Gender Inequality and the Division of Household Labor: a Comparative Study of Middle-class, Working Married Men and Women in Japan and Indonesia

Seeking to offer a balanced perspective to gender inequality and the division of household labor among middle-class, working married men and women in Japan and Indonesia, this paper examines the effects of individual-level characteristics (relative income, working hours, gender ideology) as well as the country-level factors (e.g. GEM: Gender Empowerment Measure) on the dynamics of housework distribution between spouses in both countries. Statistical analyses show a number of significant correlations between these variables, among which gender ideology seems to be of particular importance. Perhaps the most enlightening finding of all is that despite their lower GEM rank compared to Japan, the Indonesian respondents have relatively egalitarian division of labor in their households. This finding provides a new insight that GEM, which emphasizes the political economy aspects of a country, may not be sufficient to capture gender disparities without considering other socio-cultural factors in the complexity of day-to-day actual division of housework.


Introduction
With the approach of the 21 st century, the examination of gender inequality in any given society is extended to almost all areas of social studies. It is believed as one of the basic way to presage the possible emergence of a gender-equal society, a situation that has become one of the main goals of most countries in the world. To seek gender equality in all aspects of socio-cultural life, one has no option but to explore the ongoing gender inequality in their society first.
One of the most fundamental forms of gender inequality in society could be seen at the basic level of social unit, namely at the division of labor in the household. As married women's labor force participation has increased in the past several decades (Cohen & Bianchi 1999;Goldin 1990), the traditional division of household labor ---"women at home and men at work"---has been eroding. Nevertheless, these women still do the majority of the housework (Bianchi et al. 2000;Coltrane 2000), which is to say that women's primary task is still considered to be housekeeping and child rearing.
Japan, which has undergone a period of restructuring from an industrial to a post-industrial economy since the 1960s and has in particular promulgated the Equal Employment Opportunity Law in 1985, has seen a rise in women's participation in labor force, with women constituting 40.7 percent of the total paid workforce in 2000. The workforce participation rate of Japanese women has shown a steady upward trend, and approximately half of all women between fifteen and sixty-five years of age are engaged in waged labor. A significant majority of them (56.9 percent as of 2000) are married, and single women constitute only 33.1 percent of the entire women workforce (Bureau of Statistics, Japan Ministry of Public Management, 2000). On the surface, these statistics suggest that women do not regard domestic labor as their only option and play a highly significant role in the labor market.
In spite of their new role as both housewives and workers, the division of household labor remains unequal for Japanese women. Gender segregation of household task persists (Fansellow 1995; Gelb & Palley et al. 1994;Lebra 1984;Iwao 1993). Women mostly perform the "core" traditionally feminine tasks, while men concentrate on more episodic or discretionary housework tasks. On average, a man spends only 9 minutes per day on week days on such household tasks as cooking, cleaning, and washing, while a woman has to devote 2 hours and 20 minutes to household tasks (NHK Hoso Bunka Kenkyusho, 2001: 36).
Both the basis for and the pervasiveness of this gender inequality in the division of household labor such as evident in the Japanese experience have been researched extensively in recent years, especially in light of the increasing number of women's participation in the work force. However, while these previous studies focused on the effects of individuals' characteristics such as relative-resources, time available to meet the demands made at home and at work and gender-ideology, (Kiyomi 1987;Ross 1987;Coverman 1985;Takahashi 1983), the current study argues that macro-level factors are equally important in determining gender equality in the division of household labor.
Of particular importance in this regard is the fact that in cross-national comparison, differences in country-level gender inequality ---in wages, career trajectories, or political power---may influence the effect of individual level factors on the division of household labor. For example, in Japan where gender inequality in many areas of social life is relatively severe, the resources a wife brings to the home (e.g., income), how many hours she works outside the home, and what she thinks about appropriate gender roles may not matter much in determining the division of household labor. On the contrary, in Scandinavian countries, which have achieved more egalitarian social conditions, individual women may be able to negotiate the division of housework more effectively according to their individual-level characteristics (Fuwa, 2004: 751-752). Kamo's (1994) detailed study of the division of labor in industrialized countries offers further evidence in support of this thesis by pointing out that models based on individual characteristics have weaker explanatory power for Japanese couples' division of housework compared to its explanatory power for American couples. Furthermore, in studies that compared more egalitarian countries such as the United States and Sweden (Calasanti & Bailey, 1994) and Australia, Canada, Norway, Sweden and the United States (Baxter, 1997), the cross-national differences are less clear due to the studies' sole focus on individual-or couple-level.
Furthermore, as differences between Japan and other industrialized countries in terms of the division of household labor have become evident, cross-national comparison is undeniably a fundamental way to understand the pace of change and dynamics of the division of housework across countries. For instance, in a comparative study of Japan and the United States, Kamo (1994) found that in the United States, a husband's contribution to household work is 29 percent, while in Japan, a husband's contribution is only 13 percent. A broader comparison among 22 industrialized countries found a considerable difference in the husband's share of housework across countries. Husbands in Japan share housework the least, while husbands in the United States and Norway share housework the most; albeit wives still do more than their husbands even in the most egalitarian countries (Batalova & Cohen, 2002).
In this context, while cross-national comparison studies between Japan and other developed countries have been researched extensively (Kamo 1994; Batalova & Cohen 2002;Fuwa 2004), a comparative study between Japan and developing countries such as Indonesia has not been adequately examined. The aim of the current study, therefore, is two-fold. First, to give a balanced perspective to gender inequality and the division of household labor in Japan and Indonesia by examining the effects of individual-level characteristics (relative income, working hours, gender ideology) and country-level factors (e.g. GEM: Gender Empowerment Measure) on the dynamics of housework distribution between spouses. Second, it seeks to contribute to the advancement of knowledge of gender inequality in the division of household labor by bringing to light examples and practices from the developing country's experience.
Theories and hypotheses. Blumberg (1984) introduced the concept of "nesting" micro-level units such as households in a series of macro-level units ranging from community, class, state to the world economy. Blumberg suggested that "society's various micro-and macrolevels can be conceived of as a 'nesting' system in which women's relative economic power and status may vary at least somewhat independently at each level" (Blumberg, 1984:48). This multi-level conception of marital power incorporates the dynamics of power relationships in macro-level units into micro-level negotiation dynamics.
Thus, women's economic power, which can be used in negotiation with their husbands, is a function of individual power resources and macro-level male domination in economic, political, and ideological areas. Women may posses high income relative to their husbands, but their "nett economic power", when used in actual bargaining, may be reduced by male dominance at the macro level. Therefore, male control over the political economy and male dominated ideologies at the macro level may act as a "discount rate" against the power of individual women's resources (Blumberg 1984;Blumberg & Coleman 1989). Blumberg and Coleman (1989:234) maintained that "male control of the top echelons of the political economy affects the national policy agenda, the opportunity structure that women encounter, and the prevailing ideology of what women's place in that structure should be". The unequal distribution of opportunities may affect couples' calculations regarding women's present and future career trajectories. Also, dominant gender ideologies that devalue women's work may condition the negotiations over value of women's employment.
Relative resources. Relative resources perspective conceptualizes the division of labor as "an outcome of negotiation between husbands and wives who use whatever valued resources they can to strike the best deal on behalf of self-interest".The individual with more resources uses his or her advantage to avoid housework, as empirical studies have confirmed with regard to income (Bianchi et al., 2000). This perspective can be broadened by incorporating Blumberg's (1984) theory of macro-level gender inequality. Individual wives with higher earnings should have more power in the negotiation and thus share housework more equally. However, in countries where women's average wages are higher relative to men's and where women can expect more equal career opportunities and mobility relative to men, women's relative income is not as heavily "discounted". Individual women may see more opportunity for their careers, enhancing the power derived from their income in the negotiation over housework.
Time availability. This model focuses on rational time allocation, positing that the amount of time spent in domestic labor depends on the available time for housework. Many studies have found a positive relationship between wives' employment or work hours and husbands' housework (Bianchi et al. 2000, Kamo 1988). Applying Blumberg's theory of macro-level gender inequality to the time-availability perspective, this study first posits that individual wives who work more in the labor force should share housework more equally with their husbands. However, in countries where the average worth of women's work hours is higher relative to that of men's (i.e., not "discounted" at the macro-level), couples in which wives participate fully in the labor force might place higher value on the wives' work hours. Thus, such couples might rationally assign a greater share of housework to the husbands.

