Is Wife’s Marital Satisfaction Associated with Husband’s Dominance in Family Affairs? Empirical Evidence from China Is Wife’s Marital Satisfaction Associated with Husband’s Dominance in Family Affairs? Empirical Evidence from China Zhongwu Li Abstract: Employing data from the China Family Panel Studies (CFPS), the paper empirically examines the relation between husband’s dominance in family affairs and wife’s marital satisfaction. While applying the ordinal probit model and ordinary least squares (OLS) method, the paper fi nds that wife’s reported marital satisfaction is positively associated with her husband’s dominant role in family affairs. This conclusion remains valid after using an instrumental variable to deal with endogeneity and performing some robustness tests. Some heterogeneities exist: the association is particularly prominent among those women who have traditional gender norms and are living in rural areas. These women tend to embrace the traditional gender ideology which stipulates that men are the masters of the family. Keywords: Dominance in family affairs · Marital satisfaction · Socio-economic reforms · Traditional gender norms · China Family Panel Studies 1 Introduction With socio-economic reforms in 1978, e.g., the de-collectivisation of agriculture, the opening up of the country to foreign investment, and the nine-year compulsory education, Chinese women have had more opportunities to work in the labour markets for the past 40 years. Many women are now competing with their male counterparts for those jobs traditionally assigned to men. Additionally, women participate in social and political activities proactively, making their voices heard by others. All of these developments enhance the agency and capability of Chinese women greatly (MacPhail/Dong 2007). Under this social background, many women now tend to rid their well-being arising from being dependent on their husbands in order to fi nd a new form of well-being arising from being more independent (Yu et al. 2019). In the family, women no longer simply embrace husbands’ decisions, but actively participate in the decision-making process to improve their own well-being. In the eyes of women, they are able to do whatever men do, thus, it is unacceptable Comparative Population Studies Vol. 48 (2023): 47-72 (Date of release: 14.02.2023) Federal Institute for Population Research 2023 URL: www.comparativepopulationstudies.de DOI: https://doi.org/10.12765/CPoS-2023-03 URN: urn:nbn:de:bib-cpos-2023-03en9 • Zhongwu Li48 that men’s ideas dominate in family affairs. So, given the above, wife’s marital satisfaction would be negatively affected in case of husband’s sole dominance in the family. In contrast to this, traditional social norms, especially gender norms remain infl uential in Chinese society. Such norms have been there for thousands of years, which stipulate the appropriate roles and behaviours of both men and women. In the domain of family, men are traditionally regarded as the breadwinners who take the responsibility of feeding family members, while women are affi liated to their husbands for assistance (Gao 2003; Xie 1994). Correspondingly, men are granted the power over major family affairs, while women are required to accept their husbands’ decisions (Wang et al. 2019). The gender-based role division infl uences women’s role orientation greatly, so that their expectations on husbands and themselves are increasingly in line with traditional gender norms (Park/Banchefsky 2019; Yang/ Gao 2021). In this situation, the increasing match between women’s expectations and traditional gender norms contributes to their marital satisfaction and stability. So, given traditional gender norms, women tend to fi nd their well-being in being dependent on their husbands contrary to the other form of well-being arising for women from greater independence from their husbands. Consequently, wife’s marital satisfaction would increase in case of husband’s dominance in the family. Both socio-economic reforms and traditional social norms jointly determine the relation between wife’s marital satisfaction and her husband’s dominance. On one side, under the infl uence of socio-economic reforms, many women not only question the idea of husband’s dominance in the family, but they also fi ght for their own participation in family affairs. So, in this constellation, husband’s dominance would negatively affect wife’s marital satisfaction. On the other side, under the infl uence of traditional social norms, many women not only accept the tradition of husband’s dominance in family affairs, but they also alienate those who deviate from such men-centered patriarchal traditions. So given this traditional constellation, husband’s dominance