Tag Archives: Backfire effect

Fact and Friction

The alt-truth, fake news, facts schmacts world we seem to be living in rubs a lot of people the wrong way. Living in verity’s version of the upside down might salve the addled ideological souls of some, but it gives the rest of us the fantods. Especially us empirical scholars. If society decides to call off the search for verifiable truth, after all, we’re out of business. Lacking the fib and fabrication skills readily monetized in the corporate, political and entertainment worlds, we’ll be reduced to shilling empirical verification for coppers on street corners and editing Wikipedia entries on spec.

Well, good news. Despite all the hand wringing, there’s reason to believe that the reach and impact of fake news is, um, fake news. Even more cheering for those of us in the learnin’ biz, there is some persuasive evidence that facts are not quite the ideological Play-Doh some people clearly want them to be.

This isn’t to say the past couple of years hasn’t seen a particularly nasty beat down of bona fide veracity, especially by certain presidents of the United States I could name. Everyone expects a degree of truthiness from politicians, but respectable fact checking sites suggest Donald Trump is less guilty of the occasional white lie than the madcap production of technicolor extravaganzas. Politifact lists ten pages of verified false claims made by the president. The New York Times has a running tally showing Trump telling more provable falsehoods in 10 months than Barack Obama told in his entire eight-year administration.

Aiding the White House as the new home of the whopper is the full-on weaponization of social media. Russian trolls seem to work Facebook’s algorithms with impunity while Twitter enables the wholesale spraying of perfidy and perjury. For the past couple of years a careful observer could be forgiven for concluding that our political system, with premeditation and purpose, was abandoning the truth wholesale. Just how many people were consuming fake news? Did it herald the decline of mainstream media and the professional norms of journalism? Were facts being kicked to the side by voters? Have we gone completely nuts? The people tasked with sorting signal from noise and answering these sorts of question systematically are my tribe–empirical social scientists–and they operate on slower timelines than the 24/7 news cycle. The rise of alternative realities happened so fast that the only honest answer we had to these sorts of questions was, “damned if we know, but it’s pretty worrying.”

That’s starting to change. A couple of studies have recently surfaced that suggest fake news is scary but not enough to frighten the Republic into fact-addled delirium, and, even more comforting, they find that facts themselves still trump fibs, or at least give fibs a good argumentative wedgie.

The first of these studies does contain some kind of scary numbers (you can find the full study here) . Roughly a quarter of American adults, or 65 million people, visited a fake news website* in the month leading up to the 2016 election, most of them making that connection through a Facebook link. Moreover, most of these fake news consumers almost never visited reliable fact-checking websites.

The good news is that heavy consumers of fake news make up a very small proportion of Americans. Roughly 60 percent of the visits to fake news sites came from a small group (about 10 percent of adults) who were older, conservative and (very) pro-Trump. So perhaps the fact-free fabulist babble bubble everyone is so worried about is not a dome covering the Republic, some sort of hermetically sealed covering threatening to asphyxiate the electorate with the gas of toxic make-believe. Maybe it’s just a pocket-sized greenhouse in the backyard where your crazy uncle is getting light headed from inhaling alt-media political poots and discussing conspiracy theories with the geraniums.

The second study  is deep empirical dive into what’s known as the “backfire effect” (you can the full study here) . The basic idea of the backfire effect is that if you present someone with a fact that counters or corrects a politically pleasing falsehood it makes people more not less likely to support that untruth. Evidence of the backfire effect has popped up in previous research and raised some interesting questions. Are people really so committed to their political alternative realities that pointing out contradictory facts will only make them more committed to insisting on the truth of falsehoods?

Given what’s happened over the past couple of years in the political arena that’s a pretty important question. This study sought an answer by giving people factually incorrect claims made by prominent figures on the left (e.g. Hillary Clinton, Barack Obama) and the right (e.g. Donald Trump, Sarah Palin). They then randomly exposed some subjects to a factual correction of those statement, and asked everyone to evaluate the original claim. The difference in evaluations between those exposed and those not exposed provides a measure of the impact of factual information on political claims. Through five studies and more than 10,000 subjects, they did not find a single instance of backfire across more than 50 policy issues. Indeed, what they found is that people, regardless of ideological orientation, are pretty responsive to facts contradicting their political preferences, and will shift their evaluations towards the factual evidence when it is presented with them.

Neither of these studies should be considered definitive, and both come with the usual cautions and caveats of empirical social science research (my tribe’s motto is plus research opus, i.e. more research is needed). Still, given the hand wringing over fake news and the embrace of alternative realities and their potentially corrosive impact on politics, I think it is okay to view these findings with a small measure of relief. There’s a reasonable case here that most citizens are not voraciously consuming fake news inside their own political echo chambers, though Facebook and Twitter can make it look that way. And, even if they are, facts still seem capable of putting the brakes on fake. Let’s hope we get more of that sort of friction in 2018.

*What exactly constitutes a fake news website is a matter of some controversy. The authors of this study relied on previous research identifying “news” websites that repeatedly published demonstrably false stores.