Russian disinformation didn’t materially affect the way people voted in the 2016 US presidential election, according to a research study published on Monday, though that doesn’t make the effect totally inconsequential.
Boffins from New York University, University of Copenhagen, Trinity College Dublin, and Technical University of Munich analyzed more than 700,000 social media posts in April and in October 2016 from Twitter accounts associated with the Internet Research Agency, a Russian influence operation.
The IRA, along with associated business entities and numerous individuals, was indicted for election interference [PDF] in 2018. But the US government in 2020 dropped charges against two of the shell companies involved, Concord Management and Concord Consulting, after efforts by defendants to turn the judicial discovery process into a doxing exercise. The case does not appear to have progressed much since then.
A 2017 report [PDF] from the Office of the US Director of National Intelligence (DNI) found that “Russian President Vladimir Putin ordered an influence campaign in 2016 aimed at the US presidential election, the consistent goals of which were to undermine public faith in the US democratic process, denigrate Secretary Clinton, and harm her electability and potential presidency.”
As described in a paper published Monday in science journal Nature, the researchers – Gregory Eady, Tom Paskhalis, Jan Zilinsky, Richard Bonneau, Jonathan Nagler, and Joshua Tucker – found that about 1 percent of Twitter users accounted for 70 percent of exposures to the provocative posts and that most of those who saw the posts strongly identified as Republicans.
That is to say this small, highly partisan group was already inclined to vote in the way advocated by the influence operation – for Donald Trump rather than Hilary Clinton – and few minds were changed.
What’s more, the researchers found that the Russian influence campaign was drowned out by domestic news and political content in the US.
The researchers report that respondents, on average, saw about four posts from Russian foreign influence accounts per day during the last month of the election campaign, compared to 106 posts per day on average from national news media and to 35 posts per day from US politicians.
“[W]e find no evidence of a meaningful relationship between exposure to the Russian foreign influence campaign and changes in attitudes, polarization, or voting behavior,” the paper says.
The findings echo previous work from 2018 in which academics from Cyprus University of Technology, University College London, and the University of Alabama, analyzed 27,000 tweets linked to the Internet Research Agency and determined the impact was minimal.
That said, the group concludes that while there’s no evidence that exposure to social media posts from Russian influence accounts changed the way people voted in 2016, foreign meddling can have second-order effects by provoking a domestic reaction.
They note that the debate about the 2016 elections and the ensuing doubts about the legitimacy of the Trump presidency has fueled mistrust in the electoral system, which they suggest may be related to voters’ willingness to believe unsubstantiated claims of voter fraud during the 2020 election.
So Russia has seen some success, even without directly altering the outcome of the election. As the DNI observed, one of Putin’s goals was to “undermine public faith in the US democratic process.”
“In a word, Russia’s foreign influence campaign on social media may have had its largest effects by convincing Americans that its campaign was successful,” the researchers state in their paper.
Despite results that demonstrate the limits of social media meddling, the researchers argue that adversaries may learn from their failings and that foreign influence operations should still be taken seriously. ®