Intersectionality and Family-Friendly Policies in the Federal Government

Intersectionality Theory

Intersectionality discourse has burgeoned since it was first introduced by Kimberlé Crenshaw in 1989. Crenshaw illustrated through a legal analysis that race and gender are treated as mutually exclusive categories in antidiscrimination law; she further illustrated that this tendency was prevalent historically in feminist theory as well. As a result, Black women are “erased” in terms of social identity. She argues that this

single-axis framework erases Black women in the conceptualization, identification and remediation of race and sex discrimination by limiting inquiry to the experiences of otherwise-privileged members of the group. In other words, in race discrimination cases, discrimination tends to be viewed in terms of sex- or class-privileged Blacks; in sex discrimination cases, the focus is on race- and class-privileged women. This focus on the most privileged group members marginalizes those who are multiply-burdened and obscures claims that cannot be understood as resulting from discrete sources of discrimination. (Crenshaw, 1989, p. 140)

It was an excellent analysis that opened doors for study especially in feminist theory and identity politics. Black feminists who were locked out of the “mainstream” feminist movement now had a conceptual framework to theorize and examine how multiple identities create multiple facets of inequalities or oppression. Much of the published work here dissects the intricacies associated with the term intersectionality and its application to feminist inquiry. For example, Davis (2008) points out, “It is not at all clear whether intersectionality should be limited to understanding individual experiences, to theorizing identity, or whether it should be taken as a property of social structures and cultural discourses” (p. 68; also see, Crenshaw, 1991; Ferree, 2009; Staunæs, 2003; Yuval-Davis, 2006). In effect, she asks, how can a theory so vague be considered as cutting edge feminist theory? She argues that the ambiguity of the concept is what has fostered its success, because it has allowed for different applications to a phenomenon that eschews sameness and a one-size-fits-all approach. This certainly speaks volumes to the fact that different disciplines, especially in the social sciences, will adapt intersectionality in a manner that comports with their particular ethos. As Davis (2008) argues, a successful social theory is one that addresses the interests and concerns of its primary audience.

Along the same lines, many feminist and women’s studies scholars have grappled with the issue of suitable methodologies for studying intersectionality (see Ferree, 2009; Hancock, 2007; Nash, 2008; Simien, 2007). McCall (2005), for example, argues that researchers in different disciplines favor different methodologies for studying intersectionality, which produces different kinds of knowledge. She advocates for a wide range of methodological approaches to capture the “multiple, intersecting, and complex social relations” which characterize intersectionality (McCall, 2005, pp. 1772-1773). This, too, has implications for disciplinary values and suitability. Of course, it is important to stress that in all of these feminist treatises the underlying premise is that women of color see things differently due to the intersection of such factors as race and gender.

Studies by political scientists have examined the importance of intersectionality on political behaviors and equal treatment and have found that women of color are treated and perceive fairness differently (see Hedge, Button, & Spear, 1996; Simien, 2006, 2007; Smooth, 2006; Orey, Smooth, Adams, & Harris-Clark, 2007; Wilson, 2013). For example, Hawkesworth (2003) examines legislative practices during the 103rd and 104th Congresses to show that women of color are treated differently, indeed, marginalized in floor debates, committee operations, and interpersonal interactions. They in effect, become subordinate members of Congress. Their marginalization, in turn, produces a sense of frustration and anger, which then affects their policy priorities and intensity of behavior. Hawkesworth (2003) demonstrates these traits and experiences in the case of welfare reform “motivated . . . Congresswomen of color to legislate against the grain, enacting a politics of resistance that ranged from open confrontation on the House and Senate floor to the mobilization of anger in their constituencies” (p. 546).

The topic of intersectionality has been virtually absent from public administration discourse, as Bearfield (2009) and others have pointed out (see Hutchinson, 2005; Stivers, 2002). ¹ Bearfield (2009) argues that “the field must engage in research focused on the intersection of multiple identity categories such as gender and race” (p. 384). He goes on to say that research on intersectionality challenges “the idea that race, gender, and class form independent analytic categories” (p. 384). As he notes, a number of scholars in public administration have sought to promote feminist theory in the field. Most prominently, Hutchinson (2005) has called attention to this topic, asking,

Public Administration theorists, women and men, to consciously consider feminisms as a serious approach to theorizing our publics. Feminisms are not wild hair theories included in conferences and symposia to assuage the tempers of a radical few or at least I’d like to think not. (p. 359; also see Hutchinson, 2001, 2002; Hutchinson & Mann, 2004, 2006)

Stivers (2000, 2002, 2005), too, has called upon the field of public administration to focus its attention on the importance of feminist theory (also see Burnier, 2005; Rusaw, 2005; Shields, 2005). She argues that the preeminence of the male lens or framework in studying public administration has completely distorted the conception of reality in public organizations. Scholars must consider how feminism will facilitate our ability “to think differently about central questions in the field” (Stivers, 2005, p. 365).

