1161). However, like the McCullough majority, today's majority cautions that courts do not decide admissibility based solely on the existence or the strength of corroborating evidence, but rather "should be guided by 'whether the proffered expert testimony would aid a lay jury in reaching a verdict'" (id. at 8, quoting Lee, 96 N.Y.2d at 162; compare id., with McCullough, 27 N.Y.3d at 1161). I agree, but only as to the majority's reiteration of the Frye standard for admitting expert testimony generally (see Nicholson, 26 N.Y.3d at 828), which courts apply before considering alleged corroboration. I disagree to the extent the majority's rule would allow a court to consider as corroborative evidence eyewitness testimony which is the subject of the proffered expert's opinion.
B.
Whether the strength of alleged corroborating evidence is considered as a dispositive first step or one factor among others that the trial court may weigh in deciding the admissibility of proffered expert testimony on eyewitness reliability, the rule deprives a defendant of their right to challenge the eyewitness testimony before the factfinder. It thus enhances the risk of wrongful convictions that LeGrand and its progeny sought to eliminate. In McCullough, I explained why corroboration by the witness(es) offering the identification, untested by a jury, cannot serve as a ground for excluding expert testimony regarding factors that might undercut the reliability of the identification and, in turn, the corroboration itself. But, since the majority has doubled down on McCullough 's misapplication of the corroboration rule, I must explain why the Court was wrong then and, as illustrated by the increased consensus since (see supra, § II.A), is wrong now.
Permitting a trial court to consider the strength of an eyewitness' testimony for purposes of corroborating the prosecution's case when determining the admissibility of a defendant's otherwise relevant, Frye- compliant expert testimony on factors affecting the reliability of eyewitness identifications, contravenes traditional principles of evidence. First, it invades the province of the jury by requiring trial courts to engage in trial-level factfinding. Relevant evidence is admissible unless some other rule requires its exclusion (see People v Harris, 26 N.Y.3d 1, 5 [2015] ["Generally, 'all relevant evidence is admissible unless its admission violates some exclusionary rule' "], quoting People v Scarola, 71 N.Y.2d 769, 777 [1988]; Badger v Badger, 43 Sickels 546, 559, 88 NY 546, 559 [1882] ["Being evidence, relevant and material, it was admissible unless some rule of law excluded it"]). Indeed, "[t]he basic rule is that all competent, substantial, credible and relevant evidence is to be available to the courts" as "the process of investigating the truth in courts of justice is an indispensable function of society" (Sackler v Sackler, 15 N.Y.2d 40, 44 [1964]). Evidence, the Court has said, "is relevant if it has any tendency in reason to prove any material fact" (People v Alvino, 71 N.Y.2d 233, 241 [1987]). With that relevant evidence in hand, the jury's job is then to "assess the reliability of eyewitnesses, aided by cross-examination, by the arguments of counsel, and by whatever other evidence supports or contradicts the witnesses' testimony" (People v Marte, 12 N.Y.3d 583, 589 [2009]).
Other exclusionary rules of evidence that authorize courts to keep relevant evidence from the jury's view are grounded in some intrinsic quality of the evidence itself that impairs the truth-seeking function of trials. Some of the rules set threshold levels of reliability. Others aim to keep the size of the evidentiary sphere managed and the jury undistracted from its job to render a finding regarding the discrete facts at issue. Courts may, for example, curtail "the exploration of collateral issues" because it "tends 'to obscure the main issue in the minds of the jury, to lead them away from the principal matters which require their attention and to protract trials to an unreasonable extent without any corresponding advantage to any one concerned'" (People v Oddone, 22 N.Y.3d 369, 379 [2013], quoting People v Harris, 209 NY 70, 82 [1913]). Similarly, there are rules designed to mitigate the risk of faulty assumptions or ad hominem jury determinations like those forbidding" 'evidence of other crimes'" that" 'has no purpose other than to show that a defendant is of a criminal bent or character and thus likely to have committed the crime charged'" (People v Weinstein, __ N.Y.3d __, 2024 NY Slip Op 02222 at *8 [Ct App Apr 25, 2024], quoting People v Schwartzman, 24 N.Y.2d 241, 247 [1969]). More broadly, courts may exclude evidence whose probative value fails to "exceed[ ] the potential for prejudice" (Alvino, 71 N.Y.2d at 242).
Some rules permitting exclusion of relevant evidence concern the evidence's inherent lack of reliability. Courts exclude hearsay, for example because, except under certain well-known circumstances, untested, out-of-court statements are generally unreliable (see Nucci ex rel. Nucci v Proper, 95 N.Y.2d 597, 602 [2001]). The same goes for evidence offered by a party that fails to establish its authenticity, which courts may exclude because it is inherently untrustworthy (see People v Price, 29 N.Y.3d 472, 476 [2017]). And, of course, as outlined supra, courts may bar expert testimony if the party offering it fails to show that "the accepted techniques, when properly performed, generate results accepted as reliable within the scientific community generally" (Wesley, 83 N.Y.2d at 422, citing Frye, 293 F at 1014 [emphasis added]).
