Evaluating Whether Performance Pay Makes Teachers Stay

• In Event: Can a High-Quality Measure of Teacher Effectiveness Be Used to Support, Reward, and Retain Teachers?

Abstract

Prior research establishes that teacher quality is the strongest school-related predictor of student achievement (Rivkin, Hanushek & Kain, 2005; Clotfleter, Ladd & Vigdor, 2007) and suggests that financial incentives may yield workforce composition effects that improve overall teacher effectiveness (Fulbeck, 2014; Dee & Wyckoff, 2015; Glazerman & Seifullah, 2012; Lazear, 2003). In addition, there is evidence that if highly effective teachers working in the toughest circumstances are differentially rewarded, this may serve to improve retention of these high-quality faculty who are in otherwise less desirable working conditions (Ballou and Podgurksky 1997).

This paper evaluates the link between the NPS/NTU teacher contract and the associated initiatives and teacher retention. Specifically, only effective or highly effective teachers advance to a new step on the salary scale, while those rated ineffective are put on probation and can be asked to leave the district. In addition, teachers can receive bonuses for earning a highly effective rating ($5,000). Highly effective teachers who work in one of the district’s lowest-performing schools are eligible for an additional bonus ($5,000), as are those who teach in a hard-to-staff subject area ($2,500).

To estimate the impact of performance pay on teacher retention this paper provides descriptive evidence on the change in teacher retention associated with NPS performance pay, by estimating linear probability models of the change in the proportion of teachers who remain the district (and remain at their school) after implementation (Boyd et al., 2008). Second, the paper uses a regression discontinuity estimator to identify the average change in teacher retention among teachers who’s 2013-14 evaluation rating placed them near the cutoff of each evaluation rating, which provides a discontinuous change in monetary incentives (Dee and Wycoff, 2015). Differences in student, school, and teacher characteristics are included in all models.

Descriptive results show significant changes in retention due to a teacher’s prior effectiveness rating. Regression discontinuity results further buttress the robustness of these descriptive results, showing significant effects of the contract’s incentives on teacher retention especially around the cut-off between partially effective and effective in a manner that is consistent with the contract’s theory of action; the greatest, non-linear expected monetary difference in a teacher’s future earnings levied by the contract is between those who obtain a partially effective rating and those who obtain an effective rating.

Such findings will inform the expansion of similar policies and the debate around the impact of these policies on individual teacher effectiveness and the composition of the teacher workforce; as of September 2013, 27 states plus the District of Columbia (DCPS) use teacher effectiveness results as grounds for performance-related dismissal, while 40 states and DCPS had passed legislation requiring that objective measures of student achievement and growth be incorporated into teacher evaluation (National Council on Teacher Quality, 2014). Almost all states are looking for guidance about how to use measures of teaching effectiveness to manage their teaching workforce.

Authors

• Candace Hamilton Hester, American Institutes for Research

• Melissa Yisak