Association for Public Policy Analysis & Management

Pay for performance is a popular education reform for many reasons, not the least of which is that economic and management theories suggest that well-designed incentive pay programs could improve teacher effectiveness and long-run schooling outcomes. When productivity is difficult to monitor, performance incentives help to align the private objectives of workers with those of the organization as a whole. In the education sector, performance incentives are expected to reward effective teachers for the additional effort they put in to being effective, and to encourage less effective teachers to seek out more effective instructional strategies or to leave the profession. To the extent that incentives encourage teachers to upgrade their skills or adopt more effective practices, those incentives can have positive impacts that persist even if the incentive program does not. To the extent that they induce turnover among ineffective teachers, incentives can also have a positive impact long after the program expires. However, if ineffective teacher leave incentivized schools but not the teaching profession, then any persistent benefits of the program may prove illusory.

Most of the literature on performance pay in the education sector focuses on the direct relationship between teacher incentives and student outcomes.  As such, the current empirical evidence on teacher incentive pay programs tends to miss out on the indirect effect of these programs on labor market selection, or what incentive theory calls the compositional effect of incentive pay. A performance pay program will tend to retain individuals who are particularly good at the metered activity to which incentives are attached, and repel those who are not.  This effect on the workforce can be very important in explaining productivity gains. For instance, in a case study outside of the education sector, Lazear (2000) found that sorting effects were both substantial and roughly equal in magnitude to motivation effects. In other words, while the incentive system raised the productivity of the typical worker employed, it also tended to raise the overall quality of the workforce.

Tennessee provides an excellent setting for studying whether or not incentive pay impacts teacher turnover. The state received a large influx of federal dollars in 2010 from a $500 million RTT grant and a $36 million federal TIF grant.  Initiating what is arguably Tennessee’s most aggressive push for systematic compensation reform to date, a total of 14 districts and 192 public schools received funds to develop and implement strategic compensation models beginning with the 2011-12 school year. We use a rich set of panel data on individual teachers maintained by the Tennessee Consortium on Research, Evaluation, and Development (the Consortium) and difference-in-difference and instrumental variables regression techniques to examine the impact of those incentive programs on teacher retention.  We also have access to teacher effectiveness measures, which prior studies have lacked.