WASHINGTON, July 18 (UPI) -- A review of death penalty cases in which defendants may have been wrongly tied to crimes has led to changes at state and local levels, U.S. officials said.
The FBI and the Justice Department, in consulting with the Innocence Project and the National Association of Criminal Defense Lawyers, announced the review last summer.
Officials said the Texas Forensic Science Commission last week directed labs under its jurisdiction to take the first step to scrutinize hair cases, The Washington Post reported Wednesday. Texas has executed more defendants than any other state since 1982.
The federal review of old criminal cases found as many as 27 death penalty convictions resulted from FBI forensic experts erroneously tying defendants to crimes with exaggerated scientific testimony, the Post said. The review led to an 11th-hour stay of execution in Mississippi in May.
FBI officials said they intend to review and disclose problems in capital cases even after a defendant has been executed.
"We didn't do this [the review] to be a model for anyone -- other than when there's a problem, you have to face it, and you have to figure how to fix it, move forward and make sure it doesn't happen again," FBI General Counsel Andrew Weissmann said.
The Post said it isn't known how many cases involved errors, how many led to wrong convictions or how many mistakes may jeopardize valid convictions. The FBI plans to release initial report results this summer.
The capital cases are among the first 120 convictions thought to be problematic among more than 21,700 FBI lab files being examined, the Post said. At issue is a former practice in which some FBI experts exaggerated the significance of "matches" taken from microscopic analysis of hair found at crime scenes.
FBI lab chief David Christian Hassell said the review will be used to improve lab training, testimony, audit systems and research.
"One of the things good scientists do is question their assumptions. No matter what the field, what the discipline, those questions should be up for debate," Hassell told the Post. "That's as true in forensics as anything else."