Thesis advisor: Peter Gottschalk / This paper examines the impact of spatial variation in HIV rates on female marriage rates in Zambia. I formulate a search model that predicts lower marriage rates of educated females relative to uneducated females in regions with higher HIV rates. I use exogenous geographic variation in HIV rates to identify the causal effect of HIV on female marriage. The risk of HIV infection causes marriage rates to fall for educated females but rise for uneducated females. One explanation is that in high HIV regions: (1) educated females take the time to find a partner who will use condoms and get HIV tested, which delays marriage, and (2) uneducated females marry sooner because youth and virginity are prized by males, and employment opportunities are scarce. These findings imply that returns to education for young females are likely underestimated since they miss conceivably substantial health-related benefits. Is widow remarriage beneficial to child school enrollment? Women are widowed at relatively young ages in high-HIV areas of Sub Saharan Africa and are likely to have school-aged children. A main finding in the parental death literature is that the death of a mother hurts child education more so than does the death of a father. This masks important differences in child school enrollment across households who have experienced a father's death. This paper estimates the effect of widow remarriage on child school enrollment by exploiting regional variation in HIV, religion, and the sex ratio. The cross-country empirical results indicate that remarriage is detrimental to child enrollment for widows with less than six years of schooling, yet beneficial to child enrollment for widows with six or more years of schooling. This is consistent with (1) marital sorting by education (correlation=.7), (2) intra-household bargaining, and (3) differences in tastes for remarriage and schooling. A policy implication is that investing in female education in high-HIV areas - among those likely to become widows - can have multiplier effects, as there is complementarity between the returns to education on marriage market outcomes and children's education. / Thesis (PhD) — Boston College, 2011. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Economics.

This dissertation consists of three essays investigating the role of early life events, family environment and personal choices in shaping a child’s chances for human capital accumulation. The first essay examines how physical stature of a child measured in terms of age standardized height influences his/her selection for family labor activities vs. schooling in rural Ethiopia using malnutrition caused by exposure to significant weather shocks in early childhood as sources of identification for the child’s physical stature. We find no evidence that better physical stature of the child leads to his/her positive selection for full-time child labor activities. On the other hand we found reasonably strong and consistent evidence that physically more robust children are more likely to combine child labor and schooling than physically weaker children. The findings indicate that, although better early childhood nutrition leads to higher chances of attending school, it may also put the child at additional pressure to participate in family labor activities which may be reflected in poor performance in schooling. The second essay empirically investigates whether the quantity deficit in the children of the mother’s preferred gender is compensated through their favorable treatment in terms of investment in schooling and nutrition (referred to as compensating hypothesis) and to what extent the mother uses her bargaining power in the family to influence this process. We use data from siblings and twins in two rounds of the demographic and health surveys of Ethiopia with robustness checks using a similar but larger data set from India. We find the mother’s bargaining power working in the opposite direction to that of the compensating hypothesis in the case of child schooling and having no substantive role in the case of child nutritional health. Our findings for child schooling imply that mother’s empowerment could turn out to be unfavorable to a child’s attendance of schooling in the circumstances where the child is needed to help out with family activities. In the third essay we use date from the 1997 cohort of the National Longitudinal Survey of the Youth (NLSY97) to examine the extent to which high school completion (and to a limited extent college enrollment) are influenced by the choice teenagers make as to when to start dating and/or engage in sex, how many dating and/or sex partners to maintain, and how frequently to engage in sexual and/or dating activities. We use indicators of parental and peer religiosity as instruments for teenager’s involvement in sex and dating activities. While our results for teenage dating are generally weaker than those for teenage sex, the overall pattern of our estimates suggests that teenage sex and dating could have significant effects not only on high school completion but also the subsequent enrollment in a college.