Gender-ideology.
As mentioned before, the gender-role approach posits that men and women are socialized to conform to socially constructed gender roles, and housework is divided accordingly. Previous studies have found a positive relationship between husbands' and/or wives' egalitarian gender role attitudes and a more equal division of housework (Kamo, 1988). This study posits that wives who have an egalitarian gender ideology should share housework more equally. Macro-level gender inequality, however, will influence this effect. Since male domination affects the prevailing gender ideology, in countries where women's economic and political power approaches that of men's, gender ideologies that encourage the traditional division of labor will be weaker. Given weak social control of gender roles, individual women who believe in an egalitarian division of housework may realize their attitudes more effectively. To summarize, it is expected that individual-level factors will have stronger effects on the division of housework for women who live in countries with more gender equality. In contrast, for women in countries where gender inequality is more severe, individual factors should have weaker effects on the division of housework.

Methods
Data and method. The research methods used in the current study were quantitative approaches, consisting of a survey method and a collection of statistical data pertinent to the country-level political and economic gender inequality. Since this study intends to determine whether relative resources, time availability, and gender ideology affect the division of housework differently across countries, as well as whether country-level gender inequality helps explain these differences, such qualitative approaches seemed most appropriate. As an important addition to the questionnaire data, an in-depth interview that included questions not asked in the questionnaire was also given to selected informants.
In obtaining the respondents for this study, the research team relied on the resources and networks made available through the Faculty's cooperation with Ochanomizu University and Nihon University in Tokyo. Two research sites were selected: (1) Urban Tokyo and Fukuoka for gender inequality in the division of housework of the Japanese, and (2) Urban Jakarta, Tangerang and Bekasi for gender inequality in the division of housework of Indonesians. For each of the two sites, a purposive sampling method was used to select a sample of married working men and women between 25-65 years of age belonging to urban middleclass households who reside in the neighborhood of each respective campus.
As mentioned earlier, both men and women of the middle-class group were selected as respondents in this study due to the fact that in stark contrast with Japan, in which middle-class groups constitute the vast majority of its population, Indonesia has a wider degree of diversity in social class. Therefore, in order to make a compatible cross-national comparison, middle-class households living in urban areas were regarded as the most appropriate respondents for this study due to the relative similarities in their educational background, occupation, and income level. Consequently, urban Tokyo and Jakarta, as the respective capital cities of Japan and Indonesia, were selected as research sites because these areas could be regarded as the most representative living areas for the middle-class groups in both countries.
As this study is limited to men and women of the middle-class group, the operational definition of "middle class" encompasses individuals with the following occupations: university professors, company executives, entrepreneurs, professionals, teachers, and office employees.

Dependent and Independent Variables
Dependent variables. The dependent variable is the division of household labor between husband and wife. This variable comprises household tasks referred to as female tasks. Measurement of the division of housework variable was based on responses to seven questions about three types of household tasks: household chores (cooking, doing dishes, doing laundry, cleaning house, grocery shopping), child-care, and elderly-care or caring for sick family members. Each of the responses was then measured on a five-point scale: 1 (the wife always does the task), 2 (the wife usually does), 5 (about equal), and 4 (the husband usually does), and 3 (the husband always does the task). The responses to the seven questions were summed and then divided by the number of complete responses, with higher scores indicating that wives and husbands have egalitarian division of household labor.
The current study also took into account the prevalence of the extended family system and the availability of housekeepers in Indonesia, both of which may have an intervening effect on the division of household labor among the Indonesian respondents. In order to minimize such an intervening effect, a specific scoring system was implemented, in which the household tasks carried out by extended family member(s) or housekeeper(s) were assigned a score of 3 (about equal) with an additional remark as to who (i.e. the husband or the wife) is considered responsible for the task in the household.

Independent Variables
Individual-level variables. There are three main individual-level independent variables, namely, relativeresources, time-availability, and gender-ideology.
Respondents were asked to respond to questions about these individual-level independent variables. For relative-resources variable, respondents were asked to respond to questions about actual income. For timeavailability variable, respondents were asked to respond to questions about working hours, since working hours reflect the time constraints on housework more accurately. The more time a wife commits to the labor market, the less time she is available for housework. The gender-ideology variable, on the other hand, asked respondents to respond to statements such as "A job is all right, but what most women really want is a home and children", "Being a housewife is just as fulfilling as working for paid-job", "A man's job is to earn money; a woman's job is to look after the home and family", "All in all, family life suffers when the woman has full-time job", "A pre-school child is likely to suffer if his or her mother works". The responses were measured on a fivepoint scale; from strongly agree (0) to strongly disagree (4). The responses were summed and then divided by the number of complete responses to these questions, with higher scores reflecting a more egalitarian gender ideology.

