by Zoe Farkas*

Historically, eyewitness identifications have been considered the gold standard of trial evidence. There’s little that’s more convincing than a witness on the stand confidently pointing at a defendant and proclaiming, under oath, “that’s the one!” However, over the last half century it has become clear that eyewitness identification may actually be one of the most fallible evidentiary tools, despite common misconceptions of its accuracy. Even in the face of growing research demonstrating the unreliability of eyewitness identification, courts have been slow to allow experts to testify to that unreliability in the courtroom. Judges instead bar them as unqualified or unhelpful under Federal Rule of Evidence 702. This Contribution argues that these experts are not only qualified and helpful, but absolutely essential to help juries fulfill their fact-finder duties.


Federal Rule of Evidence 702 provides that a witness may only offer opinion testimony if they are an expert and sets out the standard for expert testimony, requiring that:

(a) the expert’s scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; and (d) the expert has reliably applied the principles and methods to the facts of the case.1

The Supreme Court expanded upon this rule in Daubert v. Merrell Dow Pharmaceuticals, where it established a two-prong test for district courts to apply when faced with a proposed expert witness: “whether the expert is planning to testify to (1) scientific knowledge that (2) will assist the trier of fact to understand or determine the facts in issue.”2 The Daubert Court made clear that the objective of this test was to ensure reliability and relevance of the proposed expert testimony, as such testimony can determine the outcome of a case.3

A circuit split has emerged in recent years over whether the testimony of eyewitness identification experts is admissible.4 Experts with specialized knowledge about the credibility of witness memory are often enlisted by the defense in cases where an eyewitness has identified the defendant. Historically, courts have been hesitant to allow these experts to testify.5 However, courts are gradually becoming more amenable to their testimony. As one court explained: “It may be that a door once largely shut is now somewhat ajar.”6 This Contribution will demonstrate why trial courts should open that door wider, thereby allowing juries to fulfill their role as factfinders.

Under the first prong of Rule 702, a trial court must determine whether an expert’s testimony constitutes scientific or other experience-based knowledge.7 In deciding whether the evidence the expert intends to rely upon is scientific, the trial court can consider: (1) whether the theory or technique can be tested, (2) whether the theory or technique has been subject to peer review or publication, (3) the known or potential rate of error, and (4) the particular degree of acceptance within the scientific community.8 Eyewitness reliability experts come from social science backgrounds, and therefore decisions regarding these types of witnesses vary widely because courts will often differ in how they apply the Daubert factors. Judges who choose to apply the Daubert factors strictly under the rationale that psychology is a type of science may reject an expert because, for example, they cannot test their technique on an individual basis with the eyewitness whose reliability is at issue or there isn’t a calculable rate of error on an individual level. However, in Kumho Tire Co. v. Carmichael, just six years after Daubert, the Court clarified that these were not strict elements, but rather factors to be weighed.9 The Kumho Court emphasized that the factors used in assessing reliability can be flexible based on the content and nature of the issue.10 Therefore, if certain Daubert factors are not relevant to the expertise being considered by the court, they should not be applied.11 The factors serve as guideposts and should not be treated as a one-size fits all standard, which courts rejecting these experts have failed to recognize.12

While it is true that the evidence relied upon by eyewitness reliability experts cannot be validated in the same way as traditional “hard sciences” can, the body of research in the area is now robust enough to support a general consensus in the scientific community about the limits of eyewitness testimony and exacerbating factors that worsen reliability. The Second Circuit recognized this reality in Young v. Conway, where it held that such research “has been reviewed, replicated, and retested, and is generally accepted in the scientific community.”13 The court in Conway relied upon research presented by the Innocence Project that concluded that the presence of a weapon, the stress of the situation, an eyewitness being supplied with post-incident information that could solidify false beliefs, the identification being cross-racial, the witness’s exposure to defendant through multiple identification procedures, and the passage of time between observation and identification, among other factors, can all impair a witness’s ability to accurately identify someone.14 The evidence demonstrates that specific yet common circumstances may worsen an eyewitness’s already unreliable memory; therefore, even under a stricter Daubert analysis, this science should still be admissible.

