How does fact-checking work when no one can agree on the truth?


Note: This is something I originally wrote for the daily newsletter at the Columbia Journalism Review, where I’m the chief digital writer

Facebook’s still-controversial decision to exempt political advertising on the platform from the fact-checking process has focused a lot of attention not just on the state of disinformation in general (News flash: It’s bad) but on the practice of fact-checking itself. There are more people and places than ever debunking and checking the facts on every tweet or statement from Donald Trump, and yet the volume of inaccuracies and outright falsehoods never seems to diminish. How did we get here? And what are the best practices when fact-checking in a chaotic, real-time news environment like the one we’re living in now? Can we say with any certainty that fact-checking is working at all, in the sense of correcting people’s impressions of misinformation? To explore these and other related questions, CJR convened a virtual symposium of experts and practitioners on our Galley discussion platform, including NewsGuard co-founder Gordon Crovitz, Snopes founder David Mikkelson, Jonathan Albright of Columbia University’s Tow Center for Digital Journalism, Baybars Orsek of the International Fact-Checking Network, and Renee DiResta of the Stanford Internet Observatory.

Albright, who runs the Digital Forensic Research unit at the Tow Center, says that his research shows many of the basic disinformation strategies from the 2016 election — aimed at reinforcing polarization and institutional distrust — are being leveraged this time around, built on the same wedge issues as before: religion, immigration, science. But the tactics being used are evolving quickly, he says. When it comes to political advertising, Albright said fact-checking isn’t enough. “We need a [Federal Election Commission]-style portal on how citizen data is used in political campaigns, not separate platform political ad APIs.” Orsek says that Facebook’s decision not to fact-check political ads is a mistake. “I think fact-checkers should be able to flag not only political advertisements but also political claims and statements on Facebook, not necessarily with demotion enforcement, but in a way to promote fact-checkers work on areas where there is public interest for users to know more about,” he says.

Rampant disinformation may seem like a modern invention, but Kelly Weill of The Daily Beast pointed out that “the US is a country that’s always held conspiratorial thinking close to its heart. The signers of the Declaration of Independence believed a number of falsehoods about plots by King George III against America.” Conspiratorial thinking often comes with new communication methods as well, Weill notes: the modern Flat Earth movement got its start when newspapers became widely available in the UK in the mid 1840s. Brooke Binkowski, a former managing editor at Snopes.com who now works for a fact-checking site called Truth or Fiction, said fact-checkers need to adopt a more aggressive stance for these times. “You have to be prepared to stand up for the truth and defend it, in this Disinformation Age,” she says. “This isn’t ‘view from nowhere’ journalism — you have to be willing and prepared to get into peoples’ faces a bit, to tell them they’re wrong, to point your finger at them in the public square and say, Look. This is a lie, and here is the liar who is spreading it.”

Nathan Walter, a disinformation researcher at Northwestern University, co-wrote a meta-analysis of research into whether fact-checking works or not, and says his team found evidence that fact-checking does work, in the sense that it “people’s beliefs become more accurate and factually consistent” after seeing a fact-checking message. However, Walter said the research also shows that these effects become very weak, and in fact come close to disappearing entirely, when fact-checking involves political campaign statements. And attempts to add relevant context, he says, can actually make the problem worse. Alexios Mantzarlis, who ran the International Fact-Checking Network for several years before moving to the Google News Lab, said that in the time he led the IFCN, fact-checking “went from this niche of journalism that was under-covered to being perhaps over-covered. We went from not thinking about it enough to hoping it could be a silver bullet and ultimately to fighting among ourselves because it didn’t stand up to those expectations.”

DiResta says that disinformation is “a chronic condition, and we’re now in the process of figuring out the best way to manage it.” What we need to do, she says, is to develop a much deeper understanding of how and when people internalize the messages they receive, “particularly in an era in which they’re barraged with messages and attention-grabbing content every time they pick up their phone.” Fact-checker Maarten Schenk, who runs a site called Lead Stories, said that disinformation has evolved over the past year or so. “Operations are becoming larger and more complex, with widespread use of fake or stolen accounts to spread links around, often coming from dozens of websites that are all part of the same network,” he said. Paradoxically, however, this evolution also shows that the countermeasures implemented by Facebook and others are working, he says. “It means the proverbial ‘teenager from Macedonia’ has it a lot harder now. Creating a fake Facebook account is much more difficult these days, for example, with some of them getting caught within minutes for displaying non-human behavior.”

Here’s more on disinformation and the challenges of fact-checking:

Non-responsive: Snopes founder David Mikkelson talked in a Galley interview about why his site quit working with Facebook. “We impressed upon Facebook after our last agreement with them expired at the end of 2018 that we needed them to address some issues before we could renew, and they were completely non-responsive,” he says. “It did not make sense for us as an organization to continue expending resources to benefit a platform that seemed unconcerned about the welfare of their partners or about making sincere efforts to improve.”

The long game: Angie Drobnic Holan, editor of PolitiFact at Poynter, said that she tries hard not to get disillusioned about whether fact-checking is having an impact or not. “I’ve said often to friends and family lately that I’m playing the long game with fact-checking,” she says. “My goal is to keep alive factual analysis and evidence-based methods, so they’re not lost from the world. We all need help to see the world as it is, not as we’d like it to be. I fact-check because I think the truth in and of itself is a precious thing that needs to be valued and defended.”

