Department of Quantitative Social Science Do Performance Targets Affect Behaviour? Evidence from Discontinuities in Test Scores in England

Performance targets are ubiquitous in all areas of an individual’s life such as education, jobs, sport competitions and charity donations. In this paper I assess whether meeting performance targets in tests at school has an effect on students’ subsequent behaviour. This is helpful to test whether motivation and effort by students, parents and schools that the targets may induce, contribute to explain observed behaviour. I address potentially spurious correlations between test scores and behaviour by exploiting a regression discontinuity design in tests and a linked dataset of test scores and subsequent behaviour by students in compulsory education in England. I find that meeting a target that the government sets for students at age 11 has an insignificant effect on outcomes such as the probability of absence from school or of a police warning. I also find that meeting other targets for high and low ability students decreases the probability of being bullied by up to 34% with respect to the mean probability of such outcomes. The effects are heterogeneous as they vary by gender, parents’ education level and type of behaviour. Overall, the research design offers a valuable test to assess unintended consequences that meeting the target or failing to meet it may lead to. The lack of a significant effect of targets on suspension and expulsion from school, as well as police warnings, suggests no adverse behavioural effect of performance targets, which is reassuring evidence on the design of tests in compulsory education. By using Probit estimates, one would conclude that meeting a target has an impact on behaviour. Regression discontinuity estimates show instead an insignificant effect at the expected target and a significant one at other targets for certain outcomes, although smaller than Probit estimates. JEL classification: C21, I20, I21, I28.