would instead enhance wife’s marital satisfaction. However, which factor exerts more infl uences in contemporary Chinese society? So far, there has been no relevant studies on this specifi c topic. In previous research, marital satisfaction and its correlates have been investigated almost exclusively in Western countries (Bradbury et al. 2000). Moreover, in terms of the relation between bargaining power and marital quality, several studies focused on Western countries show some mixed results: some report that shared power in marriage is linked to the partners’ higher marital quality and lower attachment insecurity over time (Leonhardt et al. 2020). But others report that bargaining power is not associated with the quality of marriage at all (Sarantakos 2000). So, it needs some new evidence on the topic for further clarifi cation, especially, in terms of non-Western countries. The contemporary China is a good case for the study given two competing forces are simultaneously infl uencing both power relations among couples and marital satisfaction. The remainder is arranged as follows. The second section reviews the previous literature and proposes the research hypotheses. Data, variables, and descriptive statistics are displayed in the third section. In the fourth section econometric Is Wife’s Marital Satisfaction Associated with Husband’s Dominance in Family Affairs? • 49 analysis is carried out and interpreted. The last section discusses the fi ndings and concludes the study. 2 Literature review and research hypothesis Our study draws on the literature on intra-household bargaining power: the determinants and the outcomes of intra-household bargaining power. Two research approaches are proposed to explain what determines one’s intra-household bargaining power. The fi rst is resource determinism, emphasising that the relative advantages in terms of education, occupation and income affects the position of family members (Agarwal 1997). Whoever has relative advantages in such resources as education, income and assets, holds stronger intra-household bargaining power, because the resources offer him/her more outside options in case of marriage dissolution. The second approach focusses on cultural norms emphasising that cultural background and institutional factors are important determinants of family decision-making outcomes (Mabsout/van Staveren 2010; van Staveren/Ode bode 2007). Such cultural norms can affect one’s exit options, one’s bargaining agency, e.g., accepting male authority when women have formally equal rights (Nikièma et al. 2008), one’s preferences, e.g., adapting to what is deemed proper for women (Sen 1990), and one’s roles in the family, e.g., limiting what can and cannot be bargained over (Cuesta 2006). Consequently, cultural norms would infl uence one’s intra-household bargaining power, and even they will override the relative resources in determining one’s bargaining power in the family (e.g., Goetz/Gupta 1996; van Staveren/Ode bode 2007). Meanwhile, previous research pointed out that bargaining power can directly or indirectly affect key personal outcomes, such as children’s health, education, family production decisions, marital violence, and women’s well-being (Deere/Doss 2006; Luke/Munshi 2011; Panda/Agarwal 2005). Further, our study also draws on the literature of marital satisfaction. It is found that many factors are related to marital satisfaction across different societies. The following part will review the central variables that are predictors of marital satisfaction. Gender: it is reported that women are less satisfi ed with their marriages compared to men (Rostami et al. 2014). Age: there is no clear association between age and marital satisfaction in the literature (Schmitt et al. 2007). Religiosity: a positive association between religiosity and marital satisfaction is found across different religious groups (Fincham et al. 2011). Duration of marriage: the literature shows either a negative or U-shaped relationship between marriage length and marital satisfaction (Karney/Bradbury 1995; Kurdek 1999). Number of children: some contradictory fi ndings are there in terms of the relation between the number of children and marital satisfaction (Onyishi et al. 2012; Twenge et al. 2003). Education: level of education is found to be positively related to marital satisfaction (Rouhbakhsh et al. 2019; Stanley et al. 2006). Economic status: low income or material hardship is associated with lower levels of marital satisfaction and stability (Kurdek 1999). Personality traits: it is found that neuroticism trait has a negative relationship with marital satisfaction, while conscientiousness trait has • Zhongwu Li50 a positive relationship with marital satisfaction (Sayehmiri et al. 2020). Moreover, the existing literature reveals that marital relationships are heavily infl uenced by culturally shaped norms, customs and expectations (Berscheid 1995). So, studying the correlates and even determinants of marital satisfaction in some other cultures apart from Western ones, would deliver new insights. Based on relevant research approaches and social development in contemporary China, this paper proposes three research hypotheses for empirical investigation. According to social role theory (Akerlof/Kranton 2000; Eagly/Wood 2016), there are generally held expectations for ideal male and ideal female roles. Such expectations, or social roles, are shaped by the values prevalent in a given society, which in turn are shaped by various social constraints. For thousands of years, Chinese patriarchal society rules that men are the masters of the family, while women are subordinate to their counterpart men (Luo/Chui 2018). All major family affairs are controlled by the husbands, and women are required to abide by their husbands’ will. As an exchange, men bear the responsibility of feeding family members and winning reputation for the family. In the male-centered social system, women’s well-being is fully dependent on their husbands’ social and economic achievements (Wang et al. 2019). As a survival strategy, women prefer to marry such men with higher socio-economic status and income (Chen 2018). Infl uenced by surrounding environments and family members, women also subconsciously believe in husband’s dominant role in family affairs. In marriage, such gender-based role division helps to form a harmonious relationship which improves women’s marital satisfaction. With deviation from the social role arrangement (e.g., men’s dominance in the family), both husbands’ and wives’ marital quality and marital stability might be hurt to some extent (Park/Banchefsky 2019). In contemporary China, women have made great achievements in various social and economic fi elds along with technological progress. Many new ideas have permeated through the whole country, so, many women are open-minded to free themselves from old ideas and concepts (Qi 2018; Zheng et al. 2011). They not only challenge the men-centered patriarchal system, but also fi ght for more bargaining power for their own well-being in the family. Overall, the socio-economic transition from a closed socialist economy to an open market economy lays the foundations for women to free their well-being from dependence on their husbands, psychologically and economically (Yu et al. 2019). However, a series of research shows that unlike the trend of gender equality in many Western societies, Chinese families are evolving slowly to give women more space, but the traditional gender norms remain dominant in determining the roles of couples (Ji/Yeung 2014; Yu/Xie 2021). This is deeply refl ected in strong expectations regarding women’s marriage and child birth at a specifi c age, otherwise, external pressures are seriously imposed on those unmarried and childless women. Once a single woman is approaching her 30s, her parents and relatives would worry about her marriage and family life. Moreover, according to Wang et al. (2019), many women heavily rely on their husbands’ political identity and social status for better well-being. Also, most women tend to marry men with higher socio-economic status and income, and the phenomenon of hypergamy prevails across different regions (Chen 2018). Relative Is Wife’s Marital Satisfaction Associated with Husband’s Dominance in Family Affairs? • 51 to women, men’s employment status plays a more important role in the well-being of the couple (Qian/Qian 2015; Zhang/Tsang 2013). Therefore, so far, traditional gender norms remain dominant in infl uencing personal social behaviours and role expectations. Following the traditional gender division of labour, we might expect that there is more harmony and higher satisfaction in marriages. Consequently, husband’s dominance in family affairs would improve his wife’s marital satisfaction, given the consistency of social role and behaviours. Regarding this, we propose the fi rst hypothesis. H1: Wife’s marital satisfaction is positively associated with husband’s dominance in family affairs. Considering that there are huge differences across people and living regions, we believe some heterogeneities exist in the relation between wife’s marital satisfaction and husband’s dominance. Among various heterogeneities, two of them deserve our detailed discussions, which can further strengthen the robustness of the conclusion. The fi rst one is personal gender ideology, as it directly affects wife’s attitudes towards her husband’s dominance. For a