Hiding in Plain Sight: Eyewitness Misidentification and its Role in Wrongful Convictions

Jane Miller, University of New Haven

Photo by engin akyurt on Unsplash

According to the National Registry of Exonerations, it is estimated that an average of 8.9 years of a person’s life is lost per wrongful conviction case (Exonerations in the United States Map, 2023). Eyewitness misidentification is the leading cause of wrongful convictions in the United States (Wells & Olsen, 2003). Despite this, only 25 of the 50 states in America have enacted sweeping core reforms when it comes to how eyewitness identifications are made and handled (Eyewitness Identification Reform, 2020).

With roughly 2 million people currently incarcerated, there is no easy way to accurately predict how many are incarcerated due to a wrongful conviction (Huff, 2003; Sawyer & Wagner, 2013). This makes it even more important to be able to decrease the number of wrongful convictions moving forward. As such, this policy brief will detail the history, current issues, and research surrounding mistaken eyewitness identifications, one of the leading causes of wrongful convictions. It will also include possible policy changes related to this issue, as well as a recommendation for which policy should be pursued moving forward.

Nature and Extent of Wrongful Convictions

Wrongful convictions occur when a defendant is convicted of a crime they did not commit (Huff, 2003). The opposite occurs in a criminal trial when a guilty defendant is not convicted. Both occurrences are inversely related, and both are miscarriages of justice. Both occurrences also result in a guilty offender remaining on the streets. Due to the number of people currently incarcerated in the United States (almost 2 million between local, state, and federal incarceration facilities), it can be hard to accurately determine how many wrongful convictions have occurred, as well as how many exonerations happen each year (Huff, 2003; Sawyer & Wagner, 2013). That being said, latest estimates by the National Registry of Exonerations indicate that 3,284 exonerations have happened since 1989 (Exonerations in the United States map, 2023).

With that many wrongful convictions being overturned, it can be staggering to think of the number of people still serving out sentences under wrongful convictions. The National Registry of Exonerations approximates that a total of 29,101 years of people’s lives have been lost to incarceration due to wrongful conviction (Exonerations in the United States Map, 2023). With the continued advances in technology, the discovery of post-conviction DNA evidence has resulted in over 375 wrongful convictions being overturned (Eyewitness Identification Reform, 2020). While there is a myriad of wrongful conviction causes, mistaken eyewitness identification is a leading cause (Wells & Olsen, 2003). Of those 375 DNA evidence-linked exonerations, eyewitness misidentifications played a role in roughly 69% of those cases (Eyewitness Identification Reform, 2020).

While there are many reforms that can be implemented to help decrease the risk of wrongful convictions in the United States, some are a lot easier and more practical than others. Since it is a leading cause of wrongful convictions, reforming the way eyewitness identifications are handled during investigations can go a long way toward decreasing the number of wrongful convictions. As of 2020, half of the states in the U.S. implemented core eyewitness reforms (Eyewitness Identification Reform, 2020). This policy brief intends to aid the remaining 25 states in starting the eyewitness identification reform process. It contains information about wrongful convictions in general, the current eyewitness identification procedures occurring in many places, descriptions of possible reforms, and policy recommendations.

Explaining Wrongful Convictions

There are a multitude of reasons why a wrongful conviction can occur. Among the most common are mistaken eyewitness identification, forensic science error, plea bargaining, prosecutorial misconduct, judicial misconduct, police misconduct, perjured testimony, false confessions, false allegations, community pressure for conviction, and tunnel vision (Gould & Leo, 2010; Leuschner et al., 2020). Cases can be the result of one or multiple causes.

As was previously mentioned, eyewitness misidentification is the most common cause of wrongful conviction in the United States (Wells & Olsen, 2003). This is because human memory, and therefore recall and eyewitness identification, are not always trustworthy. Psychologists continually warn against the criminal justice system’s reliance on eyewitness identification (Clark, 2012; Loftus, 2019; Wells, 2020). Even 30 years ago, in the late 1980s, one study found that eyewitness identification was a critical piece of evidence in 77,000 out of roughly 2,570,000 cases across 30 states. Especially in cases with little to no physical evidence, eyewitness testimony is often heavily relied upon, despite its weaknesses.

