|Home | About | Journals | Submit | Contact Us | Français|
This article explored the association between economic activity and the contribution to domestic labor in a Middle Eastern society. Analyses were carried out on cross-sectional survey data from 5,998 individuals, aged between 18 and 64 years, in three poor communities in the suburbs of Beirut, Lebanon. Domestic labor was evaluated with a composite index that takes into account both the type of task performed and the level of involvement. Housework categories included core household chores, care giving, financial management, home management, and home/car maintenance. Results showed that women continue to do most of the domestic labor in the three communities. However, women's load of domestic labor decreased as they joined the labor market, whereas men's contribution to domestic labor increased with involvement in paid work. Relatively speaking, the difference in contribution to housework between house members engaged and not engaged in paid labor was much higher for women than it was for men.
Women have been traditionally designated to perform housework (Coltrane, 2000; Grana, Moore, Wilson, & Miller, 1993; Shelton & John, 1996). Studies in the literature have highlighted a shift in gender roles and an increase in women's participation in the labor force over the past four decades (Apparala, Reifman, & Munsch, 2003; Bolak, 1997; Corrigall & Konrad, 2006; Grana et al., 1993; McKeen & Bu, 2005; Spade & Reese, 1991; Starrels, 1994). A large body of literature has addressed the gendered division of household tasks in Western societies (Coltrane, 2000; Demo & Acock, 1993; Shelton & John, 1996; Twiggs, McQuillan, & Ferree, 1999). Studies have shown an increasing contribution of men especially when women participate in the labor force (Brines, 1994; Hossain & Roopnarine, 1993; Nakhaie, 1995). Research in Western societies has shown that the relative contributions of household members to domestic labor are affected by their labor force participation, whether they generate an income or not, and how dependent they are on others for economic support (Baxter, 2002; Bianchi, Milkie, Sayer, & Robinson, 2000; Blood & Wolfe, 1960; Brines, 1994; Coltrane, 2000; Greenstein, 2000; McFarlane, Beaujot, & Haddad, 2000). It follows that women contribute less to household labor when their involvement in paid work gives them an income and decreases available time to perform housework. As more women join the work force, researchers have hypothesized that their load of household tasks lessens and is absorbed by men within the household (Baxter, 2002; Brayfield, 1992) or by paid helpers.
On the other hand, the results of other research suggest that women's economic activity and income do not reduce their contribution to household labor in societies where marital power and financial resources remain in the hands of men (Diefenbach, 2002; Greenstein, 1996; McFarlane et al., 2000; Nakhaie, 1995; Sanchez, 1993; South & Spitze, 1994; Tichenor, 1999). However, in societies in transition from traditional gender roles toward more egalitarian arrangements, a shift in relative resources between men and women apparently does influence the division of labor in housework (Diefenbach, 2002).
Social science researchers have described enduring patriarchal relations in Arab families that privilege men and elders in the domestic division of labor (Joseph, 1996). A number of small-scale anthropological studies have shown the typical division of housework in various Arab settings (Cole, 1985; El-Missiri, 1985; Rassam, 1980; Rugh, 1984), and large-scale household surveys document labor force participation in Arab countries (Administration Centrale de la Statistique, 1998; Assaad, 2002; Hanssen-Bauer, Pedersen, & Tiltnes, 1998). Little empirical research links the two. The question remains: To what extent does paid work affect the division of domestic tasks among household members in Arab societies?
This descriptive exploratory study begins an attempt to answer that question. We examined the associations between gender, economic activity, and contribution to domestic labor in three disadvantaged urban communities on the outskirts of Beirut, Lebanon. Due to the diversity in its social, religious, economical, and political ideologies, Lebanon offers an appropriate field for testing whether the enduring nature of patriarchy undermines the effect of participation in the labor market on the contributions of individuals to domestic labor. Not only has the participation of Lebanese women in the labor force increased since the early 1970s (Jamali, Sidani, & Safieddine, 2005; National Commission for Lebanese Women, 1997; United Nations Population Fund (UNFPA) & Ministry of Social Affairs, 1997), but the changing and unstable economy, which is characterized by people's difficulties in finding appropriate jobs, lack of resources, and the modern aspirations of urban dwellers has certainly led to changes in the roles and interactions within households. This has led to the frequently asked question of whether Lebanese men are willing to share the housework with women now that women are sharing the breadwinning with men (Khalaf, 1993).
