By now, you’ve probably heard of the Pew center poll that found that fewer Americans believe Obama is a Christian and more believe he is a Muslim or simply don’t know what his religion is. The question is: why have people’s beliefs changed?
That doesn’t get us very far. It seems hard to imagine that Americans suddenly got dumber between March 2009 and August 2010, when the Pew polls were conducted.
A second hypothesis comes from political scientist Brendan Nyhan, who notes that there have been numerous attempts by some media commentators and political leaders to insinuate that Obama is Muslim. He writes:
Rather than faulting the public for the weaknesses of human psychology, we should identify the elites who deceive citizens with false information and hold them accountable for their role in fostering this myth. It’s time to stop blaming the victims.
If Nyhan’s hypothesis is true, we would expect to see sharper changes over time among people who are, first, predisposed to believe bad things about Obama. This implicates Republicans, and, indeed, Pew found that Republicans registered the sharpest increase in the belief that Obama is Muslim. Second, among Republicans, we should see especially sharp changes among those who pay attention politics and the news, because these people who would be more likely to watch, read, or hear any commentators and leaders suggesting that Obama is Muslim.
Via a contact at Pew, I asked them for additional information from their March and August polls: the results broken down not only by party, but also by political attentiveness. The best measure of attentiveness in their surveys was the respondent’s level of formal education, which is a plausible but imperfect proxy for attention to politics. Nevertheless, it’s what I had to use.
Here are the trends from the March 2009 to August 2010 polls in the perception that Obama is a Muslim. I divide the sample into Democrats and Republicans. Independents who lean towards a party are counted as partisans (see here for why), so this analysis includes about 90% of the sample. I then divide the sample into the education categories that Pew provided: those with a high school degree or less, those with some college education, and those with a college degree or more.
The growth in this perception among Democrats is small and is consistent across education levels: a 2-4 increase within each level. By contrast, the growth in this perception among Republicans is more notable among those with some college education (a 19-point increase) or a college degree (15 points) than among those with a high school degree or less (9 points). In other words, better educated Republicans have changed more than the less educated Republicans. This flies in the face of the “dumb Americans” idea and provides some support for Nyhan’s hypothesis. The people most likely to hear the “Obama is a Muslim” meme are the ones whose beliefs changed most dramatically in the past 17 months.
Below is the full set of results, including the percentages who said Christian, Muslim, or expressed no opinion. (Click to make it larger.) Again, this tells a similar story: larger changes among the better educated Republicans. For example, the decline in the percentage of Republicans who believe that Obama is a Christian is larger among those with some college (-31 points) or a college degree (-24 points) than among those with a high school degree or less (-11 points).
Obviously, we cannot draw definitive conclusions from this analysis. It does not prove that some media personalities and political leaders are responsible for the increasing perception that Obama is a Muslim. But it points in that direction.
[Cross-posted at Ezra Klein’s blog]
fn1. I thank Jocelyn Kiley of Pew for providing additional data. Neither she nor the sponsors of the polls—the Pew Center for the People and the Press and the Pew Forum on Religion and Public Life—bear any responsibility for my interpretation.