Eyewitness identification usually occurs in one of three ways: a physical lineup, a showup, and a photographic lineup. A physical lineup occurs when the police present a group of people to a witness, one of which is the suspect, and the witness attempts to recognize the suspect. A showup is when a single suspect is shown to the witness, and they are asked if that is the person they saw committing the crime. A photographic lineup is similar to a physical lineup, except instead of showing the witness a group of people, the witness is presented with a group of photographs, one of which is of the suspect (Hails, 2014). Changes in how the lineup occurs (both in person and digitally) are among the most common reforms when it comes to eyewitness identifications (Eyewitness Identification Reform, 2020).

While wrongful convictions and subsequent reforms have been at the forefront of criminal justice research in the last few decades, not all states, let alone police jurisdictions, have undergone reforms. When the idea of eyewitness identification reforms first became popular, it was under the belief that while decreasing the amount of eyewitness misidentifications, reforms in the way lineups were conducted would not decrease the number of correct identifications (Clark, 2012; Wells & Olson, 2003). This was known as the no cost rule, which empirical literature has disproven (Clark, 2012; Clark et al., 2014).

Pre-Existing Policies

Traditional eyewitness identification tactics rely heavily on the police to present a witness with a suspect, either by themselves (a showup) or as part of a larger group or lineup (Hails, 2014). Because of this reliance, the police know who the suspect is, and they can advertently or inadvertently suggest to the eyewitness who they should be identifying (Bull Kovera & Evelo, 2020; Fitzgerald et al., 2018). Who is included in the lineup and how the lineup is presented also impacts the outcome and can increase the risk of a false identification. A few of the most troublesome traditional eyewitness identification practices are explained in detail below.

Single-Blind Lineup

A single-blind lineup, either digitally or in person, is when the person who is conducting the lineup knows the identity of the suspect (Bull Kovera & Evelo, 2020; Eyewitness Identification Reform, 2020). This can lead to the use of behavioral cues, either intentionally or unintentionally, that push a witness towards identifying a particular individual. Even after the initial identification is made, if the person conducting the lineup is aware of who the suspect is, they may also offer feedback that can confirm or deny the witness’ selection. This artificially inflates the confidence of the eyewitness, even if they are uncertain or wrong (Bull Kovera & Evelo, 2020). Confirmation bias is also an issue. Even if the witness is not sure about their selection, people who know who the suspect is in the lineup can feel that even hesitant identification is a positive identification of the suspect.

Perpetrator-Present Lineups

Lineups where a correct identification is made (i.e., where the suspect is present, correctly identified, and guilty of committing the crime in question) are considered perpetrator-present lineups (Clark, 2012; Clark et al., 2014). The opposite, where the actual perpetrator is not present in the lineup- whether a witness identified someone or not- is known as a perpetrator-absent lineup. If the conductors of the lineup do not explicitly state to the participating witness that the lineup does not necessarily contain the actual offender (in other words, that the current lineup could be a perpetrator-absent lineup), witnesses can become convinced that the offender is in the lineup, even if they are not (Clark et al., 2014; Eyewitness Identification Reform, 2020). This can lead to false identifications.

The Use of Fillers

Fillers are the people who are included in the lineup but are not the suspect (Colloff et al., 2016). If no fillers are present, the identification process is a showup instead of a lineup (Hails, 2014). Fillers tend to either look very similar to the suspect or completely different. Fillers who look vastly different from the suspect make the suspect stand out, which can make an eyewitness choose the subject inadvertently (Eyewitness Identification Reform, 2020). Sometimes this is done intentionally, while other times there are certain features of the suspect’s that stand out by chance. When fillers who do not match the suspect are used, it is considered an unfair lineup (Colloff et al, 2016).

