Honors Theses 2023

This paper studies the effects of eviction policies during Covid-19 on the rental market. I exploit state-level variation in eviction moratoria to compare trends in eviction filings and rents by moratorium status. I use two main difference-in-differences specifications: the first is a two-way fixed effect model where treatment varies by month and the second considers moratorium status as constant and compares the differential yearly trends across treatment and control states. I find that moratoria are associated with a 30-40% decrease in eviction filings and a 1-2% increase in rents in the average US county. However, there were no effects on rents in the most populous counties, and population-weighted regressions find that the average American lived in a county where rents did not increase in response to eviction moratoria. Considering a main motivation for these policies was to alleviate financial hardship, increased rents could counter the intended effects. However, this paper finds that the average renter was not affected by rent increases associated with these policies.

​This paper examines labor market dynamics in the post-2008 era. A period characterized by sluggish earnings growth, this paper's primary objective is to uncover metrics that help explain wage dynamics following the Great Recession. First an examination of existing literature validates that the unemployment rate is not key to understanding wage formation in the U.S. Instead, the underemployment rate, the rate of non-employment, and the inactivity rate are shown to be significant predictors of labor market slack over the time horizon of 2008-2022. There is no role for vacancies in the wage formation process. These findings hold in panel data with state and year fixed effects and are supportive of a wage curve that fits the data considerably better than a Phillips Curve. Imperfect imputed wage estimates also constitute a growing percentage of earnings data reported by the Bureau of Labor Statistics. After accounting for this measurement error, and using alternative earnings data from the ASEC March supplement, significant slack estimators become more powerful. Throughout this empirical analysis, the impact of various labor market metrics are compared to draw conclusions about the state of the macroeconomy and NAIRU post-2008.

I study the decline of neonatal immunizations of girls relative to boys over COVID- 19. Using data from the National Family Health Survey, I conduct a difference-in- differences methodology to find that the full immunizations fell by 2.11 percentage points for female infants over the pandemic relative to male infants. In 2015-'16, this gap was close to zero, with the average immunization rate of boys and girls being around 62%. I investigate the mechanism for this effect, and find that disparities in healthcare infrastructure did not cause these losses. Instead, I find that the losses for girls relative to boys were concentrated in households in high COVID-19 caseload dis- tricts. Further, underlying gender norms corroborated the losses in immunizations for girls relative to boys. I found that losses in immunizations were greatest in households in high son preference districts and wealthier households. Thus, my study provides compelling evidence that reductions in gender disparities in immunizations are sensi- tive to shocks, such as COVID-19. These results demonstrate that efforts to improve aggregate immunization access are not enough to eliminate the underlying household tendency to under-invest in female infants. Policymakers must focus on eliminating household son-preference to achieve equitable immunization coverage for all infants.

This paper studies how geography shapes the evolution of countries comparative advantages. It uses a gravity-based framework to generate productivities for the exporting country in a bilateral trading pair for a product at a given time, as well as derive potential drivers in the changes of the exporting country's comparative advantages (Costinot et. al 2012). The drivers explored are foreign demand and productivity measures interacted with a time-varying weight measure which inversely captures distances between countries - known as "dynamic" demand and learning variables. To identify the regression, I utilize air and sea distance measures to create instruments for the dynamic demand and learning variables inspired by Feyrer (2019). Geography plays a significant role in shaping countries' comparative advantages through the dynamic foreign demand variable, leading to countries specializing less in a product as their trading partner exogenously moves closer over time. This effect is heterogenous across several characteristics including country size, wealth, and product industry.

The uneven split of childcare responsibilities between men and women is often cited as a reason for the persistent gender earnings gap. I study the impact of primary school enrollment on mothers' and fathers' labor supply and investigate how this effect has changed since 1980. I use the quarter of birth of the parents' youngest child to generate exogeneous variation in that child's school enrollment. Due to primary school birth date cutoffs, some of the 5- year-olds in my samples are eligible to enroll in primary school while others are not. These cutoffs create variation in their primary school enrollment rates by birth quarter. Using data from 2006-2019, I find that as an average mother's youngest child is enrolled in primary school, her employment and labor force participation increases by 2-3 percentage points, and she works an additional hour per week. There is no significant effect on fathers, and I estimate that primary school closes the labor supply gap between mothers and fathers by anywhere from 1% to 11%. This paper also includes results using 1980 data. The effect on mothers has fallen since the 1980s, which corroborates estimates of falling child penalties in the literature. I estimate that in 1980, primary school increased mothers' employment and labor force participation by closer to 3-4 percentage points and weekly hours worked by approximately an hour and a half. I investigate the decreasing impact of primary schooling on mothers with a heterogeneity analysis that uses state-level variation in gender norms and public pre-kindergarten enrollment rates. The results indicate that converging gender norms and expanded access to public pre-K likely explain a substantial portion of the fall in effect over time. Therefore, this paper suggests that expanded public pre-K would likely increase the labor supply of mothers of preschool aged children, which could help reduce the gender-based labor supply gap earlier.

Since becoming a representative democracy, Nigerian presidential elections have been allegedly marred by various forms of malpractice. The persistence of self-serving leaders, whose main goal is to enrich themselves often serves as the basis of these allegations. Various news channels, social media posts, and online blogs repeatedly allege that state governors and the Independent National Electoral Commission (INEC) play roles in this malpractice. In this paper, I evaluate these claims using two hypotheses. The first is that state governors manipulate the presidential electoral process in favor of the presidential candidate of the same political party affiliation as them. To evaluate this claim from a causal analysis standpoint, I make use of a regression discontinuity. The second hypothesis is that the INEC is partisan, either in favor of a political party or region of the country. To evaluate this claim, I make use of a box and whisker plot analysis. In terms of results, I fail to reject the first hypothesis. Therefore, state governors likely play a causal role in electoral malpractice in Nigerian presidential elections. For the second hypothesis, I neither reject nor fail to reject the hypothesis as the box and whisker plot analysis was not sufficient to establish a causal relationship. However, the results of this analysis strongly suggests that the INEC manipulates election results. Therefore, they are likely partisan. I then propose two key policy recommendations to help address the state of electoral malpractice in Nigerian presidential elections.