In the spring of 2002, the Faculty of Health Sciences at the American University of Beirut devised a survey, referred to as the Urban Health Study (UHS), which was designed to examine the health of families in rapidly changing areas surrounding the capital city, Beirut. A household labor component was incorporated into the survey to study the division of domestic labor in this urban setting.
Most researchers to date have analyzed household tasks in relation to married couples only (Baxter, 2002; Brines, 1994; Kalleberg & Rosenfeld, 1990; McFarlane et al., 2000; Nakhaie, 1995;Sanchez, 1993). There are few studies of the division of household labor among family members other than the married couple, such as children, other adults in the household, or single parent families (Demo & Acock, 1993; Manke & Seery, 1994; South & Spitze, 1994). Rather than focus on the couple only, we considered the participation in household labor of all adult members in the household. This was especially meaningful because, in the studied communities, the elderly and the unmarried adults are not encouraged to move out and live away from the family household.
The three chosen communities, namely Hay el Sellom, Nabaa, and Burj Barajneh Camp, lie within the limits of the city suburbs, and they are characterized by their poverty and lack of adequate infrastructure. Hay el Sellom consists of Muslim Shiites displaced from their previous residence in South Lebanon and the Beqaa Valley during the 15-year long civil war (1975–1990). Nabaa is a mixed community, predominantly Christian, displaced from Mount Lebanon and other parts of the country during the civil war; it has a high population turnover and a fair amount of foreign workers. Burj Barajneh Camp consists mainly of Muslim Sunnite Palestinian refugees displaced from their homeland since 1948.
Households were randomly selected based on a two-stage cluster sample proportional to the size of the three communities. A detailed sampling frame was specifically constructed for the present study. The target population consisted of those individuals living in housing units within the boundaries of the community. Of the homes visited, 3% refused to participate and 0.5% did not complete the interview. People were not given incentives to participate. Data on indicators related to socioeconomic and demographic characteristics and domestic labor were successfully collected from 2,797 households.
Data in this survey were collected by face-to-face interviews with one proxy respondent from each household. The interviewed proxy respondents were mostly women (78% women, 22% men) over 20 years old. Specially trained field workers from the studied communities asked a male or a female member of the household to assess the contribution of all adults in the household during the past week to the 28 tasks using a four-point scale that ranged from 0=never, to 3=always. This scale had the advantage of capturing involvement of household members in performing domestic tasks in the context of the study.
For the purposes of this article, only households with individuals aged between 18 and 64 years, who were not enrolled in school at the time of the survey, were included in the analyses. Hence the analyzed sample consisted of 5,998 individuals in 2,582 households: 744 in Hay el Sellom, 1,040 in Nabaa, and 798 in Burj Barajneh Camp.
The individual characteristics in the sample by community and gender are described in Table 1. The socio-demographic characteristics of the study population provide an elaboration on the context of the study. Most of the participants were married (62% of men, 66% of women) and Muslim (67% of men, 66% of women), with a mean age of 35 years for men and 34 years for women, and mean level of education of 7 years for men and 8 years for women. Among men, 77% were engaged in paid work during the past week; this proportion dropped to 23% for women. The mean working hours for men engaged in paid work were 56 h per week, with an average monthly salary of US$ 349. As for women engaged in paid work, the mean working hours were 45 h per week, with an average monthly salary of US$ 234. The average yearly household income, US$ 6,622, was about one-half of the national Lebanese average, which is nearly US$ 12,300, according to the latest published figures from 1997 (Administration Centrale de la Statistique, 1998). At the community level, Nabaa stands out as a predominantly Christian community (79% of men, 82% of women) with slightly fewer married individuals (55% of men, 62% of women).
