Assortative Matching vs. Complementarity: Which Drives the Correlation of Hours Worked by Young Couples?

This paper starts by documenting the positive correlation between annual hours worked by young husbands and wives without children. Two possible explanations for this phenomenon are assortative matching and complementarity of leisure. This paper evaluates these two possibilities jointly to determine their relative importance. I set up an individual labor supply problem for singles and a household labor supply model with complementarity of leisure for young couples. I select a subset of married couples from the NLSY79 for whom data both before and after marriage are available. Pre-marriage data are used to recover preferences which may vary across individuals. The data show strong positive assortative matching on individual preferences. Incorporating this heterogeneity of individual preferences, the post-marriage data are used to estimate the household labor supply model. A counterfactual analysis shows that complementarity of leisure is the main factor driving the positive correlation between hours worked by young husbands and wives.