If 2016 marked the arrival of “fake news”—widely disseminated false news stories—recent events have made clear that the threat is here to stay. Recent European elections, including last week’s Italian elections, have been rocked by patently untrue stories. A slew of similarly fabricated stories is sure to hit U.S. shores ahead of the 2018 elections. Some will come from malevolent foreign actors with political motives. Others will come from shadowy companies just trying to make a buck. Still others will come from trolls who delight in mischief-making.
Whatever their origin, there’s no doubt that fake news is a serious problem. But all hope is not lost. It turns out that there’s a simple way to get people to disbelieve the fake news they encounter: Tell them the truth. Our research, forthcoming at the Journal of Experimental Political Science, makes clear that everyday people, across the ideological spectrum, are willing to reject fake news and accept a factual correction. When people who see a fake news story are told that the story they’ve read is plainly wrong, they reject the fake news story.
Indeed, this is even true when a fake story aligns with one’s political beliefs. If you’re a conservative and you’re told that a fake story flattering to conservatives was actually fabricated, you’ll likely respond by conceding that the story is fake. The same is true for liberals and independents. However, there’s one thing that proof of fake news won’t do—get people to change their political convictions.
ADVERTISEMENT
Here’s how our study worked. We recruited people to take a survey and randomly assigned them to read a fake news story. All the fake stories were taken from real life, and selected across the political spectrum. Some were of recent vintage, such as the 2016 “Pizzagate” story, which claimed Democratic operatives were operating a secret ring of pedophiles in a D.C.-area pizza parlor. Others were older, such as allegations that President Obama was not born in the United States.
We even took stories initially reported by mainstream news outlets that were later revealed to be entirely inaccurate, such as The Washington Post’s claim that the Russians had infiltrated Vermont’s power grid. And we looked at fake stories bubbling among conservative websites that credit President Trump for false accomplishments, such as a crackdown on sex trafficking.
After reading their story, some subjects were randomly assigned to read a correction. This allows us to measure changes in opinion over and above background knowledge of these stories. For example, to correct The Washington Post’s story, we relied on that newspaper’s own lengthy correction. To correct false claims about Obama’s birth, we simply showed people a replication of his birth announcement in a Hawaiian newspaper.
We then asked everyone to rate their level of agreement with the claim advanced by the fake story. We compared those who were randomly assigned to read a correction to those who were not, helping us discern whether the correction moved the needle and changed readers’ minds.
As it turned out, the average respondent was responsive to the correction, no matter their political beliefs. For every fake story, the average liberal, moderate, and conservative responded to the factual correction by rejecting the fake news. Of course, in some cases liberals were less enthusiastic about rejecting fake news stories that favored their side’s interest. In some cases, conservatives did the same. But in no case did we observe average survey participants—or even ideologically motivated respondents—rejecting the correction.
If you’re concerned about the spread of fake news, we have a few recommendations. First, tell the truth. If you encounter someone spreading fake stories, don’t shy away from reporting the facts in response. Whether it’s your racist uncle or a long-lost friend from high school, our results indicate that they’ll probably be willing to accept the facts. That is likely to be the case even if they’re coming from the other end of the ideological spectrum. Relatedly, when you rebut fake news, try to rely on non-partisan sources like Snopes.com. In your rebuttal, get to the point, and leave the partisan attacks for another time.
But be realistic about the consequences. As two of us have made clear in other research, you should not expect that telling people the truth will cause them to actually change their views about politics. While people are content to concede that a fake story is indeed fake, they are far less willing to part with the political principles they hold dear. A factual correction probably won’t change one’s preferred party, candidate or policies.
If you voted for Clinton in 2016 and aren’t sure why Trump voters don’t seem to care if Trump lies or if stories about Trump prove to be fake, consider the following hypothetical. If a pro-Clinton fake news story had circulated during the 2016 election, and you came to learn that it was fake, would you accept the facts? We think you would. But would a correction to one pro-Clinton fake story cause you to switch your vote, or even temper your enthusiasm for her compared to Trump? To put it mildly, we’re skeptical.
Americans care about their political parties, and they hold tightly to their political beliefs. But they also like being accurate. Our study suggests that there’s no reason to leave fact-checking just to those in the fact-checking business. We can all play our part, and confront untruths wherever we may find them. Of course, we’ve got our work cut out for us. Even as late as last year, more than one-third of all Americans believed Obama was born in Kenya, and 28 percent of citizens were at least willing to entertain the possibility that “Pizzagate” was real. But throwing up our hands in despair isn’t the answer. Doing the hard work of truth-telling—and setting our expectations appropriately—is all of our responsibility.
Ethan Porter is an assistant professor at George Washington University. Thomas Wood is an assistant professor at Ohio State University. David Kirby is an adjunct scholar at the Cato Institute.
