کارهای خانه، کار در بازار و "جنسیت" زمانی که رضایت زناشویی کاهش می یابد

When faced with a decline in marital satisfaction, are wives constrained from increasing their labor market work time in part because they “do gender?” One of the predictions of the human capital accumulation hypothesis, which assumes no constraints, is that housewives with little work experience will respond to a decline in marital satisfaction by increasing labor market work time (only). In contrast, the gender display hypothesis predicts that, in settings where the evaluations of marriage and wives’ work performance are closely intertwined, a decline in marital satisfaction among this group of housewives will increase both labor market work and housework—and the increase in housework serves as a constraint on the increase in labor market work. To evaluate these contrasting hypotheses, we analyze a panel survey of women in contemporary Japan. Results from multinomial logit regression models are more consistent with the gender display hypothesis than the human capital accumulation hypothesis. Housewives with relatively little work experience are 11 times more likely to increase the time spent on both labor market work and housework when the satisfaction of their marriage declines than when it does not. No evidence is found that, when marital satisfaction declines, these housewives are statistically significantly more likely to increase labor market work only.

Marriage is often viewed as an institution that improves women’s economic well-being by giving them access to men’s economic resources (e.g., Tyree and Treas, 1974 and Waite, 1995). Some family scholars note, however, that marriage also compromises women’s economic well-being by discouraging paid work while encouraging unpaid work among women (e.g., Gupta, 1999 and Hochschild and Machung, 1989). These scholars argue that marriage contains normative expectations about work appropriate for a husband or a wife. These expectations contribute to a gender asymmetric allocation of work in which husbands spend most of their work hours on labor market work while wives spend most of their time on housework. They are thought to severely disadvantage women’s economic position in the event of marital dissolution (e.g., Duncan and Hoffman, 1985, Peterson, 1996, Smock et al., 1999 and Weitzman, 1985). This presumed linkage between the normative expectations in marriage and the negative economic consequences of divorce for women has been the subject of very little scrutiny or explication. Researchers have not investigated the mechanisms underlying these consequences, even though they have measured the extent of the consequences (e.g., Peterson, 1996 and Smock et al., 1999). On a theoretical level, there are reasons to question the assumed linkage. Neoclassical economic theorization, for example, implies that economically vulnerable wives who anticipated a divorce would devote more time to labor market work and would be able to do so in an unconstrained manner. Working within this neoclassical economic framework, Johnson and Skinner (1986) provide the only previous empirical study to examine the effect of the foreseen risk of divorce on married women’s allocation of time to labor market work. They hypothesize that the foreseen risk increases labor market work hours among wives who are relatively economically vulnerable (i.e., housewives) but not among wives who are already accumulating human capital (i.e., working wives). We modify their hypothesis to develop the baseline perspective in the current study—the human capital accumulation hypothesis (See next section for Background). 1 The hypothesis posits that a decline in marital satisfaction will: (a) induce housewives with little work experience to increase the time allocated to labor market work–but not the time they devote to housework; and (b) have no major effect on the time allocation of working wives, who are already accumulating human capital. At the empirical level, however, the substantial negative economic consequences of divorce for women in a variety of industrialized countries (Burkhauser et al., 1991 and Cornell, 1990) raise questions about the neoclassical economic assumption that wives have few structural constraints on their capacity to increase labor market work hours. In this paper, we develop an alternative theoretical perspective that could help explain the disjuncture between neoclassical economic theory and the empirical observation in at least a subset of settings. This alternative view integrates two perspectives: the institutional perspective of status attainment, which emphasizes structural constraints over women’s human capital accumulation (e.g., Mincer and Polachek, 1974, Spilerman and Petersen, 1999 and Treiman and Hartmann, 1981; see Coltrane, 2000) and theories of “doing gender” or “displaying gender” (Berk, 1985, Brines, 1994, West and Fenstermaker, 1995 and West and Zimmerman, 1987). Specifically, the view is that wives’ behavioral conformity to normative expectations regarding activities appropriate to wives (i.e., doing or displaying gender) place constraints on the extent that wives increase labor market work time when they anticipate a divorce.2 These gender appropriate work activities include housework but typically exclude labor market work (Brines, 1994), and hence serve as constraints. According to Brines (1994), spouses compensate for work activities that are inconsistent with normative expectations in one aspect of their lives by “doing gender” in another aspect of their lives. Brines’ argument implies that wives may compensate for the negative evaluation by displaying gender when a decline in marital satisfaction reflects poorly on wives’ overall work performance; they may do so to avoid the negative consequences of gender non-conformity (e.g., social sanctions or disorganization in social interactions). Ideological and institutional conditions define the settings in which marital satisfaction reflects poorly on wives’ performance. Fragmentary findings in previous studies suggest that how wives display gender differs by their labor market positions, hence predicted outcomes of declining marital satisfaction vary by wives’ labor market positions. The resulting gender display hypothesis (articulated more fully in Section 2) suggests that, in settings where marital dissatisfaction reflects poorly on wives’ work performance, a decline in marital satisfaction should: (a) increase the likelihood that economically vulnerable housewives (i.e., those with little work experience) should allocate more time to both labor market work and housework; 3 and (b) reduce the likelihood that working wives allocate more time to labor market work. To empirically test the gender display hypothesis as an alternative to the human capital accumulation hypothesis, we investigate the influence of marital satisfaction decline on the subsequent chances that married women in Japan allocate more time to market work and/or to housework, conditional on their labor market position. We broaden the investigative scope relative to previous studies by simultaneously analyzing both changes in labor market work and housework time. In doing so, we assess the salience of “doing gender” in response to the perceived possibility of divorce in marriage as one form of institutionalized constraint on wives’ efforts to increase labor market work. Japanese wives provide an ideal opportunity to test the hypotheses; previous studies report that, due to the historical-political background of the country, they are expected to produce and sustain satisfactory marital life (Barlow, 2001, Cornell, 1989 and Cornell, 1990; see section titled Contemporary Japan as an Analytical Setting for additional reasons). Our analyses are based on data from the first four waves of the Japanese Panel Survey of Consumer Life, a longitudinal survey of Japanese women. This analysis extends current knowledge of the potential link between gendered normative expectations in marriage and the negative economic consequences of divorce. Evidence supportive of the human capital accumulation hypothesis would suggest that gender display is unlikely to tie normative expectations in marriage to the negative economic consequences of divorce. Such evidence would suggest that sources of wives’ notable economic disadvantage after divorce may not be located in marriage. However, evidence supportive of the gender display hypothesis in Japan would link conformity to normative expectations in marriage to the negative economic consequences of divorce in at least one industrialized country. It would imply that the extent to which wives could economically prepare themselves is constrained by gender display in ideological and institutional settings similar to Japan.