Center for Policy Alternatives
CPA - Header Photo

Eyewitness Identification

Every year, thousands of Americans are accused or convicted of serious crimes because of mistaken eyewitness identification.
An estimated 4,500 innocent people are convicted in the United States each year because of mistaken eyewitness identification.1 Researchers have long known that mistaken eyewitness identification is the leading cause of wrongful convictions.2 But the use of DNA evidence has led to a new focus on eyewitness identification by police, prosecutors and judges. Of more than 180 individuals who were convicted of crimes and subsequently exonerated by DNA evidence, more than 75 percent involved mistaken eyewitness identification.3
Studies suggest that more than one in four individuals identified as the culprit are innocent.
A series of experimental studies have found that when the perpetrator is in a traditional police lineup, witnesses correctly pick that individual about 50 percent of the time and incorrectly pick someone else about 25 percent of the time. When the perpetrator is absent from the lineup and the witness is presented with a selection of innocent individuals, witnesses identify one of the innocents as the perpetrator about 50 percent of the time.4 These rates of false eyewitness identifications remain roughly the same whether a lineup is in-person or an array of photographs.
Mistaken police lineup identifications distract law enforcement agencies from apprehending perpetrators.
Erroneous eyewitness identifications unintentionally divert police and prosecutors’ attention away from the true culprit. They also undercut the credibility of witnesses and force innocent people to defend themselves from criminal charges.
Law enforcement experts now recognize the problem of mistaken identifications and recommend solutions.
Over the past 25 years, a large body of peer-reviews, scientific research and practice shows that simple and easily implemented systemic changes in administering eyewitness identification procedures can greatly improve the accuracy of eyewitness identifications. The U.S. Department of Justice, the American Bar Association, and states across the nation have endorsed or adopted such reforms.
Four strategies substantially improve eyewitness identifications: “blind” lineup administrators, specific instructions to witnesses, collection of confidence statements, and proper composition of lineup members.
In addition to electronically recording identification procedures, the following strategies are recommended:
  • “Blind” lineup administrators—The most important reform is to ensure that the person who conducts a lineup does not know the suspect’s identity. Commonly, the person who administers a lineup is the case detective who, of course, knows the identity of the suspect. It is well-established by psychologists that a lineup administrator who knows the suspect’s identity will give inadvertent verbal or nonverbal cues that influence the witness.5 The preferred practice is also known as “double blind,” referring to the fact that neither the administrator nor the witness know who police suspect.
  • Specific instructions—The rate of inaccurate identifications is strongly affected by whether witnesses have been warned prior to viewing a lineup that the culprit might or might not appear. Witnesses tend to assume that the perpetrator must be one of the individuals presented, which is one reason 50 percent of eyewitnesses single out an innocent person when the lineup is entirely comprised of innocents. One study found the “might or might not be present” instruction reduced mistaken identifications by 42 percent.6 Witnesses should also be instructed that the lineup administrator does not know the identity of the suspect, so witnesses do not look for nonverbal cues from the administrator.
  • Confidence Statements—A confidence statement is a declaration provided by the eyewitness immediately upon identification and before any feedback is provided, in which he articulates in his own words the level of confidence he has in the identification he has made. It is critical that the eyewitness not be provided any information concerning the selection he has made before a confidence statement is obtained.
  • Proper composition of line-up members—Non-suspect (or “filler”) lineup members should be selected based on their resemblance to the description provided by the witness, yet should not stand out unduly from the suspect. Also, it is generally accepted that photographic lineups should contain at least six photographs and live lineups should contain at least five individuals.
States and localities have adopted eyewitness identification reforms.
New Jersey has adopted this reform package as standard lineup procedure. North Carolina and Wisconsin encourage law enforcement to voluntarily adopt these procedures as well. A number of cities and counties have also implemented these eyewitness identification reforms, including Boston, MA, Minneapolis-St. Paul, MN, Winston-Salem, NC, and Madison, WI.

This policy summary relies in large part on information from the Innocence Project.8

Endnotes
  1. Brian Cutler and Steven Penrod, Mistaken Identification: The Eyewitness, Psychology and the Law, 1995.
  2. Gary Wells et al., “Eyewitness Identification Procedures: Recommendations for Lineups and Photospreads,” Law and Human Behavior, 1998.
  3. Innocence Project, “Eyewitness Identification Reform,” 2006.
  4. Steven Penrod, “Eyewitness Identification Evidence: How Well Are Witnesses and Police Performing?” Criminal Justice, American Bar Association Criminal Justice Section, Spring 2003; Gary Wells and Elizabeth Olson, “Eyewitness Testimony,” Annual Review of Psychology, 2003.
  5. “Eyewitness Testimony.”
  6. Ibid.
  7. Ibid.
  8. The Innocence Project has a variety of model bills addressing eyewitness identification.
Updates