Three Essays in Health Economic

This dissertation constitutes three separate essays in health economics. The first essay examines whether Medicare Part D led to changes in the health behaviors that are essential to manage chronic diseases. Using data from the Medical Expenditure Panel Survey, I find that the implementation of Medicare Part D reduces the probability of engaging in physical exercise. The effect on dieting is inconsistent across different specifications and the effect on cigarette smoking is not statistically significant. The negative physical exercise effect of Medicare Part D is more pronounced among patients with low educational attainment.;The second essay looks into whether early health shocks persist to cause health inequality across generations. Linking the Ethiopian Socioeconomic Survey with the 1984 Ethiopian Census, I show that in utero and early childhood( age 0-3) exposure to the 1983-85 Ethiopian famine increases the probability of stunting and reduces the height-for-age z-score of the next generation. Estimates that account for the fertility response, infant and fetal culling effects of famine indicate that the baseline estimate represents the lower bound of the total effect of the famine.;Linking a village-level interpolated rainfall data to a child-level longitudinal survey--- the Ethiopian Rural Household Survey--the final essay explores whether the child health effects of drought vary across child health distribution. The correlated random effect (CRE) quantile regression for panel data model estimates shows that negative mean deviation from district-level long-term average rainfall reduces the weight-for-height and weight-for-age z-scores of children at the lower end of the anthropometric distributions. Examination of the channel of transmission indicates that the incidence of drought increases the incidence of illness and reduces per-capita consumption expenditure, and the time spent on domestic work among poor households.