'League table' information on school effectiveness in England generally relies either on a comparison of the average outcomes of pupils by school(for example, mean exam scores) or on estimates of the average value added by each school. These approaches assume that the information parents and policymakers need most to judge school effectiveness is the average

achievement level or gain in a particular school. Yet schools can be differentially effective for children with differing levels of prior attainment. We present evidence on the extent of differential effectiveness in English secondary schools and find that even the most conservative estimate suggests that around one-quarter of schools in England are differentially effective for students of differing prior ability levels. This affects an even larger proportion of children, as larger schools are more likely to be

differentially effective.