Eyewitness testimony and evidence-based policy
January 9, 2014
Great piece in Slate a few weeks ago that I’ve been remiss in letting go by this long. It’s basically your (all too typical) egregious case of somebody being convicted based on highly suspect eyewitness testimony. E.g.,
In the early afternoon of April 6, 1979, a 78-year-old white man named Jack Sasson was robbed and shot five times at close range as he sat behind the wheel of his blue Chevrolet in the carport outside his home in West Los Angeles. His wife, Renee, testified that when she found him, he was “all blood.” Three weeks later, Sasson died of his injuries.
Brenda Anderson, 19, lived in an apartment on the same street. She told police at the scene that she heard gunshots, looked out of her window, and saw a black man flee the carport, double back, fire more shots, and run off again. Three days later, detectives at the West Los Angeles Police Station showed Anderson a photo array with six pictures of young black men. Anderson, who is African-American, identified the suspect as 18-year-old Kash Register, a former classmate of hers from high school.
A second person also claimed to be an eyewitness: 31-year-old Elliott Singleton. He was shown the same photo array and also identified the suspect as Register. Singleton, who is black, said he’d been painting the house across the street from Sasson’s house when he saw the shooting, and that he chased the armed shooter for blocks, stopping only when the man turned and pointed the gun at him.
The fingerprints on Sasson’s car did not match Register’s. The police never recovered a wallet or a weapon. In a search of Register’s house, they found three brown caps, a pair of black pin-striped pants, and a burgundy shirt, similar to the clothing Singleton said the shooter was wearing. There was a speck of blood on the pants. DNA testing did not exist in 1979, so the blood was simply found to be Type O, matching Sasson, Register, and more than 3 million residents of Los Angeles.
Register, who lived nearby in an apartment with his mother and older brother, said he had spent the morning before the murder at the local unemployment office, where he had finally gotten a lead on a job, and the afternoon having lunch and watching television with his long-time girlfriend, Cheryl Perry. She was pregnant, and he was anxious to start supporting her and their baby. The unemployment office confirmed his account about the time he’d spent there, and his girlfriend backed up his alibi.
Still, Register found himself facing charges of first-degree murder for Sasson’s death—with the possibility of the death penalty. Because the physical evidence was weak, the case against Register turned almost entirely on the eyewitness identifications of Anderson and Singleton.
Oh, and Singleton was a loser drug addict who was clearly willing to tell the cops anything to avoid her own problems. Now there’s a reliable eyewitness for you. Of course, the sad thing is how not surprising this all is:
And yet, according to Brandon Garrett, a law professor at the University of Virginia and the author of Convicting the Innocent, eyewitness misidentifications have played a leading role in nearly 75 percent of 250 convictions overturned by DNA evidence between 1989 and 2010 [emphasis mine]. In more than one-half of those exonerations, the eyewitnesses start off unsure, a “glaring sign” of potential trouble as Garrett puts it, yet appear to become increasingly certain over time. This often corresponds with police practices like suggestive photo arrays, lineups, and even well-intentioned comments like “Good job!” after a witness makes an identification, however tentative. All of this can cause “contamination” of memory, Garrett says so that “there is no way to know after the fact whether the eyewitness could have actually picked the person with any degree of confidence.”
What kills me is how incredibly dumb we are about this at the policy level:
We say we have the best criminal-justice system in the world, and that’s true. But it is also flawed. One giant crack that runs through it is that courts, prosecutors, and juries routinely take eyewitness testimony at face value. Garrett describes as “toothless” the standard the Supreme Court set in 1977 for admitting eyewitness identifications as evidence: “Even in cases with eyewitnesses who were drunk, half blind, observing someone at night, from a distance, it is almost impossible to find examples where appellate judges say it was error” to allow jurors to hear their testimony, he says.
That is starting to change at the state level, but only in fits and starts…These states and a few others have revamped their police practices, trying to weed out the ones that have proven problematic. Still, in most states and the federal system, there are no standardized, evidence-based guidelines for introducing and evaluating eyewitness evidence once a case goes to trial. [emphasis mine]
In short, eyewitness testimony as a category is not at all an evidence-based best practice. In fact, it’s quite often a worst practice. The science, the psychological and neurological research, is at this point extensive and incredibly compelling. Yet it has almost no influence on how our legal system used eyewitness testimony. That’s completely nuts! Honestly, I don’t know what they do in other advanced democracies, but it surely cannot be worse. This is an absolute embarrassment and it frustrates me so that there’s just not enough people who care to change it. The persistent callousness in the obvious face of wrongful convictions– not to mention the simple-minded belief that a guilty person must really be guilty no matter what– is just infuriating. You can do better America. Damn it!
[The whole piece is an amazing narrative of this one case that I didn’t really go into at all, but you should so read it].