The dependent variable, household labor, was defined as broad categories of housework and specific chores within each category. Similar to what has been previously reported in the literature, housework was evaluated in terms of the type of tasks that both men and women traditionally do (Coltrane, 2000; Greenstein, 2000; Twiggs et al., 1999) as well as their level of involvement in these tasks. Although the concept of housework itself varies from study to study, most researchers agree on four or five main household tasks, such as housecleaning, preparing meals, doing laundry, grocery shopping, and washing dishes (Coltrane, 2000). Most researchers have collected details on each chore (Baxter, 2002; Bianchi et al., 2000; Diefenbach, 2002; Greenstein, 2000; McFarlane et al., 2000; Nakhaie, 1995; Presser, 1994; South & Spitze, 1994; Twiggs et al., 1999), and a few have summarized housework contribution into one general question (Sanchez, 1993). We tried to be comprehensive in the specification of housework by including 28 chores classified into five major categories, namely core household tasks, care giving, financial management, home management, and home/car maintenance. The core household tasks include seven specific chores that deal with cleaning (rooms, kitchen, and bathrooms separately), meal preparation (washing dishes, preparing food), and laundry (washing, ironing). The care giving tasks consist of 11 chores that measure involvement in direct care for children of different age groups, as well as elderly, sick, and disabled members of the household. In addition, this category includes picking up, accompanying, and dropping off household members, as well as any participation in school-related matters, especially helping with homework. Financial management deals with managing payments (bills, debts) and household expenses, whereas home management combines the responsibility for purchasing service water, drinking water, heating supplies, home needs, and personal items. Finally, home maintenance, car maintenance, and car washing come under the maintenance category.
Previous researchers have used different ways to assess housework (Coltrane, 2000; Nakhaie, 1995; Twiggs et al., 1999). There is evidence that time spent doing a chore by itself is not enough to analyze the contributions of men and women to household labor, but rather, in addition to the type of chore, one must also take into account the frequency with which it is performed (Baxter, 2002; Twiggs et al., 1999). Based on this argument, we opted for a composite index to measure contribution to different categories of housework in terms of the involvement in performing a household chore, as well as the usual frequency of doing that chore, as described later in this section.
Because the household tasks are not all performed with the same frequency, rather, the frequency of performance largely depends on the nature of the chore, each task was given a weight for the frequency of performance, based on feedback from a few households in the three communities under study. Women in these households were asked to report the frequency of each of the 28 chores performed regardless of which person is performing them. Hence, frequent tasks, such as meal preparation and washing the dishes, that were reported to be performed at least four times per week were given the highest weight of 8; whereas less frequent tasks, such as home repair and car maintenance, that were reported to be performed on a monthly basis or less frequently were given the lowest weight of 0.5 (see Appendix).
The four-point scale (0 to 3) was treated as a continuous scale of the involvement in performing a task. The value of involvement in performing a task was weighted by the frequency of performing it (i.e., multiplied by the weight assigned to the frequency of performance). The weighted values of involvement for all individuals in the sample were then summed across tasks within a category of housework to calculate a mean weighted value of involvement for that category. Finally a composite index for all chores, across all categories of household tasks, was calculated to measure contributions to housework in general. This index ranged from 0 (the person never does any of the housework) to 430 (the person always does all the 28 household chores) (see Appendix).
Some of the care giving tasks are not applicable to every household, and the corresponding household members were given a zero value for those tasks; for example, in households with no children less than 1 year old, household members were given a zero value for care for children less than 1 year old.
The independent variables of interest were gender, community (Hay el Sellom, Nabaa, and Burj Barajneh Camp), and involvement in paid work, which was assessed with a dichotomous variable for involvement in paid work during the past week (Yes, No). We measured the economic activity by asking questions about labor force participation (in or out of the labor force) (International Labour Organization, 1983). We asked whether a household member worked in the previous week for a wage (cash or in-kind) in any type of part-time or full-time work, including self-employment, home-based services such as childcare, and casual labor. Household members who worked for a wage were classified as ‘involved in paid work.’ Those who did not work for a wage were classified as ‘out of the labor force.’
The analysis was two-fold. First, stratified by community, the mean weighted value of involvement for each of the five categories of household tasks, and the composite index for all household tasks, were compared by gender. Second, stratified by gender and community, the mean weighted value of involvement for each of the five categories of household tasks, and the composite index for all household tasks, were compared by working status. The Student t-test was used to test all hypotheses. Statistical significance for the Student's t-test was set at p<0.05.
