THE DIVISION OF HOUSEHOLD LABOR AND ITS EFFECT ON WORK-LIFE BALANCE AMONG SPOUSES AND COHABITING PARTNERS: A COMPARISON OF GENDER RELATIONS IN THE UNITED STATES, GERMANY, THE NETHERLANDS, AND PORTUGAL A Thesis Presented to the faculty of the Department of Sociology California State University, Sacramento Submitted in partial satisfaction of the requirement for the degree of MASTER OF ARTS in Sociology by Iris Iristay FALL 2013 © 2013 Iris Iristay Doc ALL RIGHTS RESERVED ii THE DIVISION OF HOUSEHOLD LABOR AND ITS EFFECT ON WORK-LIFE BALANCE AMONG SPOUSES AND COHABITING PARTNERS: A COMPARISON OF GENDER RELATIONS IN THE UNITED STATES, GERMANY, THE NETHERLANDS, AND PORTUGAL A Thesis by Iris Iristay Approved by: , Committee Chair Mridula Udayagiri, Ph.D. , Second Reader Randall MacIntosh, Ph.D. Date iii Student: Iris Iristay I certify that this student has met the requirements for format contained in the University format manual, and that this thesis is suitable for shelving in the library and credit is to be awarded for the thesis. , Graduate Coordinator Amy Qiaoming Liu, Ph.D. Date Department of Sociology iv Abstract of THE DIVISION OF HOUSEHOLD LABOR AND ITS EFFECT ON WORK-LIFE BALANCE AMONG SPOUSES AND COHABITING PARTNERS: A COMPARISON OF GENDER RELATIONS IN THE UNITED STATES, GERMANY, THE NETHERLANDS, AND PORTUGAL by Iris Iristay With growing participation of women in the labor force, new challenges continue to rise for women to meet daily demands of the home and work environment. Social scientists have re-focused their research efforts to identify what factors are critical in helping women balance their work and family life. This study is a comparative analysis of how the division of household labor among spouses and cohabiting partners living in the United States, Germany, the Netherlands, and Portugal is a key contributing factor in balancing the work and home environment. The findings of the study based on the 2002 International Social Survey Programme (ISSP) data set indicate that individuals who believe that men and women should be treated more equally experience higher levels of work-life balance. This effect increases when individuals perceive that their spouses or v cohabiting partners spend more weekly hours on household labor. Being more financially dependent on a spouse or cohabiting partner also increases work-life balance for an individual; however, this effect decreases when the individual does less than his/her fair share of household labor. Additionally, an individual experiences less work-life balance when he/she perceives that the spouse or cohabiting partner does more than the fair share of household labor. Throughout the study, results indicate that women compared to men experience less work-life balance which points to the traditional gender role pressures within the home environment that women to this day have to face. Living in Germany or in the Netherlands compared to living in the United States increases an individual’s work-life balance while living in Portugal has no such effect. The positive country-effects suggest that there is stronger governmental support in Germany and in the Netherlands for spouses or cohabiting partners who are dual-income earners. , Committee Chair Mridula Udayagiri, Ph.D. Date vi ACKNOWLEDGEMENTS I would like to thank Dr. Udayagiri tremendously for being a continuous source of intellectual support. She has been an incredible mentor who encouraged and supported me in every step of the way to follow my research goals. I am so grateful not only for having had the opportunity to study sociological theory under her, but for experiencing academic research at its best while working for her. I would also like to thank Dr. MacIntosh immensely for being part of my thesis committee. I am so incredibly fortunate and thankful to have had Dr. MacIntosh as my statistics professor. He is a powerhouse of statistical methodology and my role model of what tough statistical research questions social scientists ought to ask. I aspire to follow in his footsteps. My further thanks goes out to Dr. Liu who was my first source of contact in the CSUS Department of Sociology. She showed me that I had come to the right place at the right time to solidify my sociological knowledge and to use it to achieve my career goals. I am ever so grateful to Dr. Liu for believing in me and for never doubting that I could land a dream job in governmental research. Thank you to Sanhita Gupta for enabling me to get into a fantastic research position and for always checking in on me during the thesis writing process. Lastly, I would like to thank my family and my dear husband Joseph for lovingly supporting me all along the way. Thank you, Joseph, for always believing in me, for being patient, for making me laugh, and for continuously nourishing my love for research. You helped me fix my wings. Now, I can fly again! vii TABLE OF CONTENTS Page Acknowledgements .......................................................................................................... vii List of Tables ......................................................................................................................x List of Figures ................................................................................................................. xiii Chapter 1. INTRODUCTION .........................................................................................................1 Significance of the Study ........................................................................................12 2. LITERATURE REVIEW ...........................................................................................20 Introduction .............................................................................................................20 Theoretical Perspectives .........................................................................................21 Work-Life Balance versus Role or Work-Family Conflict .....................................21 Role-Strain and Role-Spillover Theories ................................................................23 Border Theory .........................................................................................................26 Making the Case for Role-Strain and Role-Spillover Theories ..............................27 Concept of the Perception of Fairness of the Division of Household Labor ..........28 Legitimizing Principles ...........................................................................................32 Time Availability ...........................................................................................33 Resource Dependence ....................................................................................34 Gender Role Ideology ....................................................................................35 Social Structural Theory, Gender Stratification, and Gender Role Theory ...........36 viii Macro-level/Socio-structural Variations ................................................................42 The United States ..........................................................................................45 Germany ........................................................................................................47 The Netherlands ............................................................................................48 Portugal .........................................................................................................49 3. RESEARCH DESIGN AND METHODOLOGY ......................................................51 Hypotheses ..............................................................................................................51 Data Source .............................................................................................................54 Variable Measures ..................................................................................................56 Dependent Variable ................................................................................................56 Mediator Variables ..................................................................................................57 Independent Variables ............................................................................................59 Control Variables ....................................................................................................60 4. RESULTS AND INTERPRETATION ......................................................................67 Mean Values and ANOVA Results ........................................................................68 Crosstabulation Results for Religious Affiliation ...................................................79 Regression and Process Results ..............................................................................80 5. DISCUSSION AND CONCLUSION ......................................................................100 Strengths and Limitations .....................................................................................106 Directions for Future Research .............................................................................110 Suggestions for Improvement ...............................................................................113 References .......................................................................................................................115 ix LIST OF TABLES Tables Page Table 1: Total number of respondents for the United States, Germany, the Netherlands, and Portugal, by gender ..........................................................68 Table 2: Total mean values of work-life balance for all respondents by country affiliation ..........................................................................................69 Table 3: Independent samples t-test results with significant mean values for respondents by gender and country affiliation ...................................................69 Table 4: ANOVA and significant Scheffe post-hoc tests results for dependent and independent variables ..........................................................................................71 Table 5: Crosstabulation results for religious affiliation by country ................................80 Table 6: Multiple regression for gender role ideology on work-life balance (respondents residing in the United States, Germany, the Netherlands, and Portugal only) ..............................................................................................82 Table 7: Process mediation analysis of perception of fairness of the division of household labor with direct and indirect effects of gender role ideology on work-life balance ..........................................................................................83 Table 8: Multiple regression for time availability (respondent’s hours worked weekly) on work-life balance (respondents residing in the United States, Germany, the Netherlands, and Portugal only) ..................................................85 x Table 9: Process mediation analysis of perception of fairness of the division of household labor with direct and indirect effects of time availability on work-life balance ...............................................................87 Table 10 A: Process mediation analysis of the perception of fairness of the division of household labor with direct and indirect effects of resource dependence on work-life balance ........................................................89 Table 10 B: Effect of predictor variables and covariates from Process mediation analysis of the perception of fairness of the division of household labor with direct and indirect effects of resource dependence on work-life balance .......................................................................89 Table 11: Multiple regression for respondent’s weekly household labor hours on work-life balance (respondents residing in the United States, Germany, the Netherlands, and Portugal only) ...................................................................91 Table 12: Process mediation analysis of respondent’s weekly household labor hours with direct and indirect effects of gender role ideology on work-life balance ...............................................................................................92 Table 13: Multiple regression for spouse’s/cohabiting partner’s weekly household labor hours on work-life balance (respondents residing in the United States, Germany, the Netherlands, and Portugal only) ..................................................94 Table 14 A: Process mediation analysis of spouse’s/cohabiting partner’s weekly household labor hours with direct and indirect effects of gender role ideology on work-life balance ...........................................................................97 xi Table 14 B: Effect of predictor variables and covariates from Process mediation analysis of spouse’s/cohabiting partner’s weekly household labor hours with direct and indirect effects of gender role ideology on work-life balance ...............................................................................................97 Table 15: Decision outcomes for all tested hypotheses ....................................................99 xii LIST OF FIGURES Figures Page Figure 1: 2000-2001 marginal tax rate comparison of second earners versus single earners* in the United States, Germany, the Netherlands, and Portugal ............6 Figure 2: Days and paid percentage of maternity leave in the United States, Germany, the Netherlands, and Portugal .............................................................................7 Figure 3: Length of maternity leave in days in the United States, Germany, the Netherlands, and Portugal ...................................................................................8 Figure 4: GEM rank, value, and distinct categories for the United States, Germany, the Netherlands, and Portugal ...................................................................................9 Figure 5: Parliamentary seats held by women by all four countries .................................10 Figure 6: Women’s employment rate by country, including the United States, Germany, the Netherlands, Portugal, and the OECD total ...............................14 Figure 7: Concept map ......................................................................................................53 xiii 1 Chapter 1 INTRODUCTION Over recent decades, gender roles of both men and women have increasingly changed and moved from traditional to more egalitarian roles (Apparala, Reifman, and Munsch 2003). Since the 1960s, women have progressively entered the work force as part-time, full-time and independent workers due to groundbreaking changes in equal rights for women, especially in the work environment (Crompton and Lyonette 2006). In the United States and in many countries that are part of the Organisation for Economic Co-operation and Development (OECD), women represent over half of the labor force (OECD 2004). Yet, much of the unpaid household labor is performed still largely by women which raises questions as to how working women manage to create balance between the demands of the work and life environment. This study evaluates to which degree the perception of fairness of the division of household labor is a factor in managing work-life balance. However, including the perception of fairness of the division of household labor as a factor into work-life balance research is not meant to single out its importance. Rather, its inclusion attempts to point to the fact that a more detailed understanding of the mechanisms involved in the work-life balancing act – especially under the lens of continuous reshaping of gender role expectations – can only be reached by multi-layered, cross-country, as well as macro-and micro-level factor analysis. Before the noticeable increase of women in the work force occurred in the later 2 part of the 20th Century, the concept of work-life balance had not been of major concern for academics in the social sciences; traditional gender role ideology – in which men are viewed generally as the breadwinner and women as the caretaker of the home environment – was deeply ingrained in all levels of society (Crompton and Lyonette 2006). Historically, the division of labor was divided among men and women, with men as sole breadwinners earning money for the family by means of paid labor while women’s duties consisted of unpaid labor within the home environment (Crompton 1999; Crompton and Lyonette 2006). Since the last turn of the century, it has been evident that in many countries, the traditional breadwinner model has given way to the dual-earner model. With equal rights policies and laws put into place, women began to reevaluate their gender roles and to increase their educational levels as well as employment ambitions (Crompton and Lyonette 2006). While employment opportunities for women have become the means to supplement family income, they have also been a vehicle to a woman’s independence from the husband’s/cohabiting partner’s income as well as to her own gender empowerment. With the decrease in industrialization and the increase of the service sector, the male breadwinner model has been rapidly challenged by the influx of women workers (Crompton and Lyonette 2006). However, wages in the service sector are low, often forcing both spouses/cohabiting partners to work in order to meet financial needs. Especially minority women occupy many low-paying, un-benefitted jobs in the service industry such as in adult care-giving, childcare, and public/private housekeeping. Women employed as private domestic workers face even larger socio-structural barriers, as their 3 importance and participation in the labor force are persistently undervalued (Romero 2002). Many women face the stigma of being domestic workers and underreport the kind of domestic labor they are engaged in. Domestic labor is often an informal contract between two parties, and is not reported to the Internal Revenue Service (IRS) nor counted as part of the formal labor force or toward the Gross Domestic Product (GDP) of the United States economy (Romero 2002). Currently, the United States Census does not distinguish among the many types of domestic household work and applies a single code for all domestic work, therefore, underreporting the actual types and numbers of domestic workers. Unless female domestic workers receive the opportunity not only to declare their informal yet taxable wages without being stigmatized but to state their true occupation to the United States Census, their type of work and actual number in the work force remain largely invisible to the population and governmental institutions. Especially in the aftermath of global financial recession, unemployment in all sectors has decreased only slowly and continues to hurt many families, and makes the necessity of women to work even more evident; in the United States, many contracting and labor jobs in the construction sector had been lost due to the ‘crash’ of the housing market and financial institutions. Because many construction jobs are held by men, losing these kinds of jobs makes it even more important for women as wives or cohabiting partners to be able to financially support the family. Governments of the United States and Europe have realized the financial and societal capital that women’s employment has to offer: if women work, their financial income can be a key factor in reducing family poverty, reducing the reliance on welfare 4 programs, and redefining the value of unpaid housework by allowing others to do paid care-giving and housekeeping jobs – while the working woman allocates her time increasingly toward work (Crompton and Lyonette 2006; Orloff 2009). Yet, the increase in paid care-giving jobs does not necessarily indicate that gender ideology, care-giving, and the division of household labor have moved toward more equality (Orloff 2009). Although the United States has been one of the frontrunners in the employment rate of women since the 1960s with 43 percent in 1966, European countries as well have continuously promoted women’s employment. Among women’s employment, the United States has experienced the largest upward trend among working women who are married and have small children (Apparala et al. 2003). The Bureau of Labor Statistics (1998) reported that in the 1990s, 61 percent of married women were working and had children under the age of six years. Dual-earner households have increased in numbers, in fact, “the number of families with working couples today is double the number of families with only the male partner employed” (Saginak 2005:162). In comparison to the 1950s, households with only a male breadwinner were double the number of households with dual earners. While some women have the choice to enter the work force (because their husbands earn enough for the family’s financial needs), many other women have to work out of financial necessity (Crompton and Lyonette 2006). To this day, however, women have traditionally and disproportionately engaged in household labor, putting many unpaid hours into household chores and child-caring/care-giving activities (Braun, LewisEpstein, Stier, and Baumgaertner 2008). Although there are societal, cultural, and 5 individual differences in how spouses view the labor force participation and the division of household labor, it is argued in this study that it is especially women who are faced with a growing conflict to balance the home and work environment. The comparative research study inquires about the complex interaction of factors that affect the degree to which especially cohabiting women compared to cohabiting men in the United States, Germany, the Netherlands, and Portugal experience a conflict between the duties of the home and the duties of the work environment. Paradoxically, some women experience an increased sense of fairness of the division of household labor and, therefore, more work-life balance, even though they engage in the majority of household labor and are to varying degree financially dependent on their husbands’ income (Ruppanner 2008). The perception of fairness of the division of household labor seems to affect balance between the work and home environment. Women’s and their spouses’/cohabiting partners’ attitudes toward genderspecific division of household labor can influence perceived fairness of the household division of labor (Braun et al. 2008; Ruppanner 2008). While the attitude toward gender-specific roles is an individual characteristic, it can also be influenced by the cultural or societal norms of gender roles within a specific society. Policies and laws such as childcare support, motherhood benefits, or taxation of dual-income earners, as well as flexibility in the work environment can affect women’s work and household labor participation as well (OECD 2004). This research study takes into account these additional complex country-specific or ‘societal’ differences that can mediate women’s work-life balance (Crompton and Lyonette 2006). 6 The OECD (2004) reported (Figure 1, p. 6) that women as second earners were taxed on average at a higher rate than women who earned the same amount as single earners. Women as second earners were taxed even more when they earned 100 percent of Average Production Worker (APW) income compared to earning just 67 percent of APW income. Among the United States, Germany, the Netherlands, and Portugal, the Netherlands has the lowest and most equal tax rate ratio between second and single earners, while Germany has the highest tax rate ratio. Analyzing the tax rate ratio between second and single income earners can provide more insight on the degree to which government regulations enable spouses and partners (often women) to enter the labor force as second income earners without incurring unequal taxation. Figure 1: 2000-2001 marginal tax rate comparison of second earners versus single earners* in the United States, Germany, the Netherlands, and Portugal Women earning 67% of APW* Women earning 100% of APW Type of Second Ratio second Second Ratio second taxation earner/single earner earner/single system Single earner United Single 29 22 1.3 30 26 1.2 Optional/Joint Germany 50 34 1.5 53 42 1.3 Joint Netherlands 33 27 1.2 41 36 1.1 Separate Portugal 17 13 1.3 20 18 1.1 Joint States * The married couple has two children and the husband earns 100% of Average Production Worker earnings (APW). ** Average Production Worker earnings Source: OECD 2004 The length of maternity leave and the percentage of wages paid during maternity 7 leave are other indicators as to how governments support working mothers who just had a baby. According to Figure 2 (p. 7) and 3 (p. 8), Portugal allows the mother to take 120 days of paid maternity leave compared to the United States in which working mothers receive only 84 days of maternity leave. Whether maternity leave in the United States is paid depends on the company, the benefit package, and how long the mother worked in her current job prior to giving birth; very few companies will pay maternity leave in the United States (Institute for Women’s Policy Research 2007). Contrary to the United States, the governments of the Netherlands, Germany, and Portugal protect the rights of a working woman to take maternity leave by providing a financial safety net for the woman during her child-caring time at home. Figure 2: Days and paid percentage of maternity leave in the United States, Germany, the Netherlands, and Portugal Country Length of Percentage of wages paid in covered Provider of maternity period benefit leave in days United States 84 Not available N/A Germany 98 100 Statutory health insurance scheme, state, employer Netherlands 112 100 Social insurance Portugal 120 100 Social insurance Source: OECD 2010 8 Figure 3: Length of maternity leave in days in the United States, Germany, the Netherlands, and Portugal 140 120 120 112 98 100 84 80 60 40 20 0 Netherlands Germany Portugal United States of America Source: OECD 2010 Globally, equal opportunities for women have improved continuously, however, each country has specific policies and laws in place that either enable or prevent to various degrees married women from pursuing a career while having a satisfying family life. Equal opportunity policies are additional factors responsible for women’s mediation of work-life balance. On a social structural scale, governmental policies and laws are reflections of gender role ideologies by which a specific society is marked. Although individuals within a society can adhere to varying gender role ideologies from traditional to more egalitarian, there are scales such as the Gender Empowerment Measure (GEM) to identify the degree to which a society has removed structural/governmental barriers to promote women’s equality in the public sector. 