I Saw You but Did I Really?:

Transcription

1 I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Lori V. Berke Jody C. Corbett Berke Law Firm, PLLC 1601 N. 7th Street, Suite 360 Phoenix, AZ (602) lori@berkelawfirm.com jody@berkelawfirm.com John Wixted Department of Psychology UC San Diego La Jolla, CA (858) jwixted@ucsd.edu

2 Lori V. Berke has been practicing law in Phoenix, Arizona for 21 years, and is the founder and principal of Berke Law Firm, PLLC. She has represented public entities throughout Arizona, including cities, counties, school districts and public transportation organizations. The majority of cases she handles for public entities include 42 U.S.C civil rights defense, accident and injury litigation, premises liability, road design, construction and maintenance liability, and transit litigation. Lori is AV Rated by Martindale-Hubbell and has been designated a Southwest Super Lawyer. John Wixted received his Ph.D. in 1987 from Emory University and is now a Distinguished Professor in the Department of Psychology at UC San Diego. His research is concerned with understanding the mechanisms of human memory, focusing on the neuroscience of memory and amnesia, signal-detection analyses of recognition memory, and eyewitness memory. He is the author of more than 100 research articles and book chapters, and in 2011, he was the recipient of the Howard Crosby Warren Medal for outstanding achievement in experimental psychology.

3 I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Table of Contents I. Introduction...5 II. Science Behind Memory of Eyewitnesses...6 A. Science Behind Memories of Eyewitnesses...6 B. Best Lineup Practices...10 III. Demonstrating Validity of Eyewitness Identifications Used in Probable Cause Determination in Defense of Section 1983 Claims...11 A. Was Identification Procedure Impermissibly Suggestive? Photo Lineups In-Person Lineups One-on-One or Showup Identifications/Single Photographs...14 B. Did the Witness Exhibit Sufficient Indicia of Reliability? Opportunity to view the criminal at the time of the crime Degree of attention paid to the criminal Accuracy of the prior descriptions of the criminal Level of certainty demonstrated at the time of lineup Length of time between the crime and the lineup...16 C. Claims of Judicial Deception Under 42 U.S.C Do Results of Lineups (Positive or Negative) Need to Be Disclosed to Grand Jury/Judge for Probable Cause Determination?...17 IV. Qualified Immunity for Eyewitness Identifications...17 V. Eyewitness Identification Procedures Used During Civil Litigation by a Plaintiff to Identify Defendant Officers...19 VI. Conclusion...20 Endnote...20 I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Berke et al. 3

4

5 I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases I. Introduction Consider the following fact pattern: Officers were called to the scene of an armed robbery of a convenience store. As Officer H responded to the scene, he observed a male and female who fit the description of the armed robbery suspects walking quickly away from the convenience store. Officer H told them to stop, which they initially did, and Officer H got a good look at both of them. The male suspect, who was reported to be armed, jumped over a wall and Officer H decided to pursue him. Officer H told the female suspect to stay put and he pursued the male suspect. Approximately one hour later, the male suspect was found hiding in a residential yard, and he was taken into custody. When Officer H returned to the location where he had instructed the female suspect to remain, he discovered she had fled. The male arrestee invoked his Fifth Amendment rights and therefore did not provide police with any information about the female suspect. Officer D viewed the surveillance video from the convenience store and got a good look at the female suspect shown on the video. The following day, Officer B researched known associates of the male suspect and discovered the plaintiff, who matched the physical description of the female suspect and who had been arrested previously with the male suspect when they were together in a stolen vehicle. (Unbeknownst to police, the plaintiff was married to the male suspect at the time of the armed robbery, and had been for several years.) Officer B showed the plaintiff s Motor Vehicle Division (MVD) driver s license photograph to Officer D (the officer who had viewed the surveillance video) and Officer D advised that the plaintiff was definitely the female suspect he had seen on the surveillance video. Using the plaintiff s photograph, Officer N created a photographic lineup using the police department s computer software. The lineup contained plaintiff s photograph and five other females with similar physical characteristics and features. Officer B administered the lineup to Officer H (the officer who had stopped the suspects shortly after the robbery) and Officer H immediately identified the plaintiff as the female suspect. When he administered the lineup to Officer H, Officer B looked away so he would not inadvertently influence Officer H by looking at a particular photograph. Based on the positive identifications by Officer D and Officer H, and notwithstanding the store clerk s positive identification of a filler photograph, Officer B concluded there was probable cause to arrest the plaintiff. Officers B and N went to the plaintiff s apartment to arrest her, but she was not there. A man claiming to be her roommate gave the officers consent to conduct a protective sweep. At that time, they observed in plain view paystubs with the male suspect s name on them. The following day, Officers Z and E and Sergeant M arrested the plaintiff, using her driver s license photograph to confirm that it was her. Seven months after her arrest, the criminal charges against the plaintiff were dismissed. During the criminal prosecution, the plaintiff claimed that the male suspect s current girlfriend whose name is Christina was the female involved in the armed robbery. Detectives located Christina, who denied involvement in the armed robbery. She claimed a female named Lori last name unknown had accompanied the male suspect to the convenience store the evening of the armed robbery, and that Lori must have been the female involved in the armed robbery. Detectives working for plaintiff s criminal defense attorney claimed to have interviewed Christina s father and shown him still photos from the video of the armed robbery, and they claimed Christina s father told them the female was Christina. The plaintiff s criminal attorney s detectives testified at deposition that they did not believe this Lori actually existed. DNA had I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Berke et al. 5