Levels of Confidence

Traditional lineups often do not include any formal verification of the witness’ confidence outside of the lineup itself (Eyewitness Identification Reform, 2020). If a statement is made, but is not verifying the witness’ confidence so much as confirming that they selected the suspect, this can also artificially increase the confidence a witness has that they chose the right person, even if they did not. Witness reliability, credibility, and confidence are not always confirmed following a lineup (Huff, 2003).

Policy Options

Reforming eyewitness identification is not a new or novel concept. What has changed in the last few years, however, is how those reforms are thought about and approached. The no cost view asserts that while procedures are beginning to change, the success rate of eyewitness identifications is not changing; only the failure rate is changing (Clark, 2012; Clark et al., 2014). The latest empirical research contradicts that approach, however. Research has found that identification reforms, when done correctly, not only lower the likelihood of a misidentification, but also the likelihood of a correct identification, or any identification at all. This means that reforming the current eyewitness identification process to better protect the innocent may also protect the guilty. The question, therefore, becomes whether this is worth the risk. Outlined below are three policy options for advancing the field of eyewitness identification.

Option 1- Lineup Process Reforms  

Changing how lineups are conducted can decrease the likelihood of false identifications. Double-blind lineups (where neither the witness nor the people conducting the lineup know who the suspect is) should be used instead of single-blind lineups. Eyewitnesses should be reminded that the lineup may not contain the actual offender of the crime in question. The use of showups over lineups should be limited, and the fillers chosen for lineups should look like the suspect. At least 5 suspects should be included in each lineup.

Advantages

• Double-blind lineups and informing the witness of the possibility of a perpetrator-absent lineup decrease the risk of confirmation bias.
• Using lineups instead of showups and choosing the appropriate fillers decreases the likelihood of a false identification due to the suspect simply standing out.
• Both in-person and photo lineups can be used to make things easier, quicker, and less costly.

Disadvantages

• Changes to the lineup process also protect the guilty as well as the innocent.
• Arranging a double-blind lineup can be time consuming.
• Relying on eyewitness identification, even with reforms in place, still means relying on human memory, which is fallible even without the added stress and trauma of witnessing a crime.

Option 2- Vetting Eyewitnesses

The reliability, validity, and credibility of an eyewitness identification should be determined by both a psychologist and a judge prior to inclusion of an eyewitness identification in a trial where eyewitness identification is the main evidence against the defendant. Expert witnesses that can detail the uncertainty in human memory and eyewitness identification in general should be allowed to speak during trials relying on eyewitness identifications. Those conducting the lineup should also take a signed statement from the witness that acknowledges that they understand what they are doing, their level of confidence in their identification, and that they understand the lineup may not contain the actual offender before the lineup begins.

Advantages

• Sworn statements make a witness confront the gravity of what they are doing, which may make them less likely to identify someone if they are unsure if they are who they saw commit the crime.
• Witnesses may be less likely to incorrectly identify someone if they know their identification is going to be scrutinized by a judge.

Disadvantages

• Trauma victims or those who appear hesitant to identify the suspect out of fear may be seen as unconfident and their statement not given the proper credibility.
• Emphasizing the possible unreliability of eyewitness identifications can decrease their weight during a trial, even if the identification is correct.

Option 3- Maintain Current Procedures

Continue using the current eyewitness identification procedures. Make it easier on police to identify a suspect by allowing showups. Keep using single-blind lineups that do not require the use of an outside, unbiased, third party. Do not involve judges or psychologists when determining how much weight to put on an eyewitness identification (leave that up to the investigators). Do not make any new or additional reforms in how eyewitness identifications are performed.

Advantages

• There will not be an increase in protection for the guilty. The number of correct identifications will not decrease.
• No additional time, effort, or money will need to be dedicated to reworking current procedures.
• No new training would be needed for the police.