9 Figure 4: GEM rank, value, and distinct categories for the United States, Germany, the Netherlands, and Portugal Country GEM GEM Seats in Female Female Ratio of Rank Value parliament legislators, professional estimated held by senior officials and technical female to women (% and managers workers (% of male earned of total) (% of total) total) income Netherlands 6 0.859 36.0 26.0 50.0 .64 Germany 9 0.831 30.6 37.0 50.0 .58 United States 15 0.762 16.3 42.0 56.0 .63 Portugal 22 0.692 21.3 34.0 50.0 .59 Source: United Nations Human Development Programme 2007 Of the four countries under investigation as illustrated in Figure 4 (p. 9), the Netherlands has the highest GEM value of 0.859 and is followed by Germany with 0.831, the United States with 0.762, and Portugal with 0.692. While the United States has on average the largest percentage of female legislators, female professionals, and female to male income ratio among all four countries (Figure 4, p. 9), it has the lowest percentage of parliamentary seats held by women (Figure 5, p. 10) and ranks only at place 15 worldwide in overall gender empowerment. The degree to which women are empowered within a country by governmental laws and regulations might be linked to how well they are represented in parliament (United Nations Human Development Programme 2010). 10 Figure 5: Parliamentary seats held by women by all four countries 40 36.0 35 30.6 30 25 21.3 20 16.3 15 10 5 0 Netherlands Germany United States Portugal Source: United Nations Human Development Programme 2007 Not only societal, but cultural, as well as racial, ethnic and religious affiliations (although racial and ethnic affiliations are not addressed in this research due to such variables missing in the 2002 ISSP data set) can impact women’s attitudes and perception about gender roles and their division of labor as well. Gender ideologies shape perceptions of womanhood which vary by culture and country. This research study attempts to apply a gendered lens to the structural (country-specific) as well as individual conditions influencing women’s perception about their own gender roles and the division of household labor when they navigate their lives between the home and work environment. The complex societal and individual conditions leading to such choices are the focal point of this sociological inquiry. Therefore, the research study attempts to be comprehensive by analyzing macrolevel/societal factors (these are country-level variables in addition to descriptions of 11 explanatory factors such as the GEM scale and country-specific policies) as well as micro-level/individual factors that have shown to be of relevance in previous research of work-life balance and fairness perception in the household division of labor. The complexities in the research present themselves in the overlap of macro- and micro-level factors. In this case, egalitarian and traditional gender views can be understood in terms of belonging to both, a micro-level/individual and macro-level/societal sphere because the specific views of an individual toward gender roles can be influenced by the broader societal ideology that leans toward either an egalitarian or a more traditional gender view and division of work and household labor. Although many women support equal rights for women, they take on a ‘doubleshift’ by going to work as paid employees only to come home and engage in unpaid household labor and care-giving responsibilities at home (Hochschild 1989). At home, women also engage in the ‘triple shift’ by doing emotional work in trying to make the home a pleasant environment for all household members (Hochschild 1989). Assuming that women face increased stress of struggling to fulfill their own multiple gender roles as workers, wives, partners, caregivers, and/or mothers, to what varying degrees are their gender role ideologies and their perceptions about the division of household labor a factor in achieving balance in the work and home environment (Milkie and Peltola 1999, Peltola et al. 2004, Schieman et al. 2003)? With more women in the labor force, research on household labor and work-life balance has increased over the last decades. The main assumption under investigation posits that – due to greater women labor force participation – more men would increase their hours in household labor and 12 spouses/cohabiting partners would adopt a more egalitarian view on the division of household labor as well, ultimately pushing on the individual level in the direction of societal change, toward gender equality (Lam and Haddad 1992; Apparala et al. 2003). However, women living in more egalitarian societies may report a decreased sense of fairness in the division of household labor as well as decreased work-life balance when the male spouse/cohabiting partner is only minimally involved in household labor and child-caring activities, suggesting that other factors such as gender role ideology may be involved. The division of household labor has only disproportionately kept up with the equal rights movement and women’s increased labor force participation in what Hochschild has called the ‘stalled revolution’ (Hochschild 1989; Walzer 1996; Glavin and Schieman 2011). Analyzing the macro- and micro-level conditions which lead to the perception of equity in the division of household labor can give clues about the mechanisms that enable cohabiting women as well as men to perceive balance in their work-life in the United States, Germany, the Netherlands, and Portugal. Pinpointing the mechanism might also indicate why it has come to a ‘stalled revolution’ and why full gender equality has not been fully realized yet in paid and unpaid labor. Significance of the Study Many jobs in the service and farm industry pay very little for unskilled labor which often raises the necessity for both spouses to work in order to cover the financial needs of families, especially the ones with children (Crompton and Lyonette 2006). In the United States, half of the women are employed in non-manufacturing jobs while 65 percent of the employed women are married and live with the spouse or partner, and 47 13 percent of the overall employed women have a child or children under the age of 18 (Bond and Galinsky 2011). Among the many employed women, “two thirds of working mothers have children who are preschoolers, and many are involved in the labor force while their child or children are still infants” (Saginak 2005:162). Women who participate in the labor force enable their families to bypass or move out of poverty (Crompton and Lyonette 2006). However, government policies have to be passed which provide women with equal pay, job securities, childcare support, and favorable taxation brackets for dual-income earners – practical incentives to encourage women to become a larger part of the labor force (Guest 2002; OECD 2004). Although women participate increasingly in the work force, even to the same degree as men in the United States, they are still paid less than men on average (Bureau of Labor Statistics 2005). In recent decades, families in which both woman and man work, have increased; yet the home environment has continuously been perceived by the larger population as the responsibility of the woman (Hochschild et al. 1989). 14 Figure 6: Women’s employment rate by country, including the United States, Germany, the Netherlands, Portugal, and the OECD total 80 70 Employment Rate 60 Germany 50 Netherlands 40 Portugal United States 30 OECD total 20 10 0 1966 1973 1980 1987 1994 2001 2008 (Vertical drop-off lines indicate missing data) Source: OECD 2010 Comparing the United States, Germany, the Netherlands, and Portugal, the employment rate of adult women between 1998 and 2008 (Figure 6, p. 14) has increased most dramatically in the Netherlands at 70.2 percent with an 11.1 percentage point change, followed by Germany at 64.3 percent with an 8 percentage point change, and a moderate to weak 4.2 percentage point change to 62.5 percent in Portugal (OECD 2010). The United States has experienced a decline in women’s employment rates in 2008 with a decrease of a 1.6 percentage point change at 65.5 percent, yet the United States is at second place among the four countries in terms of the employment rate of women. Overall, despite a slow change, the average employment rate of women in OECD 15 countries has increased between 1998 and 2008 by 2.9 percentage points to 57.8 percent, indicating growth in women’s overall employment participation. However, the participation of women in the labor force has only disproportionately changed the rate men engage in household labor and child-caring/caregiving activities (Livingston and Judge 2008; Coltrane 2000; Bianchi, Robinson, and Milkie 2006). Although the breadwinner model has been challenged by women’s labor force participation, the assigned gendered duties of men seem to not have changed significantly and have largely remained confined to the work environment. According to research on fathers’ attitudes toward household labor, their participation in the actual household labor has only to a limited degree correlated with societal changes such as shifts in gender role ideology and policy change. Women, on the contrary, have taken on multiple roles as unpaid caregivers at home as well as paid laborers in the work environment. Women’s workload has increased suggesting that women have not been freed from the traditional views of women’s domestication. Hence, women presumably feel an increasingly growing conflict (simultaneously, a challenge to balance the multiple roles) between the home and work spheres (Livingston and Judge 2008). Yet, there is some evidence that women who share a more traditional view on the division of household labor will experience less of a conflict even though they work more paid and unpaid labor hours than their spouses (Braun et al. 2008). This suggests that the perception of fairness of the division of household labor may act as a powerful factor in facilitating work-life balance regardless of actual hours worked. The structural as well as individual conditions under which women perceive balance or growing conflict need to 16 be investigated with further research studies in order to inform new policies that support social change toward women’s equal treatment in the work and home environment, make it fiscally beneficial for women of dual-income families to work, encourage and give men the choice to take paternity leave at minimal pay-cut, and enable families with dual earners to balance the work and home environment life (Lachance-Grzela and Bouchard 2010; Coltrane 2010). Also, most of the previous research has focused on the difference of gender roles, and not on the complexities women have to manage in the work and home environment (Hare-Mustin 1988). When analyzing current gender roles, researchers have to increasingly operate on multiple levels. As an attempt to further explore the impact of gender ideologies on work-life balance and gender roles and responsibilities, this study will use the 2002 International Social Survey Programme (ISSP 2004) data set to highlight some of these elements in gender relations within a comparative framework. The 2002 ISSP data set is a fitting empirical data research pool because it inquires about women’s and men’s attitudes towards gender ideology, the division of household labor as well as their gender role perceptions in the home and work environment which can be used to assess how spouses and cohabiting partners mediate work-life balance. Even though women engage on average more in household labor compared to their male spouses, clearly pointing to a persisting imbalance in the division of household labor, a review of research studies on the perception of the division of household labor points to the fact that many women still report fairly low levels of feeling treated unfairly (Braun et al. 2008; Mikula 1998). In fact, only 20 to 30 percent of women expressed 17 feelings of being treated unfairly according to a literature review by Mikula (1998). The assumption that the actual labor within the household is a correlating indicator for the perception of fairness of the division of household labor needs to be reevaluated (Braun 2008). What is considered fair or unequal seems to differ from person to person, and micro-/individual factors may influence the perception of fairness scale. Yet, the interpretation of what constitutes fairness may be impacted not only by the individual’s own view on gender roles but by the gender norms of a specific group or society in which a woman or man grew up (Braun et al. 2008; Lachance-Grzela and Bouchard 2010; Coltrane 2010). After review of the research literature on work-life balance and the perception of fairness of the division of household labor, social scientists are still working on fully identifying all significant factors on the micro- as well as macro-level that influence work-life balance. One of the challenges to identifying the most significant factors to influence work-life balance is that answers to the question become a matter of perspective and focus: researchers of work-life balance have often focused on the work environment and factors such as work hours, benefits, supervisor’s support, flexibility, state’s or country’s work policies, and gender ideology. Researchers looking more closely at the home environment have focused on factors such as the perception of fairness of the division of household labor, time availability, resource dependence, gender ideology, as well as actual hours put into household work. Although some researchers such as Milkie and Peltola (1999) have used “perceived unfairness in sharing housework” (p. 476) as one of the predictors of work-life balance, the perception of fairness of the 18 division of household labor has not been emphasized enough as a key indicator for worklife balance. It is argued in this research study that the perception of fairness of the division of household labor is one of the more striking indicators for managing work-life balance. This study does not intend to single out the perception of fairness of the division of household labor as the key factor in managing work-life balance. On the contrary, it attempts to point to the fact that by its mere inclusion of the perception of fairness factor, the study emphasizes that what alleviates women’s work-life conflict may be even more complex in nature than previously expected. In order to understand the concept of the perception of fairness of the division of household labor, country-specific, socio-cultural, macro- as well as micro-level factors – which continuously shape and reshape how women and men experience and act upon gender expectations – have to be examined closely. This is not to say that it is the only concept and key factor that allows women and men to balance their work-life spheres, especially when countries’ policies are taken into account that allow women to engage in paid labor without them having to worry about promotions, childcare, and disadvantaged taxation within their dual-earner household. However, in this study, the perception of fairness of the division of household labor is emphasized as a ‘meso-marker’ that is both, a micro-level variable as well as a macrolevel type variable which is reflective of gender ideologies embedded in societal structures. Coltrane (2010) has called for more comprehensive and complex research designs that include cross-country and trend studies, as well as regional, racial/ethnic, immigrant, or gay/lesbian/transsexual groups. Also, Coltrane points out that only a few recent studies 19 such as of Lachance-Grzela and Bouchard (2010) and Cooke (2007) have begun to “take seriously the economic, political and social context under which couples divide household labor as well as paid labor” (Coltrane 2010:798). Although this study cannot include trend studies or a wide array of diverse groups – because, with exception of regional data, the 2002 ISSP data set lacks such information – it tries to be more comprehensive by using a cross-national approach to identify societal factors as well as by adding a diversifying yet little used factor of religious affiliation. One theoretical approach which Coltrane deems inescapable to better understand the linkage between micro-/individual and macro-/societal factors that mediate the division of household labor is to perceive the division of household labor as a reflection of gender inequality on a larger structural scale within that particular society. Therefore, this study aims to expand on popular theories such as time availability, resource dependence, and gender role ideology used to analyze the division of household, and to integrate social structural theory, social stratification, and gender role theory which have been used to assess women’s attitudes toward their gender roles, division of household labor, and social equality. 20 Chapter 2 LITERATURE REVIEW Introduction Work-life balance and the perception of fairness of the division of household labor have been prominent yet very distinct concepts among researchers. Some researchers have mainly focused on unpaid labor by analyzing the division of household labor while others have focused solely on paid work by analyzing organizational constraints and policies of the work environment (Coltrane 2000; Keuzenkamp and Hooghiemsra 2000; Wierda-Boer 2009). Only over the past few years, have these in itself extensive research concepts been integrated into more complex research designs aiming at the overlap of the two research topics. With the goal to build on and contribute to the research literature of the two concepts and by offering an integrative research approach of comparing the perception of fairness to work-life balance across four countries, this research focuses on clarifying to which degree the perception of fairness of the division of household labor can in fact be used as a key-indicator or ‘meso-marker’ for work-life balance. Milkie and Peltola (1999) have directly linked and identified the “perceived unfairness in sharing housework” (p. 476) to work-life balance as one significant indicator. They analyzed a sample taken from the 1996 General Social Survey of people who were married and employed. Although their research study consisted of a detailed analysis to identify gender-specific factors that mediated married women’s and men’s perceived successes in balancing work-life, they did not include a wider analysis on societal factors that could as well affect work-life balance. By including a macro-level, 21 cross-country comparison, this research study aims at investigating and describing the larger structural frameworks that influence cohabiting women’s and men’s individual characteristics, and therefore, their mediation of work-life balance. Within the sphere of work-life research, this study draws from several theoretical perspectives such as Work-Life Balance, Role or Work-Family Conflict, Role-Strain and Roll-Spillover Theory, as well as Border Theory. Within the sphere of household labor research, the study will make use of the theoretical concepts of time availability, resource dependence, and gender role ideology which are known as legitimizing principles and are commonly used in fairness perception and the division of household labor research. In addition, the study builds on social structural theory with special focus on gender stratification or ‘gentrification’ as used by Bondi (1991), and on using gender role theory to view the perception of fairness of the division of household labor and its effect on work-life balance through a gendered lens. Theoretical Perspectives Work-Life Balance versus Role or Work-Family Conflict This research study uses the term of work-life balance, rather than role or workfamily conflict, which are earlier terms used to analyze the relationships and challenges associated with the work and home environment. Work-life balance is a newer concept described as the “good functioning at work and at home with a minimum of perceived role conflict” (Clark 2001:751; Guest 2002). It is a concept which focuses on balance rather than conflict. Yet, the literature on the work-life has not agreed upon a general and consistent definition: work-life balance is used and defined differently throughout 22 research studies, and it remains in a state of flux waiting to be defined concisely in its meaning (Frone 2003; McMillan et al. 2011). Marks and MacDermid (1996) describe positive role balance within the work and family environment as “the tendency to become fully engaged in the performance of every role in one’s total role system, to approach every typical role and role partner with an attitude of attentiveness and care” (p. 421). McMillan et al. (2011) describes Marks’s and MacDermid’s conceptualization of worklife balance as “a continuum with imbalance (in either role) anchoring one end and balance (again in either role) anchoring the other end” (p. 13). However, because there is a multitude of factors that affect people’s ability to meet the challenges of the work and home environment, the degree to which people deal with the two spheres is a flux of conflict and balance. Work-life balance is still defined in terms of a lessening or absence of conflict between the roles in the home and work environment, just as conflict poses a lack of balance (McMillan 2011; Frone 2003). Although the concept and term of worklife balance has developed out of border theory (which is not utilized in this study), the study will continue to use the term work-life balance to point to factors such as gender role ideology which allows women to perceive the division of household labor as fair and to balance their work-life although they engage in the majority of the household labor. The concept of work-life balance embedded in border theory has been criticized for not accounting for the ‘double shift’ of women, especially when their roles blur (Glavin et al. 2011; Hochschild 1989). This study inquires in part about the possibility of perceiving balance between the work and family environment even if the actual division of household labor is unequal. 23 Role or work-family conflict between the work and home environment is a result of the stressors that develop when role responsibilities of the work environment infringe on the role responsibilities of the home environment and vice versa. For the reason that the home environment with household chores and care-giving duties has traditionally been assigned to a woman’s role, working women especially are expected to face more work-life stressors than men. Yet, work-life stressors affect both men and women, and work-life conflict has been identified in many research studies as being the cause for health-related physical and psychological illnesses such as depression (Glavin and Schieman 2011; Schieman, Milkie, and Glavin 2009; Bellavia and Frone 2005; Frone, Russell, and Cooper 1992). Some researchers resist using the term work-life balance for they argue that the term work-life balance fails to account for gender role differences (Caprioni 2004; Smithson and Stokoe 2005). They point to organizational research which often makes use of the term to describe rather individual and merit-based characteristics which can facilitate work-life balance while failing to account for structural inequalities imbedded in social institutions such as the workplace. The following perspectives have largely dominated work-life research: role-strain and role-spillover theory on one hand, and border theory on the other. Role-strain and role-spillover theories differ from border theory in their views on role pressure incompatibility and role-overlapping. Role-Strain and Role-Spillover Theories Role-strain and role-spillover theories describe work-life conflict as “a form of inter-role conflict in which the pressures from work and family domains are mutually 24 incompatible in some respect” (Greenhaus and Beutell 1985:77; Glavin and Schieman 2011). Role-strain theory – also known as the conflict hypothesis – proposes that the roles of the work and home environment create a conflict for a person; with the duality of paid and unpaid work, time becomes limited and energy has to be divided between the two spheres which creates a strain or conflict for the person (McMillan et al. 2011; Marks 1977; Goode 1960; Wierda-Boer et al. 2009). Conflict between multiple roles can go both ways: family roles can interfere with work, and roles at work can interfere with the family environment (McMillan et al. 2011). Glavin and Schieman (2011) argue that role-strain emphasizes role conflict and does not provide alternative explanations as to how work-life roles can be perceived differently by people. Especially with societal changes from the traditional composition of the nuclear family (with the male as the sole breadwinner) to dual earner households, and with the development of technologies that allow employees to work from home, there seems to be the need for a new theoretical model that could account for the changes: border theory. Yet, role-strain and role-spillover theories provide the needed theoretical framework for this research study to explain how conflict may be mediated or avoided by other interfering mechanisms such as gender role ideology to help working women make sense or justify their mutually incompatible roles. Role-spillover theory gives explanatory insights as to why women face more challenges and conflicts between their employee role and roles within the home environment while the traditional gender role stereotype of men as the provider of the family allows the man to engage more in work-related activities at home (Glavin et al. 25 2011). His focus on work is often assessed against the traditional stereotype and, therefore, perpetually normalized and approved (Glavin and Schieman 2011). Yet, women experience role-spillover quite differently due to the persisting female stereotype of the woman as homemaker and caregiver. Role-spillover has multiple effects on women: Women will engage in more emotional and physical work, in what Hochschild (1989) has described as the ‘double’ and ‘triple shift’ to enact their gender ideals. They may feel conflicted in that they believe they inadequately fulfill their home responsibilities due to their work demands (Glavin