HIV and fertility: long-term evidence from sub-Saharan Africa

Guenther Fink, Harvard School of Public Health
Sebastian Linnemayr, Harvard University

In this paper, we use HIV/AIDS induced variations in mortality to investigate the interactions between parental fertility choice and mortality risk at different stages of life. We develop a simple theoretical model to show that the optimal fertility adjustment to HIV is larger for more educated parents than for parents with little or no formal education. We test this prediction using a novel data set combining historical individual level data from World Fertility Surveys (WFS) with recent data from the Demographic and Health Surveys (DHS). The result that more educated women reduce fertility more than uneducated mothers in the presence of HIV appears to hold both in the longitudinal and the cross-sectional data. Our results imply that HIV is unlikely to have a significant effect on population size, but will negatively affect countries’ long term economic prospects through an adverse shift in the population’s human capital composition.