How unstable are 'school effects' assessed by a value-added technique?

This paper re-considers the widespread use of value-added approaches to estimate school 'effects', and shows the results to be very unstable over time. The paper uses as an example the contextualised value-added scores of all secondary schools in England. The study asks how many schools with at least 99% of their pupils included in the VA calculations, and with data for all years, had VA measures that were clearly positive for five years. The answer is - none. Whatever it is that VA is measuring, if it is measuring anything at all, it is not a consistent characteristic of schools. To find no schools with five successive years of positive VA means that parents could not use it as a way of judging how well their primary age children would do at age 16 in their future secondary school. Contextualised value-added (CVA) is used here for the calculations because there is good data covering five years that allows judgement of its consistency as a purported school characteristic. However, what is true of CVA is almost certainly true of VA approaches more generally, whether for schools, colleges, departments or individual teachers, in England and everywhere else. Until their problems have been resolved by further development to handle missing and erroneous data, value-added models should not be used in practice. Commentators, policy-makers, educators and families need to be warned. If value-added scores are as meaningless as they appear to be, there is a serious ethical issue wherever they have been or continue to be used to reward and punish schools or make policy decisions.