Abstract
In this paper, we investigate whether better schools can compensate for the effects of children's genetic differences. To this end, we combine data from the Norwegian Mother, Father, and Child Cohort Study (MoBa) with Norwegian register data to estimate the interaction between measures of children's predisposition to education and school quality. We use MoBa's genetic data to compute polygenic indices for educational attainment (PGI(EA)). Importantly, MoBa includes genetic data on mother-father-child trios, allowing us to identify causal genetic effects using within-family variation. We calculate school value-added measures from Norwegian register data, allowing us to causally estimate school quality effects. Leveraging the advantages of both data sources, we provide a causally identified study of gene-environment interactions in the school context. We find evidence for substitutability of PGI(EA) and school quality in reading but not numeracy: A 1 SD increase of school quality decreases the impact of a 1 SD increase of PGI(EA) on reading test scores by 6%. The substitutability arises through gains of students at the lower end of the PGI(EA) distribution. This suggests that investments in school quality may help reduce educational inequalities arising from genetic differences between students.