The three chapters of my dissertation study factors that contribute to the uneven distribution of economic activities across space. In the first chapter, I study why firms are more productive in larger cities, by focusing on a potential explanation first proposed by Adam Smith: Larger cities facilitate greater division of labor within firms. Using a dataset of Brazilian firms, I first document that division of labor is indeed robustly correlated with city size, controlling for firm size. I propose a theoretical model in which this relationship is generated by both a selection effect---firms endogenously sort across space, choosing different extents of division of labor---and a treatment effect---larger cities increase division of labor for all firms, by reducing the costs associated with greater division of labor. The model embeds a theory of firms' choice of the optimal division of labor in a spatial equilibrium model. Structural estimates derived from the model show that division of labor accounts for 16\% of the productivity advantage of larger cities in Brazil, half of which is due to firm sorting and the other half to the treatment effect of city size. The theory also generates a set of auxiliary predictions of firms' responses to a reduction in the cost of division of labor. Exploiting a quasi-experiment that changes the cost of division of labor within cities---the gradual roll-out of broadband internet infrastructure---I find causal empirical support for these predictions, validating the model. Finally, the quasi-experiment also provides out-of-sample validation for the structural estimation. The estimated model predicts changes in the average division of labor within different cities in response to the new broadband internet infrastructure, which I find are similar to the actual changes. The second chapter, co-authored with Ariel Burstein, Gordon Hanson and Jonathan Vogel, studies how occupation (or industry) tradability shapes local labor-market adjustment to immigration. Theoretically, we derive a simple condition under which the arrival of foreign-born labor into a region crowds native-born workers out of (or into) immigrant-intensive jobs, thus lowering (or raising) relative wages in these occupations, and explain why this process differs within tradable versus within nontradable activities. Using data for U.S. commuting zones over the period 1980 to 2012, we find that consistent with our theory a local influx of immigrants crowds out employment of native-born workers in more relative to less immigrant-intensive nontradable jobs, but has no such effect within tradable occupations. Further analysis of occupation labor payments is consistent with adjustment to immigration within tradables occurring more through changes in output (versus changes in prices) when compared to adjustment within nontradables, thus confirming our model's theoretical mechanism. We then use an extended quantitative model to interpret the magnitudes of our reduced-form estimates and to aggregate up the consequences of counterfactual changes in U.S. immigration from the region-occupation level to the region-level. The third chapter proposes a new channel through which improvements in transportation or communications technologies affect skill distribution across space. In this joint work with Yang Jiao, we start with the empirical observations that substantial skill and occupation relocation took place across U.S. cities during past decades. In particular, big cities attract more skilled workers and become more specialized in cognitive-intensive occupations. Motivated by empirical literature on the association between modern communications technology adoption and production fragmentation, we develop a spatial equilibrium model with domestic production fragmentation to analyze the impact of a reduction in the costs of cross-city production teams---e.g., communications cost---on spatial distribution of skills and economic activities. The model generates predictions consistent with the observed empirical patterns, including more spatial segregation of skilled and unskilled workers, and occupation specialization across U.S. cities over time. In contrast to findings in the international offshoring literature, in which there are winners and losers, we find that under regularities conditions, there are Pareto welfare gains for all agents with heterogeneous skills, together with a substantial measured labor productivity increase at the aggregate level.

