From Predictive to Protective? Is the Relationship between HIV and Education Changing?

Abstract

This study uses two rounds of HIV prevalence data from eight sub-Saharan African DHS to examine the hypothesis that positive association between educational attainment and HIV prevalence is reversing. I compare a group of high prevalence countries to a group of low prevalence countries and find that the education-HIV relationship has weakened in higher prevalence countries among the youngest cohort. This is true both across and within regions. However, the association remains statistically significantly and positive across age cohorts in both country groups. Secondarily, I test for two commonly hypothesized explanations for such a change - the erosion of educational infrastructure in high prevalence areas and the behavior change among the educated. I find evidence of educational erosion among the low prevalence group of countries only. I also find evidence of an educational gradient in protective behavior in both country groups, but it does not appear to affect the Ed-HIV association.