Essays in development economics

Abstract

This dissertation brings together three essays on the relationship between education, health, and family structure in developing countries. The first essay studies the impact of the AIDS epidemic on children's schooling in Kenya. I draw on the relationship, established in previous literature, between the lack of male circumcision and HIV prevalence. The Luo ethnic group, who does not generally practice male circumcision, had a much larger increase in the HIV prevalence rate between 1993 and 1998. I show that there was a corresponding increase in orphan rates and a decrease in educational achievement among the children in this group. This does not seem to be accounted for by alternative explanations, such as changes in the political clout of the Luo or mean reversion. The second essay examines the impact of sibling sex composition on educational outcomes of children in Tanzania. The estimates suggest that 14-year-old children with three sisters are 24 percentage points less likely to complete primary school (7th grade) after completing 6th grade than children with no sisters, and 8.4 to 9 percentage points less likely to complete primary school overall. Having two or more older sisters, however, can benefit children in completing 4th, 5th, or 6th grade of primary school. These results are robust to the inclusion of parental background characteristics and an index measuring household assets. The third essay, co-authored with Professor Esther Duflo, takes advantage of a school construction program that took place in Indonesia between 1973 and 1978 to estimate the effect of education on fertility and child mortality. Time and region varying exposure to the school construction program generates instrumental variables for the average education in the household, and the difference in education between husband and wife.

(cont.) We show that female education is a stronger determinant of age at marriage and early fertility than male education. However, female and male education seem equally important factors in reducing child mortality. We suggest that the OLS estimate of the differential effect of women' s and men's education may be biased by failure to take into account assortative matching.