Friday, June 17, 2011

Religiosity vs. State Failure: Guess the Correlation

Y axis: Fund for Peace state failure index (higher = more failure)

X axis: Religiosity (Gallup Poll, "Is religion important in your daily life?", % answering yes). (Gallup report here.)

That's a pretty good R^2:

Possible biases: the less developed (more failed) a country, the less likely it was to have religiosity data. Also China is 1/6th of the world's population and did not have religiosity data.

I haven't plotted or religiosity vs life satisfaction (read, happiness) although it would be tough to argue that happiness doesn't correlate inversely with failed-state-ness. Another interesting one, though with harder-to-find data, would be per capita economic growth by nation over the long term (in decades or centuries) compared with religiosity in the same periods compared against time (the trend would be a surface on a 3D plot). I expect there would be a big hole in it for Europe from A.D. 400-1400. Now why would that be?

[Added later.] I noticed that the scatter plot appeared (to abuse the term) bimodal - there's a break between the very religious and less religious countries. I suspected based on my previous eyeballing of the data that predominantly Muslim countries would be overrepresented among those countries most overenthusiastic about religion; after all, it's Abrahamic Religion v3.0, and it had better do a good job of getting its customers excited! So I did a little bar chart, breaking down the countries in each bin by religion (non-integer values allowed, i.e. Benin is half-Muslim, half-Christian like several other West African countries). The bins are, as above, the percentage of people in that country that replied yes to the Gallup poll "Is religion important in your daily life?"

NB5=Non-Big-5 religions (mostly local African religions), H=Hindu, EA=East Asian (Buddhist, Confucianist), M=Muslim, C=Christian. In particular note the expansion of the Muslim share at the religion=important end.

I suspect that there are better predictors of religiosity than just whether your country is Muslim (like its overall level of development; less developed Christian countries are more religious than more developed Christian ones) but it's clearly an important correlation.

No comments: