Author ORCID Identifier

Date of Award

Spring 5-8-2021

Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Dr. Shiferaw Gurmu

Second Advisor

Dr. Charles J. Courtemanche

Third Advisor

Dr. James H. Marton

Fourth Advisor

Dr. Michael F. Pesko


This dissertation consists of three essays. The overarching objectives are to provide causal evidence of the effects of health insurance program from Ghana, charter school in the United States (U.S.), and economic conditions in the U.S. on healthcare utilization, health behaviors and human capital development.

Chapter 1 uses the Demographic and Health Survey to study the healthcare utilization effects of Ghana’s 2004 adoption of a national health insurance scheme (NHIS), covering over 95% of medical expenditures. First, we find that self-reported participation in the NHIS increases twelve-month healthcare visits by 32 percentage points using the timing of the rollouts across districts as an instrument. We also show that the positive effect is larger for less-educated, poor, and rural women. Second, we find that the NHIS increases deliveries in health facilities and prenatal care visits by 6 and 7 percentage points, respectively, using a difference-in-differences strategy and women from rural Nigeria as the control group. Together, these findings are consistent with evidence from similar programs in developed countries despite numerous implementation challenges and relatively low take-up of the NHIS.

In Chapter 2, we examine the impacts of exposure to charter schools on students’ long-term outcomes using the restricted, geocoded National Longitudinal Survey of Youth data and school information from the National Center for Education Statistics. We use an instrumental variable method by constructing charter school exposure in the county of birth as an instrument for actual exposure in the county of residence. Our results suggest strong evidence of charter schools increasing four-year college completion and reducing adverse health behaviors. We also show evidence of heterogeneity. While college graduations are more pronounced among females and minorities, the reduction in binge drinking and cigarette smoking is higher among minorities and high-educated individuals. Overall, our results demonstrate that charter schools improve students’ long-term outcomes.

Finally, Chapter 3 uses individual-level data from the 1987{2019 Behavioral Risk Factor Surveillance System and state-level employment data from the Bureau of Labor Statistics to estimate the effects of macroeconomic conditions on cigarette smoking. We find that a one-point increase in the employment rate raises the current cigarette smoking rate by 0.4%. We also show heterogeneity among males, Blacks and low-educated individuals having larger impacts but no differential effects by age. A dynamic treatment effect analysis demonstrates that the procyclical relationship is declining and unstable over time, with a sharp temporal decrease during the Great Recession period and a weak countercyclical effect in 2019. Overall, the results suggest that future adverse macroeconomic shocks will increase cigarette smoking, contrary to studies that demonstrate healthy living in bad economic times.