Explore Harvard's Nieman network Nieman Fellowships Nieman Lab Nieman Reports Nieman Storyboard

Mistaken eyewitness identifications: The leading cause of wrongful convictions

ASK THIS | June 13, 2007

Are your local law-enforcement agencies using these scientifically-designed safeguards to minimize the risk of eyewitnesses mistakenly identifying innocent people? If not, why not?

By Gary L. Wells

Questions that should be asked of police and prosecutors regarding eyewitness identification evidence in your area:

Q. Does your agency have written policies and procedures on live and photo lineups?

Q. What are those policies and procedures? Can we get a copy?

Q. Do your policies and procedures follow the recommendations that have been made by the U.S. Department of Justice and by eyewitness scientists?

Q. How many fillers (non-suspects) are used in your lineups?

Q. What are the procedures used for selecting the fillers to be used in your lineups?

Q. What instructions are given to witnesses prior to viewing a lineup?

Q. Do you use an independent lineup administrator (the double-blind lineup procedure) or do you permit the lineup to be conducted by someone who knows which persons are fillers and which is the suspect?

Q. Do you secure a statement of certainty from the eyewitness at the time of the identification?

Mistaken eyewitness identification is the most common cause of the conviction of innocent people. Since 1992, there have been 200 definitive exonerations of people whose convictions were overturned using forensic DNA testing, and mistaken eyewitness testimony was involved in 154 of those cases.

Scientists who study psychology have examined the mistaken identification problem and made recommendations regarding critical safeguards when conducting police lineups that can help prevent these mistakes.

Although eyewitness identification evidence will never be totally free of error, eyewitness scientists have made a strong case that a substantial portion of eyewitness identification error is attributable to the ways that lineups are conducted. There are no laws dictating how police should conduct eyewitness identification procedures. Instead, each police jurisdiction (there are over 14,000 independent law enforcement agencies in the U.S.) sets its own policies and procedures. Many police agencies have no written procedures or policies for how lineups should be conducted.

Note that most lineups are actually done with photographs, not live lineups. When live lineups are conducted, they often are of a “confirmatory” type in the sense that the eyewitness has already identified the suspect from a photo lineup and hence are mere formalities.

In general, reform procedures that mesh science and practice include the following features:

  • Instructions to the eyewitness prior to the lineup that stress the fact that the actual perpetrator might not be in the lineup and that they should not guess. (Thereby helping relieve the witness of the natural pressures to make an identification.)

  • The use of a minimum of five lineup fillers who fit the description of the perpetrator. (Thereby helping assure that the person the police suspect does not stand out as the obvious choice.)

  • The use of a lineup administrator other than the case detective, a procedure known as a double-blind lineup. This independent administrator is someone who does not know which person is the person of interest and which ones are fillers. (This important feature of a properly-administered lineup assures that the person who administers the lineup to the eyewitness does not inadvertently cue the witness as to the “correct” choice or influence the certainty of the eyewitness.)

  • A formal securing of a statement from the witness as to how certain s/he is at the time of the identification. This certainty statement, secured at the time of the identification by the independent administrator, remains a matter of record that is discoverable at trial. (The natural tendency is for eyewitnesses to become highly certain later about their identification (e.g., at trial) after being briefed by police and prosecutors. Their certainty, however, should be based on their own assessment of their memory at the time of identification, not by pressures that occur later.)

  • Clear records must be maintained with regard to all lineups, not just those that result in a witness identifying the person who was the focus of the investigation. (In many cases there are multiple witnesses and police fail to fully document what the non-identifying witnesses said or did when they viewed the lineup.)

At this time, it is estimated that less than 15% of police jurisdictions have made substantial changes to their lineup procedures to align those procedures with the best science and practices. Jurisdictions that have made sweeping reforms to their eyewitness identification procedures include:

  • New Jersey (the entire state)
  • Boston and many surrounding areas
  • Minneapolis and many surrounding areas
  • North Carolina (nearly all major police departments)
  • Many of the departments in Wisconsin
  • Virginia Beach, VA
  • Santa Clara County, CA

Often, police themselves have a good sense of the vagaries of eyewitness identification evidence. This is because experienced detectives have had the experience of showing lineups to eyewitnesses who then identify fillers (known-innocent lineup members who are in the lineup to merely “fill it out”). Prosecutors, on the other hand, rarely know about these filler identifications because police only forward the “good” identifications. As a result, police tend to be more receptive to these reforms of lineup procedure than prosecutors. This situation inhibits the rate of change because police will not reform their evidence-collection procedures unless prosecutors are fully behind such reforms.

Gary Wells Home Page

The Wells Eyewitness Test

Innocence Project on Eyewitness Identification

Eyewitness Identification Reform Blog

The NiemanWatchdog.org website is no longer being updated. Watchdog stories have a new home in Nieman Reports.