Achievement Effects of Individual Performance Incentives in a Teacher Merit Pay Tournament

This paper estimates the effect of the individual incentives teachers face in a teacher-based value-added merit pay tournament on student achievement. We first build an illustrative model in which teachers use proximity to an award threshold to update their information about their own ability, which informs their expected marginal return to effort. The model predicts that those who are closer to an award cutoff in a given year will increase effort and thus will have higher achievement gains in the subsequent year. However, if value-added scores are too noisy, teachers will not respond. Using administrative teacher-student linked data, we test this prediction employing a method akin to the bunching estimator of Saez (2010). Specifically, we examine whether teachers who are proximal to a cutoff in one year exhibit excess gains in test score growth in the next year. Our results show consistent evidence that teachers do not respond to the incentives they face under this program. In line with our model, we argue that a likely reason for the lack of responsiveness is that the value-added measures used to determine awards were too noisy to provide informative feedback about one's ability. This highlights the importance of value-added precision in the design of incentive pay systems.