After decades of mass incarceration, it is more vital than ever for courts to take a hard look at the past failures of the criminal justice system and do what they can to ensure juries come to the verdict justice requires. Although the causes of mass incarceration are abundant and interconnected, wrongful convictions contribute significantly to this problem.15 The disturbing reality that as many as one in twenty people incarcerated in the United States are innocent16 is driven in large part by mistaken identifications.17 Many courts have recognized this injustice and have become more receptive to the science behind expert testimony on witness identification.18

The second prong of Rule 702 requires that the expert testimony be helpful to the jury.19 The emerging research in the field establishes that juries absolutely need this information to reach an informed verdict, and when they are deprived of it, the consequences are a stain on the justice system. Despite this reality, judges continue to reject these experts as improper or unhelpful to juries in discharging their duties.

One reason judges frequently cite for rejecting these experts is that determining a witness’s credibility is the responsibility of the jury alone. However, it is becoming increasingly clear that, without this information, jurors are unable to make credibility assessments of eyewitnesses. Over 50 years ago, the Supreme Court acknowledged the enduring problems with eyewitness testimony in United States v. Wade, stating that “[t]he vagaries of eyewitness identification are well-known; the annals of criminal law are rife with instances of mistaken identification.”20 The data produced since Wade has only provided further support for this concern.

Whether the substance of a proposed expert’s testimony is considered common sense is often a determinative factor for judges in their Rule 702 analyses.21 While the general notion of unreliable memory may be well known, the specific mechanics of degenerative memory, the contributing factors, and the extent to which certain factors can impact memory, are not. Factors that are known to impact the forming of memories are often present in criminal trials, for example cross-racial identifications, extreme stress, and weapons focus. The average juror would not be versed in how these factors contribute to an eyewitness’s unreliability and may even believe the opposite.22

Similarly, lay jurors are not generally aware, nor is it obvious, that there is a very weak correlation between confidence and accuracy of identifications. The common notion of unreliable memory evokes an image of someone struggling to remember what they saw, not someone confidently asserting their false beliefs under oath. For this reason, cross-examination, which is often employed to provide evidence of a witness’s lack of credibility to the jury, is insufficient.23 Assuming lay jurors know not to emphasize confidence in determining the credibility of an eyewitness is dangerous and leads to severe miscarriages of justice. According to the Innocence Project, eyewitness misidentifications were present in an overwhelming 71% of the cases in which subsequent DNA testing established a wrongfully convicted defendant’s factual innocence,24 and those are only cases where exoneration was possible.

Courts that exclude experts often reason that cross-examination and jury instructions can remedy the issue. But those less rigorous mechanisms are simply ineffective. Cross-examination will rarely be helpful because a mistaken eyewitness is wrong, but not lying. And jury instructions are simply a judge, who is a layperson in this context, telling other laypeople about complicated psychology without the benefit of expanded questioning and clarification on cross-examination. In these circumstances, there is a lot of room for mistakes, misunderstandings, miscommunications, and gaps of knowledge to wreak havoc. Having an empirically qualified expert on the stand to elucidate the information and concede weaknesses in conclusions through cross-examination from an adversary is much more effective at achieving an equitable outcome.

Another concern often cited by opponents of the admissibility of eyewitness identification experts is the “aura of special reliability and trustworthiness” surrounding an expert witness that has the potential to lend them unearned credibility with a jury.25 While there is a valid concern that juries will simply listen to an expert because they seem authoritative, the same could be said for eyewitness testimony that is already given much more credibility than it is due. In fact, “[a]ll the evidence points rather strikingly to the conclusion that there is almost nothing more convincing than a live human being who takes the stand, points a finger at the defendant, and says ‘That’s the one!’”26 It is clear that solely relying on cross-examination and jury instructions against identification witness testimony has failed. Therefore, perhaps it is time we instead rely on those mechanisms to counter the qualified experts instead, either with another expert or through thorough cross-examination.