woman with traditional gender ideology, such as men being the masters of the family, her behaviours will bear the characteristics and imprints of such ideology (Egan/Perry 2001; Nielson et al. 2020). The woman would subconsciously believe that it is her husband who should manage major family affairs. So, she tends to accept or even embrace the role arrangement of husband’s dominance in the family, despite possible welfare loss (Endendijk et al. 2018; Rittenour et al. 2014). Moreover, she would oppose against any deviation from such arrangement by alienating even condemning such deviating behaviours. Overall, the woman rarely questions the tradition of men’s dominance, and more, plays a supportive role in the family (Turner/Salemink 2015; Valutanu 2012). Therefore, her marital satisfaction will be improved if the role arrangement matches with her expectation of men’s dominant role in family affairs. Based on this, we propose the second hypothesis. H2: The positive association between marital satisfaction and husbands’ dominant roles in family affairs is particularly evident among those women with traditional gender norms. The second one is living regions (urban vs rural) . Women from rural and urban regions are fundamentally different in their decision making regarding family affairs (Cheng 2019). It is possible that the positive impact of husband’s dominance only exists in rural families, but not in urban families. The rural-urban gap is large, and continues to amplify in educational, social, economic, and ideological aspects (Meng/Zhao 2019). In rural areas, women have been disadvantaged for a long period of time, given limited education and economic opportunities (Xiao/Asadullah 2020). The economic opportunities being more concentrated in urban areas, rural villagers have to depend on limited land for livelihood. Given the physical advantages in farming, rural men naturally become heads of the family to control major family affairs. Additionally, compared to urban areas, advanced ideas and thoughts are diffi cult to penetrate in the isolated rural areas, so, traditional gender norms remain infl uential there (Song et al. 2021). Consequently, rural women personally believe • Zhongwu Li52 that women’s subordinate status is normal, and compliance with this is essential to marital quality and stability. In this case, the marital satisfaction of rural women will be largely associated with their husbands’ dominant role in family affairs. Based on this, we propose the third hypothesis. H3: The positive association between marital satisfaction and husbands’ dominant roles in family affairs is particularly evident among those women living in rural areas. 3 Data, variables and descriptive statistics 3.1 Data source The paper uses the nationally representative China Family Panel Studies (CFPS) for econometric analysis. The CFPS is a longitudinal survey of Chinese communities, families and individuals updated every two years. Launched in 2010, it aims at collecting the latest information on China’s socio-economic, demographic, educational, and health changes (Xie/Lu 2015). The CFPS is drawn through a stratifi ed three-stage sampling method, covering 14,798 households in 162 counties and 635 villages in 25 provinces in China. The CFPS respondents are tracked through 2-year follow-up surveys, with an average response rate of 79 percent (Xie/Hu 2014). So far, the latest CFPS survey has been updated in 2018, with a collection of some newly added variables. Nevertheless, only the 2014 CFPS survey contains the variables needed for the paper, such as marital satisfaction, dominant role in family affairs, gender role beliefs. Even though the 2014 CFPS survey has no advantages over later rounds of surveys in terms of timeliness, the paper identifi es further research needs. And thus, the fi ndings would lay the foundations for future search in related fi elds. Given some control variables are missing in the 2014 CFPS, we also use variables in the 2012 CFPS adult and family relationship datasets in econometric estimations. Meanwhile, to obtain a causal effect of husband’s dominance, the paper fi nds an instrumental variable in the 2010 CFPS family relationship dataset. Considering that husband’s dominance in family affairs is built based on marriage, we constrain the dataset to individuals in marriage. There remain 9,186 observations in the survey sample. After deleting the variables with missing values, non-response items and outliers, we fi nally get 7,240 observations for econometric analysis. According to our balance test, the variables used in the study have similar means and standard deviations between the analytical sample and the original sample.1 1 Results of the balance test are not shown