Search

This dissertation tries to answer some important questions in labor and transportation economics in an attempt to broaden understanding of different topics.In the first chapter “Correlated Labor Market Risk and Housing Investment” I show that households have lower levels of housing investment when they live in areas with labor markets that are more correlated with their industry of employment. I find that if a household lives in an area where many other households work in the same or similar industries, then housing may be a riskier asset as it is more correlated with labor market income. In response household decrease their investment in housing, and this This decline is driven by concentrations and riskiness of other correlated industries, suggesting agglomeration in one industry can have negative spillovers to workers of other related industries.The second chapter “Loyalty rewards and redemption behavior: Stylized factsfor the U.S. airline industry” joint with Alexander Luttmann, provides a novel identification method for frequent flyer tickets in a comment industry database, DB1B. Using the FFAs we identify, we show how the characteristics of award tickets differ from paid tickets and how these characteristics have changed over time. We then demonstrate how various market and product quality characteristics influence the share of passengers traveling on FFAs. Finally, we find that price dispersion increases on routes with larger shares of frequent flyer passengers, implying that airline loyalty programs enhance market power.In the third chapter, “Help Really Wanted? The Impact of Age Stereotypes in Job Ads on Applications from Older Workers” joint with Ian Burn, Daniel Firoozi and David Neumark, We construct job ads for administrative assistant, retail, and security guard jobs, using language from real job ads collected in a prior large-scale correspondence study. We modify the job-ad language to randomly vary whether the job ad includes ageist language regarding age-related stereotypes. In contrast to a correspondence study in which job searchers are artificial and researchers study the responses of real employers, in our research the job ads are artificial and we study the responses of real job searchers. We find that job-ad language related to ageist stereotypes, even when the language is not blatantly or specifically age-related, deters older workers from applying for jobs. The change in the age distribution of applicants is large, with significant declines in the average and median age, the 75th percentile of the age distribution, and the share of applicants over 40. Based on these estimates and those from the correspondence study, and the fact that we use real-world ageist job-ad language, we conclude that job-ad language that deters older workers from applying for jobs can have roughly as large an impact on hiring of older workers as direct age discrimination in hiring.

This dissertation examines interactions between the unemployment insurance system, the tax system, and the labor market in the United States. It focuses primarily on how these major government policies affect individual decisions, particularly viewed through households' financial constraints and available liquidity. The dissertation conducts several novel reassessments of how individuals respond to key components of the tax and social insurance system, generating new economic evidence to inform potential policy reform. The chapters offer both causal and descriptive evidence and build on previous economic research in the fields of public economics, labor economics, and household finance. Overall, the dissertation highlights the importance of using well-identified micro-level natural experiments for evaluating the impacts of government policy.

This dissertation empirically explores different aspects of the market design in centralized school choice.The first chapter studies the consequences of limited information among school choice participants, and the influence of outside options in application and enrollment decisions. 23% of the Chilean applicants who receive an offer choose to enroll elsewhere, unnecessarily blocking seats that would improve the allocation for 12% of the placed applicants and offer placement to 11% of the non-placed students. Based on a model of the joint decision of school choice and enrollment, I show that imperfect information translates into penalization on the valuation of the schools, affecting application and search behavior and decreasing the probability of enrollment. Concurrently, greater availability of outside options diminishes the incentive for search and lowers the cost of rejecting placement offers. The counterfactual analysis highlights the effect of different information campaigns and the inclusion of outside options in the centralized system, underscoring the importance of after-market design in centralized school choice systems.The second chapter shows that beliefs about admissions chances shape choice outcomes even when the school choice assignment mechanism is strategyproof. Data from a large-scale survey of choice participants in Chile shows that learning about schools is hard, that beliefs about admissions chances guide the decision to stop searching, and that applicants systematically underestimate non-placement risk. We then use RCT and RD research designs to evaluate scaled live feedback policies. 22% of applicants submitting applications where risks of non-placement are high respond to warnings by adding schools to their lists, reducing non-placement risk by 58%. The third chapter evaluates how new information influences families' applications and assignment outcomes in elementary school choice settings. Specifically, using a multi-country RCT based in Tacna, Peru and Manta, Ecuador, we examine the effect of providing personalized information on schooling alternatives and placement risk. We find that applicants who received feedback on placement risk and a suggestion of new schools added more schools to their applications and were more likely to include recommended schools than other alternatives.

The central theme of this dissertation is the investigation of information frictions on both the demand and supply sides of the labor market. On the demand side, firms must select among prospective workers with many initially unobservable characteristics, such as ability or work ethic. In the first chapter of this dissertation, I propose a new interpretation for the canonical Abowd, Kramarz, and Margolis (1999) (AKM) wage decomposition as reflecting market beliefs about workers. I leverage this decomposition to measure the differences among firms in their ability or inclination to screen workers and explore how such heterogeneity shapes market-level learning about workers. In the second chapter, I extend my analysis of AKM decompositions, with an application to unpacking the well-documented 'wage scarring' effect of recessions for labor market entrants. The third chapter flips the focus on prospective workers (graduating college students), their (mis)perceptions about the labor market, and consequences thereof. We provide causal evidence that misperceptions affect job search behavior and final wages using a randomized controlled trial (RCT) in which corrective information about past graduate outcomes is provided to some subjects.

This dissertation contains three chapters in labor economics and applied microeconomics.In Chapter 1, I provide the first empirical evidence on the effect of having a female officer in the primary unit on the discovery of domestic violence (DV) in physical abuse incidents. DV is a crime that usually has female victims and is often under-reported. This chapter studies the effect of having a female officer dispatched in the primary unit on the discovery of DV in physical abuse incidents in Milwaukee and Chicago. Using three-year calls for service data and conditional random assignment of officers in the dispatch process, the chapter finds that the existence of a female officer in the primary unit scales up the likelihood of discovering DV in physical abuse incidents by 10% in Milwaukee. Analysis of data from Chicago indicates a similar effect. These results indicate that female officers play an important role in discovering DV. In Chapter 2, we study the effect of college basketball game days on sexual assault. Basketball games are an important part of college identity and social activities. This chapter studies the effect of college basketball game days on the probability of having local sexual assault reports. Using crime data from universities with top basketball programs and local law enforcement agencies, this chapter shows that home game days have little effect on the probability of sexual assault reports, while away game days scale up the probability by 14%. This finding is different from those found for football, which likely reflects differences in viewing and partying behavior across the two sports. In Chapter 3, I study the relationship between school lunch and nutrition. School lunch is an important channel of students' nutrition intake. This chapter studies the potential mechanisms for effects of the Healthy, Hunger-Free Kids Act using metabolic analysis. Using data from the National Health and Nutrition Examination Survey, the chapter finds that this policy decreased the probability of having high total cholesterol by about 30% for students who usually eat school lunches. This result is largely driven by the significant decrease in the proportion of students with high LDL cholesterol and triglycerides. The results reported in the chapter show the powerful impact of this policy on adolescent health.

In this dissertation, I delve into the impact of both local and national shocks on California's labor market and unemployment insurance (UI) program. Initially, I analyze the consequences of mass layoffs, which serve as a local employment shock, on neighboring firms. Later, I shift my focus to the effect of the COVID-19 shock on California workers, specifically those who receive UI benefits, and how UI assisted these displaced workers.In the first chapter, I examine the spillover effects of mass layoffs on neighboring establishments, shedding light on the dynamics of agglomeration economies. I employ a difference-in-differences event study framework and leverage comprehensive administrative data encompassing all entities in California to study the indirect effects of mass layoffs on employment, earnings, and the number of nearby establishments. I exploit the geographic coordinates of establishments to define treatment and control areas based on their proximity to instances of mass layoffs. The findings reveal persistent and negative spillover effects on local employment levels, payroll, and the number of operating establishments four years after the events. However, there is no significant change in the average earnings of workers. Moreover, empirical evidence demonstrates that the spillover effects diminish with increasing spatial distance, effectively disappearing after 6km. In summary, a one percent employment shock results in a one percent indirect decrease in employment levels within a 6km radius four years later. Furthermore, I contribute to the literature on agglomeration economies by assessing the mechanism of agglomeration economies in the observed spillover effects. I use economic distance measures to show the importance of industry linkages, knowledge spillover, and thick labor market as forces behind the spillover effects. My findings show that industries closely tied to the industry of mass layoff establishments experienced a more substantial employment decline, while those economically distant from the events show minimal changes.In the second chapter, which is joint work with Alex Bell, T.J. Hedin, PeterMannino, Geoffrey Schnorr, and Till von Wachter, we answer this question: To what extent did jobless Americans benefit from unemployment insurance (UI) during the COVID-19 pandemic? We document geographic disparities in access to UI during 2020. We leverage aggregated and individual-level claims data to perform an integrated analysis across four measures of access to UI. In addition to the traditional UI recipiency rate, we construct rates of application among the unemployed, rates of first payment among applicants, and exhaustion rates among paid claimants. Through correlations across California counties and across states, we show that areas with more disadvantaged residents had less access to UI during the pandemic. Although these disparities are large in magnitude, cross-state analysis suggests that policy can play a salient role in mitigating them.In the third chapter, coauthored with Alex Bell, T.J. Hedin, Peter Mannino, Carl Romer, Geoffrey Schnorr, and Till von Wachter, we leverage California's administrative longitudinal UI data to introduce two cumulative measures of the labor market health and use them to assess the impact of the COVID-19 pandemic on California's labor market and UI system. First, on the extensive margin, we measure the share of the pre-crisis labor force that applied for UI benefits. Second, on the intensive margin, we calculate the share of UI claimants who have received more than 26 weeks of unemployment benefits in the first year of the crisis. By combining the two measures, we show that the average member of the labor force spent nearly two months receiving regular UI benefits during the first year of the COVID-19 pandemic. Finally, we look into the demographic disparities in receiving UI benefits and show that more vulnerable workers experienced more weeks on UI than the more privileged.

This dissertation consists of three chapters that study questions on take-up and impacts of programs that increase the welfare of vulnerable populations. In chapter 1, we estimate income effects on earnings using the 2021 expanded Child Tax Credit (CTC). Children born in January 2016 were eligible for a credit of $3,600, while children born in December 2015 were eligible for $3,000. We use tax data and a regression discontinuity design to estimate the earnings response to this temporary, one-year $600 increase in non-labor income. We find that employment decreased by 2.0% in 2021 and 1.1% in 2022 in low-income families, with no effect on families with higher incomes. This decrease in wage employment was offset by an increase in self-employment reported on tax returns, but this result is not robust in all specifications.In chapter 2, we experimentally test whether information frictions and costs of filing taxes explain low take-up of the expanded CTC. We designed an information and counseling intervention contacting parents through school districts, coupled with assistance from local VITA (Volunteer Income Tax Assistance) programs. According to tax records, treatment increased tax filing by 2 percentage points, increased CTC claims by (a statistically insignificant) 2 percentage points, and tripled filing through VITA. Effects were concentrated among previous non-filers and Spanish speakers, demonstrating the value of targeted outreach.In chapter 3, we provide the first causal evidence on reducing forced labor risk. 18 million people are victims of labor exploitation globally every year. In the Indian construction sector, the most common indicators include excessive hours, working on rest days, and unpaid wages. Micro-contractors (MCs), who directly control working conditions, often face financial challenges, primarily due to limited access to working capital. This paper tests if easing MC's liquidity constraints can reduce forced labor and improve working conditions. 250 MCs in Bengaluru and Delhi were randomly offered low-cost loans. Overall, working conditions did not improve 6-9 months after loans were disbursed, but we find significant variability: workers of more educated and non-migrant MCs experienced improvements in their working conditions, including timely payments and shorter working hours.

