Eyewitness testimony isn’t most reliable

By LINDA P. CAMPBELL

Think you could positively, without doubt, identify the criminal who victimized you?

Jennifer Thompson was certain she could. But she was wrong.

After Thompson mistakenly identified Ronald Cotton as having raped her, he spent 11 years in prison.

But DNA testing finally corrected that error in 1995 and identified the real perpetrator as a convicted rapist named Bobby Poole.

After Cotton’s exoneration, he and Thompson-Cannino (her married name) formed a close friendship and now speak all over the country about the procedural flaws that too often undermine the credibility of witness identifications. Working with writer Erin Torneo, they’ve put their experience of tragedy and triumph into a book, “Picking Cotton.” It’s an absorbing howdithappen — unraveling not a mystery but an injustice.

The dissection of their case illustrates how the criminal justice system can get things wrong when well-meaning people acting with the best of intentions follow procedures that are outdated and inadequate.

Photos in the book show the composite drawing that a police artist drew based on the description that Thompson-Cannino, then a 22-year-old college student in Burlington, N.C., gave the night of the attack.

Determined to find her assailant in a photo spread, she picked Cotton; she later ID’d him in a lineup where she stood across a table from the suspects.

“Ron was the only person who had been in both the photo and the physical lineups, making his face more familiar to me,” she says in the book. “Later, when I looked at the composite we created and at a mug shot of Poole, I thought I had actually done a great job. The problem was that Bobby Poole had not been in my lineup; Ronald Cotton had, and at the time Ronald Cotton most resembled that composite. The standard way eyewitness evidence was collected had failed me, and because of that, I’d failed, too.”

Seeking improvements

Thompson-Cannino and Cotton, along with the detective who handled the case, have worked to improve witness ID procedures that all too often lead to erroneous convictions.

It isn’t just about protecting defendants’ rights to due process. It’s about protecting the integrity of the system, so people believe the real wrongdoers will be punished. It’s about protecting public safety, so predators can’t continue victimizing while authorities pursue erroneous leads. It’s about using tax dollars wisely, so time and money aren’t wasted investigating, prosecuting, incarcerating — and then compensating — innocent people.

Faulty witness testimony was a factor in 33 of the 39 Texas cases since 1994 in which convicted defendants later were cleared through DNA testing (and three-quarters of such cases nationwide), according to the Innocence Project.

Since 2001, Texas has paid almost $9 million to 46 people who were found to have been wrongly convicted, the comptroller’s office said.

After surveying Texas law enforcement agencies last year, the Justice Project concluded that only 12 percent had written lineup policies in line with “best practices” for conducting eyewitness identifications.

(The group queried 1,034 agencies, and 750 responded.)

The most commonly recommended improvements are to caution witnesses that the offender might not be in the lineup and that it’s not mandatory to pick one of those in the group; to have all those in the lineup look as much as possible like the actual suspect; to record or document the lineup proceedings; and to have the procedure conducted by someone who doesn’t know the suspect’s identity.

The point is to avoid sending overt or subtle signals to the witness about whom to pick, or to cause a witness with doubts to suppress them.

Texas is now considering legislation that would require law enforcement agencies to adopt detailed, written policies for line-ups It’s essential reform designed to help get convictions right the first time.

X Linda P. Campbell is a columnist and editorial writer for the Fort Worth Star-Telegram. Distributed by McClatchy-Tribune Information Services.