The Chicago Guide to Fact-Checking

The University of Chicago Press

Contents

Why We Fact-Check

When writers present a piece as nonfiction, they create a contract with the reader. This is true whether the piece in question is a newspaper article, a magazine feature, or the script for a documentary. The writer is saying this happened. To bolster their account, they present evidence including, though certainly not limited to, quotes from experts, data, and eyewitness reports. Together, these sources give the story a foundation. The overarching argument that the writer builds on top of this foundation is important, too: it tells the reader not only did this happen, but here is the context in which you should consider what happened.

But somewhere between all the reading, interviewing, and thinking, the foundation may crack and crumble. Maybe the problem is minimal — a simple misunderstanding or copy error, like flubbing a person's official title or inadvertently transposing the digits in a number. If the crack is small and the remaining sources are solid, the story could survive. Still, it's a crack, and an observant reader starts to wonder about the rest of the structure. Take, for example, a 2012 Vogue profile of Chelsea Clinton in which Daniel Baer, who at the time was the Deputy Assistant Secretary for the Bureau of Democracy, Human Rights, and Labor at the U.S. Department of State, was identified as an interior designer. At its surface, flubbing a job title is a relatively small misunderstanding. But knowing this particular flub, do you trust the rest of the story?

Then there are the more glaring problems that shake a story's foundation: explanations oversimplified to the point that they are wrong, credulous sources, and gross misunderstandings of an event and its context. Take, for example, the 2012 Supreme Court ruling on the Affordable Care Act, when both CNN and Fox News briefly reported that a controversial and key piece — the individual mandate — had been struck when in fact it had not. Or consider a 2015 article by New York Times tech writer Nick Bilton, which suggested that wearable technologies are as bad for your health as smoking cigarettes. Science writers criticized the story, pointing out that Bilton cherry-picked a handful of studies that tried to link cell phones to cancer — ignoring a swath of research that said otherwise — and also quoted a controversial alternative medicine proponent as an expert. The mistake ultimately resulted in a response from the newspaper's public editor, Margaret Sullivan, and the online version later included a 200-word addendum.

Even worse are full-blown earthquakes where the writer commits plagiarism or publishes outright alterations or fabrications of quotes, or other lies. For examples of journalistic misconduct, look no further than Jayson Blair, who plagiarized and fabricated stories for the New York Times including a series on the Washington Sniper in 2002; Patricia Smith, who fabricated pieces of her Boston Globe column; Jonah Lehrer, who self-plagiarized several posts for the New Yorker blog and also made up several Bob Dylan quotes in his book Imagine; Stephen Glass, who not only fabricated stories but also his fact-checking notes and sources while working for the New Republic; Judith Miller, who relied on inaccurate sources in her coverage leading up to the Iraq War; or Michael Finkel, who made up a composite character, along with other fabrications, for a New York Times Magazine profile in 2002. All of these writers committed these sins in order to tell a good yarn or make a persuasive argument.

In other cases, writers may twist the ground rules with a source in order to use or attribute material that wasn't agreed upon. Usually, writers and their sources will be clear about whether the source's comments are on the record, which means their identity may be included in the story, or on background, in which it may not (for definitions and further discussion, see chapter 4). Writers should honor these and other rules of source attribution, but a look at their interview notes and recordings may reveal that they, in fact, have not.

There are also grayer areas that fall between simple errors or intentional rule breaking. A writer's own biases may sneak into the work. Writers and editors, too, while crafting a compelling and page-turning narrative, may shuffle a few facts to help with the story's flow. And when a writer has spent weeks, months, or even years on a piece, it is difficult, if not impossible, to step outside to catch these mistakes. A blind spot will probably continue to be a blind spot. Further, each newspaper, book publisher, magazine, podcast, and more has a unique worldview and the stories it shares will reflect that, which adds another layer of perspective and spin on how a story is told.

Independent fact-checkers — people who are not involved in the story's creation — temper these gray areas and catch the more obvious and easy mistakes. "Fact-checkers, we're like the janitors, the custodians. We clean up after everybody," says Beatrice Hogan, the former research chief at More magazine. Indeed, a good fact-checker goes through a story both word by word and from a big-picture view, zooming in to examine each individual fact or statement and then zooming out to see whether the story's premise is sound. The fact-checker's presence does not absolve the writer and the editor from their mistakes; the responsibility is on everyone to deliver the most accurate story possible. Still, the fact-checker will likely feel the weight of a mistake the most, particularly if it was an oversight on their part or the insertion of an error where there was none before, and not an error the team made collectively. The fact-checker is indeed like a janitor, but an especially meticulous and skeptical one.