This dissertation investigates inequality-related issues across three chapters, focusing on social media activism and disparate impacts of policies on historically marginalized groups. Chapter 1 provides the first causal evidence of slacktivism, the phenomenon whereby visible, low-effort forms of support on social media deter more vital, higher-cost actions. Using a laboratory experiment, I show that subjects who send –or “post”– a digital message to peers stating ‘I support racial justice’ are less likely to donate to related charities than those who could not publicize their support. While people believe their post is helpful, posts do not encourage others to give, and perceived effectiveness of posting is not driving results. Rather, self-interested biases explain behavior. Multiple mitigation strategies are unsuccessful, demonstrating the persistence of slacktivism.Chapter 2 shows that retaliation disproportionately disadvantages women on peer-to-peer review platforms. I leverage an exogenous policy change implemented by Airbnb in 2014 which made reviews simultaneous reveal, preventing users from seeing each other’s review before posting their own. The policy enhanced honesty by removing the ability to retaliate to negative reviews and reciprocate extremely positive ones. Analysis of review data reveals that while both male and female guests wrote less positively worded reviews, the negative shift was more pronounced towards male hosts. This indicates that reviews for male hosts were artificially elevated due to fear of retaliation, confirming that review systems with retaliatory capacity exhibit gender biases.Chapter 3 explores whether we can narrow application gaps and promote diversity by modifying language around qualification requirements in job ads. We established a non-profit company that acts as an intermediary in the job search process to conduct a large-scale, reverse audit study field experiment where we randomize the content of real job ads. We vary whether job seekers are encouraged to apply even if they don’t meet all the listed qualifications and whether they are informed that companies routinely hire individuals who do not have all qualifications. Preliminary results show this intervention encourages applications, and we hypothesize that it will have larger impacts on women and individuals from underrepresented racial groups by changing perceptions of the hiring process.

Policies are implemented targeting specific objectives, but there are both intended and unintended consequences that have been studied in applied economics. In my dissertation, I study how firms' production choices interact with policies in various contexts. These choices are especially related to drug affordability, socioeconomic equity, and environmental concerns.Chapter 1 studies the Indian import penetration in the U.S. pharmaceutical industry. After India amended its patent protection law in 2005, Indian pharmaceutical firms increased generic drug exports to Western countries, especially the United States. Drug exports were stronger for drugs that contain active pharmaceutical ingredients, which Indian pharmaceutical firms had manufacturing capabilities for before 2005. These 'treated' drugs experienced a larger price drop and increased competition from Indian firms. Then, I show that non-Indian firms exited the market for 'treated' drugs through more discontinuation and fewer new registrations of 'treated' drugs. This strategic change was stronger when a firm had larger revenue shares exposed to the trade shock. Non-Indian firms chose to keep selling drugs they already had experience producing rather than exploring new active pharmaceutical ingredients. The Indian law change also resulted in foreign supply chain disruptions, with 'treated' drugs facing more frequent shortages in the market. This chapter studies how foreign supply shocks affect competition in the pharmaceutical industry and how trade-driven competition could crowd out incumbent firms.Chapter 2 explores how affirmative action supports minority-led firms in the procurement market. While the previous literature has focused on the utilization of minority-led firms in procurement contracts, this paper examines corporate strategies among female-owned businesses. In the Korean public procurement market, there was a law revision that mandated a quota to contract with 'certified' female firms, i.e., firms confirmed by the government as being owned by female CEO(s). I find that 'certified' female firms participated in and won more bids with two channels involved: the previous contracting relationship between procuring agencies and firms and the information sharing in the network among certified female firms. I also examine if affirmative action resulted in strategic changes among non-targets in a way that male-owned firms changed their ownership to female-owned to be eligible for the preferential policy. Although previously uncertified female-owned firms opted to complete the female certification, I do not find such evidence from male-owned firms. Certified female firms acquired a larger revenue share made from both procurement and private markets. Affirmative action may increase the revenue share of its target outside the policy's realm.Chapter 3 analyzes the composition of innovation and the technology adoption of green inputs in the pesticide industry. By leveraging the announcement of a major expansion of organic food offerings in the one of the largest grocery chains in the U.S in 2006, this chapter studies how a change in the demand for organic food influenced product composition in the market of biopesticides, i.e., non-toxic pesticides approved for organic agriculture. We find suggestive evidence that the organic market expansion and the increase in demand for organic produce resulted in a higher share of newly-registered biopesticides in the pesticide market. However, many of the new biopesticides are generics, indicating a lack of innovative behavior. We also find that farms in California increased biopesticide use and used lower-toxicity pesticides on average. The change in pesticide use is not attributed to the pesticide product choice set but rather an increase in profitability to organic or mixed-use production. This chapter provides implications on how the composition of innovation is shaped by demand-side shocks and factors that affect green technology adoption.

Success in the labor market is a key driver of social mobility, and the primary focus of this dissertation is to explore how to ensure young people possess the skills and conditions necessary to achieve such success.In recent years, millions of children have been displaced, emphasizing the importance of evidence-informed public policy for migrants and recipient communities. Chapter 1 examines the sudden influx of Venezuelan migrant children into the Peruvian school system. Analyzing cross-grade within-school variation, I find that as Venezuelan migrants enter Peruvian schools, parents transfer their children to higher-quality schools with fewer migrants. While native flight may mitigate the effects of migrant influx for some students, it generally brings no gains and comes at a high cost.Chapter 2 investigates the impact of a law in Bolivia that temporarily lowered the legal working age from 14 to 10 and introduced benefits and protections for child workers. Using a difference-in-discontinuity approach, I find a decrease in work for children under 14, particularly in areas with a higher threat of inspections. However, there is no evidence of improved work safety, suggesting reductions in visible child labor may be driven by avoiding legal and social sanctions rather than increased safety measures.Chapter 3 analyzes young people's constraints as they transition into adulthood by studying programs to improve the employment prospects of underemployed Rwandan youth. I find that while cash transfers affect marriage and fertility differently for men and women, they do not significantly impact labor market outcomes differently. There are also no changes in the perception of gender roles. While financial constraints are a significant barrier for young people in Rwanda, addressing these alone does not alter deeply entrenched gender roles and cultural norms.

This dissertation investigates factors influencing economic growth in developing countries through their impact on the labor market. Chapter 1 analyzes the effects of industrial robot adoption on local labor markets in Mexico. Using data on robot imports alongside administrative, survey, and economic census data and leveraging spatial and temporal variations across commuting zones, the study finds that robot adoption does not negatively affect formal employment and average wages, unlike developed countries. Additionally, regions importing robots see increased gross value added per worker, indicating productivity gains offset potential displacement effects. Chapter 2 studies the connection between intergenerational social mobility and economic development, using a novel dataset on educational mobility for 52 Latin American sub-regions. Employing a new weighting procedure that accounts for cohort participation in the economy, the study finds that higher intergenerational mobility is strongly linked to economic growth and improved development indicators. The final chapter examines discrepancies in reported Type I errors in linked datasets, which are crucial for studying historical intergenerational mobility. The study reveals that differences arise from reporting conditional versus unconditional error rates. The higher unconditional error rate, considering all algorithm links, better reflects the true error rate and affects inferences from historical data lacking a ground truth.

This dissertation consists of three essays on human capital. In Chapter 1, I investigate the impact of collaborating with robots on human capital, focusing on professional baseball umpires who were provided with and then deprived of robot assistance. Umpires displayed enhanced accuracy with robot assistance, but experienced significant declines in performance once it was removed. I argue that these findings can extend beyond baseball, suggesting broader implications for occupations sharing similar skill sets. In Chapter 2, joint with Felipe Goncalves and Emily Weisburst, we investigate the effects of changes in workplace risk awareness on police behavior and public safety, focusing on incidents of police officer fatalities while on duty. Our findings reveal that following the death of a fellow officer, police officers reduce arrest activity for one to two months, indicating heightened fear. This reduction is most prominent for minor offenses and is more pronounced in smaller cities. Yet, we find no evidence of increased crime rates during this period, suggesting that reduction in arrests does not adversely impact public safety. In Chapter 3, joint with Anna Aizer, Shari Eli and Adriana Lleras-Muney, we utilize newly gathered data of 16,000 women who applied for Mothers' Pensions, America's first welfare program, to explore the impact of means-tested cash transfers on lifetime family dynamics and maternal welfare. In the short term, these transfers led to postponed marriage and reduced geographical mobility. However, in the long run, they had no discernible effect on remarriage probability, spouse quality, fertility, or maternal well-being, as measured by longevity and family income in 1940. With the absence of significant negative behavioral outcomes, we suggest that the benefits of such transfers may outweigh the costs, particularly if they yield even modest positive impacts on children.

This dissertation consists of three chapters. Chapter 1 presents new empirical evidence and a theoretical framework, which suggests that firing costs impede human capital accumulation within firms. To that end, I investigate the impact of a German reform in 2004 that removed firing costs exclusively for smaller establishments. Applying a difference-in-differences approach to administrative matched employee-employer data, I find that the targeted smaller establishments experienced an increase in the proportion of employees participating in training compared with others. To clarify the underlying mechanism and quantify the aggregate effect of the reform, I develop an on-the-job search model with human capital and firing costs. Reducing firing costs in this model stimulates job creation by firms, and thereby increases the probability of workers transitioning to firms in which their skills are more valuable. This encourages workers to accumulate human capital. The calibrated model suggests that the German reform could result in a 0.58% increase in aggregate productivity, with 0.49% attributed to human capital accumulation and the rest to more efficient resource allocation.Chapter 2, which is co-authored with Yasutaka Koike-Mori and Koki Okumura, develops an endogenous growth model that incorporates on-the-job search and examines the allocation of inventors across firms, knowledge diffusion, and its impact on growth. In our model, inventors play dual roles: they engage in in-house R&D and transfer knowledge from previous employers to new ones when changing jobs. Using an administrative panel dataset on German inventors matched to their employing establishments and patents, we find that, relative to general workers, inventors are more likely to transition to less productive establishments and suffer a higher wage growth via the transition. We also find that the knowledge base of establishments measured by patents grows faster when a significant proportion of their inventors originate from establishments possessing a larger knowledge base. We then calibrate the model to reflect these empirical findings and examine the effects of innovation policy. While subsidies to frontier firms discourage knowledge diffusion from these firmsto technologically lagging firms, these subsidies also encourage innovation within frontier firms. The former negative effect dominates in the short term, but the latter positive effect dominates in the long run.Chapter 3, which is co-authored with Tomohide Mineyama, estimates the user cost of labor (UCL) - an allocative labor cost measure used in analyzing economies with ongoing employment relationships—while controlling for cyclical changes in job-match quality, which has been identified as a potential problem in previous studies. We use a novel dataset that exploits school (nonemployment)-to-employment flows to control for this measurement problem. The estimated UCL remains highly procyclical after making this correction, with an estimated wage semi-elasticity of 2.69 with respect to a one percentage point change in the unemployment rate. The semi-elasticity is around twice as large as that of the new-hire wage. We also find that the corrected UCL cyclicality is asymmetric, with the UCL rising in booms but remaining flat in downturns. To account for this asymmetry, we develop a directed search model with asymmetric information in which firms use wages as a screening tool to receive applications from targeted workers. The model economy generates asymmetric labor market dynamics that are consistent with the data.

The three papers that comprise this dissertation contribute key building blocks for my analysis of migration corridors as critical spatialities with the potential to significantly rework our approach to global migration governance—including in legal, political, and scholarly discourses. The study is a multi-site ethnographic account of migration corridors—circuits of human mobility within and across national borders that are governed by nation states as well as transnational financial, political, and social forces. It examines governance of migration corridors traversed by migrant agricultural, domestic, and garment workers in relationship to three building blocks: (1) expulsions that propel migration (e.g. national/global patterns of uneven development, environmental devastation, corporate land grabs, and conflict); (2) junctions where disparate migration flows converge and are redirected, including urban production and service hubs, special economic zones (SEZs), and territorial borders; and (3) forces that direct migration flows (e.g. legal regimes, product and labor supply chains, securitization, patriarchal norms, and local processes shaped by women labor migrants, recruitment intermediaries, and kinship and social networks).