et al. 2011). In addition, they can experience the pull of what Blair-Loy (2003) describes as the ‘work-devotion schema’ which “demands that [they] give an immense time commitment and strong emotional allegiance to [their] firm or career” (p. 7). Role-spillover theory is also used in this research because the newer school of work-life balance has used the theory to strengthen the argument that balance should be perceived as a source of enrichment and integration (McMillan et al 2011). According to some work-life researchers, “integration is grounded in spillover theory, in that individuals carry attitudes and beliefs from one role to another” (McMillan et al. 2011:11). Spillover may also explain how people merge or integrate the work roles with their family roles according to their attitudes (for example, their attitudes about gender role ideology) and preferences. Attitudes that spill over from the work into the family environment and vice versa may initiate the negotiation process between work-life roles and any inequalities that may be attached to them. The varying physical and emotional responses of women to spillover raise the question as to how much gender role ideology and the perception of the division of household labor help 26 women to balance their roles of the home and work environment. The lingering of gendered and stereotyped roles of men and women points to the structural influence of larger societal gender role ideologies which transcend by means of policy mandates, socio-cultural, ethnic, and religious traditions into individual households and parents’ socialization behaviors. Looking at the mechanism of structural change, role-strain and role-spillover theories can give clues as to how individuals either perpetuate gender differences or influence structural change toward more equitable treatment of women. Border Theory Border theory is a relatively new perspective and has been used increasingly for the past ten years (Clark 2000). It is described as “a subjective, cognitive phenomenon involving perceived integration of work life and home life that is situated in a highly interdependent work-family context such as the simultaneous work and family demands that can be present when people bring their paid work into the home” (Desrochers, Hilton, and Larwood 2005:449; Glavin and Schieman 2011). This perspective allows researchers to describe people’s work-life stressors in the context of changing workfamily dynamics such as a mother working with a flexible schedule from home, yet facing interferences from the home environment such as a crying baby. In border theory, people are described as “border crossers” (Glavin and Schieman 2011:44; Clark 2000). When they work from home, they blur their roles of worker and spouse, partner, parent, caregiver, and/or homemaker (Clark 2000). With border theory, the blurring roles of the home and work spheres a person experiences are not described in terms of a role conflict, 27 but in terms of role blurring intensity (Greenhaus et al. 2006; Galvin and Schieman 2011). Also, Voydanoff (2004) has developed the concept of ‘boundary-spanning resources’ such as parental leave, family vacation time, and flexibility in the work schedule which have been identified to mediate work-life balance. For example, more flexibility to work from home allows a person to minimize the role conflicts, therefore, experiencing more work-life balance (Clark 2000; Clark 2001). Making the Case for Role-Strain and Role-Spillover Theories Border theory will not be applied in this analysis, although the term of work-life balance – which is derived from border theory – will be used: the rationale behind giving preference to role-strain and role-spillover theories is the assumption that when roles are performed, they do not necessarily blur, but occur as a distinct execution of a role separately from another role, even if they happen at the same time. Role blurring refers to the degree to which a person can jump back and forth between the roles of the home and work environment to facilitate work-life balance; this is increasingly possible through work flexibility and technology which allows people to work from the home environment. Even though the physical barriers of work and home vicinities are lifted, it is argued in this study that such removal does not eliminate the conflict that people perceive when they try to balance all roles of the work and home environment. It is agreed upon, however, that border theory allows for another new explanatory tool to describe situations in which people experience no role conflict due to the removed physical barrier between the work and home environment, made possible by flexible work schedules and technology. 28 Also, with the assumption that “integrating work and family facilitates transitions between these domains” (Desrocher and Sargent 2004:41), border theory still seems rather limited in accounting for persistent gender differences and inequalities in the roles men and women take on. Being a border crosser by working from home does not necessarily change the way a man and woman do gender. Also, many low-level, lowpaying jobs (especially in the service industry) require women and men to leave the home environment, not allowing them to have the flexibility to work from home or via technology. While it is clear that there should be different theories such as border theory to address ever-changing social realities, role-strain and role-spillover theories can account for persistent realities of gender stratification within the public and the private spheres. Especially role-spillover theory can explain how emotions and attitudes spill over from the work to the home environment and vice versa, possibly pointing to mechanisms of gender ideology transfer that link the public spheres with social institution and embedded gender stratifications to the home environment with its individual actors. Concept of the Perception of Fairness of the Division of Household Labor There has been wide-ranging research on household labor. Especially since the 1990s, there have been over 200 academic publications (Apparala 2003, Coltrane 2000). With women entering the labor force, the dynamics of the family unit has changed. Two camps have developed over the past 30 years of household labor research: some social scientists argue that there has been much change in time allocation with more working women spending less time on household labor and even having more leisure time. Contrary to that perspective, other social scientists posit that working women experience 29 more labor stresses because they not only put unpaid time into household chores and care-giving, but spend more time working as well. According to the perspective of researchers that change has occurred in the division of household labor, time-use diary research conducted from the early 1960s to the 1990s in the United States as well as in European countries has provided evidence that after using control variables which accounted for structural societal changes, men have increased their participation rate in domestic labor while women’s actual hours put into domestic labor have decreased (Sullivan 2000). Baxter (1992) has argued that by comparing working women to non-working women, working women on average do less housework. With more women in the labor force, this presumably pushes the division of household labor between couples toward more equality. According to Gershuny (1995), men increase their hours spent on household labor when women begin part-time work or when women move from part-time work into full-time employment. Sullivan (2000) points out that leisure time has increased as well for working women over the past 30 years. However, although there have been changes in the men’s participation in household labor, the amount of time put in by men has reduced employed women’s household labor hours by only a limited amount (Sullivan 2000). Therefore, overall labor of employed women has increased in what has been called the working women’s ‘double shift’ (Hochschild 1989). The rationale is explained in terms of limited changes in the assignment of household labor responsibilities to women (Sullivan 2000). Even if men engage in more household labor, and seem to accept “an equivalent amount of domestic 30 responsibility, study after study has shown that women retain responsibility for the management of domestic tasks even when men are (helping in) performing it” (Sullivan 2000:438). This may be due to two factors: the couple’s gender role ideology as well as the country-specific gender role ideologies which in turn exert influence onto the gender role ideologies held by the couple. At this point is should be pointed out that Sullivan (2000), Gershuny (1995), and Baxter (1992) share the perspective that positive change toward equality in the division of household labor and especially toward reduced time pressures has occurred. Yet, Hochschild (1989) contests the positive change toward reduced time pressures and describes it as rather minute. She stresses that women’s increased labor force participation has put more rather than less time pressure on women’s household labor. Hence, her perspective of working women facing more time management conflict due to the ‘double shift’ stands in direct contradiction to research that working women’s leisure time has increased. However, this research study views both perspectives on change in the division of household labor not necessarily as mutually exclusive; both, reduced as well as increased time pressures due to work or a ‘double shift’ may both be mediated by gender role ideology and what is perceived by the couple and the woman as fair. Therefore, the perception of fairness in the division of household labor is used as another key factor in contributing to work-life balance in addition to the often used factors of time availability and actual time spent on household labor. Still, this research includes the traditional theoretical approaches of time 31 availability, resource dependence and gender role ideology as well, because they do affect the perception of fairness of the division of household labor and work-life balance (Braun et al. 2008; Milkie and Peltola 1999). What Sullivan (2000), Gershuny (1995), and Baxter (1992) seem to omit is the fact that working women who have their own income and belong to a higher social class may be in a position to renegotiate their household labor responsibilities by outsourcing childcare and household chores to a paid third party. Outsourcing work, however, may not change the actual gender role ideology held by the couple. A financially well-to-do working woman may still view household labor and care-giving as duties essentially assigned to her or to women in general, allowing for the enacting or perpetuating of a specific gender role image which would be a reflection of her larger societal, cultural, ethnic, and/or religious background. Also, research of case studies has shown that some women who make the same or more money than their husbands will engage in household and care-giving duties in what is called “doing gender” (Ruppanner 2008:520; West and Zimmerman 125). This is the woman’s physical and emotional effort to reaffirm the traditional gender role ideology within the household in order to counteract her equal or higher socio-economic success compared to her husband’s in the work environment. For the research, it is also essential to include both the actual hours of household labor as well as the work hours of the respondents. Hochschild (1989) has often been criticized for focusing solely on the unpaid hours of women within the family environment rather than analyzing the total work load of paid and unpaid work of both men and women (Beaujot and Liu 2005; Wierda et al. 2009). Bianchi and Raley (2005) 32 have claimed that comparing women’s and men’s total paid and unpaid workload reveals that both work about the same (Wierda et al. 2009). Frone (2003) stresses that total paid and unpaid workload negatively affects work-life balance. Yet, the main question then becomes “Who does what kind of work, and for how long?” Even if men and women are engaged in the same amount of work hours, it does not imply that the division of household labor is equally divided among women and men. Inequalities in the division of labor, especially household labor, might still persist. Legitimizing Principles Research on the division of household labor has focused on three main factors: Time Availability, Resource Dependence, and Gender Role Ideology. All three factors will be used for this research study in what Braun et al. (2008) have called “Legitimizing principles” (p. 1146). Both factors – time availability and resource dependence – are grounded in equity theory proposed by Walster et al. (1978; Braun et al. 2008). Equity theory posits that fairness in the division of household labor is perceived when both spouses/cohabiting partners put in and produce equal shares of work effort. The perception of fairness of the division of household labor can be interpreted in terms of a person either under- or over-benefitting from the division of household labor (Lively et al. 2008). If a person engaged in more than his/her fair share of household labor, he/she would under-benefit from the division of household labor. If a person engaged in less than his/her fair share of household labor, he/she would over-benefit from the division of household labor. In sociological research, equity theory has been attentive especially toward women because, historically, women engage in more household labor than men: 33 therefore, continuously under-benefitting from the division of household labor. Time Availability By applying equity theory, input of work usually relates to work hours that a spouse/cohabiting partner allocates to employment. Research has shown that women who devote more time toward employment will reduce their time involved in household labor; while their household hours decrease, some studies have shown that men’s household work increases (Bianchi et al. 2000; Lewin-Epstein et al. 2006; Braun et al. 2008). Also, regardless of gender, when spouses/cohabiting partners work and have less time for the household, they may either lower the standard for household duties and/or employ a third party (if another household member beyond spouses/cohabiting partners is unavailable) to do the household labor (Bianchi et al. 2000; Lewin-Epstein et al. 2006; Braun et al. 2006; Braun et al. 2008). It can be argued that if a spouse/cohabiting partner does not work full-time, he/she has more time available, therefore, should put the available time into more household labor. Applying equity theory, the spouse who puts more and, therefore, an unequal amount into household labor, should perceive and legitimize this division as fair if the other spouse/cohabiting partner works more hours such as full-time hours in the labor market (Braun et al. 2008). However, if one of the spouses/cohabiting partners would work longer hours such as 50-plus hours in the labor market – causing the other spouse/cohabiting partner to be bound to the home environment because of caregiving activities – the division of labor would be perceived as unfair. This points to the possible necessity of a prior agreement between the spouses/cohabiting partners about the amount of hours that each attempts to invest in the labor market (Braun et al. 2008). 34 Resource Dependence By applying equity theory to resource dependence, the output or production of invested work reflects resources, typically income and/or education that a spouse/cohabiting partner can access in order to leverage relative power and to negotiate the division of household labor (Bittman et al 2003; Braun et al. 2008). Lam and Haddad (1992) use resource dependence theory with the assumption that “the social organization of married couples’ family role dynamics is based on the continual exchange or rewards and gratifications” (p. 69). When a woman earns more income and has a higher education, she is in a position of more independence, especially in regard to marriage because she will not be reliant on a husband for financial needs. Therefore, the theory puts forward the idea that with the increased labor force participation of women, the increasing relative bargaining power of women should move the division of household labor toward more equality (Apparala et al. 2003; Lam and Haddad 1992). In the notion of equity theory, a woman who does not earn as much as her husband and/or does not have a higher education to potentially earn an equal or larger amount of money, will not seek a life outside of marriage and financial security, therefore, she will be more likely to legitimize and perceive the unequal division of household labor as fair (Braun et a. 2008). Looking at both time availability and resource dependence theories, women’s actual household labor hours do not fully predict whether the division of household labor is truly unequal. Women who make an equal or larger amount of money than their husbands may outsource household labor to a third party, reducing their own household labor time. This may lead to the wrong conclusion that the overall division of household 35 labor between the wife and husband has become more equal. The division of household labor could still remain unequal due to the ‘elevator effect’ (Braun et al. 2008; Beck 1992). Beck (1992) developed this concept by analyzing why lower classes tend to accept income inequalities (Braun et al. 2008). His research indicated that as long as welfare entitlements existed to supplement the income of a lower class family, they would not contest unfair income levels or at least not perceive their income as unfair. Similar to the acceptance of income disparities by lower classes, it is argued that any remaining unequal division of household labor – for example, the wife does less or little housework while the husband does even less or no housework – is accepted or perceived as fair as long as a third party alleviates the strain of household labor and care-giving duties for the woman (Braun et al. 2008). Gender Role Ideology The hypothesis of gender role ideology becomes the missing and possibly most important link within the concept of legitimizing principles. Gender role ideology is a powerful factor in that it can act as a facilitator of actual gender equality within the division of household labor, but it can also perpetuate inequality within the division of household labor. Gender role ideology affects perception of fairness of the division of household labor to the degree that an actual unequal division of household labor is perceived as legitimate and fair; a paradox, for it would be the perception of fairness and equality of an actually unequal division of household labor. However, in the case of women who are more egalitarian in their gender ideology, they will push for a more equal division of household labor which only then will be 36 perceived as more fair. Braun et al. (2008) has stressed that egalitarian women push for more egalitarian rights and, therefore, are more successful in achieving more equal division of household labor compared to more traditional women. On the other hand, women with a more traditional gender role ideology are “likely to accept an imbalanced division of household labor as an integral part of a proper woman’s role, irrespective of the amount of their labor market employment and even if it leaves them alone with the household work” (Braun et al. 2008:1147). In this specific case, Katz and Lavee (2002) explained that traditional women view the unequal division of household labor as fair “if it (a) corresponds to what they are socialized to value in a relationship (outcome values), (b) matches their social and normative standards (on the basis of comparison referents), and (c) is perceived as justifiable or legitimate” (p. 28). Social Structural Theory, Gender Stratification, and Gender Role Theory In hunter-gatherer societies, men hunt and women stay at home. This strong bias persists in most agricultural and industrial societies, and, on that ground alone, appears to have a genetic origin … My own guess is that the genetic bias is intense enough to cause a substantial division of labor even in the most free and egalitarian of future societies … Even with identical education and equal access to all professions, men are likely to continue to play a disproportionate role in political life, business and science. (Edward Wilson 1975, quoted in Bonvillain 1998:179) Assigning gender roles to both men and women has been part of human reality. The all too familiar images of men as hunters and women as gatherers have lingered through the centuries. Assumed biological differences between men and women have 37 been grounded in essentialism. To this day, socio-biologists and groups of psychologists (especially evolutionary and cognitive psychologists) attempt to explain traditional gender roles by making the “claim that distinctions in social behavior of women and men are derived from innate genetic differences” (Bonvillain 1998:179). Gender role theory suggests that there is a polarization of gender roles in which men are identified with the work environment and women are identified with the family environment (Livingston et al. 2008). As Prince Cooke (2011) points out, the assigned role of women as housewives and caregivers has been linked to women’s biological ability to give birth. For the reason that women are the ones who are able to give birth to children, care-giving has become a responsibility assigned to women. In addition, care-giving is usually an activity executed in the household, so the household – as an extension of the ‘natural’ role of the woman – has become the domain of women in what is known as the cult of domesticity (Bonvillain 1998). Therefore, the work environment is not perceived as a ‘natural’ environment for the woman, but rather the appropriated domain of the man. Role theory used by Eagly and Steffen (1984) maintains that people make stereotyping assumptions when they observe other people in specific roles (Apparala 2003). Especially with Prince Cooke’s assertion that historically women were viewed as caregivers due to their biological ability to give birth and breastfeed, women were stereotyped into being more communal and caring for others, while men were stereotyped into self-confident and able leaders (Apparala 2003; Eagly and Steffen 1984; Hoffman 1977). This stereotyping belief is quite possibly a factor in gendered role distribution of men representing the public and women representing the private sphere. Gender 38 stereotypes manifested themselves in that roles of men were more valued than women’s, hence, explaining one angle of the gender gap in unequal wages and the undervalued, unpaid work of household labor (Bonvillain 1998). The lack of valuing household labor stems from the fact that it is unpaid (Eisenstein 1979; Bonvillain 1998). Bonvillain (1998) explains that “given capitalist values implicit in economic and social relations, social worth is measured by money obtained through one’s labor” (p. 176). When work is unpaid, it does not gain value and, therefore, is exempt from social recognition. The Gross Domestic Product (GDP) reflects the value of paid versus unpaid work: unpaid household labor is not accounted for in the GDP while paid care-giving and household labor is part of the overall GDP. This can be interpreted in Capitalistic terms: if a woman does unpaid care-giving and household labor, she is considered economically unproductive. Many industrialized countries, especially the United States with its Capitalist economy, have assigned care-giving responsibilities to the least powerful groups such as women (Prince Cooke 2011; Tronto 1993). Hence, gender stratification becomes evident within women’s care-giving activities: especially when social class hierarchies are compared, women from lower classes and minority groups will engage in the care-giving and household labor duties of women from higher classes. Also, while women from higher classes can more easily employ third-party caregivers and household aides without affecting the household savings negatively, women from middle and lower classes are forced to engage in household labor in order to abstain from depleting the household income. If household labor would be a paid activity, its monetary value would have to be at a level of other paid jobs in order for employees to cover all basic living 39 expenses (Bonvillain 1998). By not paying women for performing household labor, wages can be kept low and profits can remain high. It becomes a perpetuation of profit mechanisms in what Eli Zaretsky (1973) described as “the wage labor system is sustained by the socially necessary but private labor of housewives and mothers. [Their work] constitutes a perpetual cycle of labor necessary to maintain life in society. In this sense, the family is an integral part of the economic system under Capitalism” (p. 79). State policies can either empower or inhibit women to be part of a valued work force by determining the extent of compensation, health-care and retirement benefits, care-giving support and other subsidies for working women (OECD 2004). Further, Bondi (1991) reexamines the previous research on gender proposing that gentrification is at the center of understanding the changing roles among women. As she explains gentrification, she believes that intersections of production and consumption, agency and structure, as well as cultural and economic conditions have to be understood through the gender lens. Looking at the intersections, the researcher can gain access to the mechanisms of gender construction and its change. Her research explores the mechanisms which lead to women identifying themselves within certain categories (Bondi 1991). It is a reminder that a researcher has to look at main aspects and their intertwined complexities, such as the labor market, culture, the individual agents and the structural system within a society to get a comprehensive picture about gender ideology within a given society. Social structural theory aids in understanding gender stratification and gender role theory in that gender role differences and gender inequalities are based on larger macro- 40 level structures. Analyzing a society at the macro-level and using social and gender stratification theory, allows the researcher to identify social institutions such as the government, employment organizations, and educational institutions and the degree to which their