Anyway I'm arguing that there's a higher chance of Einstein being an atheist as atheists have higher IQ than non-atheists (whatever sematics you will use to differentiate magic believers).
I'm not sure that I entirely believe that. To the extent that it's true at all, it's most likely a crude statistic that doesn't really catch the nuances.
Here's what I mean:
Graduation from university is probably a better indicator of intelligence than simply taking an IQ test.
Here in the United States, in 2008, 27% of the over-25 population were university graduates.
But 31% of individuals who report no religious adherence are university graduates.
That's evidence that 'nones' are better educated than average, right?
Well, yes. But before we get carried away, let's look more closely at the numbers and break the figures down by several varieties of religious adherence:
Pentecostal/Charismatic - 13%
This grouping is most popular among the poor and among some racial minorities. These people often underperform in school and on IQ tests. Much of the reason for that might be socioeconomic.
Baptist - 16%
As above, but not so dramatically.
Catholic - 25%
This grouping comes in at near the national average (27%). That's not surprising, given the massive numbers (25.1% of American adults identify themselves as Catholic). My guess is that there's lots of variation within this group, ranging from Irish- and Italian-Americans who likely outperform, to many recent arrivals from Mexico who probably don't.
Mormon/LDS - 31%
The None's - 31%
Mainline Protestant - 35%
These are the Episcopalians, Methodists and Lutherans, typically the more theologically liberal denominations. Many of these people are from families who have been in the US for many generations and they are probably more upscale than average.
Jewish - 57%
This is a small group in which Torah study was traditionally considered a religious duty. As they have become secularized, their emphasis on study has broadened out to academics in general.
Eastern religions - 59%
Another small group, consisting of Asian immigrants who push their children very hard to succeed, combined with white converts who are typically drawn from better educated individuals.
The bottom line seems to be that in terms of university education, the difference between religious adherents and non-adherents is real but rather small. It's dwarfed by the variation between different kinds of religious adherents.
The numbers come from the authoritative 2008 American Religious Identification Survey (page 16 on the pdf version). If anyone is interested (it's a fascinating read) you can find the ARIS site by googling "ARIS survey".
