Abstract
Stratification research which estimates the association between family background and children’s outcomes at the mean does not take into account possible nonlinear relationships with the dependent variable. There are, however, strong theoretical reasons to expect that social origin differences in child outcomes vary across the distribution. Analyzing these differences is important for our understanding of the intergenerational transmission of advantage. We use unconditional quantile regression models to estimate how the associations between different indicators of family background (parental class, parental education, parental occupational standing, and parental wealth) and children’s educational performance vary across the distribution in Germany and in Norway. We find that social origin differences in both countries are concentrated at the bottom and in the middle but are weakest at the top of the distribution. There are no systematic differences in the variation of the associations between family background and academic performance in Germany and Norway.