This thesis explores app-based ride-share as an emblem of the digitized, post-Fordist economic transition. I review how its recent emergence connects to the 20th century taxi-dominated ride-hailing industry, and to broader post-War socioeconomic and political trends. Examining driver demographics in the context of the ride-share labor process, I assess how class, gender, racial, and other social hierarchies influence workplace outcomes. Throughout, I emphasize the modulating role of both algorithmic management and worker-to-worker exchange.

Climate change is one of the largest challenges of our time. This dissertation assesses the economics of water conservation programs in the face of increasing water scarcity and estimates the causal impact of education on people’s climate-related beliefs, behaviors, policy preferences and voting.Chapter 1 focuses on takeup of water-conserving rebate campaigns in the U.S. Southwest, estimating a demand curve, neighbor spillovers, and heterogeneity for programs replacing thirsty grass lawns with drought-tolerant landscaping. I find strong evidence that households respond to rebate generosity: a 1% increase in generosity leads to a 0.93% increase in takeup. Each conversion leads to an additional 0.61 adoptions the following year. I also find substantial heterogeneity by race and wealth, with Hispanic rebate applicants 47% more likely than White applicants to have their application rejected, a disparity that completely disappears when using a contractor.Chapter 2 uses thermal infrared satellite imagery to estimate temperature changes from such landscaping conversions in an event study analysis. Conversions increase summertime parcel temperatures by 0.6°C (1.1°F) with effects twice as large on the hottest 20% of days and for homes with the most removed vegetation. Using existing estimates from the literature, this implies annual heat costs of $1,675 per household, comprised of increased electricity usage ($48) and mortality risk ($306), as well as harder-to-quantify comfort values, diminished cognition, and costly adaptation behavior. In contrast, water savings are only $574-954 for a typical home, depending on water price.Chapter 3 estimates the causal effect of education on pro-climate survey responses using variation in compulsory schooling laws across 20 European countries. An additional year of education increases pro-climate beliefs, behaviors, and policy preferences with no detectable effect on voting for green parties. Results are robust to the functional form specification of the time trend, the selection of included countries and reforms, as well as an alternative specification of the instrument.

While higher education has been known as the great equalizer, the racial wealth gap of college graduates is overlooked: Among college-educated households (2013) the median net worth was $359,928 for white households; $250,637 for Asian households; $49,606 for Hispanic households; and $32,780 households for Black households. Although MSIs demonstrate social mobility for low-income students, racial/ethnic gaps are unclear. This specificity on students of color is foundational to Minority-Serving Institution (MSI) accountability and funding.Using the 2008-2018 and 2016-2017 Postbaccalaureate and Beyond datasets, this quantitative study integrated descriptives and logistic and multiple regression to examine racial disparities in annualized salary, including those within MSI types, and implemented propensity score matching (PSM) to simulate the treatment effects of internships. Findings showed that Asian ($91,286) and white ($78,227) graduates increasingly outpaced Hispanic ($64,142) and Black ($60,690) graduates in annualized salaries 10 years after degree completion. Among MSIs, racial groups generally fared better at Predominantly White Institutions, while some groups fared better at HBCUs and HSIs. While Black and Hispanic graduates had lower annualized salary averages compared to white and Asian graduates, multiple regressions showed that Black and Hispanic graduates had higher predicted salaries once social class entered each model. PSM results showed that participating in paid internships had higher predicted positive effects (b=2067, p

This dissertation examines the different aspects of education policies and their socioeconomic implications through three distinct research papers. The first chapter investigates the causal impact of grade retention in primary school on juvenile crime in Chile and it was written jointly with Juan Diaz, Nicolas Grau, and Jorge Rivera. Utilizing a fuzzy regression discontinuity design, the study finds that repeating an early grade in primary school reduces the probability of committing a crime as a juvenile by 14.5 percentage points. By employing a dynamic model, the research demonstrates that the observed result is primarily driven by two mechanisms associated with the timing of grade retention. First, grade retention in early grades decreases the likelihood of subsequent retention in later grades. Second, late grade retention in primary education exerts a stronger positive effect on crime reduction compared to direct effects in early grades. The findings suggest that, if grade retention remains an ongoing policy, its optimal implementation at the margin should involve retaining students in early grades to prevent retention in later ones.The second chapter explores the equity and efficiency effects of a 2013 reform in the Chilean college admissions system. This reform aimed to increase equity by introducing a third component based on the student's GPA relative to the historical average at their high school. By simulating the admission mechanism with and without the relative GPA boost, the study categorizes applicants into three groups: those who gained access to more selective programs (pulled-up), those who lost access to more selective programs (pushed-down), and those whose admission was unaffected. Employing a difference-in-differences design, the research estimates the impacts of the reform on enrollment, persistence, and graduation. Pulled-up students were able to persist in their newly accessed programs, leading to more selective degree attainment without significant effects on overall BA completion. Pushed-down students, predominantly from better-educated/higher-income families, experienced comparable reductions in the probability of graduating from selective programs, offset by gains in graduation from less selective programs. The study concludes that the reform improved equity with minimal or no loss in efficiency.The third chapter estimates gender bias in a college admissions system based on standardized test scores and was jointly written with Matias Grau, Nicolas Grau, and Damian Vergara. This research applies standard discrimination literature tools in the context of the Chilean centralized admission system. We show that marginally enrolled female applicants exhibit higher first-year grades and are more likely to graduate on time compared to marginally enrolled male applicants. By employing an outcome test model, the research translates outcome differences into disparities in selection thresholds. We interpret this result as gender bias against females who, being "equally qualified" - based on their latent college performance - are not admitted into the program. Approximately 10% of enrolled male students would not have been admitted to their programs under the female effective selection thresholds. A counterfactual exercise simulating the college assignment algorithm with gender-targeted test-score inflation to correct for estimated bias suggests that such a policy would enable an improvement for 2% of female applicants in their preferred options relative to the observed assignment. Similar size effects are found with a decrease in the test score weight assigned to the application score formula. Collectively, these research papers shed light on the complex relationships between education policies, socioeconomic outcomes, and student trajectories, emphasizing the importance of evidence-based policy recommendations to promote equitable educational systems.

In this dissertation, I study factors that shape women’s economic and social participation.I explore the role of extra-legal factors in decision-making in sexual crime cases, and studyaspects of women’s economic participation through their participation in leadership positionsand competitive environments, and the intergenerational health impacts of maternalemployment opportunities.Evidence suggests that the lack of safety pushes women to choose lower-quality educationinstitutes and drop out of the labor market. The MeToo movement took the world bystorm and brought the issue of sexual crimes to center stage. However, we do not know ifsuch global outrage systematically increases the costs of sexual crimes for the perpetratorsin form of the punishments meted out to them. In absence of granular data on case-specificjudicial outcomes and media reporting, it is hard to rigorously answer this question. Motivatedby this, in my first chapter, I investigate if judges are more likely to convict offendersof sexual crimes when they are exposed to more media reports about sexual assaults. Iuse rich administrative data on the universe of sexual assault cases filed in lower courtsin India and high-frequency data on daily media reports about sexual assaults in districtsin the country. Using a generalized difference-in-difference model, I find that judges are more likely to convict offenders in sexual assault cases that carry low punishments whenthey are exposed to more media reports. I find no impact of media reports on more serioussexual assault cases like rape cases, where judicial discretion might be lower since rapecases are filed with more forensic evidence, and convicting the offender here carries harsherpunishments. The increase in the probability of conviction for assault cases is likely to bedriven through the channel of more scrutiny by the judges in response to media outrage.Apart from safety concerns, women’s participation in the economy is also restricted due tothe burden of childcare duties that disproportionately fall on them. In my second chaptertitled “Employment Booms and Infant Health: Evidence from Ready-Made GarmentSector in Bangladesh”, I argue that having more women in the labor force can improvetheir children’s outcomes. I look at the impact of employment booms, particularly beneficialfor women, on infant health using evidence from the readymade garment industryin Bangladesh. I exploit exogenous variation in employment created in the sector due tothe removal of a trade quota, which made a spatial and temporal variation in employmentopportunities for women. Using a difference-in-difference model, I find that access tothese jobs leads to increased female labor force participation. Correspondingly, I find thatwomen delay their age at first birth, increase the birth spacing between two births, andbreastfeed longer. These channels of better health inputs and behavior collectively lead toan improved survival rate of neonates in areas with a relatively high density of these jobs.Motivated by this positive spillover of women’s economic empowerment, in the third chapterof my dissertation, I explore if women’s representation in leadership positions mattersfor educational outcomes. In this paper, I ask if constituencies with female leaders in Indiasee higher growth in the number of new schools in rural India. Using a regression discontinuitydesign, I find that constituencies, where a female leader is narrowly elected againsta man, do not see any differences in school construction compared to constituencies where a female leader narrowly lost to a man. Using detailed community development projectexpenditure data by these leaders from two states in the country, I find descriptive evidencethat, on average, female leaders get a lower number of community development projectssanctioned than male leaders. They also get less money approved for these projects. Thismay be suggestive of the lower political agency of women. These results extend the narrativeon women’s performance in competitive and male-dominated environments.

While the usual paradigm of supply and demand in a frictionless market is useful for dis- cussing many issues, plenty of important questions are not easily addressed with this approach. In this dissertation I aim to further our understanding of markets where trading frictions are relevant from two different perspectives: a search-theoretic model and labora- tory experiments.The first chapter of this dissertation focuses on a frictional labor market where firms can decide to replace current employees by randomly search for new candidates. I show that even in an economy where agents are risk neutral and face no liquidity constraints, severance payments emerge as part of an optimal contract when firms can search on-the-job. An optimal contract is composed of both the wage and a payment conditional on the worker being replaced. The size of the payment is chosen to allow the firm to internalize the externality associated so that firms take into account the total surplus effect of their replacement hire decisions. I add ex ante heterogeneity among workers which makes the pairwise optimal contract insufficient to achieve the constrained-efficient result (even under the Hosios condition) and then use a calibrated version of the model to illustrate that a lower on-the-job search cost can in fact reduce welfare.In the second chapter, I develop an experimental framework to investigate price, output and welfare consequences of implementing the optimal monetary policy in a version of the search- theoretic money model of Lagos and Wright (2005): The Friedman Rule. I aim to further understand previous experimental results by Duffy and Puzzello (2022) which are somewhat at odds with the standard theory and contrary to the optimality of the rule vis-`a-vis an inflationary scheme, which they suggest could be rationalized on the basis of liquidity constraints and/or precautionary motives due to future price uncertainty. For this, I request subjects to make predictions about market prices and include a novel treatment for the decentralized goods market to try and mitigate price uncertainty: prices are exogenously imposed on consumers so they can only select from a fixed menu of quantities (and prices) when making an offer. My results tend to be consistent with previous experimental findings, but no clear evidence for the Friedman Rule emerges. Even when prices are fixed exogenously in pairwise meetings, evidence in favor is still only mixed and high volatility of prices in the centralized market persists. When using subjects own predictions about the centralized market price to look at how they expected to rebalance their holdings conditional on their beliefs there seems to be a clear bias: subjects mostly want to increase their token holdings regardless of previous trades.In the third and last chapter, I use a semi-unstructured bargaining approach to experimentally study the effects of liquidity constraints on the determination of terms of trade. This setting is specially relevant for search-theoretic models of money: trade surplus and its di- vision are endogenously and simultaneously determined, with participants facing possible liquidity constraints due to buyer’s endogenous ex-ante choice of costly money holdings. My aim is to test the empirical relevance of two widely used axiomatic bargaining solutions: generalized Nash bargaining and Kalai’s proportional bargaining. Each bargaining solution predicts different outcomes and buyer’s anticipating their decision’s effect on the bargaining outcome may choose to additionally restrict their money holdings, which can prevent efficient outcomes from being achieved, even when money is costless to hold. A most relevant issue, since the protocol to determine terms of trade in monetary economies is critical for normative results, optimal monetary policy, and the welfare cost of inflation. By imposing different costs to holding money, I find strong evidence that costlier holdings do incentivize participants to economize on money holdings leading to a more constrained bargaining set resulting in less production and surplus, but find only mixed evidence in favor of any of the two bargaining solutions. Finally, I introduce two possible variations to the model that help better understand the data: myopic buyers and a sunk cost fallacy.

