Forensic Hair Analysis: Inadequacy and the Rise of DNA Testing

Introduction: The Problem with Forensic Hair Analysis

One June night thirteen years ago, a killer fired several shots, killing 30-year-old Perry Harder. The killer and an accomplice loaded the body into the back of a van and drove to an isolated spot outside Winnipeg, Manitoba, where they dug a shallow grave and buried the body. Three and a half months later, someone on a passing train spotted Harder's body, and police quickly identified a suspect (1). Their suspect, Jim Driskell, ran a machine shop out of a garage and had agreed to let Harder store some items there. Most of what Harder stored there turned out to be stolen property, discovered during a police raid of the garage. After the raid, police arrested both men. Weeks later, Harder was murdered, and police statements reported that Driskell had committed the murder to prevent Harder from implicating him in the thefts (2). Driskell maintained his innocence, but a jury heard testimony that three hairs tied him to the victim. Driskell was convicted of murder and sentenced to life in prison. However, in recent years, the science of hair analysis has come under scrutiny, and Driskell's conviction has been called into question.

Driskell's case is not simply an isolated incident — there are many individuals serving life sentences or sitting on death row who were convicted as a result of forensic hair analysis. In the past few years, many of those convictions have turned out to be false, setting men and women free based on the inaccuracy and inadequacy of forensic hair analysis. Years of research have led laboratory scientists to conclude that although forensic hair analysis is often instrumental to crime solving, it is essentially inadequate unless extensive nuclear and mitochondrial DNA analysis is also conducted and considered. In Driskell's case, the police had physical evidence that was hard to refute. In the back of Driskell's van, police found three hairs they believed had come from Perry Harder, caught in the van's carpet. Those hairs were scrutinized under a microscope alongside hairs taken from Harder's body (3). At trial, an expert in hair comparison noted any similarities between the samples. If there were more than twenty points of similarity, the hairs could be declared a match (4). At trial, the expert told the jury that in his judgment they did match. Driskell was convicted solely on that physical evidence found in the back of his van.

Cases such as Driskell's have revealed the problems of forensic hair testing and have prompted numerous research studies on the topic. Microscopic hair comparison matches and mistaken identification have been cited in such studies as the most common factors leading to wrongful conviction. Other forensic experts have argued that the practice does have its uses, particularly when combined with the more discriminating technique of mitochondrial testing. Forensic experts have further argued that hair analysis does not carry the same level of reliability as other tests, and that a prosecution should never be based solely on microscopic hair analysis. Microscopic hair analysis has been widely criticized as being based on subjective interpretations without adequate scientific support.

The main arguments against microscopic hair analysis are that it was never intended to identify individuals, because its limits allow it only to determine whether hairs share similar characteristics. Problems in forensic testing have arisen when analysts have overstated their findings or failed to fully explain the limitations of microscopic hair analysis. According to the FBI, a truly accurate hair comparison analysis depends in large part on the individual abilities of the analyst. Research indicates, however, that mitochondrial DNA testing renders findings different from those of a microscopic examination, because mitochondrial DNA allows for further discrimination. For example, a scientist might be able to say that a known piece of hair and an unknown piece of hair share similar characteristics, but it is misleading to suggest that this fact alone implicates a particular person in a crime.

History of Forensic Hair Analysis

The science of hair comparison has been used in thousands of criminal cases throughout the United States for decades. Hair, because of its widespread presence at crime scenes, has great evidentiary value in forensic science. Even when a suspect has attempted to clean the crime scene of personal possessions, fingerprints, or footprints, hair strands almost always remain embedded in the surrounding material. Since hair is so personal to each individual, it serves as a significant indicator of who was present. The forensic testing of hair strands has very early roots — it was first used as evidence in 1861, but did not gain scientific acceptance until after the turn of the century, and did not achieve public acceptance until the late 1950s. During this period, public acceptance of hair testing was widespread, and police received enormous support for evidence consisting of hair strands.

Forensic hair analysis as a formal discipline originated in the 1970s with forensic examiner Barry Gaudette. Gaudette released a study concluding that hair comparison was so reliable it would likely be wrong only 1 in 4,500 times. His testing of hair comparison accuracy made it the courtroom science of choice. According to his research, examiners were correct 99.99% of the time, and jurors rarely doubted the results. However, Stafford Smith, a New Orleans lawyer, discovered a critical flaw in that study. In 1994, while handling a death row case that hinged on hair evidence, Smith concluded there were serious flaws in placing so much trust and emphasis on forensic hair analysis as evidence strong enough to convict an individual. Smith, a scholar in both science and law, began a search for published research but was unable to find any scientific validation justifying the adequacy of forensic hair analysis — with the sole exception of Gaudette's report.

