Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department


Major Professor

Gabriel Picone, Ph.D.

Committee Member

Murat Munkin, Ph.D.

Committee Member

Xin Jin, Ph.D.

Committee Member

Hongdao Meng, Ph.D.


Body Weight, BMI, Income, Malaria, Education


Chapter 1 studies the effect of body weight on labor market outcomes. Using longitudinal data from the China Health and Nutrition Survey (CHNS) from 1989 to 2011, and both instrumental variable and fixed effects estimation to control for the endogeneity of body weight, I examine the wage effect of BMI by gender, type of residence area, and occupation. Results from linear OLS regression show a positive relationship between body weight and income for males in both rural and urban areas and females in rural areas, while a negative one for females in urban areas. After controlling for the unobserved individual time-invariant heterogeneity with fixed effects estimation, there is still a positive relationship for males in rural areas, which means that higher body weight is associated with higher income, and a one-unit increase in BMI increases log income by about 1.5%. For females in urban areas, there is some evidence of a wage penalty for being obese, but the results are not robust throughout all identifications. When breaking up by occupation, I find that male managers and male craftsmen experience wage penalties for being obese.

Furthermore, with a semiparametric estimation method, the conditional income functions are estimated graphically. The income – BMI functions peak at different BMI value for each sample and the nonlinear relationship differs across samples. Finally, in addition to the different effects between rural and urban areas, between males and females, across occupations and age groups, the effect over the whole income distribution for each sample is estimated with quantile regression. Results show significant differences between the mean effect and the effects at different income quantiles.

Chapter 2 examines the effect of anti-malaria campaigns on children’s educational outcomes. Malaria is one of the most life-threatening diseases in the world and presents an immense health burden in sub-Saharan Africa. Since the year 2000, international financing for malaria control has reduced malaria prevalence (measured by PfPR) significantly. At the same time, educational attainment has improved in many countries in the area. Previous literature shows that the eradication of malaria or a decrease in the prevalence level has improved children’s educational outcome. In this paper, we empirically test the effect of anti-malaria campaigns on educational outcomes for school-age children using data from the Demographic and Health Survey (DHS) combined with malaria prevalence and anti-malaria intervention data from the Malaria Atlas Project (MAP) for 14 sub-Saharan African countries from 2000 to 2015. Using a fixed effects model, the estimates from the main specification suggest that a decrease of 10% in malaria prevalence level leads to around a 1.6% increase in school attendance rate for students aged 10 to 14. In addition, younger children benefit more from the decrease in malaria prevalence. Alternative specification using interventions and pre-campaign malaria levels shows that the causal effect of ITN is significant on school attendance rate. The results correspond with other work in the literature showing the positive effect of disease eradication campaigns on the educational outcome for school-age children. Determinants of school attendance, such as quantity-quality trade-off, child labor, parents’ education, and gender differences are discussed, based on evidence from our results to enrich the literature.