There is a large amount of empirical literature reporting that people who regularly attend religious services tend to have better health outcomes. However, it remains an unanswered question as to whether the observed correlation reflects any causality. Exploiting exogenous changes in church...