Brenner’s (2009) study is one of the few in public administration that examines the issue of intersectionality. ² She examines the effects of Latina administrators’ role construction on their policy intentions. Through interviews with 16 Latina administrators, Brenner illustrates how ethnicity and gender interact to affect their policy goals. She found that three role orientations emerged for these city leaders—activists, bridge builders, and institutionalists. For example, Latina activists were very instrumental in combating domestic violence. She points out that the Latina activists addressed “issues affecting the lives of Latinas marginalized by violence, a traditional feminist concern. Activist Latinas were not constrained by existing bureaucratic norms” (Brenner, 2009, p. 844). However, Latina bridge builders serve as “cultural brokers” and rely on coalition strategies to assist Latina citizens in the community; Latina institutionalists follow organizational norms, and hence, their policy intentions comport with those of the institution. She concludes that “race and ethnicity served as their primary lens” but role construction, too, affected the policy behaviors of these Latina administrators (p. 847).

Women and Work–Life Balance

Balancing the demands of work and family is a concern for both women and men. However, given that women generally carry a disproportionate share of the responsibility for child and elderly care, the issue is particularly important to women workers. Family-friendly policies have garnered much attention in the public administration literature (see, for example, Durst, 1999; Ezra & Deckman, 1996; Facer & Wadsworth, 2008; S. Kim, 2001). ³ J. Kim and Wiggins (2011) in their insightful study examine federal employees’ satisfaction with family-friendly policies in 2004; they compare women with men. They found, for example, that women were more satisfied than men with alternative work schedules (AWS) and work–life programs but were less satisfied with child care programs as compared with men.

However, work–life balance is experienced differently by White women, compared with women of color because they face different challenges. Consider, for example, that 55% of African American children are living in single-parent homes versus just 19.9% of White children and 31% of Latino children (U.S. Census Bureau, 2013). Most of these single-parent homes are led by women. Fifty-one percent of African American children are living in female single parent-households; 18.7% for White children and 27.5% for Latino children (U.S. Census Bureau, 2014). In effect, the responsibility for child care is vastly different for women of color compared with White women. Also, African American women return to work sooner than White women after having children (U.S. Centers for Disease Control and Prevention, 2010, p. 332). In addition, the employment participation rates for African American women have historically been higher than that of any other racial or ethnic group (U.S. Bureau of Labor Statistics, 2013a). Latina and African American also earn less than both White and Asian American women. For example, in 2012, the median weekly earnings for White women was US$710; for African American women it was US$599; for Latina, US$521, and for Asian American women, US$770 (U.S. Bureau of Labor Statistics, 2013b). As many have pointed out, these patterns of inequality exist for women federal employees as well (see, for example, Guy, 1993; Riccucci, 2009; Smith & Monaghan, 2013).

Given the differences in pay structure, day or elderly care is often cost prohibitive for women of color. As such, they must find alternative ways to balance the demands of work and family. Roberts (2002) points out that “kinship care” is an important tradition for African American women, who rely on women-centered, communal child care (also see, for example, Bhavnani, 1994; Davidson, 1997). She notes that African American women rely on “cooperative networks [which] have included members of the extended family (grandmothers, sisters, aunts, and cousins) as well as nonblood kin and neighbors” (p. 279).

Similarly, Finley-Hervey and Wooten (2004) point out that the experiences of African American women are different from White women’s; African American women rely on “the family values of their African-American cultural heritage” (p. 62). In their study of university professors, they found that African American women were influenced by their cultural heritage and relied on “kinship network systems” for support when managing child or elderly care. Cultural values of African American women, they note, include respecting and caring for your elders. In addition, their support networks include church, book clubs, sororities, extended family, and friends. They attribute this to “collectivism values” within African American communities, which ultimately assists in the balancing of work and family demands.

Researchers at the Center for Women Policy Studies conducted a survey of 1,600 women of color at all different levels in private industry to determine how gender—intersected with race/ethnicity—affected their perceptions of workplace issues including their family lives. They found that the

treatment of women of color in the workplace has a profound effect on women’s ability to strike a balance between fulfilling work responsibilities and satisfying personal obligations and needs. Women of color report that limited career advancement opportunities and perceived discrimination based on race and gender, as well as excessive workloads and management expectations, have a negative impact on their work and family balancing act. (Tucker, Wolfe, Viruell-Fuentes, & Smooth, 1999, p. 24) ⁴

In sum, women of color experience and utilize family-friendly policies differently than White women. Thus, we expect to see differences in how Latina and African American women in the federal government perceive the fairness of programs aimed at work–life balance.

