Knowledge that Transforms

We introduce new measurement tools to understand the sources of earnings differences across space. Based on the natural language employers use in job ads, we develop granular measures of job tasks and of worker specialization. We find that jobs in larger commuting zones involve greater interpersonal interactions and have higher computer software requirements. Between 10% and 50% of task and technology variation between large and small commuting zones exists within occupations. Furthermore, workers in larger markets are more specialized. Tasks, technologies, and worker specialization account for a substantial portion of the market size premium even within occupations.

There is evidence showing racial bias in firms’ hiring decisions, but less is known about bias in career recognition. We construct a dataset from the second-largest US police department to estimate the Black-white gap in award nominations. Leveraging institutional features, we find that white supervisors are less likely to nominate Black officers conditional on work performance. This appears to be driven by supervisor bias in advocacy decisions rather than statistical discrimination. Given the reliance on subjective evaluations for promotions in many organizations, our findings have important implications for the Black-white promotion gap and the lack of diversity in upper-management positions.

We estimate the longer-run effects of attending an effective high school (one that improves a combination of test scores, survey measures of socioemotional development, and behaviors in ninth grade) for students who are more versus less educationally advantaged. All students benefit from attending effective schools, but the least advantaged students experience larger improvements in high school graduation, college going, and school-based arrests. Test score value-added understates the long-run importance of effective schools, particularly for less advantaged populations. Patterns suggest that this may, in part, reflect less advantaged students being relatively more responsive to non-test-score dimensions of school quality.

The income maintenance experiments have received renewed attention due to growing international interest in a basic income. Proponents of a negative income tax (NIT) viewed it as a replacement for traditional welfare with stronger work incentives. However, existing labor supply estimates for single mothers (those eligible for welfare) are uniformly negative. We reassess the experimental evidence and find randomization failure in two NITs (Gary and Seattle). In Denver and Manitoba, we find positive labor supply responses for those on welfare before random assignment. Our results provide strong evidence that an NIT can increase work activity among single mothers on welfare.

We study how a significant relaxation of internal migration restrictions affects labor market outcomes of incumbent migrants and natives, exploiting the 2014 hukou reform in China, which substantially removed the migration barriers of cities with an urban population below 5 million (nonmegacities). Using a difference-in-differences method, we find that migrants’ wages in nonmegacities experienced approximately a 2.6%–7.9% decline relative to that in megacities after the policy. The policy had nonnegative impacts on the wages of natives in nonmegacities. These results suggest that the downward wage pressure imposed by new migrants falls primarily on incumbent migrants rather than on natives.

We investigate the impact of recent gentrification shocks on minority neighborhoods in the 50 largest US labor markets. We show that household moves from a given neighborhood are concentrated to few destinations with similar minority shares and strong network ties, but those neighborhoods are farther away from downtown. Gentrification affects Black neighborhoods by raising house prices, reducing the proportion of Black households, and increasing the share of movers going to neighborhoods with network ties. However, gentrification has negligible effects on Hispanic neighborhoods. Overall labor market area segregation decreases after a gentrification shock because highly Black neighborhoods become less segregated.

This paper asks whether reducing the informativeness of college admission exams reduces inequality in postcollege earnings. I examine a 2000 reform of the Colombian college admission exam that sought to reduce bias in scores. The reform reduced test score gaps between high- and low-income students by nearly 50% in some subjects, but it also decreased the exam’s predictive power for abilities that matter in college. I find that the reform caused students to attend colleges where they were more likely to drop out, which reduced earnings for both high- and low-income students.

Tournaments are often used to improve performance in innovation contexts. Tournaments provide monetary incentives but also render teams’ identity and image concerns salient. We study the effects of tournaments on team performance in a nonroutine task and identify the importance of these behavioral aspects. In a field experiment (n>1, 700 participants), we vary the salience of team identity, social image concerns, and whether teams face monetary incentives. Increased salience of team identity does not improve performance. Social image motivates the top performers. Additional monetary incentives improve all teams’ outcomes without crowding out teams’ willingness to explore or perform similar tasks again.

The COVID-19 pandemic resulted in major declines in employment of women. We provide projections of impacts of this reduction on less educated women’s future human capital framed within the traditional Mincerian model. We find that wage losses one year out from 2020 are relatively modest on average, generally less than 1%, with the largest for married women without children in the home. But losses are greater for young married women, mothers with very young children, and those working in COVID-impacted industries. School and childcare closures increase negative wage impacts for married mothers by an additional 50%.

Using Michigan Panel Study of Income Dynamics data, we study selection bias and the gender wage gap. Employing several methods, we find large declines in the total and unexplained gender gaps in wage offers between 1981 and 2015. Under our preferred selection correction method, the median total and unexplained gaps fell by 0.378 and 0.204 log points, respectively. These are larger declines than if we had not corrected for selection and simply measured convergence in observed wage gaps. However, substantial selectivity-corrected median gender wage gaps remain in 2015: 0.242 log points (total gap) and 0.206 log points (unexplained gap).