Living arrangements have undergone considerable change in recent decades. In most Western countries marriage is no longer the exclusive context of family formation. In the former Federal Republic of Germany, about 38.3 percent of the women born between 1954 and 1958 started their rst union outside of a formal marriage. The gure increases to 67.9 percent for the cohort 1964-1968 (Le Goff, 2002). These demographic trends challenge the microeconomic literature in which couples living in consensual unions are implicitly assumed to act exactly as married couples. A closer look at the literature reveals, however, growing evidence of the link between marital status and household behavior with respect to many outcomes. For instance, DeLeire and Kalil (2005) nd that cohabiting-parent families spend a greater amount on alcohol and tobacco, and a smaller amount on education and health care than marriedparent families. Cohen (2002), Mamun (2004) and Stratton (2002) examine the wage differential for married and cohabiting men. Higher wages are observed for married men, ceteris paribus. There is also evidence, based on US cross section data, that married couples exhibit a more`traditional' division of domestic and market work than cohabitants (South and Spitze 1994, Stratton, 2005). The purpose of this paper is to examine whether the shift from cohabitation to marriage is associated with a signi cant change in household market and non-market labor supply. More speci cally, we use a long German panel (GSOEP) to test whether the transition from cohabitation to marriage reinforces the degree of specialization among couples. We estimate a model that relates married life to the female-to-male domestic and market work hours log ratios. Other regressors of the log ratios are the female relative earnings, the number of children and the duration of the relationship.We account for selection bias in the presence of endogenous regressors following the procedure advocated in Semykina and Wooldridge (2005), that we adapt to system GMM estimation. Our results suggest that marriage increases female specialization in home-based activities. Importantly, marriage leads to a fall in womenï¿ 1 2 s leisure, particularly for couples with preschool children. The results also exhibit a fall in married menï¿ 1 2 s leisure coming from either a rise in market hours or an increase in domestic hours depending on the speci cation.
