A Random Assignment Experiment to Examine the Impact of Performance Pay on Teacher Retention in High-Poverty Schools
*Names in bold indicate Presenter
The study took place in 20 K-8schools in a Midwest district. The schools have among the highest poverty levels and lowest student achievement levels in the district. Prior to the 2011-12 school year, ten of the schools were randomly assigned to have the opportunity to receive performance pay bonuses (treatment schools); teachers in the other ten (control) schools could not receive these bonuses.
Teacher performance is based both on student growth scores (on state standardized assessments) and year-end effectiveness ratings that incorporate multiple classroom observations. Teacher collaboration groups (TCGs) work together to help improve instructional practices as well as attain a TCG student growth target score; if successful, each TCG member receives a bonus. In addition, school-wide growth targets are established that, if met, enable each teacher to receive an additional bonus. Finally, individual bonuses are awarded to teachers based on their year-end effectiveness rating. At the upper limit, teachers in treatment schools can receive as much as $15,000 in bonuses per year.
Retention results were analyzed from two transition periods: from school years 2011-12 to 2012-13, and from school years 2012-13 to 2013-14.
This study addressed two research questions. The first was to examine the impact of performance pay on teacher retention. The second was to examine whether the bonus amount was related to retention in treatment schools.
Regarding the first issue, we found evidence that performance pay had an impact on treatment school teacher retention. For treatment schools during the first transition period, performance pay did not increase the odds of teacher retention (p = .79); however, for the second transition period, the odds of retention were 1.78 times larger when educators could receive performance pay (p < .05).
In studying the second issue, we found some evidence that bonus amounts were positively related to higher retention, especially for the second transition period. Specifically, the odds of retention were 3.22 times larger with the receipt of a bonus that was greater than $5,000 and up to $10,000 (p < .05). Although marginally significant, the odds of retention were 3.93 times larger with the receipt of a bonus that was greater than $10,000 and up to $15,000 (p = .08).