The mean weighted values of involvement (MWVIs) in categories of household tasks, and the composite index, by gender and community are shown in Table 2. In general, the results show that women perform most of the household labor. In accordance with other published studies, women continue to assume the primary responsibility for the traditionally feminine tasks such as cooking, cleaning, doing the laundry, and attending to children's needs (Antill & Cotton, 1988; Kroska, 2003). The exceptions are in the tasks that are more traditionally associated with men, such as home management, including purchasing drinking water and kerosene for heating purposes, and maintenance tasks, including household and car maintenance. The differences in MWVIs in categories of household tasks, by gender, are statistically significant (p<0.05). The associations are the same across the three communities with the exception of home management in Burj Barajneh Camp, which appears to be more performed by women (MWVI=4.41 for men and 7.05 for women, t=−11.97, p=0.000). It is worth noting that, in general, men and women in Hay el Sellom performed household chores more often than their counterparts in Nabaa and Burj Barajneh Camp. This is evident by the elevated MWVIs in Hay el Sellom compared to those in Nabaa and Burj Barajneh Camp. The composite index also confirms that women do significantly more housework than men do in all of the three communities (total composite index=73.22 for men and 196.38 for women, t=−78.79, p=0.000). This finding is in accordance with the literature on the existing imbalance in the division of housework, which remains primarily women's responsibility (Coltrane, 2000; Demo & Acock, 1993; Hossain & Roopnarine, 1993; Starrels, 1994;Twiggs et al., 1999; Van Willigen & Drentea, 2001). The results also confirm previous findings on the division of housework among men and women aged 5 years and older in Hay el Sellom, Nabaa, and Burj Barajneh Camp, which revealed that housework is feminized, not only in adulthood, but also in the early stage of childhood (Habib, Nuwayhid, Merhi, & Myntti, 2005).
The mean weighted values of involvement in categories of household tasks, and the composite index, by paid labor are shown separately for men and women in Table 3. The summary measures, or composite indices, show that the relationship between household chores and paid work is consistent across communities, though it varies in magnitude. The difference between the contributions of non-working individuals and working individuals is proportionally much higher for women than for men (total composite index=208.54 vs. 157.42 for women; total composite index=60.05 vs. 77.38 for men).
Men engaged in the labor force contribute significantly more than men not engaged in the labor force to all categories of household tasks, except for core household tasks, including cleaning, meal preparation, and laundry. In Hay el Sellom working men contribute more to core household tasks than their non-working counterparts do, but this difference is not statistically significant. Working men who earn an income may prefer to maintain control of the budget, spending, and expenses, and may choose the tasks to which they wish to contribute, leaving to the women the more tedious housework, such as cleaning and meal preparation (Bianchi et al., 2000; Hossain & Roopnarine, 1993; McFarlane et al., 2000; Nakhaie, 1995; Starrels, 1994).
In general our findings agree with the theory that men feel more comfortable “helping out” with the housework when they are employed and contributing to the household income, because their traditional masculine status is not threatened, whereas non-earner men contribute less to housework as a means of asserting their traditional gender role (Brines, 1994; Greenstein, 2000). It could also be that unemployed men are ill or disabled, and therefore limited in their capacity to do housework.
The relationship between paid work and housework is quite the opposite for women. Women involved in paid work do significantly less housework than women not involved in the labor market, except for the chores that are traditionally considered part of the masculine role, such as repair and maintenance (see Table 3). These tasks are done less frequently and require less time than the core household tasks such as cleaning, doing laundry, and preparing meals. Hence working women continue to do these infrequent and occasional chores, and they do less of the more time consuming urgent and routine tasks (Milkie & Peltola, 1999). In fact studies have shown that working women “maintained responsibility for running the household and making sure things get done” (Tichenor, 1999, p. 642) though they might not actually be doing the work themselves (Bianchi et al., 2000; Coltrane, 2000). Part of the housework may be left undone or delegated to other members of the household, especially female children.
It is worth noting that when we stratified the sample by age (<20, 20–39, 40–59, 60+), working men were, in general, doing significantly more housework than non-working men across all age groups (p<0.05). As for women, those in paid work did less housework across all age groups, however, in the age group 40–59, working women did more maintenance than those not in paid labor. This suggests that women with income do more repair work around their homes.
Our data show that women's income did have an effect on their contribution to household labor in the three communities. Furthermore, the composite indices show that men engaged in paid labor contributed significantly more to household tasks than men who were not engaged in paid labor (see Table 3). Given that the three communities differ in their main religious composition, the question could be asked to what extent they differ from one another on the relationship between paid and domestic work.