6 been collected from a sweatshirt that the female suspect discarded at the location she had been instructed to remain, and neither the plaintiff nor Christina could be included or excluded as a contributor of DNA on the sweatshirt. The plaintiff s criminal attorney and her detectives put an extensive presentation together in which they attempted to persuade prosecutors that Christina was the female involved in the armed robbery, and the plaintiff was not. Because the plaintiff and Christina looked so much alike, prosecutors did not believe there was a substantial likelihood of conviction, and they ultimately dismissed the charges against the plaintiff. The plaintiff filed civil lawsuits in both state and federal court against all of the officers described above claiming that there was not probable cause for her arrest. The plaintiff claimed in the lawsuits that it was improper for detectives to use her driver s license photograph in the photo lineup because at the time of the armed robbery, she weighed thirty pounds more than she did when her driver s license photograph was taken, and therefore her appearance was dramatically different. However, Officer Z used the driver s license photograph to identify the plaintiff when they were searching for her to arrest her. Thus, her appearance was not dramatically different. We successfully moved to dismiss most of the claims in the state case, and then the plaintiff ultimately agreed to dismiss the remaining claims. We also successfully moved to dismiss early in the federal case all claims other than the claim for false imprisonment. We moved for summary judgment as to the false imprisonment claim and the District Court granted the motion. The basis for the District Court s dismissal of that claim was that there was probable cause for the plaintiff s arrest. The plaintiff has appealed the Order granting summary judgment to the Ninth Circuit Court of Appeals. The appeal is fully briefed and we are awaiting a decision by the Court of Appeals as to whether the plaintiff s request for oral argument will be granted. The case agent assigned to the investigation recently retired, and the new case agent assigned to this open case requested additional DNA testing of the sweatshirt that had been discarded by the female. There was a hit and it was determined that the DNA belonged to a female named Lori Perez. On January 21, 2016, Ms. Perez pled guilty to attempted armed robbery and is currently serving a prison term of three years. All three women the plaintiff, Christina, and Lori have similar physical features. Thus, notwithstanding that the plaintiff was not involved in the armed robbery, it was reasonable for Officer H to believe that the female he selected in the photo lineup was the female he saw the night of the armed robbery. The reliability of eyewitness identifications does not only present itself in criminal cases. It also comes up in civil cases involving claims wherein a plaintiff is claiming that officers did not have probable cause for his or her arrest because the eyewitness identification was unreliable. The specific constitutional rights plaintiffs argue were violated are typically the Fourth Amendment right to be free from unreasonable seizure without probable cause and the Fourteenth Amendment right to due process/fair trial. These materials will discuss the many scientific and legal issues related to eyewitness identifications that attorneys who represent government entities and employees may face in their defense of lawsuits involving Section 1983 claims and related state law claims. II. Science Behind Memory of Eyewitnesses A. Science Behind Memories of Eyewitnesses Eyewitness memory is widely believed to be unreliable, and it is easy to understand why. Eyewitness misidentifications played a role in over 70% of the 347 DNA exonerations of the wrongfully convicted to date. In court, these misidentifications were invariably made with high confidence. (Garrett, B. (2011), Convicting the Innocent: Where Criminal Prosecutions Go Wrong, Cambridge, MA: Harvard University Press). In addi- 6 Civil Rights and Governmental Tort Liability January 2017

