Incentive Design in Education: An Empirical Approach∗

While incentive schemes to elicit greater effort in organizations are widespread, the effort-incentive strength relationship is difficult to ascertain in practice, hindering incentive design. We propose a new semiparametric method for uncovering this relationship in an education context, using exogenous incentive variation and rich administrative data. We then devise a structural estimation procedure that allows us to recover the primitives underlying the effort function, based on a model of effort setting. The parameter estimates combined with the model form the basis of a counterfactual approach for tracing the effects of various accountability systems on the full distribution of scores for the first time. We show higher average performance comes with greater inequality of outcomes for widespread fixed-target schemes, and that incentive designs not yet enacted can, at no extra cost, improve student performance while reducing test score inequality – of relevance to education reform.