The Problem With Polls About Whether Obama Is a Muslim

By asking voters what they think about the president's religion, pollsters help to perpetuate a falsehood.


Liberal blogs and sites are having a field day with new data from Public Policy Polling that shows that 52 percent of Mississippi Republicans believe President Obama is a Muslim (a comparatively slight 45 percent of Alabama GOP voters agreed with them). Huffington Post had a banner headline Monday morning branding it a "SHOCK POLL," Talking Points Memo smirked at those southern bumpkins, and Daily Kos deadpanned, "Alabama and Mississippi Republicans don't believe in evolution ... but do believe Obama is a Muslim."

PPP, which is a Democratic firm, is sometimes maligned for being an unreliable pollster, but in this case the biggest problem is that they're asking the question at all. The belief that Obama is a Muslim, like the belief that he is somehow not an American citizen, is pernicious and flatly wrong. It has also been rejected by the vast majority of the American body politic, although there are some glaring examples of politicians who flirt with it to score political points. But if the goal is to fight mistaken beliefs, this is the wrong way to do it. The Dartmouth political scientist Brendan Nyhan has researched misperceptions and conspiracy-theory belief in America politics. In particular, he and colleague Jason Reifler have found that false ideas, once introduced, are very hard to get rid of. One especially bad way to fight them is to reiterate them:

The more times a false claim is repeated, the more likely people are to be exposed to it. The fewer people exposed to a false claim, the less likely it is to spread. It is also important not to repeat false claims because people are more likely to judge familiar claims as true. As false claims are repeated, they become more familiar and thus may come to seem more true to people.

The pollsters, by asking the question, and news outlets, by gleefully publicizing the results, are playing into this vicious cycle.

Besides, there's a difference between asking whether respondents believe in interracial marriage (which PPP also did) and asking whether they believe the president is a Muslim or Kenyan. The first is a matter of opinion; the second is objectively either true or false (as it happens, both are false). The problem is that it's unclear to what extent people really believe these things. Perhaps they're responding to the stimulus of a pollster asking them -- Nyhan and Reifler also found that people's answers fluctuate depending on who is asking. PPP uses automated telephone polling in which respondents answer by pushing buttons on their phones. Julian Sanchez posits that many of these responses are fed by ideology: the people who purport to hold them are really just doing so for political reasons. For example, he says, most liberals who indulged in Bush bashing didn't literally think he was a fascist. By the same token, the slight majority of Mississippi Republicans who mashed that button for "Muslim" may not really think Obama is an adherent of Islam: They just staunchly dislike him and view his opinions as so alien to theirs that they'll say so.

This impulse, which Sanchez calls "symbolic belief," would explain why so many people still question Obama's citizenship, even after he produced his long-form birth certificate. And it also explains why no one should make too much of these results. The number that matters most is the presidential preference between Obama and whoever his Republican opponent is.

Image: Tami Chappell / Reuters