This week, USA Today reported that a study prepared for the National Institute of Justice found that most police agencies nationally lacked written policies on eyewitness identifications. This report was conducted for the NIJ by the Police Executive Research Forum. It revealed that 64% of police agencies did not have a written policy for conducting the most common type of identification procedure--the photo array.
Innocence Project suggests that 75% of wrongful convictions are the result of a bad eyewitness identification. As early as 1999, best practices on eyewitness identifications were developed, but not all of the recommended practices have actually been implemented.
Most police agencies permit the lead investigator to conduct photo array identification. (A photo array is a procedure where 5 or 6 photographs are shown to a witness who is asked to identify a suspect.) But scientific research suggests that, whether intentional or unintentional, an investigator who knows who the suspect may be more likely to give subtle clues to witnesses during a photo array. For that reason, the blind administration (by an officer not involved in the investigation) of photo arrays reduces the chance of misidentification.
Another recommendation is to make a video recording of the identification procedure. A person charged with a crime who is required to participate in a live line-up procedure has a right to the presence of counsel whose job is to help ensure that the procedure is fair. There is no right to counsel at a photo array, but photo arrays can also be conducted in an unfair way. Having a video recording of the photo array identification helps make a record of any suggestiveness in the process. Few agencies record their photo array identifications.
In Maryland, police agencies are required by law to have a written eyewitness identification policy. However, most Maryland police departments do not follow the current best practices in eyewitness identifications, such as the blind administration or recording of the procedure.