Beyond the threshold: The implications for pupil achievement of reforming school performance metrics

We study the effects of a major change to the school accountability system in England. In 2015, the leading published school performance metric was switched from a threshold measure (essentially the fraction of students above a test score level) to an average score measure. Using 7 years of data on all secondary schools in England, we show that this intervention relatively reduced the test scores of students near the threshold, in favour of groups above the threshold (marginally) and below (substantially). We check the sensitivity of our results to different decisions, and present findings on heterogeneous treatments.