Are mixed ability classes bad for school performance and educational choice? Socioeconomic and ethnic inequality in English and Swedish schools

A longstanding question in the organisation of education has been the one about learning in ability-homogenous groups.

Previous research has suggested that tracking or ability grouping, aiming at homogeneity, do not improve learning on average, but may lead to increasing inequality.

However, lack of suitable data and methodological difficulties have made results inconclusive. We contribute with a comparative study between England (with institutionalised ability grouping) and Sweden (with occasional use). We use CILS4EU data to study the effect of (a) homogeneity in ability in instructional groups on grade point averages (GPA) and academic transition to upper secondary education; (b) the effect of ability grouping within subjects on grades in these subjects.

School fixed effects, Inverse of Probability of Treatment Weighting models, and placebo analyses show little evidence of efficiency gain in homogeneous instruction groups or ability grouping, but also little evidence that such grouping has any effects on inequality; nor do we find any significant differences between England and Sweden.