Ryan J. Fitzgerald, University of Portsmouth
Heather L. Price, Thompson Rivers University
Tim Valentine, Goldsmiths, University of London

The medium used to present lineup members for eyewitness identification varies according to the location of the criminal investigation. Although in some jurisdictions live lineups remain the default procedure, elsewhere this practice has been replaced with photo or video lineups. This divergence leads to two possibilities: Either some jurisdictions are not using the lineup medium that best facilitates accurate eyewitness identification or the lineup medium has no bearing on the accuracy of eyewitness identification. Photo and video lineups are the more practical options, but proponents of live lineups believe witnesses make better identification decisions when the lineup members are physically present. Here, the authors argue against this live superiority hypothesis. To be superior in practice, the benefits of live presentation would have to be substantial enough to overcome the inherent difficulties of organizing and administering a live lineup. The review of the literature suggests that even in experimental settings, where these difficulties can be minimized, it is not clear that live lineups are superior. The authors conclude that live lineups are rarely the best option in practice and encourage further research to establish which nonlive medium provides the best balance between probative value and practical utility.

photo lineup, video lineup, live lineup, corporeal lineup, identity parade

“There seems to be a conflict— or at least a perceived conflict—between choosing the lineup medium that best facilitates accurate eyewitness identification and choosing the lineup medium that is most practical to construct and administer. Unlike live lineups, which require the lineup members (and other relevant parties) to be physically present for the identification procedure, photo and video lineups can be conveniently constructed by recording an image of the suspect and then choosing fillers from a repository of similarly recorded images. Live lineups are nevertheless preferred in some jurisdictions, which seems to be the consequence of a live superiority hypothesis: The belief that live presentation of lineup members yields the best eyewitness identification outcomes. There are theoretical grounds to predict that an eyewitness would fare best at a live lineup, where the lineup members are observed in their entirety and can even be seen walking or talking. Compare this to a photo lineup, normally composed of static mugshots, and the notion of live superiority seems all the more plausible. Even video lineups, as conventionally practiced, only show the lineup members from the shoulders up, neutrally posed, turning from side to side. To make the right decision, a witness might need access to cues available only at a live lineup. Nevertheless, in spite of its appeal to intuition, we have reservations about the notion of live superiority.” (pp. 307–308)

“We are confident that the live superiority hypothesis exists, but doubtful that it is true. Our review of five countries shows that all either once had or still have policies suggestive of live superiority. But our review of the research literature reveals that empirical tests of the hypothesis are scarce. And from the small corpus of experiments available to review, we find no strong indication that live presentation improves lineup performance. What we do find, however, are numerous factors that could compromise the reliability of eyewitness identification from live lineups.” (p. 308)

England and Wales: “In recent years, video lineups have become the preferred procedure in England and Wales. Police forces now have access to continually growing databases of videos depicting people who have volunteered to be recorded as fillers. […] Current guidelines specify that a suspect should normally appear in a video lineup, but a live lineup may be administered if a video lineup would be impractical. Although photographs may be presented at the detection phase, this is not advised at the evidentiary phase if the suspect is available for a video or live lineup.” (p. 308)

United States: “In the United States, eyewitness identification policies have been conspicuously neutral on the lineup medium. American guidelines typically include separate provisions for live and photo lineups, with no hint at a preference between the two. […] In a nationally representative survey of U.S. police agencies, 94% reported use of photo lineups and only 21% reported use of live lineups. […] At the national level, photo lineups are likely to remain the predominant medium. […] As long as photo lineups continue to be admissible in U.S. courts, it seems only a matter of time before all police agencies in the country discontinue the use of live lineups and rid themselves of the hassle of finding fillers to physically attend the identification procedure.” (pp. 308–309)

Canada: “Photo identification has become the most widely used lineup method in Canada. Surveys show that Canadian police agencies from across the country all use photo lineups, a practice enabled by the admissibility of photo lineup evidence in Canadian courts. […] To our knowledge, video lineups have not been considered for use in Canada.” (p. 309)

