Speaking in early December at a ceremony to honor Harry Reid’s retirement from the US Senate, Hillary Clinton took aim at a target that would have been totally unfamiliar to audiences as recently as the summer of 2016: fake news.
She spoke of “an epidemic” of the stuff that has “flooded social media” over the past year and “can have real-world consequences.”
This was reported largely as commentary on the Pizzagate conspiracy theory, which had recently led to an alarming armed standoff at DC’s Comet Ping Pong restaurant. But it was also pretty clearly an allusion to her own recently failed presidential campaign, especially because she spoke favorably of the idea of bipartisan legislation to curb foreign propaganda news, arguing that “it is imperative that leaders in both the private and public sector step up to protect our democracy and innocent lives.”
While it’s true that fake news appears to have circulated widely in Trump-friendly corners of the internet — possibly with some assistance from the Russian government — the idea that fake news was central to the outcome of the campaign has little basis in fact. The very nature of viral fake news is that it’s mostly likely to be shared by people who have already bought into a partisan or ideological worldview, with pro-Trump fake news largely shared by Trump supporters to other Trump supporters.
Clinton’s campaign did have a real news problem, but the problem was with the real news coverage — coverage that dwelled overwhelmingly on a bullshit email server scandal, devoted far fewer resources to investigating Trump’s shady foundation than Clinton’s lifesaving one, largely ignored Trump’s financial conflicts of interest, and almost entirely avoided discussion of the policy stakes in the campaign.
Trump ended the campaign as he began it — unpopular and viewed as unqualified by a majority of voters, with no amount of fake news stories to puff him up succeeding in moving the needle. But Clinton, who began the 2016 cycle with reasonably high favorable numbers, saw them crater under a torrent of email stories with 45 percent of voters telling exit pollsters they were bothered “a lot” by her decision to forgo a state.gov email address, of which 86 percent voted for Trump.
Whether journalists want to be proud or ashamed of the work done by mainstream press during the campaign is up to them, but it was perfectly normal stories in normal outlets that moved the needle in a major way — fake news was a total sideshow.
What is fake news?
The basic idea of fake news — take made-up stories and present them as if they are factual news accounts — is not actually particularly new, but the traditional media industry paradigm made it somewhat limited in scope. The Weekly World News had its niche as a supermarket tabloid, but fundamentally the problem with constantly publishing inaccurate stories is that you would develop a reputation as inaccurate that would kill your business model.
Made-up stories are a bad way to sell subscriptions, and it’s a terrible way to sell premium ads to brand-conscious sponsors.
The internet changes this in three big ways.
One is that the web features a lot of algorithmically targeted advertising served up by third-party companies and not sold directly by editorial brands to sponsors. These kinds of ad services make inferences about who you are based on your web browsing history, and then serve ads that are targeted to you. The sites you see the ads on are, from the standpoint of the algorithm, just irrelevant cutouts whose only job is to attract clicks.
Another is that Facebook drastically reduces the power of brand reputation over the dissemination of information. Highly engaging content is highly engaging content, and it is routed specifically to people who Facebook’s algorithms believe are likely to find it compelling.
Those two factors both operate to reduce the traditional downside to publishing inaccurate stories. The third crucial factor is that increasingly intense competition between media outlets increases the incentive to make things up. On any given day, you tend to have a bunch of different outlets all covering any major story in competition with each other. Creating a genuinely unique scoop is difficult and costly, and its key elements are going to be aggregated by rival outlets. But if you make up a story about a popular figure like Pope Francis or Tom Hanks endorsing Donald Trump, then you get a cheap and easy scoop.
Back in the early days of Vox, we accidentally took advantage of the power of getting a story wrong with a piece about how the evening of December 21, 2014, would be the longest night in the history of the Earth. This was an interesting story, but it wasn’t true — and precisely because it wasn’t true, we had the scoop all to ourselves.
Our story was a good-faith error. Everyone involved felt embarrassed about it because we are professionals who take our work seriously. What’s more, Vox Media’s business model does not support clicks over reputation as an editorial strategy. But the power of an inaccurate story in pure traffic terms underscores the extent to which publishing deliberately false stories can be a viable business in the modern world.
Donald Trump was very good for the fake news business
Back on November 3, Craig Silverman and Lawrence Alexander published a blockbuster story about the Macedonian teenagers who were running a range of pro-Trump fake news websites and making good money by doing it. Amusingly, the group was all concentrated in an obscure town called Veles, where local English-speaking youths worked together on pro-Trump propaganda.
The story made it clear that political convictions had nothing to do with the operation. Pro-Trump fake news was a moneymaking scheme and nothing more.
Trump was good for the fake news business in part for the exact same reason he’s been good for the real news business — he’s a fascinating person and a skilled entertainer, whose unlikely rise to the presidency has been just flat-out weird. But he was also good for the fake news business for the specific reason that the real news about him was generally bad. In the real world, Trump did not receive any unexpected endorsements. On the contrary, the real news was that he pointedly failed to secure the endorsements of living former Republican Party presidents, about a dozen sitting Republican senators, and a wide range of other party elders.
What’s more, Trump’s support among conservative intellectuals, writers, and pundits was shockingly low for a Republican Party nominee.
