Forensic evidence carries enormous weight with juries, yet much of it is produced by crime labs that operate inside law enforcement agencies, with uneven funding, limited oversight, and methods that have not always been scientifically validated. When forensic work is wrong, or simply too slow, the consequences can include wrongful convictions that take decades to undo.
In this TalksOnLaw interview, recorded in 2021, Joel Cohen speaks with Professor Brandon Garrett of Duke Law School, author of Autopsy of a Crime Lab, and Dr. Peter Stout, President and CEO of the Houston Forensic Science Center, about where forensic evidence goes wrong and how labs, courts, and lawyers can do better.
When Forensics Fail
Stout and Garrett walk through cases including Josiah Sutton and George Rodriguez, whose wrongful convictions led to the creation of the Houston Forensic Science Center; Lydell Grant, convicted on eyewitness testimony after an inconclusive DNA result and later exonerated through probabilistic genotyping; and Keith Harward, who spent more than three decades in prison on bite-mark testimony. Garrett describes the firearms-examination problems at the District of Columbia's crime lab and why "inconclusive" results and undocumented pattern comparisons can hide error.
Experts, Juries, and the Courts
The conversation examines why jurors treat forensic evidence as nearly infallible, how rarely defense lawyers challenge it or obtain their own experts, and why judges applying Daubert and Rule 702 seldom exclude forensic testimony in criminal cases. Garrett and Stout contrast the thin pretrial discovery in criminal cases with the expert reports and depositions routine in civil litigation, and discuss cognitive bias, false evidence ploys in interrogations, and the plea-bargaining system that leaves most forensic results untested.
Funding, Quality Control, and Reform
Stout explains how crime labs are underfunded relative to the cost of errors, as the Massachusetts drug-lab scandals showed, and why most problems arise from evidence handling and reporting rather than the science itself. The speakers discuss the limits of accreditation and licensing, the case for random audits, the Houston lab's blind quality-control program, the importance of timely results, and why structural independence and transparency help a lab earn the trust of the rest of the justice system.
What to Know Now
Federal Rule of Evidence 702 was amended effective December 1, 2023, to clarify that the proponent of expert testimony must show by a preponderance of the evidence that it is admissible, including that the expert's opinion reflects a reliable application of the method to the facts of the case, a change aimed squarely at the gatekeeping failures discussed here. In Smith v. Arizona (2024), the Supreme Court unanimously held that when a testifying expert relays an absent analyst's statements as the basis for an opinion, those statements are offered for their truth and can trigger the Confrontation Clause, limiting the use of substitute forensic witnesses. The Bureau of Justice Statistics' most recent census of publicly funded crime labs, covering 2020, counted 423 labs that received more than 3.3 million requests. The District of Columbia's Department of Forensic Sciences, whose accreditation was suspended in 2021, has since regained accreditation for several of its units. Professor Garrett's most recent book is Defending Due Process (2025).
