Promoting Patriarchy or Dual Equality? Multiculturalism and the Immigrant Household Division of Labor

ABSTRACT

In this article, we provide the first empirical analysis of the relationship between multicultural immigration policy and gender inequality within immigrant communities. A fierce sociological debate pits those who identify multiculturalism as a key obstacle to gender equality among immigrant families against those who believe multiculturalism and gender egalitarianism are “dual-equality projects.” At the core of this debate are differences over the extent to which multiculturalism impedes or promotes the transmission of gender egalitarianism from host societies to immigrant communities. To adjudicate between these two perspectives, we examine whether microfoundations of the household division of labor—relative resources, time availability, and gender ideology—reduce the share of women’s labor to a greater or lesser degree in multicultural countries. We find multiculturalism increases the egalitarian effects of microfoundations among immigrant households. Both symbolic and material forms of multiculturalism contribute to its moderating effect, but immigrant women benefit the most in countries with both types of multiculturalist policies. In highly multicultural countries, rising incomes, greater employment, and more egalitarian gender ideologies can produce dramatic reductions in housework for immigrant women. We conclude by specifying the conditions where multiculturalism reduce inequalities between immigrants and natives, and within immigrant communities.

KEYWORDS:

• Culture
• immigration
• inequality
• multiculturalism

Acknowledgments

The authors thank David Brady, Lena Hipp, Matt Huffman, Augustine Kposowa, Ruud Koopmans, Ellen Reese, Jan Stets, and participants of the Poverty, Inequality and Health Workshop, for helpful comments.

Notes

1. By South–North migration, we refer to migration between developing countries in the global South to highly developed countries in the global North.

2. Norway offers migrants a lesson in how to treat women (Higgins 2015).

3. The influence of relative resources and time availability on the HHDL can be controversial (see Bianchi et al. 2000; Hochschild and Machung 1989). However, meta-analyses across quantitative studies find that the coefficients of micro-level determinants remain largely consistent (Davis and Wills 2014).

4. While the HHDL does not capture every dimension of gender inequality, it has been shown to be consequential for inequalities between women and men, even in arenas outside the home (Cohen 2004; Cohen and Huffman 2003; Ridgeway 2011).

5. Prior studies highlight the critical importance of other state policies including child-care services and gender affirmative action in shaping egalitarian outcomes in the HHDL (e.g., Fuwa and Cohen 2007).

6. These arguments are consistent with our own data, which show that immigrants hold less gender-egalitarian ideologies than natives (on average), even when controlling for a host of sociodemographic and country variables (available upon request). Nevertheless, it is important to acknowledge that Western societies are far from perfectly gender egalitarian (e.g., Coltrane 1997; Kimmel 2000; Schafer, Caetano, and Clark 1998). Moreover, the roots for North–South differences in gender egalitarianism are tied to variation in economic development, of which unequal global power relations between the North and South are a key explanation (Mahutga and Smith 2011).

7. Decoupling—that is, gaps between formal policy and actual practice—is most often theorized at the organizational level. However, parallel processes of “identity decoupling” have been observed at the individual level (DeJordy 2008). Dasgupta (1998) finds similar results of “judicious biculturalism” among Asian-Indians in the United States, whereby immigrants actively adjust and negotiate their behaviors and identities across ethnic and native spaces.

8. It is possible that respondents live on their own instead of with their partners, which would necessarily lead to scores of either 0 or 4. However, the majority of cases are a mix of tasks done by the husband and wife. We observe 6.69 percent of cases where husbands of immigrant respondents do all the tasks (i.e., scores of 4). Only 0.56 percent of households with immigrant respondents report a division of labor where the wife completes all tasks (i.e., scores of 0). Removal of these cases does not substantively alter the results.

9. Although several countries participate in all relevant waves, only countries with immigrant respondents were included.

10. Time diaries undoubtedly provide more accurate estimates of total hours worked on different types of tasks than retrospective surveys, but we are unaware of any cross-nationally comparative data sets employing them that approach the coverage of the ISSP. Some suggest that retrospective survey questions asking how many hours of housework husbands/wives perform give more detailed estimates of the relative amount of household work done by husbands and wives (Lachance-Grzela and Bouchard 2010). The ISSP data include retrospective accounts of the absolute amount of work done by each partner in waves 2002 and 2012, but not in 1994. Because our question requires maximal observations of multiculturalism (country-years) and immigrants, we use relative over absolute measures of the HHDL—the 1994 data account for roughly a third of our sample.