Control Variables.
The control variables in this study consisted of age, gender, level of education, existence of child, and existence of additional help (from extended family member or housekeeper), as well as religion and ethnicity, particularly for the Indonesian respondents.

Country-Level Variables.
To measure a country's level of gender inequality, the Gender Empowerment Measure was taken from the Human Development Report (2009) of the United Nations Development Program (UNDP). UNDP created the Gender Empowerment Measure (GEM) to capture gender disparities in "women's participation in political decision-making (political power), their access to professional opportunities (career trajectories) and their earning power (wage)" (UNDP 2009:72). GEM was constructed from the combination of the percentage of parliamentary seats held by women, the percentage of administrators and managers who are women, the percentage of professional and technical workers who are women, and women's share of earned income compared to that of men. GEM is one of the few measurements that provide data on gender inequality in economic and political power across a broad array of countries.
Analytical model. The conceptual relationship between individual characteristics and country-level political economic gender inequality in the division of household labor in Japan and Indonesia is shown in Figure 1 and 2.

Method of analysis. A cross tabulation test and t-test
were employed to examine whether the effects of key individual-level factors differed by country and whether those differences could be explained by country-level gender inequality. As an important addition, the study incorporated qualitative data from the questionnaire and interviews into the analysis of the data.

Research limitation.
Since this is a comparative study between two countries with different socio-cultural backgrounds, there are a number of methodological limitations.
First, in the construction of the questionnaire, there were some socio-cultural discrepancies that needed conceptual adjustments, such as the social concepts of working pattern and employment status, relative income, cultural norms and values of ideology, gender, religion, ethnic group, and the existence of extended family (s) and housekeeper (s).
Second, in terms of sampling method, this study faced difficulties in finding respondents who were simultaneously caring for children and the elderly. Therefore, the respondents were divided into four groups, namely, respondents with no children, respondents with children between 0-5 years old, respondents with children between 6-18 years old, and all respondents.
With these limitations in mind, it should be noted at the outset that the findings of the study only represent the respondents with specific characteristics, and could not be regarded as a representation of Japanese and Indonesian society as a whole.

Results and Discussions
Demographic Characteristics. There were a total of 185 Japanese respondents from Tokyo, Osaka, Fukuoka, and 204 Indonesian respondents from Jakarta, Tangerang, and Bekasi in the survey conducted in from November 2009 to March 2010. As Figure 3 indicates, most of the Japanese and Indonesian respondents belonged to the age group of 35-44. This group comprised 43.6% of the total Japanese sample and 38.1% of the total Indonesian sample. Twentyfive point nine percent (25.9%) of the Japanese respondents and 27% of the Indonesian respondents belonged to the 50-59 year-old age group, while 27.6% of the Japanese respondents and 26.4% of the Indonesian respondents belonged to an age group of 30-39 years.
While there were more female than male respondents in the Japanese sample (54.1% and 45.9%, respectively), the proportions of male and female respondents in the Indonesian sample were equal (50% for each) ( Figure 4).
(a) Japanese Respondent (a) Indonesia Respondent

Figure 3. Age Group Distribution of Respondents
As shown in Figure 5, most of the Japanese and Indonesian samples indicated that their highest educational attainment was college. This group comprised 38.9% of the total Japanese sample and 54.9% of the total Indonesian sample. A large percentage of Japanese respondents (24.3%) answered that their highest educational attainment is high school, as compared to only 11.8% of their Indonesian counterparts. Figure 6 shows the frequency distribution of the respondents based on occupation. Employees comprised the largest group of respondents, both for the Japanese (59.5%) as well as the Indonesian (50%) sample. While a large number of entrepreneurs (15.7%) and civil servants (14.2%) belonged to the Indonesian sample, 12.4% of the Japanese sample was composed of professionals.
As Figure 7 indicates, most of the Japanese and Indonesian respondents belonged to the employment status of full-time employee. This employment status comprised 78.9% of the total Japanese sample and 77.9% of the total Indonesian sample. Figure 8 shows the frequency distribution of the respondents based on weekly working hours. Most of the Japanese and Indonesian respondents admitted to working for 40-44 hours per week. This group of respondents comprised 33.5% of the total Japanese sample and 44.6% of the total Indonesian sample. Figure 9 presents the frequency distribution of the respondents based on monthly income. While most of the Japanese respondents (90%) belonged to the upper middle income group (USD 3000 -USD 6000 per month) by World Bank standard, most of the Indonesian respondents (90%) belonged to the low income group (USD 400 -USD 600 per month). However, it should be noted that both groups of respondents belonged to the middle strata income according to each respective country-level standard.  One of the prominent differences between the Japanese and Indonesian sample was in religious background and ethnicity. Considering that the majority of Japanese people are ethnic Japanese --except for a number of minority groups such as Okinawans, Burakumin, Zainichi Koreans, Ainus, and foreign workers--and that the Japanese religious system is non-exclusivist, eclectic, and syncretic (Sugimoto 2007), survey on both religious background and ethnicity was limited for Indonesian respondents only. As Figure 11 and 12 indicates, the Indonesian respondents were predominantly Muslim (87.7%) and were mostly ethnic Javanese (49%). The second prominent difference between the Japanese and Indonesian sample was the presence of housekeeper (s). As Figure 13 indicates, all the Japanese respondents do not have housekeeper (s) (100%) to give them assistance with the household labor, while their Indonesian counterparts mostly have housekeeper(s) living in the same household (53.9%). Individual-level descriptive statistics in Japan and Indonesia. Table 1 shows that on average, all groups of respondents in both Japan and Indonesia have a relatively modest division of household labor. In terms of household chores, Japanese and Indonesian respondents show relatively the same mean scores, which were 1.83 and 1.84, respectively. The highest mean score of household labor belongs to the Japanese respondents with children between 0-5 years old (2.00) and the Indonesian respondents with children between 6-18 years old (2.00).
Meanwhile, the lowest mean scores of household labor were found for both Japanese respondents and Indonesian respondents with children above 18 years old, with mean scores of 1.76 and 1.77, respectively. However, since a score of 3.00 indicates an equal division of labor, this average division of labor in all groups of respondents indicates that individual men and women may relatively be able to negotiate the division of housework according to their individual-level characteristics.
The mean score of the division of household labor for the group of Japanese respondents with no children was 1.91, which is slightly higher than their Indonesian counterparts (1.83). Statistically, this number suggests that wives and husbands are sharing household labor relatively equally. In terms of relative income, there was a significant difference between the Indonesian (1.39) and the Japanese respondents (1.88). These numbers represent the means of income of both countries' respondents.
With a minimum value of 1 and a maximum value of 3, the Japanese respondents had higher levels of relative income compared to their Indonesian counterparts. In terms of time availability, there was a slight difference in the mean scores of the Japanese (1.69) and Indonesian respondents (1.77), indicating that the Japanese respondents spend more time at work compared to the Indonesian respondents. In terms of gender ideology, the Indonesian respondents (1.96) appeared to have quite an egalitarian gender ideology compared to the Japanese respondents (1.78).

