Public defenders are calling on Baltimore police to overhaul the way victims and witnesses pick suspects from photo lineups to ensure detectives can't influence the results.
Police made some revisions last month, but Baltimore Innocence Project director Michele Nethercott and city Public Defender Elizabeth L. Julian want investigators to incorporate more aggressive reforms made in other states into their policies.
City public defenders want the photos of possible suspects shown one at a time, rather than in a group of six. When viewing photos one by one, witnesses would have to repeatedly recall the event, rather than compare the pictures and choose someone that most resembles the person they saw, said Gary L. Wells, a professor of psychology at Iowa State University and expert on witness identification.
Defense attorneys also want the police to require that the person showing the photos not know the identity of the suspect. That way, there's no chance that an unintentional cue - such as a sigh of relief after a witness identifies the suspect - could increase a victim's confidence and prejudice her testimony at trial.
These changes are important, defense lawyers say, because about three-quarters of the more than 200 people exonerated by DNA evidence nationwide were convicted largely based upon witnesses' and victims' identification.
It's harder to get an identification when the person administering the test "doesn't have a stake" in the outcome, Julian said. "But it's proven to be more accurate. As defense counsel, we don't want less IDs. We want more correct IDs, and everyone should be concerned about justice."
Jim Green, deputy legal counsel for the Police Department, said the changes advocated by Julian and the Innocence Project are neither necessary nor practical.
He said the department doesn't have the resources to hire a team of independent photo array administrators.
Police policy requires:
That detectives warn witnesses before photos are shown that a suspect might not appear in the array.
That all arrays be submitted into evidence, including ones in which the suspect wasn't identified.
That the innocent people included in the array - so-called fillers - have a similar appearance to the suspect.
The Police Department "believes in its current system," and its revised policy continues to "promote fairness and accuracy," Green said, adding that he consulted with two division chiefs in the state's attorney's office before issuing the new rules.
Technique in use
New Jersey, North Carolina, Boston and Hennepin County, Minn., which includes Minneapolis, have adopted a newer, "double-blind, sequential" technique being pushed by Nethercott and Julian, Wells said. Under this system, witnesses are shown photos separately, in sequential order, and by someone who doesn't know the identity of the suspect.
Even though this method isn't policy, some city prosecutors are asking police to use it. The technique is most helpful in carjackings and robberies, in which the victim and attacker are often strangers and a correct identification is critical, said Andrea Mason, a supervisor in the state's attorney's firearms unit.
"I've had cases where the identification is suspect," Mason said. "And I don't mean to criticize detectives because the transcripts seem to suggest that unconsciously they're signaling somehow. ... There have been situations where the person ID'd couldn't possibly have done it. He was in jail. Or he was dead."
City State's Attorney Patricia C. Jessamy said she supports her prosecutors' requests for different procedures when they feel it's necessary, a spokeswoman said. Jessamy plans to discuss this issue at a meeting with police Commissioner Frederick H. Bealefeld III in January.
This year the Maryland General Assembly took a step toward uniform standards by requiring every police department in the state to write down its witness-identification procedures and submit them to the Maryland State Police. They are due Jan. 1.
Baltimore turned in its new rules this month because the agency had been researching the issue long before the legislature acted, Green said.
Last year, Green and Bealefeld flew to Chicago to hear the results of the first field tests comparing the double-blind, sequential method with the traditional approach. The study found the newer method didn't work as well as expected, but critics say the test was flawed.
"What this study has done is make the sequential aspect of this more controversial," Wells said. "What should not be controversial in the least is the double-blind. It is a staple idea in science. The only way you can assure that a test was done objectively is that it was done by someone who does not know the so-called correct answer."
The case of Jennifer Thompson Cannino illustrates what can go wrong when the test-giver knows the "correct answer." In 1984, Cannino, then a student at Elon College in North Carolina, was raped by an intruder in her off-campus apartment.