Schools strongly resist the publication of their performance, officially because they have special circumstances that justify their ranking, and those circumstances get overlooked. However, as I reported before, evidence from the Netherlands showed that publishing school rankings forced the worst school to improve, and they do improve. But maybe the circumstances in the Netherlands are special.
Simon Burgess, Deborah Wilson and Jack Worth exploit a natural experiment in Britain: Wales suppressed the publication of school rankings in 2001, while England kept it. Using a difference-in-difference analysis, they show, oh surprise, that the performance of Welsh students regressed significantly, based on national exams. Only the top 25% schools escape this debacle. In addition, they find no indication of changes in stratification across schools in Wales, which invalidates further the argument that publishing rankings exacerbates school disparities through self-sorting of students. What is then a good argument for withholding school performance information?