The effects of individual-level characteristics on the division of household labor in Japan and Indonesia:
Relative resources, time availability and gender ideology. The first objective of this study was to examine similarities and differences of the effects of relative income, time availability, and gender ideology on the division of household labor for both the Japanese and Indonesian respondents. A test of the correlation between individual-level characteristics and the division of household labor for respondents in both countries yielded a number of significant associations, some positive and others negative. This means that individual men and women are relatively able to negotiate the division of household labor according to their individual assets, though it has become evident that this does not necessarily lead to an equal division of household labor.
Furthermore, similarities and differences between the Japanese and Indonesian respondents were also found in terms of the associations of the three aforementioned variables (i.e., relative income, time availability, and gender ideology) with the division of housework. On the one hand, similarities were found on the ways in which relative resources affect negotiation over elderly care and gender ideology, as well as on the division of household chores and child care. On the other hand, differences were evident in the association between relative resources and time availability and on household chores, as well as between both time availability and gender ideology and elderly care.
Lastly, it became evident in the course of the study that two of the control variables age and presence of extended family (s) and housekeeper (s), affects the division of household labor for the Indonesian respondents, such that couples who live with their extended family (s) or who have housekeeper (s) living in the same household tend to share housework unequally. The reason for the inequality is that working women usually receive assitance from other female family members such as daughters, mothers or mothers in law, or paid domestic help. In this way, the existing gendered division of labour is neither challenged nor disrupted, as women still remain responsible for household tasks.
The Japanese Respondents: Relative income on elderly care. Japanese respondents with children above 18 years old, relative income has a negative relationship with the division of household labor in terms of elderly care (r=-.246, p= .018). This means that individual wives with higher earnings do not have more power in the negotiation over elderly care and thus do not share the labor more equally with their husbands. Although they have higher income relative to their husbands, they do not use their income as an advantage to avoid elderly care, and as a result, the division of elderly care between couples remains unequal. Thus, while a wife's decision-making power may increase in proportion to her income contribution, her ability to redistribute elderly care does not.
Time availability on elderly care. Furthermore, that for Japanese respondents with children above 18 years old, time availability has a negative association with the division of household labor in terms of emotional support for elderly care (r= -225, p= .027). This means that wives with longer working hours or who work full time do not share elderly care more equally with their husbands. Quite the contrary, they may not be able to "rationally" assign a greater share of elderly care to their husbands. In other words, the division of labor in terms of elderly care remains unequal no matter how many hours the wives work outside the home. While it does appear that husbands participate in elderly care when their wives become employed, the portion is relatively small. Indeed, a husband's relative share of elderly care increases when his wife works, but there was a strong tendency among the Japanese women to remain as the primary caregivers of impaired elderly relatives.
Gender ideology on household chore. As expected, a test of correlation between gender ideology and household chores among the Japanese respondents with children above 18 years old showed a positive correlation (r= .278, p= .000), This means that wives who have an egalitarian gender ideology share household labor more equally with their husbands. However, there is a tendency that, given strong social control of gender roles, individual women among the Japanese respondents who believe in an egalitarian division of household chores may not realize their attitudes more effectively. Macro-level gendered division of labor is argued by many scholars (Imamura 1996; Broadbent 2003; Suzuki 2007) to contribute to this effect to some extent. Since male domination affects the prevailing gender ideology in a societal level, the gender ideology of individual wives does not matter much in determining the division of household labor with their husbands, as indicated by this finding.
Gender ideology on child care. In gender ideology has a positive association with the division of household labor in terms of child care (r= .412, p= .001), especially for the Japanese respondents with children between 0 to 5 years old. This means that wives who have more egalitarian gender ideology share child care more equally with their husbands. On the contrary, husbands with more traditional gender ideology are less likely to be involved in child caring.
Gender ideology on elderly care. One of the most interesting findings is that for the Japanese respondents with children above 18 years old, gender ideology has a negative association with elderly care (r= -152, p= .052). Previous studies have shown that egalitarian gender ideology is associated with women performing a larger proportion of housework within the family, implying a positive relationship between husbands' and/or wives' egalitarian gender role attitudes and a more equal division of housework (Kamo 1988;Presser 1994). In stark contrast, the study has found that a more egalitarian ideology either had no effect on husbands' contribution or slightly decreased the amount of time spent by husbands in elderly care. Relative income on child care. It is evident that relative income has a positive association with the division of household labor in terms of child care (r=.0318, p= .015). This means that wives with higher earnings have more power in the negotiation and thus share child care more equally with their husbands.
Relative income on elderly care. In accordance with their Japanese counterparts, Indonesian respondents with children between 0-5 years old showed that their relative income has a negative relationship with the division of household labor in terms of elderly care (r= -267, p= .031). This means that individual wives with higher earnings do not share elderly care more equally with their husbands. Furthermore, this means that although they have higher income relative to their husbands, they do not use their income as a leverage to negotiate elderly care more equally. However, as explained above, the higher earnings have enabled individual men or women to hire domestic helper (s) to assist them with the care for the elderly. Such assistance makes elderly care on a daily basis more manageable, though the division remains unequal. Women still play a significant role in taking responsibilities for the day-to-day actual care.
Time availability on elderly care. Furthermore, in stark contrast with their Japanese counterparts, data from Indonesian respondents with children between 0-5 years old suggested that time availability has a positive association with the division of household labor in terms of personal support for elderly care (r= .113, p= .018). This means that wives who have longer working hours or work full time do share elderly care more equally with their husbands. In other words, the division of labor in terms of elderly care tend to be equal for wives who have longer working hours or work full time.
While it does appear that husbands participate in elderly care when their wives become employed, the portion is relatively small. Indeed, there is a strong tendency for the women among the Indonesian respondents to be assisted by extended family members (daughters, sisters) and/or their live-in domestic helper (s) in day-today actual work, as previously presented.
Gender ideology on household chores. As expected, a test of correlation between gender ideology and household chores among all the Indonesian respondents yielded a positive coefficient (r= .185, p= .003). This means that wives who have an egalitarian gender ideology share household labor more equally with their husbands.
There is a tendency that given enabling social control of gender roles, individual women among the Indonesian respondents who believe in an egalitarian division of household chores may realize their attitudes more effectively. Since males do not as strongly dominate the prevailing gender ideology at a societal level, the gender ideology of individual wives has a larger contribution in determining the division of household labor, as indicated by the current finding. Furthermore, in the Indonesian context, it can be argued that the enabling social structures characterized by the presence of extended family member(s) and housekeeper (s) are one of the important factors that enable wives to work full time and to carry out household labor simultaneously.
Gender ideology on child care. In accordance with the above finding, gender ideology has a positive association with the division of household labor in terms of child care (r= .310, p= .001) for the Indonesian respondents with children between 0 to 5 years old. This means that wives who have more egalitarian gender ideology share child care more equally with their husbands. On the contrary, husbands with more traditional gender ideology are less likely to be involved in child caring.
Discussions: Similarities and differences relative resources. Relative resources hypothesis posits that the greater the difference in favor of the husband's paycheck, the less housework he did. While relative resources were found to affect one dimension of the Japanese household labor, namely elderly care, they affect the Indonesian respondents on three household tasks: household chores, child care, and elderly care. Surprisingly, this study did not find support for the effect of relative power as measured by the difference in husbands' and wives' earnings, both for the Japanese and the Indonesian respondents. Thus, while a wife's decision-making power may increase in proportion to her income contribution, her ability to redistribute elderly care does not. In sum, for both groups of respondents, individual wives with higher earnings do not share elderly care more equally with their husbands. Furthermore, although they have higher income relative to their husbands, they do not use their income to their advantage in negotiating elderly care more equally.
In the Japanese context, the gender ideology which maintains that nurturing children, husband, and elderly parents are a woman's primary tasks plays a more major role in determining the division of housework compared to the relative resources aspect. As argued by Imamura (1996), the explanation lies on the cultural idea of women as primary caregivers of impaired elderly relatives. Elderly women are less likely to receive care from their spouses than from other female relatives, not only because they often outlive their husbands, but also because gender roles throughout the life course encourage nurturing behavior of women more so than of men.
In an exploratory study of seventy-five caregivers in Shiga prefecture, eighty-six percent of the caregivers were women (Harris and Long 1993: 97-118). These caregivers receive help from husbands (primarily for heavy work such as airing bedding, bathing elderly family members, and taking them to the doctor), children, siblings, and others, but frequently this assistance is provided only when primary caretakers must go out or are involved in the time-consuming task of feeding the elderly. But as a woman respondent commented on the questionnaire: In the Indonesian context, the division of housework is also influenced by the gender ideology factor, which might have been linked to the ways in which the concept of 'caring', especially for the elderly, is used and carried out in Indonesian society. However, before we discuss how caring for the elderly is carried out in Indonesia, one has to notice that Indonesian society has diversified types of family in terms of caring for the elderly. Some are nuclear units; others are three generational cohabitations (parents living with their children and grandchildren), mostly on a patrilocal principle. Especially among young generations, many nuclear households would balance their social norm to care for elderly parents and their desire for autonomy and privacy by making other special arrangements, including living in separate houses in the same housing complex or by living apart yet making frequent visits and hiring "non-family helpers" for their aging parents. Only a small number of households have their parents placed in nursing homes or assisted-care facilities. For three generational cohabitations, the most common pattern is that the married eldest child (whether it is the son or daughter) is obligated to reside with and to care for his/her aging parents, as placing parents in a nursing home is stigmatized and deemed immoral and irresponsible. However, the actual work of caring for aging parents is performed mostly by female children (whether she is the daughter or daughter-in-law). The son, who is the major breadwinner, mainly provides economic support for his parents while his wife offers care and service on a daily basis.
On another case, if the parents are living separately from the children, or when the daughter or daughter-inlaw has a job, the children will hire a non-family "helper" to take care of the increasingly frail parents who need daily assistance and personal care. In this case, the daughter or daughter-in-law then transfers most of her previous duties, including preparing meals, bathing, and changing diapers, to a helper. Some high-class people may even hire the so-called "elderly helpers", but their wage is usually very high (Rp1.000.000-1.500.000/month) by Indonesian standard.
