A stable family environment is of crucial importance for child development. Nonmarital cohabitation, which tends to be a less stable family structure, has become increasingly prevalent in developed countries over the past few decades. An important question is to understand how the changing family structure influences child development.
There is a growing literature particularly in sociology that examines the correlation between family structure and child outcomes. However, the existing research still cannot robustly identify the causal effect of family structure. For instance, parents who form unstable families tend to have lower income and other unobserved characteristics such as lower ability. Particularly, those unavailable characteristics hinder a causal identification analysis. Furthermore, the framework used in this literature does not explicitly delineate the economic mechanisms through which family structure can affect children’s skill formation.
We develop and estimate a tractable economic model using a large sample of disadvantaged families in the U.S.A to study the impact of family structure on children’s cognitive and non-cognitive outcomes. Most importantly, combining our model with data, we are able to infer those unobserved characteristics and control them to reveal a robust causal effect of family structure on child development. We highlight the biological father’s endogenous involvement with the child within and across relationships; the child’s skill formation is influenced by the mother, biological father, and social father over time.
We find significant heterogeneity in the returns to paternal time investment by the skill endowment type of fathers. Investment made by high-ability fathers carry positive and significant returns, whereas investment made by low-ability fathers can generate negative returns. Interestingly, the return to father’s time investment does not seem to vary by father’s education level, highlighting the importance of identifying the distribution of unobserved ability of fathers. For non-cognitive outcomes, we find that paternal time spent with children reduces externalizing (aggressive and rule-breaking) behavior, more so by high-ability fathers and fathers cohabiting with the mother. Paternal time does not seem to affect internalizing (depressive, anxious, self-contained) behavior.
Our results suggest that family policies that incentivize (or disincentivize) mother and biological father cohabitation should pay attention to the quality of the fathers and potential social fathers in families, because children benefit the most in terms of both cognitive and non-cognitive development by living with a high-quality father or social father.