Yesterday saw the release of reports documenting attempts by the Russia-based, Putin-backed Internet Research Agency (IRA) to influence the 2016 presidential election via social media messaging. Both reports were written for the U.S. Senate, one by Oxford University's Computational Propaganda Project and analytics firm Graphika (online here) and the other by New Knowledge (online here), and both were written up by major papers such as The New York Times and Washington Post. They are being used as proof positive that American social media is dominated by foreign trolls who are duping American voters in all sorts of sinister new ways that call for sweeping new regulations of virtually all aspects of online life.
But while "The IRA, Social Media and Political Polarization in the United States, 2012-2018" and "The Tactics and Tropes of the Internet Research Agency" provide exhaustive catalogs of what Russian bad actors were up to on Facebook, YouTube, Instagram, and other platforms, they utterly fail to show that any of those efforts had much, if any, impact on Hillary Clinton's unexpected and self-inflicted loss to Donald Trump. That omission—and a corresponding lack of interest in putting such efforts in a historical or contemporary context—means the studies provide little to no actual insight into electoral politics on the ground or online. We're at a moment right now where Congress and the tech sector are both calling for all sorts of regulation in ways that will transparently benefit current market leaders while decreasing online freedom. The last thing we should do is let reports like these drive us more quickly in such a mistaken direction. In a world not simply of seamless "deep fakes" and undetectable, authentic-seeming AI-generated images, the only way forward is through what might be called mass, decentralized media literacy. To paraphrase The Whole Earth Catalog's credo, "We are as gods as readers and we might as well get good at it."
Both reports note that the IRA, like traditional advertisers or political operatives, segmented potential voters into various different groups and adopted strategies mostly designed to reinforce negative perceptions toward Hillary Clinton. New Directions finds that the IRA spent a fair amount of energy targeting African Americans. Out of 81 Facebook pages the report discusses, 30 targeted black audiences and amassed 1.2 million followers. By comparison, 25 pages focused on conservatives, drawing about 1.4 million followers.
From the Times' account:
The report says that while "other distinct ethnic and religious groups were the focus of one or two Facebook Pages or Instagram accounts, the black community was targeted extensively with dozens." In some cases, Facebook ads were targeted at users who had shown interest in particular topics, including black history, the Black Panther Party and Malcolm X. The most popular of the Russian Instagram accounts was @blackstagram, with 303,663 followers.
The Internet Research Agency also created a dozen websites disguised as African-American in origin, with names like blackmattersus.com, blacktivist.info, blacktolive.org and blacksoul.us. On YouTube, the largest share of Russian material covered the Black Lives Matter movement and police brutality, with channels called "Don't Shoot" and "BlackToLive."…
While the right-wing pages promoted Mr. Trump's candidacy, the left-wing pages scorned Mrs. Clinton while promoting Senator Bernie Sanders of Vermont and Jill Stein, the Green Party candidate. The voter suppression effort was focused particularly on Sanders supporters and African-Americans, urging them to shun Mrs. Clinton in the general election and either vote for Ms. Stein or stay home.
The authors of "The IRA, Social Media and Political Polarization in the United States, 2012-2018" conclude in part,
Social media have gone from being the natural infrastructure for sharing collective grievances and coordinating civic engagement to being a computational tool for social control, manipulated by canny political consultants and available to politicians in democracies and dictatorships alike.
Well, yeah, no. Terms such as social control can't be thrown around so easily, especially when both Republicans and Democrats are getting all hot about regulating social media and Facebook and Twitter, facing declines in growth, are ready to play ball (Tim Cook, doubtless concerned about Apple's flattening market capitalization, is also musing on the "inevitable" regulation of the tech sector more generally).
In 2016, black voter turnout, key to Democratic victory, was about 60 percent, or the same level it was for John Kerry in 2004. Was that due to Russian bots telling African Americans to stay home or Hillary Clinton's patchy get-out-the-vote effort? I'm guessing it's the latter. There's no question that the leak of Clinton and Democratic National Committee (DNC) emails and former FBI Director James Comey's actions during the closing weeks of the election hurt her campaign (and that the Russians played a hand in the release of the emails). But the leaks confirmed negative perceptions about Clinton, her penchant for secrecy, and the way the DNC was rigging the game against Sen. Bernie Sanders (I–Vt.). And still, with all that in play, Trump won due to fewer than 80,000 votes cast in three states (Michigan, Pennsylvania, Wisconsin), all of which had gone for the Democratic candidate in 2000, 2004, 2008, and 2012. You can argue that without the Wikileaks dump, which the U.S. intelligence community says was facilitated by the Russians, Clinton would have won. But she loaded the gun herself—and in any case, that has nothing to do with social media efforts by the IRA.
Perspective here is key. When it comes to foreign influence, the CIA says that Moscow has been trying to influence presidential outcomes via covert propaganda since at least since 1964, when Nikita Khrushchev threw his weight behind Lyndon Johnson over Barry Goldwater due to the latter's higher level of bellicosity toward the Soviet Union and communism in general. The amount of impressions, likes, retweets, shares, and rubles that get thrown around in the reports sound fantastic until you zoom out to the bigger picture. As TechCrunch reported a year ago, for instance, Clinton and Trump spent a combined $81 million on Facebook ads while the IRA ponied up $46,000, or 0.05 percent as much. Nate Silver writes:
What fraction of overall social media impressions on the 2016 election were generated by Russian troll farms? 0.1%? I'm not sure what the answer is, but suspect it's low, and it says something that none of the reports that hype up the importance of them address that question.
— Nate Silver (@NateSilver538) December 18, 2018
Unfortunately for those of us who prefer our internet less fettered, various panics are crossing streams, all of which are pushing to freeze the status quo. In politics, the old coalitions that worked for Republicans and Democrats are falling apart, so much so that both parties are at or near recent lows in self-identification. We're in the thick of a modern "era of no decision," in which neither party can maintain unified control of the government for very long. Both Democrats and Republicans have spent much of the past year attacking social media platforms for mostly imaginary crimes and threatening regulation or worse. Only a week ago, representatives of both parties invoked antitrust remedies in a discussion of Google. That's a predictable response from politicians who feel their grip starting to slip. At the same time, a trans-ideological intellectual consensus also seems to be building that social media is harmful to young people, old people, lonely people, you name it. And then there's the hysteria, particularly strong among liberals, Democrats, and #NeverTrump Republicans, that Russia is somehow an ascendant power.
The combined result of all this, plus trepidation about the economy and the general direction of the country, is almost certain to be a spasm of regulatory gestures toward the internet, to cleanse it not just of hate speech but political trolls, fake news, and bad actors generally. Sadly, what we really need to be doing right now is formulating new forms of media literacy that fit the reality of a "post-fact" world, one in which everything we see may well be a simulation.