Our rule authorizing a trial court to consider the strength of the prosecution's eyewitness as part of the alleged corroborating evidence in deciding whether to exclude a defendant's otherwise admissible expert testimony on the reliability of that same eyewitness is an anomaly that fits within none of these categories. Permitting courts to treat as a factor the strength of the prosecution's corroborating evidence before admitting expert testimony on eyewitness identifications perversely withdraws this core jury function from jurors and reassigns it to judges. The Court has established that this type of expert evidence qualifies for a judicial imprimatur of reliability under Frye, but nonetheless has extended trial courts the discretion to exclude it anyway if, among other things, "the corroboration" of the identification offered by the prosecution "was strong enough for the trial court reasonably to conclude that the expert's testimony would be of minor importance" (Young, 7 N.Y.3d at 45). However, this analysis-evaluating the weight and significance that eyewitness testimony serves to one party or another's theory of the case-is the classic province of the jury, a factual and credibility determination it makes, guided by the court's instructions (see Marte, 12 N.Y.3d at 589).
The corroboration rule also empowers judges to place their thumbs on the scale by enabling them to credit the prosecution's alleged corroborating evidence before the jury has had an opportunity to accept or reject it. If time-tested rules of evidence and procedure applied, the jury would evaluate the strength of the testimony during its deliberations after both sides present their cases, including by testing the credibility and reliability through cross examination and expert opinion (see New York Guar. & Indem. Co. v Gleason, 78 NY 503, 504 [1879] ["Whether more weight should be given to the testimony of two accomplices, and whether and to what extent they corroborate each other are matters for the jury"]; Davis v Alaska, 415 U.S. 308, 316 [1974] [Cross-examination is the principal means by which the believability of a witness and the truth of (their) testimony are tested"]; People v Falletto, 202 NY 494, 499 [1911] [describing "cross-examination" as "the best method known to bring out the full and exact truth"]). And there is good reason to believe that, in some of the Court's eyewitness-identification-expert cases, the jury might not have credited the prosecution's evidence of corroboration had the trial court not weighed in first and excluded relevant and helpful expert opinion on the unreliability factors.
McCullough demonstrates the dangers of allowing judges to weigh the reliability of the eyewitness when considering the admissibility of a defense expert on such eyewitness identifications (see 27 N.Y.3d at 1158). There, the trial court barred the testimony, reasoning that the eyewitness's "identification of [the] defendant as a participant," in an attempted robbery during which a co-defendant shot the victim "coupled with [a getaway driver's] testimony connecting [the] defendant to the incident, would constitute sufficient corroboration" (id. at 1160). Defendant was thereafter convicted of murder (id. at 1161). On appeal, the Appellate Division reversed, concluding that the trial court had abused its discretion when it concluded that the driver's identification constituted sufficient corroboration under LeGrand to bar the expert testimony (see People v McCullough, 126 A.D.3d 1452, 1453 [4th Dept 2015]). This Court then reversed and reinstated the verdict after retooling LeGrand in the manner discussed supra (see McCullough, 27 N.Y.3d at 1161). As I noted in my dissent, several factors weighed against deeming the getaway driver's corroboration sufficient to bar the expert testimony-i.e., the driver's initial, inaccurate implication of his brother and cousin in the crime and the driver's motive to tailor his account to the prosecution's theory to keep his favorable plea deal in place (see 27 N.Y.3d at 1169-1170 [Rivera, J., dissenting]). Of course, the only way the trial court could have barred the expert testimony in McCullough was by crediting the corroboration itself before the jury had a chance to weigh it. But we cannot know whether the jury agreed with the trial court's assessment and, if it did not, then there is a substantial probability that testimony from an eyewitness-identification expert would have affected the jury's findings.
McCullough also illustrates how the corroboration rule ignores that juries do not assess each piece of evidence in a vacuum. The significance of one piece often turns on the significance of another, and juries are free to reason through bodies of evidence in different directions and assemble the pieces in different ways. Who can say that the jury in McCullough, instead of crediting the eyewitness's identification because of the driver's corroboration-as the trial court did-instead credited the driver's corroboration because of the eyewitness's identification? If the jury took this latter course, then an eyewitness identification expert might well have undermined both the eyewitness's and the driver's testimonies. In that respect, barring the defense's expert testimony on the reliability of eyewitnesses may result in improper judicial bolstering of the prosecution's evidence.
