God is in the Details: A Reexamination of the Role of Religion in Economic Growth

Recent work has suggested an important role for religious beliefs in affecting economic growth. We evaluate the robustness of such findings when other potential growth determinants, such as institutions, fractionalization, and geography, are also considered. We further evaluate the role of religion as an ultimate as opposed to proximate growth determinant. To do this, we employ model averaging methods that facilitate inferences when a researcher does not wish to assume knowledge of the true data generating process of the phenomenon under study. Our results suggest that, contrary to recent claims, religiosity variables such as belief in hell, belief in heaven, and monthly church attendance do not appear to be either quantitatively significant or important in explaining cross-country growth differences whereas index variables for religious affiliations do. This suggests that the source of previous findings of a religion/growth nexus is unobserved heterogeneity rather than religiosity per se.