This dissertation examines the role of information in influencing both individuals' college outcomes and the productivity of a higher education system. It focuses in particular on large-scale educational reforms that raise different mechanisms than those in the existing literature on the returns to college attendance and college quality. Recent work has shown that the choices of whether to attend college and which college to attend can both affect individuals' future earnings. These papers typically focus on a narrow subset of students or schools to credibly identify the effects of college choice. This dissertation instead uses data on the near universe of college students in an entire country to explore informational mechanisms that are difficult to isolate in existing work. To do this, I exploit reforms to the higher education system in Colombia that affect the information on individual ability that is transmitted to colleges, to employers, or to students themselves. This allows me to adapt traditional labor economic topics like employer learning (Jovanovic, 1979) and assortative matching (Becker, 1973) to the context of higher education. In addition, the large-scale nature of these reforms raises general equilibrium issues that may not arise from marginal changes in college admissions (e.g., Heckman, Lochner and Taber, 1998). In Chapter 1, "Assortative Matching and Complementarity in College Markets," I examine one type of assortative matching in college markets: students with high socioeconomic status (SES) are more likely to attend high quality colleges. Assortativity matters if SES and college quality are complementary educational inputs. I develop an econometric framework that provides tests for the existence and sign of this complementarity. I implement these tests by exploiting a 2000 reform of the national college admission exam in Colombia, which caused a market-wide reduction in assortative matching in some regions of the country. I find that the reform lowered average graduation rates and post-college earnings in affected regions, consistent with a positive complementarity between SES and college quality. I also find evidence of mismatch: part of these negative effects came from the low SES students who were shifted into higher quality colleges. However, both the market-wide and mismatch effects die out several cohorts after the exam reform, which suggests that complementarity may evolve with large-scale changes in assortativity. In Chapter 2, "The Big Sort: College Reputation and Labor Market Outcomes," W. Bentley MacLeod, Juan E. Saavedra, Miguel Urquiola, and I ask how college reputation affects the process by which students choose colleges and find their first jobs. We incorporate a simple definition of college reputation---graduates' mean admission scores---into a competitive labor market model. This generates a clear prediction: if employers use reputation to set wages, then the introduction of a new measure of individual skill will decrease the return to reputation. We confirm this prediction by exploiting a natural experiment from the introduction of a college exit exam in the country of Colombia. Finally, we show that college reputation is positively correlated with graduates' earnings growth, suggesting that reputation matters beyond signaling individual skill. Finally, in Chapter 3, "Time Gaps in Academic Careers," I ask if interruptions in students' academic careers can lower their overall schooling attainment. I study an academic calendar shift in Colombia that created a one semester time gap between high school and potential college entry. This brief gap reduced college enrollment rates relative to unaffected regions. Low SES students were more likely to forgo college, and individuals who did enroll after the gap chose higher paying majors. Thus academic time gaps can affect both the mean and the distribution of schooling attainment, with implications for the design of education systems and for wage inequality.

This paper uses the 2010 Mexican Population and Housing Survey to examine the role of regional return to education on migrant selection. The study uses a standard linear regression model to predict the educational attainment of migrants and compares it to the educational attainment of non-migrants in each Mexican State. It finds evidence of negative selection, that less educated Mexican citizens are more likely to migrate to the United States. It also finds little evidence of the impact of regional return to education on migrant selection. The study offers potential explanations for the lack of impact and suggests avenues for continued study.

This thesis aims to examine the impact of transfer market habits on top European soccer clubs through three measures of sporting success. Using 2013-2018 data from Transfermarkt.com, Deloitte, and Forbes, including data on teams from 9 different European countries, this paper will be used to enable regressions on UEFA competition performance, club valuation, and social media followers. Using these three linear regressions, this thesis will attempt to determine the effectiveness of transfer market spending in improving key metrics of club performance. This thesis ultimately suggests a positive relationship between transfer spending and success in UEFA competitions as well as increases in club valuation. The results in this paper also suggests that this increased success in Europe leads to greater brand value through a significant social media impact.

Workplace sexual harassment costs the government and companies millions of dollars a year. Women who experience sexual harassment in the workplace suffer from negative mental and physical health problems, lower career attainment, decreased productivity, and a higher rate of job turnover. Sexual harassment is both costly and unjust, however the exact cost to women who experience sexual harassment is unknown. This thesis will measure the impact of workplace sexual harassment on wages in different industries. Using data on claims filed with the Equal Employment Opportunity Commission, I calculate and analyze the impact of sexual harassment on wages, age, sex, and industry. I find that industries with high rates of women reporting sexual harassment have lower wages.

