In working couples, the distribution of domestic tasks remains unequal
Sophie Ponthieux, Amandine Schreiber*[1]
The distribution of domestic and family tasks between partners may be analysed in terms of specialisation, economic power or social norms. The huge participation of women in the labor market should go together with a less specialisation between partners. However, even in couples where both partners work full time, the distribution of tasks is far from equal. In working couples, however, the proportion of tasks carried out by men is dependant upon the difference between their wage and that of their spouse. When a woman earns more, her partner’s share of housework is a little higher, but it is still not enough to cancel the gap.
In France, according to the 1998 French Time Use Survey (box 1), women living in couples devote on average twice the time to domestic and family tasks than their partner (Brousse, 1999). The sharing of domestic tasks is often analysed in the framework of Human Capital and Family theories. In this framework, time spent on domestic tasks by each partner results in a division of work which maximises the family's well-being, considered to be the primary objective: each partner specialises in activities for which the results of the effort put in is relatively the highest (Becker 1985). If women’s earning potential in the labor market is lower than men’s, it is more profitable for them to specialise in domestic tasks, whilst men devote their time to paid work. All the more reason, conversely, why a woman should work outside the household as her earning’s potential approaches her partners'; in other words, the market values of the two partners' time become more similar. In this case, from a strictly economic point of view, it is not more profitable for the household if the woman is responsible for domestic tasks, and these tasks should become more equally shared (Becker, 1985). In the real world however, if working women spend a little less time on domestic tasks than those who do not work, the difference between partners remains great, even when both partners work full time: in this case, women spend on average 4 hours per day on domestic tasks, compared to 2 ¼ hours for men. Despite the large presence today of women in the workplace, inequality in the distribution of domestic tasks seems to be particularly resistant to change.
Distribution of domestic tasks: allocation of time, economic power and gender norms
For the last fifteen years, the persistence of an unequal distribution of housework between partners has been the subject of a number of studies (Greenstein, 2000; Bittman al, 2003). On one hand, the analyses are based on the premises of sociological theories of resource and exchange (Sabatelli & Shehan, 1993); the unequal sharing of domestic tasks is therefore analysed in terms of men’s and women's relative contribution to the couple's resources, in terms of women's economic dependence or even in terms of bargaining power (Sorensen & McLanahan, 1987; Brines, 1994). This notion of bargaining power is present in decision making models developed from Chiappori's studies (Chiappori, 1988). In these models, decisions actually result from negotiation based on a “sharing rule”, in which the potential income of each partner comes into play (Moreau, 2000). The paper is not developed in the perspective of theses models. On the other hand, “gender”studies emphasise the importance of norms in determining social roles assigned to men and women (Badgett & Folbre, 1999). In this case, the inequality of distribution is determined by gender norms to which individuals adhere, i.e. by the ideas they have on what the social roles of a man or woman are. Existing studies do not allow to conclude between these two interpretations, which for that matter are not exclusive. For example, in the USA and Australia, Bittman & al(2003) show that the distribution of tasks between spouses is less unequal the more women contribute to the couple's resources. Inequality decreases up to the point where both spouses contribute equally to income, but when women contribute more than their partner to these resources – which may be interpreted as a “gender deviancy”– the distribution of tasks seems to go back to more “traditional”setting.
In studies of women's participation in the labor market or on professional gender gaps, the assignment of domestic tasks and family care is considered as an explanatory factor. The approach is different here: conversely, it analyses the impact of wage inequality and social norms on the distribution of domestic tasks. Data from the 1998-99 Time Use Survey allow us to conduct an analysis of this relationship in working couples (boxes 1 and 2).
In working couples, the unequal distribution of domestic and family tasks is the norm
As is the case for couples overall, working couples generally share out domestic tasks unequally (box 3). During the week, between 7am and 9pm, at least one in five women are performing a domestic or family task, a figure which falls slightly in the middle of the day. The proportion of men carrying out one of these tasks is almost always 15% lower, except for the peak of participation observed in early evening – true also for women. At weekends, when most couples have more free time, things are different (figure 2). Even more women perform a domestic or family task: more than half of all women do this in the late morning, and more than 30% throughout the afternoon. Men's participation is also much greater than in the week; this peaks in the late morning and more than 20% perform an activity throughout the entire afternoon.
