Saturday, May 23, 2015

No Relief for Jeffrey MacDonald After FBI Declares It “Exceeded the Limits of Science” with Hair Analysis

It was not yet 3:30 a.m. on February 17, 1970, when tragedy struck Captain Jeffrey MacDonald’s family at 544 Castle Drive, Fort Bragg, North Carolina. His pregnant wife, Collete, “had both her arms broken and was stabbed repeatedly in the chest and neck with a paring knife and an ice pick” (Anthony 2013).  Five-year-old Kimberley “was beaten across the head with a club and stabbed multiple times in the neck. Two-year-old Kristen was stabbed over 30 times in the back, chest and neck ... . MacDonald himself received relatively minor injuries except for a single stab wound that punctured his lung” (Ibid.)

MacDonald, who was a surgeon with the Green Berets, spoke of an attack “by four intruders — two white men, a black man and a white woman. He said the woman held a candle and chanted ‘Acid is groovy’ and ‘Kill the pigs’. On the headboard in the marital bedroom the word ‘PIG’ was written in blood.” (Ibid.) It was eerily similar to the depraved murders of Charles Manson’s followers in Los Angeles. “At Roman Polanski's home they killed the director's pregnant wife, Sharon Tate, and with her blood smeared the word ‘PIG’ on a wall.” (Ibid.) Indeed, Army investigators found an article on the Manson murders in the living room.

After an extended preliminary hearing culminated in a report exonerating Captain MacDonald, he left the Army with an honorable discharge and moved to California. But his father-in-law’s relentless pursuit led to the case being placed before a federal grand jury in 1974. An indictment came the next year. In 1979, federal prosecutors convicted him of the three murders. Appeals and post-conviction motions ensued. The case generated a “small library of books, a TV mini-series, countless documentaries and a forest of newsprint.” (Ibid.)

The latest opinion in this “wilderness of error” (to use the title of the most recent book on the case) is from a federal district court in North Carolina. The court issued this opinion last week, in the midst of an ongoing investigation into FBI reports and testimony about hair comparisons in thousands of cases before 2000. MacDonald’s case is now one of many in which the Department of Justice has confessed error in the presentations of its FBI laboratory personnel who compared hair samples from crime scenes to those of suspects.

Thus, last year, the Department advised MacDonald’s counsel that:
We have determined that the microscopic hair comparison analysis testimony or laboratory report presented in this case included statements that exceeded the limits of science and were, therefore invalid: (1) the examiner stated or implied that the evidentiary hair could be associated with a specific individual to the exclusion of all others—this type of testimony exceeded the limits of science; (2) the examiner assigned to the positive association a statistical weight or probability or provided a likelihood that the questioned hair originated from a particular source, or an opinion as to the likelihood or rareness of the positive association that could lead the jury to believe that valid statistical weight can be assigned to a microscopic hair association—this type of testimony exceeded the limits of science. (A copy of the documents upon which our determination is based is enclosed.) We take no position regarding the materiality of the error in this case.
According to the court, the FBI and the Innocence Project (IP) identified three errors based in the lab reports or trial testimony. None of them prompted the court to change an earlier order denying him post-conviction relief.

In light of the perception of award-winning journalists that the FBI “faked an entire field of forensic science” (Lithwick 2015), that the Bureau placed “pseudoscience in the witness box” (ibid.), and that it performed “virtually worthless” analyses (Blakemore, 2015), it is worth looking carefully at the descriptions of the self-reported “invalid” science. Not having the FBI-IP report cited by the court at my disposal, I rely solely on the court’s description of it. If this description is accurate and if the report on MacDonald's case is representative, one may want to exercise some caution with respect to the surprising number of FBI reports that are said to exude "junk science" (Editorial 2015).

Hair analysis figured into the MacDonald case in an unusual way. It was not performed to associate MacDonald with the crime scene. He was lying in the house, wounded and apparently floating in and out of consciousness. Hairs in the house — especially ones on or around the bodies of the victims — were significant only because they might have come from the invading Manson-like killers. But visual and microscopic inspections of various hairs from the house did not seem to support MacDonald's extraordinary story. Instead, the features seen in the hairs were consistent with hairs sampled from the MacDonalds themselves.


A bedspread on the floor of the master bedroom of the MacDonald home contained a hair entangled with a purple cotton thread. An FBI lab technician mounted the hair on a slide marked “Q96 H (from thread).” Paul Stombaugh, who was in charge of the Chemistry Branch of the Chemistry and Physics Section of the FBI crime laboratory, examined the Q96 thread and hair, and wrote:
Light brown to blond head hairs that microscopically match the K1 head hairs of COLLETE MACDONALD were found in specimens ... Q96.... The Q96 hair was found entangled around a purple cotton sewing thread like that used in the construction of the Q12 pajama top [belonging to defendant]. Further, this hair had bloodlike deposits along its shaft.
The 2014 report found no errors in this 1974 laboratory report or in Stombaugh’s testimony at the 1979 trial that “this hair—in conducting a comparison examination with the comparison microscope—microscopically matched the head hairs of Colette MacDonald.