It's also worth noting that a fact-checking department is only as good as its media outlet allows it to be. If everyone involved in a story, from the writer to the editors to the art department, respects the craft of fact-checking, this support only furthers the cause. If the staff doesn't care for the fact-checking process, or if checkers feel that they can't speak up when they see a story's foundation crumbling, the entire process is doomed.

Fact-checking is also important when it comes to the Internet, both in using online sources and consuming online media. This is particularly true in an age where hoaxes make it into national news reports unchecked. There is nothing inherently bad about online information, and it isn't necessarily unique from a historical perspective. Hoaxes, sensationalism, and other wildly inaccurate accounts were around long before we went digital. Take, for example, the Winsted Wildman of Connecticut, an old Bigfoot hoax, or the yellow journalism of the 1890s, like the time William Randolph Hearst used his newspapers to fan the flames of the Spanish-American War.

"I very much reject the idea that the Internet is full of errors any more than print," says Adrienne LaFrance, a senior editor at the Atlantic, who used to run a column at Gawker called "Antiviral," where she debunked fake online stories. "I think the major difference now versus the pre-Internet is that everyone is a publisher, but not everyone has to stake their professional reputation on accuracy."

Indeed, there are two key differences between the past and the present. The first, as LaFrance points out, is that the digital age made it possible for virtually anyone with an Internet connection to publish their writing. As such, we circulate a glut of information daily — some of it great, some mediocre, and some terrible. While this has democratized journalism and publishing, making it more accessible to people who previously weren't able to participate in it, it's also led to a virtual avalanche that is hard to sift. Second, we are able to amplify any of these stories by effectively republishing them on Twitter, Facebook, Tumblr, and a slew of other social media platforms. It's so easy to click a headline and then click again to like or Tweet or reblog it, that we're collectively bombarded with inaccurate stories — whether intentionally so or because of honest mistakes.

Internet fact failures take many forms. In one category, legacy news outlets and newer media sites alike post stories without fact-checking them, essentially publishing hoaxes as news. This phenomenon is, in part, an unfortunate result of aggregation, in which online outlets pick up stories from one another without adding any original reporting. Take this example from 2013, when bored Americans celebrating Thanksgiving were riveted by a drama unfolding on several news websites. A man named Elan Gale was live-Tweeting an apparent feud with a woman named Diane on a US Airways flight. Gale claimed he had seen Diane act rudely to a flight attendant and proceeded to send her alcoholic drinks and notes essentially telling her to lighten up. She wrote back. Gale took photos of the notes and broadcast the images to his Twitter followers, which ballooned from a reported 35,000 to 140,000 following the exchange. The story hit its climax after the plane landed, when Diane approached Gale in the airport and slapped him. Many news outlets picked up the story, including ABC, Business Insider, BuzzFeed, CBC, Fox, and the New York Daily News.

But the story wasn't true. Gale — who had previously mostly posted jokes on Twitter — made it up in order to entertain his followers during the holiday travel. In a follow-up interview with ABC, Gale reportedly said of the initial media coverage: "My thought was I can't believe anyone is taking this seriously. I thought, 'Why isn't anyone doing any fact checking?' Then I saw it was on the evening news in Sacramento and it became this totally absurd thing."

Publishing a hoax weakens readers' trust in an outlet's ability to report the news, which is a problem on its own. But publishing incorrect information in more dire circumstances has even more harmful consequences. Take the media frenzy during Hurricane Sandy in 2012, the Boston Marathon bombing in 2013, or any other catastrophic event that has unfolded in the age of social media. In these cases especially, outlets are under pressure to publish — and publish fast. And it means that a lot of that published information ends up being wrong. During Hurricane Sandy, for example, Reuters picked up a rumor circulating on Twitter that nineteen Con Edison workers were trapped in a power station, surely an upsetting report for anyone with family or friends who work for the utility. And after the Boston Marathon bombing, reporters from outlets including CBS and BuzzFeed retweeted a message that said: "Police on scanner identify the names of #BostonMarathon suspects in gunfight, Suspect 1: Mike Mulugeta. Suspect 2: Sunil Tripathi." The trouble was neither man was actually a suspect, and there is no evidence that the Boston police even mentioned the latter's name. Tripathi, a Brown University student, not only had nothing to do with the bombing — he had been missing for a month and his family had been frantically searching for him. The false accusations were painful on their own, but, worse, he was later found dead from an apparent, and unrelated, suicide.