Both during the week and at weekends, these differences reflect partly the “specialisation”of each partner in tasks of different nature (Zarca, 1990; Brousse, 1999). The “core tasks”of cooking, washing-up, housework and washing remain on the whole jobs performed by women and give a better idea of daily habits than the gardening and DIY, the only two tasks in which men participate more than women (figure 3).
One hour of free time per day less for women
Overall, all days included, women spend an average of 3 and a half hours on domestic tasks, against 2 hours spent by their partners (figure 4). During the week, less time is spent on domestic tasks by both partners, but the difference remains the same. At weekends, men's increased participation in domestic tasks does not bridge the gap that separates the two partners.
At weekends, women spend on average 4 hrs 20 minutes on domestic tasks. Taking into account time spent looking after children or other members in the household only widens the gap further: 20 minutes per day for men, double this figure for women. Ultimately, in couples where both partners are employed, and despite spending less time on average than men on professional work, women end with half an hour less of free time in the week and 4 hours in total each weekend (figure 4).
A distribution of tasks that is barely more equal when women earn more than their partner
In one in four couples, the woman's wage is at least equal to that of her partner (figure 5). In these couples, men are on average more likely to participate in domestic and family tasks. They carry out a higher proportion of it than in other couples, even though the total time spent by both partners on domestic tasks is not less than average (figure 5). This observation, in principle, conforms to an analysis in terms of relative economic power (box 2): the greater the difference between each partner's contribution, the more the person who contributes the most retains power with respect to the other partner, and the more he/she will be able to “negotiate”, in particular if it is a women, a more equal distribution of domestic tasks.
Nevertheless, couples where the woman earns more are on average younger, are less often married, have fewer children than other couples and tend to have a slightly higher standard of living than the average, all characteristics which may also influence the distribution of tasks (figure 6). But these couples can be distinguished above all by the professional characteristics of the partners: the level of education of women is on average higher than that of their partner – and also higher than the average level for women overall. These women are also more likely than their partners to occupy a “managerial”or middle management position. They are more likely to work in industry or construction than other women, sectors that are mainly male-dominated, whilst their partners, who, are less likely to hold managerial positions than other men, are more likely to work in a female-dominated sector such as education or health. Finally, these women work on average longer than their partners. Although these various factors explain why they retain more economic power than their spouses, they perhaps also reveal that these couples adhere to gender norms which are less traditional than the average couple.
A statistical analysis neutralising the impact of the couples' other characteristics allows us to study the specific influence of women's relative economic power on the proportion of domestic and family tasks performed by their partners.
To evaluate this specific effect, a Tobit model has been estimated; in order to take into consideration couples in which the man did not perform any domestic or family tasks on the diary day (this is the case of 198 couples out of the 1506 working couples questioned). The model analyses the influence of women's economic power – as defined in box 2 – on the proportion of tasks carried out by the man. A large number of other characteristics are taken into account. Each partner is thus described by his/her socio-professional category, sector of activity, the fact of having or not having irregular working hours, age, level of education, state of health, and the number of domestic or family activities carried out in other households during the previous four weeks. A specific indicator is introduced – just for men – of the frequency of their participation in domestic tasks during the preceding four weeks. Any possible particularities on the diary day are also taken into account (whether it was a weekday, a weekend day, a work day, an unusual day, etc.). Each couple is described according to its standard of living, marital status, number of children (total and number under 3 years old), type and size of their town or city of residence, occupancy status, dwellings’particularities (garden, workshop), use of paid housework services, and total time spent by the two partners on domestic and family tasks.
Only a small number of variables prove to be significant; the meaning of their influence is detailed in figure 7. The specific effect of women's relative economic power on the proportion of domestic tasks performed by their spouse is significant at the threshold of 1% and positive: based on given characteristics of a couple and the partners therein, a given standard of living, and taking into consideration the total time spent by the couple on domestic and family tasks, the more the woman's economic power increases, the higher the proportion of domestic tasks her partner takes on.
However, the variability in men's participation in domestic tasks is very small. In couples where the woman earns less than her partner, the woman's wage represents an average of 63% of her partner's, and men perform an average of 32% of domestic tasks (figures 5 and 6). In couples where the woman earns at least as much as her partner, the woman's wage represents on average 128% of her partner's, whilst men perform an average of 37% of domestic tasks. The influence of women's “economic power”seems therefore to be significant, but pretty small.
