Religion exerts a powerful influence on many people’s lives. In their new paper, NOUS member Niclas Berggren and Martin Ljunge investigate how self-assessed religiosity affects self-assessed health in Europe. Their sample consists of individuals with a native father and an immigrant mother from another European country. This sample allows for a causal interpretation since we can use the religiosity of the mother’s birth country as an instrument for individual religiosity in the first stage of a 2SLS regression analysis, which is related, in the second stage, to the individual’s health assessment. They find that the more religious are substantially more likely to report bad health. Several robustness tests offer a strong confirmation of the negative relationship between self-assessed religiosity and self-assessed health. Notably, this negative relationship is concentrated among women. The analysis indicates that religious constraints on women’s autonomy can impair their health.