By Natasha Vedananda
An unfortunate reality of our criminal justice system is that sometimes, evidence presented as both scientific and reliable turns out to be neither. Sadly, the weight of this poor evidence skews against the criminal defendant and is extremely difficult to overcome on appeal. Thus substandard quasi-scientific data may lead to a loss of liberty or life, an indefensible imbalance of the scales of justice. Recent years have hosted shocking discoveries about the unreliability of microscopic hair comparison evidence used in criminal prosecutions and the damage to life and liberty caused by reliance on this shoddy pseudo-science. In particular, revelations about the FBI’s use of hair microscopy in federal prosecutions over the last three decades reached scandalous proportions as it became clear that perhaps thousands of innocent people had been wrongly convicted on this faulty evidence. This article will explore the history of the FBI’s support for this controversial and dangerous form of evidence, as well as the recent efforts to review cases in which hair evidence was material to the conviction.
The Writing on the Wall
The FBI Crime Lab has a history of questionable performance. For ten years in the 1990s, Frederic Whitehurst, a chemist and lab supervisor at the FBI Crime Lab accused the lab of producing and manufacturing biased evidence to help prosecutors. Though ignored for years, eventually Mr. Whitehurst’s accusations led to a Justice Department investigation. In April 1997, the Justice Department released a scathing 500 page study of the crime lab. The report cited flawed and inaccurate scientific work and recommended major reforms. Despite this seeming interest in reform, in 2003, there was yet another Justice Department probe into the FBI lab, this time prompted by a lab technician who consistently failed to follow DNA protocol, yet went undetected for two years.
Between 1996 and 2004, the Justice Department reviewed about 6,000 cases that had been handled by thirteen discredited FBI agents. The review was purportedly to look for exculpatory evidence requiring disclosure to defendants. Over 250 convictions were identified in which the FBI agents’ work was critical to the conviction or was otherwise problematic. The Justice Department notified only the prosecutors in these cases. By 2003, it was revealed by an unnamed government official that of the over 250 convictions critically reviewed, only 100 to 150 of the defendants had been notified. The Justice Department insisted that they met their legal and constitutional obligations by informing the prosecutors and had no obligation to inform defendants of forensic problems.
A facet of these earlier reviews concerned hair evidence. In 2002, the FBI analyzed DNA in 80 hair cases. The results showed that FBI agents reported false matches over 11% of the time. A review of the testimony given in hair cases prior to and around this time shows a pattern of disturbing overstatement of the reliability and accuracy of the hair evidence.
Given the secrecy and collusion of government entities, coupled with the exclusion of defense attorneys from earlier, limited review processes, it should come as no surprise that in 2012, yet another enormous FBI crime lab scandal was set to implode.
Scholarly Background on Hair Microscopy
The winter 1996 Columbia Human Rights Law Review held within its pages a prescient article entitled “Forensic Hair Comparison Analysis: Nineteenth Century Science or Twentieth Century Snake Oil?” by Clive A. Stafford Smith and Patrick D. Goodman. Nearly 20 years prior to the eventual unearthing of the FBI hair evidence scandal, two perceptive law students understood the frightening consequences of relying on erratic microscopic hair comparison evidence in criminal cases. The authors pointed out the near complete lack of empirical evidence to confirm the reliability of hair comparison analysis and suggested more stringent criteria for admissibility of hair evidence. They cite the only two studies done on hair microscopy, dated 1973 and 1984. The FBI would later cite these same two outdated and scientifically inaccurate studies in its own scholarly works aimed at justifying and normalizing the use of microscopic hair comparison evidence.
The July 2000 issue of Forensic Science Communications held an article entitled “Part I: Hair Evidence” by FBI Trace Evidence Unit Chief Douglas Deedrick. The article discussed the significance of hair evidence and the manner in which hair evidence was microscopically analyzed. Included were sections about the unique properties of hairs found on different parts of the body, as well as a discussion of the differences between “Caucasoid,” “Mongoloid,” and “Negroid” hairs.
As described in the article, hair comparison analysis involved a two step process: 1) identification of questioned hairs, and 2) comparison of questioned hairs to known hairs. Hairs recovered at a crime scene were known as questioned hairs, while hairs removed from a suspect were called known hairs. The known and questioned hairs were analyzed under two compound light microscopes in a process known as hair microscopy. The known and questioned hairs were placed on individual slides then viewed simultaneously under a magnification range of 40X to 400X.
Three possible conclusions may be reached following microscopic analysis of known and questioned hairs:
The questioned hair exhibits the same microscopic characteristics as the hairs in the known hair sample and, accordingly, is consistent with originating from the source of the known hairs
The questioned hair is microscopically dissimilar to the hairs found in the known hair sample and, accordingly, cannot be associated to the source of the known hairs.
Similarities and slight differences were observed between the questioned hair and hairs in the known hair sample. Accordingly, no conclusion could be reached as to whether the questioned hair originated from the same source as the known hairs.