This dissertation includes three chapters in applied labor economics. The first chapter considers the growing trends of colleges and universities making the submission of SAT and ACT scores optional for undergraduate admissions to bolster racial diversity. It uses a panel of liberal arts colleges from IPEDS and applies a two-way fixed effects approach to determine whether this test-optional policy is effective at achieving this goal. It also estimates the impact of the policy on the graduation rates for underrepresented minority (URM) students. This chapter finds that the policy bolsters freshman URM enrollment among test-optional institutions throughout the sample as a whole, regardless of admissions selectivity and early versus late treatment timing. The effects of this policy on the URM 4-year and 6-year graduation rates are heterogeneous across colleges by their selectivity in admissions. While the most selective colleges in the panel experience no change in URM graduation rates, less-selective colleges experience declines in these graduation rates.The second chapter, joint with Peter Kuhn, uses a vignette-based survey experiment on Amazon’s Mechanical Turk to measure how people’s assessments of the fairness of race-based hiring decisions vary with the motivation and circumstances surrounding the discriminatory act and the races of the parties involved. Regardless of their political leaning, survey subjects react in very similar ways to the employer’s \textit{motivations} for the action, such as the quality of information on which statistical discrimination is based. Compared to conservatives, moderates and liberals are much less accepting of discriminatory actions and consider the discriminatee’s race when making their fairness assessments. This chapter also describes four pre-registered models of fairness -- utilitarianism, race-blind rules (RBRs), racial in-group bias, and belief-based utilitarianism (BBU) -- and shows that the latter two are inconsistent with major aggregate patterns in the survey data. Instead, it argues that a two-group framework, in which one group (mostly self-described conservatives) values employers’ decision rights and the remaining respondents value utilitarian concerns, explains the main findings well. In this model, both groups also value applying a consistent set of fairness rules in a race-blind manner.The third chapter estimates the effects of the Federal Perkins Loans Program on students' non-academic work status and labor supply, as well as their choice of STEM major and post-graduation occupation via matching. Perkins loans were unique because they converted into grants if a student recipient entered a designated career path. Therefore, this chapter leverages this conditional loan cancellation feature to determine whether recipients enjoyed the benefits often associated with grant programs (e.g., reduced incentive to work) and motivated them to enter loan-forgiving occupations. It finds that Perkins loans reduce students' non-academic work activity and grades. The effect size for the latter outcome, however, is marginally small. And this chapter finds that the program had little impact on their likelihood of pursuing a STEM major or loan-forgiving occupation.

Interactive, online sex work is a relatively new and undertheorized industry in the sex trade. Research in sex work has uncovered rich and complicated histories of Black sexualities and their connections to global capitalism and imperialist projects. However, understanding Latinas and Latinidad in the context of sexual labor and global economies has not been adequately captured. Research on online sex work has begun to scratch the service of our understanding of how racialized sexualities transcend borders through virtual space to meet the high demand for “exotic” sexualities. In this interdisciplinary study, I examine the intentional creation of "The Latina Body" in the translational context to understand how trade agreements and public policy help shape a desire for The Latina Body. I use sex work studies as an essential framework here because, like Black sexualities, Latina sexualities are entwined with legacies of imperialism and economical extraction and exploitation. My study adds to this conversation by theorizing The Latina Body as a strategically manufactured imaginary that erases Blackness and Indigeneity from the mythology of Latinidad to create a palatable desire from the United States into the Global South. These desires also create industries based on this logic, such as sexual economies like digital sex work. I am intervening primarily in decolonial theory, political economy, labor studies, Chicanx studies, and sex work studies to create a sub-area that I call the political economy of Latina sexualities. Through this framework, I study The Latina Body in popular culture and online pornography to answer the following research questions: 1) How does U.S. foreign policy aid in creating The Latina Body? 2) How does The Latina Body become a consumable product born from colonialism and imperialist logic? 3) How are racialized online sexual economies created in countries like Colombia and the U.S.? 4) Why do Latina sex workers invoke or refuse narratives of racialized sexualities in their online sex work? For the first chapter, I conducted a content analysis of U.S. trade policies with Latin American countries. These policies are accompanied by media campaigns that position Latina sexualities as metaphors for bountiful land. I also analyze media representations of Latina sexualities to understand the creation of The Latina Body in the public imaginary. The second chapter examines AfroLatina rap artist Cardi B and her social media presence as a former dancer in New York City. I conducted a discourse analysis of her video narratives on the social media platform, Instagram, to theorize her explicit sex worker discourse through her lived experience as a Black Latina. I note how she signals Black Latinidad and desires through her work to disrupt notions of respectable femininity and signal often-incomprehensible defiance of gendered labor. In the third chapter, I look at digital sex workers on the subscription-based website Onlyfans.com to analyze the surveillance mechanisms of online pornography. I interviewed both cis and trans-Latinas who sell content online to understand how they navigate their gender, sexuality, race, ethnicity, and national identities in their work. The fourth chapter turns to the adult webcam platform Chaturbate.com. I utilize online Virtual Ethnography to travel to Colombia, the world’s second-largest producer of webcam models per capita. Colombia experiences economic instability due to U.S. trade policies. Meanwhile, the webcam industry illuminates a desire from U.S. consumers to travel virtually to Colombia and consume The Latina Body. The webcam industry is a stabilizing industry. "The Latina Body" is the contradiction that allows global capitalism to continue to seem logical. Its creation has facilitated exploitative trade agreements and has contributed to exponential economic growth in the camming industry. Drawing on decolonial theory, Latin American studies, theories of political economy, sociology, sex work studies, Chicana feminisms, and feminist studies, this study introduces new perspectives and areas of research for Latina sexualities and argues for the integration of sex work in our understandings of racialized sexualities. It is one of the few studies that approach Latina sexualities through the lens of sex work to better understand the global processes of capitalism and economic expansion. I posit that sex work as a labor category can teach us about how Latinas interface with global operations and how they learn to flirt with, to fuck, and to taunt the power dynamics they interface with. This study moves beyond the debate of sex work as either exploitative or empowering to advance our understanding of the legacies of colonialism and how these legacies influence our relationships with global markets.

This dissertation analyzes factors that affect labor market outcomes of individuals in the United States. The first and second chapter examines the relationship between children's access to public health insurance and parental labor supply. The third chapter analyzes how racial discrimination affects labor market decisions of racial minorities. The first chapter of the dissertation exploits variation resulting from a series of federal and state Medicaid expansions between 1979 and 2014 to estimate the effects of children’s access to Medicaid on the labor market outcomes of mothers. The results imply that the extended Medicaid eligibility of children leads to positive maternal labor supply responses at the extensive and intensive margins. The effects are concentrated among single non-white and/or Hispanic mothers. The analysis of mechanisms suggests that for non-white and/or Hispanic mothers, Medicaid is less likely to work through marital and educational outcomes and that the effects are driven by the head of the family. For white non-Hispanic mothers increased educational attainment may explain the small and imprecise labor supply responses. In the second chapter I examine the relationship between children's access to Medicaid and paternal labor market outcomes exploiting Medicaid expansions between 1979 and 2014. I find that fathers increase labor supply at the intensive margin as a result of extended Medicaid eligibility of their children and that the effects are concentrated among married white non-Hispanic fathers. The estimated effects on earnings, however, respond in areas outside of the maximum Medicaid eligibility cutoffs. To corroborate this finding, I explore various explanations and conclude that the most likely explanation is the correlation between the simulated eligibility measure and unobservable characteristics. The third chapter co-authored with Kerwin Kofi Charles, Hani Mansour, Daniel I. Rees, and Bryson M. Rintala analyzes the relationship between racial discrimination and labor supply of racial minorities. Between 2001 and 2014, more than 6,500 American soldiers died while serving in Afghanistan and Iraq. Drawing on data from the Defense Manpower Data Center, which contains information on each of these soldier’s home state as well as the exact date on which he died, we estimate the relationship between home-state fatalities and the labor market outcomes of first- and second generation Arab and Muslim men working in the United States. Because home state does not influence when, where or how the U.S. military deploys its soldiers, news of a soldier’s death can be thought of as producing a temporary, state-specific shock to the degree of prejudice faced by Arab and Muslim men working in the United States. We find that home-state fatalities are, in general, unrelated to wages and employment status. However, we find a strong, negative relationship between hours worked by Arab and Muslim men and whether a U.S. soldier from their state of residence died in the weeks preceding their interview. We argue that this result is consistent with customer taste-based discrimination but inconsistent with statistical models of discrimination.

Can personalized mentorship by experienced workers improve young job seekers’ labor market trajectories? To answer this question, in the first chapter of this dissertation I study “Meet Your Future”, a mentorship program we designed and randomized which assisted a subset of 1,112 vocational students during their school-to- work transitions in urban Uganda, where youth unemployment is high. The program improved participants’ labor market outcomes. Relative to the control, mentored students were 27% more likely to work three months after graduation; after one year, they earned 18% more. Call transcripts from mentorship sessions and survey data reveal that mentorship primarily improved outcomes through information about entry level jobs and labor market dynamics, and not through job referrals, information about specific vacancies, or through building search capital. Consistent with this finding, mentored students revise downward their overly optimistic beliefs about starting wages and revise upward beliefs about the returns to experience. As a result, they lower their reservation wages and turn down fewer job offers. The results emphasizes the role of distorted beliefs among job seekers in prolonging youth unemployment and proposes a cost effective and scalable policy with an estimated internal rate of return of 300%. In the second chapter of this dissertation, I move to investigate whether hiring processes themselves can disadvantage women and consequently explain part of the gender wage gap and the occupational segregation documented in many labor markets across the world. Specifically, I look at referrals, a significant factor in hiring decisions and one of the primary ways to land a job. With my coauthor, we conduct a correspondence experiment to examine how referrals by firm employees may perpetuate occupational gender segregation among Uganda’s skilled workers. We start by presenting pairs of gender-differing profiles of potential candidates to workers in a wide range of industries, and ask who they would refer to their firm for an internship we subsidize. We randomize the gender of the high-experience profile to elicit discriminatory preferences while mitigating endogeneity in network formation. To validate the findings in this anonymous setting we subsequently offer participants the possibility to use their networks as referral choice sets, the lifelike setting. We further randomize the disclosure of the referral source’s name to the employer. We document three facts. First, discrimination in referrals exists against both genders and is correlated with subjects’ gender and the gender dominance of their sector; however, discrimination against the non-stereotypical gender is more prevalent in male-dominated sectors. Second, the intrinsic preferences of employees are a significant driver of their discrimination in referrals, which, in general, do not simply reflect passthrough from employers’ preferences. Thirdly, when the referral is private, subjects in male dominated sectors are more likely to refer women, indicating that beliefs regarding employer preferences are a significant driver of pro-male bias in these sectors. In the last chapter I investigate gender disparities in the effect of COVID-19 on the labor market outcomes of skilled Ugandan workers. Leveraging a high-frequency panel dataset, my coauthors and I find that the lockdowns imposed in Uganda reduced employment by 69% for women and by 45% for men, generating a previously nonexistent gender gap of 20 p.p. Eighteen months after the onset of the pandemic, the gap persisted: while men quickly recovered their pre-pandemic career trajectories, 10% of the previously employed women remained jobless and another 35% remained occasionally employed. Additionally, the lockdowns shifted female workers from wage-employment to self-employment, relocated them into agriculture and other unskilled sectors misaligned with their skill sets, and widened the gender pay gap. Pre-pandemic sorting of women into economic sectors subject to the strongest restrictions and childcare responsibilities induced by schools’ prolonged closure only explain up to 65% of the employment gap.

