Do facts matter in politics? The cynic says not. Politicians seem to lie with impunity, brushing off the disapproval of the press; voters cocooned in ideology will believe anything that reaffirms their worldview and ignore anything that contradicts it. Lawmakers like Michele Bachmann, the Minnesota congresswoman notorious for her estrangement from the truth, keep getting elected. When a presidential nominee's pollster declares, "We're not going to let our campaign be dictated by fact-checkers," would-be referees seem helpless to set the record straight.
A couple of political scientists, Brendan Nyhan of Dartmouth and Jason Reifler of the University of Exeter, recently set out to test this despairing view with a field experiment. Their findings suggest it's not so hopeless after all: Politicians actually do seem to care whether they get caught lying, and they lie less when they know they're being watched.
During last year's election, Nyhan and Reifler picked nearly 1,200 state legislators in states with active affiliates of PolitiFact, the nonpartisan website based in Florida that seeks to evaluate politicians' claims and rate their validity. To one-third of the lawmakers, chosen at random, Nyhan and Reifler sent a vaguely threatening letter. It alerted the lawmakers that PolitiFact was monitoring them and speculated about the potential consequences to their careers:
Politicians who lie put their reputations and careers at risk, but only when those lies are exposed. That’s why we are especially interested in the consequences of PolitiFact verdicts and other fact-checking efforts in your state. Here are examples of the types of questions we are interested in:
• Are “false” or “pants on ﬁre” verdicts damaging to the reputation or political support of political candidates?
• Do election campaigns use “false” or “pants of ﬁre” verdicts in their advertising to attack their opponents?
• Will state legislators lose their seats as a result of fact-checkers revealing that they made a false statement?
Another one-third of the legislators got a "placebo" letter: It told them they were part of a political-science experiment "studying the accuracy of the political statements made by legislators," but no more. (The purpose of this was to distinguish between the potential effect of getting any letter at all and the effect of the wording of the treatment letter.) The final one-third got no letter.
At the end of the election, the researchers looked at the politicians' record. How many had been called out for lying, either by their state's PolitiFact affiliate or in a news story? The results were impressive: The politicians who didn't get reminder letters were more than twice as likely to be criticized for inaccuracy than those who did. "Our results indicate that state legislators who were sent letters about the threat posed by fact-checkers were less likely to have their claims questioned as misleading or inaccurate during the fall campaign—a promising sign for journalistic monitoring in democratic societies," the researchers concluded.
The effect of the letters was significant, but the number of legislators who got caught lying was strikingly small: just 21 in the control and placebo groups (2.7 percent) and four in the group that got the reminder letters (1 percent). The researchers had chosen to focus on state legislative races because there are so many of them and because they probably rely more on media coverage, and less on paid advertising, than bigger, better-funded campaigns. But it turns out PolitiFact doesn't pay much attention to state legislative races, either. In two of the states, Florida and Virginia, PolitiFact didn't evaluate any statements by state legislators at all, possibly because both states had active Senate and presidential contests. In New Jersey, five of the seven ratings issued were of a single lawmaker, state Senator Joseph Kyrillos, who was running for U.S. Senate against Bob Menendez. Kyrillos, who lost, got one "false" rating, two "mostly falses," one "half true" and one "mostly true."
The results, however limited, do suggest that politicians are more likely to watch their facts when they're threatened in advance with having their statements vetted. But the really interesting questions are the ones posed by Nyhan and Reifler's treatment letter. Do candidates' reputations suffer when they're shamed for fudging facts? When opponents turn these ratings into attacks, do candidates lose elections as a result?
Those are harder questions for political science to parse, Nyhan told me. "The problem is the lack of experimental control," he said. "If we had a way to randomly assign fact-checks, we could see if politicians got a lower share of the vote, but absent that, it's hard to rule out the possibility that any differences in vote share we observe are the result of the selection process used by the fact-checkers (e.g., targeting competitive races)." By asking suggestive questions, the researchers implied the politicians would suffer at the polls for their lies, but that hasn't been proven.
Bachmann, who unsuccessfully sought her party's presidential nomination last year, has long been a favorite target of fact-checkers. For years, she boasted a PolitiFact record that was perfect in its way: Every statement of hers the website checked rated either "false" or "pants on fire." Once, at a debate in Iowa, she boasted, "PolitiFact came out and said that everything I said was true." PolitiFact, which had done no such thing, rated that statement "pants on fire." Finally, on the 23rd try, Bachmann got her first "true," but it was a rare exception. By the time she announced her retirement in May, PolitiFact had evaluated 59 of her statements; 75 percent of them were "mostly false," "false," or "pants on fire."
But Bachmann did seem to pay a price for her record of factlessness. National news outlets ran articles on her tendency to err; her opponents aired ads highlighting it. In 2012, she very nearly lost reelection despite representing the most Republican-heavy district in Minnesota, and she decided to retire rather than face voters again.
Fact-checkers alone certainly didn't end Bachmann's career, but it's hard to imagine they didn't play a role. When she announced she was stepping down, PolitiFact ran an article commemorating its long relationship with her, seeming almost to mourn its favorite target. But the article ended on a hopeful note: "Bachmann isn't leaving immediately, however," it noted, "so we will keep the Truth-O-Meter turned on."