Essays on the economics of health insurance, labor markets, and migration

Abstract

This dissertation contains three chapters, two which focus on health insurance and one focusing on migration. The first chapter examines how a policy expanding public health insurance for young children affected their parents' labor market and health insurance outcomes. I use variation in the initial income thresholds, children's age cutoffs and timing of implementation across states to estimate the effect of a person's youngest child gaining access to public health insurance on self-employment. I find that having a child become Medicaid eligible increases a father's self-employment and increases his business income. I find no significant effect on self-employment for mothers, but I find that the increasing eligibility is associated with a large negative effect on their probability of remaining in a wage job. The second chapter examines how expanding dependent health insurance for young adults affects the health insurance and labor market outcomes of those young adults and their parents. I exploit two sources of variations in the age at which young adults age out of their parents' health insurance: i) state reforms passed between 2000 and 2010 that extended the maximum age of health insurance dependents beyond 18 and ii) the Affordable Care Act that extended coverage for all young adults in the United States until their 26th birthdays. Using regression discontinuity, I find evidence that the policies increased young adult dependent coverage. Dependent coverage for eligible young adults increased by 8 percentage points over ineligible young adults, while health insurance in the young adults' own name decreased by 6.5 percentage points. I also see evidence that parents of eligible young adults responded by changing their own coverage. The final chapter investigates the relationship between children and migration using data from the American Communities Survey. To address the issue that both migration and fertility might be correlated with unobserved variables I use twin births as an instrumental variable for the number of children. I find that that an additional child decreases migration by 0.6 percentage points and decreases the probability that a woman lives in her birth state by 1.4 percentage points. This suggests that more children hinder migration.