Cohabitation, nonmarital childbearing, and the marriage process

Past work on the relationship between cohabitation and childbearing shows that cohabitation increases fertility compared to being single, and does so more for intended than unintended births. Most work in this area, however, does not address concerns that fertility and union formation are joint processes, and that failing to account for the joint nature of these decisions can bias estimates of cohabitation on childbearing. For example, cohabitors may be more likely to plan births because they see cohabitation as an acceptable context for childbearing; alternatively, they may be more likely to marry than their single counterparts. In this paper, I use a modeling approach that accounts for the stable, unobserved characteristics of women common to nonmarital fertility and union formation as a way of estimating the effect of cohabitation on nonmarital fertility net of cohabitors’ potentially greater likelihood of marriage. I distinguish between intended and unintended fertility to better understand variation in the perceived acceptability of cohabitation as a setting for childbearing. I find that accounting for unmeasured heterogeneity reduces the estimated effect of cohabitation on intended childbearing outside of marriage by up to 50%, depending on race/ethnicity. These results speak to cohabitation’s evolving place in the family system, suggesting that cohabitation may be a step on the way to marriage for some, but an end in itself for others.