(Un)Intended Consequences of a Teacher Performance Pay Program

I use a sharp regression discontinuity design (RDD) to estimate the causal effect of a group performance-pay program in secondary education that is long-lived and universal in nature. The program design ensures internal and external validity, which is rare in studies that rely on RDD. By combining three Chilean administrative datasets into a unique longitudinal data, I can follow all the teachers in the system and four student cohorts. I find evidence consistent with teaching to the test practices and teacher retention efforts. I also find that increasing the bonus does not benefit the students differently. The results have direct policy implications.