Intriguing, and take it as exploratory.
Evidence of relationships between religious affiliation and the African AIDS pandemic is found in the medical, religion, and sociology literature. In particular, studies have shown that predominantly Christian countries tend to have higher HIV rates than predominantly Muslim countries. These relationships have been largely unexplored by economists and we seek to identify underlying institutions using a panel of up to 43 sub-Saharan African countries for 1990-2010. Catholic antagonism towards condom use has often but proposed, but we report that the protestant (rather than the Catholic) population share drives the Christianity/HIV correlation. (Also, condom use actually correlates positively with HIV prevalence, though reverse causation likely plays a role). Male circumcision rates have a large negative effect on HIV prevalence. While male circumcision has been linked to Islam in this context, we report that the male circumcision effect is robust to controlling for the Christian population share while the correlation of HIV prevalence and the Muslim population share is not. There is no significant relationship between an index of social regulation of religion and HIV prevalence.
From a new paper by Amanda Mandzik & Andrew T. Young.
