A meticulous analysis of online activity during the 2016 campaign makes a powerful case that targeted cyberattacks by hackers and trolls were decisive.
By Jane Mayer
The Newyorker
Donald Trump has adopted many contradictory positions since taking
office, but he has been unwavering on one point: that Russia played no
role in putting him in the Oval Office. Trump dismisses the idea that
Russian interference affected the outcome of the 2016 election, calling
it a “made-up story,” “ridiculous,” and “a hoax.” He finds the subject
so threatening to his legitimacy that—according to “The Perfect Weapon,” a recent book on cyber sabotage by David Sanger, of the Times—aides
say he refuses even to discuss it. In public, Trump has characterized
all efforts to investigate the foreign attacks on American democracy
during the campaign as a “witch hunt”; in March, he insisted that “the
Russians had no impact on our votes whatsoever.”
Few people,
including Trump’s opponents, have publicly challenged the widespread
belief that no obtainable evidence can prove that Russian interference
changed any votes. Democrats, for the most part, have avoided
attributing Hillary Clinton’s defeat directly to Russian machinations.
They have more readily blamed James Comey,
the former F.B.I. director, for reversing Clinton’s thin lead in the
final days of the campaign by reopening a criminal investigation into
her mishandling of classified e-mails. Many have also expressed
frustration with Clinton’s weak performance as a candidate, and with her
campaign’s tactical errors. Instead of investigating whether Russia
tipped the electoral scales on its own, they’ve focussed on the
possibility that Trump colluded with Russia, and that this, along with
other crimes, might be exposed by the probe being conducted by the special counsel, Robert Mueller.The U.S. intelligence community, for its part, is prohibited from investigating domestic political affairs. James Clapper, the former director of National Intelligence, told me, “We try not to spy on Americans. It’s not in our charter.” He emphasized that, although he and other intelligence officials produced—and shared with Trump—a postelection report confirming an extensive cyberattack by Russia, the assessment did not attempt to gauge how this foreign meddling had affected American voters. Speaking for himself, however, he told me that “it stretches credulity to think the Russians didn’t turn the election.”
Ordinarily, Congress would aggressively examine an electoral controversy of this magnitude, but the official investigations in the House and the Senate, led by Republicans, have been too stymied by partisanship to address the ultimate question of whether Trump’s victory was legitimate. Although the Senate hearings are still under way, the Intelligence Committee chairman, Richard Burr, a Republican, has already declared, “What we cannot do, however, is calculate the impact that foreign meddling and social media had on this election.”
Even the Clinton campaign has stopped short of attributing its loss to the Russians. Joel Benenson, the campaign’s pollster, told me that “a global power is fucking with our elections,” and that “every American should be outraged, whether it changed the outcome or not.” But did the meddling alter the outcome? “How will we ever know?” he said. “We probably won’t, until some Russians involved in it are actually prosecuted—or some Republican, in a moment of conscience, talks.”
Politicians may be too timid to explore the subject, but a new book from, of all places, Oxford University Press promises to be incendiary. “Cyberwar: How Russian Hackers and Trolls Helped Elect a President—What We Don’t, Can’t, and Do Know,” by Kathleen Hall Jamieson, a professor of communications at the University of Pennsylvania, dares to ask—and even attempts to answer—whether Russian meddling had a decisive impact in 2016. Jamieson offers a forensic analysis of the available evidence and concludes that Russia very likely delivered Trump’s victory.
The book, which is coming out less than two months before the midterm elections, at a moment when polls suggest that some sixty per cent of voters disapprove of Trump, may well reignite the question of Trump’s electoral legitimacy. The President’s supporters will likely characterize the study as an act of partisan warfare. But in person Jamieson, who wears her gray hair in a pixie cut and favors silk scarves and matronly tweeds, looks more likely to suspend a troublemaker than to be one. She is seventy-one, and has spent forty years studying political speeches, ads, and debates. Since 1993, she has directed the Annenberg Public Policy Center, at Penn, and in 2003 she co-founded FactCheck.org, a nonpartisan watchdog group. She is widely respected by political experts in both parties, though her predominantly male peers have occasionally mocked her scholarly intensity, calling her the Drill Sergeant. As Steven Livingston, a professor of political communication at George Washington University, puts it, “She is the epitome of a humorless, no-nonsense social scientist driven by the numbers. She doesn’t bullshit. She calls it straight.”