And yet from the moment that Trump secured the GOP nomination, it was inevitable that tens of millions of Americans would vote for him even if he lost. That essentially created an underserved market of people psychologically predisposed to click on and share Trump-friendly news stories that outlets committed to accurate reporting were not generating.
After the election, a follow-up report from Silverman concluded that the most popular fake news stories had actually been more popular than the most popular real news about the election.
“Fake news” rapidly became partisan and watered-down
The revelation that fake news stories had become very popular during the election and that successful fake news stories were disproportionately pro-Trump, along with Trump’s unexpected election win, led to an enormous surge of interest in the topic of fake news.
One consequence of this was a growing interest in the idea that Facebook could and should do something to make its platform less polluted by fake news stories. Facebook itself has indicated that it is going to take steps in this direction, and it’s at least somewhat likely that we will end up looking back on the summer of 2016 as a temporary heyday of publishing deliberately false stories as a business strategy.
At the same time, “fake news” as a concept metastasized in two different directions.
One, exemplified by Craig Timberg’s November 24 Washington Post story “Russian propaganda effort helped spread ‘fake news’ during election, experts say,” moved away from a narrative in which Donald Trump was central to the fake news business model and toward a narrative in which fake news was central to Trump’s political success. Timberg’s story was heavily based on a report from a somewhat mysterious organization called PropOrNot, which “identifies more than 200 websites as routine peddlers of Russian propaganda during the election season.” Timberg reported that “PropOrNot estimates that stories planted or promoted by the disinformation campaign were viewed more than 213 million times.”
This rather striking conclusion turns out to have been generated by describing a fairly wide array of left-wing websites — including Naked Capitalism, Truthout, Truthdig, Consortium News, and CounterPunch — as belonging on a list alongside Russian state-owned media and fake news sites.
That led to the other direction, in which the term “fake news” has (like its predecessor “clickbait”) rapidly been watered down into a general term of disapprobation. Many jokingly referred to Timberg’s own story as an example of “fake news.” And now on Twitter, the hashtag #FakeNews is used indiscriminately, primarily by Trump fans, to refer to any news or opinion piece that they dislike.
The notion of an actual fake news story in which a person deliberately makes up a false story calculated for maximal virality has rapidly faded into the background.
Clinton’s problem was real news
While the Trump campaign was great for the fake news industry, the implication that fake news was critical to Trump’s electoral victory seems unsupported. The election was close enough that everything mattered, so it’s certainly possible that white Catholics who were duped into believing that Pope Francis had endorsed Trump swung three crucial swing states.
But it also seems clear that the impact of fake news, whatever it may have been, was minor compared with the impact — for good or for ill — of the traditional news media.
A team of researchers working for Gallup found that what Americans heard about Clinton during the campaign was overwhelmingly information related to emails. By contrast, they found, “Americans' reports of what they have read, seen or heard about Donald Trump over this same period have been more varied and related to his campaign activities and statements.”
The stories about Clinton’s email server, the separate spate of stories about Clinton Foundation emails revealed through Freedom of Information Act requests, and the third spate of stories about emails stolen from John Podesta’s email account were not fake news.
They were very real stories that totally normal mainstream media organizations chose to make the focal point of their coverage of the 2016 campaign. This coverage, though extremely extensive, did an extraordinarily poor job of explaining the actual legal issue at stake in the server matter. Network television newscasts from ABC, NBC, and CBS chose to devote three times as much airtime to Clinton’s email server as they gave to all policy issues combined. The Associated Press ran a major investigative story into Clinton Foundation influence peddling that treated a meeting with a Nobel Peace Prize winner as evidence of an insidious pay-to-play scheme. The New York Times did a Clinton Foundation investigation that treated Bill Clinton successfully rescuing American hostages from North Korea as scandalous. The fact that public health experts believe the Clinton Foundation saved millions of lives, by contrast, played extremely little role in 2016 campaign coverage.
Fake news is a comforting villain
For Clinton and her closest allies and supporters, “fake news” makes a convenient boogeyman because it paints her essentially as a victim of circumstances beyond her control.
Whatever one makes of the media coverage of her email server and of the continued operations of the Clinton Foundation, those are both things that she really did do and that if she could go back in time she almost certainly would have handled differently.
By the same token, for journalists working at mainstream establishments, the focus on “fake news” is a convenient way to connect the dots between an electoral outcome most journalists deplore and larger trends in technology and media economics that most journalists also deplore.
But it was real news from establishment outlets that made the difference in this campaign. It was CNN that decided to sideline its usual stable of conservative pundits in favor of a Trump-friendly roster of Jeffrey Lord, Kayleigh McEnany, and Corey Lewandowski. It was network television news that decided Clinton’s private email server was a more important thing to cover than the policy stakes in the 2016 election. It was the New York Times that decided to dedicate 100 percent of its above-the-fold space to articles about James Comey’s content-free letter updating Congress on the discovery of new evidence in the email case that turned out to amount to nothing.
The sum total of this media coverage — real stories based on editorial decisions about how to weight and present real facts — was to give the public the impression that two similarly ethically flawed candidates were running against each other in an election with low policy stakes. The reporters and editors responsible for that coverage can reasonably (if a bit absurdly) consider themselves proud of the work that led the public to that conclusion, or they can consider themselves ashamed of it. But the idea that voters were moved by fake stories about the pope rather than all-too-real ones about email servers is a preposterous evasion.