Disadvantages  

• Eyewitness misidentification will likely continue to be the leading cause of wrongful convictions.
• Innocent people will not have any additional protection against false identification or tunnel vision on the behalf of the police.
• Unfair lineups, even if they do lead to correct identifications, run the risk of being thrown out in court as awareness and support for reforms spread.

Recommendations

Much of the research pertaining to wrongful convictions acknowledges that reforms have started to occur in the last few decades (Bull Kovera & Evelo, 2020; Clark, 2012; Loftus, 2019; Fitzgerald et al., 2018; Wells & Olson, 2003). The Innocence Project, a well-known organization dedicated to research, prevention of wrongful convictions, and the support of exonerees, suggests that true eyewitness identification reforms have only occurred in roughly half of the U.S. states (Eyewitness Identification Reforms, 2020). After careful review of the research, and consideration of the advantages and disadvantages of the policy options available to the states that have not yet initiated widespread reform through the use of court action, legislation, and individual jurisdiction voluntary action, it is the recommendation of this policy brief that sweeping lineup process reforms be put into place.

When the idea of eyewitness identification reforms first took hold in the United States, it was accompanied by the idea of a no cost view. This was the idea that while reforms would decrease the likelihood of false identifications, it would not change the likelihood of correct identifications (Clark, 2012; Clark et al., 2014; Wells & Olsen, 2003). This has since been shown to not be the case. Since reforms would likely decrease the number of eyewitness identifications overall, they would also decrease the number of correct identifications. This means that offenders who actually committed the crime could evade conviction because they were either included in a lineup and did not get identified, or they were never included in a lineup in the first place (Huff, 2003).

While this means that offenders who committed a crime have the possibility of going free, it also means there are more protections for the innocent inherently built into the system. When a wrongful conviction occurs, not only is an innocent person convicted of a crime they did not commit, but the true perpetrator of that crime goes free. It is the position of this policy brief that if there is a chance that perpetrators go free either way, it is better to have more protections in place for the innocent along the way.

Research has shown that conducting double-blind lineups, as opposed to single-blind lineups, reduces the risk of confirmation bias (Bull Kovera & Evelo, 2020). While this might make the construction of the lineup slightly more time consuming, it does not have to prevent a lineup from occurring. Simply having a police officer from another unit conduct the lineup, while having no association with the case or any knowledge of who the suspect is in the lineup, can decrease the chance of the use of conscious or subconscious cues to the eyewitness as to who they should identify. This is a small change that could have lasting benefits, both for the reduction of eyewitness misidentifications (and therefore wrongful convictions), as well as the reliability and credibility of correct eyewitness identifications in court.

Letting the witness know that the lineup potentially could be a perpetrator-absent lineup can help to reduce any pressure placed on an eyewitness to identify someone (Clark, 2012; Clark et al., 2014; Eyewitness Identification Reform, 2020). The manner in which police speak to witnesses before a lineup can greatly impact how the witness approaches the lineup. How the people included in the lineup are determined and how the lineup is presented to the eyewitness is also important (Colloff et al., 2016; Fitzgerald et al., 2018). In addition to the use of fillers who are similar in appearance to the suspect, there is no concrete research showing the superiority of an in-person lineup over a photographic lineup (Fitzgerald et al., 2018). This means that even if the police are unable to get a large enough number of fillers (at least 5) that look similar to the suspect in a lineup in person, they can use a photographic lineup that meets the same criteria in order to still create a fair lineup (Clark, 2012; Fitzgerald et al., 2018). This can help to reduce the cost and time-consumption issues of organizing a fair lineup.

