The Minnesota Lawyer (Minneapolis, MN)
September 1, 2003
Monday Significant changes in the way some jurisdictions conduct eyewitness identification procedures may be within view in Minnesota.
A number of police chiefs in Hennepin County are interested in a pilot project using "double-blind sequential lineups" -- a process in which the investigator conducting the lineup is not told who the suspect is and eyewitnesses are shown suspects one at a time rather than simultaneously.
The theory behind double-blind sequential lineup procedures is that witnesses will focus on their actual recollection of the event and perpetrator rather than comparing the individuals in the lineup. Because the investigator will not know which member of the line up is the suspect, he or she cannot influence the result by making suggestive statements to the witness.
To explore the issue, local prosecutors and police met last week with Dr. Gary L. Wells, a professor at Iowa State University and a leader in the area of eyewitness identification who conceived the double-blind sequential method. Together with representatives of the Innocence Project of Minnesota, Wells spoke at the Criminal Justice Institute on Aug. 26 in Bloomington. (The Innocence Project of Minnesota is a nonprofit organization that seeks to provide a forum for education and dialog among prosecutors, defense attorneys, law enforcement, judges and legislative representatives regarding the common causes of and remedies for wrongful conviction.)
Hennepin County Attorney Amy Klobuchar expressed support for trying the new procedure.
"Our job as prosecutors is not to just get a conviction but to make sure that we have the right person," she observed. "If in fact the pilot shows that [this new system] helps us to get better identifications, then we should do it."
Klobuchar said that Hennepin County prosecutors had a case in New Hope where a witness positively identified a man who looked similar to the actual perpetrator. "The person was charged, then later the DNA showed it wasn't the right person," Klobuchar said. "The case was immediately dismissed and the right person was charged. It turned out that defendant committed three rapes, not just one. That's an example where we do see some false identification cases. Usually we catch them, but you don't want to get to that point."
Wells told Minnesota Lawyer that case law including the DNA exoneration cases that have been brought nationwide by the Innocence Project, have shown that the biggest single cause of the conviction of innocent persons is mistaken eyewitness identification. Eighty percent or more of the DNA exoneration cases have turned out to involve mistaken identification, Wells said.
Long before DNA techniques came to be used, psychologists began researching eyewitness identification and found that mistaken identifications occur surprisingly often and appear just like accurate identifications, according to Wells.
"It turns out one can be wrong and also positive because it's a genuine error," he observed. "It's not something that cross-examination is going to reveal."
Safeguarding the process
While there will always be some possibility of mistaken identification, the criminal justice system has some control over the extent of it, Wells said, estimating that half the mistaken identifications that occur are preventable if certain changes are made to safeguard the process.
Research shows that witnesses choose a person who looks most like the perpetrator relative to the other members of the lineup, Wells said. That works out if the real suspect is in the lineup, but if not, then the witness still may identify someone who looks like the perpetrator, he observed, adding that in virtually every DNA exoneration, the witness never saw a lineup that included the real perpetrator.
"It is a part of human nature to compare one thing to another," Wells noted. "It's actually a very difficult task to recognize the absence of the perpetrator. You have to have a pretty good memory to look at a lineup and say: 'None of these are the guy I saw.'"
Additionally, witnesses often think their task is to identify someone, he continued.
In response to these identification issues, Wells and others working with him have proposed two reforms: the double-blind process and the sequential lineup.
The double-blind lineup -- where the administrator does not know who the suspect is -- is the single most important reform than can be made, said Wells. In a typical lineup, the case detectives know which photo or person is the suspect and there are many ways in which they can unintentionally influence the witness, said Wells.
"For example, let's say the detective knows the suspect is number three and the witness identifies number two," Wells observed. "The detective then may say, 'Make sure you look at all the pictures.' But if the witness said 'number three,' the detective may follow up with, 'That's who we thought it was.'"
If the detective doesn't know who are the suspects and who are the fillers, the detective can't possibly influence the witness either verbally or nonverbally, said Wells.
In the sequential lineup, witnesses see one person or photo at a time and decide each time if the person is the perpetrator. The method is designed to prevent the witness comparing one person to another to decide who looks most like the perpetrator, explained Wells. "This procedure does lower the likelihood that they will pick an innocent person," he said. "If the real person is not in the sequence, they tend to get through the whole sequence without picking anybody."
Still not infallible
These proposed reforms do not make eyewitness identification infallible, stressed Wells. "There is nothing that the legal system can do about bad witnessing conditions or the stress on the witness," he said. "But the legal system can implement these better procedures for the identification that will help protect the innocent. If you get an identification you can more readily trust it."
The countervailing concern to these propositions is that the witness will fail to identify a guilty person, said Wells. Some witnesses who might have been able to identify the actual perpetrator with a simultaneous procedure might not be able to identify him using the sequential procedure, he said. "That's a balance of interests, costs and benefits: It's a policy decision."
A witness who has been unable to identify a perpetrator in a sequential lineup can be shown the suspects all together, Wells stated. "It would be matter of record though, that they could not do it with the sequential procedure, so it would have a lesser status in terms of an identification," he said.
Wells does not believe there are any good arguments against using the double-blind procedure. The only argument that can be made is an expense argument because another officer would have to be involved in the case and testify at trial. The expense is minimal and justifiable because the conviction would be more solid, he said.
The double-blind method has been used in New Jersey for about two years and the expense has been absorbed easily into the state's budget, according to Wells. It's less than an appeal and much less than damages for a wrongful conviction, he noted.
"Having the real person out there committing crimes is very costly, and there is the intangible cost to the credibility of the system," he said.