Recruitment, Effort, and Retention Effects of Performance Contracts for Civil Servants

Skilled and intrinsically motivated public sector employees are critical to state capacity and the provision of key services, including education. Yet the capability to recruit, motivate, and retain this profile of civil servants remains a challenge in many developing countries like Rwanda, where rates of teachers quitting are high and vacant positions are challenging to fill.

This project tests whether performance contracts for teachers in Rwanda can improve teacher performance and student learning by inducing greater effort from existing teachers, or supporting the recruitment and retention of skilled and motivated teachers. A two-tiered randomized, controlled trial that distinguished these two types of effects was embedded within the Government of Rwanda’s rollout of the Supporting Teacher Achievement in Rwandan Schools (STARS) program, which pays the top 20 percent of teachers a bonus above their usual salary based on student learning gains as well as teachers presence, preparation, and pedagogy. The evaluation measures the impacts of the STARS pay-for-performance contract relative to a fixed-wage contract on applications to teaching positions, learning outcomes, and teacher retention.

Results show that performance pay changed the composition of the workforce, drawing in individuals who were more money-motivated (based on performance in a dictator game designed to measure intrinsic motivation), but who went on to teach at least as well. Newly recruited teachers working under the performance-pay scheme also attended class more often, had more effective classroom teaching practices, and were not more likely to quit during the two years of the experiment, than teachers working under fixed-wage contracts. Teachers recruited and working under the performance-pay scheme delivered 0.16 standard deviations of learning more than counterparts recruited and working under fixed wages. This means that in the second year, a student from the median (50th percentile) moved up to the 56th percentile of students.

Read the published paper in the American Economic Review here. You can also read more details about this project on the IPA website.