How Not To Measure the Popularity of a Conspiracy Theory

Defining "birther" down.


Superman's no birther!
DC Comics

"Sixty-three percent of registered voters in the U.S. buy into at least one political conspiracy theory," a team of pollsters at Fairleigh Dickinson report. That may seem absurdly low—I would think the number is somewhere in the high nineties—but of course the pollsters don't mean just any theory that involves politicians conspiring; they mean the theories that are disreputable. Specifically, they found that 20 percent of the country think it "probably true" that Obama stole the 2012 election, 23 percent that Bush stole the 2004 election, 25 percent (including 36 percent of Democrats) that Bush knew about 9/11 in advance, and 36 percent (including 64 percent of Republicans) that "President Obama is hiding important information about his background and early life."

The pollsters helpfully note that the last statement "would include what's often referred to 'birtherism.'" Sure enough, it would. It would also include a wide range of theories that are not birtherism; indeed, phrased that way, it could include the vague suspicion that Obama still has some skeletons in his closet, something that could very well be true of any prominent politician. Nonetheless, some readers seem intent on interpreting the answer as a mass endorsement of the Orly Taitz/Donald Trump birth-certificate obsession. ThinkProgress greeted the results with the carefully phrased headline "As Many As 64 Percent Of Republicans Are Birthers, Poll Finds." Salon declared that "The birther myth is the most widely believed political conspiracy theory in America," which is doubly inaccurate: It assumes that all of those 36 percent are birthers, and it assumes that none of the conspiracy theories that aren't in the survey are more popular. (Just a decade ago, an ABC News poll showed 70 percent of the country believing some sort of conspiracy was behind JFK's death. And then there's the conspiracy theories that aren't usually classified as conspiracy theories—say, "Saddam's Iraq and its longtime rival Iran were secretly allied," an idea that a president of the United States endorsed when he declared that the countries were part of an "axis of evil.")

The poll also included some basic questions about American politics, to see how a belief in conspiracies correlates with political knowledge. The pollsters' conclusions:

No wife, no horse, no mustache.
Discordian Society

In general, higher levels of actual knowledge about politics tends to reduce belief in conspiracy theories….However, the relationship between current events knowledge and belief in conspiracy theories is conditional on partisanship. Among Democrats, each question answered correctly reduces the likelihood of endorsing at least one of the conspiracy theories by seven points. Among independents, each additional question reduces it by two points. For Republicans, though, each additional question answered correctly tends to increase belief in at least one of the theories by two points.

"There are several possible explanations for this," said Cassino. "It could be that more conspiracy-minded Republicans seek out more information, or that the information some Republicans seek out just tends to reinforce these myths."

Yes, those are possible. It is also possible that the effect would disappear if you either narrowed that broad Obama question down to a specific theory ("President Obama is covering up the circumstances of his birth") or balanced it with a similarly vague statement about a prominent member of the other party ("Mitt Romney is hiding important information about his background and early life"). Before you try to forge a theory about your results, it's a good idea to figure out just what you've measured.