Indeed, when I met recently with Jamieson, in a book-lined conference room at the Annenberg Center, in Philadelphia, and asked her point-blank if she thought that Trump would be President without the aid of Russians, she didn’t equivocate. “No,” she said, her face unsmiling. Clearly cognizant of the gravity of her statement, she clarified, “If everything else is a constant? No, I do not.”
Jamieson said that, as an academic, she hoped that the public would challenge her arguments. Yet she expressed confidence that unbiased readers would accept her conclusion that it is not just plausible that Russia changed the outcome of the 2016 election—it is “likely that it did.”
An airtight case, she acknowledges, may never be possible. In the introduction to her new book, she writes that any case for influence will likely be similar to that in a civil legal trial, “in which the verdict is rendered not with the certainty that e=mc2 but rather based on the preponderance of evidence.” But, she points out, “we do make most of life’s decisions based on less-than-rock-solid, incontrovertible evidence.” In Philadelphia, she noted to me that “we convict people on probabilities rather than absolute certainty, and we’ve executed people based on inferences from available evidence.” She argued that “the standard of proof being demanded” by people claiming it’s impossible to know whether Russia delivered the White House to Trump is “substantially higher than the standard of proof we ordinarily use in our lives.”
Her case is based on a growing body of knowledge about the electronic warfare waged by Russian trolls and hackers—whom she terms “discourse saboteurs”—and on five decades’ worth of academic studies about what kinds of persuasion can influence voters, and under what circumstances. Democracies around the world, she told me, have begun to realize that subverting an election doesn’t require tampering with voting machines. Extensive studies of past campaigns, Jamieson said, have demonstrated that “you can affect people, who then change their decision, and that alters the outcome.” She continued, “I’m not arguing that Russians pulled the voting levers. I’m arguing that they persuaded enough people to either vote a certain way or not vote at all.”
The effect of such manipulations could be momentous in an election as close as the 2016 race, in which Clinton got nearly 2.9 million more votes than Trump, and Trump won the Electoral College only because some eighty thousand votes went his way in Wisconsin, Michigan, and Pennsylvania. In two hundred and twenty-four pages of extremely dry prose, with four appendixes of charts and graphs and fifty-four pages of footnotes, Jamieson makes a strong case that, in 2016, “Russian masterminds” pulled off a technological and political coup. Moreover, she concludes, the American media “inadvertently helped them achieve their goals.”
When Jamieson set out to research the 2016 campaign—she has researched every Presidential election since 1976—she had no intention of lobbing a grenade. She was spending a peaceful sabbatical as a fellow at the Shorenstein Center, at Harvard’s Kennedy School of Government, exploring a rather narrow topic: the 2016 Presidential debates. She’d chosen this subject because, having devoted decades to examining the impact of advertising and other forms of persuasion on voters, she believed that most of the big questions in the field of political-campaign communications had been answered. Also, she admitted, “I have what you could call a debate fixation. Every year since 1996 I’ve done some kind of social-science look at the effects of debates.”