7 tion, early laboratory studies by eyewitness identification researchers often reported a weak relationship between eyewitness confidence and accuracy even when memory was tested under ideal conditions in the research laboratory. Based on these earlier studies, a weak confidence-accuracy relationship meant that highly confident witnesses were only slightly more accurate than the error-prone witnesses who make an identification with low confidence. Researchers believed that if highly confident witnesses were unreliable even under ideal testing conditions (e.g., under the non-stressful conditions of a simulated crime study conducted in the lab), it stood to reason that they would be even less reliable in the real world. Finally, researchers in these early studies identified a host of variables (known as estimator variables ) that they found to impair eyewitness identification accuracy. These variables included lighting at the time of crime, exposure duration, retention interval (i.e., the time delay between the witnessed event and the first eyewitness identification test), the correspondence between the race of the witness and the perpetrator, the presence or absence of a weapon during the commission of the crime, and the level of stress experienced by the witness. More specifically, accuracy has been found to be reduced when lighting conditions were poor, the exposure duration was short, the retention interval was long, the witness and the perpetrator were of different races, a weapon was present, and stress was high. Because one or more of those accuracy-reducing conditions is likely to apply to almost every eyewitness identification case, the message seems clear: eyewitness identification is an inherently unreliable form of evidence. In light of these considerations, eyewitness identification experts often testify in court that jurors should disregard eyewitness confidence and instead pay close attention to the estimator variables that may reduce the accuracy of an eyewitness identification. However, more recent research suggests a new understanding of the reliability of eyewitness identification. With regard to eyewitness confidence, many studies published since 1995 including both mock crime studies conducted by experimental psychologists (e.g., Brewer, N., & Wells, G. L. (2006), The confidence-accuracy relation in eyewitness identification: Effects of lineup instructions, foil similarity, and target-absent base rates, Journal of Experimental Psychology: Applied, 12, pp ; Palmer, M., Brewer, N., Weber, N. & Nagesh, A. (2013), The confidence-accuracy relationship for eyewitness identification decisions: Effects of exposure duration, retention interval, and divided attention, Journal of Experimental Psychology: Applied, 19, pp ) and real-world police department field studies involving actual eyewitnesses to a crime (e.g., Behrman, B.W., & Davey, S.L. (2001), Eyewitness identification in actual criminal cases: An archival analysis, Law and Human Behavior, 25, pp ; Wixted, J. T., Mickes, L., Dunn, J. C., Clark, S. E. & Wells, W. (2016), Estimating the reliability of eyewitness identifications from police lineups, Proceedings of the National Academy of Sciences, 113, pp ) have established beyond any reasonable doubt that eyewitness confidence associated with a stranger ID made from a properly administered lineup is strongly related to accuracy. This means that a low-confidence eyewitness ID (e.g., I think that s him, but I m not sure ) is associated with relatively low (but still above-chance) accuracy, whereas a high-confidence eyewitness ID (e.g., that s him, there is no doubt in my mind ) is associated with high accuracy. The field s relatively recent understanding of the informational value of eyewitness confidence contrasts with its previous but incorrect view (a view still held by some scientists) that confidence is at best only modestly related to accuracy even when memory is tested under ideal conditions (e.g., Sporer, S. L., Penrod, S., Read, D., & Cutler, B. (1995), Choosing, confidence, and accuracy: A meta-analysis of the confidence accuracy relation in eyewitness identification studies, Psychological Bulletin, 118, pp ). The research described above focused on lineups, but the police often need to use a showup instead. A recent National Research Council report summarized the typical use of a showup procedure in the following way: A showup is a police-arranged identification procedure in which the police show one person to a witness and ask if she or he recognizes that person. This procedure typically is used when the police locate a suspect shortly after the commission of a crime and within close proximity to the I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Berke et al. 7

8 scene. Case law limits the time and distance from a crime during which such a procedure will pass legal standards. In response to such case law, police typically restrict showups to a twohour time period after the commission of a crime. Ideally, officials take the witness to the location where the suspect has been detained and do not display the suspect in a suggestive manner (e.g., not in a police car, not handcuffed, without drawn weapons). However, as chases, fights, or disarmaments frequently precede showups, the apprehension of a suspect can raise safety issues that make it difficult to adhere to recommended procedures. Further, the nature of a showup does not lend itself to the use of a blinded procedure. A showup is designed to promptly clear innocent suspects, thereby sparing them from a prolonged period of detention as the investigation continues. Delaying the showup to locate an uninvolved officer may defeat that purpose. While some law enforcement agencies use a standard procedure with written instructions when conducting a showup, there is no indication that such procedures are used uniformly. Courts consider showups highly suggestive, and prosecutors urge the police to exercise caution when conducting them. (National Research Council (2014), Identifying the Culprit: Assessing Eyewitness Identification, Washington, DC: The National Academies Press, pp ). Even for showups, confidence is a reliable predictor of accuracy, just as it is for lineups. For example, one recent lab study (Mickes, L. (2015), Receiver operating characteristic analysis and confidence-accuracy characteristic analysis in investigations of system variables and estimator variables that affect eyewitness memory, Journal of Applied Research in Memory and Cognition, 4, pp ) found that showup IDs made with high confidence (~80% correct) were much more accurate than showup IDs made with low (~40% correct). Although showups will often need to be used, it is nevertheless better for the police to use a lineup whenever possible because a high-confidence ID from a lineup is even more reliable (often ~95% correct in lab studies). What about the influence of various estimator variables on the reliability of an eyewitness identification? The field has recently acquired a new understanding of this issue as well. Critically, initial highconfidence IDs are reliable even if the conditions under which memory of the perpetrator was formed were unfavorable (high stress, poor lighting, etc.). It is true that these poor witnessing conditions can make it difficult to form a strong memory of the perpetrator. However, weak memories do not lead to high-confidence positive IDs of innocent suspects. Instead, they result in low-confidence false IDs or to failures to identify anyone (precisely because memory is weak). It is in that sense that poor estimator variables reduce accuracy. In other words, they make it less likely that a witness will be able to identify the actual perpetrator from a lineup. They do not necessarily reduce the accuracy of a high-confidence ID when it occurs. Recent research clearly shows that high-confidence IDs, while being less common when (1) viewing conditions are poor, (2) attention is distracted, or (3) the retention interval is long, remain highly accurate when they do occur. (Palmer, M., Brewer, N., Weber, N. & Nagesh, A. (2013), The confidence-accuracy relationship for eyewitness identification decisions: Effects of exposure duration, retention interval, and divided attention, Journal of Experimental Psychology: Applied, 19, pp ). The same appears to be true of cross-race IDs. (Nguyen, T. B., Pezdek, K. & Wixted, J. T. (in press), Evidence for a Confidence-Accuracy Relationship in Memory for Same- and Cross- Race Faces, Quarterly Journal of Experimental Psychology). It is important to emphasize that eyewitness confidence in a stranger ID is a reliable indicator of accuracy only on an initial memory test, not on any later memory test (such as a second lineup containing another photo of the suspect, or the ID that occurs during the trial when the witness is asked to identify the defendant). The reason is that the very act of testing memory (e.g., when a witness picks a suspect from a lineup) contaminates memory. At a minimum, having seen and identified the suspect once in a lineup, the 8 Civil Rights and Governmental Tort Liability January 2017

