Everybody needs good neighbours?: evidence from students' outcomes in England

We use administrative data to estimate the effect of neighbourhood composition on teenagers’ educational and behavioural outcomes in England. We exploit a unique research design based on changes over time in neighbourhood composition experienced by residentially immobile students, where these changes arise purely through residential migration amongst other students in our dataset. The complete coverage of our data allows investigating heterogeneity and non-linearities in the effect of neighbourhood composition at an unprecedented level. Our results show that changes in neighbourhood composition have no effects on test scores but some effects on behavioural outcomes, which are heterogeneous for boys and girls