Estimating Causal Effects of Fertility on Life Course Outcomes : Evidence Using A Dyadic Genetic Instrumental Variable Approach

The causal effects of fertility are a central focus in the social sciences, but the analysis is challenged by the endogeneity of fertility choices. Earlier work has proposed several "natural experiments" from twin births or gender composition of earlier births to assess whether having more children affects adults' outcomes, though there are limitations to using rare (twins) and weak (gender composition) instrumental variables for fertility. This paper proposes a new "natural experiment" approach to assessing the causal effects of fertility by measuring the combination of couples' genetics in predicting fertility--a dyadic genetic instrumental variable, where the key idea (exclusion restriction) is that the interactions of the couple's genetics that shift the likelihood of fertility is unknown to the couples. We use a nationally representative sample of couples to examine the long-lasting effects of fertility on older adults' life outcomes, including labor market outcomes, personality traits, and subjective wellbeing. We find that fertility reduces females' extraversion and years of working and some evidence indicates that fertility reduces both males' and females' lifetime number of jobs worked