August 26, 2013
Most law enforcement agencies in Virginia continue to use outdated eyewitness identification procedures two years after the state recommended a model policy, a study by a University of Virginia law professor found.
Brandon Garrett released findings of his survey of police lineup procedures Monday. Only 6 percent of the 144 police departments and sheriff's offices that supplied their policies have implemented the "best practices" adopted by the state Department of Criminal Justice Services in 2011, Garrett said. The vast majority are still using outdated policies from 2005 and 1993.
One-fifth of the 201 agencies that responded to the survey had no written policy at all, in violation of a 2005 state law.
According to Garrett, the failure to adopt best practices can have dire consequences. Thirteen of the 16 people wrongly convicted in Virginia and later exonerated by DNA evidence originally were misidentified by eyewitnesses, he said.
One of the study's most troubling findings, Garrett said, is that most agencies still don't use "blind" lineups in which the officer in charge doesn't know which person in the live or photo lineup is the actual suspect. Only 42 agencies require blind lineups, while 16 make them optional.
Garrett said blind lineups are central to the model policy and are the most crucial reform agencies can make.
Garrett said, "It's really hard not to give unintentional cues. Especially if you are doing it one at a time. You can say "is this the one?", "Is this the one?". You might linger over a photograph without even meaning too. It's really simple. All you have to do is have six photographs, shuffle them, and give them to someone else to show the eyewitness and that's it."
The study found that 33 agencies require or allow the use of sequential lineups — when pictures are shown to a witness one at a time — but without making sure the officer doesn't know which photo shows the suspect. Garrett said such a procedure is especially error-prone because it prolongs interaction with the administrator, providing more opportunity for the witness to pick up cues from the officer running the lineup.
Also, only nine agencies authorize the use of a "folder shuffle" method recommended in the DCJS model policy for small agencies that may not have an officer who doesn't know the identity of the suspect. Under that method, the lineup is made blind by putting photos in folders, adding a few blanks, shuffling them and allowing the eyewitness — but not the administrator — to look inside the folders.
The General Assembly has resisted mandating specific eyewitness ID procedures, trusting the agencies to adopt best practices. Garrett concluded that if police don't improve their procedures, stronger action should be taken.