This paper documents a new stylized fact about the United States labor market: internal migration rates are dramatically different across college majors. For some college majors, migration rates are even lower than those without a college degree. I relate major migration rates with majors' spatial concentration and find that a major's spatial concentration explains about one fourth of the cross-major variation in migration rates. With this descriptive evidence as a guide, I estimate a structural model of locational choice where college graduates have heterogeneous preferences---at the detailed major level---for living close to home, and for working in a location with a high concentration of their fellow majors. Using estimates of the structural model, I decompose the cross-major migration rates into supply and demand factors and find that supply factors (i.e. moving costs) explain the vast majority of differences in migration rates across majors. My findings underscore the difficulty in attracting college majors to a particular location using demand-side investments. My results also highlight the importance of place in determining the labor market outcomes of college majors.
While the severity of armed conflicts has been increasing recently, there is little evidence on the causal impacts of the end of conflicts on local labor market. This paper provides the first evidence on the local labor market effects of peace exploiting the sudden and unexpected end of the Aceh Insurgency in Indonesia. I find that the end of the Aceh Insurgency significantly reallocated short-run female labor market activity to male labor market activity within the household in two years. The effects were driven by the decrease in violence and economic hardship after the end of the Aceh Insurgency. The negative effects of peace on female labor market activity were stronger for young females and the females in households with a farm. The 2-year short-run effects were not significantly different from the 5-year and 10-year effects of peace. The evidence suggests the importance of the household-level labor market in the post-conflict economy.
This paper studies the effects of teachers’ gender biases on students’ long-term outcomes, including high school completion, college attendance, and formal sector employment. I measure teacher bias using differences in gender gaps between teacher-assigned and blindly-graded tests, and validate the assessment-based measure with novel data on teachers’ attitudes, as captured by the Implicit Association Test (IAT). I develop a large-scale online portal available to teachers and students in Peruvian public schools to collect IAT scores nationwide. This analysis provides evidence that math teachers who strongly associate males with scientific disciplines give higher scores to male students, when compared to blindly-graded test scores, while language arts teachers who strongly associate females with humanities-based disciplines award higher grades to female students. Next, using graduation, college enrollment, and matched employer-employee data on 1.7 million public high school students who were expected to graduate between 2015 and 2019, I find that female students who are assigned to more biased teachers are less likely to complete high school and apply to college than male students. Moreover, female students assigned to more biased teachers in high school are less likely to hold a job in the formal sector after graduation and have fewer paid working hours relative to their male classmates. Exposure to gender-biased teachers also leads to monthly earnings losses for women, further exacerbating the gender pay gap.
This paper studies the limitations of political affirmative action policies. In India, certain state legislature seats are restricted for the historically-discriminated lower castes (Dalits). Dalits are a large and heterogeneous group and there is little understanding of how different castes have been impacted by such enfranchisement, due to a lack of data on the individual caste of beneficiaries. Exploiting the link between names and caste membership, I create a new dataset including the caste of workers involved in a public workfare program (NREGA). Because constituencies are reserved for Dalit legislators based on a population cutoff rule, I use a regression discontinuity design to estimate the effect of having a Dalit state representative on the timing of payments to low-caste laborers in NREGA. I explore this effect on all Dalit workers and differentially by the individual caste of the worker. I find that Dalit workers represented by a Dalit state legislator experience a 12% higher probability of receiving their payments late. This effect is constant across all individual castes, except when considering constituencies won by parties that expressly favor Dalit voters. In this instance, I estimate that Dalit workers receive earlier payments in reserved constituencies and that those belonging to the state’s largest Dalit caste are even more advantaged. The deleterious effects of having a Dalit representative on Dalit workers are borne entirely by areas where the legislator has lower bargaining power over the local bureaucrat who directly manages the processing of payments. Given the high desirability of stable public employment, often these bureaucrat postings attract people from a more advantaged social background, relative to the Dalit legislators. Hence, my findings point to the importance of considering vertical power structures when designing policies aimed at empowering under-represented minorities around the world.