Infectious info: Researchers at Stanford University are tracking the spread of viral disinformation using tools designed to track infectious diseases like Ebola. One of the researchers said he wasn’t concerned about large fake news events, but was more worried about “death by a thousand cuts –– that this contributes to the erosion and the undercutting of the institutions of democracy, but it does so slowly and over time, so that we don’t recognize the gravity of what’s going on until perhaps it’s too late.”

Other notable stories:

Two students were killed and three other teens were wounded in a shooting at Saugus High School in Santa Clarita on Thursday morning. The attack began just after 7:30 am, when students were supposed to be in their first classes of the day. The sheriff’s department arrived at the school to find six students with gunshot wounds, two of whom later died. Police initially thought the shooter had fled the scene, but later determined that one of the injured students was the perpetrator, who had shot himself in the head. According to several reports, it was the shooter’s 16th birthday.

Wikipedia co-founder Jimmy Wales has quietly launched what he hopes could someday become a rival to Facebook and Twitter. WT:Social allows users to share links to articles and discuss them in a Facebook-style news feed, and will rely on donations from a small subset of users to allow the network to operate without the advertising that Wales blames for encouraging the wrong kind of engagement on social media. “The business model of social media companies, of pure advertising, is problematic,” he told the Financial Times. “It turns out the huge winner is low-quality content.”

The Daily Beast set up a test of the advertising rules at the major social platforms by submitting anti-vaccination ads to Google, Facebook, Instagram, YouTube, Twitter, and Snapchat. According to the site, most of the services rejected ads with fake health claims, but Google and Twitter both approved ads that repeated widely debunked claims about the risks of vaccination, and linked to conspiracy theory websites. Google not only approved the two advertisements with blatant language—“Don’t get vaccinated” and “Vaccines aren’t safe”—but even sent multiple prompts via email to tell the publication how to better optimize them.

A single US website that specializes in anti-vaccination misinformation accounts for almost a third of all the anti-vaxx propaganda found on social media in Brazil, where 13 percent of people don’t vaccinate themselves or their children, according to a new study from the Brazilian Society of Immunizations and Avaaz, a non-profit human rights activist network. The site had its account removed by Facebook, and was blocked by both YouTube and Twitter, but the misinformation it published has continued to be distributed in Brazil and a number of other countries.

Emily Tamkin, CJR’s public editor for CNN, writes about the network’s use of the “chyron” or superscript/caption portion of the TV screen to convey fact-checking information about its news programming. “The chyrons are clever. They’re cute. They’re wry. I am amused by the chyrons,” Tamkin writes. “But the chyron undermining Bannon’s claim that a civil rights hero would support a profoundly divisive president does not change the fact that CNN is still covering Bannon’s words.”

Isaac Bailey wrote an open letter to the editor of the Northwestern University student paper, which recently apologized for the way it reported on a student protest. Bailey, the first black primary columnist for a daily newspaper in Myrtle Beach, South Carolina, told editor Troy Closson that “your instinct to empathize with marginalized groups who have for too long been overlooked or demonized or misconstrued throughout the history of the American media is a strong one. Don’t lose that.”

Reuter’s news article that called Wednesday’s impeachment hearing “dull” sparked a social media revolt, former Intercept writer Dan Froomkin notes at Presswatchers. The Reuters report drew criticism from New York University journalism professor Jay Rosen, as well as Atlantic writer James Fallows — who said that “public-affairs writing suffers when it’s similar to theater reviews” — Sara Kendzior, host of the Gaslit Nation podcast, and New York Times Magazine writer Nikole Hannah-Jones, who said that the report “is emblematic of a deep need for a reset in political reporting.”

Apple signed on 200,000 subscribers to Apple News+ in its first 48 hours after launching in March, but has been stuck in neutral since that time, according to a report from CNBC. Apple News+ includes magazines such as People and Vanity Fair, newspapers such as the Los Angeles Times and the Wall Street Journal, and online publications like Vox, New York Magazine and theSkimm. Bloomberg News reported Thursday that Apple is considering bundling Apple News+ with Apple Music and Apple TV+ as soon as next year.

Media analyst Ken Doctor writes that the combined company that results from the merger of Gatehouse Media and Gannett could be looking for as much as $400 million in cost savings to justify the deal. “What does that mean? Almost certainly, even more reduction in headcount than had been anticipated. How much? In any room of eight people at a current GateHouse or Gannett operation, one is likely to see her job gone in 2020,” Doctor says in a report for the Nieman Journalism Lab.

The News Media Alliance says that a study it did of Spanish media sites following the departure of Google News from that country in 2014 shows that “the closure of Google News in Spain [was] not detrimental to the Spanish news publishing industry as a whole.” The organization says the Spanish news publishers included in their analysis “were minimally affected and that the reduction in traffic following the closing of Google News was, if anything, low and temporary.” Google has said it will restrict the amount of information it provides for links to French publishers because of the introduction of a new copyright law similar to the one Spain enacted in 2014.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s