Most forensic science isn’t real science. Try telling that to the criminal justice system.

Earlier this week, the Justice Department announced that Attorney General Jeff Sessions was ending the National Commission on Forensic Science and suspending a review of controversial evidence techniques, opting instead for a new in-house strategy. While Sessions praised the commission’s work, his decision has been widely interpreted as a rebuff to Obama-era efforts to bring higher scientific standards to the forensic techniques that are used in the criminal justice system. David A. Harris, the John E. Murray Scholar at the University of Pittsburgh’s law faculty, has written “Failed Evidence: Why Law Enforcement Resists Science” on the debates over forensics and scientific method. I conducted a short email interview with him on controversies over applying scientific knowledge to forensics and what abolishing the commission means.

HF: People tend to think of police forensics as highly scientific and of confessions and eyewitness testimony as highly reliable. Your book, however argues that the “bottom line” is that “most of forensic science lacks a basis in science,” a finding that has a National Academy of Sciences report to back it up. You also point out problems with eyewitness identification and confessions to police. What has gone wrong with the use of evidence in the criminal justice system?

DAH: Most of what we call forensic science did not begin in any kind of traditional scientific setting, using the scientific method; these forensic procedures came from police investigation efforts, refined in police laboratories. Therefore, most of them bore none of the safeguards and guarantees of traditional science, and did not function that way. They were based on human experience and judgment, not data-grounded work; they were intuitive, rather than experimentally based. The conclusions of the National Academy’s 2009 report were only surprising for people outside the scientific community: With the exception of DNA-based identification, procedures that had always been called forensic science were not science at all, in any real sense. The emperor had no clothes.

Because there was no scientific basis for most of these procedures, forensic testing was not anything like as reliable as most people had thought; errors were made and not caught; new (and ultimately completely unreliable) forensic methods, such as bite mark analysis and hair analysis, were introduced and used, without any kind of validation, with disastrous results. In addition, we have learned much about the errors unintentionally introduced by traditional law enforcement procedures using traditional methods of eyewitness identification and suspect interrogation.

Read the full piece at The Washington Post