Abstract
On the basis of simple rational-choice theory, it is argued that Christian religious involvement influences the divorce, abortion, non-payment, and children-born-out-of-wedlock rates negatively. This influence stems from an expectation that a god punishes certain forms of action and from social reprimands of devotees. Regression analysis is then used to test the theory in the case of Sweden in the year 1990, and the importance of religion, in its negative effect on all of these rates, is strongly confirmed. A particularly interesting result: the marginal effect on three of these rates of increasing the fraction of people active in the Church of Sweden is clearly larger than the marginal effect of increasing the fraction of more conservative free-church members.