This expertise helped Jamieson notice something odd about the three debates between Trump and Clinton. As she told me, “The conventional wisdom was that Hillary Clinton had done pretty well.” According to CNN polls conducted immediately after the debates, she won all three, by a margin of thirteen per cent or greater. But, during the period of the debates, Jamieson and others at the Annenberg Center had overseen three telephone surveys, each sampling about a thousand adults. In an election that turned more than most on judgments of character, Americans who saw or heard the second and third debates, in particular, were more likely than those who hadn’t to agree that Clinton “says one thing in public and something else in private.” Jamieson found this statistic curious, because, by the time of the first debate, on September 26th, Clinton’s reputation for candor had already been tarnished by her failed attempt to hide the fact that she’d developed pneumonia, and by the revelation that, at a recent fund-raising event, she’d described some Trump supporters as “deplorables”—a slur that contradicted her slogan “Stronger Together.” Other Annenberg Center polling data indicated to Jamieson that concerns about Clinton being two-faced had been “baked in” voters’ minds since before the first debate. Clinton “had already been attacked for a very long time over that,” Jamieson recalls thinking. “Why would the debates have had an additional effect?”
After insuring that the surveys had been properly conducted, Jamieson analyzed whether this change in a voter’s perception of Clinton’s forthrightness predicted a change in his or her candidate preference. To her surprise, she found that it did: as she put it to me, there was a “small but significant drop in reported intention to vote for her.” This statistic, too, struck Jamieson as curious; she knew from years of scholarship that Presidential debates, barring major gaffes, typically “increase the likelihood that you’re casting a vote for, rather than against,” a candidate.
Last year, while Jamieson was trying to determine what could have caused viewers’ perception of Clinton’s character to fall so consequentially, the Washington Post asked her to write an op-ed addressing whether Russian operatives had helped to elect Trump. Jamieson agreed to do so, but, she admitted to me, “I frankly hadn’t thought about it one way or the other.”
amieson is scrupulously nonpartisan in her work. Beth Myers, who helped lead Mitt Romney’s Presidential campaigns in 2008 and 2012 and worked with Jamieson on a bipartisan project about Presidential debates, told me, “If Kathleen has a point of view, I don’t know what it is. She’s extraordinarily evenhanded. She is fair and fearless.” Anita Dunn, a Democratic adviser to Barack Obama, agrees. She, too, worked with Jamieson on the Presidential-debates project, and she studied with her as an undergraduate. Jamieson, she says, “is constantly pointing out what the data actually shows, as opposed to those of us who just assert stuff.”
Jamieson began her study of the 2016 election with an open mind. But, in the fall of 2017, as she watched the House and the Senate hold hearings on Russia’s social-media manipulations, and reviewed the sampling of dozens of Facebook ads released by the House Intelligence Committee—all paid for by Russians during the Presidential campaign—she developed suspicions about the reasons behind Trump’s victory. Before the hearings, Facebook’s chairman and C.E.O., Mark Zuckerberg, had maintained that the amount of Russian content that had been disseminated on social media was too small to matter. But evidence presented to the Senate committee revealed that material generated by the Kremlin had reached a hundred and twenty-six million American Facebook users, leading Senator Dianne Feinstein to call the cyberattack “cataclysmic.”
House Democrats later released not only the ads but also their “targeting data”—the demographics and the geographic locations of users receiving them—which indicated to Jamieson “whom the Russians were going for.” Among other things, she could discern that the Russians had tried “to minimize the vote of African-Americans.” Bogus Kremlin-sponsored ads that had circulated online—including one depicting a black woman in front of an “AFRICAN-AMERICANS FOR HILLARY” sign—had urged voters to tweet or text rather than vote, or to “avoid the line” and “vote from home.”
Jamieson’s Post article was grounded in years of scholarship on political persuasion. She noted that political messages are especially effective when they are sent by trusted sources, such as members of one’s own community. Russian operatives, it turned out, disguised themselves in precisely this way. As the Times first reported, on June 8, 2016, a Facebook user depicting himself as Melvin Redick, a genial family man from Harrisburg, Pennsylvania, posted a link to DCLeaks.com, and wrote that users should check out “the hidden truth about Hillary Clinton, George Soros and other leaders of the US.” The profile photograph of “Redick” showed him in a backward baseball cap, alongside his young daughter—but Pennsylvania records showed no evidence of Redick’s existence, and the photograph matched an image of an unsuspecting man in Brazil. U.S. intelligence experts later announced, “with high confidence,” that DCLeaks was the creation of the G.R.U., Russia’s military-intelligence agency.