Australia: “Australia is unique in that live lineups are preferred in policy, but photo lineups are common in practice. The Evidence Act, 1995 states that lineups should be presented live, provided that (a) the suspect is in police custody at the time of the identification procedure and (b) the suspect does not waive their right to participate in a live lineup. […] The Act makes no reference to video lineups but advises that photo lineups should be inadmissible unless a suspect is unwilling to participate in a live lineup, the suspect’s appearance has changed since the offense, or it would not be “reasonable” to hold a live lineup. To assess whether it would be reasonable, the Act permits consideration of the availability of fillers, the severity of the offense, and the importance of the identification to the case. These contingencies may partly explain the divergence between policy and practice in Australia. Another factor is that suspects are explicitly instructed that they can refuse a live lineup and request a photo lineup instead.” (p. 309)

South Africa: “Although photo lineup evidence has been admitted in South African courts, live lineups remain the common practice. […] The argument is not that photo lineup evidence should be excluded in all circumstances, but rather that live lineups are the better option and that evidence from photo lineups, if presented at trial, should be treated with caution and skepticism.” (p. 310)

Experimental research, Live Versus Photo/Video: “Live lineups are a rarity in eyewitness dentification experiments. Much like the trend toward nonlive procedures in criminal cases, eyewitness scientists have relied almost exclusively on photo lineups (though video lineups are becoming increasingly common, particularly in U.K. research). The paucity of experimental research presents a challenge for the live superiority hypothesis; even if every experimental result supported it, the evidence would not be compelling. […] The experimental literature lends no empirical support for the live superiority hypothesis. If anything, the data converge on the conclusion that witnesses are less likely to identify any of the lineup members if they are observed in person than if they are viewed with videos or photos. Keep in mind, however, that the small number of empirical comparisons involving live lineups limits what can be inferred from this literature.” (pp. 310, 313)

Experimental research, Photo Versus Video: “In our review of empirical comparisons between photo and video lineups, we found some indicators of a benefit for video lineups, but such experiments were exceptions rather than the rule. Further, none of the experiments favoring video lineups had the sample size needed for confidence in the findings. […] In sum, the empirical literature provides no compelling evidence in favor of either photo or video lineups.” (p. 316)

“We can draw three main conclusions from this review. First, international practices and preferences regarding the identification medium are anything but uniform. […] Second, based on the current state of knowledge, the live superiority hypothesis at present is merely a belief—not a fact. The experimental literature provides no clear direction on which medium, if any, is inherently better than the others. Ignoring the obvious practical constraints of organizing a live lineup, we considered whether live identification tests would be superior to nonlive identification tests if all external factors could be neutralized. In spite of our efforts to be as charitable as possible to the live superiority hypothesis, we found little reason to support it. The empirical evidence is inconclusive, and most of the theoretical mechanisms that could be considered consistent with live lineup superiority, such as availability of body and motion cues, could be incorporated into modifications of current video lineup practices. […] Third, a policy preference for live lineups is untenable, due to (a) the lack of empirical support for the live superiority hypothesis; (b) the difficulties of administering a fair live lineup; and (c) the inherent practical advantages of nonlive procedures. Photo and video lineups are more practical, fairer, and seem to be no less reliable than live lineups. Live lineups are hard to organize and difficult to control. The inability to recruit suitable fillers or to prevent any of the lineup players from emitting unwanted cues could easily compromise the fairness of a live lineup. And the proposed safeguards that have been associated with live lineups (right to observe, rogues’ gallery, and transference of familiarity) do not hold up to scrutiny. Contrary to the live superiority hypothesis, we identified several real-world confounds associated with live lineups that could reduce their reliability.” (pp. 320–321)

“In the absence of clear empirical guidance on which medium best facilitates eyewitness identification, police might choose to adopt the most practical medium that the courts in their jurisdiction will admit as evidence. Live lineups, which require witnesses, suspects, fillers, legal representatives, and police personnel to all appear in the same place at the same time, are unquestionably the least practical option.” (p. 316)

