Professor Emeritus Joseph "Jay" Kadane will present his argument about the circumstantial nature of fingerprint evidence at the 69th Annual Scientific Meeting of the American Academy of Forensic Sciences on Thursday. Photo: Courtesy of Carnegie Mellon University

Fingerprints left behind at a crime scene are one of the best possible identifiers of a suspect or perpetrator, rivaled only, in popular opinion, by DNA evidence. However, fingerprint analysis, which has been used in criminal investigations for over a century, may not be as infallible as believed by law enforcement, court officials and the media, according to statistics expert Joseph “Jay” Kadane, a Carnegie Mellon University professor emeritus and upcoming presenter at the 69th Annual Scientific Meeting of the American Academy of Forensic Sciences.

Circumstantial vs. direct

“The idea of identification—that such an expert can say ‘this is the person who left this mark’—is beyond what science can support,” Kadane told Forensic Magazine. According to Kadane, fingerprint evidence should be considered circumstantial, rather than ironclad.

“It’s not that I don’t think that it’s evidence—it is. But it’s not identification evidence,” he says. “It’s circumstantial.”

As Kadane explains, a fingerprint analyst may compare a mark left at a scene with a fingerprint collected from a suspect and point to certain similarities between the two, but just because two prints have similarities does not mean they can only have come from the same person.

“The question is—how many people share those characteristics?” he asks. “At present, with our current scientific knowledge, I don’t think we have a way of estimating that either locally, nationally (or) worldwide.”

The professor emeritus further illustrates his point by comparing fingerprint evidence to another form of circumstantial evidence—vehicle identification. 

“It’s similar to ‘I saw a white Chevy drive away from the scene’—well, there are lots of people with white Chevys,” he explains. “But if you add to that a whole bunch of other circumstantial evidence, it gets to the point where people think ‘yeah, this is probably the guy who did it.’ But that judgement is a fact-finder judgement for the jury or the judge and not a matter for a fingerprint analyst to conclude.”

Accuracy called into question

Kadane is not the first to call the validity of fingerprint analysis as direct evidence into question. In 2002, a federal judge in Philadelphia became the first to rule that fingerprint analysis did not meet the Supreme Court’s standard for admissible scientific testimony as set forth by Daubert v. Merrell Dow Pharmaceuticals in 1993, the New York Times reported.

The “Daubert Standard” calls for judges to be the gatekeepers of scientific evidence in the courtroom by weighing several different factors to determine the validity of any scientific theory, method or technique.

Pollack ruled that fingerprint analysis had not been tested enough to meet the standard, saying witnesses could speak to two prints’ similarities, and to the fact that fingerprints are unique to each person, but could not testify that “a particular latent print is in fact the print of a particular person,” according to the New York Times.

Pollack later reversed his opinion, the New York Times subsequently reported, citing, in part, the widespread acceptance of fingerprint evidence in British courts. But since then, fingerprint analysis has continued to fall under some degree of scrutiny by experts in both the legal and scientific communities.

In 2003, law professor Jennifer L. Mnookin published an article in Issues in Science and Technology journal titled “Fingerprints: Not a Gold Standard.” In the article, Mnookin agreed with Pollack’s first ruling on fingerprints, saying “there has been woefully little careful empirical examination of the key claims made by fingerprint examiners.” She also pointed to a 1995 test in which 34 percent of analysts misidentified prints.

More recently, in 2011, Mnookin and a group of 12 other experts in the fields of law and science coauthored an article in the UCLA Law Review that encouraged the development of a “research culture” in forensic science, and which critically examined several different types of forensic evidence, including fingerprint analysis.

“Even if every set of ten prints is different from every other, two specific portions of two prints from different individuals might be extraordinarily similar to one another,” the authors pointed out, among several other potential pitfalls of attempting to “match” prints.

“The claim that friction ridge patterns are highly variable might be a necessary precondition for fingerprint identification, but it does not establish fingerprint analysts’ ability to make a match,” the article states.

The article also mentions the case of Brandon Mayfield, an attorney from Oregon who was falsely tied, due to fingerprint misidentification, to the scene of the 2004 Madrid train bombings, in which over 190 were killed and about 2000 more injured, NBC News reported. The FBI issued an apology to Mayfield, who was detained for two weeks and said he felt targeted for being Muslim, after an error by their fingerprint matching computer system, and subsequent mistakes by FBI examiners, led to his accusation and arrest.

“Three experienced FBI examiners said ‘yes, his fingerprints are a match to this bag we found in Madrid,’” Kadane says. “The defense hired a fourth independent expert who corroborated that. But the fact is that Mayfield had nothing to do with it.

“Those were presumably the best we’ve got for fingerprint analysis,” Kadane adds.

Validity upheld by the courts

Despite criticism of fingerprint evidence, it is widely defended and a frequently utilized form of forensic evidence, with many experts saying fingerprint analysis is valid and can be used to definitively identify a suspect or perpetrator.

In his book Courtroom Testimony for the Fingerprint Expert, retired FBI fingerprint examiner Gary Jones rejects the claim that fingerprint analysis is “junk science” and argues that it passes on every criterion of the Daubert test.

“As of the date of publication of this book (in 2007), the science of fingerprints has prevailed over every challenge made to it by the standards set forth in Daubert,” a passage from the book states.

Additionally, an article published last year in Mitchell Hamline Law Review outlined several cases in recent history where fingerprint evidence admission was challenged by defendants in court but upheld by judges. In one example from 2011, a challenge by accused San Diego courthouse bomber Donny Love was rejected by the judge, who cited a study from May of that year in which only six false positives occurred out of 4,083 fingerprint comparisons.

In a report published by the President’s Council of Advisors on Science and Technology last September, the council called latent fingerprint analysis “a foundationally valid subjective methodology—albeit with a false positive rate that is substantial and is likely to be higher than expected by many jurors based on longstanding claims about the infallibility of fingerprint analysis.”

The report goes on to point out the possibility of confirmation and contextual bias in fingerprint analysis, and suggests latent print analysis should change “from a subjective method to an objective method.”

“It's very hard to know what in the future we may discover,” Kadane says, about any potential advancements in fingerprint identification and technology. “But in terms of right now, it seems to me that identification is simply beyond what, as a science, we can defend.”

Kadane will give his presentation, “Fingerprint Science,” at the 69th Annual Scientific Meeting of the American Academy of Forensic Sciences on Thursday, Feb. 16.