Identifying the causal effect of income on religiosity using the Earned Income Tax Credit

How do economic conditions – income and the generosity of social insurance–affect religious behavior? This paper provides novel evidence for the United States using variation in Earned Income Tax Credit benefits across states, over time, and by number of children. In the raw data, income and religiosity have a trivial relationship. In contrast, the instrumental variables strategy shows that increases in income from the EITC are found to sizably reduce religious attendance among low-income individuals. Among individuals who attend, the results indicate that an additional $1000 in annual income is associated with a decrease of roughly 1 service per year (as compared to an average of 16 services per year). The effect arises mainly along the intensive margin, decreasing devout behavior and increasing more marginal attendance, but having little impact on whether an individual participates at all or on beliefs. There are also minimal effects on the occurrence and amount of contributions to religious organizations. The results extend our understanding of the determinants of religiosity, the relationship between governmental income support and religious behavior, and the social effects of the Earned Income Tax Credit.