The FBI’s Hairy Evidence Sends Floridians to Prison on Unscientific Testimony

By Martin Dyckman

Once again, Florida leads the nation in a sobering statistic. At least 42 of its people have gone to prison, including one who was executed and as many as nine who remain on death row, after trials in which FBI “experts” gave testimony that the agency concedes to have been unscientific.

No other state comes close. Pennsylvania was next with 21 cases, followed by Texas with 20, including five executions. There is no immediately obvious reason why Florida should have so many, but that’s something the Legislature should want to know.

The data, announced April 20 at Washington, are the initial results of an unprecedented FBI review of some 2,500 trials, prior to the year 2000, in which its examiners testified that hair from defendants matched hair found on victims or at crime scenes.

The FBI acknowledged claims unsupported by science in 257 of 268 trials, involving 284 defendants nationwide.

The examiners asserted degrees of probability, or even certainty, for which no scientific proof existed. One common error was to say that the hair in evidence could have come only from the defendant and no one else.

To claim such certainty merely from visual microscopic analysis is “junk science” – supposition made to sound like science.

The fallibility was exposed in dramatic fashion a few years ago when DNA testing of hairs in question exonerated three men convicted at Washington, D.C., who had already served long prison terms.

That dramatically confirmed a 2009 National Academy of Sciences finding that “no scientifically accepted statistics exist about the frequency with which particular characteristics of hair are distributed in the population.”

The suspect cases all predate the now-standard use of DNA testing.

The magnitude of the problem is even greater than the data suggest. They involve only cases in which FBI technicians testified or wrote reports, not any in which the witnesses were state employees who had been trained by the FBI.

These findings don’t necessarily mean that the defendants were innocent, but they do warn that junk science contributed at least in some degree to their convictions.

There’s no guarantee of new trials. Some prosecutors will agree. Others will argue that there was enough other evidence to prove guilt. Judges will have to sort it all out.

The trouble with the “harmless error” argument, as my friends at the Innocence Project point out, is the widespread public belief that the FBI can do no wrong. So if an FBI “expert” said the hair sample had to be from Defendant X, jurors may have given the prosecution the benefit of the doubt on other disputed evidence as well.

I wish I could put names to the Florida data – especially concerning the person who was put to death – but there’s a confidentiality agreement among the FBI, the Justice Department, the Innocence Project and the National Association of Criminal Defense Lawyers (NACDL). It calls for informing only the defense attorneys and/or families of the convicted people, who can then reveal the names if they wish. None of the agencies will.

The NACDL is helping to find lawyers for prisoners who don’t have them, according to Norman Reimer, its executive director, who explained why the organization agreed to withhold names.

“This is an unprecedented review,” he told me. “We’re not attempting to make a decision as to the materiality of the evidence.”

There’s a hint, however, as to the identity of at least one of the Florida defendants. The statistics say there was flawed hair analysis in the case of one Floridian who was later exonerated.

That may describe Wilton Dedge, who spent 22 years in prison for a rape that DNA testing eventually proved he did not commit. According to the Innocence Project’s web site, a hair expert – it isn’t clear whether from the FBI or a state crime lab – testified that a pubic hair found in the victim’s bed was consistent with his. It was the only physical evidence against him. Prosecutors also relied on shaky eyewitness identification, a since-discredited dog handler’s testimony and a notorious jailhouse snitch whose perjury was exposed by the DNA test.

Dedge’s ordeal reflects a horrifying reality. When innocent people are convicted, there is often more than one reason why justice miscarried. Eyewitness misidentification is the most common, according to the Innocence Project, a factor in 235 of the first 325 DNA exonerations. Improper or invalid forensic testimony, such as microscopic hair analysis, was involved in nearly half of them. False confessions figured in 88 cases and lying informants in 48.

It bears remembering, though, that the 325 people whose innocence was proven by DNA are only those in whose cases there were physical samples to test. The prisons abound with people who don’t belong there but can’t prove it.

“The law holds it better that ten guilty persons escape, than that one innocent person suffer,” wrote the English scholar William Blackstone.

It does so in theory. In practice, perhaps not.

Martin Dyckman is a retired associate editor of the St. Petersburg Times and author of Floridian of His Century: The Courage of Governor LeRoy Collins. He lives in Western North Carolina.