13:34 PM

Striking the right balance: Mitigating the effects of eyewitness misidentification in Missouri

Vol. 75, No. 1 / January-February 2019


In January 1984, a robbery and sexual assault victim positively identified Robert Nelson from a police lineup as one of her two attackers.[2] Jackson County prosecutors charged Nelson with both crimes, and at trial the victim’s identification served as the chief evidence against him. A jury convicted Nelson of both charges, and he was sentenced to 98 years in prison. There was only one problem – the victim had misidentified her attacker.

Abigail Twenter[1]

In January 1984, a robbery and sexual assault victim positively identified Robert Nelson from a police lineup as one of her two attackers.[2] Jackson County prosecutors charged Nelson with both crimes, and at trial the victim’s identification served as the chief evidence against him. A jury convicted Nelson of both charges, and he was sentenced to 98 years in prison. There was only one problem – the victim had misidentified her attacker.Eyewitness Misidentification

Nelson, who maintained his innocence, eventually filed a successful pro se motion to obtain DNA testing and was appointed a lawyer from the Midwest Innocence Project. In August 2012, the police department reported that DNA evidence extracted from the victim’s robe matched the profile of two other convicted felons. The Midwest Innocence Project and the prosecutor’s office filed a joint motion to vacate Nelson’s sentence and dismiss the charges, stating the eyewitness had misidentified Nelson as her attacker. The court issued an order granting the motion on the grounds that Nelson was innocent. Nelson was released after serving nearly 30 years in prison.

Nelson’s case is hardly an exception. There have been 354 post-conviction DNA exonerations in the United States.[3] In more than 70 percent of these cases, the exoneree was misidentified.[4] In fact, “eyewitness misidentification is . . . responsible for more wrongful convictions than all other causes combined.”[5] Closer to home, the problem is even more prolific. In Missouri, “every wrongful conviction” case overturned “by DNA evidence involved eyewitness misidentification.”[6]

In the last 30 years, there has been a significant effort to better understand why eyewitnesses are prone to misidentifying their assailants, with more than 2,000 studies on the subject.[7] These studies reveal how the scientific understanding of memory has developed over time, with the consensus among researchers that memory is fallible, capable of being influenced by a number of factors. Scientific literature generally divides factors affecting eyewitnesses into two categories: system variables and estimator variables.[8] System variables include factors “within the control of the criminal justice system” and relate to the identification procedure itself, such as the police lineup construction or the wording of questions to eyewitnesses.[9] Estimator variables “relate[] to the witness, the perpetrator, or the event itself” and thus cannot be corrected by the State.[10] A few examples of estimator variables recognized as having a negative effect on witnesses’ ability to make an accurate identification include: (1) high levels of stress; (2) witness inattention to the event; (3) a shorter duration of exposure; (4) environmental conditions (e.g., weather conditions or distance and lighting); (5) witness intoxication; and (6) cross-race effect – the idea that witnesses are notably better at identifying members of their own race than members of other races.[11]

Though there is agreement among experts in recognizing these variables, research indicates laypersons do not intuitively understand how memory works or the various factors that might affect it.[12] One study reported that 63 percent of respondents believed . . . “memory works like a video camera, accurately recording the events we see and hear so that we can review and inspect them later.”[13] Another study indicated more than half of the approximately 1,000 respondents surveyed did not understand how system and estimator variables influence the memories of eyewitnesses.[14] As a result, jurors often find eyewitnesses extremely credible,[15] especially if they appear confident.[16] As Justice William Brennan wrote in a 1981 dissent, “[T]here 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!’”[17]

Given both the persuasive nature of eyewitnesses and the common misconceptions jurors have regarding the reliability of memory, state judiciaries are tasked with defining the role of courts in ensuring that juries give appropriate weight to eyewitness testimony. Some state judiciaries embrace a hands-off approach, confident the judicial system provides enough procedural safeguards as-is, while others are more amenable to extensive judicial intervention. Missouri falls somewhere in the middle. Overall, the Missouri judiciary seems to be open to incrementally adopting techniques to prevent convictions based on eyewitness misidentification, but wary of overcorrections. This article explores Missouri’s current stance on the two most popular methods employed by state judiciaries to educate jurors on the fallibility of eyewitness memory: jury instructions and expert testimony.