Box 1The Time Use Survey
The latest French Time Use Survey was conducted amongst 8,000 households from February 1998 to February 1999. It has already led to a large number of studies, the first of which were a major part of a 2002 issue of Économie et Statistique (352-353).
The survey brings together information on a number of different levels: a “Households”questionnaire provides information on households characteristics using socio-economic indicators (composition, characteristics of housing and equipment, financial resources); it also includes more specifically targeted information, particularly on the use of domestic services, but also on the exchange of services between households. Each person aged over 15 completes an “Individuals”questionnaire, including questions focusing on the one hand on professional activities and working conditions, and on the other hand on their use of time and various activities (cultural, sporting and recreational, but also in the home), as well as on their health status. In addition to this questionnaire, the originality of this survey consists in the logging by all members of the family of all their activities during the same given day, at 10-minute intervals. The main advantage of this method of collection, known as “the day after”method, is that it does not rely on the memory or judgement of the people surveyed, as was the case in previous surveys. When individuals make a note themselves retrospectively of the time spent doing various activities, they may, consciously or not, state that they have spent more time than they expected or desired to spend on a particular activity. Gender norms may influence time estimations that both men and women make when noting down time spent on domestic tasks. Finally, the survey has the advantage of not favouring one specific activity, allowing those surveyed to describe freely their various activities during the day.
Box 2Scope of analysis: couples where both partners are employed and relative economic powerMost studies on the distribution of housework in couples analyse either all couples, couples where the man is employed or, more rarely, couples where both partners are employed.
In this paper, the scope of analysis is restricted to couples where both partners are employed as waged and salaried workers, for the following reasons:
- The analysis of the distribution of domestic tasks performed by all couples multiplies structural differences (caused by the presence of retired couples, couples where both partners work, couples where just one partner works) and analytical approaches. The impacts of generation, activity type (freelance or employed) and occupation status cause different problems. In the case of occupation status for example, an unemployed women’s situation is not the same as that of a housewife (Letrait, 2002).
- Furthermore, it is difficult to obtain information on partners’contributions to household resources except for their salaries: the survey does not provide detail of income from inheritance or conveyance; data on income from self-employed workers are unreliable, and also difficult to individualise if both partners are self-employed.
Relative economic power expresses the difference between the contribution of one individual and that of his/her spouse to the couple's overall income. Numerous studies use an indicator, originally suggested by Sorensen & McLanahan (1987). Formally, RCW = (wagew – wagem) / (wagew + wagem).
Since women, in most cases, earn a lower wage than their spouse, the indicator is calculated from the women's viewpoint; for this reason it is often refered to as the “level of economic dependence”. The higher the indicator, the more economic power the woman has. In theory, the result may vary from -1 (the man is the sole contributor, the woman is “dependent”) to +1 (the woman is the sole contributor), with 0 meaning contributions are equal.
In cases where both partners are employed, the variation is never extreme. The data in the Time Use Survey for this type of couple produces an indicator which varies from -0.84 to +0.54, with an average of -0.167. This figure corresponds to the average difference between the woman's (0.416) and the man's contribution (0.583) to the couple's total wage.
Figure 1 – Domestic time on an average day according to women's activity status
Women in employment / Not employed / OverallFull time / Part time
Women / 3h55 / 4h29 / 6h45 / 5h01
Men / 2h14 / 2h14 / 1h52 / 2h07
W/M Ratio / 1.8 / 2.0 / 3.6 / 2.4
Including Child Care
Women / 0h31 / 1h18 / 1h19 / 0h46
Men / 0h13 / 0h15 / 0h15 / 0h14
W/M Ratio / 2.4 / 5.2 / 4.6 / 3.3
Scope of analysis: couples in which the man is employed
Source: Insee, French Time Use Survey 1998-1999
Figure 2 – Percentage of men and women performing a domestic task at different times of the day
Scope of analysis: couples where both partners are employed.Source: Insee, Time Use Survey,1998-1999
Box 3Domestic and family activities
In this study, domestic and family tasks are understood in the widest sense possible, including all the tasks in the following specification table. Some activities accounted for are actually at the very meeting point of work and leisure; this is the case for “shopping”, in that the data does not allow us to distinguish the type of shopping. This is also the case for some family activities; for example, looking after children includes both supervision of homework and more fun activities.