Data and Method

This exploratory study relies on data provided by the 2014 FEVS. The FEVS is a self-administered, needs-based web survey that measures employees’ perceptions of conditions within their agencies that contribute to the organization’s overall success. The Office of Personnel Management (OPM) and agency leaders are able to use responses to help develop policy and plan actions to improve organizational performance. The survey response rate was 46.8% (392,700 respondents from the 781,047 employees contacted). Items on the survey include 14 demographic question and 84 indicators of employees’ perceptions of how their agencies manage their workforce. Eight main topic areas were focused on in the 2014 FEVS: personal work experiences, work unit, agency, supervisor/team leader, leadership, satisfaction, work/life programs, and demographics. Some limitations that should be noted in the publicly available data are that some demographic data are combined to protect the identities of respondents. The original intention of this study was to compare women of color with all women and with White women. Data broken down to that level are not made publically available so we only have respondents’ self-identification as minority or non-minority for our analysis. Lumping all races into a “minority” category, however, is a serious shortcoming of the FEVS (see, for example, Fernandez, Resh, Moldogaziev, & Oberfield, 2015).

To determine a measure of work–life balance, we created an index variable for respondents’ satisfaction with work–life programs in their agency. Work–life programs in this index include telework, AWS, health and wellness programs (e.g., exercise, medical screening, quit smoking programs), employee assistance programs (EAP), child care programs (e.g., day care, parenting classes, parenting support groups), and elder care programs (e.g., support groups, speakers). Responses are on a 0 to 5 scale, 0 meaning do not know, 1 = very dissatisfied, 2 = dissatisfied, 3 = neither satisfied nor dissatisfied, 4 = satisfied, and 5 = very satisfied.

It is important to note that the use of minority as a grouped category can be somewhat misleading. The minority category consists of African American, Hispanic, American Indian/Alaska Native, and Asian American/Pacific Islanders is not an option because of the small sizes of the group and need to protect individual identity. African Americans are the largest group in the minority category accounting for double of any other ethnic group.

Of the respondents in the survey, 44.3% identified as male and 49.6% identified as female (see Table 1). The only race variable available to us is whether or not the respondent is minority or non-minority. In our analysis, 49.4% of respondents are identified as non-minority and 42.1% as minority. An interaction term is created to determine the effects of race and gender on individuals’ work–life balance satisfaction.

OPEN IN VIEWER

Table 1.

Descriptive Statistics of Variables.

Variables	n	M	SD	Minimum	Maximum
Supervisor	364,683	1.208	0.406	1	2
Gender	360,446	1.481	0.5	1	2
Minority status	349,746	1.656	0.475	1	2
Education	366,570	2.023	0.811	1	3
Pay category	366,969	1.742	0.735	1	3
Federal tenure	367,674	2.226	0.78	1	3
Retirement intention	363,391	1.751	0.432	1	2
Disability	363,796	1.865	0.342	1	2
Age	385,428	2.451	0.989	1	4
Military	362,820	1.707	1.164	1	4
Work–life satisfaction	284,786	3.862	1.177	0	5
Race and gender interaction	346,158	2.422	1.056	1	4

Other independent variables of interest include supervisory status, pay category (1 = General Schedule (GS) 7-12; 2 = GS 13-15; 3 = other), planning to retire (1 = within 5 years; 2 = not within 5 years), and age group (1 = below 40; 2 = 40-49; 3 = 50-59; 4 = 60 and older). Education is the respondent’s highest level of education (1 = education prior to bachelor’s degree; 2 = bachelor’s degree; 3 = post-bachelor’s degree). Military and disability status is also reported.

We do not intend to present a study that limits race and gender to single binary terms. In this analysis of work–life balance, we are taking steps to see how race or gender influences perception on a large scale and understand how some addition of qualitative analysis would compliment this work. We recognize that some scholars, especially feminist and antiracist theorists, object to the development of analytic categories to construct multiple social identities (see, for example, McCall, 2005; Nash, 2008; Wilkinson, 2003). Nonetheless, as different disciplines seek to study and conceptualize intersectionality in accordance with their own relevant epistemic traditions, we offer a composite measure that can set a foundation for continued development.

Findings

Data were analyzed using ordinary least squared regression analysis. In our base model, we find that minority status, education, and time to retirement were all found to be statistically significant and related to work–life satisfaction (see Table 2). Non-minorities were found to more likely have higher levels of work–life balance program satisfaction (p < .01). As an individual’s education level increases, the level of work–life program satisfaction increases (p < .01). As respondents get closer to 5 years of retirement, they also have increased levels of work–life program satisfaction (p < .05).

OPEN IN VIEWER

Table 2.

Work–Life Policy Satisfaction and Gender.