Jury Instructions on Eyewitness Identification Testimony

Missouri is among several states that have adopted jury instructions sensitizing jurors to eyewitness testimony of questionable reliability. In December 2015, the Supreme Court of Missouri adopted Missouri Approved Instruction (MAI) 310.02. This instruction, titled “Eyewitness Identification Testimony,” states “[e]yewitness identification must be evaluated with particular care” and includes several factors for the jury to consider in assessing the credibility of an eyewitness, such as “the passage of time between the witness’s exposure to the person in question and the identification of the defendant,” “whether the witness was affected by any stress or . . . the presence of a weapon,” and “whether the witness and the person in question are of different races or ethnicities.”[18] In a note on the instruction, the Court stated its adoption of MAI 310.02 addresses the concerns raised in Perry v. New Hampshire, in which the U.S. Supreme Court expressly recognized the value of state courts adopting instructions that “warn the jury to take care in appraising [eyewitness] identification evidence.”[19]

The instruction that was ultimately adopted and effective in January 2016 was not the first draft considered by the Court. In April 2015, the Court announced it was adopting a draft prepared by its Committee on Procedure in Criminal Cases.[20] The committee’s draft of the instruction included much more extensive descriptions of the various estimator and system variables that a court could instruct jurors to consider in determining “whether the witness identification of the defendant is reliable and believable or whether it is based on a mistake.”[21] For instance, rather than simply instructing jurors to pay attention to “whether the witness and the person in question are of different races or ethnicities,”[22] the April 2015 version included a long explanation on “Cross-Racial Identification.”[23]

Some people have a greater difficulty in accurately identifying members of a different race. You should consider whether the fact that the witness and the person identified are not of the same race has affected the accuracy of the witness’ identification. You should also consider whether there are other factors present in this case that overcome any such difficulty of identification.[24]

This earlier version of the instruction more closely resembled the enhanced research-based jury instruction adopted by the New Jersey Supreme Court after it “appointed a Special Master” to conduct an exhaustive survey of the current state of scientific evidence on eyewitness reliability in State v. Henderson.[25] Perhaps the most influential jury instruction on the subject – and certainly the most comprehensive[26] – the Henderson instruction is several pages long with different versions to be given depending on whether the identification took place in court, out of court, or both.[27] It reads more like expert testimony, heavily relying on specific studies and statistics to detail the three stages of memory and a comprehensive list of system and estimator variables. The New Jersey Supreme Court directed trial courts to tailor the variables to the specific facts of the case and give the charge to the jury at the close of evidence and, if warranted, during trial.[28] Advocates for the draft instruction adopted by the Supreme Court of Missouri in April 2015 felt it would more comprehensively eliminate pre-conceived notions by jurors and maximize juror competence by giving jurors all of the information on shortcomings of eyewitness memory necessary to appropriately weigh eyewitness evidence.

There were those who opposed the adoption of the April 2015 instruction, however, and the Court took comment from prosecutors, law enforcement officials, judges, and victim service providers.[29] The Missouri Association of Prosecuting Attorneys (MAPA) filed a position paper stating the new instruction “reads more like a closing argument, sowing seeds of doubt before the attorneys even have a chance to argue the case.”[30] It urged the “Court to avoid [improperly] placing a thumb on the scales of justice” by allowing trial courts to abrogate their inherent duty to refrain from commenting on the evidence in criminal jury trials.[31] In sum, MAPA believed the proposed jury instruction would so favor criminal defendants that it would have the unintended consequence of allowing guilty individuals to escape justice.