Finally, we have required strict adherence to these same traditional evidence principles before in this very same area of the law. In Boone, we held that, "when identification is an issue in a criminal case and the identifying witness and defendant appear to be of different races," trial courts must, upon request, instruct the jury regarding the cross-racial effects on eyewitness identifications given the higher "likelihood of misidentification" and the "significant disparity between what the psychological research shows and what uninstructed jurors believe" regarding the effect's impact (30 N.Y.3d at 526, 535-536). We recognized that "[t]he instruction would not be required when there is no dispute about the identity of the perpetrator" (id. at 536). We cautioned, however, that
"a trial judge may not decide the purely factual question of whether a crime was so prolonged and a witness's exposure to the perpetrator so extended that the witness would have adequately learned the perpetrator's features so as to be able to recognize him accurately. It is the jury's task to examine and evaluate the many factors upon which the accuracy of [identification] testimony turns including, among others, the... opportunity and capacity to observe and remember the physical characteristics of the perpetrator at the time of the crime. A judge should not engage in a weight analysis as to the quantitative degree of risk of misidentification at this threshold stage" (id. at 536 n 6 [emphases added] [internal quotation marks and citations omitted]).Boone' s core holding that relevance alone controls the availability of an instruction on cross-racial effect does not support a trial court's consideration of alleged corroboration evidence "at this threshold stage" of determining the admissibility of a defense expert on eyewitness identifications (id.). No principled justification exists for, on the one hand, allowing judicial weight analyses to bar a defense expert from offering relevant, Frye- compliant expert testimony regarding eyewitness identifications while, on the other hand, expressly disapproving of the same for assessing the need for an instruction on cross-racial effect. In addition, consideration of disputed eyewitness testimony as self-corroborating may also be unconstitutional. The Constitution leaves courts with ample breathing room to regulate the entry and exclusion of evidence (see Crane v Kentucky, 476 U.S. 683, 689-690 [1986]). But it also enshrines some limits on this discretion. "Whether rooted directly in the Due Process Clause of the Fourteenth Amendment or in the Compulsory Process or Confrontation Clauses of the Sixth Amendment, the Constitution guarantees criminal defendants 'a meaningful opportunity to present a complete defense'" (id. at 690, quoting California v Trombetta, 467 U.S. 479, 485 [1984]). This constitutional guarantee precludes federal and state courts from: (1) applying the rules of evidence "mechanistically to defeat the ends of justice" (Chambers v Mississippi, 410 U.S. 284, 302 [1973]); and (2) devising such rules that are "arbitrary or disproportionate to the purposes they are designed to serve" (see Rock v Arkansas, 483 U.S. 44, 56 [1987] [invalidating rule prohibiting defendants' hypnotically-refreshed testimony wholesale]).
Indeed, this leads to the absurd result where a defendant might be entitled to an instruction on the impact of cross-racial effect on eyewitness identification under Boone (30 N.Y.3d at 535-536) but judicially barred from educating the jury about cross-racial effect through an expert on the subject, based on a nonjury finding that the corroboration was strong. The record here shows that this is not a mere thought exercise. Despite our clear position in Boone, the trial court had reservations about the necessity of the expert testimony regarding the reliability of cross-race identifications, commenting that the "expert... is going to be teaching the jury something that they know" and that it "d[id]n't find it a big loss to [defendant] if this witness didn't testify at all" but, as a purported act of grace, allowed the testimony "to give the defense latitude."
The Second Circuit has held that a trial court's exclusion of defense evidence violates the Constitution any time "the omitted evidence evaluated in the context of the entire record creates a reasonable doubt that did not otherwise exist" (Washington v Schriver, 255 F.3d 45, 56 [2d Cir 2001]).
The Supreme Court has sent a strong signal that the Constitution forbids courts from treating the strength of the prosecution's case at the time the defendant attempts to introduce evidence supporting a defense as a factor in that evidence's admissibility. In Holmes, the Court unanimously held unconstitutional a rule announced by the South Carolina Supreme Court under which, in considering whether to admit third-party culpability evidence proffered by defendants, trial courts had discretion to consider whether there was" 'strong evidence of [the defendant's] guilt'" and could exclude the defendant's evidence suggesting an alternative perpetrator if it" 'did not raise 'a reasonable inference' as to [the defendant's] own innocence' '" (547 U.S. at 328-329, quoting State v Gay, 343 S.C. 543, 550, 541 S.E.2d 541, 545 [2001], quoting State v Gregory, 198 S.C. 98, 104 16 S.E.2d 532, 534 [1941]). In the case on appeal to the Supreme Court, the South Carolina Supreme Court articulated the rule to be that, "where there is strong evidence of [a defendant's] guilt, especially where there is strong forensic evidence, the proffered evidence about a third party's alleged guilt does not raise a reasonable inference as to the appellant's own innocence" which vested the trial court with discretion to exclude it (State v Holmes, 361 S.C. 333, 342-343 [2004]; see also Gay, 343 S.C. at 550).
