Educational Opportunity Does Not Skip Generations: Testing Lundberg’s (2020) Intergenerational Transmission Framework

Lundberg (2020) showed that significant cousin correlations in educational attainment need not imply direct influence from the extended family. Instead, such correlations may arise from two sources: generational variation in parent–child transmission strength (“dynamic transmission”) and attenuation due to classical measurement error. Building on this framework, I use data from the National Longitudinal Surveys to test whether observed sibling and cousin correlations are consistent with a first-order Markov process that incorporates generation-specific transmission parameters and known reliabilities of educational measures. After accounting for both dynamic transmission and measurement error, the model fully explains the observed cousin correlations in the NLSY. Any residual extended-family influence is negligible. These findings reinforce Lundberg’s central insight: cousin correlations in educational attainment among the cohorts studied can be entirely attributed to two-generation transmission dynamics, rather than any substantive extended-family effect.