Numerous studies have been conducted on the efficacy of the Henderson instruction, and some corroborate MAPA’s theory that more comprehensive instructions can “overcorrect” the problem.[32] For example, a team of researchers at the University of Arizona recently tested the instruction by having 335 mock jurors watch a simulated murder trial and return a verdict.[33] Jurors who received the standard instruction “were more than twice as likely to convict” than those who received the Henderson instruction, but the team ultimately concluded that while the Henderson instruction “substantially reduced juror reliance on weak identification evidence,” it “also equally reduced juror reliance on strong identification evidence” because jurors were effectively so desensitized to the subject that they discounted eyewitness testimony altogether.[34] In other words, the study indicated the jury instruction was more likely to indiscriminately increase the rate of exonerations for all defendants, not just those who are innocent.[35]

In any event, it is still widely accepted that some sort of instruction is beneficial to the jury in weighing eyewitness reliability. The pared-down instruction that was ultimately approved seems to strike a compromise between advocates for and against an eyewitness instruction. MAPA ultimately issued a statement that concluded “the revised instruction, while not perfect, will go much further to assist jurors in understanding and applying the law and making informed decisions.”[36] Indeed, it will be interesting to learn whether Missouri has found the right balance in its new instruction as it continues to be used in trial courts, or whether it will require further revisions to maximize juror understanding.

II. Expert Testimony on Eyewitness Identification

While the Supreme Court of Missouri has unequivocally adopted an instruction on the subject, it is less clear whether the Court would be amenable to permitting expert testimony on the reliability of eyewitness identification. In 1988, the Court held it was not an abuse of discretion for a trial court to exclude eyewitness expert testimony in its unanimous decision, State v. Lawhorn.[37] The Court reasoned such expert testimony impermissibly invaded the province of the jury because it related to assessing the credibility of other witnesses without providing any helpful information beyond that which exists in the “general realm of common experience of members of a jury.”[38] It emphasized the due process safeguards guaranteed to defendants to challenge witness credibility and concluded, “[T]he introduction of expert testimony [on this subject] would be ‘a superfluous attempt to put the gloss of expertise, like a bit of frosting, upon inferences which lay persons were equally capable of drawing from the evidence.’”[39] In the year following Lawhorn, the Court reaffirmed its holding in State v. Whitmill.[40]

While Lawhorn and Whitmill remain the two leading Missouri cases on this issue, in the 30 years since they were decided an increasing number of courts across the United States have favored the admission of expert testimony on eyewitness identification. With the advent of DNA profiling leading to an increase in the rate of exonerations and the growing consensus among psychologists on the fallibility of memory, numerous state courts have walked back prior decisions deeming such testimony unreliable or unhelpful.[41] Many of these opinions are accompanied by detailed descriptions of the relevant scientific research, and several emphasize the value expert testimony can have in aiding jurors.[42] As the Supreme Court of Illinois noted, the last few decades “have seen a dramatic shift in the legal landscape, as expert testimony concerning the reliability of eyewitness testimony has moved from novel and uncertain to settled and widely accepted.”[43]

While Missouri’s decades-old precedent on this issue may appear to leave it well behind the latest trend, it is important to note the Court’s holdings in Lawhorn and Whitmill did not create a complete bar to such testimony. Rather, both opinions narrowly held that trial courts have the discretion to exclude expert testimony on eyewitness identification and, under the particular facts of those cases, the trial court did not abuse its discretion by doing so.[44] Thus, the Court left open the possibility for defense counsel to forcefully argue trial courts also have the discretion to admit such testimony without violating its precedent. To successfully do so, litigants would have to convince the trial court that the average juror has misconceptions about memory and the reliability of eyewitness testimony, and that an expert would aid the jury in correcting those misconceptions. This would undoubtedly be a difficult task, especially given the Court’s express findings on the subject in Lawhorn and Whitmill, but the wide acceptance of such expert testimony by both state and federal courts in the last 30 years could be an indication a Missouri trial court would be willing to allow it. Particularly since Missouri’s recent adoption of the Daubert standard, trial judges are given wide discretion in deciding motions to exclude expert testimony.[45] In fact, the Western District of the Missouri Court of Appeals even indicated such testimony is permissible in State v. Allen, stating “it would not have been an abuse of discretion for the trial court to have allowed [the defendant’s] expert testimony.”[46] The Eastern District, citing Allen, similarly found that while “defendants are not entitled to [provide] expert testimony as to the general” reliability of eyewitness testimony, trial courts have “the sound discretion” to admit it.[47]

