Statistics show that mistaken eyewitness identification is responsible for about 50% of wrongful convictions.  George Howell almost certainly doesn’t fall into the latter category.  Nonetheless,  the 8th District’s affirmance last week of his conviction and 34-year sentence for a convenient store robbery is vaguely disquieting in its treatment of identification issues.

Two brothers, Sammie and Khalid Fattah, were working at a mini-market in Cleveland when a man walked in, pointed a gun at Khalid, and demanded money.  Most of his face was covered.  A struggle ensued, with Khalid being shot twice and Sammie firing several shots at the robber’s left shoulder and chest.  Khalid and the robber fell out of the front door, and the covering on the robber’s face slipped, allowing Khalid a brief glimpse before the robber ran off.

About an hour and a half later, Howell arrived at a hospital with gunshot wounds to his left shoulder and chest, and a rather implausible story about how he’d acquired them.  His physical description matched that of the mini-market robber, and by coincidence Khalid was in the same hospital, with Sammie visiting him.  The detective told Sammie to take a stroll by Howell’s room, and Sammie confirmed that Howell was the robber.  Howell was arrested, and a gunshot residue test on his right hand proved positive.  Four days later, Khalid identified Howell from a photo array, and also reported that the robber’s jeans had a distinctive rainbow design on the back pockets.  So did Howell’s, it turns out.

The court easily rejected Howell’s claim that the identification procedure was improperly suggestive, but the opinion (or the record) leaves some questions unanswered.  What exactly did the detective tell Sammie when he instructed him to walk by Howell’s room?  Keep in mind that Sammie hadn’t been the one to see the robber’s face when the covering slipped off.  Howell was lying down in a hospital bed, presumably making it difficult to gauge his height.  He was also presumably wearing a hospital gown, not the clothes the robber had worn.  (Howell had actually been transferred from the hospital he had originally gone to.)  Was Sammie really identifying anyone?  Could he have identified anyone?  Or was this simply a situation where the detective said something along the lines of, “They brought in some guy who was shot.  See if that’s the same guy who robbed you”?

Khalid’s identification is more solid, because there’s nothing to indicate that the lineup was conducted in a suggestive manner, and although it’s certainly likely that Sammie and Khalid discussed the robber and Khalid’s impending identification, it’s more difficult to see how Sammie could have provided sufficient detail to allow Khalid to make an ID if he couldn’t have on his own.  Difficult, but not impossible.

Perhaps of greatest concern is that the opinion faithfully recites and analyzes the factors in determining whether the identification was reliable:

a court must consider (1) the witness’s opportunity to view the suspect at the time of the incident, (2) the witness’s degree of attention, (3) the accuracy of the witness’s prior description, (4) the witness’s certainty when identifying the suspect at the time of the confrontation, and (5) the length of time elapsed between the crime and the identification.

These factors were articulated way back in 1972 by the Supreme Court in Neil v. Biggers, and they’re talismanically invoked in virtually each and every case involving identifications.

There’s one problem with them:  they’re bullshit.  They were based upon common-law cases, not hard data, and the empirical research done over the past four decades has shown them to be meaningless.  The witness’ opportunity to view is broadly useless in these types of situations; studies show that witnesses are more likely to recognize a face if they’ve made abstract inferences about it (is this person honest?) than if they make physical judgments (does this person have a big nose?).  Degree of attention is a mixed bag; inattention obviously causes problems, but too much attention, caused by emotional distress, can distort perception, too.  And if there’s a gun, attention is diverted to that; inaccurate identifications are much more common in cases involving a perpetrator with a firearm.  Length of time between the event and the identification might seem significant, but if it’s more than a day, it really doesn’t matter; ability to recall details falls drastically after twenty-four hours.  Level of detail isn’t helpful, either; in fact, studies show that there may even be a negative correlation between recall of peripheral details and the accuracy of the identification.  As for the accuracy of the prior description, research indicates that there’s no “appreciable relationship between a person’s prior description of a face and the person’s accuracy in identifying the face.”

What about certainty?  The literature demonstrating that certainty has no bearing on accuracy is so overwhelming that some states, like Massachusetts and Georgia, have told trial courts to stop instructing jurors to take into account the witness’ professions about the strength of the identification.

Again, this isn’t to criticize the court of appeals.  There was ample other evidence to tie Howell in with the crime, like the distinctive jeans, the unexplained gunshot wounds, the GSW test.  In fact, what might have happened here is what often does in these cases:  the problems with the identification process are shrugged off because of the strength of other evidence.

The real problem here is that too few defense attorneys attempt to present the trial court (and the jury) with the problems of identification testimony, thereby building a record for the appellate courts.  There’s plenty of information out there.  This petition for certiorari by the Innocence Network, seeking to have the Supreme Court review the Biggers factors (and we know how that turned out, don’t we?), is one place to start.  This article from the NACDL has some good pointers, and cites to various scientific articles on the subject.  This article contains a wealth of information on the empirical data.  Ask for an expert witness on identification procedures.  Ask for a Telfaire instruction.  (But be careful; the latter includes an instruction to use the “strength of the identification” in determining its accuracy.  You may want to do some pruning.)

But do something.  Maybe George Howell wasn’t convicted of a crime he didn’t commit, but other people are, all too frequently, and we’ve got to do more to stop it.