Although all three communities share many cultural values, Nabaa stands out as less traditional than Hay el Sellom and Burj Barajneh Camp on several key variables. Nabaa is predominantly Christian, but has a greater religious diversity than the other communities; it also has a larger proportion of women involved in paid labor (32.2%) (see Table 1). In addition, 6% of non-working women in Nabaa attributed their non-participation to social restrictions (their spouse or parents disagree with their involvement in the labor force), whereas this figure is doubled in Hay el Sellom (12%) and Burj Barajneh Camp (15%). Also, 23% of women aged 35 to 39 years in Nabaa have never been married, whereas that figure is almost one-half (13%) in Hay el Sellom and Burj Barajneh Camp. The average household size in Nabaa was 4.3 persons per household, whereas it was 4.9 in Hay el Sellom and 4.6 in Burj Barajneh Camp. As for the total fertility rate, it was 1.9 children per women in Nabaa, whereas it was 4.1 in Hay el Sellom and 3.2 in Burj Barajneh Camp (Unpublished Urban Health Study data, 2002). Despite these differences, there was no significant difference in contribution to household labor by community for women engaged in paid work (composite index in Hay el Sellom=167.46; in Nabaa=151.92; in Burj Barajneh Camp=161.63) (see Table 3).
It is usually presumed that sectarian differences determine social and cultural differences in Lebanon. The assumption is that Christians may be more egalitarian in ideology and lifestyle and that Muslim groups, in particular the Shiites, may be more ‘traditional’ and live by stricter gender norms and rules of patriarchy. Contrary to what could be expected, our results showed that men's involvement in housework, whether or not they engaged in paid labor, was highest in the ‘traditional’ community of Hay el Sellom.
A comparison of the composite indices across communities showed that among those engaged in paid work, men in Hay el Sellom contributed significantly more to household tasks (composite index=99.29, t=4.40, p=0.000) than did men in Nabaa (composite index=74.41; t=3.64, p=0.000) and in Burj Barajneh Camp (composite index=64.82; t=4.81, p=0.000). The same pattern was observed among men not involved in paid work when compared by community (see Table 3).
Our study has limitations. The use of a proxy respondent to get information on the contribution of all household members to housework may have biased the results in either direction. This “reporting bias,” which results from “uneven social perceptions of the appropriate roles of women and men in the domestic division of labor,” has previously been reported in the literature (Press & Townsley, 1998, p. 190). In fact, a small proportion of the respondents in our study were men (22% men and 78% women), and the men were found to report higher involvement in housework than women typically reported for the men in their households. This could mean that we overestimated men's contribution to housework. In addition, discrepancies between couples on how much domestic labor they say they perform could not be ruled out. Therefore, the interpretation of our results should take these issues into consideration. Because the three studied communities were uniformly disadvantaged, we were unable to examine how class affects the relationship between paid and domestic work. This exploratory research cannot test hypotheses, but has value in forcing us to question basic assumptions about gender, patriarchy, and contributions to housework.
To date, most of the research on the association between housework and paid work has been conducted in Western industrial countries (Baxter, 2002; Kalleberg & Rosenfeld, 1990; Nakhaie, 1995; Twiggs et al., 1999). The present study has shed light on the association between housework and paid work in disadvantaged communities in the Arab Middle East, where cultural norms give authority to men and gender roles are clearly established.
Our findings both complement and challenge the existing literature on the relationship between paid and domestic work. In our three studied communities, women do less housework when involved in paid labor, and working men do more housework than those with no incomes. Nevertheless, women continue to do the larger share of housework regardless of their involvement in paid labor. In the Muslim community, where more traditional norms guide many aspects of family life, we found that men contribute a greater share of housework, which is contrary to general expectations.
This project was supported by grants from the Wellcome Trust, Mellon, and Ford Foundations. The authors thank Huda Zurayk and Cynthia Myntti for their comments on earlier drafts of this article.
|Household tasks||Weight||Maximum score|
|Core household tasks||144|
|Direct care—care of children <1 year old||8|
|Direct care—care of children 1–3 years old||8|
|Direct care—care of children 4–9 years old||8|
|Direct care—care of children 10–14 years old||8|
|Direct care—care of elderly >60 years old||8|
|Direct care—care of disabled||8|
|Direct care—care of sick member||8|
|Transport—dropping off, picking up member||8|
|Transport—accompanying other members||2|
|School—minding school issues||0.5|
|School—assisting with homework||8|
|Payment of bills and debt||0.5|
|Managing household expenses||2|
|Shopping for home needs||4|
|Buying service water||2|
|Buying drinking water||2|
|Buying heating supply||0.5|
|Purchasing personal items||0.5|
Weights assigned for the frequency of performance:
A task that is performed less than once per week was given a weight of 0.5.
A task that is performed once per week was given a weight of 2.
A task that is performed two to three times per week was given a weight of 4.
A task that is performed four or more times per week was given a weight of 8.