Topic: Media Research Center
Robert Epstein is the Media Research Center's favorite Google-hating researcher. His claim to be a Hillary Clinton supporter gives him cover to be a conserative darling pushing claims that Google is biased against conservatives -- which conveniently fits in with the MRC's anti-media agenda.
A September 2018 post by Corinne Weaver touting the anti-Google film "Creepy Line" noted how Epstein supplied the film's title: “Google crosses the creepy line every day. ... Google can shift the voting preferences of undecided voters by up to 80 percent in some demographic groups just by altering the order in which search results are shown."
A couple days later, Weaver cited how Epstein "told the Media Research Center that Google was building personal profiles of users mainly through Gmail" and that "conservatives have 'special reasons to be concerned' based on the leaked emails and videos from Google concerning the 2016 election." Weaver added that "Epstein clarified that he was not a conservative."
A few days after that, Weaver got another "exclusive statement" from Epstein dismissing a Stanford University study highlighting the large number of conservative websites that push fake news, echoing Weaver's suggestion that Stanford has a pro-Google bias because Google's founders were Stanford graduates and have donated money to the school, though neither offered any specific evidence that this particular study was influenced by Google money. Weaver even pushed an article Epstein did for the pro-Trump Epoch Times that purported to explain "10 ways that Google and Facebook could affect the midterm elections.
In a November post, Weaver intoned: "The scariest thing about Google Search is the algorithms and changes that users don’t see. Dr. Robert Epstein of AIRBT (sic: the American Instistute for Behavioral Research and Technology, of which Epstein is the apparent sole staff member) has spent years studying Google and the algorithms behind it." In a July 11 post, Alexander Hall gushed that the "prolific American psychologist, journalist, and author" Epstein "will soon unleash a report that he claims will bring down Google." And an Aug. 20 item reprinted MRC chief Brent Bozell's letter to Google, signed by other "conservative leaders," citing Epstein's claims of "pro-liberal bias" in a demand that Google "explain reports and allegations that the search engine is attempting to block conservative sites and exclude voices that don’t fit the liberal narrative."
But the MRC has a conservative narrative to push, and Epstein is a big part of that. So when Hillary Clinton and other media outlets pointed out the flaws in Epstein's research, the MRC rushed to his defense.
In an Aug. 21 post, Weaver claimed that Epstein was "smeared" by Clinton and CNN over his finding that "Google had helped influence 2.6 million votes in favor of Hillary Clinton in 2016," gushing that Epstein "responded with his first ever 'twitter storm,' in essence factually disarming Clinton’s comment."
Well, not so much. Clinton had claimed that Epstein's research "was based on 21 undecided voters," which is effectively true. Epstein's paper -- which examined whether Google search results could influence the voting preference of undecided voters in the 2016 presidential election -- says that his research was based on the work of 95 people, 21 of whom were undecided, meaning that his conclusions were, in fact, based on how those 21 undecided voters reacted. In that tweetstorm, Epstein objected by citing the "3,207 election-related searches & the 98,044 web pages linked to those searches," which he claimed showed "substantial pro-Hillary bias."
But as the Washington Post pointed out, the paper doesn't explain how it determined whether a given website exhibited "pro-Hillary bias" -- which makes us wonder if it's using methodology from groups like AllSides that push the right-wing narrative that mainstream media outlets, which typically populate news searches, are reflexively "liberal" -- nor did it describe how those "election-related searches" were conducted. Epstein also apparently threw out results that were unbiased based on a conspiracy theory that "perhaps Google identified our confidants through its gmail system and targeted them to receive unbiased results."
Weaver also bashed a CNN article debunking Epstein's study, complaining that it "cited the opinions of two academics who disagreed with Epstein," then huffed: "The issue is that in the world of academic studies, one can always find two academics to either agree or disagree with you. That’s not a litmus test for accuracy or truth."
Of course, Weaver did not mention the nature of those academics' objections. One noted that even if such search bias did exist, "Epstein has failed to establish that any such biases have had anywhere near the magnitude of impact on American presidential voting that Epstein suggests," while the second pointed out that the study "did not take into account how much a voter might care about a particular subject" or "how people's voting preferences might have been affected by other technological platforms, such as Facebook, which he said was 'quite clearly gamed by third parties' in 2016."
Weaver further repeated Epstein's defense of himself:
Epstein also defended the integrity of his work, which Clinton dismissively referred to as a “debunked study.” He asked, “If my work has been "debunked," why was it included in a volume just published by #Oxford U.? Why have I been invited to speak about it at prestigious venues worldwide: #Stanford U., #Yale Law School (where both you & Bill went), even our #Senate.”
Weaver would never admit that getting published in a prestigious journal is no guarantee of credibility, as The Lancet learned when it published since-discredited research on vaccines and autism. But Epstein has his own issues in that department as well. Slate reported that Epstein supplied seven pages of citations to support his congressional testimony in July in which he rehashed the above study, "but all of them are papers or op-eds he wrote or co-wrote himself" and only one of those products was peer-reviewed, though "even that study didn't demonstrate that this has actually happened."
Yes, there are obvious problems with Epstein's research. But because his work advances the MRC's narrative, it won't tell you about them.