here but it is available on request. Is Wife’s Marital Satisfaction Associated with Husband’s Dominance in Family Affairs? • 53 3.2 Variable selection The dependent variable is wife’s marital satisfaction (Marital_ satisfaction). It is assessed by the question of “to what extent are you satisfi ed with your current marriage?” The respondent can choose between 5 options: very dissatisfi ed, quite dissatisfi ed, neutral, quite satisfi ed, and very satisfi ed, with assigned values of 1, 2, 3, 4 and 5. A higher value indicates the wife has a higher level of marital satisfaction. The major independent variable is husband’s dominance in family affairs (Husband_ power). It is assessed by the question of “who plays a dominant role in fi ve important family matters”: household expenditure allocation (Expenditure), household saving and investing (Save_ investment), house construction and purchase (House_ purchase), children’s education (Child_ edcuation), and high- priced consumer goods purchase (High_ priced_ purchase). The variable is equal to 1 if the husband plays a dominant role in the specifi c family matter, 0 if it is negotiated together, and -1 if the wife plays a dominant role. Each variable has the value of -1, 0 and 1. Then, we sum the scores of the above fi ve sub-items to construct a comprehensive indicator, which refl ects the husband’s dominant power. The variable of Husband_ power ranges from -5 to 5, and a larger value indicates that the husband plays a more dominant role in major family affairs. However, the fi ve items are heterogeneous, which becomes particularly evident when referring to children’s education and household fi nancial affairs. To solve this issue, this paper also examines the fi ve aspects separately in the following econometric analysis. Additionally, this paper controls for a series of potential confounding factors, which are associated with both husband’s dominance and wife’s marital satisfaction. They include individual characteristics of women, i.e., whether she lives in an urban area (Urban, 1 if urban, 0 if rural), years of education (Education), natural logarithm of annual income (Income),2 age (Age), age squared (Age2), type of job (Occupation, 0 if unemployed, 1 in case of an agriculture-related job, 2 if employed in companies, 3 in case of entrepreneurship and self-employment). In contemporary China, entrepreneurship and self-employment is the most valued occupation, and employment in companies is more valued than agriculture-related jobs, and unemployment is the least valued, so, Occupation is an ordinal variable from the perspective of social prestige. Considering that gender ideology may directly affect wife’s attitudes towards husband’s dominance and her perceived marital quality, wife’s gender ideology (Gendernorm) is controlled in the model. It is assessed by the following four questions: “1. Do you agree that men focus on the career and women focus on the family?” “2. Do you agree that marrying well is better for women than fi nancial independence?” “3. Do you agree that women should have children throughout their life?” “4. Do you agree that men should do half of the housework?” The respondent has 5 options, for the fi rst three questions, with 1 representing strongly disagree and 5 representing strongly agree; for the last question, with 1 representing strongly agree and 5 representing strongly disagree. Then, the scores 2 Considering the zero point of annual income, Income = log(1 + annual income). • Zhongwu Li54 for the above four questions are summed to construct a variable refl ecting attitude toward traditional gender norms. A higher value indicates more traditional gender role attitudes. Additionally, non-cognitive ability (personality traits) is found to explain about 50 percent of individual differences in life satisfaction (Diener 1984). So, non-cognitive ability measured by the Big Five personality traits in the CFPS 2012 is further controlled. Among them, Neuroticism is assessed by the following four statements: “I feel depressed”, “I feel down”, “I feel scared”, “I feel sad”; Agreeableness is assessed by the following two statements: “I don’t think people are nice to me”, “I don’t think people like me”; Conscientiousness is assessed by the statement: “I have trouble concentrating when I am doing things.” For each statement, the respondent has 4 options: most of the time, often, sometimes, almost never with assigned values of 1, 2, 3, and 4. Then, we sum the scores of all statements to construct a variable of non-cognitive ability (Noncognitive). It ranges from 7 to 28, and a larger value indicates a stronger level of non-cognitive ability. Meanwhile, a signifi cant spillover effect in individual well-being is found