This dissertation consists of two essays on job search and job choice. In Chapter 1, I examine how on-the-job search (OJS) effort of employed workers varies over the business cycle. First, I document new evidence that aggregate OJS effort rises during a recession, as more workers start searching on-the-job and average search intensity increases when unemployment rises. Next, I account for compositional changes in the pool of employed workers and job seekers over the business cycle and find that workers change their search behavior in response to changing economic conditions. In particular, workers are more likely to search due to fear of job loss and search for an additional job when unemployment is higher. In addition, I find that job seekers increase their search intensity when unemployment rises during a recession. In Chapter 2, I estimate the value that workers place on non-wage job characteristics and assess their impact on compensation inequality in Germany. First, I evaluate the incidence of four key job attributes and find large disparities in the prevalence of job attributes by gender, education, and age. Next, I analyze an experiment given as part of a national survey in Germany to estimate the value that workers place on eight non-wage job attributes. In particular, I find that workers are willing to pay 31% of their wage to have a permanent employment contract, 13% of their wage for good promotion opportunities, 10% of their wage for schedule flexibility, and 8% of their wage to avoid overtime work requirements. Finally, I find that accounting for the incidence and valuation of non-wage job characteristics widens compensation inequality in Germany.

In Chapter 1 of this dissertation, I study the effect of student debt on the post-schooling migration decisions of high-skill workers in the United States. Over the past 40 years, the U.S. has experienced significant skill-based geographic sorting, with high-skill workers increasingly concentrating in large cities. During the same period, the growth of student debt has far exceeded the rate of inflation. In this chapter, I document a link between these two facts. I first estimate the causal effect of student debt on post-schooling location choices by exploiting an expansion of federal student loan limits. Using a difference-in-differences framework, I find that $10,000 of additional debt increases the probability that individuals locate in large metropolitan counties by 6.5 percentage points. By incorporating student debt into a standard spatial equilibrium model, I find that the rise in student debt from 1980 to 2019 can account for 5-19 percent of the increase in skill-based sorting over this period. Counterfactual simulation of three policy proposals – debt forgiveness, tuition-free college, and income-driven repayment – show that only income-driven repayment can eliminate distortions to location choices while improving welfare.The remaining chapters focus on public-sector employee pension systems in the United States and the over $3 trillion in debt associated with them. In Chapter 2, I consider the political economy problem of setting pension benefit levels, where politicians balance the demands of general voters and public-sector unions. More specifically, I empirically show that expanded collective bargaining rights for public-sector employees significantly increased pension plan generosity in the 20th century, and is associated with higher levels of unfunded liabilities in the 21st century. I also provide descriptive evidence that increased plan generosity resulted in higher levels of unfunded liabilities because local governments shirked their expected contributions to pension funds and made overly optimistic assumptions on investment returns, both of which were made possible by systematic information asymmetries around public pension plans.In Chapter 3, I examine the implications of public-sector pension debt for the local economy. I exploit plausibly exogenous shocks to the reported levels of unfunded pension liabilities in a difference-in-differences framework to investigate the speed and extent to which debt shocks are capitalized into house prices. I find that increases in public debt depress local house prices relatively quickly (within 9 months). Additionally, this effect is driven by responses in the price of single-family homes, owners of which may be more likely to rely on public goods that are subject to cuts following spikes in reported pension under-funding.

The credibility revolution in applied microeconomics research has led to sharp improvements in the identification of causal relationships, but risks leading the field toward interesting designs rather than interesting questions. In this work, I use credible identification strategies like random assignment and regression discontinuity designs to answer important questions in labor and public economics with a special focus on the intersection of inequity, higher education, and political economy. Chapter 1 shows that means-testing institution-specific merit scholarships can enhance equity and efficiency, with benefits accruing to colleges and students alike. Chapter 2 demonstrates that when the Appalachian Regional Commission designated the poorest decile of counties as economically distressed, voters were more likely to support Democratic candidates for US President and the US House of Representatives. Chapter 3 illustrates that conferring an admissions advantage to UC applicants in the top four percent of their high school class reduced the share of students who eventually registered to vote as Republicans and increased independent or Democratic registration, primary voting, and Democratic presidential primary turnout.

The energy industry is undergoing a rapid transformation in the United States. Climate change is leading to many of the changes, including market policy development, and the transition away from old technologies to new. The changes are leading to new challenges, real or perceived. These speed of these changes is resulting in the need for policy makers and electricity system planners to make decisions under uncertainty and modify their behavior as they learn from their policies and actions. In this dissertation I strive to understand how climate policies effect regions without climate policies, and the impacts of changing energy technologies on society. The first chapter of this dissertation takes up the effect of regional climate policies, and the second two take up the energy transition.In the fist chapter of this dissertation,I conduct the first reduced form accounting analysis of leakage in the California electricity sector following the enactment of the Cap-and-Trade program. Leakage is the result of incomplete regulation, where in this context the Cap-and-Trade program only regulates a subset of the electricity market, leading to the unintended increase in carbon dioxide emissions through the substitution of generation from the regulated region to the unregulated region. The paper finds near-term emissions are below that of unaccounted for emissions, but utilization of coal is greater than accounted for by regulators. I then find across the enactment date of the Cap-and-Trade program emissions from unregulated entities increased by 26 percent as opposed to falling by 32 percent as reported. This implicates the need to find stronger methods to reduce imported emissions from unregulated entities. In the third chapter of this dissertation,in collaboration with James Bushnell and David Rapson, we study the coal industry and labor. Coal plays an important role in the social zeitgeist to ensure communities maintain a quality labor force. We conduct a study on the impact of coal power plant retirements on local employment from 2010-2018, and find on average there is no impact to employment within the labor market. We implement an event study and synthetic control approach to estimate the dynamic effect of the retirement and find the accusations that the loss of a coal facility are largely attributable to secular trends. We further study the effect of labor market thickness on the impact of the retirements to find there is no statistical relationship with the labor market size. We find there is heterogeneity in the effect of a retirement, but there are no predictive variables to allow us to discern why there may be a large positive or negative effect. The findings are robust to multiple empirical approaches and specifications. In the third chapter of this dissertation, also in collaboration with James Bushnell and David Rapson, we take advantage of energy technology deployment across California to study the relationship between emerging energy technologies. Specifically electric cars and residential solar, and their impacts on the reliability of the distribution grid. The results reveal that there is no impact associated with the deployment of new technologies on the grid. The findings suggest that grid planners are able to make the necessary investments to avoid adverse impacts from emerging technologies on grid reliability, and deployment should not be slowed. Our results find that there are minimal effects immediate impacts following deployment of these distributed technologies, but there is no relationship with the long-term impact from these technologies.

My first chapter studies the origins of gentrification. I propose a mechanism through which a wave of gentrification can be triggered in a neighborhood: the opening of large offices of technology firms. Using information on nine such events and transaction-level housing data, I develop a difference-in-differences strategy that compares house prices in the vicinity of the new office to those in closely-matched neighborhoods slightly further away. I find that property prices rise 11% in treatment areas relative to control areas within two years after the opening. This difference subsides somewhat but remains at +8% five years later. These findings are substantially stronger than in the existing literature and thus suggest that such office openings can have a major impact on their neighborhood. I investigate two mechanisms, agglomeration forces and the development of consumption amenities, through which the impact of a single establishment opening can be amplified and sustained.My second chapter analyses the causes of the shift towards low-income service jobs that American non-college workforce has experience in recent decades. Together with the rise of college-educated workers’ incomes, this has strongly contributed to the overall increase in wage inequality. Two main explanations, routinization and consumption spillovers, have been proposed to explain this occupational shift. Although the determinants of these two theories are highly spatially correlated, studies that have exploited regional variations to identify these mechanisms have so far only considered each explanation in isolation, raising confounding concerns. I highlight these concerns and provide reduced-form evidence that both theories operate simultaneously to drive growth in service employment. To strengthen my case, I extend the structural framework proposed in Autor and Dorn (2013) to include consumption spillovers through non-homothetic preferences. I estimate key parameters and assess the relative importance of each theory using simulations from the model. Relative to a model featuring homothetic preferences, my specification yields 57% more regional disparities in the growth of service occupations, which can be interpreted as the contribution of consumption spillovers. While the routinization hypothesis quantitatively dominates, my reduced-form and structural evidence point to sizeable consumption spillover effects that cannot be neglected.My third chapter studies the effects of liberalizing the use of short-term contracts on the labor market outcomes of all workers. It specifically examines the impact of a 2003 change French jurisprudence. This decision from the French Civil Supreme Court — hereafter called the “reform” — extends the scope of a specific type of temporary contract in France, the CDDU (contrat � dur�e d�termin�e d’usage), to jobs that are not necessarily temporary by nature. This type of contract is allowed in 16 service sectors and is not restricted in terms of length or number of renewals, giving a lot more flexibility to employers than the standard temporary contract does (CDD, contrat � dur�e d�termin�e). I find that this change is associated with an increase in the share of temporary contracts of 2.9 percentage points in the CDDU sectors, while other service sectors only show a 0.7 point increase. This result echoes a frequent finding of temp for permanent substitution in the literature about the deregulation of temporary contract. A second and novel finding is that this reform weighs on young permanent workers’ wages, who experience drop of nearly 4% in their wages over the two years following the reform, and little catch up afterwards. I show some evidence that something changed in the relationship between employers and employees in CDDU sectors: the employer’s bargaining power seems to have risen as the opportunity to substitute temporary contracts for costly and protected permanent ones has increased.

This dissertation consists of three essays in the area of Labor Markets. In Chapter 1, I use a merger in the retail pharmacy sector in Brazil to study the effects of concentration on labor market outcomes. I observe a larger reduction in wages of salespeople than pharmacists. The results are consistent with the idea that salespeople have strong preferences for jobs or accumulate some industry-specific human capital, and that pharmacists are better organized into unions. In Chapter 2, co-authored with Ryan Boone, we use worker flows across occupations, industries, and geographies to define better approximations to labor markets. The defined markets maximize a measure of the density of links within markets. In Chapter 3, co-authored with Ariadna Jou Fuya, we investigate whether the flattening in the college premium in Brazil is due to changes in the average quality of college graduates. We show that the supply of workers with college degree has increased, but much of this increase came from newer, lower ranked and lower wage-premium universities. The college premium has actually increased when we hold constant a set of universities. There are more workers with a college degree, but with lower quality degrees, which reflects into lower average wages.

In chapter 1, this dissertation investigates how promotion policies impact advanced education attainment on employees throughout a career. Using administrative data from the U.S. Department of Defense (DoD), I leverage a natural experiment where academic credentials were permanently masked from promotion consideration. I show the policy influenced the premium of a graduate degree for career advancement opportunities. I then exploit an event study and difference-in-differences design to find employees are 14 percentage points less likely to complete a master's degree when education credentials are no longer considered for promotion. I also show attendance at for-profit institutions declined substantially after the policy reform. In chapter 2, I examine workplace peer effects on advanced education attainment. Using administrative data from the DoD and conditional exogeneous assignment of employees to workplace, I provide empirical evidence that coworkers’ education investment behavior increases the likelihood of using Employer-Provided Educational Assistance (EPEA) benefits to start and eventually complete a graduate degree. In chapter 3, I analyze trainees' ordinal rank across cognitive and physical ability within an initial job training program. Using administrative data from the DoD and conditional random assignment of trainees to peer groups, I test whether rank effects vary across contemporaneous training and long-term career outcomes. I find cognitive ordinal rank, measured by an individual's score on the Armed Forces Qualification Test (AFQT), has a meaningful impact on completing initial training into the U.S. Air Force (USAF). I also show physical ordinal rank, measured by an individual's initial fitness score, also affects initial job training performance. Both sets of ranking effects impact follow-on disciplinary outcomes and vary by gender.