9 suspect will seem more familiar on any later identification test (inappropriately increasing confidence that the suspect committed the crime). Indeed, confidence in an ID can increase for a variety of reasons following an initial identification even though accuracy does not increase. (E.g., Wells, G. L., & Bradfield, A. L. (1998), Good, you identified the suspect : Feedback to eyewitnesses distorts their reports of the witnessing experience, Journal of Applied Psychology, 83, pp ). However, the key point is that on an initial memory test, confidence in a stranger ID is strongly indicative of accuracy (Wixted, J. T., Mickes, L., Clark, S. E., Gronlund, S. D. & Roediger, H. L. (2015), Initial eyewitness confidence reliably predicts eyewitness identification accuracy, American Psychologist, 70, pp ), and this remains true even when exposure duration was brief, the retention interval was long, it was a cross-race ID, and so on. On an initial test, low confidence implies low accuracy, whereas high confidence implies high accuracy. That is the new understanding of the reliability of eyewitness identification. A recent analysis of DNA exoneration cases conducted by Brandon Garrett as reported in his 2011 book Convicting the Innocent: Where Criminal Prosecutions Go Wrong, Cambridge, MA: Harvard University Press, suggests that a shift in focus from courtroom confidence to initial confidence (and only initial confidence) could avert many wrongful convictions. Garrett analyzed trial materials for 161 DNA exonerees who had been misidentified by one or more eyewitnesses in a court of law. His key finding was described as follows: I expected to read that these eyewitnesses were certain at trial that they had identified the right person. They were. I did not expect, however, to read testimony by witnesses at trial indicating that they earlier had trouble identifying the defendants.... Yet in 57% of these trial transcripts (92 of 161 cases), the witnesses reported that they had not been certain at the time of their earlier identifications. (P. 49, emphasis in original). Because eyewitness certainty at the time of the initial ID is diagnostic of guilt (i.e., high confidence = high accuracy and low confidence = low accuracy), then the implication of Garrett s discovery is that in at least 57% of these cases, expressions of uncertainty on the initial memory test clearly signaled that these eyewitnesses were at high risk of having made an error. It was only later, in court, that confidence became inappropriately inflated. This unfortunate state of affairs was true even in one of the most famous cases of eyewitness misidentification, one often used to illustrate the unreliability of eyewitness memory. During a trial that was held in 1985, Jennifer Thompson confidently identified Ronald Cotton as the man who had raped her. Cotton was convicted largely on the basis of her testimony, but he was later exonerated by DNA evidence after spending more than 10 years in prison. Long before the trial, however, Thompson s initial identification of Cotton from a photo lineup was characterized by a prolonged period of hesitation and indecision that lasted for nearly five minutes and that ended with a low-confidence verbal identification consisting of the words I think this is the guy. (Thompson-Cannino, J., Cotton, R., & Torneo, E. (2009), Picking Cotton: Our memoir of injustice and redemption, New York, NY: St. Martin s Press p. 33). However, after confirmatory feedback from the police, Thompson became increasingly confident that Cotton was the rapist. The mistake was to rely on confidence expressed at the time of the trial (after it had become improperly inflated) instead of relying on confidence expressed at the time of the initial ID (before memory contamination had a chance to play a significant role). Merely asserting that eyewitness memory is unreliable overlooks this key consideration. Eyewitness memory is highly reliable the first time it is tested. It is unreliable when it is tested again because by then, memory has been contaminated. All forensic evidence not just eyewitness evidence is unreliable when it is contaminated. In that sense, there is nothing special about eyewitness identification evidence. It is reliable when it is not contaminated and unreliable when it is (just like DNA evidence). I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Berke et al. 9

