Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
In this paper, we show that the public provision of information on Australian schools’ average national test score outcomes via the My School website, launched in 2010, resulted in poorly performing schools testing fewer students in subsequent years. This increased non-participation in testing was driven primarily by formal parental withdrawal, and poorly performing students were much more likely to be withdrawn from testing. This phenomenon is consistent with schools attempting to ‘game’ the system to improve published test scores. We also provide weaker evidence that withdrawal responded more strongly to initial poor performance in independent private schools than in government schools.