Because the state has no right to appeal adverse decisions not to suppress evidence in criminal cases,[48] however, Missouri appellate courts have only been faced with cases where trial courts granted a prosecutor’s motion to exclude this testimony.[49] Other state appellate courts have found a trial court abused its discretion by granting a prosecutor’s motion to exclude eyewitness expert testimony, typically when attempting to abolish the state’s per se ban on such testimony.[50] Though Lawhorn and Whitmill fall short of establishing a per se ban, the appellate courts have understandably been wary of finding any abuse of the wide discretion the Supreme Court of Missouri affords trial courts in excluding this testimony.[51] Most recently, the Western District rejected a defendant’s argument that “many years have passed since Lawhorn and Whitmill were decided, [and] scientific study and research now reveals that jurors do not understand the problems associated with eyewitness identification and suggests that Lawhorn and Whitmill should no longer be followed.”[52] Rather, the Court of Appeals held it is “constitutionally bound to follow the last controlling decision of Missouri’s Supreme Court, regardless of how many years have passed since that decision was rendered.”[53] The Supreme Court of Missouri is not likewise constitutionally bound. Accordingly, it is the most likely venue for ambitious attorneys to cite the national shift toward accepting eyewitness expert testimony and successfully argue a trial court abused its discretion by disallowing it. To do so, attorneys would either have to convince the Court to grant their petition to transfer or invoke the Court’s mandatory jurisdiction by appealing a death penalty case.

Of course, the Court could find these arguments unpersuasive and stand by Lawhorn and Whitmill should they again hear a case where a defendant seeks to introduce an eyewitness expert. This is especially true given the adoption of MAI 310.02 since the two cases were decided, as the judges could find the model instruction does enough to correct any misconceptions by jurors. Such is the case in New Jersey, which will likely rarely see such expert testimony in its trial courts because, in Henderson, the New Jersey court indicated enhanced jury instructions are a preferred substitute for expert testimony, finding the instructions are “authoritative,” “cost-free,” and “spare the jurors” from considering “battling experts” on the subject.[54] Further, there are the concerns raised in Lawhorn and Whitmill. Is this a matter of the credibility rather than the reliability of witnesses? Are jurors already privy to information an expert might provide? Finally, like the debate surrounding the instruction on eyewitness identifications, there is much debate as to whether allowing expert testimony on the subject would result in more sophisticated jurors or whether it would result in jurors placing too little weight on eyewitness identifications.


1 AbigailAbigail Twenter Twenter is an associate in the litigation practice group of Armstrong Teasdale in St. Louis. Formerly she was a judicial clerk at the United States District Court, Eastern District and also clerked for Judge Laura Denvir Stith at the Supreme Court of Missouri.

2 The case number for Nelson’s criminal case is 16-CR-84000682. See also The Nat’l Registry of Exonerations (database updated March 2018), https://www.law.umich.edu/special/exoneration/Pages/casedetail.aspx?caseid=4201.

3 See The Innocence Project, https://www.innocenceproject.org/dna-exonerations-in-the-united-states/ (last visited December 13, 2018).

4 Id.

5 People v. Lerma, 47 N.E.3d 985, 993 (Ill. 2016) (citing State v. Dubose, 699 N.W.2d 582, 591-92 (Wis. 2005)).

6 Tricia Bushnell & Amol Sinha, Show Me Real Eyewitness ID Reform, St. Louis Post Dispatch, Jan. 22, 2016, http://www.stltoday.com/opinion/columnists/show-me-real-eyewitness-id-reform/article_74e36781-0ca1-5636-ab48-29e071ee9682.html.

7 Adam Liptak, 34 Years Later, Supreme Court Will Revisit Eyewitness IDs, N.Y. Times, Aug. 22, 2011, https://www.nytimes.com/2011/08/23/us/23bar.html.

8 State v. Henderson, 27 A.3d 872, 878 (N.J. 2011).

9 Id.

10 Id.

11 Henderson at 904-11; see State v. Lawson, 291 P.3d 673, 687-88 (Or. 2012).

12 See Melissa Hogenboom, Why Does the Human Brain Create False Memories?, BBC, Sept. 29, 2013, http://www.bbc.com/news/science-environment-24286258 (“Christopher French of Goldsmiths University in London says there is still a lack of awareness of how unreliable human memory is, especially in the legal system. ‘Although this is common knowledge within psychology and widely accepted by anybody who has studied the literature, it’s not widely known about in society more generally,’ he says.”).