This thesis focuses on how health care policies affect the labor supply of physicians and beneficiaries. Further, I examine how the labor supply responses of physicians vary based on the level of competition. In the first chapter, I focus on the labor supply response of physicians to two large public health insurance expansions, the State Children’s Health Insurance Program (SCHIP) and the Affordable Care Act (ACA). These insurance programs have significantly increased the number of patients with public health insurance and the demand for medical services, but it is not clear whether providers will supply additional services for newly-insured patients. In response to the introduction of SCHIP, my estimates suggest that physicians reallocate their total working hours between patient care and non-patient care activities. The size of the impact was greater in areas with high level of physician concentration prior to the expansion. Physicians in high concentration areas tend to decrease time spent on direct patient care, but increase hours on non-direct patient care. In response to the ACA, physicians’ working hours did not increase, but working hours and the probability of being employed increased for registered nurses. This suggests that physicians might utilize other healthcare providers to accommodate increases in demand for medical services after the expansion. In the second chapter, we analyzed the impact of expanding Medicaid on health insurance coverage and labor market outcomes. Expansions of public health insurance have the potential to reduce the uninsured rate, but also to reduce coverage through employer-sponsored insurance (ESI), reduce labor supply, and increase job mobility. In January 2014, twenty-five states expanded Medicaid as part of the Affordable Care Act to low-income parents and childless adults. We compare the changes in insurance coverage and labor market outcomes over time of adults in states that expanded Medicaid and in states that did not. Our estimates suggest that the recent expansion significantly increased Medicaid coverage with little decrease in ESI. Overall, the expansion did not impact labor market outcomes, including labor force participation, employment, and hours worked. In the third chapter, I examined the impact of competition among dentists on the labor supply of dentists. I focus on how dentists’ working hours will changes when the level of competition increases by examining the effect of the National Health Service Corps (NHSC). The NHSC was created to increase the supply of rural physicians, which might increase the competition in rural areas. I examine the number of dentists (extensive margins of labor supply) and the change in the working hours of dentists (intensive margins of labor supply) in response to the increased level of physician competition. I found that 1 percent increase in NHSC-approved sites increases 5.4% increases in the number of providers and 0.2% of competition in a rural county. In addition, I found that there is a positive relationship between the number of NHSC-approved sites and providers’ working hours. If the competition among dentists increases about 1, then working hours of providers increase about 6 hours per week.

Previous studies have determined that teachers with strong subject backgrounds in math or science have a positive effect on student achievement. Using data from the American Community Surveys, I find that nation-wide, teachers who studied math or science in their undergraduate degrees receive a roughly 1% increase in salary over teachers that studied other subjects. I find that private schools do not reward teachers with a math or science background more than public schools do, but that medium-poor states as a group reward teachers with math or science backgrounds while richer states do not.

Income variability explains a significant fraction of the increase in annual income inequality. Chapter 1 considers the impact of variability on tax unit inequality. Using income tax return panel data, I estimate that between a tenth and a quarter of the increase in top one percent income shares between the early 1980s and 2000s was caused by variability. Increased income variability over this period resulted from mean-reverting fluctuations in the bottom quintile and top one percent. Variability in the top of the distribution seems partly driven by permanent income shifting in response to the Tax Reform Act of 1986. Chapter 2 examines the individual earnings distribution. Using Social Security Administration earnings panel data, I estimate that variability explains half of the increase in annual inequality in the bottom half of the distribution between 1973 and 1985. When workers with years of zero earnings are included, increasing earnings variability explains almost all of this group's increase in inequality. The increase in earnings variability appears to be explained by an increased fraction of working age men with years of zero earnings. Annual individual earnings inequality in the bottom half of the distribution not only increased with variability in the 1970s and 1980s, but also fell with variability in the 1950s and early 1960s. This suggests that the U-shaped trend in income inequality observed over these four decades was partly caused by first a fall and then a rise in earnings variability. Between 1985 and 2000, falling variability caused most of the decline in annual earnings inequality within the bottom half of the distribution. Within the top of the distribution, earnings inequality increased over this period because of changes in permanent earnings and not increasing variability. Income variability means that in a progressive tax system annual and lifetime federal tax rates can diverge. Chapter 3 shows that on an annual basis, those at the bottom of the distribution pay little or no federal income taxes, while on a lifetime basis they pay average tax rates about five percentage points higher. Income variability also means there is a trade-off between vertical and horizontal equity.