However, in contrast with many societies' traditional ideal of caregiving as related to the hierarchical concept of "serving", caregiving in Indonesian society is associated with the more egalitarian notion of "caring". Therefore, unlike many societies that hold the traditional stereotype that a caregiver is a female relative in a subordinate position (i.e., a wife serving her husband or a daughter-in-law serving her parents-in-law), Indonesian society is less likely to have the so-called traditional stereotype of a caregiver. The act of caregiving is also not as strongly tied to the ideal of womanhood, and the failure to fulfill these gendered responsibilities does not always incur social stigma. Therefore, the nature of "caring" for the elderly could be done by any family member, and in some cases is supported by the nonfamily "helper" as well. Since a large number of Indonesian respondents (53.9%) have housekeepers to assist with daily house labor, including caring for the elderly, individual wives are able to redistribute the housework to the hands of other family member (s) or domestic helper (s). Such assistance makes it possible to carry out elderly care on a daily basis, although it should be noted that the division remains unequal. Women still play a larger role in taking responsibilities for the day-to-day actual care compared to men.
Given these two circumstances from the Japanese and Indonesian respondents, there was no evidence to support the hypothesis that the higher the wives' proportional contribution to overall income, the more husbands engage in domestic labor. The Japanese cultural ideal of female nurturing and the absence of a traditional stereotype of caregiving for the elderly in Indonesia may have precisely inverse effects on the division of household labor in terms of elderly care. Furthermore, the presence of a support system provided by "helpers" for the elderly could potentially explain why higher incomes on the wives' part do not result in a more equal division of elderly care between husbands and wives. In accordance with the negative effects of relative income on elderly care, relative income also has a negative effect on household chores for the Indonesian respondents. Individual wives with higher earnings do not share household chores more equally with their husbands. One of the factors that seem to bear upon this exclusion is the presence of housekeeper (s) living in the same household. Given the presence of housekeeper (s) to assist in daily household tasks, individual wives are able to redistribute the housework to the hands of other domestic helper (s). However, as will be explained in the next section, rather than relative resources, the Indonesian respondents' situation of division of household chores is largely influenced by the gender ideology of couples.
Time Availability. The number of hours worked has a negative effect on elderly care for the Japanese respondents. Quite on the contrary, the number of hours worked has a positive effect on elderly care for the Indonesian respondents. In the Japanese respondents' context, the division of labor in terms of elderly care remains unequal no matter how many hours the wife works outside the home. As stated above, while it does appear that husbands participate in elderly care when their wives become employed, the portion of the husbands' participation is relatively small. Indeed, a husband's relative share of elderly care may increase when his wife works, but there is a strong tendency among the female Japanese respondents to remain the primary caregivers of impaired elderly relatives. This study argues that one of the factors that seem to bear upon this situation is rooted in the cultural ideal of Japanese women as nurturers. In other words, gender ideology persists in Japanese society The working women surveyed in this study were all primarily responsible for elderly care. In response to the survey question "Who performs most of the care for aging parents?", 93.8% of them claimed that as women, they were solely responsible. When asked if the amount of time spent on elderly care had decreased since they began paid work, slightly more than half (51.9%) responded that the amount of time spent on caring for the aging parents had not decreased.
Based on commentaries cited from the questionnaires, this study found that the working women were faced with "the double burden" of elderly care, in that they were the primary caregivers of the aging parents. One middle-aged woman wrote, "The extent of my husband's participation in elderly care was far from an equal share, although my working hours in the paid job are the same amount as his…." Therefore, it could be argued that while previous studies have shown that a wife's work hours decrease the time she spends in household tasks (Coverman 1989), the Japanese female respondents' employment status does not significantly influence the gender division of labor in terms of elderly care. This holds true regardless of the wife's type of job, as career women do not differ significantly from other full-time employed wives in the amount of hours dedicated to elderly care. In other words, the number of hours the wives worked has virtually no effect on their husbands' participation in elderly care, contrary to what the time-availability hypothesis would predict.
In stark contrast, 95% of the women among the Indonesian respondents said that their work hours decrease the time they spend in elderly care. However, this is only manifested in families with no housekeeper (s) or domestic helper (s). As mentioned above, since a large number of Indonesian respondents (53.9%) have housekeeper (s) to assist with daily house labor, including caring for the elderly, individual wives are able to redistribute the housework to the hands of their housekeeper (s). Their assistance makes elderly care possible to be carried out on a daily basis, regardless of the amount of their work hours. Therefore, it still could be concluded that the division of elderly care remains unequal. Wives' number of work hours has virtually no effect on husbands' participation in elderly care, and women still play a bigger role compared to men in taking the responsibility of the day-to-day actual elderly care.
Gender Ideology. As can be seen in the findings, the gender ideology hypothesis receives the greatest support among the Japanese and Indonesian respondents. While for the Japanese respondents gender ideology is significant on all three dimensions of household labor, it is significant for the Indonesian respondents on only two of three household tasks: household chores and childcare. It appears that men who have more egalitarian attitudes are more engaged in a variety of household tasks. In stark contrast, men who have more traditional attitudes are less likely to be involved in household labor. For the young Japanese respondents in terms of household chores and child care, nearly 85% of them did not agree with the statement that "A man's job is to earn money; a woman's job is to look after the home and family". A similar proportion of respondents believed that a father should actively participate in raising his children. This suggests support for the idea that women are not necessarily the sole caretakers of the home and family, but that men should participate in nurturing and disciplining children as well. Some husbands in the survey, particularly those who belong to "dual career families", share housework with their wives and even exceed the wives' share in cooking, child caretaking, or home management in general. An exceptional case is shown by a young couple whose husband not only works as a government official, but also shares more housework than she expected---preparing breakfast, giving the infant child a bath, cleaning the house, taking the trash out for collection, grocery shopping, dishwashing, and so on.
However, while there is a change in the young Japanese respondents' attitudes that would refute the gendered division of labor, the findings showed that the actual work of men and women in household chores and childcare are far from equal. Men performed household tasks and childcare only 10% of the total number of time. Given the small proportion of husbands' share in household labor and childcare, this study argues that there is a gap between the changes in gender role attitudes among young couples and their actual life, or between an ideal and a lived experience. There exists a contradiction, in that the home remains a place of inequality, in spite of the intensification of the value of gender equality and the rise of egalitarian gender-role attitudes that were observed in the study. Regardless of their work status, both women and men among the Japanese respondents indicated that the wife does nearly all of the housework and child rearing, as noted by some women respondents as follows: Yet, a small proportion of these women responded that certain tasks are primarily done by others, namely extended family members. When women are assisted in balancing domestic and paid work, the assistance is usually from other female family members such as daughters, mothers or mothers-in-law. Or in other cases, working mothers could put their young children in the care of childcare facilities.
Nevertheless, due to the widespread unavailability of childcare facilities or economic reasons for employing outside housekeeping assistance, wives continue to do most of the housework, holding the major responsibility of childrearing and seemingly remaining to provide primary care for the elderly. In sum, despite the variation found in the reality of day-to-day lives of contemporary Japan, the ideal of egalitarian gender-role attitudes may not have been achieved yet.
Also, it may be tempting to argue that men spend little time on housework and related activities because their paid work hours are longer than women's, and they are presumed to commute further than women. However, this study found that the average amount of working hours of women with full-time employee status only differs from those of men by 90 minutes per week. Given the extra-work activities employers expect of Japanese men, it is not surprising that working women arrive home before their partners. Several factors seem to influence the ways in which the ideal is played out in daily housework, childrearing, and elderly care as will be explained bellow.
Control Variables. Consistent with past studies, age is evidenced as having a positive effect on an egalitarian division of domestic labor for both groups of respondents in the study. First, age may relate to gender-role ideology; this study has found younger men to be more egalitarian. Second, older men generally have higher earnings and occupational prestige. For both reasons, the older the man, the less likely he is to engage in domestic labor. Indeed, this study has found the relationship between age and household labor among men. By the same token, while age may be important for women's genderrole ideology, it is less likely to influence her occupation and earnings, given labor market discrimination. If we find that only a man's age is significant, it would again indicate the greater importance of the characteristics of men in the division of domestic labor, a finding which would point to the influence of patriarchy.
The Effects of Gender Empowerment Measure in the Division of Household Labor in Japan and Indonesia. The second aim of the study was to examine whether macro-level economic and political gender inequality affect the similarities and differences between the two countries. It was hypothesized that the effect of the three individual-level factors would be stronger for wives in more gender-egalitarian countries, since in these countries, the effects of individual-level factors are less likely to be "discounted" by macro-level economic and political power differences (Blumberg 1984). The results, suprisingly, do not sufficiently support this hypothesis of the study. As shown in the Table 2, Indonesia has a relatively modest division of housework in spite of their lower GEM score. On the contrary, Japan, regardless of their higher GEM score, does not suggest a more egalitarian division of household labor.
The results of t-tests confirmed these findings. There are no differences between the average division of labor in Japan and Indonesia in the three dimensions of household labor, namely household chores (p= .075), child care (p= .247) and elderly care (p= .130). This means that GEM does not affect the division of household labor both in Japan and in Indonesia.
According to Blumberg (1984), individual-level factors will have a stronger effect on the division of household labor for women who live in countries that adopt more gender equality, while for women in countries where gender equality is more severe, individual factors will have a weaker effect on the division of household labor. However, the current study's findings show that both Japan and Indonesia have roughly similar numbers portraying the effect of individual factors on division of household labor. Based on the GEM rank report, Japan holds a much higher rank than Indonesia; it follows that the higher the rank, the higher its level of gender equality. Nevertheless, this study found that GEM rank did not have significant role in addressing the strength of the effect of individual factors on the division of household labor.
Wives in Indonesia --a less gender-egalitarian country, are more likely to be burdened with a traditional division of household labor yet benefit more from their individuallevel assets in the negotiation over housework. While in Japan, which should be regarded as a more genderegalitarian country judging from its GEM score, wives benefit less from their individual-level resources. This suggests that macro-level gender inquality in economic and political power may not be enough to determine the effect of individual-level factors on negotiation over housework in family units. Therefore, in addition to individuallevel factors and country-level factors that emphasize political economic aspects, this study argues that there seems to be other factors that play a significant role in determining division of housework in both countries.
The question is, why do GEM scores not seem to affect the division of household labor between couples in Japanese and Indonesian household units? What, then, are the other factors that play more signifant roles in determing the division of housework?
This study argues that beyond macro-level gender inequality in economic and political power, the analysis should focus on other social and cultural factors as well. Although macro-level gender inequality in economic and political power helps to explain couples' division of housework in both countries to some extent, the cultural context and social system are also important influences on couples' housework division. In other words, there is a mutual interaction between gender ideology and social structures operating in the division of household labor in Japan and Indonesia.
The Japanese Context.