10 These considerations explain why Wixted et al. (2015) made the following recommendation: Jurors should consider the level of certainty expressed by an eyewitness during the initial identification (at which time confidence is likely to be a reliable indicator of accuracy) while disregarding the level of certainty expressed at trial (because, by then, confidence may no longer be a reliable indicator of accuracy). Note that this is the exact opposite of the advice that the field of psychology has sent to the legal system for many years. For many years, that advice has been to ignore confidence (except, perhaps, as a way to discredit the witness should it become inflated over time) and to pay attention to estimator variables. B. Best Lineup Practices The police are not legally obligated to use eyewitness identification procedures that scientists have deemed to be best practices. Indeed, scientists often disagree with each other about what those best practices are. Under such conditions, the police cannot reasonably be expected to read the scientific literature and figure out for themselves which procedures are best. Similarly, they cannot reasonably be held accountable for not using procedures that some scientists believe are the best ones to use. How are scientific findings actually translated into policy recommendations that the police can use (even if they are not legally obligated to do so)? The National Academy of Sciences in Washington D.C. is an organization consisting of the nation s most prominent scientists. One of its major functions is to help the government translate science into policy implications. As noted on their web site, Our peer-reviewed consensus reports bring together the world s foremost scientists, engineers, and health professionals to address some of society s toughest challenges. Each year, more than 6,000 of these experts volunteer to serve on hundreds of study committees that are convened to answer specific sets of questions. One such committee was recently formed to address the policy implications of eyewitness identification research. Their report ( Identifying the Culprit ) was published in November of (National Research Council (2014), Identifying the Culprit: Assessing Eyewitness Identification, Washington, DC: The National Academies Press). This report was issued after the committee listened to testimony from various scientists who strongly disagree with each other about the utility of various eyewitness identification practices (e.g., simultaneous vs. sequential lineups, the information value of initial confidence, etc.). This was the first time that our nation s most prestigious scientific organization made direct policy translations of the relevant scientific research on this topic. Their report recommends that initial eyewitness confidence be recorded, that the initial eyewitness identification procedure be videotaped, that lineups be administered in double-blind fashion, that only one eyewitness identification test be performed per suspect, etc. Other guidelines have been provided over the years even though they are less authoritative than the guidelines provided by the National Academy for the first time in The first official science-based guidelines for eyewitness ID procedures by any national government organization came in The 1999 National Institute of Justice Eyewitness Evidence A Guide for Law Enforcement ( nij/ pdf) indicates that This Guide differs from earlier efforts in several fundamental ways: This Guide is supported by social science research (page 1). These guidelines made numerous recommendations (e.g., that there be only one suspect per lineup), many of which were similar to the recommendations that were later made by the National Academy of Sciences committee. (See also, Technical Working Group for Eyewitness Evidence (1999), Eyewitness evidence: A guide for law enforcement, Washington, DC: United States Department of Justice, Office of Justice programs, available at: The key point to appreciate is that now that the National Academy of Sciences has weighed in with its best practices recommendations, a police department may have a harder time defending itself against a lawsuit filed by, for example, a former prisoner who was convicted, in part, based on eyewitness evidence, 10 Civil Rights and Governmental Tort Liability January 2017

11 and who was freed for one reason or another after serving a considerable prison sentence. A common complaint made in such lawsuits is that the police used inappropriate eyewitness identification procedures that contributed to what the plaintiff considers to be a wrongful conviction. To support that claim, the plaintiff may hire an eyewitness identification expert who will testify about what science has learned about the unreliability of eyewitness identifications. As an example, scientific research clearly shows that using multiple identification procedures with the same witness and the same suspect (e.g., first, a photo lineup, then a live lineup) is inappropriate because, for reasons explained above, only the initial memory test is known to provide reliable information. However, back in 1995 (for example), the police commonly had witnesses make a photo ID of a suspect (e.g., from a mug book) and then later placed that same suspect in a lineup. If the witness picked that same suspect from a lineup with high confidence, that evidence would be used against the suspect at trial. Back then, however, it was not as clear as it is today that such double-testing is dangerous. Double-testing obviously does not appear on the list of best practices recommended by the National Academy of Sciences. Instead, the National Academy report recommends against multiple testing involving the same suspect and same witness. If the identification procedures in question were administered in 1995, the police can reasonably claim that such procedures were not known to be unreliable because the science had not yet been translated into policy recommendations by the appropriate agency of the government. In the future, that defense is less likely to be persuasive to a jury. Imagine, for example, a police identification procedure administered in 2016 that results in a suspect ID that, in turn, results in a wrongful conviction that is overturned by DNA evidence 10 years later. While it is true the National Academy best practices recommendations are not mandated by law, it is also true that police departments that choose not to use those best practices have decided not to use eyewitness identification procedures that, according to scientific research, best protect the innocent while at the same time maximizing the odds that a guilty suspect with be identified. An argument might be made by a particular police department that the costs of implementing the best-practice eyewitness identification procedures would be prohibitive. Consider, for example, the recommended use of blindly administered photo lineups, where the administrating officer does not know who the suspect is. Photo lineups are often administered in the home of a witness, who might live far from the police station. For small departments, the goal of having two officers available (one to blindly administer the lineup, and the other to further investigate the case) might be very difficult to achieve. Under such conditions, an approximation to the double-blind procedure might be necessary and could presumably be justified. One such approximation would be to use a blinded procedure in which the officer knows who the suspect is but shuffles the photos before giving them to the witness to examine. My point is that, more so than in the past, police departments that choose not to implement the procedures recommended by the National Academy of Sciences report had best have a clear explanation of why they chose to use a different approach even though it is known to potentially increase the risk of an innocent suspect being misidentified. A safer strategy would be to implement the recommended best practices procedures whenever possible. III. Demonstrating Validity of Eyewitness Identifications Used in Probable Cause Determination in Defense of Section 1983 Claims Courts in both criminal cases and civil cases use the same analysis to determine if an eyewitness identification is admissible (criminal case) or valid to support probable cause for an arrest (civil case). There is a two-part analysis to make that determination. The first question is whether the officers employed an identification procedure so impermissibly and/or unduly suggestive as to give rise to a substantial likelihood of misidentification. Manson v. Brathwaite, 432 U.S. 98, (1977); Foster v. California, 394 U.S. 440, 442, (1969); Grant v. City of Long Beach, 315 F.3d 1081, 1086 (9 th Cir. 2002). If the procedure was not impermissibly I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Berke et al. 11

