The Gendered Division of Labor and Family Outcomes in Germany
Cooke, Lynn Prince, Journal of Marriage and Family
The literature on the predictors of the division of household labor continues to expand, but the effect of this division on family outcomes has not been explored. Using the German SocioEconomic Panel (N = 628), I analyze the effect of men s participation in housework and child care on the likelihood of second birth and divorce. Fathers' greater relative child-care time increases couples' odds of second birth, attenuating the negative effect of mothers' employment. Husbands' relative housework time is insignificant in predicting second birth or divorce among couples with at least one child, but increases the likelihood of divorce among childless couples. This is evidence that the division of domestic labor affects family outcomes, but effects differ depending on the outcome and presence of children.
Key Words: divorce, fertility, gendered division of labor, longitudinal analysis.
Wives' increasing labor force participation was expected to lead to a revolution in the gendered division of labor, a revolution that "stalled" at the door of most households (Hochschild, 1989). Although the division of paid labor continues to dissolve, the division of domestic labor remains firmly gendered and female (Blossfeld & Drobnic, 2001; Gershuny, 2000). In industrialized countries, husbands' average domestic participation is about one third of the time contribution of wives (Gershuny).
One explanation for why the gendered division of domestic labor persists is that couples "do" gender to reflect normative expectations of social identities as women and men within the intimate sphere of marriage (Berk, 1985; West & Zimmerman, 1987). For example, U.S. married women contribute more and married men fewer hours toward domestic tasks than either does in his or her single state (South & Spitze, 1994), and where U.S. women are the primary breadwinners, they are even more likely to do a greater share of domestic tasks (Brines, 1994; Hochschild, 1989).
Families, however, are a locus of struggle, not an uncontested, stable gendered domain (Hartmann, 1981). Consequently, the gendered division of domestic labor might have repercussions for subsequent family outcomes. Although some research has looked at the effects of the division of housework on individual outcomes such as wages and work effort (Becker, 1985; Bielby & Bielby, 1988; Coverman, 1985; McAllister, 1990); marital satisfaction (Perry-Jenkins & Folk, 1994; Robinson & Spitze, 1992; Yogev & Brett, 1985) or conflict (Blair, 1993; Lye & Biblarz, 1993); and physical (Ross & Bird, 1994) or psychological wellbeing (Bird, 1999; Glass & Fujimoto, 1994; Mirowsky, 1996; Ross, Mirowsky, & Huber, 1983), none has explored whether the division of domestic labor affects future family decisions. Gupta (1999) analyzed the causal link between marital status and housework time, but not vice versa. So it remains unknown whether husbands' even modest contributions to domestic tasks have a significant effect on the likelihood of two key family outcomes: fertility and marital stability.
Here, I use event history analysis of the German SocioEconomic Panel to explore whether the division of domestic labor alters the likelihood of either second birth or divorce. Germany is chosen for three reasons. First, the gendered division of labor has been encouraged by German labor market policy, family transfers, and other social provisions (Ostner, 1993). It is therefore the ideal country in which to test the fundamental assumptions of the dominant family model described in the next section. second, because the household division of labor changes as family members move into and out of employment and as family demands increase and decrease, effects on subsequent family transitions are best analyzed with longitudinal data where both adults are interviewed at regular intervals, which is the case with the German SocioEconomic Panel. U.S. panels are either conducted infrequently or have limited measures of domestic tasks and rely on a single household respondent, whereas other national panels have started more recently. …