<p> In this dissertation I present the first county-level estimates of deaths in the Confederate Army for eight of the former Confederate States (Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, and Virginia). As described in Chapter 2, I estimate the number of deaths by Confederate company (a unit of roughly 100 men) and map these back to the company's county of origin. Counties' death rates were driven by the battles in which their men fought, determined by generals for strategic reasons. This produces a wide distribution in county-level death rates, and it allows for causal inference in assessing the impacts of these losses on counties' later development.</p><p> In Chapter 3, I estimate the long-run effects of population loss on the economic geography of the South. Populations in counties with higher death rates caught up to neighboring areas within 15 years after the war, but then they kept growing. These increases were caused by migration, especially by African Americans: counties with ten percentage-point higher death rates had 14% larger black populations in 1900 and 27% larger in 1960. Migrants also increasingly went to counties that were less advantaged in Southern economy before the Civil War. The economic geography of the American South was thus changed significantly after the institutional shock from the Civil War. </p><p> In Chapter 4, I estimate the effects of relative labor scarcity on racial violence and political participation in the American South from 1865 to 1900. I find counties with 10 percentage-point higher death rates in the Civil War had 24-33% fewer lynchings of African Americans from 1866 to 1900. They also had 3.6-5.6% higher voter turnout despite a larger fraction of their population being black. These effects persisted for at least two decades after the counties' relative labor scarcity disappeared. However, in the very long run (100 years), counties with greater Civil War deaths saw a reversal, with much worse discrimination by the Civil Rights Era, likely due to their larger black populations and absence of economic incentives to prevent discrimination. This suggests relative levels of discrimination were not culturally determined and can change fairly quickly.</p>

This dissertation contains three chapters that examine the effect of price in higher education. The first chapter considers the effect of community college tuition on college enrollment using a natural experiment in Texas where discounts for community college tuition were expanded over time and across geography. Additionally, the long-term effects of community college are examined including transfer to universities and graduation with a bachelor's degree. This chapter uses Texas administrative data from 1994-2012 on the universe of high school graduates and their college enrollment and graduation. For high school graduates, community college enrollment in the first year after high school increased by 7.1 percentage points for a \$1,000 decrease in tuition. Lower tuition also increased transfer from community colleges to universities. There is also marginally statistically significant evidence that attending a community college increased the probability of earning a bachelors degree within eight years of high school graduation by 23 percentage points. The second chapter examines whether students respond to immediate financial incentives when choosing their college major. From 2006-07 to 2010-11, low-income students in technical or foreign language majors could receive up to \$8,000 in Federal Science and Mathematics Access to Retain Talent (SMART) Grants. Since income-eligibility was determined using a strict threshold, this chapter determines the causal impact of the grant on student major with a regression discontinuity design. Using administrative data from public universities in Texas, it is estimated that income-eligible students were 3.2 percentage points more likely than their ineligible peers to major in targeted fields. Brigham Young University had a larger impact of 10.1 percentage points. The third chapter considers the effect of financial aid arising from students being declared financially independent on educational outcomes including reenrollment, credits attempted, and graduation. Students who are 24 at the end of the calender year cannot be declared dependent while students who are 23 at the end of the year can be. This sharp change in eligibility is leveraged to compare dependent students to independent students in a regression discontinuity framework. The analysis uses administrative data from from all public universities and colleges in Texas from 2003-04 to 2013-14. Financial independence is associated with modest changes in educational outcomes. / text