I study the overstatement of GDP growth in autocratic regimes by comparing the self-reported GDP figures to the night time lights (NTL) recorded by satellites from outer space. I show that the NTL elasticity of GDP is systematically larger in more authoritarian regimes. This autocracy gradient in the elasticity is robust to multiple changes in data sources, econometric specification or sample composition and is not explained by potential differences in a large set of country characteristics. The gradient is larger when the incentive to exaggerate economic growth is stronger or when the constraints on such exaggeration are weaker. The results suggest that autocracies overstate yearly GDP growth by as much as 35%. Adjusting the GDP data for the manipulation taking place in autocracies leads to a more nuanced view on the economic success of non-democracies in recent decades and affects our understanding of the effect of changes to foreign aid inflows on income per capita.
We show that sugar-rich diet early in life has large adverse effects on the health and economic well-being of adults more than fifty years later. Excessive sugar intake early in life led to higher prevalence of chronic inflammation, diabetes, elevated cholesterol and arthritis. It also decreased post-secondary schooling, having a skilled occupation, and accumulating above median wealth. We identified elevated sugar consumption across lifespan as a likely pathway of impact. Exploiting the end of the post-WWII rationing of sugar and sweets in 1953 in the United Kingdom, we used a regression discontinuity design to identify these effects.
Early consumption of a product often benefits later consumers by revealing quality information. Inefficiency arises, however, because early consumers do not internalize the social value of their consumption. How could a platform that intermediates information between early and late consumers mitigate such an inefficiency by designing its recommendation policy? In a model with binary product quality and general post-consumption signals, I show that the optimal design features simple threshold policies. The product should be recommended when the platform’s current belief of high quality is above a certain time-specific threshold, which varies in a U-shaped pattern over the product’s life. Characterizations of the recommendation dynamic and comparative statics about the recommendation standards are also provided. My analysis also illustrates the usefulness of a Lagrangian duality approach for a class of dynamic information design problems.
Between 1973 and 1978, the Indonesian government engaged in one of the largest school construction programs on record. Combining differences across regions in the number of schools constructed with differences across cohorts induced by the timing of the program suggests that each primary school constructed per 1,000 children led to an average increase of 0.12 to 0.19 years of education, as well as a 1.5 to 2.7 percent increase in wages. This implies estimates of economic returns to education ranging from 6.8 to 10.6 percent.
We examine the impacts of wearing Adidas shoes on economic outcomes in low-income countries. To study this, the research will use a variety of data sources, including administrative data on income and employment from national statistics agencies, survey data from individuals and small businesses, and economic indicators such as GDP and poverty rates. To identify the impacts of wearing Adidas on economic outcomes, the research will use a fixed-effects model, controlling for individual-level characteristics such as education level and occupation. This will allow us to isolate the effects of shoe brand on economic outcomes, accounting for any unobserved differences between individuals who own and regularly wear shoes from a particular brand and those who do not. Overall, this research will contribute to our understanding of the potential impacts of shoe brand on economic development and inform policy debates on the role of fashion in promoting economic opportunity in low-income countries.
The idea for this research is to examine the possible effects of a universal basic income (UBI) on labor market outcomes. The research will employ a quasi-experimental design, using a sample of individuals from a country or region that has implemented a UBI program and comparing their labor market outcomes to a control group of individuals from a similar country or region that has not implemented a UBI program. To identify the effects of UBI on labor market outcomes, the research will use a difference-in-differences approach, which involves comparing changes in outcomes for the treatment group (those receiving UBI) to changes in outcomes for the control group over the same time period. This will allow us to isolate the effects of UBI on labor market outcomes, controlling for other factors that may affect these outcomes. To measure labor market outcomes, the research will use a variety of indicators, including employment rates, wages, and hours worked. The research will also consider potential spillover effects of UBI on non-labor income sources, such as entrepreneurial activity and asset accumulation. Overall, this research will contribute to our understanding of the potential impacts of UBI on labor market outcomes and inform policy debates on the feasibility and effectiveness of UBI as a means of addressing income inequality and poverty.