Variables	Work–life policy satisfaction
Gender	0.001 (0.005)
Minority status	0.014*** (0.005)
Education	0.009*** (0.003)
Pay category	−0.002 (0.003)
Supervisor	0.001 (0.006)
Federal tenure	0.002 (0.004)
Retirement intention	0.017** (0.007)
Disability status	−0.004 (0.007)
Age	−0.026*** (0.003)
Military	−0.000 (0.002)
Constant	3.861*** (0.029)
Observations	241,765
R 2	.001

Note. Standard errors in parentheses.

*

p < .1. **p < .05. ***p < .01.

When we introduce an interaction term between race and gender, we find very similar results (see Table 3). We find the interaction between race and gender does have a negative relationship with respondents’ satisfaction with work–life program satisfaction (p < .1). Non-minorities are also found to have higher satisfaction with work–life policies (p < .05). Individuals that are both female and minority have lower satisfaction with federal work–life balance programs. As in the first model, both education and proximity to retirement are positively related to work–life balance satisfaction (p < .01 and p < .05, respectively).

OPEN IN VIEWER

Table 3.

Work–Life Policy Satisfaction and Gender (Interaction).

Variables	Work–life policy satisfaction
Race and gender interaction	−0.016 (0.010)*
Gender	0.028 (0.018)*
Minority status	0.038** (0.016)
Education	0.009*** (0.003)
Pay category	−0.002 (0.003)
Supervisor	0.001 (0.006)
Federal tenure	0.002 (0.004)
Retirement intention	0.017** (0.007)
Disability status	−0.004 (0.007)
Age	−0.026*** (0.003)
Military	−0.000 (0.002)
Constant	3.821*** (0.039)
Observations	241,765
R 2	.001

Note. Standard errors in parentheses.

*

p < .15. **p < .05. ***p < .01.

We acknowledge the low R² for both models and believe some of the data limitations are behind the low numbers. R-square is important for statistical significance but does not always inform the substance of the findings (Lewis-Beck & Skalaban, 1990). We would expect more detailed demographic data would lead to further explanation of work–life balance program satisfaction. Participation and awareness of work–life balance programs would also contribute to success rates. There is no way for us to know if participation and satisfaction may be limited to organizational promotion of work–life balance programs. Due to the exploratory nature of this study and the use of secondary data, we have imposed limitations on this analysis.

Conclusion and Discussion

The importance of workers’ ability to successfully balance their work and family lives cannot be overstated. It certainly has significance for both individual and organizational outcomes (see Özbilgin, Beauregard, Tatli, & Bell, 2011). Work–family conflicts can lead not only to job dissatisfaction (see Bruck, Allen, & Spector, 2002; Ernst Kossek & Ozeki, 1998; Grandey, Cordeiro, & Crouter, 2005) but also to low organizational commitment and attachment (see Thompson, Beauvais, & Lyness, 1999) and to employee burnout (see Westman, Etzion, & Gortler, 2004). Thus, it is in the best interest of organizations to develop family-friendly policies. However, as Özbilgin et al. (2011) state, to genuinely improve work–life benefits, organizations must consider the need of their employees. They argue that “if initiatives are designed and implemented without considering the needs and household structures of large proportions of the population, those initiatives will necessarily be less than effective” (p. 178).

The U.S. Equal Employment Opportunity Commission (2010) reported that of the 2.6 million individuals employed in the federal sector, 5.9% are Asian American and Pacific Islanders. This is helpful for our analysis as most of the intersectionality literature focuses on African American women. As we begin to talk about the influence of work–life balance in employee satisfaction, groups within the minority category could have very different experiences. They also suggest there is a need for a more inclusive definition of diversity to address the needs of different groups and the types of challenges they may face in the workplace and community (The U.S. Equal Employment Opportunity Commission, 2010). Research on other women of color and work opportunities is rapidly frowning. Flippen (2014) found immigrant Latina women have interrelated constraints on employment usually stemming from a culture in which wives and mothers disproportionally shoulder the household responsibilities. Asian immigrant women were also found to be mostly employed in low-paid and socially devalued jobs that have lingering effects on future employment opportunities. Chun, Lipsitz, and Shin (2013) advocate of an intersectional approach to understanding Asian women and their workforce challenges.

Although this study is exploratory in nature, it does begin to illustrate the significance of intersectionality between race and gender for satisfaction with efforts to balance work and life responsibilities. Our findings show that minority women are less satisfied with federal work–life balance programs. If there are deficiencies in the availability and access to family-friendly policies, organizations may find certain groups are disproportionately affected by the policy outcomes. For example, if minority women have jobs that are not available for telework and compressed work schedules, they could be systematically excluded from family-friendly policies.

One of the findings that held across models was proximity to retirement and increased satisfaction with work–life policies. As baby boomers continue working past retirement age, they are often faced with dual responsibilities of caring for parents and children or grandchildren. Future work in this area could be promising and help us better understand why people may be able to work longer and still balance responsibilities that come with an aging workforce.