“From an investigative perspective, the appeal of photo and video lineups requires no stretch of the imagination. The ability to create and electronically store vast databases of photos and videos transforms the filler recruitment process from hunting for physical people who resemble the suspect (or the culprit’s description) to the mere act of browsing through a computer database filled with images. Once an image of the suspect has been obtained, the primary reason for live lineup cancellations—suspects not showing up—is no longer a concern. Witnesses are also more likely to appear if they do not have to be in the same proximity as the perpetrator. Photo or video lineups can also be presented with a computer, which could be used to deliver clear and consistent instructions, record the witness decision automatically, and visually represent response options that might not be obvious to the witness, such as “not here” or “not sure”” (p. 316)

“In spite of their practicality, the use of nonlive lineups in some jurisdictions runs the risk that the evidence will be given less weight (or even excluded) at trial. Photo lineups are widely used in Canada and the United States because they are practical and judges in these jurisdictions have a history of accepting them as legitimate evidence. But in other jurisdictions, judges consider photo identification to be inferior to live identification. This preference for live lineups has been primarily supported by two overarching claims: (a) identifications from live lineups are more reliable than from nonlive lineups, and (b) live lineups are fairer than nonlive lineups.” (p. 316)

“There are intuitive reasons to believe that, all else equal, live presentation would be the best method to facilitate an accurate eyewitness identification. […] If a crime has been witnessed live, benefits may come from an identification procedure that is also administered live. When the cognitive processes at encoding and retrieval overlap, transfer appropriate processing is theorized to occur. From this perspective, the processes engaged while witnessing a live event would be more likely to be reengaged at a live lineup than at a photo or video lineup. Matching the test medium with the event medium would also conform to the encoding specificity principle, which emphasizes the correspondence between the encoding and retrieval contexts. […] Identification success should be enhanced by the availability of encoded details at test. The availability of dynamic cues, in particular, may increase through live presentation of the lineup members. […] Another potential benefit of live lineups is that the lineup members’ bodies are normally in view, whereas in photo and video lineups usually only the head and shoulders are visible. […] Although the points discussed in this section may provide ammunition for believers in the live superiority hypothesis, we nevertheless remain skeptical of its application to lineups in practice. In addition to the absence of experimental data in support of live lineup superiority, there have been few experimental tests even on the broader issue of live identification superiority. […] Even when issues of fairness or practicality are not considered, the live superiority hypothesis has few legs to stand on. And when these factors are taken into account, the case for live lineups over other procedures falls apart.” (p. 317)

“Live lineups have been frequently cited as the fairest identification procedure. […] We believe that photo and video lineups are at minimum as fair as live lineups (and more likely, fairer). […] Live lineups do not ensure the critical aspects of the testing conditions will be observed. The only meaningful difference between lineup medium types from a right to observe standpoint is that the suspect is present at a live lineup. However, witnesses are commonly permitted to make the identification from behind a one-way screen. And even if the suspect is able to observe the identification, it would be naïve to expect a nonexpert to know about and detect the myriad of factors that could compromise the procedure. From our perspective, the right to observe would be best served by video-recording the lineup procedure whatever the medium—and preserving the conduct of the identification proceeding […] The rogues’ gallery effect is another idea that may have been a legitimate concern in the past, but seems no longer relevant in the digital age. Rogues gallery presumes that if witnesses know that police possess a photograph of the suspect, they will infer that the suspect has a criminal record. But photographs can now be obtained in a variety of ways. […] the rogues’ gallery can be avoided with easily implemented safeguards and the use of photographs in itself need not imply that the lineup members have criminal pasts.” (p. 318)