Since Gaudette's report was the only scientific data Smith could locate, he re-ran Gaudette's test with the assistance of a data analyst at Columbia University. The results, published in the prestigious Columbia Law Review, concluded that Gaudette's 1-in-4,500 finding was so seriously flawed it amounted to modern-day snake oil (5). This research served as an eye-opener to those in the detective and forensic professions who had relied so overwhelmingly on Gaudette's report. Gaudette's inaccurate conclusions had been accepted as foolproof even by courts and the broader judicial system. Smith's testing led forensic examiners and police to question the accuracy of forensic hair testing as a sole means of proving guilt. The Boston Globe reported in 2003 that at least four American states were taking a closer look because DNA testing had revealed the overwhelming inaccuracy of forensic examiners. One such examiner in Oklahoma had worked on more than 3,000 cases, yet DNA testing revealed she had been wrong in several of them.

In the past decade, more than twenty people in the United States convicted on the strength of hair analysis have been set free (6). Driskell was among them: upon re-testing, the hairs found in the back of his van did not come from Perry Harder. Further DNA analysis revealed that those hairs did not even match each other, and that they had come from three different people. As a result, the number of invalid convictions based on microscopic hair analysis may be far larger than known. Forensic analysts and scientists have therefore labored to find new methods of analysis upon which to base reliable conclusions.

Hair Analysis: The Process

Hair grows from the hair follicle, and as the hair is being formed in the generative zone of the follicle, it is in contact with and receives nourishment from the blood. As the hair moves upward through the follicle, it no longer receives nourishment from the blood and consists of dead cells (7). Once the hair reaches the surface of the skin, it comes into contact with everything that touches the skin. Because of the original contact with the blood, some researchers believe the chemical makeup of the hair reflects what was present in the blood and the timing of a toxicant or other substance in the bloodstream (8). While this may be true to a very limited extent, several factors make hair an unreliable clinical indicator for certain chemicals that have been in the body (9). Hair comes into contact with numerous chemicals present in the air and in water used to wash it (10). Hair is also in contact with chemicals in shampoos and any dyes, gels, sprays, or other cosmetics that may be applied to it (11).

Since there is no standardized method for removing these external contaminants from hair prior to analysis, the potential for inaccurate results due to external contamination is widespread. There is no way to determine in the laboratory whether a chemical is contained within the hair — and therefore originated from inside the body — or is on the surface of the hair and did not come from within the body (12). An extensive body of scientific research has indicated that hair analysis is unreliable as a diagnostic tool in crime solving. For example, in one study, researchers took hair from the head of a single individual and sent portions of the sample to six different laboratories; the results varied widely from laboratory to laboratory (13). In another report released by the U.S. Agency for Toxic Substances and Disease Registry (ATSDR), the agency concluded that for most substances, the presence of a substance in hair may indicate both internal and external exposure, but that such exposure does not necessarily indicate the source. That report also criticized the lack of standard procedures for sample collection, the absence of standardization of methods and quality assurance among laboratories, and the possible over-interpretation of results well beyond the current body of scientific data and given the limitations of available techniques (14). These findings underscored the early need for a more suitable method of hair testing in addition to traditional microscopic methods.

Conclusion

The future of all DNA analyses looks promising as techniques for smaller-scale and higher-throughput testing become available. While mtDNA analyses do not provide the discrimination potential of some nuclear DNA tests, mtDNA data are often the only information that examiners can gather from degraded evidence that is either old or has been exposed to the environment for a significant period of time. The development of forensic mtDNA sequencing over the past decade has proved helpful to many past cases and will continue to provide useful information to the law enforcement community. In this way, crimes for which innocent people have served time — such as Jim Driskell — will become part of a correctable past. Additionally, forensic hair analysis in combination with nuclear and mitochondrial DNA testing will enhance the speed with which suspects are arrested and charged, as demonstrated by the murder of X and Y.

Although forensic hair analysis has often been instrumental to crime solving, years of research have indicated its significant error potential. The field of crime solving is not one that can tolerate inaccuracy, for human lives depend on its very reliability. In cases involving human lives, even a 95% accuracy rate is widely regarded as completely unacceptable. New methodologies such as nuclear and mitochondrial DNA analysis have, however, given rise to resolutions of both new and old criminal cases. Forensic hair analysis combined with extensive DNA testing is ultimately the more accurate method, as indicated by a review of the literature and research studies in the area of forensic testing. New technological advances in DNA testing offer hope for victims, police detectives, and government officials engaged in crime solving in the United States and internationally. Perhaps there will come a time when innocent individuals cease to be convicted on faulty forensic hair evidence.