On the causal effect of religion on life satisfaction using a propensity score matching technique

Using the British Household Panel Survey (BHPS) data set, we investigate the effect of religion on subjective well-being (SWB), specifically taking into account the implication of selection effects explaining religious influence. In order to measure the level of religious involvement, we construct different indices on the base of individual religious belonging, participation and beliefs. By applying a Propensity Score Matching (PSM) estimator, we find evidence that the causal effect of religion on SWB is better captured than through typical regression methodologies focusing on the mean effects of the explanatory variables. Our results show that religious active participation plays a relevant role among the different aspects of religiosity; moreover, having a strong religious identity such as, at the same time, belonging to any religion, attending religious services once a week or more and believing that religion makes a great difference in life, has a high causal impact on subjective well-being. Our findings are robust to different aspects of life satisfaction.