In this post, I criticized psychologist Satoshi Kanazawa’s widely discussed recent article claiming that more intelligent people are more likely to be politically liberal. Kanazawa also claims that his data shows that more intelligent people are more likely to be atheists. This claim is better supported than the article’s argument about political liberalism. Nonetheless, the evidence isn’t nearly as strong as Kanazawa suggests. My own conjecture is that Kanazawa may be right about countries such as the United States, where atheists are a small minority, but less likely to be correct about the many nations where atheists are a much larger fraction of the population.
On the plus side, Kanazawa’s measure of atheism (survey responses stating a lack of religiousity and lack of belief in God) is better than his extremely dubious definition of liberalism. Even this part of his analysis is not completely airtight, since lack of religiosity doesn’t necessarily equate to lack of belief in God. Still, the two are at least highly correlated, and Kanazawa gets similar results for a General Social Survey question that directly asks respondents whether they believe in God. Controlling for various other variables, including education, gender, and race, more intelligent respondents are more likely to say that they don’t.
Unfortunately, however, Kanazawa improperly generalizes from the US results. Only about 2 to 10 percent of Americans are atheists and agnostics. In an overwhelmingly religious society, it is probable that relatively more intelligent people would be more likely to question conventional wisdom – especially since many of the arguments for atheism are counterintuitive. Kanazawa wrongly assumes that atheism is just as uncommon in the rest of the world as in this country. He cites the fact that The Encyclopedia of World Cultures refers to atheism in its descriptions of only 19 cultures, [...]