The content of The Marshall Project
The Marshall Project is a nonprofit newsroom covering the U.S. criminal justice system. Sign up for our newsletters to receive all of our stories and analysis. In the three decades since Florida sent Gerald Delane Murray to death row for murder, he has watched the case against him crumble. DNA evidence prosecutors used to link him to the crime was handled so badly that the state’s highest court threw it out — twice. A jailhouse witness who testified that Murray confessed to the rape and murder later said he may really have learned about the crime from watching “America’s Most Wanted.” Now, after four separate trials and multiple appeals, Murray and his lawyers argue that the main thing keeping him behind bars is a single piece of evidence: pubic hair found at the scene of the crime. Back in the early 1990s, an FBI analyst looked at the hair under a microscope and found it “consistent” with Murray’s, meaning the hairs shared some visual characteristics. At trial, the agent said that in his experience, it was “highly unlikely” that the hair came from someone other than Murray. But Murray says — and many scientists and lawyers agree — that this kind of testimony is nonsense. That’s because hair, unlike DNA, doesn’t have enough unique characteristics to identify a particular person. “The scientific community says it’s ‘beyond the limits of science,’” Murray told me when I visited him in prison in rural Raiford, in northern Florida. Murray maintains he’s innocent. His lawyers are fighting to get him a new trial, arguing that his conviction hinged on flawed and false testimony. “It comes down to the microscopic hair analysis and a jailhouse rat,” Murray said. “That’s it.” A spokesman for the prosecutors said they wouldn’t comment because the litigation is active. In the past, they have argued that Murray has already had the chance to challenge the hair evidence. In courtrooms across America, “scientific evidence” used to imprison people for heinous crimes has been increasingly discredited. Blood-spatter patterns, arson analysis, bite-mark comparisons, even some fingerprint evidence have all turned out to be unreliable. A quarter of the 3,439 exonerations tracked by the National Registry of Exonerations involved false or misleading forensic evidence. But these exonerations are only the tip of the iceberg, some experts say. Many more people remain incarcerated despite questions about the forensic analysis of evidence used against them. Cases are not automatically reopened when a field of forensics is questioned or even discredited. That’s true of hair analysis, which has been under scrutiny for decades: Government studies have found that in hundreds of cases, hair analysts from the Federal Bureau of Investigation exaggerated their findings in reports and court testimony. A new report by the exoneration registry found 129 cases in which people were falsely convicted at least partly because of flawed hair analysis and testimony. Fifteen of the defendants were sentenced to die. Exonerees lost almost 2,000 years of their lives in prison and cost taxpayers hundreds of millions of dollars. And there may be many more people behind bars who were convicted because of bad hair evidence. “I am willing to speculate that — because the process of exoneration is so difficult — those exonerees could represent between two and 10 times as many wrongly convicted people,” said Simon Cole, a professor at the University of California, Irvine, who is the director of the registry and an author of the study. Many of them “aren’t ever going to be able to prove their innocence to the state’s satisfaction and become exonerated,” he said, noting that overturning a verdict requires very persuasive evidence, such as DNA, and a lot of time, money and legal help. All sentences that involved hair analysis should be reviewed, he said, “to see what other evidence, if any, supported the conviction.” The FBI, which did thousands of hair analyses, continues to say that the technique is valid, and that problems involved exaggerated testimony. But its policy since 2000 has limited the use of hair microscopy to preliminary screening that must be backed up by DNA, a spokeswoman said. From barely visible peach fuzz to velvety locks, wiry mustaches and curly pubic strands, hair covers almost all of our bodies. It’s constantly shedding. The average human head alone contains as many as 120,000 hairs, and loses about 100 a day. As hair falls out, it can transfer to people and clothing, which is why it so often becomes evidence in crime investigations. The first reported use of hair in a U.S. murder case was in 1855, according to a 2016 article in the Virginia Journal of Criminal Law. John Browning and his son were tried for the murder of a cotton plantation manager in the Mississippi Delta, whose injuries included a broken neck. A search of the Browning home revealed a rope tied into a noose containing several drops of blood and a few hairs. Prosecutors presented evidence that the hairs appeared to match the color and length of the victim’s hair. But according to the authors, Chris Fabricant of the Innocence Project and Tucker Carrington of the University of Mississippi Law School, a state supreme court justice found the evidence insufficient, and the Brownings weren’t convicted. The science of hair analysis ascended in the late 1970s after the FBI published the 53-page “Microscopy of Hairs: A Practical Guide and Manual,” identifying characteristics — too small to see with the naked eye — that can be used to compare hairs. Agents trained for a year to learn how to discern the differences and similarities among hairs collected at crime scenes and samples plucked from suspects. By examining characteristics like pigment distribution and structure, the agents could purportedly determine whether a strand came from an animal or a human. Other characteristics helped form narratives about a crime: the person’s race, the area of the body where the hair originated, and whether it was pulled, cut, or fell out. The manual said hair analysis was “not a positive means of personal identification.” But it also said that when two hairs were randomly selected from a population of people, it was rare that analysts couldn’t tell them apart. What the manual didn’t say is that there have never been any scientific studies to show whether three, 300, or 3 million people have the same hair characteristics, legal experts say. Even the hairs on a person’s head can vary — the grays on a man’s temples, for example, look different from the black hairs on his crown. Still, hair microscopy became instrumental in more than 20,000 cases before the year 2000, and led to countless convictions. For decades, FBI agents gave two-week hair-analysis training courses to hundreds of state and local hair examiners, raising concerns that flawed hair testimony may have tainted thousands more cases. The “science” of microscopic hair analysis started unraveling in the 1990s, and not just because DNA analysis emerged on the scene.