12 suggestive, the identification evidence presents no due process obstacle to admissibility (criminal) and does not demonstrate that the officer could not use the identification to establish probable cause (civil), and thus no further inquiry is required. Id.; Foster v. California, 394 U.S. 440, 442 n. 2 (1969). If the court finds, however, that the procedures were impermissibly suggestive, it must then determine whether the identification was nonetheless independently reliable, considering five different factors: [1] the opportunity of the witness to view the criminal at the time of the crime, [2] the witness degree of attention, [3] the accuracy of the witness prior description of the criminal, [4] the level of certainty demonstrated by the witness at the confrontation, and [5] the length of time between the crime and the confrontation. Neil v. Biggers, 409 U.S. 188, (1972); see also Manson, 432 U.S. at 114. If the five factors weigh in favor of reliability, then the eyewitness identification is admissible (criminal) and sufficient to support probable cause to arrest (civil). Because both criminal and civil courts use the same analysis, courts deciding civil cases will often cite criminal cases as authority regarding this analysis. Since there are many more criminal cases that analyze these issues, these materials cite to many criminal cases as well as civil cases in the discussion of the various factors. With regard to claims brought pursuant to 42 U.S.C. 1983, and related state law claims, plaintiffs attack eyewitness identifications under two different constitutional provisions. First, plaintiffs allege that their Fourth Amendment right to be free from a seizure not based on probable cause was violated when an unreliable lineup was used. Second, plaintiffs allege that their Fourteenth Amendment due process right to a fair trial was violated by the admission of evidence derived from an unduly suggestive identification. However, as explained above, an unduly suggestive identification, in and of itself, is not unconstitutional. Rather, the identification must also be so unreliable under the totality of the circumstances that it could not support a finding of probable cause. Celestin v. City of N.Y., 581 F. Supp. 2d 420, (E.D.N.Y. 2008); Yattoni v. Oakbrook Terrace, 801 F. Supp. 140, 147 (N.D. Ill. 1992), aff d, 14 F.3d 605 (7 th Cir. 1993). Likewise, a plaintiff s constitutional right to a fair trial is violated only if the identification is both unduly suggestive and unreliable and then is admitted as evidence at the plaintiff s criminal trial. Manson v. Brathwaite, 432 U.S. at 113, n. 13; see also Sejnoha v. City of Bisbee, 815 F. Supp. 1300, 1303 (D. Ariz. 1993). Following is an analysis of each of the factors courts consider in evaluating the different types of eyewitness identifications for suggestiveness and reliability. A. Was Identification Procedure Impermissibly Suggestive? There are numerous ways that police officers conduct eyewitness identifications. These include single photographs, photographic lineups, in-person lineups, and one-on-one identifications or showups. 1. Photo Lineups There are several methods for administering photographic line-ups. These include a photo array with a set number of photographs on one page (such as six sometimes referred to as a six-pack or simultaneous ) and a stack of photographs shown to the witness individually (also known as sequential ). Whatever method is used, a photographic lineup is impermissibly suggestive if either the photographic display itself or the method of administering it emphasizes focus on a single photograph. United States v. Bagley, 772 F.2d 482, 493 (9 th Cir. 1985). Likewise, it is impermissibly suggestive if the officer administering it makes any comment or takes any action that directs the witness to a certain photograph. See United States v. Boston, 494 F.3d 660, 666 (8 th Cir. 2007) (holding that photographic lineup was not impermissibly suggestive because there was no evidence that the police officer made any suggestive sounds or gestures during the photographic lineup). a. Simultaneous Photo Arrays/ Six-Packs Modern photo arrays are often created using computer programs that search databases containing photographs, such as the department of motor vehicles, for photographs of individuals with similar physical 12 Civil Rights and Governmental Tort Liability January 2017

