Dissertation Title:  "Essays in Health Economics: Using Evidence from Sub-Saharan Africa"

Paper 1: The effect of price on demand and use of bednets: Evidence from a randomized experiment

An on-going debate currently exists about whether health products, like insecticide-treated bednets (ITNs), should be distributed for free or at a positive price to maximize ownership and use of these goods. This study evaluates a randomized experiment implemented in Madagascar where net price was randomly assigned at the village level. The results show a significant negative effect of pricing both on household demand and use of ITNs. Net ownership and use are maximized when price is zero or close to zero. The study fails to reject a selection effect of zero, meaning that households which are more at risk of malaria will not be more likely to own or use nets. The study also fails to reject a sunk cost effect of zero which means that households do not exhibit the hypothesized behavioral response of having higher rates of use for goods that they pay more for. Smaller sunk cost and selection effects may exist but the study is not powered to detect such effects. These results suggest that even partial subsidization of health goods like ITNs significantly limits their ownership and use and that free or small nominal prices will maximize ownership and effective coverage.

Paper 2: Long-term effects of in-utero rainfall on maternal mortality: Evidence from sub-Saharan Africa

Although recent estimates of maternal mortality rates show that progress has been made in reducing maternal deaths in many countries, the countries in sub-Saharan Africa continue to have some of the highest rates of maternal mortality worldwide. One possible explanation may come from conditions during a girl’s in-utero period, which may have long-term consequences on her maternal survival later in life. I examine this hypothesis by using data from sub-Saharan African countries on rainfall in-utero and during young ages, which I relate to subsequent maternal mortality. The results show that a positive rainfall shock during the in-utero period reduces the probability of pregnancy-related deaths by 60%. I investigate the plausible mechanisms to link rainfall shocks in-utero to maternal mortality and find the strongest evidence in favor of an effect from positive rainfall shocks in-utero on the probability of being moderately or severely anemic as an adult which is associated with the main cause of maternal deaths. I also find evidence for a significant effect from positive rainfall shocks in-utero on the probability of delivering with a skilled-birth attendant, which is a key factor for improving maternal survival. This study’s main findings suggest that if better outcomes associated with positive rainfall shocks could be achieved, maternal mortality ratios would be 16% closer to the Millennium Development Goals for maternal health.

Paper 3: Effects of in-utero, early-life and childhood rainfall on women’s educational attainment: Evidence from sub-Saharan Africa

Children in sub-Saharan Africa have among the worst educational outcomes world-wide and significantly lag behind other regions of the world in terms of gender parity for educational attainment. This paper evaluates the role that early-life and childhood conditions play in influencing educational outcomes for women in sub-Saharan Africa. The methodological approach uses rainfall during the period when a woman was in-utero, and rainfall during her early years of life, as proxies for nutritional conditions during her early-life, and estimates their effects on educational attainment. The results show that early-life and childhood rainfall is predictive of educational outcomes. A 1 standard deviation increase in rainfall during the in-utero period, from ages 1-3, and from ages 6-9 increases education for women by 0.04 years, 0.04 years, and 0.05 years. There is also suggestive evidence that rainfall during the first year of life also affects educational attainment. Subsequent analyses fail to find a detectable effect from early-life rainfall on economic outcomes. However, there is strong evidence that women who experienced better conditions during the in-utero period have greater decision-making power within their household. In addition, women who experienced better conditions during the in-utero period or during ages 1-3 marry men with higher levels of education. Overall, improvements in early-life conditions have relatively small effects on women’s educational attainment and will not significantly help SSA countries achieve the MDG of universal primary school completion by 2015. However, the downward trend in rainfall in SSA over the last decades implies that worsening early-life conditions may decrease future potential progress towards improving women’s educational outcomes.