This dissertation focuses on the job searching and matching in online labor markets and explores research topics including workers' job search behaviors, employers' recruitment decisions, and the role of internet job platforms in online job matching, based on the field experimental data and internal data from job boards. In the first chapter, joint with Peter Kuhn and Kailing Shen, we study how explicit employer requests for applicants of a particular gender enter the recruitment process on a Chinese job board, focusing on two questions: First, to what extent do employers’ requests affect the gender mix of a firm’s applicant pool? Second, how ‘hard’ are employers’ stated gender requests-- are they essential requirements, soft preferences, or something in between? Using internal data from a Chinese job board, we estimate that an explicit request for men raises men’s share in the applicant pool by 14.6 percentage points, or 26.4%; requests for women raises the female applicant share by 24.6 percentage points, or 55.0%. Men (women) who apply to gender-mismatched jobs also experience a substantial call-back penalty of 24 (43) percent. Thus, explicit gender requests do shape applicant pools, and signal a substantial but not absolute preference for the requested gender. The second chapter, based on joint work with Peter Kuhn, Taoxiong Liu and Kebin Dai, studies how workers make voluntary wage disclosure decisions in the job search process using internal data from a leading online Chinese job board, Liepin.com. We find that on average, workers' disclosure decisions are consistent with a model in which high current wages are seen as "good news" by prospective employers: Workers are more likely to disclose their wages when their wages are higher than might be expected, based on the worker's resume and where they applied. Employers' responses to workers' resumes, however, are hard to reconcile with these disclosure patterns: While employers respond positively to workers with higher-than-predicted current wages, they do so equally, regardless of whether those wages have been disclosed. This suggests that firms can infer the unobserved ability associated with a worker's current wages from other aspects of her resume and application behavior. Finally, the act of disclosing one's current wage --regardless of its level-- appears to reduce firm's interests in hiring a worker. Disclosures of low wages (which are rare) appear to be mistakes (because they reduce both application success rates and offered wages); disclosures of high wages may, however, benefit workers by filtering out unwanted low-wage job offers.The third chapter investigates gender bias in job recommender systems. By conducting an algorithm audit in four Chinese job boards, I find that gender-specific jobs, which are only displayed to one gender, account for 9.72% of the total recommended jobs to identical male and female applicants. Gender-specific jobs differ in both the job's explicit quality and the words used in job descriptions: Compared to jobs that are only recommended to men, only-to-women jobs propose lower wages, request fewer years of working experience, are more likely to require literacy skills and administrative skills, and tend to contain words related to feminine personality, which reflect gender stereotypes in the workplace. Item-based collaborative filtering, content-based recommendation algorithms and the hiring agents' behaviors incorporated in job recommender systems are the possible drivers of the gender bias in job recommendations.To social science researchers, the recommendation algorithms used by job boards to recommend jobs to workers are a proprietary ‘black box’. To derive insights into how these algorithms work, we conduct an algorithmic audit of four Chinese job boards, where we create fictitious applicant profiles and observe which jobs are recommended to profiles that differ only in age and gender. We then estimate the cumulative effect of pursuing these recommended jobs by applying to them in up to three rounds of successive applications. Focusing on the jobs that were recommended to just one of the two genders that applied, we find that only-to-women jobs propose lower wages, request fewer years of working experience, and are more likely to require literacy skills and administrative skills. Only-to-women (men) jobs also disproportionately contain words related to feminine (masculine) personality characteristics, as measured by three distinct approaches for identifying such characteristics. Finally, we assess the patterns in the recommendations generated by our audit study for their consistency with four processes the algorithms could be using: item-based collaborative filtering, content-based matching, matching based on recruiters’ profile views, and rules-based matching based on employers’ stated gender preferences. We find evidence suggesting that the algorithms are relying on all but the last of these processes.

Managers are important for firm productivity (Bloom and Van Reenen, 2007, Lazear, Shaw and Stanton, 2015). Consequently, firms must decide how to assign the right workers to higher-level jobs (Rosen, 1982, Holmstrom and Tirole, 1989). However, little is known about how firms’ organizational design affects different groups of workers and whether it can explain why women continue to be underrepresented in higher-level positions. In this dissertation, I collect a unique dataset in collaboration with a large multinational firm in order to provide new insights on the key impediments that women face in their career progression.In the first chapter, I focus on the role of managers for workers’ career progression, which is motivated by the fact that most organizations rely on managers to identify talented workers for promotions. However, managers who are evaluated on team performance have an incentive to hoard workers. I find that talent hoarding reduces workers’ applications for promotions by more than half, with particularly large effects on women. Marginal female applications who would have not applied under talent hoarding are twice as likely than their male counterparts to land a promotion and three times as likely to outperform their team at a higher level. Talent hoarding thus contributes to misallocation of talent and perpetuates gender inequality in representation and pay.The second chapter builds on these findings and attempts to identify when and why gender differences in representation first emerge in the leadership hierarchy. The rich personnel records allow me to construct a new and granular measure of job hierarchy that captures career progression throughout the job ladder. I find that the transition to first-level leadership positions represents a key bottleneck in women’s career progression. This early leadership gap is not fully explained by employee characteristics and is driven by internal promotions, not differential entry to or exit from the firm. Women who make it to the first-level leader- ship level are not less likely to get promoted than men, which contrasts the common notion of a glass ceiling.The third chapter demonstrates why women are less likely to advance to first-level leadership positions than men. Identifying the drivers of female underrepresentation is difficult, because the promotion gap can arise due to both labor supply and labor demand factors. By combining rich personnel records and the universe of application and hiring decisions at a large multinational firm, I am able to analyze employees’ labor supply decisions separately from the firm’s labor demand decisions. I find that women at lower hierarchy levels are less likely to apply for promotions to first-level leadership positions than observationally similar men, but do not experience lower hiring likelihoods than men. Using detailed information on every internal job opening in employees’ choice sets, I show that preferences for leading a team are a key determinant of the gender gap in applications for promotions. These gender differences in preferences for team leadership are not fully explained by other factors, including correlated job features such as flexibility and skill requirements or the gender composition of the coworkers associated with a job opening.Together, these findings imply that the organizational design within firms may have large impacts on gender differences in career progression, deterring high-quality women from climbing the job ladder and contributing to misallocation of talent. While all three chapters use the same setting and data, each chapter is intended to be a stand-alone set of research questions, so the respective description is included within each chapter.

This dissertation consists of three essays that study higher education effectiveness and budget decisions. In the first chapter, I evaluate whether instructional spending in colleges, the main input in the human capital production process, is effective in promoting student earnings after graduation. Using a nationally representative dataset, I find a positive elasticity between instructional spending and student earnings of around 2 percent which is robust to various specifications and potential confounding factors. The effects are mostly driven by private institutions and four-year institutions. The effects are slightly lower for well-established public institutions. Cost-benefit analysis reveals that a student has to work for over 40 years until her increased earnings can cover the cost of the spending, indicating the cost-ineffectiveness of increasing instructional spending from an investment perspective.In the second chapter, we analyze one driving factor for college budget decisions - the competition for good students. We construct a model to show that colleges choose their share of spending on educational inputs and on amenities to attract good students. Their optimal decision depends on their competitors' choices. Simulation results reveal that over the range where budget choices are commonly made, colleges respond positively to their competitors' decisions. In addition, when the competitive pressure from competitors weakens, such responsiveness also declines. Empirical evidence generally supports the predictions from the simulation exercise.In the third chapter, we propose a method that bounds the treatment effect for the general population of interest under randomized controlled trials. Our bounds depend on a set of mild assumptions that are different from existing methods in the literature. Hence, we view our method as an alternative for applied researchers to choose from. Applying our method to analyze the effectiveness of a first-year learning program on retention rates, our bounds suggest a possible negative population effect and restrict the possibility of a large positive population effect, despite the estimated positive effect for the experimental sample. We also show that assumptions required for other methods in the literature to be applicable are unlikely to hold in this context, highlighting the importance of the availability of our new method.

In this dissertation I study the long-term effects of moving to a high-poverty neighborhood on earnings and schooling using evidence from a slum clearance program implemented in Santiago, Chile, between 1979 and 1985. During the country’s dictatorship, the government mandated the eviction of entire slums and their relocation to public housing in low-income areas: Two-thirds of slums were relocated to new housing projects on the periphery of the city, and the rest received housing at their initial location. This dissertation is divided into three chapters. In Chapter 1, I estimate a displacement effect by comparing the outcomes of displaced and non-displaced children and their parents 20 to 40 years after the end of the policy. I construct a novel data set that combines archival records with administrative data containing 19,852 homeowners matched to 55,343 children. I show that displacement is unrelated to families’ demographics or neighborhood attributes prior to eviction, and I find negative effects on children and families: Displaced children have 10% lower earnings and 0.5 fewer years of education as adults than non-displaced. Moreover, displaced children are more likely to work in informal jobs and their parents are more likely to die after the intervention. In Chapter 2, I study the mechanisms that explain the negative effects experienced by displaced children when young. I find that destination characteristics mediate my results: Lower social cohesion in destination projects reduces children’s schooling, and their earnings are also affected by worse labor market access. Finally, in Chapter 3, I describe the data collection process in detail. I explain the processes to construct family composition and how I link parents and children. I also describe how I match families to slums of origin, and I discuss potential selection issues due to missing data.

This dissertation consists of three essays in labor economics. Broadly, they explore the labor market responses of firms and workers to various economic policies. The first chapter uses a lab experiment to study the impacts of salary history inquiry bans on wage bargaining. I find that these laws may not work as well as intended. The second chapter uses text analysis of reviews to explore the impacts of minimum wage changes on restaurant prices, hygiene, staff friendliness, wait times, and portion sizes. I find evidence of a rise in prices and improvements in staff friendliness with higher wages. While there is some weak evidence for deteriorating hygiene standards, there are no detectable impacts on portion sizes or wait times. Overall, these changes are associated with a small but significant drop in restaurant ratings. The third chapter explores the early impacts of the Opportunity Zone program on residents of targeted areas. The Opportunity Zone program, created by the Tax Cuts and Jobs Act in 2017, was designed to encourage investment in distressed communities across the U.S. We leverage restricted-access microdata from the American Community Survey and employ a matching approach to estimate causal reduced-form effects of the program. Our results point to little or no evidence of positive effects of the Opportunity Zone program on the employment, earnings, or poverty of zone residents.

In my first chapter, I construct a novel measure of misogyny using Google Trends data on searches that include derogatory terms for women. I show that misogyny is an economically meaningful and statistically significant predictor of the wage gap, and use it to test the predictions of two influential labor market discrimination models. I find that the gender wage gap is inconsistent with the Becker model of taste-based discrimination, but that it fits Black's search model of discrimination which allows for discrimination from even a small group of misogynists to result in a wage gap. In my second chapter, I investigate the impact of COVID-19 lockdowns on domestic violence by leveraging Google Trends data on search interest in domestic violence resources. I find that COVID-19 lockdown orders result in a meaningful and statistically significant increase in search interest in domestic violence hotlines but a decrease in search interest for other resources such as protective orders and domestic violence shelters, which require victims to leave their homes in the midst of a global pandemic. Moreover, I find that increased exposure between victim and perpetrator exacerbate domestic violence and that economic hardship and a lack of independent earning potential make women especially vulnerable. Finally, I find that search interest in terms likely to be searched by third-party reporters drastically decrease as a result of COVID-19 lockdowns, suggesting that external mechanisms for identifying and reporting domestic violence are affected by lockdown orders.

