Test Applied to DNA Isn't Always A-OK

Case closed. Go to commercial.

DNA evidence has been acclaimed as the magic bullet of crime fighters for nearly two decades. Along the way, William Thompson has become one of its principle skeptics, questioning its infallibility by studying the poor marksmanship that can result when forensic technicians pull the trigger.

"I wouldn't call myself a critic of DNA evidence," said Thompson, a UC Irvine professor of criminology, law and society who is one of the nation's leading experts in the field. "I think DNA evidence is great. I'm a critic of bad DNA evidence."

Thompson's review last year of Houston's crime lab led to its closure and a reconsideration of a decade's worth of criminal convictions that were based on DNA matches from the lab.

In March, a 21-year-old Texas man serving a 25-year sentence for a 1998 rape was released from prison after Thompson's review of DNA evidence in the case. Prosecutors argued at trial that Josiah Sutton's DNA was a 1-in-694,000 match with the rapist's. It took a jury just two hours to convict him. In fact, Thompson was able to show that Sutton's DNA wasn't found in any of the police samples.

None of this surprised Thompson, an attorney who also has a doctoral degree in psychology.

Although cases of innocent people falsely convicted by bad DNA evidence are rare, Thompson sees ambiguous results, sloppy lab work or overstated findings in a quarter of the cases he examines.

The problem, he says, lies in human error and the biases of forensic scientists who work for police agencies.

"I think it's a service industry for law enforcement," he said. "They provide a service to clients, and the clients are law enforcement. Because of that, the culture that develops in police labs is inconsistent with upholding the best scientific values."

That rough critique has earned Thompson praise from an unusual quarter: the Assn. of Forensic DNA Analysts and Administrators.

"I think he's excellent, although some feel that he's a lot more rambunctious than most defense attorneys," said association Chairman Ron Rubocki, a forensic scientist at the University of Nebraska Medical Center.