Preaching to the choir

To many people outside of the criminal justice system, eyewitness testimony is considered the gold standard of evidence. When the general public talks about how sure they want to be of someone's guilt, they always mention eyewitnesses. But the truth about eyewitness testimony is very, very different from the public perception. I believe it is vital that we, the defense bar, get the message out about how very fallible eyewitnesses are, no matter how well-intentioned they may be.

First, let's be clear. There are two kinds of eyewitnesses: people known to the perpetrators and strangers. Obviously, if the eyewitness is the perpetrator's boss, sibling, spouse, etc., we aren't terribly concerned that the eyewitness has misidentified the perpetrator. Those witnesses may have credibility problems, but they're not wrongly saying the person known to them did it when it was really someone else entirely.

Stranger eyewitness identifications are where the problems I'm interested in come in to play. Quite frankly, stranger eyewitnesses are probably wrong as often as they are right. But they're still relied on as strong evidence of guilt. In the book Actual Innocence by Barry Scheck and Peter Neufeld, they looked at 62 cases of wrongful convictions and found mistaken identifications in 52 of them. That book was published in 2000, so we know even more now about how frequently mistaken identifications factor into wrongful convictions.

The case that has me thinking about this today is of Patrick Waller. He was convicted of crimes from a 1992 incident. A young married couple was abducted from the street at gunpoint by two men, robbed, and taken to an abandoned building where the woman was raped by both kidnappers. One of the two kidnappers left DNA evidence behind. An unlucky couple happened upon the building and were also taken inside. Patrick Waller was identified by a cop as one of two men who fled from police in that same area days after the crime. That cop suspected that the men who ran might have been the kidnappers. Police then showed the victims photo line-ups and 3 of the victims picked Waller out. Seems pretty unlikely that 3 of the victims would have picked out the same wrong guy, right?

But we now know that the unlikely scenario is exactly what happened. After years of fighting with the DA's office, defense attorneys finally succeeded in getting DNA tests done that were not available in 1992. All we knew in 1992 was that the blood type of the DNA sample was consistent with Waller's. In 2008, we finally learned that the DNA did not match Waller, but instead matched Byron Bell, who was then serving a 45-year sentence for a burglary that occurred months after the abduction in Waller's case. So it seems that Bell must have been one of the two perps. But that didn't clearly exonerate Waller as there were two kidnappers.

Someone from the DA's Conviction Integrity Unit finally interviewed Bell this spring. By now, the statute of limitations on the incident had expired so Bell faced no possibility of prosecution. So Bell freely confessed to his involvement in the crime. Then came the big question. "Who was your partner in crime?" To everyone's surprise, Bell said it was some guy named Mondo who he hadn't seen since the night of the crime. Sure enough, police tracked down a man named Lemondo Simmons, who also could not now be prosecuted for the crime. So Mondo also freely confessed his involvement in the crime.

What we know, then, is that 3 victims identified the wrong guy. Those victims had no incentive to pick the wrong guy. Surely they all wanted the men who attacked them to be found and put behind bars. What could give those victims greater peace than knowing that their attackers could not hurt them or anyone else? But well-meaning as they were, they were all wrong. How could that be?

That it could be doesn't surprise me all that much. A well-intentioned officer thought those two guys who ran from him might have something to do with the crime. That officer picked out Waller as one of the men who ran. Then well-intentioned officers put Waller's photo into a line-up and showed that line-up to the victims. That well-intentioned officer probably knew which photo was the suspect. Without even realizing, that officer probably provided subtle clues to the victims that helped lead the victims to picking out Waller. We know from studies that this happens, despite the officer's best intentions not to lead the eyewitness. That's why it's always best that the officer administering the line-up not know which photo is the official suspect. Perhaps Waller did closely resemble one of the actual perps.

Then, invariably, once the eyewitness has misidentified a suspect once, that identification becomes cemented in his or her mind. Memory is really fuzzy that way. When the eyewitness sees the defendant in court, s/he will have no hesitation about identifying the defendant as the perpetrator because s/he recognizes the defendant from that photograph. The eyewitness' brain knows it's seen the defendant's face before. So that is how 3 victims can all be absolutely sure in their testimony that the defendant was one of the attackers and be dead wrong at the same time.

What frustrates me is the way police and prosecutors continue to treat eyewitness testimony as if it is the greatest evidence. We know better. Case after case has shown how wrong we all were to rely so much on eyewitnesses. So stop fighting us so much on it. Stop opposing our use of experts to tell the jury about some of the pitfalls of stranger eyewitness identification. We know that a particular subset of cross-racial identifications are particularly suspect, so let us explore those problems in front of the jury. We know through research that the certainty an eyewitness expresses about an identification has absolutely no correlation to the accuracy of that identification. Let us change the standard jury instruction that tells juries they should consider the certainty of the witness. Why on earth would we instruct juries to consider a factor that we know is invalid? Stop stalling about instituting new policies in your jurisdiction about how photo and live line-ups should be conducted. We know through research that there are better ways to conduct those line-ups that produce more reliable results, so why aren't all police forces throughout the country using those procedures?

And for goodness sakes, stop using one-person show-ups! To avoid forever tainting your eyewitness, just wait. Don't bring your newly arrested, handcuffed suspect to the emotionally-charged scene of the crime where your victim or eyewitness is undoubtedly still feeling the shock of being victimized or seeing it happen to someone else. Just wait. It may seem easy just to bring the suspect by and see if the eyewitness confirms your suspicion, but you'll get a much more reliable result if you let it sit for a while and show the eyewitness several people, not just your suspect.

Finally, since we do know how frequently misidentifications have played a role in convicting the wrong guys, the next time you have a defendant requesting new DNA testing on old evidence, don't fight it just because 3 victims identified the guy. Look what happened when you used that excuse this time: the wrong guy spent an extra 7 years in prison and the right guys got off scot free. Everybody loses in that scenario. Well, except the actual bad guys.