within the family (Wang et al. 2019), so, some family-level factors are also controlled. They include educational difference (Education_diff, the difference between years of education of husband versus wife), income difference (Income_diff, the difference between husband’s annual income and wife’s annual income), non-cognitive ability difference (Noncognitive_diff, the difference between husband’s non-cognitive ability and wife’s non-cognitive ability), husband’s type of job (S_Occupation, same defi nition with Occupation) and the family’s social status in local community (Family _ status, 1-5, a larger value indicates a higher family social status). Considering regional differences, we further add provincial dummies to the model. 3.3 Descriptive Statistics Table 1 lists the descriptive statistics of the main variables. The mean of Marital_ satisfaction is 4.395, indicating that women in the sample are generally satisfi ed with their marriages. The mean of Husband_ power is 1.201, indicating that in most cases, the husband is the decision maker of important family affairs. For the mean value of 1.201 (with SD of 3.689), it is worrying that one could also conclude that in Chinese society, husband’s dominant power is non-existent given the Husband_ power ranges from -5 to 5. Regarding this, we display the distribution of Husband_ power in Table 2. The percentage of husbands dominating in all the fi ve family matters is about 34.65 percent, which is twice larger than that of wives’ dominance (about 16.53 percent). So, it further indicates that the husband plays a dominate role in major family affairs in Chinese society. For other sub-indicators of husband’s dominance, their mean values are greater than zero. It indicates that the husband also dominates in the fi ve specifi c family affairs. However, the mean of Child_ education is the smallest, but that of House_ purchase is the largest, Save_ investment is the second largest. So, the husband has the greatest decision-making power in house purchase, then, saving and capital investment. However, the wife also actively participates in the children’s education despite husband’s dominance. Additionally, 48.5 percent of the women live in urban areas. The average years of education are Is Wife’s Marital Satisfaction Associated with Husband’s Dominance in Family Affairs? • 55 about 6; The average age is about 49 years. In terms of gender difference, it is shown that men have more advantages in education, income, and non-cognitive ability Variable N Mean SD Min p50 Max Marital_ satisfaction 7,240 4.395 0.933 1 5 5 Husband_ power 7,240 1.201 3.689 -5 2 5 Expenditure 7,240 0.229 0.890 -1 1 1 Save_ investment 7,240 0.274 0.876 -1 1 1 House_ purchase 7,240 0.398 0.818 -1 1 1 Child_ education 7,240 0.072 0.912 -1 0 1 High_ priced_ purchase 7,240 0.228 0.883 -1 1 1 Urban 7,240 0.485 0.500 0 0 1 Education 7,240 6.157 4.891 0 6 22 Income 7,240 3.287 4.065 0 0 12.61 Noncognitive 7,240 24.830 3.151 7 26 28 Occupation 7,240 1.143 0.890 0 1 3 Gendernorm 7,240 14.00 2.542 4 14 20 Age 7,240 48.94 13.39 19 49 89 Education_ diff 7,240 1.765 4.303 -16 0 16 Income_ diff 7,240 1.379 4.793 -12.61 0 12.35 Noncognitive_ diff 7,240 0.835 3.341 -18 0 20 S_Occupation 7,240 1.458 0.879 0 1 3 Family _ status 7,240 3.193 0.947 1 3 5 Tab. 1: Summary statistics of key variables Note: N: the number of observations, Mean: mean value, SD: standard deviation, Min: minimum value, p50: median value, Max: maximum value. Source: Own calculations based on CFPS 2014, 2012 Tab. 2: Distribution of Husband_ Power Husband_ power Frequency Percent -5 1,197 16.53 -4 61 0.84 -3 359 4.96 -2 78 1.08 -1 441 6.09 0 834 11.52 1 628 8.67 2 153 2.11 3 799 11.04 4 181 2.50 5 2,509 34.65 Source: Own calculations based on CFPS 2014 • Zhongwu Li56 than their counterpart women. These are roughly in line with the actual situation of Chinese society in 2014,3 thus ensuring the representativeness of the sample, and laying a foundation for empirical analysis. 4 Econometric analysis and interpretation 4.1 Econometric model In order to examine the relation between husband’s dominance in family matters and his wife’s marital satisfaction, we take Marital_ satisfaction as the dependent variable and Husband_ power as the main independent variable to establish an econometric model as follows: Marital_ satisfactioni = β0 + β1 * Husband_ poweri + Covariates *β3 + μi 1, Marital_ satisfactioni ≤ r1 2, r1 < Marital_ satisfactioni ≤ r2 Marital_ satisfactioni = 3, r2 < Marital_ satisfactioni ≤ r3 4, r3 < Marital_ satisfactioni ≤ r4 5, r4 < Marital_ satisfactioni Among them, Marital_ satisfactioni is the latent variable of female marital satisfaction, r1, r2, r3, and r4 are the cut-off points, with r1 < r2 < r3