Income and Religiosity
The most recent Pew Global Attitudes Survey report (The Pew Global Attitudes Project, October 4, 2007) has this neat graphic (click on it to see a much larger version):
The report explains that respondents were given one point if they believed faith in God is necessary for morality, one point if they said religion was very important in their lives, and one point if they prayed at least once a day. The religiosity score was the sum - so it could range from zero to three for each respondent. I assume the dots show country averages.
Here's the summary of the Pew Center's analysis:
Global publics are sharply divided over the relationship between religion and morality. In much of Africa, Asia, and the Middle East, there is a strong consensus that belief in God is necessary for morality and good values. Throughout much of Europe, however, majorities think morality is achievable without faith. Meanwhile, opinions are more mixed in the Americas, including in the United States, where 57% say that one must believe in God to have good values and be moral, while 41% disagree.
The survey finds a strong relationship between a country’s religiosity and its economic status. In poorer nations, religion remains central to the lives of individuals, while secular perspectives are more common in richer nations. This relationship generally is consistent across regions and countries, although there are some exceptions, including most notably the United States, which is a much more religious country than its level of prosperity would indicate. Other nations deviate from the pattern as well, including the oil-rich, predominantly Muslim – and very religious – kingdom of Kuwait.
The report itself has a much more detailed discussion. The graphic doesn't answer the causation question.

I wonder whether it is declining income that encourages religiosity or whether it is the other way round. Or it could just be a coincidence
Posted by: Tejvan Pettinger | January 16, 2008 at 03:39 AM
There are obviously quite a few confounding variables here. I don't think there is a causal link between the two in either direction, just an illuminating correlation. For one, I would guess education is probably the biggest confound. Higher education rates have an adverse effect on religiosity and probably a positive effect on GDP.
Posted by: Andrew | February 07, 2008 at 12:26 AM