Overall, the suggested policy reforms are small, easy-to-implement steps that can have a major impact on the reduction of wrongful convictions. While it is true that reform is no longer a no cost endeavor, the benefits still outweigh the costs (Clark, 2012; Clark et al., 2014). Maintaining current procedures will not increase the likelihood of a guilty offender going free, but it also will not provide more protection for innocent people in the eyewitness identification process. While reforming the eyewitness identification process is not going to halt all wrongful convictions in the United States, it is a substantial first step toward decreasing their frequency. Therefore, it is the recommendation of this policy brief that any state that has not begun eyewitness identification reforms does so immediately.  

Annotated Bibliography

Bull Kovera, M., & Evelo, A. J. (2020). Improving eyewitness-identification evidence through double-blind lineup administration. Current Directions in Psychological Science, 29(6), 563-568. https://doi.org/10.1177/0963721420969366

This article looks at the use of single- versus double-blind lineups for eyewitness identification, and the impact both have on false identifications. It looks at eyewitness identification through a psychological lens, including the possible body language cues that can be given in a single-blind lineup. This was useful for summarizing the current standard of single-blind lineups and the policy recommendation for double-blind lineups.

Clark, S. E. (2012). Costs and benefits of eyewitness identification reform: Psychological science and public policy. Perspectives on Psychological Science, 7(3), 238-259. https://doi.org/10.1177/1745691612439584

This review analyzes the possible means of eyewitness identification reform, as well as the costs and benefits of various approaches. It details the changes that have occurred in the reform literature, as well as the resulting public policy implications. Recommendations for changes to the current lineup process were pulled from this review.

Clark, S. E., Moreland, M. B., & Gronlund, S. D. (2014). Evolution of the empirical and theoretical foundations of eyewitness identification reform. Psychonomic Bulletin & Review, 21, 251-267. https://doi.org/10.3758/s13423-013-0516-y 

This theoretical review takes an in-depth look at the no-cost view of eyewitness identification reform. It also provides more background on the history of scientific and public opinion surrounding eyewitness identification and its use in criminal cases. Information about the no-cost view was retrieved from this source, as well as details about perpetrator-present and perpetrator-absent lineups.

Colloff, M. F., Wade, K. A., & Strange, D. (2016). Unfair lineups make witnesses more likely to confuse innocent and guilty suspects. Psychological Science, 27(9), 1227-1239. https://doi.org/10.1177/0956797616655789

This empirical article focuses on unfair lineups and the effects of those on eyewitness identification. It describes the experiment conducted by the authors comparing fair and unfair lineups. Descriptions of what constitutes an unfair lineup, as well as descriptions of fillers, were pulled from this article.

Exonerations in the United States Map. (2023). The National Registry of Exonerations. Retrieved April 24, 2023, from https://www.law.umich.edu/special/exoneration/Pages/Exonerations-in-the-United-States-Map.aspx

The National Registry of Exonerations keeps an up-to-date, detailed listing of the known exonerations in the United States. Exonerations prior to 1989 do not have as detailed records, but case details for anything post-1989 are available for analysis. The best estimate for the current number of exonerations, as well as the damage done by wrongful convictions the United States, was retrieved from here.

Eyewitness Identification Reform. (2020). The Innocence Project. Retrieved April 19, 2023, from https://innocenceproject.org/eyewitness-identification-reform/

This website is run by The Innocence Project, a well-known advocate for exonerees and public policy reform that could decrease the number of wrongful convictions. The specific article that I used from the website talks about eyewitness identifications. This source was used for information about traditional lineup procedures, as well as the current state of policy reform.

Fitzgerald, R. J., Price, H. L., & Valentine, T. (2018). Eyewitness identification: Live, photo, and video lineups. Psychology, Public Policy, and Law, 24(3), 307. https://doi.org/10.1037/law0000164

How a lineup is conducted is just as important as who is included in it. This article looks at the different types of lineups, and whether or not live, in-person lineups are superior to other forms. It was used to focus on the pre-existing policies and what types of lineups are best used moving forward.

