The Impact of Classroom Peer Groups on Pupil GCSE Results

The effect of a more able peer group on a child's attainment is considered an integral part in estimating a pupil level educational production function. Examinations in England at age 16 are tiered according to ability, leading to a large stratification of pupils by ability. However, within tiers, there is a range of policies between schools regarding setting, ranging from credibly random to strict setting by results from examinations at age 14. We use this variation to estimate ordinary least squares (OLS) estimates, with school and teacher fixed effects, of the effect of a more able peer group using a subset of schools that has apparently random allocation of pupils. As a robustness test of the apparently random setting results, we use an instrumental variables (IV) methodology developed by Lefgren (2004b). We find significant, positive, and non-trivial effects of a more able peer group using both the OLS and IV estimations for English and mathematics.