13 See Daniel J. Simons & Christopher F. Chabris, What People Believe About How Memory Works: A Representative Survey of the U.S. Population, PLoS ONE 6(8): e22757 (2011), https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0022757.

14 Elizabeth F. Loftus, Juries Don’t Understand Eyewitness Testimony, N.Y. Times (Sept. 1, 2011, 11:32 AM), https://www.nytimes.com/roomfordebate/2011/08/31/can-we-trust-eyewitness-identifications/juries-dont-understand-eyewitness-testimony.

15 See id; Kate A. Houston et al., Expert Testimony on Eyewitness Evidence: In Search of Common Sense, Behav. Sci. & L. 637, https://onlinelibrary.wiley.com/doi/abs/10.1002/bsl.2080; Liptak, supra note 6.

16 There is much debate as to the correlation between eyewitness confidence and accuracy. See Brandon L. Garrett, Eyewitnesses and Exclusion, 65 Vand. L. Rev. 451, 469 (2012), https://wp0.vanderbilt.edu/lawreview/wp-content/uploads/sites/89/2012/03/Garrett_65_Vand_L_Rev_451.pdf.

17 Watkins v. Sowders, 449 U.S. 341, 352 (1981) (Brennan, J., dissenting) (emphasis in original omitted) (quoting Elizabeth F. Loftus, Eyewitness Testimony 189-190 (1979)).

18 MAI 310.02 Eyewitness Identification Testimony (Approved December 16, 2015), https://www.courts.mo.gov/file.jsp?id=95086.

19 565 U.S. 228, 246 (2012) (addressing the “safeguards built into our adversary system that caution juries”). Id. at 245.

20 Order dated April 2, 2015: re: Additions to MAI-CR 3d, https://www.courts.mo.gov/sup/index.nsf/9f4cd5a463e4c22386256ac4004a490f/9e2d3a1ae4baba6e86257e1b00644d44?OpenDocument.

21 Id. at 1.

22 MAI 310.02 Eyewitness Identification Testimony (Approved December 16, 2015), https://www.courts.mo.gov/file.jsp?id=95086

23 Order dated April 2, 2015: re: Additions to MAI-CR 3d, https://www.courts.mo.gov/sup/index.nsf/9f4cd5a463e4c22386256ac4004a490f/9e2d3a1ae4baba6e86257e1b00644d44?OpenDocument.

24 Id.

25 State v. Henderson, 27 A.3d 872, 895 (N.J. 2011).

26 Amy D. Trenary, State v. Henderson: A Model for Admitting Eyewitness Identification Testimony, 84 U. Colo. L. Rev. 1257, 1301 (2013).

27 See New Jersey Model Criminal Jury Charges (2018), https://www.judiciary.state.nj.us/attorneys/criminalcharges.html.

28 Henderson, 27 A.3d at 924.

29 Statement from Kevin S. Hillman, president of the Missouri Association of Prosecuting Attorneys (Jan. 22, 2016), http://www.prosecutors.mo.gov/files/MAPA%20PR%20Eyewitness%20ID%20Jury%20Instruction%20(1).pdf.

30 Position Paper on Proposed MAI 310.02 Relating to Eyewitness Identification, Missouri Association of Prosecuting Attorneys, http://www.prosecutors.mo.gov/files/Eyewitness%20ID%20Instruction%20Position%20Paper.pdf.

31 Id.

32 See Marlee Kind Dillon, et al., Henderson Instructions: Do They Enhance Evidence Evaluation?, 17 J. Forensic Psychol. Research & Prac. 1 (2017); Athan P. Papailious et al., The Novel New Jersey Eyewitness Instruction Induces Skepticism but Not Sensitivity, PLOS│ONE (Dec. 9, 2015), https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0142695; Angela M. Jones & Steven Penrod, Improving the Effectiveness of Henderson Instruction Safeguard Against Unreliable Eyewitness Identification, 24 Psychol., Crime & L. 177 (2017).