Academic research has also shown that political messages tend not to change the minds of voters who have already chosen a candidate; they are most likely to persuade undecided voters. And in 2016 an uncommonly high percentage of voters liked neither candidate and stayed undecided longer than usual. By some counts, about thirty-seven million Americans—fifteen per cent of the electorate—were still undecided in the final weeks before the election.
Jamieson argues that the impact of the Russian cyberwar was likely enhanced by its consistency with messaging from Trump’s campaign, and by its strategic alignment with the campaign’s geographic and demographic objectives. Had the Kremlin tried to push voters in a new direction, its effort might have failed. But, Jamieson concluded, the Russian saboteurs nimbly amplified Trump’s divisive rhetoric on immigrants, minorities, and Muslims, among other signature topics, and targeted constituencies that he needed to reach. She noted that Russian trolls had created social-media posts clearly aimed at winning support for Trump from churchgoers and military families—key Republican voters who seemed likely to lack enthusiasm for a thrice-married nominee who had boasted of groping women, obtained multiple military deferments, mocked Gold Star parents and a former prisoner of war, and described the threat of venereal disease as his personal equivalent of the Vietcong. Russian trolls pretended to have the same religious convictions as targeted users, and often promoted Biblical memes, including one that showed Clinton as Satan, with budding horns, arm-wrestling with Jesus, alongside the message “ ‘Like’ if you want Jesus to win!” One Instagram post, portraying Clinton as uncaring about the 2012 tragedy in Benghazi, depicted a young American widow resting her head on a flag-draped coffin.
Another post displayed contrasting images of a thin homeless veteran and a heavyset, swarthy man wearing an “UNDOCUMENTED UNAFRAID UNAPOLOGETIC” T-shirt, and asked why “this veteran gets nothing” and “this illegal gets everything.” It concluded, “Like and share if you think this is a disgrace.” On Election Day, according to CNN exit polls, Trump, despite his political baggage, outperformed Clinton by twenty-six points among veterans; he also did better among evangelicals than both of the previous Republican nominees, Mitt Romney and John McCain.
In her Post article, Jamieson wrote that it was “hard to know” if Russian propaganda and dirty tricks—including the steady release of hacked e-mails, starting with Democratic National Committee correspondence that was leaked just before the Party’s convention—had made a decisive difference in 2016. Nevertheless, she argued, the “wide distribution” of the trolls’ disinformation “increases the likelihood” that it “changed the outcome.”
After the article’s publication, she returned to her sabbatical project on the debates, with a newly keen eye for Russian trolls and hackers. After reviewing the debate transcripts, scrutinizing press coverage, and eliminating other possibilities, Jamieson concluded that there was only one credible explanation for the diminishing impression among debate viewers that Clinton was forthright: just before the second debate, WikiLeaks had released a cache of e-mails, obtained by Russian hackers, that, it said, were taken from the Gmail account of Clinton’s campaign chairman, John Podesta. They included excerpts from speeches that Clinton had given to banks, for high fees, and had refused to release during the campaign. The speeches could be used by detractors to show that, despite her liberal rhetoric, she was aligned with Wall Street. The hacked content permeated the discourse of the debates, informing both the moderators’ questions and the candidates’ answers. All this, Jamieson writes, gave legitimacy to the idea that Clinton “said one thing in public and another in private.”
During the second debate, on October 9th, before 66.5 million viewers, one of the moderators, Martha Raddatz, relayed a question submitted by a voter: Did Clinton think that it was acceptable for a politician to be “two-faced”? The question referred to a leaked passage from one of Clinton’s previously unreleased paid speeches; Russian hackers had given the passage to WikiLeaks, which posted it two days before the debate. In the speech, Clinton had cited Steven Spielberg’s film “Lincoln” as an example of how politicians sometimes need to adopt different public and private negotiating stances. The point was scarcely novel, but the debate question—which took her words out of context, omitted her reference to the movie, and didn’t mention that Russian operatives had obtained the speech illegally—made Clinton sound like a sneaky hypocrite. When Clinton cited “Lincoln” in order to defend the statement, Trump pounced.