This dissertation employs quasi-experimental methods and recent advances in econometrics to uncover the causal links and policy levers that are relevant to gendered criminal justice and racial bias in policing. Chapter 1 evaluates the consequences of a policy that deputizes landlords to internalize public safety externalities; nuisance property ordinances. These ordinances label a property as a nuisance when police respond to a home a set number of times within a certain period. After having a property declared a nuisance, property owners who do not abate the nuisance can face fines and criminal charges. Therefore the concern is that this policy may incentive landlords to retaliate against tenants who report crimes, thus raising the cost of crime reporting which is especially worrisome for domestic violence where the cost of reporting is already high. I collect data on implementation of these ordinances across major cities in the U.S. and consider what happens to domestic assaults when these ordinances are passed. I exploit the variation in timing of the implementation of these ordinances and find that this policy reduces the rate at which domestic assault is reported to police. Another wing of my work examines variation across a variety of criminal justice contexts local to Nashville, TN. The first of these papers, Chapter 2 (joint with Emily Owens and Kerri Raissian) provides one of the first credibly causal evaluations of specialty courts in the criminal justice context, and there are two key innovations. Relative to the small amount of existing research, which explores how cases change after the introduction of a court, we are able to identify how decision making and case outcomes change in cases set before a well-established fully operational specialty court. Second, we are able to examine not only the impact of these courts on offenders, but also on victims. We find that examining both reoffending and revictimization provides a substantially different picture of the impact of these courts than reoffending alone; what initially appears to be a “null” impact on recidivism is revealed to be the result of both an increase in victim safety and willingness to involve the court when future violence does occur. In Chapter 3, joint with Yoonjung Kim and Jaclyn Rosenquist, we utilize a unique data set and exploit the variation of arrestees’ location to study the causal effect of being charged with the drug-free school zone enhancement. we find that enhancement charges against eligible offenders are more likely to be applied to Black arrestees than White arrestees, leading to lasting impacts across races in sentencing outcomes. We present preliminary evidence of White defendantsbeing offered better plea deals as the mechanism driving these results.

In this project I investigate three sets of policy changes related to overtime penalties, school bond measures, and police training. I study the effects that these policy changes have had on labor market and other outcomes. In the first chapter, I examine how the demand for labor and the categorization of labor change in response to a change in penalties for overtime violations. I propose a model for determining labor demand for fully compensated overtime hours, undercompensated overtime hours, and total hours when employers rationally undercompensate overtime labor. I find evidence that, following a decrease in the expected penalty for state overtime violations in Massachusetts, the average number of undercompensated and fully compensated overtime hours worked increased, while the average number of non-overtime hours worked decreased. There is little evidence that employers responded to the change in penalties by reclassifying workers as exempt or nonexempt from the overtime law.In the second chapter I use multiple methods to examine how school bond measures, and the accountability measures associated with the bond measures, affect home prices and teacher labor market outcomes. I find evidence that school bonds passed with a 55% passage threshold are correlated with increases in home prices and average teacher salaries. The findings are reversed for bonds passed with a two-thirds passage threshold, which are correlated with reductions in home prices and teacher salaries. The findings suggest that the accountability measures for school bond measures do influence the way the bonds are used, and have downstream effects on home prices and teacher labor market outcomes.In the final chapter, I evaluate how different types of police training may affect the conduct of police officers. I find that increased Peace Officer Standards and Training (POST) funding leads to a small temporary decrease in civilian complaints sustained. Increased POST funding is also correlated with decreases in police shootings in each of the following three years, though these results are statistically indistinguishable from zero. Differences between how POST and non-POST training programs cover various topics—such as those related to mental health, professionalism, and rights—may explain these results. Policymakers may look to investing in increased police training funding from the state to improve policing outcomes.

This dissertation contains three essays studying topics in applied microeconomics. The first chapter is a co-authored paper in which we use daytime satellite imagery and convolutional neural networks to model economic growth at the neighborhood level. In the second chapter, I use this model to examine the spatial distribution of residential impacts from fracking. The third chapter investigates methods of measuring skill distance between occupations and proposes a new method which matches patterns of observed occupational transition. Each chapter uses unconventional data sources and machine learning techniques to contribute to central questions in labor economics research and policy. In the first chapter we apply deep learning to daytime satellite imagery to predict changes in income and population at high spatial resolution in US data. Our model predictions achieve R2 values of and 0.32 to 0.46 in decadal changes, which have no counterpart in the literature and are 3-4 times larger than for commonly used nighttime lights. Our network has wide application for analyzing localized economic shocks. One such application is my second chapter, which studies changes in total neighborhood income and population in areas near fracking extraction and shale reserves. My microspatial approach identifies that fracking exposure as far as 20 miles away leads to a 2 percent decline in neighborhood income. The spatial gradient and associated mechanisms of this effect indicate that it is driven by local industrialization rather than direct environmental externalities. Examination reveals margins of policy and labor conditions which attenuate the observed impacts. In the third chapter I show that a regression framework generates a novel, empirical occupational skill distance norm which is disciplined by observed occupation switching patterns. This approach relieves key limitations of existing measures such as linearity and symmetry. It also allows for an analysis of which skill dimensions relate to the portability of human capital, and which do not. Implications for existing results on skill portability are discussed, along with immediate policy applications on employee adjustment costs.

This dissertation reflects research on the US labor markets over the course of large macroeconomic events. Each chapter studies a particular facet of how to understand movements in labor markets and individual lives as the macroeconomy shifts. In my first chapter, I introduce and motivate each of the subsequent chapters as essential for furthering our collective understanding of human well-being in an ever evolving and fluctuating macroeconomy. In the second chapter of my dissertation, I study the impact of technological change on workers by examining the adoption of the tractor in American agriculture. The paper uses complete count census data at both the county and individual level to study the impact of farm mechanization in the first half of the twentieth century. I examine the individual-level outcomes and adjustment mechanisms for impacted individuals. I find that the generation of farm workers that was older than 18 when the tractor was adopted saw large disemployment effects and an exodus from the labor market, though no strong migratory patterns. In contrast, youth who grew up on farms were likely to stay in farm work and do not show strong propensity to migrate out of their county of origin. I conclude that changing technology requires new tasks and skills, making it difficult for incumbent workers to adjust. New entrants, however, have the capacity to adapt to the changing work requirements.In the third chapter, co-authors and I show that the standard implementation of X13-ARIMA yields predictable revisions to initial releases of employment growth. At the onset of severe recessions, initial growth rates tend to be revised downward. This causes the severity of recessions to be underestimated in real-time-- a major concern for policy-makers. The problem arises from the combination of concurrent seasonal adjustment, and the standard forecasting procedure used as part of the seasonal adjustment algorithm, when applied to abrupt movements such as those that occur at the onset of a recession. This problem was avoided in the COVID crisis only due to the triggering of outlier detection clauses, which counteracted this effect.In the fourth chapter, I study the empirical accuracy of home production in macroeconomic models. I begin by building a simple RBC-style model with home production and two income earners. I solve and simulate the model to demonstrate how shocks to productivity may impact time allocation between two spouses in the home. I then use the American Time Use Survey to study time allocation in response to the Great Recession. I use state-level variation in the unemployment rate over the Great Recession to measure how men and women's time spent on market work, home production, childcare and leisure varied with labor market conditions. I find that men's market work does decline with state-level unemployment and that men's home production and childcare do increase. Women's time use does not appear to change. Finally, I test the model by estimating how men's and women's time in home production and market work relate. I find that contrary to the prediction of the model, men's and women's time in home production do not co-move. This finding calls for further work to understand how to better model time allocation and substitution within a household.

This dissertation studies the implementation of large-scale public policies in developing countries. Chapter I focuses on how collective action can shape public policy. In particular, I study how organized groups of workers affected the land reform implemented by Salvador Allende ́s government in 1970 in Chile. Using an event-study design, we find that the local political action of workers - proxied by land invasions - affected the intensity and location of expropriations. In Chapter II, I change time and location and study how political favoritism affected the expansion of the electric grid in Kenya around the 2013 and 2017 presidential elections. While the aggregate political bias we estimate is meaningful, it is smaller in magnitude than that documented previously, suggesting that increasingly active independent media scrutiny as well as increasingly robust democratic institutions, expanding constraints on executive power, and donor oversight may have partly curbed favoritism. Finally, in Chapter III I study the effects on welfare of a large-scale public investment: the expansion of the subway network in Santiago, Chile between 2002 and 2020. We find that when workers receive access to the subway network, they start taking jobs further away. We see an increase in wages for benefited workers, including those workers who do not switch firms, suggesting a reduction in the labor market power firms have over workers. Our model estimates suggest incorporating labor market power can make an important difference in welfare estimations.

My dissertation's primary contribution is identifying factors that affect differences in labor market outcomes among various demographic groups.In the first chapter, I designed an experiment to examine the effects of stereotype threat, a concept from psychology where people conform or feel at risk of conforming to negative stereotypes about their group, under competition. I found that providing information about performance differences as a stereotype threat changed the subjects’ preferences to favor the tasks that their gender is perceived to perform relatively better in but did not affect performance outcomes. The implication is that even if stereotype threat does not affect one’s ability to perform, it may affect education and career choices that could contribute to group differences in the labor market.The second chapter was written jointly with Dr. David Neumark and published in Demography. This paper analyzed the effects of cohort sizes on the labor force participation and wages of older workers in the United States. Older workers increased their labor force participation as their relative size to the working age population increases, which is contrary to the standard labor supply hypothesis. Additionally, when using a richer model that accounts for the size of older workers relative to younger workers, we found the demand for older workers was high when their cohort size is large relative to prime age workers, suggesting that older workers enter in more flexible working arrangements in their later age.Finally, the last chapter analyzed the effects of expanding public transit infrastructure on labor market outcomes in Los Angeles. I use panel data on tracts, treating route placement as endogenous, which is then instrumented by the distance from the centroid of each tract in LA to a hypothetical Metro route. Overall, I find proximity to Metro stations increases labor force participation and employment for residents, which is robust to using both a binary and continuous measure of distance. Additionally, I find evidence that increased job density in neighborhoods near new transit stations is contributing to the employment increase.

This dissertation is comprised of four chapters. Chapter 1 is a reprint of my job marketpaper “Estimation and Inference with Transfer Learning.” In this chapter we propose andstudy machine learning algorithms which utilize “transfer learning,” and their application toimproving semi-parametric inference in cases of insufficient sample size. We consider multiplemachine learning algorithms, including elastic net and deep feedforward neural networkswith rectified linear unit (ReLU) activation functions, and allow for transfer through an ℓ1penalty term in the loss function which shrinks the estimates towards auxiliary estimates.Novel results on error bounds and convergence rates are established to justify the usage ofthese algorithms as nuisance function estimators for double machine learning. We evaluatethe usage of these algorithms in conducting valid inference on treatment effects with MonteCarlo simulations and an empirical application on the Job Corps training program. Ournumerical results show that transfer learning can substantially improve estimation in thepresence of sizeable missing data.Chapter two is comprised of an analysis of the effects of minimum wages on unemploymentduration and re-employment outcomes. Using the Survey of Income and ProgramParticipation, we build a sample of unemployment spells to study how the minimum wageaffects several outcomes of the unemployed. We establish that the policy matters for theunemployed in two ways. A higher initial level of the minimum wage (at the start of a spell)leads workers to abandon their job search but has mostly null effects on other outcomes.However, being unemployed at the time the minimum wage is raised is associated withlonger spells, a higher rate of search quitting, and fewer working hours after re-employment.Chapter three is derived from a joint work of mine with Ying-Ying Lee entitled ”DoubleMachine Learning Nonparametric Inference with Continuous Treatment.” In this chapterwe numerically evaluate the effectiveness and characteristics of the double machine learningcontinuous treatment estimator discussed in Colangelo and Lee 2020. We conduct simulationsusing a variety of machine learning algorithms such as lasso, random forests andneural networks (and variations of these), and also provide an empriical application usingdata from the Job Corps program. The estimator is fairly robust to the choice of machinelearning algorithm in the presence of cross-fitting, with all algorithms attaining near perfectcoverage rates and unbiasedness. Some algorithms perform well even without cross-fitting,indicating that the procedure can be skipped in particular cases. The empirical results aresimilar regardless of which algorithm is used, and are consistent with previous research onthe Job Corps program.Chapter 4 discusses a new estimator for estimation and inference in long term panelswith interactive effects. We modify the Common Correlated Effects (CCE) approach ofPesaran (2006) to produce a simpler and more computationally expedient estimator thanthe original CCE estimator. While CCE substitutes the factors with cross sectional means,the new method which we call Two Way CCE (TWCCE) substitutes out the individualspecific factor loadings in addition to the factors themselves. This conveniently reducesthe estimation problem to simple least squares without the need to estimate heterogeneouscoefficients on each factor. We investigate the performance of TWCEE in comparison withthe other most common factor model estimators such as the Interactive Effects Estimator(IFE) of Bai (2009) and the Augmented Mean Group Estimator (AMG) of Eberhardt andTeal (2010). We show that TWCEE has similar performance to the other methods, whilealso demonstrating the least computation time.