The Cultural Aspects: The cultural ideal of women as primary caregivers and state gender ideology.
This study argues that in addition to the earlier explanation of the cultural ideal of women as primary caregivers, state gender ideology plays a significant role in producing a constraining social system that disadvantages Japanese women. First and foremost, it is not the individual gender ideology that matters much in the division of labor in the Japanese context, but instead it is the "state gender ideology" that is formed through welfare and social policies (social system) that institutionalize the gendered division of labor by privileging spouses/families where women remain dependent on male income. In the early Meiji period, the family (i.e.), functioned as an administrative organ of the government. Under the Meiji Civil Code, an individual's life was totally ruled by the family system, and the head of family had the authority to govern the behavior of all family members. State authority supported this patriarchal structure, as the head of family was obligated to notify authorities of changes in the status of all family members (Yoshizumi 1995:187-188). In theory and policy, the family functioned symbolically, as the larger nation-family either the Emperor occupying the role of patriarchal head. The family was chosen as a means of social control, and in order to achieve this goal, the government needed to control and regulate relationships within the family. To this end, "Motherhood was emphasized within the context of a social hierarchy structured to support the state" (Ohinata 1995:200). Ohinata argues that the emphasis on motherly love and raising children properly focused on the mother's role as "breast feeder" and coincided with the development of capitalism and moves by working-class families into industrial jobs (1995:2000).
As argued by many scholars on Japan (Broadbent 1992, Imamura 1996, Suzuki 2006), Japanese policies fail to challenge existing patterns of the division of labor in the household. Women in Japan, while encouraged to become a part of the paid workforce, are not encouraged to remain in the workforce during the childrearing years when family demands keep them in an exclusively domestic role. In emphasizing industrial development, governments in Japan have created a welfare state with "the family" as its foundation. The family has been the provider of this successful development, and as corporate Japan demands commitment from males, the role of reproducing labor through care and services for husbands, children and dependants falls to women.
Legislation designed to address equity issues and social welfare policies, such as tax and pension, institutionalize and systematize the gendered division of labor. Women in Japan have been constituted as wives and mothers responsible for domestic work, care of children and aged relatives, as well as a range of other responsibilities expected of a Japanese housewife and mother. Welfare policies such as insurance, health, pension, and dependant allowances, as well as the income tax threshold, are premised on the notion that women are dependent on males.
The legislations initially proposed to improve conditions for women workers such as the EEOL (Equal Employment of Law), the Parental Leave Law (Child Care Leave Law), and the Part-time Workers' Law have not addressed fundamental issues such as discriminatory employment conditions and wages as well as parenting responsibility.
Moreover, as women entered the workforce since the mid of 1970s, whereas greater numbers of married women or women with dependants in full-time paid work have contributed to pressure governments into providing more extensive welfare services and have jeopardized employers' ability to continue guaranteeing full-time male workers' employment conditions, the increasing numbers of working women complements both governments and employers goals.
As will be explained bellow, some social mores and obstacles for women who continue with employment after marriage, including the short hours of childcare facilities, imply that children are cared for at home. This reduces the incentive for governments to extend childcare hours. In addition, the scarcity of elderly care services and facilities means women are forced to quit their jobs to care for elderly dependants when the time arises; again, there is no incentive for governments to address issues relating to services for elderly care. The analysis of the construction of women within the Japanese state is clearly applicable to explain the contradictive findings of this study regarding the low correlation between GEM rank and the average division of household labor in Japan. This study suggests that the Japanese government has glorified mothers without glorifying women, providing an ideological instrument for subordinating all women. The underlying assumptions that determine the construction of the role of women and its impact on the division of household labor, as well as welfare policies encouraging women to enter or exit the paid workforce have been modified depending on specific historical circumstances. As working women, such as the ones the surveyed in this study, are central to the restructuring of the domestic Japanese economy, the constitution of their role has remained focused on the function they fulfill as caregivers of husbands, children, and aging parents. To conclude, it is becoming evident that in the Japanese context, in spite of the current trend in women's employment in the workforce and the changes in gender role attitudes among young Japanese, role-sharing in terms of household labor is still far from gender-equal. The findings show that division of housework is largely influenced by mutual interactions of cultural and social factors, namely the state gender ideology and the constraining social structures.
The Constraining Social Structure Based on the Gendered Division of Labor: Workplace organization, the care insurance system, child-care leave. It is argued that the state gender ideology has powerfully formed a division of labor that put men in an advantegous position over women. This division of labor along gender lines tenaciously exists not only in the individual psyche, but also in families, enterprises, and society; moreover, there is a continuation of practices in line with it. The gendered division of labor has been reproduced and maintained amidst these practices. Even now, in the twenty-first century, there continues a vicious cycle in practices connected with the gendered labor division in the home, and the gendered division of labor in the home has transformed the gendered labor division in the business world into an impregnable system (Suzuki 2006a).
Though there are more women continuing paid work even after giving birth, there are still many women who have to quit their jobs in spite of the desire to stay at work and to be financially independent, because conditions are not in place to enable them to keep working upon marriage or childbirth, let alone if there are elderly family members they have to take care of. This fact, again, illustrates that there are still insufficient arrangements in the working environment aimed at gender equality in both the family and at work.
While there has been a certain amount of progress in legislative improvements that would make it easier for married women to continue working, (e.g., the revised equal employment opportunity law prohibiting gender discrimination and direct discrimination against men and women coming into effect since April 2007), the social system, workplace organization, workplace climate, and workplace support and services do not constitute a framework commensurate with these legal developments.
There are contradictions between the laws and various systems. Examples of the latter that have been set up include the care insurance system, under which the elderly will not receive sufficient care unless families also look after them because the system alone is inadequate, and the systems of taxation, pensions, and wages that are more advantageous to homemaker wives than to working wives.
Because these systems are modeled upon a 'standard household' (husband, homemaker wife, and two children) where the gendered division of labor in the home is established, it is of greater advantage to married women to work in a financially independent manner. For that reason, from before World War II through to the present day, the work of married women in Japan has often been for the sake of supplementing household finances. In other words, it has been non-regular employment for such purposes as children's educational expenses or housing loan repayments, and moreover, women have stayed in jobs for a shorter period than men (Suzuki 2007:6). This has made women come to fulfill the function of an economic control-valve. In addition, women who are able to work long hours similar to men can become regular employees, but women who cannot work such hours are in a position where they have no alternative but to take jobs as non-regular employees.
The constraining social structure does not place the burden on the woman's side alone. In recent years, due to the impact of protracted economic stagnation from the 1990s, there has been an increase in corporate structuring, which results in increasing overtime work as the number of regular employees fell. Moreover, there still remains a deep-seated value defining a loyal company employee as one who puts in long hours at work. Though there has been an increase in men who do wish to balance their work and home lives (Sorifu 2000), in reality, however, most males likely to be socially dominant over females are also constricted by job-centered values such as those epitomized by a large amount of overtime and long hours of work. It is hard to say that these men have autonomously established an affluent life career while prioritizing the private life (Suzuki 2006).
Another important system to discuss is the child-care leave. There is a fact that in stark contrast with other developed countries, in Japan it is still difficult for men and women to take child-care leave.