“She got caught in a total lie!” Trump said. “Her papers went out to all her friends at the banks—Goldman Sachs and everybody else. And she said things, WikiLeaks, that just came out. And she lied. Now she’s blaming the lie on the late, great Abraham Lincoln!”
The dynamic recurred in the third debate, on October 19th, which 71.6 million people watched. When Trump accused Clinton of favoring “open borders,” she denied it, but the moderator, Chris Wallace, challenged her by citing a snippet from a speech that she had given, in 2013, to a Brazilian bank: “My dream is a hemispheric common market with open trade and open borders.” Again, there was no mention of the fact that the speech had been stolen by a hostile foreign power; Wallace said that the quotation had come from WikiLeaks. The clear implication of Wallace’s question was that Clinton had been hiding her true beliefs, and Trump said to him, “Thank you!” His supporters in the audience laughed. Clinton said that the phrase had been taken out of context: she’d been referring not to immigrants but to an open-bordered electric grid with Latin America. She tried to draw attention to Russia’s role in hacking the speech, but Trump mocked her for accusing Putin, and joked, “That was a great pivot off the fact that she wants open borders.” He then warned the audience that, if Clinton were elected, Syrians and other immigrants would “pour into our country.”
The fact-checking organization PolitiFact later concluded that Trump had incorrectly characterized Clinton’s speech, but the damage had been done. Jamieson’s research indicated that viewers who watched the second and third debates subsequently saw Clinton as less forthright, and Trump as more forthright. Among people who did not watch the debates, Clinton’s reputation was not damaged in this way. During the weeks that the debates took place, the moderators and the media became consumed by an anti-Clinton narrative driven by Russian hackers. In “Cyberwar,” Jamieson writes, “The stolen goods lent credibility” to “those moderator queries.”
As Jamieson reviewed the record further, she concluded that the Russian hackers had also been alarmingly successful in reframing the American political narrative in the crucial period leading up to the second debate. On Friday, October 7th, two days before it took place, three major stories landed in rapid succession. At 12:40 P.M., the Obama Administration released a stunning statement, by the Department of Homeland Security and the director of National Intelligence, accusing the Russian government of interfering in the election through hacking. This seemed certain to dominate the weekend news, but at 4:03 P.M. the Washington Post published a report, by David Fahrenthold, on an “Access Hollywood” tape that captured Trump, on a hot mike, boasting about grabbing women “by the pussy.” Then, less than half an hour later, WikiLeaks released its first tranche of e-mails that Russian hackers had stolen from Podesta’s account. The tranche contained some two thousand messages, along with excerpts from the paid speeches that Clinton had tried to conceal, including those that would be mentioned in the subsequent debates. (Julian Assange, the head of WikiLeaks, has denied working with the Russian government, but he manifestly despises Clinton, and, in a leaked Twitter direct message, he said that in the 2016 election “it would be much better for GOP to win.”)
If the WikiLeaks release was a Russian-backed effort to rescue Trump’s candidacy by generating a scandal to counterbalance the “Access Hollywood” tape and the intelligence report on Russian interference, Jamieson writes, it worked splendidly. The intelligence community’s report faded from the headlines; that Sunday morning, none of its authors were invited on any major talk show. Instead, the programs breathlessly discussed the “pussy” tape and the Clinton campaign’s e-mails, which were portrayed as more or less exposing both candidates as liars. Jamieson notes, “Instead of asking how we could know that the Russians were behind the hacking, the October 9 Sunday show moderators asked what effect the disclosures would have on the candidates’ respective campaigns and what the tape and speech segments revealed about the private versus public selves of the contenders.” If not for WikiLeaks, she writes, the media discourse in those crucial days likely would have remained locked on two topics advantageous to Clinton: Russian election subversion and Trump’s treatment of women.