Garrett: Supreme Court Unlikely to Solve Problems with Unreliability of Eyewitness Identification
Professor Brandon Garrett, author of "Convicting the Innocent: Where Criminal Prosecutions Go Wrong,” spoke about the problems of eyewitness identification evidence on Thursday.
The U.S. Supreme Court signaled last week that it is unlikely to address some of the key questions surrounding the unreliability of eyewitness identifications, but certain states — most notably New Jersey — are beginning to take steps that reduce the possibility of wrongful criminal convictions based on eyewitness testimony, University of Virginia law professor Brandon Garrett said Thursday.
The Supreme Court on Wednesday heard oral arguments in Perry v. New Hampshire, the court's first case in 34 years involving questions of procedure and due process of eyewitness identification. During arguments, many of the justices seemed reluctant to change how eyewitness testimony is handled in criminal trials.
"The case involves some fairly technical state action issues and [the Supreme Court] probably won't touch on the larger issues we're talking about," said Garrett, speaking at a brown-bag lunch talk on eyewitness identification sponsored by the American Constitution Society.
Garrett's latest book, “Convicting the Innocent: Where Criminal Prosecutions Go Wrong,” examines what happened in the cases of the first 250 wrongfully convicted people who were exonerated by post-conviction DNA testing.
In 190 of the 250 cases, Garrett found, an eyewitness misidentified the wrongfully convicted person. Sixty-eight of those cases involved multiple eyewitnesses.
“Without exception almost, these witnesses were absolutely sure at the time of trial they had picked out the right person,” Garrett said. “So how did that false confidence come about?”
About 2,000 social science experiments have been conducted on issues related to eyewitness identification. The research has found that when a person sees a stranger and forms a memory of the stranger’s face, that memory can be altered when the person is asked to retrieve it.
Police procedures, Garrett said, can not only change an eyewitness’ memory but also boost the eyewitness’ confidence, even when wrong. “It doesn’t take intentional police misconduct or intentional suggestion, to cause a wrongful identification,” he said.
“We all know that it’s hard to see people from far away and we all know that it’s hard to remember faces of strangers. But the idea that the words that police use when they show someone a lineup could affect someone’s memory? That’s not particularly intuitive … Jurors think that [eyewitnesses] who are confident should be trusted and are more reliable. But, in fact, confidence is the most malleable feature of eyewitness memory.”
In some cases, Garrett said, the police used improper eyewitness identification procedures, such as convening a lineup in which the suspect was the only person matching the description.
“We’ve all heard of cases, you know, like in ‘The Wire,’ where someone taps on the photo in the array and says, ‘Take another look at No. 4,’” he said.
However, most of the eyewitness misidentifications did not involve reports of blatant suggestion amounting to misconduct, but rather suggestive procedures that might have seemed like a good idea to police that were not well trained, like conducting a repeat lineup to be sure, or trying to reassure an eyewitness by telling him he did a good job and picked out the suspect, Garrett said.
In evaluating eyewitness evidence, courts rely on a test outlined in the Supreme Court’s 1977 decision in Manson v. Brathwaite that asks whether the police procedures were “unnecessarily suggestive” and whether the identification was nevertheless “reliable,” based on factors including the witness’ level of certainty, degree of attention, accuracy of prior description, opportunity to view and length of time between the crime and the identification.
“You very, very rarely see a court exclude an eyewitness identification applying this test,” Garrett said. “They look at the totality of circumstances. It is easy to find reports of cases with very troubling facts, where judges still say, ‘Well, I think it’s nevertheless reliable, I’ll let it in.’”
The New Jersey Supreme Court recently issued new rules for handling eyewitness identifications in court that incorporate much of the social science about unreliability of eyewitnesses. The court’s decision cited Garrett’s research, as well as that of Virginia Law professor John Monahan.