Research

I study racial disparities in the criminal justice system by analyzing abnormal bunching in the distribution of crack-cocaine amounts used in federal sentencing. I compare cases sentenced before and after the Fair Sentencing Act, a 2010 law that changed the 10-year mandatory minimum threshold for crack-cocaine from 50g to 280g. First, I find that after 2010, there is a sharp increase in the fraction of cases sentenced at 280g (the point that now triggers a 10-year mandatory minimum), and that this increase is disproportionately large for black and Hispanic offenders. I then explore several possible explanations for the observed racial disparities, including racial discrimination that occurs after entry into the criminal justice system. I analyze data from multiple stages in the criminal justice system and find that the increased bunching for minority offenders is driven by prosecutorial discretion, specifically as used by about 20-30% of prosecutors. Moreover, the fraction of cases at 280g falls in 2013 when evidentiary standards become stricter. Finally, the racial disparity in the increase cannot be explained by differences in education, sex, age, criminal history, seized drug amount, or other elements of the crime, but it can be largely explained by a measure of state-level racial animus. These results shed light on the role of prosecutorial discretion and racial discrimination as causes of racial disparities in sentencing. 

Many claimants of Social Security Disability Insurance (SSDI) retain legal representation to help with the approval process. The Social Security Administration imposes strict rules on representative compensation. Representatives are only paid if claimants are awarded disability, and they are paid the lesser of 25 percent of the claimant's past due benefits or a pre-specified maximum fee ($6,000 since 2009). Because past due benefits are a function of the number of months claimants wait to be awarded, representatives face incentives to delay case resolution until past due benefits push the representative fees past the fee ceiling. We use difference-in-differences to evaluate how these incentives impact SSDI claimant wait times. After the fee ceiling increased in 2002, average wait times increased by 0.85 months among claimants for whom the fee threshold is more binding, implying a 2.6-5.6 month increase for claimants with representatives. This indicates that the structure of representative compensation does matter for case outcomes, and highlights the importance of interactions with auxiliary agents so common in modern social programs.

We show that the quality of police hires varies over the business cycle. Officers hired when the unemployment rate is high have fewer complaints, disciplines, and are less likely to be fired than officers hired when the unemployment rate is low. Effects are larger for younger workers who have weaker outside options in recessions. We find that the size and quality of the applicant pool increases in high unemployment years--more people take entry exams and a smaller fraction fail the exam. Our findings shed light on how outside options affect police hires and speak to policy questions about police recruitment.

Recent evidence suggests the U.S. business environment is changing, with rising market concentration and markups. The most prominent and extensive evidence backs out firm-level markups from the first-order conditions for variable factors. The markup is identified as the ratio of the variable factor’s output elasticity to its cost share of revenue. Our analysis starts from this indirect approach, but we exploit a long panel of manufacturing establishments to permit output elasticities to vary to a much greater extent - relative to the existing literature - across establishments within the same industry over time. With our more detailed estimates of output elasticities, the measured increase in markups is substantially dampened, if not eliminated, for U.S. manufacturing. As supporting evidence, we relate differences in the markups’ patterns to observable changes in technology (e.g., computer investment per worker, capital intensity, diversification to non-manufacturing), and we find patterns in support of changing technology as the driver of those differences.

In 1974, a federal court ordered that public schools in Jefferson County, KY be desegregated. To achieve racial integration, students were assigned to a busing schedule that depended on the first letter of their last name. This led to quasi-random variation in the number of years of busing and, for the initial cohorts, whether individuals were bused at all. We use this variation to estimate the long-run impact of busing on political participation and preferences. Focusing on white males, we do not detect any effect on voter turnout. We do, however, find that busing significantly increases Democratic party affiliation more than forty years later. Consistent with the idea that attending a formerly black, city school causes a change in the broad ideological outlook of whites, we also find that bused individuals are less likely to donate to organizations that advocate for conservative causes.

I estimate the effect of school desegregation on long-run economic outcomes by studying a natural experiment in Jefferson County, KY. In 1975, the district, under a court order, developed a unique busing assignment plan to merge the majority-white County district and the majority-black City district. Under this plan, students were assigned to be bused to new schools (versus stay at their home school and have new students bused in) based on their race and the first initial of their last name. Using this plausibly conditional random assignment and confidential data from the US Census Bureau, I find black students assigned busing to former County schools live in better neighborhoods (e.g. neighborhoods with higher tract-level income) at adulthood than black students assigned to remain in former City schools. This effect is strongest for students bused in earlier grades and is increasing in the total number of years a student is assigned busing. Busing assignment has small to zero effect on white students. I explore the implications of white disenrollment from the district (i.e. “white flight”) by using a novel dataset of archival yearbook records. I find the effect for white students remains small even after preliminary accounting for disenrollment. These results suggest that school desegregation in this setting had positive long-run effects for black students by giving them access to better schools (e.g. schools with more capital investment, more credentialed teachers, lower drop-out rates, etc.). 

I estimate the effect of access to Food Stamps on criminal recidivism. In 1996, a federal welfare reform imposed a lifetime ban from Food Stamps on convicted drug felons. Florida modified this ban, restricting it to drug traffickers who commit their offense on or after August 23, 1996. I exploit this sharp cutoff in a regression discontinuity design and find that the ban increases recidivism among drug traffickers. The increase is driven by financially motivated crimes, suggesting that the cut in benefits causes ex-convicts to return to crime to make up for the lost transfer income. 

Local governments rely heavily on sales tax revenue. We use national bankruptcies of big-box retail chains to study sudden plausibly exogenous decreases in this type of revenue. Treated localities respond by reducing spending on law enforcement and administrative services. We further study how cities with different degrees of autonomy vary in their response. Cities in home rule states, who have greater autonomy, react more swiftly by raising other types of revenue. A regression discontinuity analysis of cities in Illinois, where home rule status is triggered by crossing a population threshold, shows that this effect of local autonomy is causal: home rule leads to smaller revenue drops and stronger bond ratings. 

An important gap in most empirical studies of establishment-level productivity is the limited information about workers’ characteristics and their tasks. Skill-adjusted labor input measures have been shown to be important for aggregate productivity measurement. Moreover, the theoretical literature on differences in production technologies across businesses increasingly emphasizes the task content of production. Our ultimate objective is to open this black box of tasks and skills at the establishment-level by combining establishment-level data on occupations from the Bureau of Labor Statistics (BLS) with a restricted-access establishment-level productivity dataset created by the BLS-Census Bureau Collaborative Micro-productivity Project. We take a first step toward this objective by exploring the conceptual, specification, and measurement issues to be confronted. We provide suggestive empirical analysis of the relationship between within-industry dispersion in productivity and tasks and skills. We find that within-industry productivity dispersion is strongly positively related to within-industry task/skill dispersion.