Cross-national reports of housework: An investigation of the gender empowerment measure

Abstract

This paper analyses the relationship between country-level gender empowerment and individual-level divisions of housework. Pairing the 2004 United Nations gender empowerment measure (GEM) with individual-level (n = 18,560) data from the 2004 European Social Survey, the author compares the relationship between a country’s GEM score, both as an index and as disaggregated measures, and respondents’ housework hours and housework proportions. The GEM index has a positive and linear relationship with men’s housework hours and a positive and non-linear relationship with men and women’s housework proportions and with women’s housework hours. For the disaggregated GEM measures, women’s representation in parliament is positively associated with men’s housework hours and proportions and women’s housework hours. Women’s labor market status, including the percent of women in professional positions and female–male wage ratios, is negatively associated with women’s housework hours and proportions. Finally, the cross-level interactions demonstrate theoretically important relationships to the housework literature.

Introduction

The persistence of unequal divisions of housework between spouses has intrigued scholars for decades. At the individual-level, theoretical explanations of spouses’ unequal housework have centered on spouses’ resources (Becker, 1991, Brines, 1994, Lennon and Rosenfield, 1994) and on gender role expectations (Becker, 1991, Berk, 1985, West and Zimmerman, 1987). Recent scholarship has emphasized cross-national comparisons of spouses’ unequal housework divisions (Batalova and Cohen, 2002, Baxter, 1997, Bittman et al., 2003, Cooke, 2006, Fuwa, 2004, Fuwa and Cohen, 2007, Hook, 2006, Geist, 2005). From this body of comparative research, many macro-level characteristics have been shown to be significantly associated with housework, including the gross domestic product (Knudsen and Wærness, 2008, Fuwa, 2004), rates of female labor force participation (Batalova and Cohen, 2002), the skill specificity required for employment (Iversen and Rosenbluth, 2006), the divorce culture (Yodanis, 2005), and the welfare state (Geist, 2005, Fuwa, 2004). Throughout much of this cross-national multi-level research, one macro–micro relationship consistently emerges: couples living in countries with stronger gender empowerment, measured through the United Nations Development Report’s gender empowerment measure (GEM), have more equal divisions of housework (Batalova and Cohen, 2002, Hank and Jürges, 2007, Knudsen and Wærness, 2008, Fuwa, 2004).

The significant relationship between country-level gender empowerment and couples’ divisions of household labor is an important contribution to the housework literature and to understanding gender equality more generally, but the GEM has been used relatively uncritically. The GEM is a composite measure of four indicators: women’s representation in parliament; the percent of women as legislators, senior officials and managers; the proportions of women employed as professionals; and the female–male wage ratio. As a composite index, this measure has been shown to be significantly correlated with couples’ divisions of household labor (Batalova and Cohen, 2002, Hank and Jürges, 2007, Knudsen and Wærness, 2008, Fuwa, 2004), but certain components of the GEM may be more important predictors of spouses’ housework hours and proportions than others. For example, female parliamentarians are more likely to pass legislation, including paternal leave, that encourages male participation in the home (Swers, 1998). In addition, women in countries with lower female–male wage gaps may use their resources to encourage a more equal division of housework. In this respect, understanding the separate components of the GEM is important to understanding individual divisions of housework.

Disentangling the separate country-level effects can have important theoretical and policy implications – not all forms of gender empowerment may be equal. Referring to the GEM index, Fuwa (2004) argues, “Research that differentiates these specific aspects of macro-level gender inequality will improve our understanding of contextual effects in relation to public policies.” Most previous comparative housework research on the GEM index has utilized one dataset (the International Social Survey Programme) and has focused exclusively on the relationship between GEM and couples’ divisions of housework tasks (Batalova and Cohen, 2002, Fuwa, 2004) at the expense of each spouses’ hours (see Knudsen and Wærness, 2008 for an exception). In this respect, the significant relationships between GEM and housework hours may hinge on both the data and the dependent variable. To address these limitations, this study applies a new dataset, the 2004 European Social Survey (ESS), and expands the analyses to include the respondents’ housework hours and housework proportions. For this study, the housework proportion measure is a ratio of the total weekly housework the respondent performs to the total weekly household housework. Specifically, this research asks: how does the effect of the GEM compare to the effects of the separate GEM measures on respondent’s housework hours and housework proportions?

To address this question, I pair the 2004 United Nations Development Report’s GEM scores with individual-level data from the 2004 European Social Survey for respondents in 25 countries. To address simultaneously individual-and country-level effects, I analyse the data using hierarchical linear modeling (HLM). Since the division of household labor is a highly gendered process, I run the models separately for men and women to compare the relationship between the aggregated and disaggregated GEM scores and the respondents’ housework hours and proportions. I then test for significant cross-level interactions for three strategically selected individual-level measures. The results demonstrate complex relationships between the level-1 and level-2 variables for men and women.