Association for Public Policy Analysis & Management

Low-income girls and women have made remarkable strides over the past few decades in educational attainment, but a significant proportion of low-income boys have not kept up with their counterparts’ advancements, at all levels of schooling. Women now outpace men in obtaining high school diplomas, enrolling in college, and completing post-secondary degrees across the SES spectrum. Focusing on the low-income population, this paper explores a potential factors affecting differential educational achievement and attainment by gender: cognitive development as impacted by family structure and family instability – specifically the biological parents’ relationship over time.

In this paper, we use data from the Fragile Families and Child Wellbeing Study (FFCWS) and a novel identification strategy to investigate how the evolution of the biological parents’ relationship status over time affected cognitive test scores for urban children at age 9 by gender. While controlling for other relationship status and trajectory characteristics, such as the presence of stepfathers (both married and cohabiting), we center our analysis on the biological parents’ relationship under the theory that biological parents will contribute, on average, the most to the development of their biological children in light of prevailing social norms (Becker, 2009; Cherlin, 2004; Hofferth & Anderson, 2003). Moreover, since we find that having continuously married biological parents produces the highest cognitive test scores, on average, we explore deviations from this relationship, while controlling for the re-partnering history of the mother.

Our primary contribution is the robust finding that, unlike equivalent boys, low-income urban girls’ cognitive development is not sensitive to their biological parents’ relationship status over time. In our preferred models – when we use the biological parents’ relationship both at birth and at age 9 to define categories of roughly equivalent youth – we find that the cognitive development of low-income urban girls, as measured by test scores, is unaffected by their parents’ relationship status, after controlling for a limited set of demographic characteristics. However, the story for boys is much more nuanced: while boys with parents who were single at the time of their birth but married to each other by the child’s 9^th birthday perform approximately 0.4 standard deviations better than girls in that same family structure category on cognitive test scores, boys in three other categories – biological parents cohabitating continuously, biological parents divorced, biological parents never married or cohabiting – perform roughly 0.4 standard deviations worse than equivalent girls.

In addition, when accounting for evidence that boys impact their biological parent’s relationship at birth – driven in part because fathers are more likely to remain partnered with a woman who bears his male child – we find that young boys perform even worse. Finally, even with smaller sample sizes – which inhibits statistical precision and inference – evidence suggests that Black, Hispanic, and white boys are all adversely impacted by any deviation from the two-biological parent continuously married family structure. Thus, while girls appear to be unaffected, low-income boys are unilaterally impacted by changes in their biological parents’ relationship status over time.