policies enable or restrict social classes or gender groups to be valued participants in the public environment. Risman (2004) uses gender as a social structure to analyze such gender perceptions within societies. She proposes that the complexities of women’s gender roles become apparent when the researcher investigates how gender role ideology and inequalities are reflected through the social structure; gender issues rise to the level of economy and politics. As individuals, women continuously live and redefine their gender roles; they influence the social structures while the structures provide the constricting framework for gender role variation. Understanding the intersections between agency and structure can bring about the changing mechanisms to recreate gender roles. In this study, looking at the intersections of agency and structure – how women (agency) within a specific society (structure) experience their femininity through the lens of being a woman, worker, wife, mother, and/or caregiver – will reveal by crosscountry comparison women’s attitudes towards their gender and how they mitigate conflicts between home and work spheres. Social structural theory can explain the influential processes at the macro- (social institutions), meso- (social networks), and micro-level (influence of socialized normative behaviors of individuals) that can define a specific country and its inhabitants. In this notion, the division of labor is such a social construct. Depending on how paid and unpaid labor is divided among gender, it reflects the social roles of men and women and, 41 therefore, the gender inequalities within a specific society (Eagly and Wood 1999). Rachel Hare-Mustin (1988) stresses that the research in gender theory has failed to address the complexities of women’s roles in the family and work environment. She points out that gender theory such as sex role theory which has informed some Feminist agendas, has put continuous focus on the differing spheres of men and women. Yet, women in the United States have to juggle between work and family roles. The author argues that gender roles have to be reevaluated for modern times and that the multitude of female roles requires a new kind of approach and theory which accounts for the work and family intersections of women’s lives. Hare-Mustin’s (1988) work suggests a new discourse of action to accommodate the varying and complex roles of women in modern times. The discourse provides a historical and theoretical account which is essential to gain a solid understanding of the varying gender theories used to analyze changing gender roles. Therefore, this research study makes use of an integrative approach by selecting multiple yet complementing theories to account for the complexities of women’s roles. As Hare-Mustin suggests as well, gender roles have to be embedded into a larger socio-cultural context, hence, contextually descriptive macro-level variations among specific countries are used in conjunction with the selected theories. Social structural theory, gender stratification, and gender role theory inform this research study in that they help explain how gender roles and ideology within a specific country affect women’s perception of fairness of the division of household labor and their ability to balance the home and work environment. Change toward equality among men and women is desirable, and social role theory “predicts that in countries where women 42 have a large presence in the paid labor force, this distribution of social roles will lead citizens to perceive men and women on more equal terms and thus to endorse egalitarian attitudes” (Apparala 2003:191; Eagly and Steffen 1984; Eagly and Wood 1999). Macro-level/Socio-structural Variations Crompton (2006) has pointed out that in order to understand how both women and men mediate work-life balance in different countries, the focus of analysis needs to be on “gendered allocation of work within the household, as well as the market” (p. 385) of a specific country under investigation. Crompton and Lyonette (2006) investigated the effects of governmental policies on work-life balance spouses experience in dual-earner households. The study used the 2002 ISSP data set and compared findings in Britain, France, Finland, Norway, and Portugal. In addition to a comparative approach, the authors used the ‘societal’ approach which analyzes specific cultural differences in values, ideologies, and in which kinds of policies are implemented. Their research becomes invaluable for this research study because it provides guidance for conducting comparative social research while using the same ISSP data set. Further, by measuring the actual hours a couple puts into household labor, the degree of equal division of household labor and the direct power relationship between the couple can be analyzed. However, the degree of equality in the division of household labor will not reveal how the couple perceives such division. The individuals’ household needs analysis in a larger country-specific, societal context in order to gain an increased understanding about the perception of fairness of the division of household labor. As Sironi et al. (2010) point out “gender inequalities are embedded in social institutions, 43 which affect the roles individuals are expected to inhabit in those institutions” (p. 5). The division of household labor becomes a reflective indicator of gender inequalities within public institutions in terms of financial resources availability, power displays, and normative values of gendered labor arrangements (Curtis 1986; Katz 1991, Davis 2004, Sironi et al. 2010). The degree to which labor is divided among men and women “are not just labor ‘rational’ and efficient economic strategies, but also take into consideration cultural, moral, ideological and historical implications” (Sironi et al. 2010:5-6; Davis 2004). By using a gendered lens, it can be argued, therefore, that the division of household labor on the individual/micro-level is to a degree a reflection of gender role ideology and gender stratifications at the macro-level. While societal ideology on the division of household labor can certainly have an effect on individuals via early socialization processes and via later economic opportunities to gather financial resources, individual gender role behavior will in turn perpetuate either traditional models of gender roles such as the woman as homemaker and the man as breadwinner or move toward more egalitarian gender roles with a more equitable division of household labor. Apparala et al. (2003) propose that countries which become increasingly egalitarian will presumably have a more equal division of household labor. Therefore, the use of crosscountry comparisons may help in analyzing the “links between household labor and other cultural, institutional, and structural factors” (Coltrane 2000:1218). Yet, Braun et al. (2008) have criticized that limiting the analysis to only a few countries would provide only a partial and at best limited explanation of how societal and individual-level factors contribute to the perception of fairness of the division of 44 household labor. Braun et al. (2008) compared data of 25 countries surveyed in the 2002 ISSP data set to research the degree to which women perceive equity in the division of household labor. The authors use ‘legitimizing principles’ such as the dependence of resources among spouses/cohabiting partners, how much time they have available, and the spouses’ ideology of gender roles. While legitimizing principles might explain the differences in fairness perception of the division of household labor, societal and cultural differences within each country are also assumed to influence how individuals view and evaluate fairness, and manage actual household labor inequality. However, Ruppanner (2008) investigated the fairness levels in the division of household labor perceived among women living in Sweden, the Netherlands, and Hungary by using only a three-country comparison. Her research results proved to be significant on the country-level. She used the 2002 ISSP module in a quantitative approach to analyze how married women manage the home and work environment while putting each country in a short historical context in terms of gender role ideology, women’s labor force participation, and policy implementation. Her research reaffirms that using only a few countries in an analysis can be an effective research approach to analyze the concept of the perception of fairness of the division of household labor. Hence, this quantitative analysis compares only four countries in the OECD by integrating societal factors (country-specific variables as well as descriptive content of the four countries’ political economies) with individual-level variables; it is argued that it may be more efficient to first investigate the hypotheses in a smaller four-country comparison, and later on, expand on the hypotheses in future research studies by using a 45 large-scale country comparison of all countries that participated in the 2002 ISSP. The United States The United States is a liberal welfare state which ranks fairly low among other OECD countries in terms of respondents who perceive the division of household labor as equal. Only 41.0 percent of American women perceived the division of household labor as fair (ISSP 2004; Braun et al. 2008). In the recent 2008 National Study of the Changing Workforce (NSCW), the Families and Work Institute reported that in the United States, 75 percent of employed parents in 2008 felt that they lacked time to spend with their children; this trend increased within sixteen years by 9 percent from 66 percent in 1992 (Bond and Galinsky 2011). The trend also increased from 50 percent in 1992 to 64 percent in 2008 for working couples who felt they lacked time to spend with their partners. The study suggests that there has been a continuous upward trend of United States employees feeling a “Time Famine” (Bond and Galinksy 2011:2) which can also affect how working married couples and unmarried cohabiting partners manage to balance their work and life environments. Compared to Northern European countries, the United States did not have specific laws addressing family and medical leave before 1993. Between 1998 and 2008, there have been no significant changes in maternal and paternal leave policies; in fact, a decline from 27 percent in 1998 to 16 percent in 2008 in providing full-time pay for new mothers who have to temporarily take leave from work due to a birth-related disability has been reported (Galinsky et al. 2008). Waldfogel (1996) indicated in his research that the Family and Medical Leave Act (FMLA) enables approximately half of the employees 46 in the United States to take advantage of family leave. In 1993, the federal FMLA was passed which mandates that male and female employees have the right to take at least twelve weeks of leave for women to give childbirth with job-guarantee upon return to work; the leave is unpaid and is only granted when the employee has worked at least 1,250 hours in the year before taking the leave. However, in a study conducted by the Family and Work Institute, 18 to 21 percent of 1,100 investigated small and large businesses did not comply with the FMLA law (Galinsky et al. 2008). The Institute for Women’s Policy Research (2007) reported that the amount of companies in the United States which offer paid maternity and paternity leave is quite minute, even among companies with the best benefit packages. Thoits (1991) argued that working couples in the United States who have children and do not receive paid family and medical leave may not be able to experience work-life balance. Especially working married women will be affected by the paid leave inaccessibility for household labor and care-giving responsibilities often fall on their shoulders (Thoits 1991). In addition, couples who work in low-income jobs frequently do not take unpaid leave which FMLA provides because they cannot afford the lack of income during the leave (Waldfogel 2001). This situation leads to the argument that FMLA is not effective as a family-friendly policy for population groups that work in lowincome jobs (McDowell et al. 2005). Evans (2002) pointed out that women who work for smaller companies are likely to have less income and no benefit packages which may be due to larger societal gender stratifications within the job environment. Also, women are disadvantaged because they often have no choice but to take unpaid leave to give birth. 47 After childbirth, some women will choose to move from full-time into part-time employment to bring more balance into their work-life (Orloff 2009). Yet, part-time employment is rarely paid well, secure, or does not offer medical and retirement benefit packages (Evans 2002). However, with the Obama administration, new efforts have been made to support women in particular with specific tax relief measures (White House 2010). President Barack Obama has put forth policy measures which extend tax benefits such as the Earned Income Tax Credit (EITC) as well as the Child Tax Credit (CTC) that are geared to help especially working single and married mothers. Extending these tax credits to working families will help approximately 60 percent of working mothers to take care of their children and families, according to White House assessments. The White House estimates that about four million single-mother households benefit from the tax credit extensions. Germany Germany belongs to the countries that have traditionally embraced the male breadwinner model, with the man being responsible for the income and the woman taking on the household and care-giving responsibilities (Wierda et al. 2009). In this notion, Germany is also categorized as a fairly traditional welfare state because men and women have less roles to act upon: the man will engage in the worker role and less in the houseman role, while the woman is more likely to be a housewife and only to a limited degree a full-time worker (Wierda 2009). In a study of British and German households and the use of time, Gershuny (1995b) concluded that in both countries, couples began to 48 rearrange the division of household labor toward more equality when the women became employed: men increased their household labor when women moved from staying at home to working or moving from part-time to full-time work. Yet, overall, Germans had fairly traditional gender views in 2002: according to the 2002 ISSP, every fourth respondent in Germany thought it was a man’s duty to work, and a woman’s duty to manage the household and care-giving responsibilities (Wierda 2009; ISSP 2004). Only 42 percent of German women perceived equity in the division of household labor (ISSP 2004; Braun et al 2998). Also, over 50 percent of Germans believed that mothers needed to stay home when they had children of pre-school age. For the reason that many Germans seem to adhere to a more traditional gender ideology, it could be argued that the unequal division of household labor reflects the countries traditional ideology. If both partners have more traditional ideologies, women might engage in more unpaid and less paid work (while having less resources to negotiate the division of household labor); therefore, they may perceive the division of household labor as increasingly fair, helping them to experience more work-life balance. The Netherlands Traditionally, the Netherlands is a conservative welfare state, which has also relied on the male breadwinner model; yet, it is a country that exhibits more gender equality similar to Finland and Sweden (Wierda et al. 2009). Therefore, Dutch men and women engage in multiple roles such as housewife/man and worker in fairly equal manner: the government has promoted the equal division of household labor (Duyvendak and Stavenuiter 2004). This is not quite reflected in the 2002 ISSP data which indicates 49 that only 31.6 percent of Dutch women perceive fairness in the division of household labor (ISSP 2004; Braun 2008). The lower household labor equity score, however, could possibly be an indication that Dutch women – with a more egalitarian mindset and/or who have internalized the governmental support toward more egalitarian rights – are increasingly voicing their dissatisfaction with the unequal division of household labor; consequently, Dutch women would report lower scores in household labor equity perception. In addition, Dutch men and women are often employed part-time, for parttime work in the Netherlands is considered a legal right, making the Netherlands a parttime society (Keuzenkamp and Hooghiemstra 2000; Wierda 2009; Treas and Widmer 2000; Ruppanner 2008). Also, the legal right to part-time work has helped facilitate the positive change in attitude toward mothers who work (Portegijs et al. 2006; Wierda 2009). When Dutch mothers were surveyed in the 2002 ISSP, only 29 percent of respondents agreed that mothers should stay home if they had preschool-aged children. Portugal Portugal is considered a familistic welfare state (Esping-Andersen 1999; Crompton and Lyonette 2006). In 2008, 62.5 percent of Portuguese women were in the labor force, which was 4.7 percent higher than the OECD employment rate for women of 57.8 percent. Many women are employed full-time, and among dual-earner households, 67 percent of the households have couples who both are employed full-time (Crompton and Lyonette 2006). Before 1974, the corporatist regime under Prime Minister António de Oliveira Salazar deemed wives to be under the legal directive of husbands and responsible for all household labor and care-giving activities; they were also prohibited 50 from being employed in many vocational jobs (Crompton and Lyonette 2006). Due to Portugal being a country that is historically marked by male emigration, many women had to enter the work force to supplement family income and prevent the household from slipping into poverty. Yet, to this day, the government does not provide sufficient welfare support, making the family’s well-being dependent on the income of relatives and spouses. The government does not provide policies that help women in the work environment. Although there have been a few policy changes, childcare is only minimally offered by the state while employers provide fairly decent levels of childcare support. Portugal also has one of the highest equity perceptions of the division of household labor among women with 63.2 percent (ISSP 2004; Braun et al. 2009). Yet, women in Portugal spend about 20 hours on household labor (Crompton and Lyonette 2006). Crompton and Lyonette (2006) point out that in Portugal “there is no evidence of state policies that explicitly encourage men to take on a larger share of domestic work” (p. 383). This research study attempts to analyze by means of role-strain and role-spillover theories the mediating factors for work-life balance. With special focus on social structural theory, gender stratification, sex role theory, and the country-specific socioeconomic, historical, and gender ideological conditions, this research proposal intends to dissect the various factors – especially the perception of fairness of the division of household labor and the legitimizing principles – which influence work-life balance. 51 CHAPTER 3 RESEARCH DESIGN AND METHODOLOGY Hypotheses (Figure 7, p. 53) Hypothesis HIA – Gender role ideology is significantly correlated to work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Mediating Hypothesis HIB – The perception of fairness of the division of household labor mediates the degree to which gender role ideology affects work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Hypothesis H2A – Time availability is significantly correlated to work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Mediating Hypothesis H2B – The perception of fairness of the division of household labor mediates the degree to which time availability affects work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Mediating Hypothesis H3 – The perception of fairness of the division of household labor mediates the degree to which resource dependence affects work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Resource dependence is positively correlated with work-life balance. 52 Hypothesis H4A – Respondent’s weekly household labor hours are significantly correlated to work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Mediating Hypothesis H4B – Respondent’s weekly household labor hours mediate the effect of gender role ideology on work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Hypothesis H5A – Spouse’s/cohabiting partner’s weekly household labor hours are significantly correlated to work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Mediating Hypothesis H5B – Spouse’s/cohabiting partner’s weekly household labor hours mediate the effect of gender role ideology on work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. 53 Figure 7: Concept map 54 Data Source The research study builds on quantitative research designs and theoretical approaches proposed from both work-life and household labor research in an attempt to reduce research gaps by focusing on the underutilized yet critical variable of perception of fairness of the division of household labor; this approach creates an alternative method for the investigation of work-life balance. The quantitative analyses include Analysis of Variance (ANOVA), multiple regression, and Process for path analysis-based mediation computation in combination with Hayes’s ‘Conditional Process Modeling’ (Hayes 2013). The study relies on a quantitative research approach by using the data set of the 2002 International Social Survey Programme: Family and Changing Gender Roles III (ISSP 2004). The 2002 ISSP data set is used to analyze and compare cross-culturally the degree to which women in particular experience work-life balance in the United States, Germany, the Netherlands, and Portugal. The data set is an ideal source for scientific inquiry because it is based on respondents’ attitudes toward a variety of topics concerning the home environment. Created in 1983, the program is a continuous and annual research effort with the intent to collect relevant scientific data on various social topics by either adding a supplemental questionnaire to an existing national survey or by conducting a separate survey in each participating country. For almost thirty years, the ISSP has been made possible by the global collaboration of about 40 countries, among them the United States, Germany, the Netherlands, and Portugal. In the 2002 ISSP, 34 countries participated. Each participating country independently funds and collects the data by sampling methods ranging from 55 simple random, systematic random, and clustered, to stratified random sampling, as well as by either in-person interviews, questionnaires by mail, or self-enumerated questionnaires (ISSP 2004). After each country has conducted and processed the survey data, the complete data set from all participating countries is produced and archived by the Zentralarchiv für Empirische Sozialforschung at the Universität Köln. The first complete edition of the 2002 ISSP was published September 2004. The Inter-university Consortium for Political and Social Research (ICPSR) at the University of Michigan is responsible for the distribution of the complete ISSP data set; the 2002 ISSP data was made available in the first ICPSR version, number 4106, November 2004. In Germany, the 2002 ISSP data was administered and analyzed by INFAS, the Institut für angewandte Sozialwissenschaft GMBH (ISSP 2004). The data was gathered between February and August of 2002 by supplementing ALLBUS, the German General Social Survey with a self-guided 2001/2002 ISSP questionnaire in addition to an interviewer asking questions in person about demographic variables. The sample size consisted of 1,367 received and eligible responses, with 936 responses from West Germany and 431 responses from East Germany. The data set was not weighted because the sample from East Germany was over-sampled. In the Netherlands, NIPO was the fieldwork institute in charge of the 2002 ISSP data collection and analysis (ISSP 2004). By method of cluster sampling of addresses, the surveys were sent out and collected between October 2002 and January 2003 and consisted of two self-administered questionnaires. However, in some cases, NIPO used interviewers who would bring by and collect the responses themselves. The sampling size 56 amounted to 1,268 eligible responses, and the sample was not weighted (ISSP 2004). The fieldwork institute Motivacao-Estudos Psico-sociologicos was in charge of collecting and analyzing the 2002 ISSP data for Portugal (ISSP 2004). Fieldwork was conducted between February and July 2003. Using stratified random probability sampling, the survey consisted of interviews conducted in person with the use of visual aids. The sample size amounted to 1,092 responses, but the sample had to be weighted because the rate of non-responders and people who refused to answer the questionnaire was too high for an unweighted sample. The demographic variables skewing the sample consisted of gender, age, and education. In order to collect 2002 ISSP data for the United States, the National Opinion Research Center (NORC) used a supplemental questionnaire at the end of the 2002 General Social Survey (ISSP 2004). Survey addresses were determined by using multistage area probability sampling; the survey fieldwork lasted from February 2002 to June 2002, and was self-administered after the interviewer ended the questioning portion of the General Social Survey. The sample size consisted of 1,171 responses and was not weighted because men were under-sampled. Variable Measures Dependent Variable Work-Life Balance: The dependent variable is Work-Life Balance. The variable was created by merging and scaling four variables from the 2002 ISSP data set. The variable was recoded to provide the most valid (at least one valid response on the four questions) mean values regardless of the number of missing cases. Higher means 57 represent greater work-life balance. The four variables used were V48 Too tired from work to do duties at home: In the past three months it has happened that I have come home from work too tired to do the chores which need to be done, V49 Difficult to fulfill fam. [family] responsibility: In the past three months it has happened that it has been difficult for me to fulfill my family responsibilities because of the amount of time I spent on my job, V50 Too tired from hhwork [householdwork] to function i [in] job: In the past three months it has happened that I have arrived at work too tired to function well because of the household work I had done, V51 Difficult to concentrate at work: In the past three months it has happened that I have found it difficult to concentrate at work because of my family responsibilities. The choices of answers were (1) Several times a week, (2) Several times a month, (3) Once or twice a year, or (4) Never. Lower scores (answers 1 and 2) indicate less work-life balance while higher scores (answers 3 and 4) indicate more work-life balance. Mediator Variables Perception of Fairness of the Division of Household Labor: The mediator variable is Benefit Distribution in the Division of Household Labor. In the 2002 ISSP, the variable V38 Sharing of hh [household] work between the partners was used by asking respondents the question “Which of the following best applies to the sharing of household work between you and your spouse/partner.” Response categories were (1) I do much more than my fair share of the household work, (2) I do a bit more than my fair share of the household work, (3) I do roughly my fair share of the household work, (4) I do a bit less than my fair share of the household work, (5) I do much less than my fair 58 share of the household work. The response categories do not reflect a variable that is a linear measure of fairness: the response score of 3 indicates a fair distribution of the division of household labor in which neither respondent nor spouse/cohabiting partner under- or over-benefit from the division of household labor. Therefore, the division of household labor is perceived as fair and equal. The lower scores of 1 and 2 signify that the respondent under-benefits from the division of household labor: he/she does more than his/her fair share of household labor. The higher response scores of 4 and 5 indicate that the respondent over-benefits from the division of household labor; therefore, engaging in less than his/her fair share of household labor. Previous researchers such as Ruppanner (2008) and Braun et al. (2008) used the same variable to confer the concept of perception of fairness of the division of household labor. Although the response categories of Benefit Distribution in the Division of Household Labor are interpreted in terms of a respondent under- or over-benefitting from the division of household labor, the variable is clearly used in this study to signify that the concept of the perception of fairness of the division of household labor is under investigation as a mediating factor in work-life balance. Actual Hours of Household Labor for Respondent or Spouse/Cohabiting Partner Respondent’s Weekly Household Labor Hours: The variable V36 How many hours do you spend on hh [household] work was used; respondents were asked: “On average, how many hours a week do you personally spend on household work, not including childcare and leisure time activities?” The categories were (01) 1 hour or less than 1 hour through (95) 95 hours and more (in the United States: 96 hours and more). 