“The last of the three arguments rests on the assumption that live identifications are inherently superior to photo identifications. Transference of familiarity, which occurs when the memory of a person identified at a photo identification procedure replaces the memory of the criminal observed at the witnessed event, is a legitimate concern associated with repeated identification procedures. Not much is gained from a live lineup if a photo identification has already been conducted because at the second lineup it is never clear whether the suspect was remembered from the witnessed event or from the first lineup. If live identifications were more reliable than photo identifications, there could be an argument against administering a photo lineup because it would taint the witness’s memory and ruin the opportunity to obtain evidence via the more reliable, live procedure. Here lies the problem: the reliability of live over photo identification has not been established. Without clear evidence of live superiority, transference of familiarity has no relevance to the question of whether photo lineups are a suitable alternative to live lineups.” (pp. 318–319)

“A central concern about the fairness of live lineups is that suspects are likely to emit cues and, contrary to nonlive procedures, there is no opportunity for a second take to prevent the witness from observing these cues. […] With video and photo lineups, lineup administrators have greater control over the behaviors of suspects and fillers. If a filler or suspect emits a cue during a live lineup, there are limitations in what can be done to minimize the cue’s influence on the identification decision. […] The risk of the administrator leaking cues to the eyewitness may also be heighted at a live lineup because simple methods of reducing administrator influence like computer-based administration are more difficult than with photo or video lineups. […] Computer administration can reduce the possibility of influence even further by minimizing the social interaction between the administrator and witness or, in the case of self-administration, eliminating interaction altogether while the images are in view. […] The suitability of fillers is central to the fairness of a lineup. Although common practice in the field is to select fillers who resemble the suspect in physical appearance, researchers have proposed that a fair lineup comprises members who all match the witness description of the culprit. […] Constructing a lineup with implausible fillers has the effect of increasing suspect identifications, regardless of whether the suspect is guilty or innocent. Whether trying to match fillers to the witness description or to the suspect’s appearance, the objective would be more easily met by selecting images from an electronic database than by recruiting locally for a live procedure.” (p. 319)

“Previously we considered whether live identifications might be more reliable than video or photo identifications, all other factors held constant. It is important to be mindful, however, that external factors are not held constant in the field and that in practice the lineup medium is inherently confounded. […] First, live lineups are likely to increase witness stress and anxiety. Victims of violent offenses probably do not ever want to see the perpetrator again, but this is precisely what happens at a live lineup. Even witnesses who were not victimized may find live lineups stressful. […] Although witnesses may feel threatened by the prospect of identifying the perpetrator irrespective of the medium, the thought of a live procedure seems to amplify such concerns. […] The second factor, which also has clear implications for memory performance, is cue consistency. Photo and video images provide a record of the suspect at the time of arrest, which may help to increase the consistency between the cues encoded at the event and those available at retrieval. Suspects who are released from custody have the opportunity to make an intentional appearance change prior to appearing at a live lineup. […] Third, live lineups are particularly susceptible to long delays. […] Archival studies suggest that the majority of nonlive lineups occur within a month of the event, but this rarely happens for live lineups. […] Numerous characteristics of live lineups make them difficult to organize quickly: fillers must be recruited to appear in person; a time when all the relevant players are available must be scheduled; and everyone must show up. Timely identification procedures are desirable not only for the efficiency of the justice system, but also because reliability is increased if witnesses are tested when their memory of the perpetrator is fresh.” (p. 320)

“We recommend against live lineups, but believe more evidence is needed before a preference between photo and video lineups can be established. Photo lineups are the most practical option, but the availability of dynamic information in video lineups may improve identification outcomes. More research is needed to determine whether video lineups lead to outcome benefits that justify their practical costs.” (p. 321)

“One avenue for future research would be to modify existing video techniques to maximize their potential benefits. […] There have long been calls to test video lineups that show the entire body and provide additional cues of gait or voice. […] Virtual or augmented reality may be another technological advance with a potential application for lineups. Although the realism of immersive environments could be problematic for victims of traumatic events, some witnesses could benefit from the increased viewing angles and physical context reinstatement that would be possible. Another consideration is whether ethnicity plays any role in lineup medium effects. For example, image quality and the availability of dynamic cues associated with a medium may be more critical when the witness and lineup members have different ethnic backgrounds.” (p. 321)

As always, please join the discussion below if you have thoughts or comments to add!