Of course, all issues of admissibility must always be governed by the circumstances of the case at issue to determine whether the evidence is substantially more prejudicial than probative, as Rule 403 requires.27 An expert’s testimony may, for example, not prevail in a Rule 403 analysis if they are only speaking to general knowledge but cannot point to any specific factor influencing the contested witness’s memory.28 Rule 403 is a rule of inclusion, not exclusion, and when the minimal potential prejudice is weighed against the probative value of this testimony, admissibility should prevail. Accordingly, trial courts should generally find that the Daubert test permits qualified, reliable eyewitness identification experts who will testify about the implications of pertinent factors present in the case by default, and only exclude such experts if the circumstances make their testimony substantially more prejudicial than probative.

The consequences of rejecting such an application of Daubert are particularly grave because once a trial judge makes a ruling regarding these experts, there is little recourse for a convicted defendant to appeal this determination. Trial judges are meant to serve a gatekeeping role in applying Rule 702 and appellate courts defer to trial court findings unless it rises to an abuse of discretion.29 This high bar requires that the trial judge’s use of discretion be “manifestly erroneous” in order to be overruled,30 and, naturally, few cases have met this high bar for reversal.

The burden is always on the government to convince a jury of a defendant’s guilt beyond a reasonable doubt. Growing evidence now definitively shows that this cannot be achieved without allowing juries to have access to the information only an expert can sufficiently provide. Eyewitness identification experts are a vital tool for defendants to get a fair trial and should be presumed essential, not prejudicial.


* Zoe Farkas is a J.D. Candidate (2023) at New York University School of Law. This Contribution is a commentary on the problem at the 2022 Prince Evidence Competition hosted by Brooklyn Law School. The question presented was whether expert testimony on the reliability of eyewitness identifications is admissible under Federal Rules of Evidence 702 and 403. This contribution presents a distillation of the side of the argument assigned to the author in the Prince Evidence Competition, and the views expressed herein do not necessarily reflect the views of the author.

1. Fed. R. Evid. 702.

2. 509 U.S. 579, 592 (1993); see also Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999).

3. Kumho Tire Co., 526 U.S. at 152.

4. Compare United States v. Lumpkin, 192 F.3d 280, 289 (2d Cir. 1999) (affirming the exclusion of expert testimony on the unreliability of eyewitness identification because it would usurp the jury’s role in determining witness credibility), with United States v. Stevens, 935 F.2d 1380, 1383 (3d Cir. 1991) (upholding the inclusion of an expert on eyewitness testimony and finding it to be neither highly suggestive nor misleading).

5. E.g., United States v. Fosher, 590 F.2d 381, 382 (1st Cir. 1979); United States v. Purham, 725 F.2d 450, 454 (8th Cir. 1984); United States v. Langan, 263 F.3d 613, 624 (6th Cir. 2001).

6. United States v. Brien, 59 F.3d 274, 277 (1st Cir. 1995) (deciding not to “adopt a blanket rule that qualified expert testimony on eyewitness identification must routinely be admitted or excluded”).

7. Daubert, 509 U.S. at 592; see also Kumho Tire Co., 526 U.S. at 151 (holding that expert knowledge can be based on experience as long as it is reliable).

8. Daubert, 509 U.S. at 593–94.

9. 526 U.S. at 150.

10. Id. at 141 (“[A]s the Court stated in Daubert, the test of reliability is ‘flexible,’ and Daubert’s list of specific factors neither necessarily nor exclusively applies to all experts or in every case.”).

11. Id.

12. See, e.g., United States v. Hudson, 884 F.2d 1016, 1024 (7th Cir. 1989) (holding eyewitness expert testimony will not aid the jury because it addresses issues of which the jury is already generally aware, despite the presence of stress on the identifying witness and a cross-racial identification).

