Assessing Forensic Science

In Issues’ Summer 2002 issue, we had the opportunity to introduce this journal’s readers to several topics that raise complex questions at the intersection of science and the law. That issue dealt with a trio of recent Supreme Court rulings (Daubert, Kumho, and Joiner) on the admissibility of scientific and technical expert testimony; the relationship between the legal protection of intellectual property and the advancement of scientific and technical research; public access to scientific information used in federal regulatory decisions; and balancing the protection of individual privacy and the beneficial use of medical records in public health research. Many of these topics have been analyzed further by the National Academies’ new Science, Technology, and Law Program, which we cochair.

This year, we continue our exploration of issues where science and the law converge by turning our attention to the implications of the Supreme Court’s trio of rulings on the forensic sciences. For 70 years, U.S. courts relied on the standard enunciated in Frye v. United States to determine the admissibility of expert testimony. Under Frye, expert testimony is admissible only if it is “generally accepted” in the relevant scientific community. In Daubert v. Merrell Dow Pharmaceuticals, the Supreme Court, relying on the new Federal Rules of Evidence, declared that scientific expert testimony must be grounded in the methodology and reasoning of science. To determine whether expert testimony meets the Daubert standard, the Court provided trial courts with the following criteria:

• whether the theories or techniques on which the testimony relies are based on a testable hypothesis;
• whether the theory or technique has been subject to peer review;
• whether there is a known or potential rate of error associated with the method;
• whether there are standards controlling the method; and
• whether the method is generally accepted in the relevant scientific community.

These criteria are flexible, and no single one alone would be dispositive. Indeed, the Court recognized that some would be inappropriate under certain circumstances. A few years later, in Kumho Tire, the Supreme Court extended the Daubert standard to apply to expert testimony based on a wide range of technical or specialized disciplines while also recognizing that criteria for admission may differ across areas of expert testimony.

Recently, expert testimony based on forensic evidence has been challenged under the Frye standard, which still governs in many state courts, or under the Daubert standard in federal courts. In Ramirez v. State, the Supreme Court of Florida rejected expert testimony asserting that a knife belonging to the defendant’s girlfriend was the instrument used to inflict a fatal stab wound in the victim’s body, holding that the expert’s methods did not meet the Frye standard. In January 2002, an eminent federal district judge, Louis Pollak, using the Daubert standard, granted a defense motion to preclude expert testimony that purported to identify a specific individual on the basis of matching fingerprints. However, Judge Pollak granted a motion for reconsidering his order and ultimately allowed the prosecution to present identification testimony based on the matching fingerprint. Judge Pollak’s initial ruling, along with challenges to other kinds of forensic evidence, have increasingly led to suggestions that the scientific foundation of many common forensic science techniques may be open to question.

The increased use of DNA analysis, which has undergone extensive validation, has thrown into relief the less firmly credentialed status of other forensic science identification techniques (fingerprints, fiber analysis, hair analysis, ballistics, bite marks, and tool marks). These have not undergone the type of extensive testing and verification that is the hallmark of science elsewhere. These techniques rely on the skill of the examiner, but since the practitioners have not been subjected to rigorous proficiency testing, reliable error rates are not known.

Advances in the forensic sciences have generally emerged to address the needs of the criminal law community. Most of the research has been sponsored by federal agencies whose missions include law enforcement and prosecution, but relatively little science. Scant funding has been provided for competitive basic academic research, and very few, if any, doctoral programs in forensic sciences exist. The culture of academic research, with the free and open exchange of ideas, peer review of research findings, and rigorous disciplinary programs, has not been the norm for the forensic science community.

The challenge of Daubert should lead us to ask how scientific principles can be appropriately applied throughout the forensic sciences, how academic research in the forensic sciences can be promoted, and what the research agenda in this area should be. In the wake of Daubert, the community of forensic scientists may well be pressed to answer these questions in order to maintain their prominent role in U.S. courts.

In this issue, four articles examine the science behind forensic sciences. D. Michael Risinger and Michael J. Saks explore an issue that has already perplexed the Science, Technology, and Law Program: How does the source of funds for and the conduct of research influence the integrity of its outcome? Jennifer Mnookin provides an historical look at the increasing acceptance of fingerprint evidence in law enforcement and in society as well, and explores the implications of the recent legal challenges. David L. Faigman, Stephen E. Fienberg, and Paul C. Stern examine another technique that has long been used in the law enforcement and national security communities: polygraphs. Paul C. Giannelli reviews some of the problems in crime laboratories and makes the case for mandatory accreditation of all forensic labs.

In assessing admissibility under the Daubert standards, courts are seeking a better understanding of the scientific bases of forensic analysis. Courts are inquiring into the relative frequencies at which the identifying traits occur in the general population and the probability of a coincidental match with a crime scene sample. Courts are questioning the standards to which the experts making the identification are held; whether identification is based on objective criteria; and whether standardized minimum criteria must be met for a positive identification. Recognizing that no science consistently produces certain results, courts are also questioning the error rates associated with forensic identification techniques. In these ways, courts are actively seeking an improved understanding of the scientific basis of forensic science and of the body of research required to support expert testimony. We hope the academic and law enforcement communities will do the same.