59 Spouse’s/Cohabiting Partner’s Weekly Household Labor Hours: The variable V37 How many hrs [hours] spouse, partner works on hh [household] was used; respondents were asked: “And what about your spouse/partner? On average, how many hours a week does she/he personally spend on household work, not including childcare and leisure time activities?” The categories were (01) 1 hour or less than 1 hour through (95) 95 hours and more (in the United States: 96 hours and more). Independent Variables Legitimizing Principles: Gender Role Ideology, Time Availability, and Resource Dependence Gender Role Ideology: The variable V11 Men’s job is work, women’s job household was used and renamed Gender Role Ideology. Respondents were asked the question “A man’s job is to earn money; a woman’s job is to look after the home and family.” The responses were (1) Strongly agree, (2) Agree, (3) Neither agree nor disagree, (4) Disagree, (5) Strongly disagree. Time Availability: The variable Time Availability is based on the variable V240 ˂WRKHR ˃ R: Hours worked weekly. Respondents were asked about the “Hours worked weekly.” In Germany, the question was asked “How many hours per week do you normally work in your main job, including overtime?” In the United States, the question was asked “How many hours did you work last week, how many hours do you usually work a week, at all jobs?” The responses were coded as (01) one hour to (96) 96 hours or more. Resource Dependence: The variable V43 Who has the higher income? was used. 60 Respondents were asked “Considering all source of income, between you and your spouse/partner, who has the higher income?” The response categories for the variable Resource Dependence are (1) My spouse/partner has no income, (2) I have a much higher income, (3) I have a higher income, (4) We have about the same income (5) My spouse/partner has a higher income, (6) My spouse/partner has a much higher income, (7) I have no income. The response category of 1 indicates no resource dependence and the response category of 7 indicates complete resource dependence of the respondent. Control Variables Gender Being Female: The variable V200 ˂SEX˃ R: Sex was used with the respondent being asked about his/her sex. The response categories were recoded into (0) Male and (1) Female. Marital Status Not Married/Cohabiting Partner and Not Married/Not Cohabiting Single For an appropriate sample size, all married respondents, cohabiting partners, and non-cohabiting respondents without a partner were selected from the 2002 ISSP data set. The variable V203 <COHAB> R: Steady life-partner was used; respondents were asked “(If married) Do you have a steady life partner? (If not married) Do you live together with a steady partner?” Response categories were (1) Yes (2) No (9) NA, refused (0) NAP (married, living together with spouse). Category (0) was reinstated with an explicit code and became the reference category of Married and Cohabiting for two dummy variables, Not Married/Cohabiting Partner and Not Married/Not Cohabiting Single. 61 Not Married/Cohabiting Partner: The variable Not Married/Cohabiting Partner had the response categories of (0) Married and cohabiting and (1) Not married, but cohabiting partner. Not Married/Not Cohabiting Single: The variable Not Married/Not Cohabiting Single had the response categories of (0) Married and cohabiting and (1) Not married and not cohabiting single. Age Age: The variable V201 ˂AGE˃ R: Age was used; the question was asked about “Age of Respondent.” The categories were (16) 16 years of age through (94) 94 years of age. The variable Age was not recoded to eliminate ages 66 and older. In this study, respondents of older ages (past 65 years of age) are viewed as a valuable pool of respondents because increasing numbers of older people remain active in the work force, and therefore, need to manage work-life balance. Education Years of Education: The variable V204 ˂EDUCYRS˃ R: Education I: years in school was used; respondents were asked about their “Education I: Years in school.” In Germany, the respondents were asked “How many years from school to university education but without vocational training?” In the Netherlands, the respondents were asked “Years of schooling following R’s [Respondent’s] sixth birthday included vocational training, part time courses excluded.” In the U.S., respondents were asked “What is the highest grade in elementary school or high school that you finished or got credit for?” The categories were (06) 1 year or less than 1 year to 6 years (7) 7 years and 62 consecutively through (19) 19 years or more. Number of Children by Age Categories Number of Children (0 to 5, 6) and Number of Children (6, 7 to 17) Number of Children (0 to 5, 6): The variable V67 Number of people in hh [household]: kids up to 5, 6 was used; respondents were asked “And what about children up to the age of 5/6 years (depending on the start of compulsory schooling in respondent’s country?” After reinstating the missing category of (00) No children – due to the high number of cases that belonged to the category – into a valid category of (0), the categories were recoded into (0) No children and consecutive numbers up to (10) 10 children. Number of Children (6, 7 to 17): The variable V66 How many people in hh [household]: 6, 7 – 17 yrs [years] was used; the respondents were asked “And what about children between 6/7 (depending on the start of compulsory schooling in respondent’s country) and 17 years of age.” In Germany, the age cut-off range was 6-17 years, while in the Netherlands the age cut-off range was 4-17 years. After reinstating the missing category of (00) – due to the high number of cases – into a valid category of (0), the categories were (0) No children, followed by consecutive ages (01) 1 child through (10) 10 children. Employment Status Part-time Employment, Housewife/man, and Other Employment Status To ensure an appropriate sample size, all full-time, part-time employed, housewife/man categories, and other employment status categories were pre-selected 63 from the original employment variable in the ISSP data set. Of these categories, three dummy variables were created with full-time employment as the reference category: Part-time Employment: The dummy variable is based on the original variable V239 ˂WRKST˃ R: Current employment status. Respondents were asked about “Respondent: Current employment status – current economic position, main source of living.” The response categories were (1) Full-time employed, main job (in the Netherlands more than 35 hours), (2) Part-time employed, main job (in the Netherlands 15 to 35 hours), (3) Less than part-time (in Germany, the Netherlands, and Portugal less than 15 hours, in the United States temporarily not working), (4) Helping family member, (5) Unemployed, (6) Student, (7) Retired, (8) Housewife (man), home duties, (9) Permanently disabled, (10) Other, not labor force (in Germany army/civil service). The variable Part-time was recoded into response categories of (0) Full-time employment and (1) Part-time employment. Housewife/man: The variable Housewife/man was recoded into the response categories of (0) Full-time employment and (1) Housewife/man. Other Employment Status: From the original employment variable, categories (3) through (7) and (10) were combined to create the response category (1) Other employment status. The additional response category of (0) was created for the reference category of full-time employment. The dummy variable for Other Employment Status had to be included in order to maintain an acceptable sample size for all employment dummy variables. 64 Income Income: For the variable Income, the original variable V249 <RINCOME> R: Earnings was used; the question was asked about “Respondent’s earnings.” In Germany, the “monthly net income after the deduction of taxes and social insurance in Euro” was used while in Portugal the “R’s [Respondent’s] monthly average net income in Euro (midpoints)” was used as the income recording standard. In the United States, the “R’s [Respondent’s] earnings from all jobs in 2000 before taxes or other deductions in $ (midpoints)” was the income recording guideline. The income variable was recoded into group mean centered values for comparative use across the four countries. However, after running preliminary analysis, the income variable had to be excluded as a control variable because its integration resulted in reducing the overall sample size due to its large number of missing cases. Continued integration of the variable would have led to biasing the analytical results. Religious Main Groups Roman Catholic, Protestant, and Other Religion The variable V289 <RELIGGRP> R [Respondent]: Religious main groups (derived) was used to create three dummy variables for Roman Catholic, Protestant, and Other Religion. For each dummy variable, the reference category is no religion. The variable V289 <RELIGGRP> R [Respondent]: Religious main groups (derived) was derived from the variable V288 <RELIG> R [Respondent]: Religious Denomination. In Germany, the question for variable V288 was asked “Which religious group do you belong to?” In the Netherlands, the question was asked “Do you consider yourself to 65 belong to a religious group or church? If yes, which of those groups or churches do you consider yourself to be a member of?” In the United States, the question was asked “What is your religious preference? Is it Protestant, Catholic, Jewish, some other religion, or no religion (If Protestant) What specific denomination is that?” Roman Catholic: The variable Roman Catholic was recoded into response categories (0) No religion and (1) Roman Catholic. Protestant: The variable Protestant was recoded into response categories (0) No religion and (1) Protestant. Other Religion: After running an initial crosstabulation of the variable V289 for the four countries, it was determined that other religious denominations such as Christian Orthodox, Jewish, and Islam had to be recoded into the dummy variable Other Religion to avoid a lack of sufficient cases. The response categories of Other Religion are (0) No religion and (1) Other religion. Macro-/Country-level Variables Germany, the Netherlands, and Portugal The macro/country-level dummy variables were selected from the country variable V3: Country with the response categories of (02) Germany (West) (D-W), (03) Germany (East) (D-E), (06) United States (USA), (11) Netherlands (NL), and (30) Portugal (P). When the three dummy variables for Germany, the Netherlands, and Portugal were recoded, the United States became the reference category. Germany: The dummy variable Germany was recoded with the response categories of (0) The United States and (1) Germany. The response category of (1) 66 combines all cases for West and former East Germany. The Netherlands: The response categories of The Netherlands were (0) The United States and (1) The Netherlands. Portugal: The response categories of the dummy variable Portugal were recoded into (0) The United States and (1) Portugal. 67 Chapter 4 RESULTS AND INTEPRETATION In this chapter, the following tables highlighting ANOVA, regression, and Process results will be presented and discussed. Interpretation will focus especially on mean differences between countries and gender differences within countries to explore in greater detail the socio-structural dynamics and gender role interactions in the United States, the Netherlands, Germany, and Portugal. Mean differences are illustrated for the dependent variable work-life balance and all mediator and independent variables. Regression results for the independent variables of gender role ideology, time availability, respondent’s weekly household labor hours, spouse’s/cohabiting partner’s weekly household labor hours and control variables are analyzed for significance levels and interpreted accordingly for their effects on work-life balance. For Process results, special attention is given to the mediator variables perception of fairness of the division of household labor, respondent’s weekly household labor hours, and spouse’s/cohabiting partner’s weekly household labor hours. The statistical analyses are based solely on the data set of the 2002 International Social Survey Programme: Family and Changing Gender Roles III (ISSP 2004). Table 1 (p. 68) illustrates that each of the four countries has an effective sample size of at least 1,000 respondents (male and female respondents combined) who are either married with a cohabiting spouse, not married with a cohabiting partner, or single (total n of all four countries: n=4,873, United States: n=1,171, Germany: n=1,364, the 68 Netherlands: n=1,249, Portugal: n=1,089). Respondents who are single (not married/not cohabiting) were included in the analysis by the dummy variable Not Married/Not Cohabiting Single to contrast the reference category of married/cohabiting respondents. Table 1: Total number of respondents for the United States, Germany, the Netherlands, and Portugal, by gender United States Germany The Netherlands Portugal Count per country (n=4,873 for total of four countries ) 1,171 1,363 1,249 1,089 Female (%) 680 (58.1%) 693 (50.8%) 645 (51.6%) 647 (59.4%) Male (%) 491 (41.9%) 671 (49.2%) 604 (48.4%) 442 (40.6%) Mean Values and ANOVA Results Work-Life Balance Table 2 (p. 69) provides total mean values of work-life balance for all respondents by country affiliation while Table 3 (p. 69) highlights mean results from independent samples t-tests for equality of means which were conducted to determine gender differences within each country for work-life balance. Only the USA and Portugal had statistically significantly different mean values for work-life balance among men and women (Table 3, p. 69). In the USA, women had little work-life balance averaging a mean of 2.99 (SD=0.74) while men had more work-life balance averaging a mean of 3.09 (SD=0.69). In Portugal, the significant mean difference was even higher between men and women with men experiencing more work-life balance averaging 3.22 (SD=0.66) while women had little work-life balance by averaging the lowest mean value of 2.98 69 (SD=0.74). Table 2: Total mean values of work-life balance for all respondents by country affiliation Germany Mean Work-Life Balance Gender Role Ideology Time Availability Resource Dependence Perception of Fairness of the Div. of Hh. Labor R’s Weekly Hh. Labor Hrs. Spouse’s/Cohabiting Partner’s Weekly Hh. Labor Hrs. USA Mean 3.21 Std. Dev. .614 3.65 The Netherlands Mean 3.03 Std. Dev. .715 1.203 3.72 42.11 11.875 3.77 Portugal Mean 3.27 Std. Dev. .607 3.08 Std. Dev. .713 1.439 3.77 1.004 3.42 1.274 41.38 14.531 30.70 13.437 42.37 13.113 1.728 3.81 1.697 3.91 1.862 4.02 1.656 2.83 1.166 2.62 1.100 2.76 1.278 2.92 1.084 14.07 12.299 12.42 15.257 13.33 15.966 26.55 26.767 15.47 15.407 17.31 26.193 15.32 19.235 36.41 36.232 Table 3: Independent samples t-test results with significant mean values for respondents by gender and country affiliation Germany Work-Life Balance Gender Role Ideology USA Netherlands Portugal Male 3.19 Mean (SD) Female 3.24 Mean (SD) Male Female 3.09 2.99* Mean (SD Male Female 3.30 3.24 Mean (SD) Male Female 3.22 2.98*** (.62) (.61) (.69) (.74) (.60) (.61) (.66) (.74) 3.54 3.77*** 3.48 3.90*** 3.67 3.87*** 3.34 3.49*** (1.19) (1.21) (1.44) (1.42) (1.00) (1.00) (1.25) (1.29) 70 Time Availability 45.78 37.16*** 44.70* 38.37*** 36.12 24.46*** 45.05 40.07*** (10.46) (11.89) (14.60) (13.81) (12.51) (11.66) (13.34) (12.48) Resource Dependence 2.73 4.83*** 2.88 4.50*** 2.49 5.37*** 2.85 4.90*** (1.19) (1.54) (1.31) (1.62) (1.16) (1.19) (1.13) (1.43) Perception of Fairness of the Div. of Hh. Labor 3.60 2.09*** 3.21 2.18*** 3.56 1.96*** 3.60 2.45*** (.89) (.89) (.97) (.98) (1.02) (.97) (.88) (.95) R’s Weekly Hh. Labor Hrs. 7.50 20.01*** 11.65 12.99 10.47 16.22*** 25.53 27.31 (6.37) (13.31) (18.45) (12.34) (18.34) (12.50) (35.14) (18.10) Spouse’s/ Cohabiting Partner’s Weekly Hh. Labor Hrs. 21.17 9.17*** 18.35 16.53 18.38 12.21*** 29.63 41.52*** (14.73) (13.59) (20.93) (29.51) (12.63) (23.77) (22.55) (43.12) *p<.05; **p<.01; ***p<.001 (two-tailed tests) Table 4 (p. 71) highlights the results from the one-way Analysis of Variance (ANOVA) with Scheffe post-hoc test for work-life balance by country affiliation suggesting that country affiliation poses a measurable influence on respondents’ worklife balance. There was a statistically significant difference between countries as determined by one-way ANOVA (F(3, 3,130) = 22.466, p<.001). The Scheffe post-hoc test (Table 4, p. 71) revealed that work-life balance was statistically significantly higher for respondents living in Germany (3.215±.614) compared to respondents living in each, the USA (3.039±.715, p<.001) and Portugal (3.088±.713min, p<.01) (Table 2, p. 69). There were no statistically significant differences between Germany and the Netherlands 71 (3.274±.607, p<.347) or between the USA and Portugal (p<.596) (Table 2, p. 69). Table 4: ANOVA and significant Scheffe post-hoc tests results for dependent and independent variables Dependent and Independent Variables Work-Life Balance Gender Role Ideology Time Availability Resource Dependence Perception of Fairness of the Div. of Hh. Labor R’s Weekly Hh. Labor Hrs. Spouse’s/Cohabiting Partner’s Weekly Hh. Labor Hrs. Mean Difference (I-J) USA .17594*** Portugal .12680** USA Netherlands -.23452*** Netherlands Portugal .18538*** Germany Portugal .23000*** USA Portugal .29891*** Netherlands Portugal .34552*** Germany Netherlands 11.415*** USA Netherlands 10.680*** Netherlands Portugal -11.670*** Germany Not significant for mean differences Germany USA .21503** USA Portugal -.30768*** Germany Portugal -12.476*** USA Portugal -14.126*** Netherlands Portugal -13.217*** Germany Portugal -20.943*** 72 USA Portugal -19.108*** Netherlands Portugal -21.100*** *p<.05; **p<.01;***p<.001 (two-tailed tests) Gender Role Ideology Compared to Portugal (3.428±1.274) with neither a too traditional nor too egalitarian gender role ideology score, each of the countries Germany (3.658±1.203, p<.001), the Netherlands (3.774±1.004, p<.001), and the USA (3.727±1.439, p<.001) had significantly higher mean values signifying more egalitarian gender role ideologies (Tables 4/p. 71 and 2/p. 69). Independent samples t-tests showed that all countries had statistically significantly different mean values for male and female respondents’ gender role ideology (Table 3, p. 69). Women of all four countries had more egalitarian gender role ideologies than the male respondents. The USA (t(1,155)=-4.961, p<.001) had the highest mean difference between male and female respondents with men averaging 3.48 (SD=1.44) and women averaging 3.90 (SD=1.42). Portugal (t(1,080)=-1.966, p<.050) was the country with the smallest mean difference between male and female respondents for gender role ideology. Portuguese men also had the lowest average mean of 3.34 (SD=1.25) while the women averaged a mean of 3.49 (SD=1.29); both means signify neither a traditional nor egalitarian gender role affiliation. Yet, Portugal is a “familistic welfare state” (Crompton and Lyonette 2006:382) which has a history of having a more traditional gender role ideology. Perception of Fairness of the Division of Household Labor One-way Analysis of Variance (ANOVA) with Scheffe post-hoc test for the perception of fairness of the division of household labor by country affiliation suggests 73 that country affiliation is also a factor in influencing the perception of fairness of the division of household labor, but only for Germany, the USA, and Portugal (Table 4, p. 71). There was a statistically significant difference between the three countries as determined by one-way ANOVA (F(3, 3,022) = 8.245, p<.001). The Scheffe post-hoc test revealed that the perception of fairness of the division of household labor was statistically significantly higher for Germany (2.837±1.166) compared to the USA (2.622±1.100, p<.01) and lower for the USA (2.622±1.100) compared to Portugal (2.929±1.084, p<.001) (Table 2, p. 69). After conducting independent samples t-tests for the perception of fairness of the division of household labor by country and gender, all countries showed statistically significant mean differences between male and female respondents (Table 3, p. 69). All mean calculations by gender affiliation show that men throughout all four countries tended to perceive themselves as engaging in roughly the same amount of household labor or slightly over-benefitting from the division of household labor than their female counterparts. The largest mean difference for gender affiliation appears in the Netherlands (t(799)=22.746, p<.001) with women having the lowest mean value with an average of 1.96 (SD=0.97) for under-benefitting from the division of household labor, therefore, doing more than what they perceive as their fair share of household labor. Due to the Netherlands being a country with a more egalitarian gender role ideology, it can be argued that Dutch women will be more aware about the distribution of the division of household labor, and will more likely criticize and report any unequal division of household labor. Dutch men averaged a mean value of 3.56 (SD=1.02) for doing less household labor, therefore, slightly over-benefitting from the 74 division of household labor. The smallest mean difference between men and women appears in the USA (t(629)=13.705, p<.001) with women having a mean value of 2.18 (SD=0.98) for under-benefitting from the division of household labor while men perceive to be doing roughly their fair share of household labor – neither over- nor underbenefitting from the division of household labor – with an average mean value of 3.21 (SD=0.97). Time Availability In Table 4 (p. 71), one-way Analysis of Variance (ANOVA) (F(3, 3,094) = 144.477, p<.001) with Scheffe post-hoc test for time availability (for the respondent’s hours worked weekly) by country affiliation pointed to statistically significant differences between the four countries. A Scheffe post-hoc test revealed that the hours worked weekly by the respondent was statistically significantly higher for each, Germany (42.11±11.875, p<.001), the USA (41.38±14.531, p<.001) and Portugal (42.37±13.113, p<.001) compared to the Netherlands (30.70±13.437) (Table 2, p. 69). After conducting independent samples t-tests for the respondent’s hours worked weekly by country and gender (Table 3, p. 69), all countries showed statistically significant mean differences between male and female respondents. All mean calculations by gender affiliation show that men throughout all four countries tend to work more hours weekly than their female counterparts. The largest mean difference for gender affiliation appears in the Netherlands (t(809)=13.771, p<.001) with women working weekly an average of 24.46 (SD=11.66) hours – which is considered part-time in the Netherlands – while Dutch men work weekly an average of 36.12 (SD=12.51) hours – which is considered full-time work 75 status in the Netherlands. Dutch men and women work the least per week with an average of 30.7 hours weekly (SD=13.43) compared to any male and female respondents of the other three countries (Table 2, p. 69). According to Table 2, their combined mean average for weekly work hours is considered part-time. The smallest mean difference of weekly work hours between men and women appears in Portugal (t(799)=5. 