Despite the wordiness of these conclusions, the article states little about the requirements to reach them. Which and how many “same microscopic characteristics” are required for a finding of consistency? What are the “similarities and slight differences” which would lead to an inconclusive result? How does this differ from a finding of dissimilarity? These questions remain unanswered. Curiously, the article also included the following passage:
The hair examination process involves many different steps, the first of which is to determine whether the hair in question originated from an animal or a human being. If the hair originated from an animal, it is possible to further identify it to a particular type of animal. Although certain hairs can be attributed to species, it is not possible to identify hairs to a specific animal to the exclusion of other similar animals. An example of this occurs when dog hairs can be associated to a particular breed but cannot be identified to a specific dog within that breed.
Despite this clear statement about the inability of this hair analysis technique to determine whether a particular dog hair belongs to a particular dog, the FBI relied on this technique to attribute particular human hairs to particular humans in criminal cases.
The FBI analysts continued to publish papers aimed at legitimizing microscopic hair comparison analysis in the criminal context. The January 2004 issue of Forensic Science Communications included an article entitled “Microscopy of Hair Part 1: A Practical Guide and Manual for Human Hairs” by Douglas Deedrick, now FBI Supervisory Special Agent in the Scientific Analysis Section and Sandra Koch, Physical Scientist and Forensic Examiner in the FBI Trace Evidence Unit.
The article further details the structure of hair and the process of microscopic hair analysis, including specifying fifteen characteristics which should be considered in hair comparison. These characteristics are: race, body area, color, length, tip, root, diameter, cuticle, scales, pigment, medulla, cortex, artificial treatment, damage, and special characteristics. Still, this article is no clearer than the earlier one about how many of these characteristics must be consistent to determine a match. With regard to trial testimony, the FBI authors advise:
Expert witness testimony should include an education component on hair evidence for the jury and judge and a statement of the results as reported. The witness should be prepared to discuss the procedures used in reaching the conclusion(s) and to defend opinions. An expert witness should endeavor to promote a better understanding of the methods of examination, the theory of the transfer of trace materials, and the strengths and limitations of the science.
Clearly, in 2004, the FBI advocated the use of this flawed science in criminal proceedings. A section entitled “Significance and Value” cites the dated 1973 and 1984 studies mentioned above. The section also makes some interesting conclusions about the reliability of a hair comparison analysis:
Human hair identifications are subjective interpretations of objective criteria. The variability and distribution of the microscopic characteristics are useful in determining whether or not a questioned hair could have originated from a particular individual.
It is recognized that hair comparisons do not constitute a basis for absolute personal identification. Whereas hairs cannot be positively identified as originating from a particular individual, it is unusual to find different people having the same hair characteristics. This is based on evidentiary samples received in casework and on proficiency tests prepared in the laboratory.
Despite the seemingly scientific rhetoric employed throughout the article, the absolute lack of any useful detail or data is striking. How many matching criteria determine a legally significant match? How can it be “unusual to find different people having the same hair characteristics” when it is accepted that hairs from the same head may have differing characteristics? Once again, the FBI’s scholarly work on hair comparison analysis begs more questions than it answers.
The lack of any reliable empirical underpinnings to hair microscopy did not go unnoticed by the bench. In 2009, the Widener Law Journal published an article entitled “Twenty-First Century Forensic Science Challenges for Trial Judges in Criminal Cases: Where the ‘Polybutadiene’ Meets the ‘Bitumen’” by Honorable Donald E. Shelton. Judge Shelton devoted a section of the article to hair evidence. Of hair microscopy, he states:
Although used by investigative agencies, including the FBI, the most that it can show is that a hair is consistent with having originated from a particular person but would also be consistent with many others. Prosecutorial and expert claims of a “match” based on hair analysis are clearly overstatements of the capabilities of microscopic hair analysis. And it is clear that substantial errors have occurred using microscopic hair analysis.
Judge Shelton suggests the use of mtDNA analysis of hair instead, stating “[t]he trend is clearly that mtDNA evidence has become the standard for hair analysis testimony and that microscopic analysis alone may no longer be sufficiently reliable for admission.”
Unearthing the Hair Evidence Scandal
Despite years of criticism of the FBI’s use of hair microscopy in criminal cases, the FBI continued to rely on microscopic hair comparison evidence. Then, between 2009 and 2012, three convictions in Washington D.C. were overturned using DNA evidence. Santae A. Tribble was convicted of killing a taxi driver in 1978; Kirk L. Odom was convicted of a sexual assault in 1981; and Donald E. Gates was convicted of the rape and murder of a Georgetown University student in 1982. It so happened that all three defendants had been convicted on FBI microscopic hair comparison evidence which was proven to be completely incorrect. Tribble and Odom were not part of the FBI’s earlier review of cases, while Gates, whose case had been reviewed, spent an extra 12 years in prison because prosecutors did not bother to tell him that an FBI analyst’s flawed work had been material to his conviction. Together these three innocent men spent nearly a century behind bars.
