Hundreds of state criminal cases are under review after the FBI discovered the widespread misrepresentation of microscopic hair analysis.
In 2015, the FBI concluded an evaluation of federal analysts’ testimony that revealed false statements were made in at least 90 percent of trial transcripts analyzed by the Bureau, and 26 of 28 FBI analysts were found to have provided inaccurate testimony or lab reports.
Analysts at the state and local level were also trained by the FBI in hair microscopy, prompting former Director James Comey last year to contact governors across the country, including Gov. Doug Ducey. Comey encouraged state level reviews to ensure the Bureau’s approach did not corrupt state practices.
‘Hair is not an absolute means of identification’
In August, the Arizona Department of Public Safety began the process of assessing roughly 2,300 cases dating back to the late 1980s. And by the end of December, 218 cases were identified for further scrutiny by partners with the state Attorney General’s Office, Arizona State University’s Post-Conviction Clinic and the Court of Appeals.
“The ultimate question that’s trying to be answered is, ‘Is there any situation out there where somebody needs post-conviction relief?’” said Scott Rex, DPS central regional crime lab manager. “And if there is a case where the hair results in some way influenced the case inappropriately – whether it was through court testimony or through the attorneys misrepresenting it, for example in closing arguments – that information needs to be found.”
Rex explained the issue would not be with the hair analysis itself; hair microscopy is the fairly straightforward practice of comparing the characteristics of a suspect’s hair sample to one perhaps found at the scene of a crime or on a piece of evidence.
But the science behind it is not capable of absolutely identifying someone.
Rex, who has short, dark hair, said a long, blonde hair would exclude him from its potential donor pool. But the blond hair could not possibly be narrowed down to a single blond-haired person.
If court testimony or closing statements made by a prosecutor suggested otherwise, that would be entirely inaccurate.
Though DPS analysts have been trained by the FBI, Rex said he has no concerns they would have provided any misleading testimony, unintentionally or otherwise.
“From the beginning, we’ve had a statement that basically says that hair is not an absolute means of identification,” he said. “We have that directly on the report, so it is there for the jury to see. It is there for the attorneys to see. It is very clearly delineated for them.”
Though Arizona Justice Project Executive Director Lindsay Herf commended Rex and DPS for their cooperation, she said the disclaimer that made Rex so confident is not a guarantee.
Herf worked on the FBI’s review for two years and is now on the Arizona team responsible for digging through court records on the cases that need legal consideration.
“The same thing happened with the FBI review, where the disclaimer was on a lot of those reports and errors still came out,” she said, adding some analysts said “some really ridiculous things” while others tried to be careful but still went beyond the bounds of science.
“A lot of these disciplines, like hair microscopy but more so like tire print evidence and that type of thing, are disciplines that weren’t created in a science lab for biology,” she said. “They’re created by law enforcement for use in law enforcement, so there’s a lack of foundation in so many of these disciplines.”
And sometimes mistakes are made.
‘The case where just about everything that could go wrong went wrong’
Herf recalled only one Arizona case in which hair analysis has been conclusively proven wrong.
Ray Krone, whose case is better known for the vast overstatement of bite mark evidence, was twice convicted for the murder of Kim Ancona, a Phoenix bartender found naked and stabbed to death on the floor of the men’s bathroom floor.
She had bite marks on her chest and neck, marks an expert witness wrongfully testified could have only come from Krone.
But a Phoenix crime lab hair analyst also provided inaccurate testimony at the 1992 trial.
Herf said the analyst told the court hairs found on the victim did not exclude Krone, a Caucasian man. In 1996, an FBI analyst offered a different opinion, testifying the hairs not only excluded Krone but came from someone of Native American or Asian heritage. Six years later, DNA analysis proved the FBI analyst right.
Kenneth Phillips was the killer. He lived 500 yards from the bar, Herf said. He had been at the bar that night. Ancona had kicked him out because he was too drunk. And then he came back.
Krone was exonerated. And Herf said the same analyst who testified in Krone’s case could have made other, potentially influential errors.
“Just because looking under a microscope won’t give you the perfect answer, it’s possible that that analyst – with good intentions, doing what they were trying to do – still got it wrong,” she said. “And maybe that’s the only mistake this analyst ever made. Or maybe this analyst made 10 other similar mistakes.”
‘It should match’
Part of the problem may simply be human error, sometimes driven by extraneous information that works its way into labs.
“You might find out that the analyst has information from the detective or prosecutor that, oh, this guy confessed and three witnesses have identified him,” she said.
She doesn’t believe analysts are dishonest, but there could be a little voice reminding them, “It should match. It’s the gun from the guy who confessed. It should match.”
It’s hard to know how many cases like Krone’s have been missed, but Herf’s team is determined to find them if they’re out there.
In May, Herf applied for a grant to fund the monumental job ahead – poring over pages upon pages of testimony.
Crime labs in Tucson, Scottsdale and Mesa have also completed reviews that will add to Herf’s caseload. The Phoenix crime lab, with its vast case history, has not yet performed its own analysis, but some of the grant money would go toward hiring someone to do so.
If the grant does not come through, Herf said the review team is invested enough to see this through.