Eyewitness IDs Can Be Made Better
It's business as usual for many police agencies, even after bungled eyewitness procedures led to high-profile exonerations.
A year ago April, Alan Northrop and Larry Davis walked out of a Clark County, Wash., courthouse on the north banks of the Columbia River, across from Portland, Ore. The two had just been cleared by DNA evidence after serving 17 years in prison for rape and burglary. The real perpetrators have not been found.
In the year since their exonerations, Davis and Northrop have been virtually forgotten.
They were convicted in 1993 based entirely on the eyewitness identification of the victim, who said she was blindfolded and never got a clear view of her attackers. Detectives eventually got enough of a description from her to make a composite drawing of one of the suspects. After several photo arrays and live lineups later, she positively identified both men.
"In this case, they made what's considered a classic error, which is to show the victim a photo montage that contains a suspect, notably Alan Northrop," said Jackie McMurtrie, director of the Innocence Project Northwest, based at the University of Washington.
"She (the victim) doesn't make a selection, and then they show her a lineup where the only person who's the same is Alan Northrop. It's suggestive. Her memory is going to be that she's seen this person before. But she has seen him before in previous lineups."
"The eyewitness identification was as bad as it gets," said John Pantazzis, an attorney on a fellowship at the time with the Innocence Project Northwest Clinic, which represented the men.
He added that the outcome of the case was the kind of thing young lawyers dream about in law school. "For Larry and Alan, finally being vindicated after so long, it was an absolutely terrific experience."
But as the limelight faded, so did the hopes that what led to this horrific mistake could be corrected, or at least prevented from happening to others. A bill to provide compensation to the wrongfully convicted has died in the state Legislature. And no one is looking to change much, if anything, of the police procedures that led to the largest wrongful-conviction case in Washington state history.
Mistakes in eyewitness identifications — like those that put Davis and Northrop in prison for most of their adult lives — occur in 75 percent of convictions overturned by DNA evidence nationwide, according to the Innocence Project, which has helped free 267 people so far. And even as recently as last week, the Los Angeles Police Department exonerated a suspect in the beating of Giants fan Bryan Stow. Investigators could not corroborate eyewitness testimony with forensic evidence.
In that case, at least they made the effort.
"Unfortunately, it's pretty typical, even in this day and age, for police departments to not even have written procedures for conducting eyewitness identifications," said Brandon Garrett, a law professor at the University of Virginia who recently wrote the book Convicting the Innocent: Where Criminal Prosecutions Go Wrong.
"It's somewhat shocking, given how much we know after three decades of pretty intensive laboratory research, looking at eyewitness testimony and the lessons from so many of these cases," Garrett said.
Virginia and North Carolina each passed laws following high-profile exonerations in those states that standardized eyewitness identification procedures. Attorneys general in other states, such as Wisconsin, and police chiefs in various cities, such as Philadelphia, have enacted what the U.S. Department of Justice and the Innocence Project consider model reforms. So far, nothing similar has occurred in Washington, where many police agencies don't even have detailed written guidelines for eyewitness identifications.
Decades of research, led most notably by Gary Wells, a distinguished psychologist in the field of criminal justice, show that the human memory does not recall images like a video camera. Memory is much more malleable and can be strongly influenced by suggestion, no matter how inadvertent, Wells and others have found.
"The primary reform is blind administration of lineups," said Rebecca Brown, the Innocence Project's senior policy advocate for state affairs. That's where someone who doesn't know who the subject is conducts the lineup. "A lot of the research speaks to the fact that there are inadvertent cues that take place all the time."
Brown said the logic is similar to drug trials, where administrators don't know which subjects are receiving a placebo. "All we're suggesting here is that you apply that fundamental scientific principle to the criminal justice setting," Brown said.
In Clark County, double-blind lineups did not occur in the Davis and Northrop cases, and they virtually never occur today, said John Fairgrieve, chief deputy prosecuting attorney. "I haven't seen that," he said.
In 1999, the Department of Justice said police should use blind administrators for all eyewitness lineups, according to the report "Eyewitness Evidence: A Guide for Law Enforcement," produced under then-Attorney General Janet Reno. In 2004, the American Bar Association also endorsed double-blind lineups.
Sequential, rather than simultaneous, lineups are also extremely important, according to Brown. In both photo arrays and live lineups, experts say presenting suspects one at a time lowers the chances of mistaken identity. But it was also found to lower the number of positive identifications, which is why many police agencies, including those in Clark County, are reticent to use it.
Many advisories, including those already mentioned, fall short of advocating directly for sequential lineups because without blind administration, they have the potential to be even more suggestive, Brown said.
"Almost uniformly in our jurisdiction, detectives and officers are using simultaneous photo laydowns, usually with six photos," said Fairgrieve, who cited a 2007 paper in the journal Law and Human Behavior that reported on lower accuracy rates of sequential presentations.
Fairgrieve said a good identification, no matter how it's performed, comes down to how well an officer documents the procedure. In the Davis and Northrop cases, detectives withheld evidence that the victim failed to identify Northrop the first time. Curt Shelton, Northrop's original trial attorney, said he learned this after the trial and presented it to Judge Robert Harris in a motion for retrial. Both the judge and subsequent appellate courts denied the motions.
"In 25 years of law practice, I can't remember another case where a trial judge was so hell-bent on seeing a defendant convicted, regardless of how little evidence there was," Shelton said.
While the recent exonerations haven't changed directly the way officers conduct eyewitness identifications in Clark County, they have made prosecutors more careful, Fairgrieve said.
"It has strengthened our prosecuting, and it's called into question in certain cases the reliability of eyewitness testimony," he said. "We all have to be sensitive to that."