Goldstein, A. G., Chance, J. E., & Schneller, G. R. (1989). Frequency of eyewitness identification in criminal cases: A survey of prosecutors. Bulletin of the Psychonomic Society, 27(1), 71-74. https://doi.org/10.3758/BF03329902

While older than the rest of the included articles, this research article helps highlight the continued reliance on eyewitness testimony in criminal cases. It looked at 150 prosecutors across 30 states to see the prevalence of eyewitness identification in 1989. The article was used to demonstrate the historical support for eyewitness identification, despite its challenges and flaws.

Gould, J. B., & Leo, R. A. (2010). One hundred years later: Wrongful convictions after a century of research. The Journal of Criminal Law and Criminology, 825-868. http://www.jstor.org/stable/25766110

This review looks at the changes in wrongful convictions over time. It also includes a list of the most common causes of wrongful convictions and what the current research says about the field. The article was used to provide background information on wrongful convictions.           

Hails, J. (2014). Criminal Evidence (8^th ed.). Cengage Learning.

This textbook includes detailed descriptions of the intersection between criminal justice, forensic science, and law. Chapter 16 talks about types of identifications, and illustrates the differences between lineups, showups, and photographic lineups. It was used in the background section of this policy brief.  

Huff, C. (2003). Wrongful conviction: causes and public policy issues. Criminal Justice, 18(1). 15-19.

This article details the state of wrongful conviction research and public awareness in the early 2000s. It also contains a long list of policy suggestions for improving wrongful conviction rates, some of which have been more successful than others. This was useful in the initial stages of researching wrongful convictions reforms, as well as seeing what has changed in the past 20 years.

Leuschner, F., Rettenberger, M., & Dessecker, A. (2020). Imprisoned but innocent: Wrongful convictions and imprisonments in Germany, 1990-2016. Crime & Delinquency, 66(5), 687-711. https://doi.org/10.1177/0011128719833355

Even though it is not focused on wrongful convictions in the United States, this article is important because it provides perspective on the state of international wrongful convictions. It also helps to explain possible causes of wrongful convictions. Information from this article is included in the background section of this paper.  

Loftus, E. F. (2019). Eyewitness testimony. Applied cognitive psychology, 33(4), 498-503. https://doi.org/10.1002/acp.3542

This article provides a more recent look at the reliance of law enforcement on eyewitness testimony. It also looks at the dissidence in the research about identification reforms. Some of the support for the idea that while some changes have occurred, more are required came from this source.

Sawyer, W., & Wagner, P. (2023). Mass incarceration: The whole pie 2023. Prison Policy Initiative. Retrieved April 24, 2023, from https://www.prisonpolicy.org/reports/pie2023.html#:~:text=Together%2C%20these%20systems%20hold%20almost,centers%2C%20state%20psychiatric%20hospitals%2C%20and

This website shows current information about the state of mass incarceration in the United States. It breaks down how many people are incarcerated overall, as well as what kind of incarceration facility they are in (local, state, federal, etc). Statistics from this website were used to highlight the sheer number of people who are currently incarcerated in the US.

Wells, G. L. (2020). Psychological science on eyewitness identification and its impact on police practices and policies. American Psychologist, 75(9), 1316. http://dx.doi.org/10.1037/amp0000749

This is another article that draws on the psychological aspect of eyewitness identification, as well as its pitfalls. The authors detail the fact that concerns over false identification are nothing new, despite the recent push in support of reforms. It also provided connections between the need for DNA exonerations and previous eyewitness identifications for this brief.

Wells, G. L., & Olson, E. A. (2003). Eyewitness testimony. Annual Review of Psychology, 54(1), 277-295. https://doi.org/10.1146/annurev.psych.54.101601.145028

This is the first review of eyewitness identification research published in the Annual Review of Psychology. It focuses on the current literature pertaining to eyewitness identification, the role of DNA exonerations on wrongful convictions, and how often DNA exonerations overturn cases where eyewitness misidentification plays a role in conviction. It is used to discuss the shifts that have been occurring in eyewitness identification in the last few decades.