The Enabling Social Structure with Support System from Extended Family (s) and Domestic Helper (s).
Furthermore, for Indonesian respondents, the flexibility of their social structure may produce a support system for childcare tasks. Similar to Japan, parents are prominent figures in childcare tasks, but there are other figures that can help the parents ensure the childcare tasks are run perfectly. These figures mainly include extended family members and household helpers hired to help with household chores and childcare tasks. It is common, especially among urban people, to entrust part of their household labor to other people. As the economical demands get higher, it is also not unusual for both men and women to work to support their family. Household labor holds as important a role as workplace task, and for the people who work outside the home, to be able to run one's work and household labor perfectly at the same time would require good time management skills.
In consequence, some Indonesian respondents hire household helpers to do parts of the household chores and to baby sit their children. Otherwise, during working hours, some parents leave their children in the supervision of their family members such as their parents (i.e., grandparents of the children), siblings, or even distant relatives such as cousins, uncle, and aunts. Unlike Japan, there are norms in Indonesia that allow/support extended family members to help their relatives in caring for their children. With the existence of these helpers (i.e., extended family and household helpers), parents can allot a proportional amount of time for both work and household labor, which influences gender equality such that all they need to do is divide a small set of chores through compromise because some of the household labor are handled by other people.
Results of the compromises the husband and wife make will determine the level of equality between them.
Owing it to the structural reasons that exist within the Japanese and Indonesian contexts, it can be implied that GEM Rank may be implacable for certain societies. In particular, for both Japanese and Indonesian people, there are cultural and structural aspects that may reduce the numbers that appear on the GEM Rank.