555, p<.001) with women having a mean value of 40.07 (SD=12.48) hours while men work a weekly average of 45.05 (SD=13.34) hours (Tables 4/p. 71 and 3/p. 69). Portuguese women work the most per week, on average full-time, compared to any other women of the other three countries. Portugal is considered a “familistic welfare state” (Crompton and Lyonette 2006:382) in which family members are supposed to take care of each other within the family. Also, Portugal has been a country of male emigration, therefore, making it necessary for women to increasingly work in order to fill needed positions and to help the family stay out of poverty. Due to the government not providing enough social welfare support, it becomes even more necessary for women to participate in employment. Resource Dependence One-way Analysis of Variance (ANOVA) for resource dependence within marriage or partnership (for income distribution between spouses or cohabiting partners) initially showed statistically significant differences among the four countries with the ANOVA result (F(3, 2,998) = 2.857, p<.05) (Table 4, p. 69). The Scheffe post-hoc test, however, revealed that the mean differences for resource dependence were not statistically significantly different (p>0.05) for the four countries. 76 Respondent’s Weekly Household Labor Hours Further one-way Analysis of Variance (ANOVA) with Scheffe post-hoc tests was conducted for the R’s Weekly Hh. Labor Hrs. (for hours the respondent spends weekly on household labor), and the results showed statistically significant differences between the four countries as determined by one-way ANOVA (F(3, 3,003) = 92.935, p<.001) (Table 4, p. 71). A Scheffe post-hoc test revealed that the hours the respondent spent weekly on household labor were statistically significantly lower for each, Germany (14.07±12.299, p<.001), the USA (12.42±15.257, p<.001) and the Netherlands (13.33±15.966) compared to Portugal (26.55±26.767, p<.001) (Table 2, p. 69). The results may reflect that in Germany, the USA, and the Netherlands there may be greater egalitarianism in gender roles than in Portugal. More societal egalitarianism would suggest that women engage in less and men in more weekly household labor. After conducting independent samples ttests for the respondent’s hours spent weekly on household labor by country and gender (Table 3, p. 69), however, only Germany and the Netherlands showed statistically significant mean differences between male and female respondents. All mean calculations by gender affiliation show that men throughout all four countries tended to spend less hours weekly on household labor than their female counterparts. The largest mean difference for gender affiliation appears in Germany (t(684)=-18.152, p<.001) with women spending an average of 20.01 (SD=13.31) hours on household labor while German men spend an average of 7.50 (SD=6.37) hours on household labor. This confirms to a larger degree that although Germans spend on average less time on household labor than men and women in Portugal, Germany embraces the male 77 breadwinner model which assigns the man the duty to provide income and the woman the duty of taking care of the home environment. Therefore, German men and women will still engage in an unequal division of household labor, with a woman engaging in almost four times more the amount a man spends on household labor. Female respondents in the Netherlands (t(797)=-5.170, p<.001) spend an average of 16.22 (SD=12.50) hours on household labor, while male respondents spend an average of 10.47 (SD=18.34) hours on household labor. Although the Netherlands is a fairly egalitarian country which has actively promoted gender equality, the historic male breadwinner model may continuously exert influence in the home environment when men and women engage in household labor: Dutch men still participate in less household labor than women. Spouse’s/Cohabiting Partner’s Weekly Household Labor Hours After running the one-way Analysis of Variance (ANOVA) with Scheffe post-hoc tests for the Spouse’s/Cohabiting Partner’s Weekly Hh. Labor Hrs., the results for the ANOVA (F(3, 2,966) = 116.081, p<.001) also showed statistically significant differences between the four countries (Table 4, p. 71). The Scheffe post-hoc test revealed that the hours the spouse or cohabiting partner spent weekly on household labor were statistically significantly higher for each, Germany (15.47±15.407, p<.001), the USA (17.31±26.193, p<.001) and the Netherlands (15.32±19.235, p<.001) compared to Portugal (36.41±36.232) (Table 2, p. 69). After conducting independent samples t-tests for the respondent’s spouse’s or cohabiting partner’s hours spent weekly on household labor by country and gender (Table 3, p. 69), Germany, the Netherlands, and Portugal showed statistically significant mean differences between male and female respondents. Mean 78 calculations by gender affiliation showed that female respondents in Germany and the Netherlands reported that their spouses or cohabiting partners tended to spend less hours on household labor than their female counterparts, while female respondents in Portugal estimated that their spouses or cohabiting partners spent much more weekly hours on household labor than the women themselves. The largest mean difference for gender affiliation appears in Germany (t(862)=12.463, p<.001) with spouses of female respondents spending a weekly average of only 9.17 (SD=13.59) hours on household labor while spouses of male respondents spend a weekly average of 21.17 (SD=14.73) hours on household labor. The second largest mean difference by gender was among Portuguese respondents (t(570)=-4.493, p<.001). Female respondents in Portugal said that their spouses spent a weekly average of 41.52 (SD=43.12) hours on household labor – which is nearly twice as many household labor hours (25.53) that Portuguese male respondents reported for themselves – while male respondents said their spouses spent a weekly average of 29.63 (SD=22.55) hours on household labor. Portuguese men were close in their estimates on how many weekly hours their spouses or cohabiting partners spent on household labor when the mean value was compared to the mean hours that female respondents estimated for themselves (27.31). The lowest statistically significant mean difference (t(596)=4.557, p<.001) was among Dutch male and female respondents. Female respondents said their spouses spent on average 12.21 (SD=23.77) hours weekly on household labor while male respondents stated that their spouses spent a weekly average of 18.38 (SD=12.63) hours on household work. On average, respondents in each of the four countries do overestimate the hours of involvement of the spouse or 79 cohabiting partner in weekly household labor, but the large overestimation of Portuguese respondents on their spouses’ or cohabiting partners’ involvement in household labor is intriguing. It could be argued that due to Portugal’s history of being a familistic welfare state, stronger emphasis on the family and household environment may be key to the Portuguese engaging in more household labor hours. There may also be additional cultural factors that spur women to overemphasize their spouses’ or cohabiting partners’ involvement in household labor: for example, stressing the involvement of a spouse or cohabiting partner could signify the well-functioning of a family unit or bolster a family’s reputation. On the contrary, a more accurate estimation of household labor hours may also require a more critical estimation of how much a spouse or cohabiting respondent puts into household labor hours. Countries such as Germany, the USA, and the Netherlands that have a more egalitarian gender role ideology than Portugal may encourage women to develop a heightened consciousness on how much a spouse or cohabiting partner is involved in the household labor, therefore, leading to more accurate estimations for the spouses’ or cohabiting partners’ household labor involvement. Further research would be required to correlate specific factors to Portuguese women’s large overestimation of weekly household labor hours spent by their spouses or cohabiting partners. Crosstabulation Results for Religious Affiliation Crosstabulation comparison of percentages between religious main groups for each country (Table 5, p. 80) showed the prevalence of no religious affiliation among Dutch respondents with 65.1 percent, Protestant affiliations in Germany with 37.0 percent 80 (the highest percent within the country) as well as in the USA with 53.4 percent, and a strong Catholic affiliation in Portugal with 87.3 percent. Table 5: Crosstabulation results for religious affiliation by country Countries n=4,855 USA No Religion 14.5% Roman Catholic 24.7% Protestant 53.4% Other Religions 7.4% Germany 29.5% Netherlands Portugal Total 100.0% 29.7% 37.0% 3.8% 100.0% 65.1% 18.5% 12.0% 4.4% 100.0% 9.9% 87.3% .6% 2.1% 100.0% This research study intended to investigate whether religious affiliations had any influence on a respondent’s work-life balance. However, religious affiliations used as dummy variables in following regression analyses proved statistically insignificant. It is intriguing that religious affiliations have not had any kind of effect on managing worklife balance, which may suggest that gender role ideology (among other mentioned independent factors) may be one of the more critical factors in affecting work-life balance; possibly overwriting any degree of correlation between religious affiliation and work-life balance. This study will abstain from any further discussion on religious affiliation. Regression and Process Results Hypothesis H1A: Gender role ideology is significantly correlated to work-life balance after controlling for gender, marital status, employment status, education, age, number of children, religion, and country affiliation. Table 6 (p. 82) lists the multiple regression coefficients and standard errors for 81 work-life balance with time availability as the main predictor variable after controlling for gender, age, education, marital status, number of children ages 17 and under, religion, and country affiliation. The result indicates that gender role ideology is correlated with work-life balance. Respondents who have a more egalitarian gender role ideology experience more work-life balance. Work-life balance, however, is decreased by 0.108 points for female respondents compared to male respondents. Also, compared to being a married respondent, work-life balance decreases by .111 points for a respondent who has a cohabiting partner. Among the positive correlations, a one-year increase in the respondent’s age increases work-life balance by .003 points. Working part-time increases work-life balance by .126 points compared to full-time employment, while having another employment status with less than part-time hours also raises work-life balance by .100 points. On a country-level comparison, living in Germany compared to living in the United States increases work-life balance by .180 points. Living in the Netherlands compared to living in the United States increases work-life balance even more, by .207 points. There is, however, no statistically significant country-level effect of living in Portugal compared to the United States on work-life balance. Although the United States has the second highest average for a more egalitarian gender role ideology, respondents experience the least work-life balance. Female respondents in the United States reported a higher egalitarian gender role ideology compared to female respondents from the other three countries. Some of the explanations for this contradictory condition could lie within the societal structure of the United States. Work conditions with sometimes limited health 82 care and caregiving benefits as well as decreased accessibility to social insurance for the general population may be factors in work-life balance management. Table 6: Multiple regression for gender role ideology on work-life balance (respondents residing in the United States, Germany, the Netherlands, and Portugal only) Independent Variable: Gender role ideology Dependent Variable: Work-life balance B Beta .037*** (.011) .065 -.108*** (.026) .003* (.001) .001 (.001) -.111** (.040) -.044 (.029) -.032 (.028) -.024 (.014) .126*** (.035) -.008 (.076) .100* (.042) -.004 (.033) .008 (.035) -.114 (.061) -.081 .180*** (.034) .207*** (.037) .071 (.042) .119 Control Variables: Being Female (male=0) Age Years of Education Not Married/Cohabiting Partner (married and cohabiting=0) Not Married/Not Cohabiting Single (married and cohabiting=0) Number of Children (0 to 5, 6) Number of Children (6, 7 to 17) Part-time Employment (full-time=0) Housewife/man (full-time=0) Other Employment Status (full-time=0) Roman Catholic (no religion=0) Protestant (no religion=0) Other religion (no religion=0) .051 .019 -.055 -.031 -.021 -.031 .070 -.002 .051 -.003 .005 -.035 Country-level Variables: Germany (United States=0) Netherlands (United States=0) Portugal (United States=0) .139 .040 83 Constant 2.851*** (.079) .043 R-square Note: n=3,059; b=unstandardized regression coefficient with standard error in parenthesis; Beta= standardized regression coefficient. *p<.05; **p<.01; ***p<.001 (two-tailed tests) The hypothesis that gender role ideology is significantly correlated to work-life balance is supported. Mediating Hypothesis H1B: The perception of fairness of the division of household labor mediates the degree to which gender role ideology affects work-life balance after controlling for gender, marital status, employment status, education, age, number of children, religion, and country affiliation. As shown in Table 7 (p. 83), mediation analysis indicates a direct effect at .0298 (p<0.05) of gender role ideology on work-life balance. Table 7: Process mediation analysis of perception of fairness of the division of household labor with direct and indirect effects of gender role ideology on work-life balance Dependent Variable (Y)=Work-life balance Independent Variable (X)=Gender role ideology Mediator (M)=Perception of fairness of the division of household labor Direct effect of X on Y Effect Std. Error p LLCI ULCI .0298 .0129 .0214 .0044 .0551 BootLLCI BootULCI -.0014 .0027 Indirect effect of X on Y Effect Perception of Fairness of the Div. of Hh. Labor .0003 Boot Std. Error .0010 84 There is, however, no indirect effect of gender role ideology on work-life balance through the mediator variable perception of fairness of the division of household labor. The bias-corrected 95% bootstrap confidence interval for the indirect effect includes zero. The hypothesis that the perception of fairness of the division of household labor mediates the degree to which gender role ideology affects work-life balance is not supported. Hypothesis H2A: Time availability is significantly correlated to work-life balance after controlling for gender, marital status, employment status, education, age, number of children, religion, and country affiliation. Table 8 (p. 85) lists the multiple regression coefficients and standard errors for time availability on work-life balance after controlling for gender, age, education, marital status, number of children ages 17 and under, religion, and country affiliation. The result indicates that time availability (respondent’s hours worked weekly) is negatively correlated to work-life balance. With a one-hour increase in work time per week, the respondent’s work-life balance decreases by .008 points. Work-life balance further decreases by .138 points for female respondents compared to male respondents. Also, compared to being a married respondent, work-life balance also decreases by .091 points for a respondent who has a cohabiting partner. On a country-level comparison, living in Germany compared to living in the United States increases work-life balance by .142 points. Living in the Netherlands 85 compared to living in the United States increases work-life balance as well by .126 points. Living in Portugal compared to the United States has no country-level effect on work-life balance. In the Netherlands, the significant results may point again to a ‘societal’ effect because the Dutch have the least weekly work hours compared to the other three countries. In addition, the Netherlands and Germany are both countries with social insurance which may also be a contributing factor for allowing respondents from these two countries to experience more work-life balance. Further, Germany is also a country which has more paid holidays and vacation days than any of the four countries. This structural condition can also help Germans to better manage their work-life balance even though they work more weekly hours than respondents from the United States. Table 8: Multiple regression for time availability (respondent’s hours worked weekly) on work-life balance (respondents residing in the United States, Germany, the Netherlands, and Portugal only) Independent Variable: Dependent Variable: Work-life balance B Beta Time Availability -.008*** (.001) -.163 -.138*** (.026) .002 (.001) .433 (.001) -.091* (.041) -.028 (.030) -.035 (.028) -.023 (.014) -.107 Control Variables: Being Female (male=0) Age Years of Education Not Married/Cohabiting Partner (married and cohabiting=0) Not Married/Not Cohabiting Single (married and cohabiting=0) Number of Children (0 to 5, 6) Number of Children (6, 7 to 17) .037 .001 -.047 -.020 -.024 -.032 86 Part-time Employment (full-time=0) Housewife/man (full-time=0) Other Employment Status (full-time=0) Roman Catholic (no religion=0) Protestant (no religion=0) Other religion (no religion=0) .010 (.041) .597 (.446) .057 (.082) -.008 (.034) -.002 (.036) -.096 (.062) .006 .025 .019 -.006 -.001 -.031 Country-level Variables: Germany (United States=0) Netherlands (United States=0) Portugal (United States=0) Constant R-square .142*** (.036) .126** (.039) .010 (.042) 3.407*** (.090) .054 .095 .089 .006 Note: n=3,059; b=unstandardized regression coefficient with standard error in parenthesis; Beta= standardized regression coefficient. *p<.05; **p<.01; ***p<.001 (two-tailed tests) The hypothesis that time availability is significantly correlated to work-life balance is supported. Mediating Hypothesis H2B: The perception of fairness mediates the degree to which time availability affects work-life balance after controlling for gender, marital status, employment status, education, age, number of children, religion, and country affiliation. In Table 9 (p. 87), mediation analysis points to a direct negative effect at -.0098 (p<.05) of time availability, the respondent’s hours worked weekly, on work-life balance. There was, however, no indirect effect of the respondent’s hours worked weekly on 87 work-life balance through the mediator variable perception of fairness of the division of household labor. The bias-corrected 95% bootstrap confidence interval for the indirect effect includes zero. Table 9: Process mediation analysis of perception of fairness of the division of household labor with direct and indirect effects of time availability on work-life balance Dependent Variable (Y)=Work-life balance Independent Variable (X)=Time availability Mediator (M)=Perception of fairness of the division of household labor Direct effect of X on Y Effect -.0098 Std. Error .0015 p LLCI ULCI .0000 -.0127 -.0069 Indirect effect of X on Y Effect Perception of Fairness of the Div. of Hh. Labor .0002 Boot Std. Error .0001 BootLLCI BootULCI .0000 .0005 The hypothesis that the perception of fairness mediates the degree to which time availability affects work-life balance is not supported. Mediating Hypothesis H3: The perception of fairness of the division of household labor mediates the degree to which resource dependence affects work-life balance after controlling for gender, marital status, employment status, education, age, number of children, religion, and country affiliation. Resource dependence has a positive effect on work-life balance. Due to previous research pointing to resource dependence within marriage or 88 partnership (income distribution between spouses or cohabiting partners) as a factor in the perception of fairness of the division of household labor, the relationship of the two variables and their effect on work-life balance was explored through mediation analysis in this study. Process analysis tested the positive effect of resource dependence on worklife balance mediated by the perception of fairness of the division of household labor. In Tables 10 A and B (p. 89), results indicate that increasing resource dependence is directly related to more work-life balance. The positive effect is .0215 (for a one-tailed test at p<0.05). The positive effect, however, is offset when the respondent over-benefits from the division of household labor and does less than their fair share in the household. The indirect effect of resource dependence on work-life balance through the perception of fairness of the division of household labor is -.0040. One viable explanation as to why over-benefitting for the division of household labor can negatively affect the positive relationship between resource dependence and work-life balance is that a respondent may have feelings of guilt that he/she is engaged in less than his/her fair share of household labor, while the spouse or cohabiting partner has a higher income than the respondent. Higher income may be ascribed a higher value in the relationship, and to balance out the skewed income distribution between the spouses or cohabiting partners, the respondent believes that he/she ought to engage in more household labor. Not doing more household labor, however, could evoke feelings of guilt, ultimately making it tougher for the respondent to experience work-life balance. 89 Table 10 A: Process mediation analysis of the perception of fairness of the division of household labor with direct and indirect effects of resource dependence on work-life balance Dependent Variable (Y)=Work-life balance Independent Variable (X)=Resource dependence Mediator (M)=Perception of fairness of the division of household labor Direct effect of X on Y Effect Std. Error p LLCI ULCI .0215 .0118 .0700 -.0018 .0447 BootLLCI BootULCI -.0084 -.0013 Indirect effect of X on Y Effect Boot Std. Error Perception of Fairness of the Div. of Hh. Labor -.0040 .0017 Table 10 B: Effect of predictor variables and covariates from Process mediation analysis of the perception of fairness of the division of household labor with direct and indirect effects of resource dependence on work-life balance Independent Variables: Dependent Variable: Work-life balance B Std. Error LLCI ULCI Perception of the division of household labor Resource dependence .0485** .0154 .0183 .0786 .0215* .0118 -.0018 .0447 Female -.0726 .0412 -.1533 .0081 Age .0040** .0015 .0011 .0069 Education/years of schooling .0003 .0014 -.0025 .0031 Cohabiting partners/not married (married=0) Single (married=0) -.0739 .0460 -.1641 .0163 -.0675 .0832 -.2306 .0956 Children in household up to 5, 6 years of age .0025 .0299 -.0562 .0611 90 Children in household 6, 7-17 years of age Part-time (full-time=0) -.0225 .0163 -.0544 .0094 .1580*** .0450 .0698 .2462 Homemaker (full-time=0) .0601 .0819 -.1005 .2207 Other employment status (full-time=0) Roman Catholic (no religion=0) Protestant (no religion=0) .1643** .0553 .0558 .2727 -.0111 .0411 -.0917 .0694 -.0134 .0424 -.0965 .0696 Other religions (no religion=0) Germany -.0954 .0739 -.2403 .0494 .1787*** .0407 .0988 .2585 Netherlands .2070*** .0456 .1176 .2965 Portugal .0251*** .0527 -.0783 .1285 Constant 2.6993*** .1077 2.4881 2.9106 R-square .0557 *p<.05; **p<.01; ***p<.001 (two-tailed tests) The hypothesis that the perception of fairness of the division of household labor mediates the degree to which resource dependence affects work-life balance while resource dependence has a positive effect on work-life balance is supported. Hypothesis H4A: Respondent’s weekly household labor hours are significantly correlated to work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. In Table 11 (p. 91), regression analysis pointed to no statistically significant relationship between the hours a respondent spent weekly on household labor and worklife balance (p<.89). 