In Sickness and Wealth:Three Essays on Health Human Capital and HIV in sub-Sahran Africa

Abstract

This dissertation concerns itself with the effects of education and wealth on health in Africa. The first two chapters focus on the impact of adult education levels on the HIV epidemic in four sub-Saharan countries. The last chapter turns to children's health and examines whether the urban advantage in health persists despite rapid urbanization in South Africa.

Chapter one employs longitudinal HIV data from Mali, Tanzania, Kenya and Zambia to examine whether the positive relationship between educational attainment and HIV prevalence is changing. I find evidence that the relationship between HIV and education has begun to reverse. Although it remains positive at the regional level, its much weaker for the youngest cohort. Furthermore, I find no association between HIV and education at the individual level among the youngest cohort. Secondarily, I test two explanations for change- erosion of educational infrastructure and adoption of protective knowledge among the educated. I find evidence consistent with the hypothesis that education is becoming protective as the epidemic matures; regions with higher average adult education at baseline experience larger drops in HIV prevalence.

Chapter two builds on these findings by examining whether the behavioral response to HIV is stronger among the more educated. I find a robust positive association between education and condom use, HIV testing, and age at marriage, with evidence that younger cohorts may be reducing age at marriage. I also find that more educated individuals are increasing their rates of HIV testing and reducing age of marriage more than the less educated.

Finally, I use anthropometric scores from two national surveys from South Africa to examine changes in urban and rural children's health over 15 years. I find that the urban health advantage disappears despite urban children retaining advantages in average household wealth. I then explore several common explanations for this pattern, including the growth of particularly vulnerable urban populations or deepening urban poverty. I find no evidence of deteriorated circumstances for the urban poor, although urban-rural migrants have begun to show a health disadvantage. I find that the differential gains are likely due to improvements made by very poor rural households.