Conclusions
This study is among the first to examine the effect of individual-level characteristics and country-level gender inequality on the division of housework in Japan and Indonesia. The effects of relative income, time availability, and gender ideology on the division of housework in both countries were observed to be very dynamic, even contradictive, and among which gender ideology was found to be of particular importance. Perhaps the most enlightening finding of all is that despite their lower GEM rank compared to Japan, Indonesian respondents have a relatively egalitarian division of labor in their households. It was hypothesized that the effect of the three individual-level factors would be stronger for wives in more gender-egalitarian countries, considering that in such countries the effects of individual-level factors are less likely to be discounted by macro-level economic and political power differences (Blumberg 1984).
Yet the results do not entirely support this hypothesis. The effects of relative income, time availability, and gender ideology were not statistically significant. As stated above, wives in Indonesia-a less genderegalitarian country-who are supposedly more likely to be burdened with a traditional division of household labor, benefit more from their individual-level assets in the negotiation over housework. While women in Japan, which according to its GEM score should be regarded as a more gender egalitarian country, benefit less from their individual-level resources. This suggests that macro-level gender inequality in economic and political power may not be enough to determine the effects of individual-level factors on negotiation over housework in family units. Therefore, in addition to individuallevel factors and country-level factors that emphasize political economic aspects, this study argues that there seems to be other factors that play a significant role in determining division of housework in both countries.
The alternative factor lies in the cultural and sociostructural contexts of both countries. These include the cultural ideal of women as primary caregivers, the state gender ideology and the constraining social structure based on the gendered division of labor in the Japanese context, as well as the diversified gender ideology and the enabling social structure that gives room for support system from extended family member(s) and domestic helper (s.) in the Indonesian context. Such factors are also important influences on couples' housework division. These findings suggest that incorporating theories of micro-macro interaction might provide some insight into the dynamics of gender relationships and help refine our investigative scope regarding the division of housework. Such an approach bridges the gap between micro and macro emphases in research on social inequality. Furthermore, these findings provide a new insight that GEM, which emphasizes political economy aspects of the country, may not be sufficient to capture gender disparities without considering other socio-cultural factors in the complexity of day-to-day actual division of housework.