91 Table 11: Multiple regression for respondent’s weekly household labor hours on work-life balance (respondents residing in the United States, Germany, the Netherlands, and Portugal only) Independent Variable: R’s Weekly Hh. Labor Hrs. Dependent Variable: Work-life balance B Beta .000 (.001) -.003 -.094** (.033) .004* (.001) .001 (.001) -.080 (.046) -.038 (.080) -.003 (.030) -.028 (.016) .158*** (.044) .087 (.081) .180** (.056) -.014 (.041) -.002 (.042) -.085 (.075) -.072 .201*** (.041) .225*** (.045) .053 (.053) 2.915*** (.086) .138 Control Variables: Being Female (male=0) Age Years of Education Not Married/Cohabiting Partner (married and cohabiting=0) Not Married/Not Cohabiting Single (married and cohabiting=0) Number of Children (0 to 5, 6) Number of Children (6, 7 to 17) Part-time Employment (full-time=0) Housewife/man (full-time=0) Other Employment Status (full-time=0) Roman Catholic (no religion=0) Protestant (no religion=0) Other religion (no religion=0) .066 .014 -.043 -.011 -.002 -.039 .092 .025 .077 -.010 -.001 -.027 Country-level Variables: Germany (United States=0) Netherlands (United States=0) Portugal (United States=0) Constant .155 .031 92 R-square .051 Note: n=3,059; b=unstandardized regression coefficient with standard error in parenthesis; Beta=standardized regression coefficient. *p<.05; **p<.01; ***p<.001 (two-tailed tests) The hypothesis that respondent’s weekly household labor hours are significantly correlated to work-life balance is not supported. Mediating Hypothesis H4B: Respondent’s weekly household labor hours mediate the effect of gender role ideology on work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. Because the study explores the relationship among specific variables and their potential influence on work-life balance, the hours a respondent spent weekly on household labor were further investigated in Process analysis of the mediator variable and its effect on the relationship between gender role ideology and work-life balance. Table 12: Process mediation analysis of respondent’s weekly household labor hours with direct and indirect effects of gender role ideology on work-life balance Dependent Variable (Y)=Work-life balance Independent Variable (X)= Gender role ideology Mediator (M)= Respondent's weekly household labor hours Direct effect of X on Y Effect Std. Error p LLCI ULCI .0306 .0131 .0194 .0049 .0562 93 Indirect effect of X on Y Effect R’s Weekly Hh. Labor Hrs. .0002 Boot Std. Error .0009 BootLLCI BootULCI -.0016 .0021 In Table 12 (p. 92), mediation analysis points to a direct effect at .0306 (p<0.05) of the respondent’s gender role ideology on work-life balance. There was, however, no indirect effect of the respondent’s gender role ideology on work-life balance through the mediator variable respondent’s hours spent weekly on household labor. The bias-corrected 95% bootstrap confidence interval for the indirect effect includes zero. The hypothesis that respondent’s weekly household labor hours mediate the effect of gender role ideology on work-life balance is not supported. Hypothesis H5A: Spouse’s/cohabiting partner’s weekly household labor hours are significantly correlated to work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. As the ISSP data set offers a variable highlighting the hours spent on household labor by the respondent’s spouse (as estimated by the respondent), the same regression and Process analyses were run to determine any statistical significance of the spouse’s household labor hours on the respondent’s work-life balance. Regression analysis indicates a statistically significant relationship between the hours spent on household labor by the respondent’s spouse and the respondent’s work-life balance (Table 13, p. 94 94). Table 13: Multiple regression for spouse’s/cohabiting partner’s weekly household labor hours on work-life balance (respondents residing in the United States, Germany, the Netherlands, and Portugal only) Independent Variable: Dependent Variable: Work-life balance B Beta Spouse’s/Cohabiting Partner’s Weekly Hh. Labor Hrs. Control Variables: -.002** (.001) -.062 Being Female (male=0) -.098** (.033) .004** (.001) .000 (.001) -.076 (.046) .009 (.082) .002 (.030) -.027 (.016) .146** (.044) .098 (.082) .156** (.055) -.008 (.041) -.001 (.042) -.087 (.074) -.075 .193*** (.041) .230*** (.045) .069 (.053) 2.933*** (.086) .133 Age Years of Education Not Married/Cohabiting Partner (married and cohabiting=0) Not Married/Not Cohabiting Single (married and cohabiting=0) Number of Children (0 to 5, 6) Number of Children (6, 7 to 17) Part-time Employment (full-time=0) Housewife/man (full-time=0) Other Employment Status (fulltime=0) Roman Catholic (no religion=0) Protestant (no religion=0) Other religion (no religion=0) .074 .004 -.041 .003 .002 -.038 .084 .028 .067 -.006 -.001 -.028 Country-level Variables: Germany (United States=0) Netherlands (United States=0) Portugal (United States=0) Constant .158 .041 95 R-square .053 Note: n=3,059; b=unstandardized regression coefficient with standard error in parenthesis; Beta=standardized regression coefficient. *p<.05; **p<.01; ***p<.001 (two-tailed tests) Table 13 (p. 94) lists the multiple regression coefficients and standard errors for the spouse’s/cohabiting partner’s weekly household labor hours on work-life balance after controlling for gender, age, education, marital status, number of children ages 17 and under, religion, and country affiliation. The result indicates that more hours spent weekly on household labor by the respondent’s spouse are negatively correlated to the respondent’s work-life balance. With a one-hour increase in the spouse’s or cohabiting partner’s weekly household labor time, the respondent’s work-life balance decreases by .002 points. If the respondent is disproportionately engaged in household labor – possibly due to longer weekly work hours – the spouse or cohabiting partner would need to invest more weekly household labor hours. Under this scenario, it would be very likely that the respondent experiences less work-life balance. Although following interpretation would require further testing, the result could also indicate that respondents may experience feelings of guilt when they estimate that their spouses or cohabiting partners engage in more household labor hours. Work-life balance further decreases by .098 points for female respondents compared to male respondents. Also, with each year of age, the respondent’s work-life balance increases by .004 points which could indicate that respondents who get older might adopt additional coping strategies that allow them to better manage work-life balance. Being a respondent who is employed part-time compared to full-time, increases 96 work-life balance by .146 points. Compared to respondents with full-time employment status, having another employment status with less than part-time hours increases worklife balance as well by .156 points. On a country-level comparison, living in Germany compared to living in the United States increases work-life balance by .193 points. Living in the Netherlands compared to living in the United States increases work-life balance as well by .230 points – slightly more than for respondents who live in Germany. There is, however, no statistically significant country-effect of living in Portugal compared to living in the United States on work-life balance. Compared to the United States, Germany and the Netherlands are two countries which have more employment fringe benefits, paid holidays and vacations, and social insurance which are all factors in better managing work-life balance. The hypothesis that spouse’s/cohabiting partner’s weekly household labor hours are significantly correlated to work-life balance is supported. Mediating Hypothesis H5B – Spouse’s/cohabiting partner’s weekly household labor hours mediate the effect of gender role ideology on work-life balance after controlling for gender, age, education, marital status, number of children, employment status, religion, and country affiliation. As highlighted in Tables 14 A and B (p. 97), Process analysis indicated a direct effect at .0264 of gender role ideology on work-life balance (p<.0433). As the previous correlation had confirmed as well, higher scores toward more egalitarian gender role 97 ideology is correlated and has a direct positive on work-life balance. Table 14 A: Process mediation analysis of spouse’s/cohabiting partner’s weekly household labor hours with direct and indirect effects of gender role ideology on work-life balance Dependent Variable (Y)=Work-life balance Independent Variable (X)=Gender role ideology Mediator (M)= Spouse’s/cohabiting partner’s weekly household labor hours Direct effect of X on Y Effect .0264 Std. Error p LLCI ULCI .0131 .0433 .0008 .0520 BootLLCI BootULCI Indirect effect of X on Y Effect Boot Std. Error Spouse’s/Cohabiting Partner’s Weekly Hh. Labor Hrs. .0025 .0014 .0005 .0067 Table 14 B: Effect of predictor variables and covariates from Process mediation analysis of spouse’s/cohabiting partner’s weekly household labor hours with direct and indirect effects of gender role ideology on work-life balance Independent Variables: Dependent Variable: Work-life balance B Std. Error LLCI ULCI Hours respondent’s spouse spent on household labor Gender role ideology -.0017** .0006 -.0029 -.0004 .0264* .0131 .0008 .0520 Female -.1147*** .0334 -.1802 -.0492 Age .0045** .0015 .0016 .0075 Education/years of schooling .0000 .0014 -.0028 .0028 Cohabiting partners/not married -.0813 .0463 -.1721 .0094 98 (married=0) Single (married=0) .0253 .0832 -.1378 .1884 Children in household up to 5, 6 years of age Children in household 6, 7-17 years of age Part-time (full-time=0) .0061 .0302 -.0532 .0654 -.0264 .0163 -.0584 .0055 .1481*** .0446 .0606 .2355 Homemaker (full-time=0) .1057 .0826 -.0562 .2676 Other employment status (fulltime=0) Roman Catholic (no religion=0) .1616** .0555 .0527 .2705 .0045 .0415 -.0768 .0858 Protestant (no religion=0) .0070 .0428 -.0769 .0908 Other religions (no religion=0) -.0655 .0750 -.2126 .0816 Germany .1955*** .0412 .1146 .2763 Netherlands .2320*** .0455 .1428 .3213 Portugal .0701 .0532 -.0343 .1745 Constant 2.8169*** .1024 2.6161 3.0176 R-square .0559 *p<0.05; **p<0.01; ***p<0.001 (two-tailed tests) There is also a positive indirect effect at .0025 of gender role ideology on worklife balance through the weekly hours spent on household labor by the respondent’s spouse. The bias-corrected 95% bootstrap confidence interval for the indirect effect does not include zero. The effect of gender role ideology on work-life balance increases when respondents perceive that their spouses or cohabiting partners spent more hours on household labor. Work-life balance also increases when the respondents are either older, have a part-time job, less than a part-time job, live in Germany, or live in the Netherlands. Work-life balance decreases, however, when respondents are female compared to male which suggests that women have to assume multiple and demanding roles of workers, mothers, caregivers, and home-managers. Additional research needs to be conducted to test for a possible interaction between the respondent’s gender and the spouse’s/cohabiting partner’s weekly household labor hours and their effects on the 99 relationship between gender role ideology and work-life balance. This could point to deeply entrenched traditional gender role ideologies within the individual, his/her surroundings, and/or the larger socio-structural context which posit that women are assigned the duties of the home environment while men are assigned the duties of the work environment. The hypothesis that spouse’s/cohabiting partner’s weekly household labor hours mediate the effect of gender role ideology on work-life balance is supported. Table 15 (p. 99) provides a complete list of decision outcomes for all hypotheses tested in this research study. Table 15: Decision outcomes for all tested hypotheses HYPOTHESIS DECISION H1A Supported H1B Not Supported H2A Supported H2B Not Supported H3 Supported H4A Not Supported H4B Not Supported H5A Supported H5B Supported 100 Chapter 5 DISCUSSION AND CONCLUSION Regardless of gender, almost any employed worker faces to some degree the challenges of balancing the work and home environment. Work-life balance has been a newer concept with special focus on the well-functioning of the work and home environment by minimizing role conflict (Clark 2001; Guest 2002). Women, however, have been associated historically with the home environment, and by women making up a continuously growing part of the labor force, they are confronted with the ‘double shift’: making it harder especially for them to minimize role conflict and to balance the work and home environment (Hochschild 1989). While research results for men have been described in this study and are treated as an essential component in understanding gendered relationships, the attempt has been made to apply a gendered lens throughout the study, emphasizing women’s complex duties as wives, partners, mothers, caregivers, home-managers, and workers. Research on work-life balance has often focused on paid work rather than unpaid work within the home environment and how factors within the paid work environment such as workplace policies, flexibility or fringe benefits can help balance work-life spheres (Coltrane 2000; Kreuzenkamp and Hooghiemsra 2000; Wierda-Boer 2009). In order to widen the pool of potential factors influencing work-life balance, this study has drawn from research on the division of household labor to identify independent factors such as gender role ideology, time availability, and resource dependence, as well as mediating factors such as the perception of fairness of the division of household labor as a potential ‘meso-marker’ and the weekly household labor hours 101 which potentially affect work-life balance. Milkie and Peltola (1999) conducted one of the first research studies which focused directly on the perception of fairness of the division of household labor and its effect on work-life balance. Although their study included gender-specific mediation factors that pointed to gender differences in how employed married couples managed their work-life balance, the study did not include broader macro-level/country-specific variables. Hence, this study also attempted to create statistical measures to account for macro-level/country-level effects on work-life balance. The data set used for the analyses stems from the 2002 International Social Survey Programme: Family and Changing Gender Roles III (ISSP 2004). Research results identified a more egalitarian gender role ideology correlated with more work-life balance. Also, work-life balance decreased for women or for respondents who had a cohabiting partner (but were not married). This suggests that marriage as a legal bond may provide spouses with a sense of security that alleviates some of the pressures on work-life balance. By investigating the effects of gender role ideology on work-life balance, the analysis confirmed that part-time or less than part-time employment are also factors in having more work-life balance. On a macro/country-level, only respondents from Germany and the Netherlands compared to respondents from the United States exhibited more work-life balance. This may be due to specific policies that are in place in Germany and the Netherlands which allow for better management of work-life balance (Voydanoff 2004). Although in many countries gender roles have moved from traditional (male breadwinner) to more egalitarian (dual-income earner) roles, there still remain structural barriers such as unequal or low pay, and limited social 102 or health insurance for working women. While Germany and the Netherlands are labeled conservative welfare states which historically embrace the male breadwinner model, both countries rate high on the GEM index which points to less structural barriers (such as more equitable pay) in the work environment (United Nations Human Development Programme 2007; Ruppanner 2008). In Germany, tax rates for second income earners are highest compared to the Netherlands, the United States, and Portugal, yet days of paid maternity and paternity leave are high (OECD 2010). Even if dual-income earners are disadvantaged to some degree, Germany has made efforts in increasing work-life balance by offering not only longer periods of paid maternity leave but extending the leave to fathers as well; therefore, allowing women to quickly re-enter the workforce after giving birth to minimize gaps in employment and/or jeopardizing career opportunities (Voydanoff 2004). In the Netherlands, the right to part-time work is institutionalized, therefore, prevalent as a preferred employment status (Ruppanner 2008; Moors 1995). Many Dutch women compared to Dutch men will take advantage of part-time work, especially after giving birth (Ruppanner 2008). Hence, it can be inferred that living in the Netherlands – as a society that has normalized part-time employment and that has a more egalitarian gender role ideology – has a positive “societal” (Crompton and Lyonette 2006:389; Ruppanner 2008) effect on work-life balance. Yet, a more egalitarian gender role ideology within a country does not automatically suggest that the division of household labor has also moved toward a more equitable household division of labor between men and women (Orloff 2009). Even when a country such as the Netherlands exhibits an 103 increased egalitarian gender role ideology and has governmental policies in place which allow women and men to better manage their work-life, institutionalized part-time work may lead women to push their male spouses less to engage in a more equitable division of household labor (Ruppanner 2008). Further, more work hours affect work-life balance negatively, while being a woman or having a cohabiting partner decreases work-life balance. Again, this may signify that women are especially affected by the ‘double’ and ‘triple shift’ and struggle to achieve work-life balance when they work longer hours only to come home to resume responsibilities in the home environment (Hochschild 1989). Due to part-time work being institutionalized in the Netherlands compared to the United States, respondents in the Netherlands can take advantage of working less per week, and therefore, experience the country-effect that the Netherlands has on increased work-life balance. Although German respondents work slightly more than respondents in the United States, Germany not only has statutory health insurance but is one of the countries with the most holidays and paid vacations, especially when compared to the Unites States (OECD 2010; Voydanoff 2004). These employment and health benefits within the German system may enable Germans to better manage and perceive more work-life balance while working more hours on a weekly average than respondents from the United States. For the effects of gender role ideology or time availability on work-life balance, the perception of fairness of the division of household labor has not been a mediating factor. Mediation analysis, however, indicated that increased resource dependence on the spouse’s or cohabiting partner’s income is positively correlated with work-life balance. 104 This effect is reduced, however, when the respondent over-benefits from the division of household labor by doing less than his/her fair share of household labor. This indirect effect could be explained by potential guilt that a respondent might experience, knowing that the spouse or cohabiting partner earns more money while the respondent him/herself is keenly aware that he/she does not engage adequately in household labor (Greenhaus and Beutell 1985; Greenhaus et al. 2006). According to Braun et al. (2008), resource dependence equates to relative bargaining power in negotiating the amount of household labor in which a spouse or cohabiting partner engages. Using equity theory and the legitimizing principle, if a spouse or cohabiting partner is more dependent on resources from the other party, he/she will legitimize his/her greater involvement in household labor, perceiving the division of household labor as fair even though the division of actual household labor hours is unequal (Braun et al. 2008). Resource dependence, hence, can cause a respondent to assign more value to the spouse who earns the higher income, therefore, believing that he/she should engage in more household labor. If the respondent, however, fails to do his/her perceived fair share of household labor, he/she most likely transposes feelings of guilt onto his/her work-life balance. Using role-spillover theory, an individual’s attitudes and emotions can spill over from the work to the home environment and vice versa. For the reason that gender role ideologies can transfer and link the public spheres with social institution and embedded gender stratifications to the home environment with its individual actors, individuals can struggle – through an emotion such as guilt – to reconcile inter-role conflict through complex gender role pressures within the work and home environment (Greenhaus and Beutell 1985; Greenhaus et al. 105 2006). Further research needs to be conducted in order to apply a more detailed gender lens and identify specific gender differences for the mediating effect of the perception of fairness of the division of household labor which negatively offsets the effect of resource dependence on work-life balance. Weekly hours a respondent spent on household labor did not affect the respondent’s work-life balance, nor did they have any mediating indirect effect of the positive correlation between a more egalitarian gender role ideology and work-life balance. Weekly household labor hours spent by a spouse or cohabiting partner, however, affected the respondent’s work-life balance negatively. This negative correlation might point to a respondent working longer hours which would create an unequal division of household labor between the respondent and the spouse; consequently, this would decrease the respondent’s work-life balance. In addition, the respondent may even experience feelings of guilt for his/her own disproportionate work and household labor involvement. Further, the correlation between gender and work-life balance was stronger for female compared to male respondents. The fact that the effect on work-life balance is weaker for a female respondent points to women’s multiple, demanding, and often competing roles such as workers who may not receive strong enough governmental and/or organizational support to offset their duties at home as mothers, caregivers, and home-managers. Traditional gender roles and expectations supported by the larger society can make the balancing of the work and family environment even more challenging for women, especially when working women compare themselves to other women in a traditional society. This can lead women to “doing gender” (Ruppanner 106 2008:520; West and Zimmerman 1987:125), to engage in even more household labor in order to signify their support for a traditional gender ideology and division of household labor. When the spouse’s weekly household labor hours were identified as indirectly increasing the positive effect of a more egalitarian gender role ideology on work-life balance, this might also suggest two competing gender role views on a gender-specific individual level which are not further investigated in this research study but touched upon in the following hypotheses: For one, even more egalitarian men may still have remnants of more traditional gender role ideologies internalized, and would not find fault in their wives or female cohabiting partners engaging in more weekly household labor hours. Secondly, more egalitarian women would welcome it if their husbands or male cohabiting partners engaged in more weekly household labor hours, therefore, alleviating some of the women’s stresses of the home environment, and increasing their work-life balance. Strengths and Limitations This study has several strengths but also limitations which should be discussed by reiterating briefly the data set that was used for the analyses. One of the strengths of using the 2002 ISSP data set is that yearly data sets from the International Social Survey Programme are cross-national with the intent to collect data that can be used for replications on social science issues across all participating countries (ISSP 2004). The survey data is collected and processed with proper weighting measures from a reputable institution in each participating country. The sample sizes for each country are large enough to use the analytical results for representative samples of each country under 107 investigation (ISSP 2004). The data set is made available for public use by the Zentralarchiv für Empirische Sozialforschung at the Universität Köln. The 2002 ISSP survey with the module’s focus on family and changing gender roles has been an ideal data set for the researcher to investigate which additional factors drawn from the division of household labor affect work-life balance. The data set offers more variables than were used for this research study, therefore, leaving opportunities for continued exploration of other factors that may influence work-life balance. In addition, research studies that have used the same perception of fairness of the division of household labor variable have often interpreted the categories on a scale of a less fair to a more fair division of household labor. The value of a fair division of household labor lies in the middle score of 3; scores of 4 or 5 state that the respondent perceives to be doing a bit less or much less than their fair share of household labor. The strength of this research study, however, consists of the differing and possibly improved interpretation of the categories, with lower scores of 1 and 2 meaning the respondent under-benefits from the division of household labor, and higher scores of 4 and 5 signifying the respondent over-benefits from the division of household labor. On a conceptual and theoretical level, the perception of fairness of the division of household labor is complicated in so far as perception of fairness can have varying meanings to spouses or cohabiting couples. This has become evident in the descriptions of the legitimizing principles which are used to assess how fair a specific division of household labor is perceived by men and women (Braun et al. 2008). Using equity theory and the perception of fairness of the division of household labor in a research study point to the 108 mechanisms that legitimize an unequal division of household labor. Although time-use diary research has been an alternative to investigating the division of household labor and has shown that men’s participation in domestic tasks has increased, women still remain largely in charge of managing domestic tasks – even if men execute them (Sullivan 2000). Also, even if men put equal hours into household labor, the type of task may be more gender-specific such as mowing the lawn or washing the car; both being public outdoor activities which reinforce more traditional gender roles (Bonvillain 1998; Frone 2003). This suggests that lingering traditional gender role ideologies of a country and/or of an individual continue to exert influence on the management and division of household labor among couples or cohabiting spouses. There are a few disadvantages to the 2002 ISSP data set. Racial and ethnic variables are not included in the data set. This exclusion, however, could be justified by putting forward the notion that cross-country comparisons would be too limited: only very few races and ethnicities would overlap between countries if the variables were included. Also, racial and ethnic variations are often very complex within a country and any attempt to engage in cross-country comparison would have to be conducted in the utmost sensible manner. Yet, it appears to be of great disadvantage to not have racial, ethnic, and country of origin variables included. Women from lower classes and minority groups are often the most vulnerable to experience gender stratification and its effects on the division of household labor and work-life balance. Due to the ‘elevator effect’, these groups may also be the ones more likely to accept unequal treatment within the work and the home environment (Braun et al. 2008; Beck 1992). Especially when looking at the 109 United States and other industrialized countries, care-giving responsibilities in the public sphere have been assigned predominately to underprivileged minority groups, increasing the stresses of the ‘double shift’ even more for women of these groups; working long, underpaid hours especially in the service sector with limited access to good health and social benefits while assuming long, unpaid hours of household labor and care-giving activities when they get home (Prince Cooke 2011; Tronto 1993). Although adding these critical variables to the ISSP data set would make the information gathering and weighting process more intricate, it would provide an even more powerful research tool to investigate in greater