Search

This paper develops the first quantitative framework for analyzing distributional effects of incentive schemes in public education. The analysis is built around a hump-shaped effort function, estimated semi-parametrically using exogenous incentive variation and rich administrative data. We identify key primitives that rationalize this effort function by estimating a flexible teacher effort-choice model. Both the model and parameter estimates are necessary components in our counterfactual framework for tracing the effects of alternative accountability systems on the entire test score distribution, with effort adjusting endogenously. We find widespread schemes that set a fixed target for all students give rise to a steep performance-inequality tradeoff. Further, counterfactual incentive policies can outperform existing schemes for the same cost -- reducing the black-white test score gap by 7% (via student-specific bonuses), and lowering test-score inequality across students by 90% (via student-specific targets). Our quantitative approach opens up new possibilities for incentive design in practice.

This paper provides the first causal evidence about how elected local school boards affect student segregation across schools. The key identification challenge is that the composition of a school board is potentially correlated with unobserved determinants of school segregation. We overcome this issue using a regression discontinuity design at the electoral contest level, exploiting quasi-random variation from narrowly-decided elections. Such an approach is made possible by a unique dataset, which combines matched information about North Carolina school board candidates with time-varying district-level racial and economic segregation outcomes. Focusing on the political identity of school board members, regression discontinuity estimates reveal that (relative to their non-Democratic counterparts) Democratic board members decrease racial segregation across schools. Our findings suggest that school boards realize such reductions in segregation by shifting attendance zones (which we infer without the need for exact geocoded boundaries) and that white families differentially exit the traditional public school system for local charter schools in response.

It is well understood that adverse economic shocks affect workers nonuniformly. We explore a new channel through which unequal employment outcomes may emerge during a downturn: displacement through the extensive margin of establishment deaths. Intuitively, workers who are concentrated in less resilient establishments prior to an economic decline will be disproportionately affected by its onset. Using rich administrative employment and establishment data for the United States, we show that Black workers bore the brunt of the Great Recession in terms of within-industry employment changes arising from establishment deaths. This finding has important implications for the evolution of worker disparities during future downturns.

We develop a novel strategy to identify the relative importance of school and neighborhood factors in determining school segregation. Using detailed student enrollment and residential location data, our research design compares differences in student composition between adjacent Census blocks served by different schools to analogous differences between those schools. Our findings indicate that neighborhood factors explain around 62% of racial segregation and 44% of economic segregation across all schools, playing an even more pronounced role in urban areas, where school segregation has been especially acute. These findings suggest that the involvement of urban planners is essential when attempting to confront inequality of opportunity through education.

Teacher value-added is not a technological ‘primitive.’ Instead, it increases within-teacher when accountability incentives are strengthened, as we show. This evidence motivates a framework in which teacher value-added depends on two unobserved inputs to education production: teacher ability and incentive-varying teacher effort. We present a strategy to identify these two inputs and their distinct effects on test scores for the first time, using exogenous incentive policy variation and rich longitudinal data from North Carolina. The estimates indicate that both the ability component of teacher value-added and teacher effort raise current and future scores. We also find that effort responds systematically to incentives, underlining the agency of the teaching force as an important factor in education production. To explore the policy implications of this economic agency, we then use our estimates to compare the cost effectiveness of incentive-oriented education reforms with policies that target teacher ability. Incentive reforms come out ahead in a variety of plausible cases, given their potential to influence all teachers rather than a subset.

While incentive schemes to elicit greater effort in organizations are widespread, the incentive strength-effort mapping is difficult to ascertain in practice, hindering incentive design. We propose a new semi-parametric method for uncovering this relationship in an education context, using exogenous incentive variation and rich administrative data. The estimated effort response forms the basis of a counterfactual approach tracing the effects of various accountability systems on the full distribution of scores. We show higher average performance comes with greater score dispersion for a given accountability scheme, and that incentive designs not yet enacted can improve performance further, relevant to education reform.

This paper provides the first evidence that value-added education accountability schemes induce dynamic distortions. Extending earlier dynamic moral hazard models, I propose a new test for ratchet effects, showing that classroom inputs are distorted less when schools face a shorter horizon over which they can influence student performance. I then exploit grade span variation using rich educational data to credibly identify the extent of dynamic gaming, fi nding compelling evidence of ratchet effects based on a triple-differences approach. Further analysis indicates that these effects are driven primarily by effort distortions, with teacher reallocations playing a secondary role.