This dissertation focuses on the effect of government policies on disadvantaged groups. In the first two chapters, I study the effect of occupational licensing policies that disparately impact immigrants. In the third chapter, I study the effect of Medicaid expansion under the Affordable Care Act on the use of online medical crowdfunding platforms.Occupational licensing is a labor market institution that restricts the right to practice an occupation to workers in possession of a recognized credential (Friedman 1962). Kleiner and Kruger (2013) estimate that almost 30% of the workforce is licensed. While licensing has been justified on the grounds of asymmetric information about work quality, most research finds that licensing restricts labor supply, with ambiguous effects on quality. Moreover, occupational licensing may have a disparate effect on disadvantaged groups. A growing literature examines the effects of licensing on different types of workers, such as immigrants, women, and African Americans (Dorsey 1983, Law and Marks 2009, Blair and Chung 2018, Gomez et al 2015, Federman et al 2006, Kugler and Sauer 2005, Peterson et al 2014, Runst 2018, Brucker et al 2018, Cassidy and Dacass 2019, McDonald et al 2015, Tani 2018). In the first chapter of my dissertation, I exploit a recent California law to study the effects of licensing restrictions on undocumented immigrants. In the US, undocumented immigrants are barred by federal law from obtaining occupational licenses. This is a much stronger barrier to entry than occupational licenses represent for natives and legally present immigrants. In 2016, the State of California began to allow undocumented immigrants to apply for occupational licenses. I combine this policy variation with a difference-in-difference methodology and survey and administrative data to estimate the effect of legal access to licenses for undocumented immigrants. I find that licensing reform increased access to licenses. Using administrative data from the California Department of Consumer Affairs, I find that after 2015, the number of licenses granted to people with Hispanic names increased 10% relative to non-Hispanic whites. Using data from the Current Population Survey, I find that self-reported licensing increased 2 percentage points among likely undocumented workers in California relative to other groups. Access appeared to increase earnings for undocumented workers: comparing California to other states, earnings increased more after the reform for likely undocumented immigrants working in more heavily licensed occupations. Moreover, this increase in earnings was not at the expense of other groups; there was no concurrent decrease in earnings for natives or naturalized immigrants. My results have two important implications. First, undocumented immigrants are meaningfully constrained by federal licensing policy. Second, state governments may be able to expand access to licenses for undocumented immigrants without adversely affecting other workers.These findings are useful for policymakers working on occupational licensing reform. In 2019, Nevada and Oregon passed laws similar to California’s policy, with New Mexico and New Jersey following suit in 2020. Many other states have enacted policies providing access to occupational licenses specifically for DACA recipients. Moreover, occupational licensing in general remains an active area for state legislatures, with 449 laws relating to occupational licensing enacted by state legislatures in 2019 (National Congress of State Legislatures 2019). Understanding the effects of these policies is essential to creating effective occupational licensing policy in the future. In the second chapter, I estimate the effect of occupational licensing laws on immigrant participation in various occupations in the US from 1850 to 1940. The effect of occupational licensing on access to professions for disadvantaged groups is theoretically ambiguous. While economists typically view licensing as a barrier to entry which should reduce labor supply, some scholars suggest that licensing may make professions more accessible to disadvantaged groups by creating clear and objective standards for entry (Redbird 2017). Law and Marks (2009) find that black and female employment increased in an occupation after it became licensed, though these results have been subject to some debate (Klein et al. 2012, Law and Marks 2012). It is an open question whether occupational licensing facilitated discrimination perpetuated by licensing boards or rather mitigated it.Using the staggered adoption of licensing for fifteen occupations initially licensed in most states during the Age of Mass Migration, together with difference-in-differences empirical strategies and historic Census data, I estimate the effect of licensing on the share of foreign-born workers in a given occupation-state pair. I find that, after licensing, the share of foreign-born workers increases 3-5 percentage points. These results are fairly consistent across alternative estimation strategies, including recent difference-in-differences methodologies that are robust to heterogeneous and dynamic treatment effects (Chaisemartin and D’Haultfœuille (2021)). I also examine the heterogeneous effects of occupational licensing across occupations. I use linear probability models to examine the effect of licensing on the probability that an immigrant works in an occupation relative to native workers. Consistent with conflicting theories of the effect of licensing on occupational attainment, I find mixed results. For some relatively small occupations (accountants, beauticians, dentists), I find that licensing reduces immigrant representation in the occupation. However, for larger occupations (physicians and possibly teachers and barbers), I find some evidence that licensing increases the participation of immigrants. In contrast with prior research (Law and Marks 2009), I do not find any evidence that occupational licensing has more positive effects on immigrants in occupations that feature asymmetric information. In the third chapter, I address the question: do Medicaid expansions crowd-out private donations for medical expenses? Using a novel data set of over 180,000 fundraisers scraped from GoFundMe, a crowdfunding website, I compare changes in crowdfunding activity in states that passed Medicaid expansions with states that do not. I find no effect of Medicaid expansion on intensive margin fundraising outcomes (funds raised and fundraiser goal). I do find large and statistically significant effects of Medicaid expansion at the extensive margin; after expansion, fundraisers declined by nearly 20% in states that expanded Medicaid. These results are robust to estimation methods that allow for heterogeneous treatment effects and staggered treatment adoption (Chaisemartin and D’Haultfœuille’s 2021). This research is important for understanding the role of crowdfunding in the medical finance ecosystem, and more broadly for understanding how big the problem of public transfers crowding out private insurance is for the ACA Medicaid expansion.

This dissertation investigates key aspects of the gender wage gap in the U.S.The first chapter studies whether there is a gender \ask gap", that is whether womenwith similar resume to men's ask for lower wages. This analysis relies on the innovativerecruitment process of an online recruitment platform for full-time engineering jobs in theUnited States. To use the platform, job candidates must post their resume information aswell as an ask salary, stating how much they want to make in their next job. Using dataon more than 120,000 candidates over several years, I first document a 7.2% raw ask gap onthe platform. After controlling for all the candidates' resume characteristics, the ask gap is3.3%. In other words, women with resumes comparable to those of men ask for 3.3% less.The second chapter investigates the relationship between the gender ask gap and thegender wage gap. In this chapter, I leverage the fact that, on the recruitment platform, firms apply to candidates by offering them a bid salary, solely based on the candidate's resumeand ask salary. Additionally, if the candidate is hired, a final salary is recorded. Using dataon more than 510,000 bids, I find a raw bid gap on the platform of 3.4%. Adjusting forcandidates' resume characteristics but excluding their ask salary leaves a 2.4% residual bidgap. When candidates' ask salaries are included as a control, this residual bid gap disappears.In other words, while resume characteristics can only reduce the bid gap by 30%, genderdifferences in ask salaries can essentially explain 100% of the bid gap. For the sub-sampleof 8,333 hired candidates, gender differences in ask salaries explain nearly all of the gap infinal offers. In particular, while conditioning on resume characteristics only narrows the finaloffer gap to 1.8%, adding the ask salary to the controls yields an insignificant final offer gapof -0.5%. I find no evidence of discrimination against women at the extensive margin. Infact, conditional on their resume characteristics, women get slightly more bids than menand, conditional on interviewing, women are just as likely as men to get a final offer.The third chapter estimates the market-level effects of an increase in women's ask salaries,leveraging an unanticipated feature change on the platform. In mid-2018, Hired.com abruptlychanged the way that some candidates were prompted to provide their ask salary. From thefirst year of the data to mid-2018, candidates stated their ask salary by filling out an emptytext box. Starting in mid-2018, the answer box for San Francisco software engineers was prefilled with the median bid salary over the past 12 months for the candidate's combination ofdesired location, job title, and experience. In effect, this change gave candidates informationon the typical offers received by similar candidates on the platform and provided them withan anchor to benchmark their own ask salary. Using an interrupted time series design, I showthat the new framing of the ask salary elicitation eliminated the ask gap and rendered thebid gap insignificant. These results are mostly driven by women asking for higher salariesafter the reform. Further, I find no discernable impact on the number of bids that womenreceived or the time it took women to receive their first bid, suggesting that women hadeffectively been leaving money on the table.

The last two centuries have seen dramatic shifts in labor and demography: a precipitous drop in fertility rates, a dramatic increase in human capital, and the economic empowerment of women. Economic theorists often hypothesize that these transformations played a vital role in the transition to modern economic growth. This dissertation uses historical vital records from the Canadian province of Quebec and applied microeconomic methods to test the assumptions that underpin these theories. First, it shows that parents did not target a family size before the demographic transition. Second, it shows that family size only had a modest impact on the accumulation of human capital. Third, it shows that assortative marriage, unlike family size, mattered long before many might expect. Quebec, with its unusual demographic regime and exceptionally high quality demographic data, provides solid empirical evidence to support these claims. The IMPQ (l’Infrastructure intégrée des microdonnées historiques de la population québécoise) is a large new database of family reconstitutions from baptism, burial, and marriage records (IMPQ 2020). It integrates two previous databases, the BALSAC database and the RPQA (Project Balsac 2020, PRDH 2020). While the dataset is still being extended, as of writing it contains 1.8 million unique births, 0.8 million unique deaths, and 4.2 million unique marriages from 1481–1992 (though births and deaths are limited to a particular region after 1849, there are very few records from before 1620, and the births and marriages are only available through 1971). Moreover, in those records a total of 2.8 million other individuals are mentioned besides the main participants, which provides additional observations over time for many people besides their own vital events. The first chapter of my dissertation, co-authored with Gregory Clark and Neil Cummins and published in Demography, considers the demographic transition. Was this decline in fertility due to changing economic incentives or was it a more fundamental change in the ability or willingness of parents to control fertility? To bring contemporary methods to this old debate, we use the natural experiment of twins to test if pretransition parents targeted an optimal family size. We find strong evidence of no fertility response to an additional child in England, France, and Quebec before the demographic transition. This research led directly to the second chapter of my dissertation. A dramatic increase in human capital was crucial for the transition to modern economic growth. It is often theorized that a key mechanism behind this increase was a trade-off between the quantity of children and the quality of their human capital. However, there are few papers that have estimated the size of the trade-off in a society just on the cusp of modernity. I use twins as a source of exogenous variation in quantity to estimate the trade-off in Quebec 1620–1850. I find that one additional child born decreased the literacy rate (proxied by signatures) of their older siblings by 0.6 percentage points. While statistically significant and robust, this estimate is too small to permit a large increase in human capital. The trade-off present before Quebec industrialized appears insufficient to initiate modern economic growth by itself. While the findings of the first two chapters suggest that parents had little influence on child outcomes through fertility decisions, the third chapter shows that they had a significant influence through their choice of spouse. The final chapter in my dissertation asks the question: when did marriage become assortative on economic ability? While past research has considered if rising female labor force participation increased assortment, we know surprisingly little about marriage matching from before the late 20th century. Using a large new dataset of vital records from Quebec, I find that marriage was as assortative in the 1830s as it was in the 1960s. Not only was this assortment strong, it was on the human capital of individual men and women, not merely between families or social classes. Finally, I show that the human capital of both men and women mattered equally for child outcomes, confirming that assortment mattered for social mobility long before the late 20th century.