33 Athan P. Papailious et al., The Novel New Jersey Eyewitness Instruction Induces Skepticism but Not Sensitivity, PLOS│ONE (Dec. 9, 2015), https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0142695

34 Id.

35 Id.

36 See Statement from Kevin S. Hillman, president of the Missouri Association of Prosecuting Attorneys (Jan. 22, 2016), http://www.prosecutors.mo.gov/files/MAPA%20PR%20Eyewitness%20ID%20Jury%20Instruction%20(1).pdf.

37 762 S.W.2d 820, 822 (Mo. banc 1988).

38 Id. at 823.

39 Id. (citing State v. George, 481 A.2d 1068, 1075 (Conn. 1984)).

40 780 S.W.2d 45 (Mo. banc 1989).

41 See National Research Council, Identifying the Culprit: Assessing Eyewitness Identification 38 (2014).

42 Id.

43 People v. Lerma, 47 N.E.3d 985, 993 (Ill. 2016).

44 In fact, in Commonwealth v. Walker, 92 A.3d 766, 782-83 (Pa. 2014), the Supreme Court of Pennsylvania collected cases from 44 states, the District of Columbia, and 10 federal circuit courts, noting “there is a clear trend among state and federal courts permitting the admission of expert eyewitness testimony, at the discretion of the trial court, for the purpose of aiding the trier of fact in understanding the characteristics of eyewitness identification.” Interestingly, it included Missouri as a state following this trend, citing State v. Ware, 326 S.W.3d 512 (Mo. App. S.D. 2010). In that case, however, the Southern District did not permit the admittance of such testimony, but rather, citing Lawhorn and Whitmill, found no abuse of discretion in a trial court’s decision to grant a prosecutor’s motion in limine to exclude “testimony from [defendant’s] expert witness concerning the fallibility of eyewitness identification testimony.” Id. at 526, 529-30.

45 Section 490.065, RSMo 2017; see also Daubert v. Merrell Dow Pharm., 509 U.S. 579 (1993). Prior to Missouri’s adoption of the Daubert standard, trial courts had to first find the underlying science was generally accepted as reliable by the relevant scientific community in order to allow the expert to testify in a criminal case. See State v. Keightley, 147 S.W.3d 179, 187 (Mo. App. S.D. 2004). Now, the trial judge has broad discretion in determining the reliability of scientific knowledge after considering a number of factors.

46 274 S.W.3d 514, 527 (Mo. App. W.D. 2008).

47 State v. Body, 366 S.W.3d 625, 629 (Mo. App. E.D. 2012).

48 Section 547.200, RSMo 2017.

49 State v. Naylor, 505 S.W.3d 290, 298 (Mo. App. W.D. 2016); State v. Ware, 326 S.W.3d 512, 528-29 (Mo. App. S.D. 2010); State v. Allen, 274 S.W.3d 514, 526-27 (Mo. App. W.D. 2008); State v. Cunningham, 863 S.W.2d 914, 923 (Mo. App. E.D. 1993); State v. Donnell, 862 S.W.2d 445, 450-51 (Mo. App. W.D. 1993); State v. Hill, 854 S.W.2d 486, 487-88 (Mo. App. E.D. 1993); State v. Simpson, 793 S.W.2d 182 (Mo. App. E.D. 1990). See also Morgan v. Steele, No. 4:07-CV-0877-TCM, 2010 WL 3564425, at *9 (E.D. Mo. Sept. 7, 2010) (finding there was no due process violation by the exclusion of expert testimony on the reliability of eyewitness identification).

50 See e.g., State v. Carr, 331 P.3d 544, 690 (Kan. 2014) (finding “it is an abuse of discretion for . . . trial [courts] to automatically exclude [such] testimony”); People v. Lerma, 47 N.E.3d 985, 996-97 (Ill. 2016); Commonwealth v. Walker, 92 A.3d 766, 792-93 (Pa. 2014) (finding such testimony is no longer per se admissible); State v. Copeland, 226 S.W.3d 287, 304 (Tenn. 2007).

51 State v. Henderson, 27 A.3d 872, 895 (N.J. 2011).

52 Naylor, 505 S.W.3d at 298.

53 Id.

54 Andrew E. Taslitz, “Curing” Own Race Bias: What Cognitive Science and the Henderson Case Teach About Improving Jurors’ Ability to Identify Race-Tainted Eyewitness Error, 16 Legis. & Pub. Pol. 1049, 1056 (2013).