"Under this rule, the trial judge does not focus on the probative value or the potential adverse effects of admitting the defense evidence of third-party guilt. Instead, the critical inquiry concerns the strength of the prosecution's case: If the prosecution's case is strong enough, the evidence of third-party guilt is excluded even if that evidence, if viewed independently, would have great probative value and even if it would not pose an undue risk of harassment, prejudice, or confusion of the issues" (id. at 329).
The Court further summarized the invalid South Carolina rule's "logic" as such:
"Where (1) it is clear that only one person was involved in the commission of a particular crime and (2) there is strong evidence that the defendant was the perpetrator, it follows that evidence of third-party guilt must be weak" (id. at 330).
The majority cites Holmes to support its proposition that a trial court may not decide admissibility "based solely on the existence or the strength of corroborating evidence," but may instead treat it as a factor (majority op at 9). But that squarely conflicts with the Holmes Court's reasoning. Holmes held that the South Carolina rule violated the Constitution because it "depends on an accurate evaluation of the prosecution's proof, and the true strength of the prosecution's proof cannot be assessed without considering challenges to the reliability of the prosecution's evidence" (id.). "Just because the prosecution's evidence, if credited, would provide strong support for a guilty verdict," the Court continued, "it does not follow that evidence of third-party guilt has only a weak logical connection to the central issues in the case" (id.). Furthermore-and perhaps most critically-"the strength of the prosecution's case cannot be assessed without making the sort of factual findings that have traditionally been reserved for the trier of fact" (id.). In other words, "by evaluating the strength of only one party's evidence, no logical conclusion can be reached regarding the strength of contrary evidence offered by the other side to rebut or cast doubt" (id. at 331). Certainly, there is no persuasive justification for departing from Holmes for the type of expert testimony at issue here. Like the rule at issue in Holmes, the corroboration approach the majority endorses (see majority op at 7-10) is at odds with Holmes's core concerns about judges-instead of juries-evaluating the strength of the prosecution's case (see 547 U.S. at 330-331).
As one New York jurist has suggested, Holmes throws the constitutionality of permitting consideration of the prosecution's corroboration completely into doubt (People v Santiago, 75 A.D.3d 163, 177 [1st Dept 2010] [McGuire, J., concurring], revd, 17 N.Y.3d at 673 [holding, preMcCullough, that the prosecution's corroborating evidence was not "sufficient to obviate the second stage of the LeGrand analysis"]).
Dispensing with corroboration as a factor in the admissibility of this expert evidence would have no effect on the ability of courts to disallow such expert testimony under circumstances where the normal rules of evidence would justify exclusion. As an easy example, if the defense seeks to admit expert testimony on the reliability of eyewitness identifications where the prosecution has no evidence of such an identification to offer, the court can exclude it as irrelevant (see Harris, 26 N.Y.3d at 5). So too can courts insist on live testimony from such experts under the hearsay rule, thereby giving the prosecution a fair opportunity to cross-examine them (see Nucci ex rel. Nucci, 95 N.Y.2d at 602). Courts simply would be prohibited from considering the strength of the prosecution's alleged corroborating evidence as a factor in admissibility (see Holmes, 547 U.S. at 330-331) in the same manner as in the cross-racial identification context, where the Court has appropriately cautioned that "[a] judge should not engage in a weight analysis as to the quantitative degree of risk of misidentification" when deciding whether to admit expert testimony from a defense witness on the subject (Boone, 30 N.Y.3d at 536 n 6).
V.
In a case that turned on eyewitness identification, the trial court here precluded defendant's expert from helping the jury to understand factors present in this case that the scientific community agrees have a counterintuitive impact on human memory and thus impair the reliability of eyewitness testimony. The factors at issue here-the witness's stress, weapon focus, and confidence relative to accuracy-had been recognized by courts years before defendant's trial, as proper bases for expert testimony. The trial court's exclusion of it was therefore an error that warrants reversal and a new trial.
The majority's affirmance here is contrary to law and modern science and, in turn, greatly disserves our system of justice and undermines this Court's role in minimizing the leading cause of wrongful convictions: mistaken identifications. The responsibility now falls to trial judges on the frontlines to exercise their discretion to ensure that juries are well-informed about the potential unreliability of eyewitness testimony.
Order affirmed. Opinion by Judge Troutman. Chief Judge Wilson and Judges Garcia, Singas, Cannataro and Halligan concur. Judge Rivera dissents in an opinion.