11. Previous work suggests that the division of household labor may be more equal in households in which the husband is native (Blau, Kahn, and Papps 2011). Unfortunately, the ISSP only captures the immigrant status of the respondent, which is based on reported national origin. However, this interaction likely captures the primary mechanism linking egalitarian divisions of labor to the native status of husbands (gender ideology).

12. Although the age and number of children would be the ideal measure to include in the model, prior surveys from ISSP 1994 do not directly ask the respondents for the number of children. Rather, respondents are asked “Did you work outside the home when a child was under school age?” Available responses included “yes (worked full-time),” “yes (worked part-time),” “no (stayed at home),” or “does not apply.” We coded any of the first responses as 1 (have children) and the last response as 0 (no children).

13. On average, countries within our pooled sample scored 4.33 on the multicultural scale. The measure can be accessed at http://www.queensu.ca/mcp. This measure necessarily averages across regions within countries that may vary greatly in their degree of gender egalitarianism, which should attenuate any moderating effects of multiculturalism that we observe.

14. Values ranged from 0 to 100 and higher scores on the GEM index indicate greater gender empowerment. The index score of GEM is based on the average score across four equivalent percentage indexes: (1) the male and female percentage share of parliamentary seats; (2) the male and female shares of positions as legislators, senior officials, and managers; (3) the male and female shares of professional and technical positions; and (4) women’s and men’s estimated earned income, to capture multiple dimensions of sociopolitical power.

15. Results are substantively identical using HLM estimators, which leverage between-country variation.

16. For example, estimating a model including interactions between multiculturalism and all three micro-level determinants simultaneously would necessitate controlling for all eight of the constituent terms in Table 3, as well as the following “tacit” interactions: Relative resources × gender ideology, relative resources × time availability, gender ideology × time availability, relative resources × gender ideology × time availability, relative resources × gender ideology × time availability × multiculturalism (Braumoeller 2004).

17. Models replacing immigrant stock with immigrant flow (measured as net migration as a percent of the total population) produced similar results and are available in Table 1A (Models 1–3) in the Appendix.

18. Measuring time availability as full-time employment and part-time employment (reference group: unemployed) produce substantively similar results, which can be found in Table A1 of the Appendix (Models 4–6).

19. One anonymous reviewer wondered whether multiculturalism promotes gender egalitarianism among natives. In the sample of natives for which complete data are available, we observe a positive but nonsignificant partial association between multiculturalism and gender-egalitarian ideology. We also observe nonsignificant effects of multiculturalism on the slope of the microfoundations. Multiculturalism appears to play a more important role in the gender performance of immigrant households than those of natives.

20. To be sure, the distinction between “material” and “symbolic” is not entirely sharp, since many policies that provide explicit material support (e.g., funding) are in and of themselves symbolic. Nevertheless, policies 6–8 involve explicit financial commitments of the state, while the others do not.

21. The jackknife point estimate is simply the mean of the coefficients estimated across each jackknife sample. Jackknife standard errors are equal to , where θ is the coefficient calculated with the ith jackknife sample, and k is the number of jackknife samples.

Additional information

Notes on contributors

Ronald Kwon

Ronald Kwon is a Ph.D. candidate at the University of California, Riverside.

Matthew C. Mahutga

Matthew C. Mahutga is Associate Professor of Sociology at the University of California, Riverside. His research examines global and institutional determinants of economic organization and various types of inequality. His work appears in interdisciplinary outlets including Europe-Asia Studies, Global Networks, Review of International Political Economy, Social Forces, Social Networks, Social Problems, Social Science Research, Urban Studies, and elsewhere, and has been supported by the National Science Foundation.

Amanda Admire

Amanda Admire is a Ph.D. candidate at the University of California, Riverside and a National Science Foundation Fellow. Her research examines gender inequality, policing, and intimate partner violence.