Eyewitness Misidentification
Eyewitness misidentification is the single greatest cause of wrongful convictions nationwide, playing a role in 75% of convictions overturned through DNA testing.
While eyewitness testimony can be persuasive evidence before a judge or jury, 30 years of strong social science research has proven that eyewitness identification is often unreliable. Research shows that the human mind is not like a tape recorder; it neither records events exactly as it sees them, nor recalls events like a tape recorder that has been rewound. Instead, witness memory is like any other evidence at a crime scene; it must be preserved carefully and retrieved methodically, or it can be contaminated.
When witnesses get it wrong
In case after case, DNA has proven what scientists already know – that eyewitness identification is frequently inaccurate. In the wrongful convictions caused by eyewitness misidentification, the circumstances were different, but witnesses, law enforcement officials and juries all relied on testimony that could have been more accurate if reforms proven by science had been implemented. The Innocence Project has worked on cases in which:
- A witness made an identification in a “show-up” procedure from the back of a police car hundreds of feet away from the suspect in a poorly lit parking lot in the middle of the night.
- A witness in a rape case was shown a photo array where only one photo – of the person police suspected was the perpetrator – was marked with an “R.”
- Witnesses substantially changed their description of a perpetrator (including key information such as height, weight and presence of facial hair) after they learned more about a particular suspect.
- Witnesses only made an identification after multiple photo arrays or lineups – and then made hesitant identifications (saying they “thought” the person “might be” the perpetrator, for example), but at trial the jury was told the witnesses did not waver in identifying the suspect.
Variables impacting accuracy of identifications
Leading social science researchers identify two main categories of variables affecting eyewitness identification: estimator variables and system variables.
Estimator variables are those that cannot be controlled by the criminal justice system. They include simple factors like the lighting when the crime took place or the distance from which the witness saw the perpetrator. Estimator variables also include more complex factors, including race (identifications have proven to be less accurate when witnesses are identifying perpetrators of a different race). The presence of a weapon during a crime and the degree of stress or trauma a witness experienced while seeing the perpetrator are also estimator variables that have been proven to reduce the accuracy of identifications.
System variables are those that criminal justice can and should control. They include all of the ways that law enforcement agencies retrieve and record witness memory, such as lineups, photo arrays and other identification procedures. System variables that substantially impact the accuracy of identifications include selection of “fillers” (or members of a lineup or photo array who are not the actual suspect), instructions to witnesses before identification procedures, administration of lineups or photo arrays, and communication after witnesses make an identification.
Click here to learn about reforms the Innocence Project advocates for individual law enforcement agencies and state legislatures.
Decades of solid scientific evidence supports reform
As far back as 1932, experts have known that eyewitness identification is all-too-susceptible to error, and that scientific study should guide reforms for identification procedures. When Yale law professor Edwin Borchard studied 65 wrongful convictions for his pioneering 1932 book, “Convicting the Innocent,” he found that eyewitness misidentification was the leading cause of wrongful convictions.
Since then, hundreds of solid scientific studies (particularly in the last three decades) have affirmed that eyewitness identification is often inaccurate – and that it can be made more accurate by implementing specific reforms to identification procedures.