13. 698 F.3d 69, 79 (2d Cir. 2012) (quoting State v. Henderson, 27 A.3d 872, 916 (2011) (referring to a comprehensive 86-page report based on seven experts’ testimony and hundreds of scientific studies, including many of those cited by the Innocence Project, as the gold standard in applying social science research to the law)).

14. Id. at 78­–79.

15. Claire Gilbert, Beneath the Statistics: The Structural and Systemic Causes of Our Wrongful Conviction Problem, Georgia Innocence Project (Feb. 1, 2022), https://www.georgiainnocenceproject.org/2022/02/01/beneath-the-statistics-the-structural-and-systemic-causes-of-our-wrongful-conviction-problem/ (finding as high as 6% of people incarcerated in the United States prisons are actually innocent).

16. Id.

17. Eyewitness Identification Reform, The Innocence Project, https://innocenceproject.org/eyewitness-identification-reform/ (finding mistaken identifications to be a leading cause of wrongful convictions); Richard A. Wise et al., An Examination of the Causes and Solutions to Eyewitness Error, Frontiers in Psychiatry, Aug. 2014, at 1 (estimating that as many as one in three eyewitnesses make an erroneous identification).

18. E.g., Brien, 59 F.3d at 277 (“There is more expert literature on the subject [of witness identification], more experts pressing to testify, and possibly more skepticism about the reliability of eyewitnesses.”); United States v. Stevens, 935 F.2d at 1384 (concluding that expert testimony about the weak relationship between confidence and accuracy in eyewitness identifications was admissible under circuit precedent and would have been helpful to the jury).

19. Daubert, 509 U.S. at 592.

20. 388 U.S. 218, 228 (1967).

21. See e.g., Young, 698 F.3d at 79 (observing that “many of these factors are counterintuitive and not coterminous with common sense”); United States v. Nolan, 956 F.3d 71, 81–82 (2d Cir. 2020) (“[I]t appears to us unlikely that [cross-racial identifications being less accurate than intra-racial identifications] is common knowledge among lay jurors.”).

22. Innocence Staff, How Eyewitness Identification Can Send Innocence People to Prison, (Apr. 15, 2020), https://innocenceproject.org/how-eyewitness-misidentification-can-send-innocent-people-to-prison/ (“[M]ost people believe that highly stressful events are particularly memorable . . . . However, the high levels of stress that most of us would naturally experience during a crime actually impede reliable memory formation.”); Young, 698 F.3d at 79 (stating these factors “are counterintuitive and therefore, not coterminous with ‘common sense’”).

23. Contra Fosher, 590 F.2d at 382 (1st Cir. 1979) (holding that issues concerning eyewitness credibility can be explicated more efficiently and less prejudicially defense’s thorough cross examination of the eyewitness and the court’s comprehensive jury instruction).

24. Nolan, 956 F.3d at 75 (citing Eyewitness Identification Reform, The Innocence Project, https://innocenceproject.org/eyewitness-identification-reform/).

25. Fosher, 590 F.2d at 383.

26. Watkins v. Sowders, 449 U.S. 341, 352 (1981) (Breyer, J., dissenting) (internal citation omitted).

27. Fed. R. Evid. 403.

28. See, e.g., Fosher, F.2d at 382 (rejecting an expert on memory unreliability in part because the written offer of proof proclaimed that the expert would not comment on the testimony of any named witness in the trial); cf. Brien, 59 F.3d at 276 (barring an expert witness from testifying about memory because he intended to testify generally with few details about the case and disclaimed any ability to determine if a particular witness is credible or accurate in making an identification).

29. Gen. Elec. Co. v. Joiner, 522 U.S. 136, 139 (1997) (holding abuse of discretion is appropriate standard of review for trial court decisions to admit or exclude expert testimony under Daubert).

30. Id. at 142.