Putting Performance Pay to the Test: Effects of Denver ProComp on the Teacher Workforce and Student Outcomes. Issue 12 (December 2020)

• Record Type: Journal Article
• Authors: Atteberry, Allison; Lacour, Sarah E.
• Is Part Of: Teachers College record. Volume 122:Issue 12(2020)
• Page Start: 1
• Page End: 48
• Publication Date: 2020-12
• Subjects: Education -- Periodicals; Enseignement -- Périodiques; Éducation -- Périodiques; Education; Periodicals; 370.5
• DOI: 10.1177/016146812012201208
• Languages: English
• ISSNs: 0161-4681

Abstract

Context: In 2005-06, Denver became one of the first U.S. districts to implement a pay-for-performance (PFP) compensation system, and Denver's ProComp is now the longest-running PFP policy in the country. The national proliferation of PFP systems in education has been controversial, with mixed evidence and competing narratives about its impacts. During Denver's 2019 strike, disagreements arose about whether 13 years of ProComp have helped or harmed efforts to retain effective teachers to improve student outcomes. This article addresses this policy debate. Research Questions: We use a 16-year panel to analyze the effects of ProComp on both student and teacher outcomes. We focus on the onset of the second version of the policy, ProComp 2.0, in 2008. Intervention: The ProComp policy is a pay-for-performance teacher compensation system, which includes ten distinct financial incentives, some of which are awarded schoolwide. Annual payouts represented 12% of base-pay, on average, among full-time teachers. Research Design: We use comparative interrupted time series (CITS) to examine pre/post ProComp trends in outcomes in DPS relative to similar districts across the same period. When CITS is not possible, we conduct interrupted time series (ITS) analysis in DPS using a panel up to five years pre-PC1 and up to ten years post-onset. Results: ProComp may have had a positive effect on ELA, math, and writing achievement that was not evident in comparable districts with similar achievement trends prior to 2005-06. We also find descriptive evidence that more effective teachers were recruited to DPS once ProComp began and that the overall decline in teacher retention across districts in this time period was less precipitous among DPS' highly effective teachers during ProComp. Conclusions: Our results can help reflect on which of the hypothesized mechanisms undergirding PFP policies find empirical support in the field. The onset of ProComp shifted the composition of the DPS teacher workforce through recruitment and retention of certain kinds of teachers. These results at first appear to contradict teacher perceptions that the program coincided with a dramatic decline in teacher retention and was thus ineffective. However, retention did, in fact, decline throughout the period. Yet DPS retention patterns were not that different from other comparable Colorado districts during this period. Thus, while teachers' perceptions of reduced teacher retention were accurate, it would be very difficult to see from within DPS that retention rates were not necessarily distinct from secular trends outside DPS.