Cumulative Effect of Family Structure on Educational Attainment

Hope Harvey, Harvard University

Scholars have long recognized the importance of family structure, but studies often “control away” effects that operate through time-varying characteristics like income. Using the NLSY79, I employ inverse probability treatment weighting and marginal structure models to examine the effects of family structure on educational attainment. These methods allow for dynamic selection, in which family structure affects time-varying characteristics that are in turn associated with future family structure. I find that compared to an additional year with married biological parents, a year with a single mother is associated with a 6.1% reduction in the odds of graduating high school and 3.1% reduction in the odds of attending college, and a year with a cohabiting social father is associated with a 13.5% reduction in the odds of graduating high school and 12.4% reduction in the odds of attending college. Totaled across childhood, family structure can substantially shape children’s life chances.