13 characteristics (height, weight, skin color, distinguishing features). The detective is then able to go through the photographs and select similar looking individuals to use in the array. The main issues that arise with photo arrays is whether they are created so that the suspect s photograph is not too different from the others. For photo arrays not to be impermissibly suggestive, the photographs should be an equal size and equally spaced with similar backgrounds. However, a photo array is not unconstitutional simply because the suspect is the only person wearing a particular item of clothing or because the backgrounds are different colors. See Taylor v. Evans, 2009 WL , at *5 (E.D. Cal. Apr. 20, 2009), report and recommendation adopted, 2009 WL (E.D. Cal. July 29, 2009). The suspect s photograph must be so different from all of the others that it stands out for the witness to pick only it to be impermissibly suggestive. b. Sequential Photo Lineups There are several scientific articles arguing that sequential photo lineups are less suggestive than simultaneous arrays. However, there is conflict among scientists on this issue and there is no case law requiring that sequential lineups must be used instead of simultaneous lineups in order to defeat a claim that the photo lineup was impermissibly suggestive. See Black v. Warren, 2015 WL , at *3 (D.N.J. Apr. 27, 2015), certificate of appealability denied (Dec. 15, 2015) (ruling that the habeas corpus petitioner failed to demonstrate how simultaneous, as opposed to sequential, display of the photos prejudiced the eyewitness identification and stating, Although sequential display is now the required procedure, it was not in 1996, and we see nothing inherently suggestive about the simultaneous displays in this case ); 1 United States v. Velasquez, 2013 WL , at *6 (N.D. Ind. Aug. 8, 2013) (ruling that the use of the non-sequential photo array was not in and of itself suggestive. The witnesses were admonished that the shooter might or might not appear in the photos that they were about to view. The photo array contained images of six Hispanic men with similar complexions, each with a similar build, dark hair that was somewhere between ear-and shoulder-length, a thin mustache, and a goatee ); Barker v. Galaza, 2003 WL , at *5 (N.D. Cal. Aug. 12, 2003), aff d, 113 F. App x 754 (9 th Cir. 2004) (stating that even though the habeas corpus petitioner argued that the photo lineup was simultaneous rather than sequential and that scholarly literature establishes that simultaneous viewings are inherently suggestive, he failed to cite to any controlling federal precedent on this issue and the argument was properly rejected); United States v. Johnson, 282 F. Supp. 2d 808, 811 (W.D. Tenn. 2003) (ruling that the criminal defendant s objection to the use of a photographic array instead of a sequential showing of photographs lacked merit because the defendant did not offer any evidence showing that a sequential showing is any less likely to result in a misidentification than is an array). Thus, the use of sequential lineups is not legally required. c. Double-blind Finally, the National Academy of Science s best practice of using a double-blind procedure, i.e. having an officer who does not know which photograph depicts the criminal suspect conduct the photo lineup, is also not legally required by United States Supreme Court precedent or by any circuit court. Instead, courts have stated that each photographic lineup must be considered on its own facts and that as long as the officer conducting the lineup does not make any suggestive comments, then the officer who created the lineup or who knows which photograph is the photograph of the suspect can conduct the lineup. See Simmons v. United States, 390 U.S. 377, 384 (1968) (holding that each case must be considered on its own facts, and that convictions based on eyewitness identification at trial following a pretrial identification by photograph will be set aside on that ground only if the photographic identification procedure was so impermissibly suggestive as to give rise to a very substantial likelihood of irreparable misidentification ); Hart v. Mannina, 798 F.3d 578, 588 (7th Cir. 2015), reh g denied (Oct. 16, 2015) (holding that there was no evidence that the officer coached I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Berke et al. 13

14 the witnesses or led them to identify the plaintiff as the suspect in any way and thus the photo lineup was not impermissibly suggestive); Taylor v. Evans, 2009 WL , at *5 (E.D. Cal. Apr. 20, 2009), report and recommendation adopted, 2009 WL (E.D. Cal. July 29, 2009) (holding that the lineup procedure was not impermissible suggestive because the officer conducting the photo lineup did not make any suggestive comments, but simply admonished the witnesses that the person may or may not be in the lineup and that they were under no obligation to identify anyone). Therefore, again, there is no legal requirement that a doubleblind procedure be used by police officers in conducting photo lineups. 2. In-Person Lineups In-person lineups are the type of lineups that Hollywood and average citizens often equate with eyewitness identifications. These consist of the suspect plus several other individuals with similar characteristics being viewed by the witness in person (usually on the other side of one-way glass). Similar to photo lineups, there is no required procedure that the lineup be conducted in a certain way, as long as the procedure is not impermissibly suggestive. In Jenkins v. City of N.Y., 478 F.3d 76, (2 nd Cir. 2007), the court held that even though both witnesses were informed that their stolen property had been recovered from a person who was in police custody and both were asked to come to the station for an in-person lineup, this knowledge was not impermissibly suggestive. In so holding, the court stated, When a complainant is brought to a police station to view a lineup, it is implicit that the lineup will contain at least one suspect, otherwise there would be no point whatever in conducting the lineup. Id., quoting People v. Green, 14 A.D.3d 578, 579 (N.Y. App. 2005). The court concluded that the in-person lineup procedure was not so flawed as to undermine probable cause and summary judgment in favor of the defendants on the plaintiff s state and federal false arrest claims for the period after those identifications was appropriate. Id.; see also Thompson v. City of N.Y., 603 F. Supp. 2d 650, 657 (S.D.N.Y. 2009) (holding that the in-person lineup that the plaintiff contended used fillers who did not look like him was not impermissibly suggestive so as to preclude the officer from relying on it to establish probable cause). There is likewise no legal requirement that an in-person identification be conducted instead of a photo array if the suspect is in custody and available for an in person identification. See United States v. Lee, 2016 WL , at *2 (2 nd Cir. Aug. 24, 2016) (holding that there is not a per se rule requiring identification by lineup whenever a suspect is in custody and that even when a suspect is in custody and available for an in-person lineup, a photo array may be used for an identification as long as the photo array is not tainted ). 3. One-on-One or Showup Identifications/Single Photographs As discussed above in Section II(A), police officers commonly use one-on-one identifications (also known as showups or confrontations) immediately or shortly after a crime has occurred when possible suspects and witnesses are still generally in the area where the crime occurred. Again, based on the recognition that these types of identifications are most accurate the closer they occur to the time of the crime, many police departments have policies that dictate the maximum amount of time after a crime has taken place that a one-on-one identification can be conducted. For example, the City of Phoenix Police Department s policy on one-on-one identifications states, As a general rule, confrontations should occur within two hours of the time of the crime. (Phoenix Police Department Operations Order 4.19(6)(A)(2)). However, regardless of the length of time that has occurred between the crime and the identification, courts have held that one-on-one identifications or showups as well as the use of a single photograph, are not impermissibly suggestive in and of themselves depending on the circumstances. For example, in Neil v. Biggers, 409 U.S. at , over the course of seven months the victim viewed photographs of suspects in her home and viewed suspects at the police station in lineups, in showups and photographs, but she never identified anyone as the suspect. Finally, 14 Civil Rights and Governmental Tort Liability January 2017

