Skip to main content

Will 'DNA Phenotyping' Lead to Racial Profiling by Police?

Bioethicists, geneticists, attorneys, and cops weigh the pros and cons of a radical new science.
(Photo: KreativKolors/Shutterstock)

(Photo: KreativKolors/Shutterstock)

An unknown serial killer wreaking havoc in Baton Rouge, Louisiana, in the summer of 2002 led local police and the FBI to employ a novel and controversial approach.

They sent a sample of DNA that had been found at the scene of one of the crimes to a be analyzed using a new test designed by a molecular biologist. Tony Frudakis promised that, with his “DNAWitness” test, he could identify the suspect’s race within a small margin of error. Following a witness description, the cops had previously been searching for a white male. But Frudakis identified the suspect as being black. Because of those results, the investigation switched course, and eventually produced the killer.

Even though the science led to an arrest, one of the prosecutors in Baton Rouge, Tony Clayton, expressed his wariness about its emphasis on race and ethnicity; he said he didn’t trust anything that implies all people don’t “bleed the same blood.” Clayton told a Wired reporter in 2007: “If I could push a button and make this technology disappear, I would.” The Wired piece concluded that, despite the technology’s accuracy, “police won’t touch it” for just that reason, and that Frudakis’ company would likely go out of business. (It did.)

“How far do we want the dragnet to go? Do we want the science police knocking on everybody’s door?”

In the years since, law enforcement seems to have grown less squeamish. An extensive New York Times article this week details all the different companies—some with peer-reviewed methods, some without—that are currently cashing in on a process now known as “forensic DNA phenotyping,” the approximate identification of a suspect’s ancestry, eye color, and hair color.

DNA has long been used to implicate or exonerate individual suspects, as the Times’ Andrew Pollack explains, but phenotyping is a more predictive method by which investigators can narrow or broaden their suspect searches, based on general categories of physical traits. Law enforcement can now publicly announce that they are looking for a suspect who “probably” has dark skin, or who “probably” has blue eyes, and so on.

Like Clayton, the Louisiana prosecutor, a Stanford anthropologist is quoted in Pollack’s article worrying that this methodology will lead to racial profiling. They are not alone in having this concern. In a chapter she wrote in the book What’s the Use of Race? Modern Governance and the Biology of Difference, bioethicist Pamela Sankar expressed concern about the simultaneous vagueness and (perceived) unassailability of DNA results. “It seems possible that instead of making suspect searches more exact, the vagueness of FDP descriptions might make them more vulnerable to stereotyping,” Sankar wrote. “Of course, the same might be said of most descriptions the police are handed. Other descriptions, however, are not based on genetics.”

By contrast, geneticists Manfred Kayser and Peter Schneider argued in Forensic Science International that the potential for discrimination and stereotyping is actually less present with a DNA phenotype than with eyewitness testimony. “In principle, we do not see a difference in the consequence for the public reaction, as the same situation could arise from an eyewitness statement indicating that a member of a particular population or minority group was seen near the crime scene,” they wrote. “In fact, an eyewitness statement ... may be less accurate and more biased than the results of a DNA test whose reliability can at least be supported by statistical evidence....”

Still, the fact that race, ethnicity, and ancestry are inherently fraught categories led Bert-Jaap Koops and Maurice Schellekens to wonder in the Columbia Science and Technology Law Review about the impact on the public after the results. Would making DNA phenotype results public lead to a kind of guilt-by-association of entire ethnic groups within communities? The authors, who study the intersection of technology and law, also wrote of “the risk of establishing a link in people’s minds between certain minorities and crime [as] greater, i.e. ‘it’s in their DNA.’” As to the latter problem, the authors proposed that law enforcement present the results of the DNA tests (as in, “the suspect is probably light-skinned”) to the public without necessarily including the source of the information (“...which we know because we tested DNA left at the crime scene”).

Koops and Schellekens also wrote about the potential for law enforcement to develop “tunnel vision” during an investigation, and focus too exclusively on a suspect’s race, if this were the only known factor. But at the end of their article, despite the pitfalls, they still concluded that DNA phenotyping is a positive development, because the dangers it presents are not much worse than those presented by previous types of identifiers or investigative methods. Yes, the news-reading public may discriminate, and law enforcement could misuse the DNA results, but this is true of any other information involved in ongoing crime investigations too.

It’s ultimately up to law enforcement officers to use any tool at hand responsibly, and not as an excuse for racial profiling. Certainly, information about a perpetrator’s race or ethnicity that was provided by witnesses or victims can also be misused. In 2013, the victim of a sexual assault in Ontario told police that her attacker had dark skin; armed with that information, the Ontario Provincial Police responded by subjecting 100 migrant workers to a DNA “testing sweep” to see if any of them were a match, according to a damning article in the Toronto Star.

Ricardo Federico, an attorney interviewed for the story, asked, “How far do we want the dragnet to go? Do we want the science police knocking on everybody’s door?”

True Crime is Lauren Kirchner's weekly column about crime and criminal justice issues.