Another case where the pressure to publish breaking news played an unfortunate role was in the aftermath of a 2011 shooting in Tucson, Arizona, where a gunman opened fire on a crowd, hitting eighteen people, including U.S. Representative Gabrielle Giffords. Six people died. Early reports claimed Giffords was one of them, from outlets including CBS, CNN, Fox, the Huffington Post, the New York Times, NPR, and Reuters. In fact, Giffords wasn't dead but had been shot in the head and was rushed into surgery. The false news and Twitter announcements made an already frightening and sobering situation all the more difficult for Representative Giffords's family and constituents, as well as the nation as a whole.

A year after the Arizona shooting, Craig Silverman — who at the time covered errors, corrections, fact-checking, and verification at the Poynter blog "Regret the Error" and has a book by the same name — recounted the debacle: "Twitter gave me a window into the captivating mixture of urgency, confusion and information that emerges when major news breaks and the story takes off." Indeed, this process used to happen in the newsroom as reporters decided how and when to report breaking news. In the Internet era, the often-messy aspects are aired in the open, which is unfortunate because this is where readers can access the information and assume it is true. As Silverman notes in the same post, local Arizona news outlets got the Giffords story right. This provides insight on how to judge news when you're Tweeting from a distance: look for publications as geographically close to the story as possible, because they have the best vantage point (for more tips, see Silverman's Verification Handbook: A Definitive Guide to Verifying Digital Content for Emergency Coverage.)

Most breaking-news outlets don't employ independent fact-checkers, partly because of time constraints; television and web reporters who are covering these sorts of stories have to rely on their own sourcing because they're on short deadline. Still, examining where stories like this can go wrong is instructive: look for the potential weak spots to see how you would have checked it, if you had the opportunity. In the Gale Thanksgiving hoax, a fact-checker might have confirmed the story not just with Gale, but also with the airline and, if possible, "Diane." In the stories about Hurricane Sandy, the Boston Marathon, and the Arizona shooting, a checker should know that a rumor circulating on Twitter must be confirmed before it is repeated or put in a story — this could have been accomplished through a phone call to Con Ed or to the Boston Police Department or to the hospital, respectively.

* * *

Beyond these philosophical and practical reasons for fact-checking, there are also the legal incentives. If a story is wrong, it damages not only the writer's reputation but also that of the publisher, particularly for controversial or investigative pieces where the stakes are high. Factual errors may open an outlet to lawsuits, which can run into damages worth millions of dollars. Fact-checkers should be aware of several areas of law — including defamation, copyright, and invasion of privacy — and be especially vigilant with relevant sources. Necessary disclaimer: What follows is not intended as legal advice. If you're unsure about whether your work opens you or your publication to lawsuits, consult an attorney.

Defamation comes in two flavors: slander, which is spoken, and libel, which is written (although, confusingly, defamation in news broadcasts and the like is usually considered libel). In either case, defamation involves making a false statement about a person or company — one that is done so with fault and damages a reputation. Usually, a plaintiff — the person bringing the suit — has to prove the information was indeed inaccurate and that it harmed them in some tangible way, such as losing a job or anything else that can be connected to money (although the injury could also be the emotional stress that results from a damaged reputation).

Defamation laws are different for public and private citizens. Public figures include celebrities, government officials, politicians, and more, all of whom lose some of their expectations for privacy when they enter the public sphere. In order to successfully sue for defamation, a public figure must prove that the information was published or spoken with actual malice. This doesn't mean the publisher or speaker harbored actual ill will against the public figure, but that they knew the information was incorrect and published it anyway. It's up to the plaintiff to prove actual malice.

In 2011 former Chicago Bulls player Scottie Pippen sued NBC Universal Media LLC and CBS Interactive Inc. because of an inaccurate story that suggested he had filed for bankruptcy. But although the reporter had documentation proving a Scottie Pippen had declared bankruptcy, it turned out that it was not the basketball player but a different man. The U.S. Supreme Court ultimately declined to hear the case while a lower court dismissed it, because Pippen could not prove that the publications knew they had the wrong person and thus had no evidence of actual malice. Even though the case was dismissed, it would have been better to avoid it to begin with. If faced with a story like this, a fact-checker would want better confirmation that the Scottie Pippen on the bankruptcy documents was indeed the basketball player, perhaps by comparing addresses and other identifying information in those documents, or contacting the athlete through his booking agent or other representative.