Semester

Summer

Date of Graduation

2023

Document Type

Dissertation

Degree Type

PhD

College

Chambers College of Business and Economics

Department

Economics

Committee Chair

Josh Hall

Committee Co-Chair

Brad Humphreys

Committee Member

Bryan McCannon

Committee Member

Christopher Carpenter

Abstract

The first chapter examines the role that same-sex marriage legalization had on the number of adoptions of children from foster care in the United States. We do so by employing a synthetic difference-in-differences estimator which leverages both the differential timing of these laws across states and the subsequent wave of state-level legal protections which give foster-care agencies the right to deny service to same-sex couples based on religiously-held beliefs. Using highly detailed, county-level data of nearly 20 million children in the foster care system from 1995-2020, our findings reveal that same-sex marriage legalization led to a 3.8%-5.9% increase in the annual number of adoptions. We show that this is driven by an asymmetric substitution in the composition of adoptive household types away from unmarried couples and single women and towards recognized families. Back-of-the-envelope calculations suggest that 1) same-sex marriage indirectly caused an additional 10,000- 17,500 additional adoptions within an average four-year time window after its passage and 2) that there would have been 2,900-4,500 more adoptions from 2016 to 2019 without laws which grant foster-care agencies the right to refuse service to same-sex couples.

The second chapter examines the role that teacher union strength played in the reopening decisions of public schools in Connecticut throughout the ’20-’21 academic school year. To do so, I construct a measure of union strength that combines the number of pages in a union’s collective bargaining agreement with a text analysis that measures the expansiveness of their contract’s leave policies. I find that school districts with strong unions were less likely to offer in-person classes throughout the 2020-2021 academic year. However, this effect becomes insignificant when accounting f or t he local incidence of COVID-19. Instead, I find that differences in the vote share for the Democratic presidential nominee of 2020 and the number of COVID-19 cases were stronger predictors of reopening decisions. The results contradict the consensus within the literature that unions and political attitudes prevented schools from reopening. These results suggest that using more geographically granular data and a longer time horizon to define reopening decisions describes the differences in results.

The third chapter examines the effect of negative health shocks on the retirement decisions of peers. I do so by gathering detailed data on players from the National Football League (NFL) from 1980-2019. The introduction and advancement of research into chronic traumatic encephalopathy (CTE), coupled with increasing early-life mortality among NFL players, has significantly increased the salience of American football’s long-term negative health effects. I exploit the quasi-random timing of early-life, CTE-related mortality of both former and current players on the retirement decisions of their teammates. I demonstrate that these players are significantly more likely to retire than comparable peers in the seasons following one of these deaths. This magnitude of this effect larger for players with fewer years of college attendance, but is uncorrelated with the amount of time shared as teammates.

Download

Share

COinS