15 after seven months, officers called her to the police station to view the criminal defendant. Id. The officers attempted to compile a lineup, but there was no one else in the jail that fit the criminal defendant s unusual description, so two officers just walked the criminal defendant past the victim and had him say, Shut up or I ll kill you. Id. The victim immediately identified him as her attacker and stated that she had no doubt that it was him. Id. The court held that under these circumstances, the showup identification did not violate due process because even though the showup was arguably suggestive, the other factors considered by the court, including her opportunity to view the criminal at the time of the crime, her degree of attention, her accurate prior description of the criminal, and her certainty in identifying the suspect, weighed in favor of the identification being reliable. Id. at 201. Therefore, even a one-on-one showup can be sufficient to support probable cause to arrest under the right circumstances. B. Did the Witness Exhibit Sufficient Indicia of Reliability? If, and only if, a court determines that the procedure used by the police officer for the eyewitness identification was impermissibly suggestive, does the court undertake the second inquiry, which is whether the witness exhibited sufficient indicia of reliability to support the integrity of their identification. Neil, 409 U.S. at ; Grant, 315 F.3d at Even if the identification is impermissibly suggestive, if the witness exhibits sufficient indicia of reliability, the identification can still form a proper basis for a finding of probable cause. Id. The five factors to be considered by the court in making that determination are each analyzed as follows. 1. Opportunity to view the criminal at the time of the crime Courts have found that this factor is met if the witness had time to observe the suspect and was close enough with conducive lighting conditions and no obstructions to get a good look at the suspect. See, e.g., United States v. Hammond, 666 F.2d 435, 440 (9 th Cir. 1982) (holding that the two eyewitnesses were able to see the bank robber from a distance of only a few feet for several minutes and the bank robber was acting suspiciously when he first entered the bank prompting one of the witnesses to activate the bank s security cameras). However, even if the witness only had a limited opportunity to view an individual during a crime that does not preclude an ultimate conclusion of constitutional reliability where there are other sufficient indicia of reliability. U.S. ex rel. Kosik v. Napoli, 814 F.2d 1151, 1157 (7 th Cir. 1987) (stating that this factor only negates the reliability of the eyewitness identification if the witness had no opportunity to view the criminal at the time of the crime). Therefore, as long as a witness had some opportunity to view the suspect, this factor should support reliability. 2. Degree of attention paid to the criminal A witness who can provide a minimal physical description of a suspect is typically found to have paid sufficient attention to the suspect to provide a reliable identification. See United States v. Hill, 967 F.2d 226, 232 (6 th Cir. 1992) (holding that a bank teller who kept her attention on the bank robber who was constantly yelling at her from no more than two to three feet away for most of the time had paid sufficient attention to the robber to make a reliable identification). In contrast, if a witness cannot provide any description of the suspect, it is clear that he or she did not pay sufficient attention to the suspect to make a reliable identification. See Velez v. Schmer, 724 F.2d 249, (1st Cir. 1984) (holding that the witnesses were able to give very detailed descriptions of the car and the gun involved in the crime, but were not able to describe the driver of the car and this demonstrates that they did not pay sufficient attention to the driver to provide a reliable identification). Thus, for this factor to weigh in favor of reliability, there must be some evidence that the witness paid sufficient attention to the suspect to identify him or her. I Saw You but Did I Really?: Eyewitness Identification Issues in Civil Cases Berke et al. 15