Religion and the Poor

Tyler Cowan posts a link to an interesting paper on religion.

What is the role of religious institutions and religious workers in the racial earnings gap in the United States? In this paper we explore the relationship between childhood exposure to religious density, as measured with the number of religious workers at the state level, and the labor market outcomes of the worker thirty years later. We use data that spans over fifty years to identify changes in earnings due to early exposure to religion: our first source of identification uses changes in these two variables within states, and our second source of identification uses states’ differences by following workers who moved to a different state. Our results suggest that living in a state with a an extra clergy member for each 1,000 habitants increases the earnings of black workers by 1.7 to 3.6 percentage points relative to white workers.. In addition we show that this relationship is robust to different measures of exposure to religious density, and that these estimates increase to 7.6 percentage points when the change on religious density is defined exclusively increasing an extra black religious workers for each 1,000 habitants. Finally, we estimate a series of robustness tests that suggest that these results are not due to spatial sorting across states, nor to secular time trends associated with changes in labor market outcomes for black American workers.

 It’d be interesting knowing how religion affects wages independent of race.

2 thoughts on “Religion and the Poor”

  1. This is a very rough estimate. The direct effect of clergy on earnings in this analysis is strongly negative and can be directly seen in Table 1 (as well as the cross-term which is the effect you quoted). But I think this is just picking up the simple fact that poorer (“red”) states are more religious and that richer (“blue”) states have more urban poverty that is religiously concentrated. (I couldn’t find any attempt to control for regional income levels.)

    Also, and partly because of this, my confidence is quite low about this really picking up much of a causal effect. Still interesting though.

Leave a Reply

Your email address will not be published. Required fields are marked *