Conventional twin studies overestimate the environmental differences between families relevant to educational attainment

Estimates of shared environmental influence on educational attainment (EA) from the Classical Twin Design (CTD) are often interpreted as evidence for unequal environmental opportunity. However, key CTD assumptions appear violated for EA. Using data from 982 German families (TwinLife), the authors compared CTD estimates to those from a Nuclear Twin and Family Design (NTFD) that accounts for assortative mating and twin-specific shared environments. The CTD estimated shared environment at 43%. The best-fitting NTFD model (accounting for assortative mating) estimated total shared environment at 26%, decomposed into twin-specific shared environment (16%) and sibling-shared environment (10%). Only the latter reflects between-family environmental differences, implying CTD overestimates between-family inequality in educational opportunity. Additionally, once genetic influences were accounted for, parental education had no environmental effect on offspring education.