Religiosity and Social Capital - Causal Heterogeneity in a Double Comparative Perspective

This paper examines the context dependency of religion’s effect on social capital, ie. civic engagement and social trust. I aim to explain the causal heterogeneity between citizen’s individual religiosity and their civic attitudes and behavior. This will be done from a double comparative perspective. Religiosity’s effect on social capital may vary a) across religious groups and b) across national countries. Methodologically, I translate this theoretical problem into a Non-nested Varying Intercept Varying Slope Multilevel Model. The idea behind this approach is to model two different context levels which are not nested in a strict hierarchy but rather overlap, ie. religious groups exist in many countries, and one country is often home to more than one religion. The models are estimated using Bayesian inference. Results based on the European Social Survey indeed suggest that there is causal heterogeneity between religiosity and social capital that varies both across religious groups and nations. The exact pattern of variation however differs depending on the dimensions of religiosity and social capital examined. Furthermore, it is possible to explain this variation. While religiosity has a stronger impact on civic engagement in more secular societies, social trust is only fostered by religion if state and church are separated. This leads to the paradoxical conclusion that religion matters more for social capital if its general societal statues is weak.