On cross-examination, however, defense counsel suggested that it was peculiar that the thread and the hair “were still wrapped around together after four years of having been in the laboratory custody.” He asked, “Doesn't it make a difference to you to find out what treatment or handling a hair would have had before you examined it in the laboratory?” Stombaugh replied that “The hair was not mounted sir, as were many other ones in this submission. We opened the vials up and identified what was inside. If they were hairs, we would mount it on a slide and then they were compared.” The following exchange then occurred:
Q. Mr. Stombaugh, the question was: weren't you concerned with what might have been done to that hair that might possibly lead you to a wrong conclusion unless you found out what they had done with it?
A. Sir, the only conclusion on the hair examination that I was going to make was its origin.
Q. That is pretty serious about whose hair it is. That is a fundamental question you were being asked.
A. That is correct.
This last exchange is what, in the eyes of the Inspector General and the FBI and IP reviewers, moved Stombaugh’s testimony beyond the limits of science—he said he was examining the hair to reach a “conclusion” of some sort about “its origin” and that this was a “fundamental question.” But he never presented any definitive conclusion of identity. Neither did he try to quantify the probability of identity. To be sure, he did state that the hairs had matching colors and microscopic features. But the reviewers did not deem this conclusion improper or unacceptable. Somehow the conclusion became “invalid” because Stombaugh explained that he was not overly concerned with how the hair had come to entangled with the thread. This event, he said, was not a problem for him to consider because his task was strictly limited to ascertaining whether there was a possible association between that hair and the sample of known hairs from the defendant. Considering this testimony about “the origin” in context, it hardly seems like an egregious example of “pseudoscience” or the like.


The second instance of “invalid science” reported in 2014 was a 1999 laboratory report of Robert Fram, an examiner in the FBI Lab Hairs and Fiber Unit. At this point in the post-conviction proceedings, the district court had ordered the FBI to ship the hairs to the Armed Forces DNA Identification Laboratory for mitochondrial DNA testing. Fram documented the contents of the slides and sample being packed up and sent. During this process, he examined a glass microscope slide marked “19 1/2 L2082 Q96 PMS,” which contained four hairs. He observed that:
A forcibly removed Caucasian head hair found on one of the Q96 resubmitted glass microscope slides . . . exhibits the same microscopic characteristics as hairs in the K2 specimen. Accordingly, this hair is consistent with having originated from KIMBERLY MACDONALD, the identified source of the K2 specimen.
Fram also stated in the report that “[h]air comparisons are not a basis for personal identification.

Again, condemning these observations as erroneous seems harsh. Although the phrase “consistent with” is far from ideal, no one seems to doubt that the hair truly was “consistent with” the little girl’s, and MacDonald did not contend that it originated from anyone else.


In response to MacDonald’s original 1990 Petition for Post Conviction Relief, FBI laboratory analyst Michael Malone studied one hair found near Colette MacDonald. Malone was to become notorious for giving false or dubious testimony in other cases (Earl 2014). In this phase of the MacDonald case in 1991, however, he simply wrote that:
This hair [Q79] was compared to the pubic hair sample of JEFFREY MACDONALD (specimen K22). This hair exhibits the same individual microscopic characteristics as the pubic hairs of JEFFREY MACDONALD, and accordingly is consistent with having originated from JEFFREY MACDONALD.
Like Fram, he added a qualification. But where Fram cautioned that “[h]air comparisons are not a basis for personal identification,” Malone noted that “hair comparisons do not constitute a basis for absolute personal identification.

Despite the addition of the word “absolute,” on their face, these statements do not seem to “state[] or impl[y] that the evidentiary hair could be associated with a specific individual to the exclusion of all others,” and they do not “assign[] to the positive association a statistical weight or probability or provide[] a likelihood that the questioned hair originated from a particular source.” Finding matching physical features is consistent with the proposition that the hair was MacDonald's. At the same time, “hair comparisons do not constitute a basis for absolute personal identification” -- the match does not exclude everyone else in the world. Thus, Malone's statements do not seem to be scientifically invalid (at least with respect to the two criteria in the DOJ's letter).

Rather, the legitimate concern is psychological -- without a literal statement that the observed similarities are also consistent with the possibility that the hair was not MacDonald's, the reader might give the match more weight than it logically deserves. This misconstruction of the report by a lay reader is certainly possible, and I would not want reports about hair matches to be written like Malone's and Fram's were. But this objection is different than dismissing the findings as invalid on the theory that the statements in the report logically imply that the only individual in world who could have been the source of the hair was MacDonald. In reaching the latter conclusion, the DOJ may have gone too far.

Microscopic hair comparison is only a rough indicator of identity. Many people could share the same characteristics. But this limitation does not make the field fraudulent. Many disease symptoms, for example, are overinclusive when used to make a diagnosis, but that fact does not render them invalid or worthless as diagnostic criteria.

Likewise, the consistency that Malone reported was not sufficient to establish to a near certainty that the hair was MacDonald’s rather than an intruder’s. In fact, the later mitochondrial DNA testing excluded MacDonald, his wife, and his children as the source of the hair. Consequently, Malone’s reported similarity could have been false (if Malone did not make accurate observations, or if he lied about what he observed). Or, perhaps the Q79 hair was physically similar to MacDonald’s, as Malone said, but it nevertheless originated from someone else. As MacDonald and Fram explicitly stated, physical similarity alone is probative but not definitive of identity.


Related Postings