Essays in Public Policies and Health Economics.

다국어 초록 (Multilingual Abstract)

This dissertation encompasses three chapters that are centered around the relationship between different transformative public policies and their - often unintended - impact on healthcare and socioeconomic conditions, both in the U.S. and internationally. All three articles use detailed administrative and restricted-use data and employ rigorous identification strategies to answer the questions of interest. The major interventions analyzed cover in-kind transfers in the form of land restitution, school closures during the COVID-19 pandemic, and access to healthcare when insurance companies are terminated.Chapter 1 explores the direct and indirect effects of the land restitution policy in Colombia on mental health conditions and wellbeing. The indirect effects (spillovers) on depression, anxiety and substance abuse are estimated at the municipality level exploiting the comprehensive intervention from the government that protects the land of victims who have been displaced and dispossessed by illegal armed groups in a difference-in-differences strategy. I show that the restoration of property rights is associated with a reduction in overall mental health conditions. A decrease in conflict, more productive investment, and coverage of the health system suggest that the interventions to secure property rights boost the trust and expectations of citizens during times of war. The direct effects take advantage of the legal sentences that restitution judges issue and restore property rights at the individual/household level. Using administrative data, I examine how relying on welfare, poverty and healthcare change for beneficiaries who have enjoyed their returned property the longest.Chapter 2, joint with Sumedha Gupta and Kosali Simon, looks at how parental mental health in the US was disrupted once schools closed during the early and later phases of the COVID-19 pandemic. Using a large medical claims dataset that covers around 20% of the commercially insured population nationwide, leveraging variation from cohorts of parents more and less exposed to the pandemic and computing the relative change in mobility to schools (elementary, middle and high) at the zip code level, we find an increase in maternal and paternal antidepressant use in zip codes with above median school closures. The positive effect on the demand for anti-depressant prescriptions is persistent over time after Fall 2020, even as in-person schooling resumed in 2021. We do not see changes in the demand for therapy, worsened conditions for mothers with previous antidepressant use or effects on mothers of middle/high school-aged children, but we do find positive effects from elementary school closures, which is a sign of how women during this economic downturn bore a disproportionate burden with increased housework and care hours, especially when it came to young children. Overall, the results suggest that the school system plays an important role in maintaining population mental well-being and in helping families cope with stress.Chapter 3 investigates the association between the termination of the largest health insurance company in Colombia, SaludCoop, and access to mental health services. After years of investigations and poor management, the government decided to terminate SaludCoop due to the mishandling of health resources by its board of directors. The government decided then to transfer the nearly 5 million SaludCoop beneficiaries (10% of the national covered population) to a smaller insurer called Cafesalud, which by that time had less than 5% of the market. I use several administrative data sources that record every contact that citizens have with the healthcare system, in addition to vital statistics, and leverage variation from the temporal and geographical presence of the insurer in a difference-in-differences empirical strategy. I find a decrease in the overall utilization of mental health services and consultations once SaludCoop exits the insurer market, at the same time that mortality linked to alcohol increases. I find these effects are persistent over time and are likely explained by an increase in the congestion of the health system, as the remaining insurers experienced greater demand from beneficiaries.