As a result of this trio of flawed FBI hair evidence-based exonerations, the FBI conducted an internal review of cases involving hair microscopy. The findings were shocking. According to the FBI’s own estimates, around 21,000 cases involving hair microscopy would need to be reexamined. The Justice Department announced the review in July 2012. This time, defense attorneys would not be excluded from the process. Instead, the historic 21,000 case review would be in conjunction with the Innocence Project and the National Associate of Criminal Defense Lawyers (NACDL).
The review process begins internally with the FBI. The FBI first identifies cases with a positive hair analysis association. Then, the FBI reviews the lab reports and FBI examiner trial testimony (if applicable), identifying errors, such as overstatement of the conclusions or statements exceeding the limits of science. The results of this internal review are then sent to the Innocence Project or the NACDL for independent review. As of last year, no fewer than 74 defendants whose convictions involved hair evidence were exonerated as a result of post-conviction DNA testing. Some of these convictions date back to 1989. Combined, these exonerated defendants served 1,056 years of unjust, undeserved prison time.
While large scale institutional changes are required now, as they have been for decades, there is one particular reform that defense attorneys may implement immediately. A troubling but significant revelation derived from case reviews is the underutilization of expert witnesses by criminal defendants. In a review of 137 post-conviction DNA exoneration cases, a defense expert testified in just 19 of these. This explains how shoddy, overstated, pseudo-scientific, perhaps even wholly fictional evidence can play such a critical role in convictions. If the scientific claim is not immediately debunked before the jury, it can grow into a lethal piece of evidence. Defense attorneys, armed with knowledge of the danger posed by inaccurate scientific testimony, may put on their own experts to rebut untenable scientific conclusions. By ensuring fairness at the trial level, prudent defense attorneys can secure their clients’ liberty well in advance of the next FBI scandal.
For a comprehensive review of the Microscopic Hair Comparison Review Project, including up to date information on recent exonerations, visit https://www.nacdl.org/haircomparison/.
For a detailed list of exonerated defendants organized by state, see the Innocence Project’s report Not a Strand of Evidence: DNA Exonerations Involving Unvalidated or Improper Microscopic Hair Comparison Evidence, available at https://www.scribd.com/document/333192688/not-a-strand-of-evidence-updated.
Associated Press, Probe of FBI’s DNA Lab Practices Widen, Fox News (Apr. 28, 2003), http://www.foxnews.com/story/2003/04/28/probe-fbi-dna-lab-practices-widen.html.
Clive A. Stafford Smith and Patrick D. Goodman, Forensic Hair Comparison Analysis: Nineteenth Century Science or Twentieth Century Snake Oil?, 27 Colum. Hum. Rts. L. Rev. 227 (1996).
Douglas W. Deedrick and Sandra L. Koch, Microscopy of Hair Part 1: A Practical Guide and Manual for Human Hairs, Forensic Science Communications, Jan. 2004, https://archives.fbi.gov/archives/about-us/lab/forensic-science-communications/fsc/jan2004/research/2004_01_research01b.htm.
Douglas W. Deedrick, Part I: Hair Evidence, Forensic Science Communications, July 2000, https://archives.fbi.gov/archives/about-us/lab/forensic-science-communications/fsc/july2000/deedric1.htm.
Hon. Donald E. Shelton, Twenty-First Century Forensic Science Challenges for Trial Judges in Criminal Cases: Where the “Polybutadiene” Meets the “Bitumen”, 18 Widener L.J. 309 (2009).
Innocence Project, Not a Strand of Evidence: DNA Exonerations Involving Unvalidated or Improper Microscopic Hair Comparison Evidence, available at https://www.scribd.com/document/333192688/not-a-strand-of-evidence-updated.
- Norton, W. E. Anderson, and G. Divine, Flawed forensics: Statistical failings of microscopic hair analysis, Significance, 13: 26–29 (Apr. 8, 2016), http://onlinelibrary.wiley.com/doi/10.1111/j.1740-9713.2016.00897.x/full.
Microscopic Hair Comparison Review Project, NACDL, https://www.nacdl.org/haircomparison/.
Outside Experts to Review F.B.I. Crime Lab, New York Times, (Sept. 19, 1995), http://www.nytimes.com/1995/09/19/us/outside-experts-to-review-fbi-crime-lab.html
Sean Allocca, Researchers ‘Unpick’ the Flawed FBI Hair Evidence, Forensic Magazine (Apr. 27, 2016), http://www.forensicmag.com/article/2016/04/researchers-%E2%80%98unpick%E2%80%99-flawed-fbi-hair-evidence.
Spencer S. Hsu, Convicted defendants left uninformed of forensic flaws found by Justice Dept., The Washington Post, Apr. 16, 2012, https://www.washingtonpost.com/local/crime/convicted-defendants-left-uninformed-of-forensic-flaws-found-by-justice-dept/2012/04/16/gIQAWTcgMT_story.html?utm_term=.4fb51c9d0908.
Terry Friedan, FBI whistle-blower leaves, gets $1.16 million, CNN (Feb. 27, 1998) http://www.cnn.com/US/9802/27/fbi.whitehurst/.