There's a story today in the
Washington Post headlined, "
Russian trolls on Twitter had little influence on 2016 voters." The article reports that: "Russian influence operations on Twitter in the 2016 presidential election reached relatively few users, most of whom were highly partisan Republicans, and the Russian accounts had no measurable impact in changing minds or influencing voter behavior, according to a study out this morning." However, "the study doesn’t go so far as to say that Russia had no influence on people who voted for President Donald Trump."
Here's a link to the study itself. I haven't read it yet, but based on the story in the
Post, it seems the headline is misleading.
Remember the word cloud that showed the word "email" completely dominated the public's perception of Hillary Clinton in 2016? A lot of that was driven by Russia, via Wikileaks, releasing DNC emails in August and Clinton campaign emails in October. (It didn't matter that the August emails showed nothing worse than political infighting and the October emails were entirely innocuous. Because the FBI director had publicly criticized how Clinton managed her email as Secretary of State, the repeated references to "email" by the media for the entire month leading up to the election created perception that she'd done something wrong.) Those leaks were part of Russia's disinformation campaign. To the credit of the
Post, the article notes that the study "doesn’t examine other social media, like the much-larger Facebook" and that it doesn't "address Russian hack-and-leak operations. Another major study in 2018 by University of Pennsylvania communications professor Kathleen Hall Jamieson suggested those probably played a significant role in the 2016 race’s outcome." (
Here's a link to Jamieson's 2018 analysis. And
here's a study from Columbia University last year that found that betting markets swung toward Hillary Clinton at times when Russians were likelier to be on vacation, i.e., when the Kremlin's troll farms would be less active.) The
Post article also acknowledges the possibility that given the "small margins of victory in some states for Trump ... even a small number of people who changed their attitudes as a result of Russian influence operations online [could] have swayed the vote," but goes on to say that the "sample size of the Twitter study suggests not," and quotes one of the researchers saying, "we’ll never really know."
Edited to add: the study's authors have
posted some caveats about their work.
Edited further to add: Matt Yglesias makes a decent argument
here about the relative influence of different parts of the 2016 Russian influence campaign vs. media coverage of the race, which he more briefly summarizes thusly: "The most serious Russian influence on the 2016 race was itself entirely mediated by mainstream news organization choosing of their own free will to get very heavy coverage to often trivial stories based on John Podssta’s hacked emails." He adds:
At the time, I was really furious about this. Having mellowed some, I now believe there’s an element of scope to this — nobody held a gun to Democrats’ heads and made them nominate someone who was known to have bad relationships with the press. Part of the responsibility of a political party is to select strong candidates, and this was a failure in that regard.
My problem with this description is that I think no Democratic primary voter picked Bernie Sanders or Martin O'Malley instead of Hillary Clinton on the grounds that Sanders or O'Malley were more liked by reporters!
