The September PCAST Report: The State of DNA in Criminal Courtrooms and the Need for a Well-Informed, Disciplined Judiciary

By Madeleine Jennings
Associate Editor, Vol. 22

On September 20, the President’s Council of Advisors on Science and Technology (PCAST) released a report on the state of the forensic sciences in criminal courts. In recent years, PCAST has issued similar reports on many issues in science and technology, including hearing technologies, education technology, antibiotic resistance, and the future of cities, to name a few. The September report discusses the criteria through which courts assess the scientific validity of certain forensic methods and offers recommendations for their improvement and implementation. It provides a welcome resource for forensic scientists and the judiciary, and its issuance is significant, both practically and symbolically.

The increasing scrutiny and changing consensus surrounding forensic methods, such as bite-mark analysis, ballistics testing, and microscopic-hair-comparison, demonstrate the fragile and evolving character of the forensic sciences. Until recently, DNA technology has been largely immune from these inspecting inquiries.

Discovered by British geneticist Alan Jeffreys in 1984, DNA technology was and continues to be powerful. It is, at this point, an evolving and highly varied technology. Many of the early concerns regarding sample contamination, as made famous in the 1995 O.J. Simpson trial, were soon eliminated or reduced. This reduction occurred with the advent of the FBI’s lab accreditation process, which standardized collection and storage techniques, and with increased media attention to lab conditions and the tightening relationships between labs, police, and prosecutors.

Where DNA was once collected as blood or semen (frequently, the only fluid samples large enough to permit RFLP testing), it is now routinely extracted in extremely small quantities. Many of these small, often degraded samples are examined through mitochondrial DNA analysis. Since a person’s mitochondrial DNA is inherited through maternal bloodlines, it is not specific to a single individual, making its use dubious in a law enforcement context. Further, many of the samples collected today contain DNA from multiple individuals. This often takes the form of “touch DNA,” which can be as little as a few skin cells left on a bicycle’s handle—a handle likely used and touched by multiple individuals.

In analyzing complex mixtures from multiple contributors, analysts often employ low copy number analysis. Through this process, they derive a full DNA profile from extremely small quantities of genetic material by copying DNA fragments and amplifying them into a sample large enough for testing. But stochastic effects appear at much greater frequencies, affecting the reliability of DNA typing results, and studies have shown the extent to which subjectivity and bias are present in the analysis of complex mixtures. Because this method is more error prone and because the errors are less easily identifiable, analysts resort to the use of statistical models to provide likelihood ratios. These models attempt to answer this question: “What is the probability that a person other than the suspect, randomly selected from the population, will have this profile?”

The tools used to provide these likelihood ratios have unavailable or inadequate error rates and, given that there is more genetic variability across (rather than within) racial groups, it can be difficult for analysts to determine the appropriate statistical model, as that determination might depend on knowledge of a suspect’s race. For these reasons, the FBI does not include low-copy number profiles in its DNA database. There is no clear case law on the merits of low-copy number analysis, so judges, often unapprised of the technique’s particularities, are left to assess admissibility on a case-by-case basis. Given the great potential for error and the opacity surrounding the statistical models used, consideration of exculpatory evidence is paramount in cases where such evidence is admitted. This is unfortunately and probably less likely to occur where the defendant is a minority.

In voir dire, where jurors are often asked about their television preferences and evaluated on their penchant for crime TV, prosecutors capitalize on jurors’ unexamined faith in the infallibility of forensic science. CSI has been repeatedly named The Most-Watched Show in the World. Last year, the series finale of the fifteen-year-old show drew 12.2 million viewers and, according to a 2006 Neilson rating, 70 million people watched at least one of the three CSI franchises on one night in 2006.

Studies have examined whether jurors, influenced by crime television, are more likely to acquit criminal defendants when not presented with scientific evidence. Perhaps more important is whether jurors are too quick to convict based on the mere presence of DNA evidence, regardless of the validity of the type and method used. Though the presence of DNA evidence is neither necessary nor always sufficient to establish a defendant’s culpability, jurors often view it as both.

These realities are noteworthy and dangerous since police officers now obtain DNA samples, with increasing frequency, from individuals during the course of routine traffic stops. Law enforcement officers in Florida, North Carolina, Connecticut, and Pennsylvania have justified the practice under the theory that all samples are consensually obtained. But these samples are obtained from unwary individuals, sometimes as young as fifteen. Given the character of the interaction—a uniformed officer, a traffic stop, and a minor (not infrequently, a minority minor)—has the individual really consented to the acquisition of this sample?

The fact that black drivers are pulled over in traffic stops more frequently than their white and Hispanic counterparts and that those stops culminate in more searches of black and Hispanic drivers than white drivers lead easily to the inference that this sample collection practice will disproportionately affect minority populations, not unlike former iterations of stop-and-frisk practices in New York City.

There is a compelling need to better educate laypeople, our once and future jurors, about changes in DNA technology and analysis. Similarly, there is a concurrent need for a well-informed, well-disciplined judiciary, and the September PCAST report acknowledges that, if nothing else.