School accountability: Incentives or sorting?

We exploit a nested school accountability reform to estimate the causal effect on teacher mobility, sorting, and student achievement. In 2003, lower-secondary schools in Oslo became accountable to the school district authority for student achievement. In 2005, information on school performance in lower secondary education also became public. Using a difference-in-difference-in-difference ap-proach, we find a significant increase in teacher mobility and that almost all non-stayers leave the teaching sector entirely. The impact is larger on high-ability teachers following the second part of the reform. Non-stayers are largely replaced by high-ability teachers, indicating a positive sorting effect. We find a small, positive effect on student achievement after the second part of the reform, thus the mechanism in place seems to be positive teacher sorting rather than teacher incentives.Keywords: school accountability regimes, design of incentives, teacher turnover, teacher quality, teacher sorting, student achievement, difference-in-difference-in-difference.