detail and under a racial and ethnic lens the complexities of social interactions between individuals, societies, and countries. Within the data set, the income variable has been a problematic measure because the overall sample size indicated too many missing cases. An income measure is often a necessary covariate factor which ought to be included in regression and Process analyses. Yet, the income variable had to be excluded because it would have introduced bias into the research results. To mitigate this limitation, the education variable became the proxy for the income variable, but continuously had no effect on work-life balance in this study. Previous research studies have picked one variable over the other; regardless though of this approach, it is ideal to include both variables, income and education, into this kind of research design. In equity theory, fairness of the division of household labor is assessed by inputs and outputs of each spouse or cohabiting partner; weekly hours worked would be the input, and income the resulting outcome of the effort (Braun et al. 2008; Walster and Berscheid 1978). Further, the 2002 ISSP data set is an older data set. Within the past 110 decade, changes in attitudes among respondents from the United States, Germany, the Netherlands, and Portugal may have occurred. Although there is a newer module 2012 International Social Survey Programme: Family and Gender Roles IV currently in the process of being published, it is not publicly available yet. If a newer data set would have existed, the researcher could have approached the study from a methodological point of trend analysis, including changes in attitudes from the last three modules of 1994, 2002, and 2012. Directions for Future Research This research study investigated division of household labor factors as predictors for work-life balance among respondents in only four countries, the United States, Germany, the Netherlands, and Portugal. Although using only four countries in this analysis is a methodologically viable approach in order to test hypotheses on a small cross-country scale, future research should also include large-scale country comparisons of all countries that participated on the 2002 ISSP data set. In addition, trend analysis of changing attitudes about gender roles within the work-life context could offer more details in social processes and how changes in gender role ideologies over time have affected respondents’ views on the division of household labor and on other critical attitudes that can affect their work-life balance. As Coltrane (2010) has stressed, there is a great need for more comprehensive and complex research designs that not only use crosscountry comparisons and trend analysis but take into account the “economic, political and social context under which couples divide household labor as well as paid labor” (Coltrane 2010:798). For the reason that this research study did not focus on what kind of 111 household labor respondents and spouses/cohabiting partners of respondents were engaged in and for how long, future research should also include more of such measures to get detailed accounts on how household labor is divided among spouses or cohabiting partners and who takes responsibility in the household for the management of such tasks. Even if spouses or cohabiting partners engaged in equal amounts of household labor, the kind of household labor they are involved in could still perpetuate more traditional gender role ideologies: the man mowing the lawn or cleaning the car, while the woman cooks and cares for the children (Bonvillain 1998; Frone 2003; Sullivan 2000). So, even with an equal amount spent by both on cumulative paid and unpaid household labor, the woman could still be assigned more traditional care-giving duties (Wierda et al. 2009). In addition, this research study did not further investigate the interaction effect of gender on the relationship between gender role ideology and work-life balance. This, however, raises the question: Would egalitarian male respondents, compared to egalitarian female respondents, experience more work-life balance when their spouses/cohabiting partners engaged in more household labor? The results could confirm or refute the notion that traditional gender roles are still deeply imbedded within men and women, even if they described themselves as having a more egalitarian gender role ideology. Previous research on fairness perception of the division of household labor has pointed to the paradox that women who have a more traditional gender role ideology can perceive the division of household labor as fair even though they engage in paid labor as well as unpaid labor in the form of doing more than their fair share of household labor, therefore, under-benefitting from the division of household labor (Braun et al. 2008). 112 Under these conditions, this raises the question whether women are potentially engaging (consciously or subconsciously) in a thought process that could paradoxically enable them to also experience more work-life balance. This would be, however, counterintuitive to the research finding in this study that respondents with more egalitarian gender role ideologies experience more work-life balance, while the perception of fairness of the division of household labor has no mediating effect on that relationship. In this regard, the findings may deserve further scientific inquiry by using Process analysis with multiple mediation variables to explore the complex relationships between gender, gender role ideology, time availability, resource dependence, respondent’s and spouse’s/cohabiting partner’s weekly household labor hours, and perception of fairness of the division of household labor on work-life balance. Also, an attempt in future research should be made to expand on the theoretical concept of equity theory and resource dependence by including both income and education into quantitative analyses (Braun et al. 2008; Walster and Berscheid 1978). Both variables can point to the complexities of mechanisms that are in place which determine spouses’ or cohabiting partners’ perception of fairness of the division of household labor. According to Ruppanner’s research (2008), women in the Netherlands, for example, perceive more fairness in the division of household labor when they have more years of education. This is clearly the opposite of what would be expected since education is treated as a proxy for resource dependence and gives women a legitimizing mechanism to perceive the division of household labor as less fair (Braun et al. 2008). Women’s higher income has been linked to a decreased perception of fairness of the 113 division of household labor. These legitimizing principles are intricately linked with performing gender roles. The fact that spouses engage in “doing gender” (Ruppanner 2008:520; West and Zimmerman 1987:125) becomes even more complex when both financially dependent and independent women engage in more household labor – reaffirming traditional gender roles – to counteract resource inequities between themselves and their husbands or male cohabiting partners. Suggestions for Improvement While research on work-life balance has pointed to factors such as more flexibility in work hours, options to work from home, better fringe benefits, improved maternity and paternity leave, and daily childcare and breastfeeding support at the workplace that help women and men balance the work and home environment, governmental institutions can also implement laws and regulations which support more gender equality within a society (OECD 2010, Voydanoff 2004). Allowing the husband or male cohabiting partner to take time off to care for an infant or child, and the government improving the tax rate for second income earners are both steps toward a more equal division of labor within the work and home environment. Institutionalized paternity leave would support a gender role ideology that moves the father into the home environment, encouraging men to take on a more equitable care-giving responsibility without jeopardizing job loss or career advances. By means of structural governmental changes in paternal leave policies, role-spillover could have positive effects in allowing more male spouses or cohabiting partners to engage in physical and emotional work by caring for newborns (Glavin and Schieman 2011; McMillan et al. 2011). Such structural 114 country-level changes in care-giving policies carry the potential to move gender role ideology of a country in a more egalitarian direction, challenging the traditional breadwinner model and enabling more men not only to perform ‘female tasks’ within the home environment but to assume management of such housekeeping and care-giving tasks (Sullivan 2000). In this notion of receiving structural governmental support, better tax rates for second income earners would allow more women to leave the home environment to engage in the employment market without incurring higher tax rates for dual-income families. Historically, women have been both assigned the duties of the home environment and disadvantaged in the work environment. To this day, women are facing the ‘double’ and ‘triple shift’ making it especially hard for them to manage work-life balance (Hochschild 1989). While women on an individual level can stand up for more equality in the division of household labor and at the workplace, it is also the society’s responsibility to institutionalize gender equality. A continuous effort has to be made to identify remaining inequalities and to keep them visible until they are addressed by the governmental institutions. Further research studies are part of this continuous effort to identify individual conditions and attitudes within a larger social and economic countrylevel context under which spouses or cohabiting partners divide their paid as well as unpaid household labor and perceive work-life balance (Coltrane 2010). This research study has attempted to point to some of the complexities of gender roles and how remaining inequalities within the division of household labor affect male and female spouses and cohabiting partners in managing their work-life balance. 115 REFERENCES Adler, Emily S., and Roger Clark. 2008. How It’s Done: An Invitation to Social Research. Belmont, CA: Thomson Wadsworth. Apparala, Malathi L., Alan Reifman, and Joyce Munsch. 2003. “Cross-National Comparison of Attitudes Toward Fathers’ and Mothers’ Participation in Household Tasks and Childcare.” Sex Roles 48(5/6):189-203. Baxter, Janeen. 1992. “Power, Attitudes and Time: The Domestic Division of Labour.” Journal of Comparative Family Studies 2:165–82. Beaujot, Roderic, and Jianye Liu. 2005. “Models of time use in paid and unpaid work.” Journal of Family Issues 26:24-946. Beck, Ulrich. 1992. “Risk society: Toward a new modernity.” London: Sage. Bellavia, Gina M., and Michael R. Frone. 2005. “Work-Family Conflict.” Pp. 113-148 in Handbook of Work Stress, edited by Julian Barling, E. Kevin Kelloway, and Michael R. Frone. Thousand Oaks, CA: Sage. Berg, Bruce L. 2007. Qualitative Research Methods for the Social Sciences. 6th ed. Boston, MA: Pearson. Bianchi, Suzanne M., Melissa A. Milkie, Liana C. Sayer, and John P. Robinson. 2000. “Is anyone doing the housework? Trends in the gender division of household labor.” Social Forces 79:191-228. Bianchi, Suzanne M., and Sara B. Raley. 2005. “Time allocation in families.” Pp. 21-42 in Work, family, health, and well-being, edited by Suzanne M. Bianchi, Lynne M. Casper, and Rosalind Berkowitz King. Mahwah, NJ: Erlbaum. 116 Bianchi, Suzanne M., John P. Robinson, and Melissa A. Milkie. 2006. Changing Rhythms of American Family Life. New York: Russell Sage. Blair-Loy, Mary. 2003. Competing Devotions. Cambridge, MA: Harvard University Press. Bond, James T., and Ellen Galinksy. 2011. “Workplace Flexibility In Manufacturing Companies.” National Study of the Changing Workforce. Families and Work Institute. Retrieved January 20, 2011 (http://familiesandwork.org/site/research/ reports/WorkFlexAndManufacturing.pdf). Bondi, Liz. 1991. “Gender Divisions and Gentrification: A Critique.” Transactions of the Institute of British Geographers 16(2):190-219. Bonvillain, Nancy. 1998. Women and Men: Cultural Constructs of Gender. New Jersey: Prentice Hall. Braun, Michael, Noah Lewin-Epstein, Haya Stier, and Miriam K. Baumgaertner. 2008. “Perceived Equity in the Gendered Division of Household Labor.” Journal of Marriage and Family 70:1145-1156. Bureau of Labor Statistics. 2005. Women in the labor force: A databook. Washington, DC: U.S. Department of Labor. Caproni, Paula J. 2004. “Work/life balance: You can’t get there from here.” The Journal of Applied Behavioral Science 40(2):208-218. Clark, Sharon C. 2001. “Work cultures and work/family balance.” Journal of Vocational Behavior 58(3):348-365. 117 Clark, Sue C. 2000. “Work/Family Border Theory: A New Theory of Work/Family Balance.” Human Relations 53:747-70. Coltrane, Scott. 2000. “Research on household labor: Modeling and measuring the social embeddedness of routine family work.” Journal of Marriage and the Family 6: 1208-1233. Coltrane, Scott. 2010. “Gender Theory and Household Labor.” Sex Roles 63:791-800. Creswell, John W. 2003. “A Framework for Design.” Chapter 1 from Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. Sage Publications: Thousand Oaks, CA. Crompton, Rosemary, and Clare Lyonette. 2006. “Work-Life ‘Balance’ in Europe.” Acta Sociologica 49(4):379-393. Curtis Richard. (1986), Household and Family in Theory on Inequality. American Sociological Review 51:168-183. Davis Shannon. 2004. “Is Justice Contextual? Married Women’s Perceptions of Fairness of the Division of Household Labor.” Journal of Comparative Family Studies 41:19-39. Desrochers, Stephan, Jeanne M. Hilton, and Laurie Larwood. 2005. “Preliminary Validation of the Work-Family Integration-Blurring Scale.” Journal of Family Issues 26:442–66. Duyvendak, Jan W., and Monique Stavenuiter. 2004. Working fathers, caring men. Reconciliation of working life and family life. The Hague, the Netherlands: Ministry of Social Affairs and Employment/Verwey-Jonker Institute. 118 Eagly, Alice H., and Valerie J. Steffen. 1984. “Gender stereotypes stem from the distribution of women and men into social roles.” Journal of Personality and Social Psychology 46:735–754. Eagly, Alice H., and Wendy Wood. 1999. “The origins of sex differences in human behavior: Evolved dispositions versus social roles.” American Psychologist 54:408–423. Eisenstein, Zillah. 1979. “Developing a theory of capitalist patriarchy and socialist feminism.” Pp. 5-40 in Capitalist Patriarchy and the case for Socialist Feminism, edited by Zillah Eisenstein. New York: Monthly Review Press. Esping-Andersen, Gosta. 1999. Social Foundations of Postindustrial Economics. Oxford: Oxford University Press. Evans, John M. 2002. “Work/family reconciliation, gender wage equity and occupational segregation: The role of firms and public policy.” Canadian Public Policy, 18(Suppl.):187- 216. Frone, Michael R. 2003. “Work-family balance.” Pp. 143-162 in Handbook of occupational health psychology, edited by James C. Quick, and Lois E. Tetrick. Washington, DC: American Psychological Association. Frone, Michael R., Marcia Russell, and M. Lynne Cooper. 1992. “Antecedents and outcomes of work-family conflict: Testing a model of the work-family interface.” Journal of Applied Psychology 77:65-78. 119 Galinksy, Ellen, James T. Bond, Kelly Sakai, Stacy S. Kim, and Nicole Giuntoli. 2008. “2008 National Study of Employers.” Families and Work Institute. Retrieved January 22, 2011 (http://familiesandwork.org/site/research/reports/2008nse.pdf). Gershuny, Jonathan I. 1995. “Change in the Division of Domestic Work: Microsociological Evidence.” Discussion paper no. 107, Deutsches Institut für Wirtschaftsforschung, Berlin. Gershuny, Jonathan. 1995b. “Time Budget Research in Europe”. Statistics in Transition 24:1-23. Giddens, Anthony. 1984. The Constitution of Society: Outline of the Theory of Structuration. Berkeley: University of California Press. Gillespie, Rosemary. 2003. “Childfree and Feminine: Understanding the Gender Identity of Voluntarily Childless Women.” Gender and Society 17(1):122-36. Glavin, Paul, Scott Schieman, and Sarah Reid. 2011. “Boundary-Spanning Work Demands and Their Consequences for Guilt and Psychological Distress” Journal of Health and Social Behavior 52(1)43-57. Goldin, Claudia. 1990. Understanding the Gender Gap. New York: Oxford University Press. Goode, William J. 1960. A theory of role strain. American Sociological Review 25:483496. 120 Greenhaus, Jeffrey H., Tammy D. Allen, and Paul E. Spector. 2006. “Health Consequences of Work-Family Conflict: The Dark Side of the Work-Family Interface.” Pp. 61-98 in Research in Occupational Stress and Well-Being, edited by Pamela L. Perrewe and Daniel C. Ganster. Amsterdam, the Netherlands: JAI Press/Elsevier. Greenhaus, Jeffrey H. and Nicholas J. Beutell. 1985. “Sources of Conflict between Work and Family Roles.” Academy of Management Journal 10:76-88. Hare-Mustin, Rachel T. 1988. “Family Change and Gender Differences: Implications for Theory and Practice.” Family Relations 37(1):36-41. Hayes, Andrew F. 2013. Introduction to Mediation, Moderation, and Conditional Process Analysis A Regression-Based Approach. Retrieved March 2, 2013 (http://www.guilford.com/cgi-bin/cartscript.cgi?page=pr/hayes3.htm). Hayes, Andrew F., and Kristopher J. Preacher. 2008. “Asymptotic and Resampling Strategies for Assessing and Comparing Indirect Effects in Multiple Mediator Models.” Behavior Research Methods 40(3):879-891. Hochschild, Arlie R. 1989. The Second Shift: Working Parents and the Revolution at Home. Berkeley: University of California Press. Hoffman, Lois W. 1977. “Changes in family roles, socialization, and sex differences.” American Psychologist 32:644-657. Institute for Women’s Policy Research. 2007. “Maternity leave in the United States.” Retrieved October 13, 2010 (http://www.genderprinciples.org/resource_ files/Maternity_Leave_in_the_United_States_Fact_Sheet.pdf). 121 International Social Survey Programme (ISSP). 2004. “International Social Survey Programme: Family and Changing Gender Roles III, 2002.” [Computer file]. ICPSR version. Cologne, Germany: Zentralarchiv für Empirische Sozialforschung [producer], 2004. Cologne, Germany: Zentralarchiv für Empirische Sozialforschung/Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributors], 2004. Katz Elizabeth. 1991. “Breaking the Myth of Harmony: Theoretical and Methodological Guidelines to the Study of Rural Third World Households.” Review of Radical Political Economics 23:37-56. Katz, Ruth, and Yoav Lavee. 2002. “Perceived Fairness, and Marital Quality: The Effect of Gender Ideology.” Journal of Marriage and Family 64(1):27-39. Keuzenkamp, Saskia, Erna Hooghiemstra, Erna. 2000. De kunst van het combineren. Taakverdeling onder partners [The act of combining. Task sharing between partners]. The Hague, the Netherlands: Social and Cultural Planning Office. Lachance-Grzela, Mylene, and Genevieve Bouchard. 2010. “Why Do Women Do the Lion’s Share of Housework? A Decade of Research.” Sex Roles 63:767-780. Lam, Lawrence, and Tony Haddad. 1992. “Men’s participation in family work: A case study.” International Journal of Sociology of the Family 22:67-104. Lewin-Epstein, Noah, Haya Stier, and Michael Braun. 2006. “The division of household labor in Germany and Israel.” Journal of Marriage and Family 68:1147-1164. 122 Lively, Kathryn J., Brian Powell, Claudia Geist, and Lala Carr Steelman. 2008. “Inequity Among Intimates: Applying Equity Theory to the Family.” Justice: Advances in Group Processes 25:87-115. Livingston, Beth A. and Timothy A. Judge. 2008. “Emotional Responses to Work– Family Conflict: An Examination of Gender Role Orientation Among Working Men and Women.” Journal of Applied Psychology 93(1):207-216. Livingston, Mary M., Kim Burley, and Thomas P. Springer. 1996. “The Importance of Being Feminine: Gender, Sex Role, Occupational and Marital Role Commitment, and Their Relationship to Anticipated Work-Family Conflict.” Journal of Social Behavior & Personality 11(5):179-92. Lynch, Karen D. 2008. “Gender Roles and the American Academe: A Case Study of Graduate Student Mothers.” Gender & Education 20(6):585-605. Marks, Stephen R., and Shelley M. MacDermid. “Multiple Roles and the Self: A Theory of Role Balance.” 1996. Journal of Marriage and Family 58(2):417-432. McDowell, Linda, Kath Ray, Diane Perrons, Colette Fagan, and Kevin Ward. 2005. “Women’s paid work and moral economies of care.” Social & Cultural Geography 6(2):219-235. McMillan, Heather S., Michael Lane Morris, and E. Kate Atchley. 2011. “Constructs of the Work/Life Interface: A Synthesis of the Literature and Instruction of the Concept of Work/Life Harmony.” Human Resources Development 10(1):6-25. Mikula, Gerold. 1998. “Division of Household Labor and Perceived Justice: A growing Field of Research.” Social Justice Research 11:215-241. 123 Milkie, Melissa A., and Pia Peltola. 1999. “Playing All the Roles: Gender and the WorkFamily Balancing Act.” Journal of Marriage and Family 61(2):476-490. Moen, Phyllis, Mary A. Erickson, and Donna Dempster-McClain. 1997. “Their Mother's Daughters? The Intergenerational Transmission of Gender Attitudes in a World of Changing roles.” Journal of Marriage and Family 59(2):281-93. Moors, Hein. 1995. “The Netherlands: Great Tolerance but Little Solidarity.” Volume I. in Population, Family and Welfare: A Comparative Survey of European Attitudes, edited by Moors, Hein and Rossella Palomba. Clarendon Press: Oxford, England. OECD. 2004. “Female Labour Force Participation: Past Trends and Main Determinants in OECD Countries.” OECD Economics Department. OECD. 2010. “Employment Rates.” In OECD Factbook 2010: Economic, Environmental and Social Statistics. OECD Publishing. Retrieved August 23, 2011 (http://www.oecd-ilibrary.org/sites/factbook-2010en/06/01/01/index.html?contentType=/ns/StatisticalPublication,/ns/ Chapter&itemId=/content/chapter/factbook-2010-45-en&containerItemId =/content/serial/18147364&accessItemIds=&mimeType=text/html). OECD. 2011. “Employment rate of women.” No. 5 in Employment and Labour Markets: Key Tables from OECD. Retrieved August 23, 2011 (http://www.oecd-ilibrary.org/employment/employment-rate-ofwomen_20752342-table5). Orloff, Ann Shola. 2009. “Gendering the Comparative Analysis of Welfare States: An Unfinished Agenda.” Sociological Theory 27(3):317-343. 124 Peltola, Pia, Melissa A. Milkie, and Stanley Presser. 2004. “The ‘Feminist’ Mystique: Feminist Identity in Three Generations of Women.” Gender & Society 18(1):122144. Portegijs, Wil, Marielle Cloı¨n, Ingrid Ooms, and Evelien Eggink. 2006. Hoe het werkt met kinderen. Moeders over kinderopvang en werk. [How it works with children. Mothers, childcare, and employment]. The Hague, the Netherlands: Social and Cultural Planning Office. Risman, Barbara J. 2004. “Gender as a Social Structure: Theory Wrestling with Activism.” Gender and Society 18(4):429-450. Romero, Mary. 2002. Maid in the U.S.A. New York: Routledge Ruppanner, Leah. 2008. “Fairness and Housework: A Cross-National Comparison.” Journal of Comparative Family Studies 39(4):509-526. Schieman, Scott, Debra Branch McBrier, Karen Van Gundy. 2003. “Home-to-Work Conflict, Work Qualities, and Emotional Distress.” Sociological Forum, 18(1):137-164. Schieman, Scott, Yuko Kurashina, and Karen van Gundy. 2006. “The Nature of Work and the Stress of Higher Status.” Journal of Health and Social Behavior 47:242– 57. Schieman, Scott, Melissa A. Milkie, and Paul Glavin. 2009. “When Work Interferes with Life: Work-Nonwork Interference and the Influence of Work-Related Demands and Resources.” American Sociological Review 74:966-88. 125 Sironi, Maria, Letizia Mencarini, and Carlo F. Dondena. 2010. “Happiness, Housework and Gender Inequality in modern Europe.” Retrieved September 4, 2013 (http://www.bing.com/search?q=Sironi%2C+Maria%2C+Letizia+Mencarini%2C +and+Carlo+F.+Dondena.+2010.+%E2%80%9CHappiness%2C+Housework%09 and+Gender+Inequality+in+modern+Europe.%E2%80%9D+Political+Science&f orm=IE10TR&src=IE10TR&pc=MDDCJS). Smithson, Janet, and Elizabeth H. Stokoe. 2005. “Discourses of work-life balance: Negotiating ‘genderblind’ terms in organizations.” Gender, Work and Organization 12(2):147-168. Sullivan, Oriel. 2000. “The Division of Domestic Labour: Twenty Years of Change?” Sociology 34(3):437-456. Retrieved August 3, 2011 (http://club.fom.ru/books/sullivan_O.pdf). The White House. 2010. “A Win for Women, Mothers and Working Families.” Council on Women and Girls. Retrieved August 11, 2011 (http://www.whitehouse.gov/blog/2010/12/10/a-win-women-mothers-andworking-families). Thoits, Peggy A. 1991. “On merging identity theory and stress research.” Social Psychology Quarterly 54(2):101-112. Treas, Judith, and Eric Widmer. 2000. “Married Women’s Sense of Fairness.” Journal of Family 12:181-196. Tronto, Joan C. 1993. Moral Boundaries: Political Argument for an Ethic of Care. New York: Routledge. 126 United Nations Human Development Programme. 2007. “Human Development Report 2007/2008. Fighting Climate Change: Human solidarity in a divided world.” Retrieved September 10, 2011 (http://hdr.undp.org/en/media/HDR_20072008_EN_Complete.pdf). United Nations Human Development Programme. 2010. “The Real Wealth of Nations: Pathways to Human Development.” Retrieved September 10, 2011 (http://hdr.undp.org/en/media/HDR_2010_EN_Complete_reprint.pdf). Voydanoff, Patricia. 2004. “The effects of work demands and resources on work-tofamily conflict and facilitation.” Journal of Marriage and Family 66(2):398-412. Waldfogel, Joel. 1996. The impact of the family and medical leave act on coverage, leave-taking, employment, and earnings. New York: Mimeo. Walster, Elaine, G. William Walster, and Ellen Berscheid. 1978. Equity: Theory and Research. Boston: Allyn & Bacon. Walzer, Susan. 1996. Thinking about the Baby: Gender and Transitions into Parenthood. Philadelphia, PA: Temple University Press. West, Candace, and Don Zimmerman. 1987. “Doing Gender.” Gender and Society. 1: 125-151. Wierda-Boer, Hilde, Jan Gerris, Ad Vermulst, Kaisa Malinen, and Karen Anderson. 2009. “Combination strategies and work-family interference among dual-earner couples in Finland, Germany, and the Netherlands.” Community, Work & Family 12(2):233-249. 127 Wilcox, Clyde, and Ted G. Jelen. 1993. “Catholicism and opposition to gender equality in Western Europe.” International Journal of Public Opinion Research 5:40-57. Zaretsky, Eli. 1973. “Capitalism, the family, and personal life.” Socialist Revolution 13/14:69-125.