It was a strange moment of triumph against racism: The gun-slinging white supremacist Craig Cobb, dressed up for daytime TV in a dark suit and red tie, hearing that his DNA testing revealed his ancestry to be only “86 percent European, and … 14 percent Sub-Saharan African.” The studio audience whooped and laughed and cheered. And Cobb—who was, in 2013, charged with terrorizing people while trying to create an all-white enclave in North Dakota—reacted like a sore loser in the schoolyard.
“Wait a minute, wait a minute, hold on, just wait a minute,” he said, trying to put on an all-knowing smile. “This is called statistical noise.”
Then, according to the Southern Poverty Law Center, he took to the white nationalist website Stormfront to dispute those results. That’s not uncommon: With the rise of spit-in-a-cup genetic testing, there’s a trend of white nationalists using these services to prove their racial identity, and then using online forums to discuss the results.
But like Cobb, many are disappointed to find out that their ancestry is not as “white” as they’d hoped. In a new study, sociologists Aaron Panofsky and Joan Donovan examined years’ worth of posts on Stormfront to see how members dealt with the news.
It’s striking, they say, that white nationalists would post these results online at all. After all, as Panofsky put it, “they will basically say if you want to be a member of Stormfront you have to be 100 percent white European, not Jewish.”
But instead of rejecting members who get contrary results, Donovan said, the conversations are “overwhelmingly” focused on helping the person to rethink the validity of the genetic test. And some of those critiques—while emerging from deep-seated racism—are close to scientists’ own qualms about commercial genetic ancestry testing.
Panofsky and Donovan presented their findings at a sociology conference in Montreal on Monday. The timing of the talk—some 48 hours after the violent white nationalist rally in Charlottesville, Va.—was coincidental. But the analysis provides a useful, if frightening, window into how these extremist groups think about their genes.

RECKONING WITH RESULTS

Stormfront was launched in the mid-1990s by Don Black, a former grand wizard of the Ku Klux Klan. His skills in computer programming were directly related to his criminal activities: He learned them while in prison for trying to invade the Caribbean island nation of Dominica in 1981, and then worked as a web developer after he got out. That means this website dates back to the early years of the internet, forming a kind of deep archive of online hate.
To find relevant comments in the 12 million posts written by over 300,000 members, the authors enlisted a team at the University of California, Los Angeles, to search for terms like “DNA test,” “haplotype,” “23andMe,” and “National Geographic.” Then the researchers combed through the posts they found, not to mention many others as background. Donovan, who has moved from UCLA to the Data & Society Research Institute, estimated that she spent some four hours a day reading Stormfront in 2016. The team winnowed their results down to 70 discussion threads in which 153 users posted their genetic ancestry test results, with over 3,000 individual posts.
About a third of the people posting their results were pleased with what they found. “Pretty damn pure blood,” said a user with the username Sloth. But the majority didn’t find themselves in that situation. Instead, the community often helped them reject the test, or argue with its results.
Some rejected the tests entirely, saying that an individual’s knowledge about his or her own genealogy is better than whatever a genetic test can reveal. “They will talk about the mirror test,” said Panofsky, who is a sociologist of science at UCLA’s Institute for Society and Genetics. “They will say things like, ‘If you see a Jew in the mirror looking back at you, that’s a problem; if you don’t, you’re fine.'” Others, he said, responded to unwanted genetic results by saying that those kinds of tests don’t matter if you are truly committed to being a white nationalist. Yet others tried to discredit the genetic tests as a Jewish conspiracy “that is trying to confuse true white Americans about their ancestry,” Panofsky said.
But some took a more scientific angle in their critiques, calling into doubt the method by which these companies determine ancestry—specifically how companies pick those people whose genetic material will be considered the reference for a particular geographical group.
And that criticism, though motivated by very different ideas, is one that some researchers have made as well, even as other scientists have used similar data to better understand how populations move and change.
“There is a mainstream critical literature on genetic ancestry tests—geneticists and anthropologists and sociologists who have said precisely those things: that these tests give an illusion of certainty, but once you know how the sausage is made, you should be much more cautious about these results,” said Panofsky.

A COMMUNITY’S GENETIC RULES

Companies like Ancestry.com and 23andMe are meticulous in how they analyze your genetic material. As points of comparison, they use both preexisting datasets as well as some reference populations that they have recruited themselves. The protocol includes genetic material from thousands of individuals, and looks at thousands of genetic variations.
“When a 23andMe research participant tells us that they have four grandparents all born in the same country—and the country isn’t a colonial nation like the U.S., Canada, or Australia—that person becomes a candidate for inclusion in the reference data,” explained Jhulianna Cintron, a product specialist at 23andMe. Then, she went on, the company excludes close relatives, as that could distort the data, and removes outliers whose genetic data don’t seem to match with what they wrote on their survey.
But specialists both inside and outside these companies recognize that the geopolitical boundaries we use now are pretty new, and so consumers may be using imprecise categories when thinking about their own genetic ancestry within the sweeping history of human migration. And users’ ancestry results can change depending on the dataset to which their genetic material is being compared—a fact which some Stormfront users said they took advantage of, uploading their data to various sites to get a more “white” result.
J. Scott Roberts, an associate professor at the University of Michigan, who has studied consumer use of genetic tests and was not involved with the study, said the companies tend to be reliable at identifying genetic variants. Interpreting them in terms of health risk or ancestry, though, is another story. “The science is often murky in those areas and gives ambiguous information,” he said. “They try to give specific percentages from this region, or x percent disease risk, and my sense is that that is an artificially precise estimate.”
For the study authors, what was most interesting was to watch this online community negotiating its own boundaries, rethinking who counts as “white.” That involved plenty of contradictions. They saw people excluded for their genetic test results, often in very nasty (and unquotable) ways, but that tended to happen for newer members of the anonymous online community, Panofsky said, and not so much for longtime, trusted members. Others were told that they could remain part of white nationalist groups, in spite of the ancestry they revealed, as long as they didn’t “mate,” or only had children with certain ethnic groups. Still others used these test results to put forth a twisted notion of diversity, one “that allows them to say, ‘No, we’re really diverse and we don’t need non-white people to have a diverse society,'” said Panofsky.
That’s a far cry from the message of reconciliation that genetic ancestry testing companies hope to promote.
“Sweetheart, you have a little black in you,” the talk show host Trisha Goddard told Craig Cobb on that day in 2013. But that didn’t stop him from redoing the test with a different company, trying to alter or parse the data until it matched his racist worldview.