After 19 Years in Prison, Man May Be Freed after Sole Eyewitness Recants

The crime happened like this:

A gang-member in South LA was shot dead at 17th and Rimpau. Two neighborhood men, perhaps members of a rival gang, went to look at the scene of the crime. While there, they were both shot in revenge for the original killing. One died, the other survived.

John Edward Smith, a member of the avenging gang, was eventually tried and convicted of the crime, mostly because he couldn't prove he didn't do it.

Lawyer Deirdre O'Connor, founder of Innocence Matters -- a non-profit law firm combatting wrongful convictions -- has fought for Smith for the last two years. She said Smith was only convicted because he was a "good enough fit" for the crime: "Mr. Smith had an alibi, but it wasn't an airtight alibi -- it relied on family members -- and so it was good enough."

It was good enough because, when police conduct an investigation, said Alex Yufik, an expert on forensic evaluations and false testimony, "they are not there for the purpose of determining what actually occurred, they are they to confirm a hypothesis."

Another problem is the eyewitness was escorted to the police station in handcuffs and then shown autopsy photos of his friend who died in the shooting. He would have been very susceptible under that stress to even unintentional cues from the police.

Said Yufik, "If we're shown a number of pictures, we automatically assume that one of the people on the photos has to be the perpetrator. Unless the police give specific instructions.

Those instructions would be something like "these photos may or may not show the face of a person who may or may not have committed this crime." But if police don't give these kinds of instructions, eyewitnesses may recast their memories to match the photos they're looking at.

Even though there's a large and growing body of research showing the unreliability of witness testimony, juries, police, and prosecutors tend to place a lot of weight on eyewitness identifications. Certainty plays well, too.

But, say Yufik, "Research shows that the amount of certainty has nothing to do with the amount of accuracy. In fact, the more certain you appear, the more likely that you're actually incorrect."

The California Commission on Fairness and Justice (CCFAJ) was created by the state legislature in part to study the problems in eyewitness testimony. It released extensive recommendations in 2008.

But police and other investigative services are still inconsistent in implementing the suggestions. That may be due in part to budgetary constraints, said Yufik, but also it's just a cultural shift that's slow to happen.

Deirdre O'Connor of Innocence Matters said the problem with inconsistent implementation is that "if we aren't willing to do it right in every case, we are signing up for a collection of innocent people in custody, because once the system starts to function that way for disfavored people, it will function that way for all of us."

A judge will decide tomorrow afternoon if Smith will be exonerated. The County District Attorney's Office has filed paperwork with the court, concluding, in a rare move, that the witness' recantation severely erodes the case against Smith. O'Connor said it's unlikely that Smith won't be freed in tomorrow's hearing.