An Intersectional Analysis of Women’s Social Role Engagement and Mental Health

Abstract

Objective:

Guided by role theory and the intersectionality framework, this study assesses whether social role volume, role type, and role configuration influence the mental health of Non-Latina White, African American, Afro-Caribbean, Mexican, Cuban, Puerto Rican, Chinese, Filipina, and Vietnamese American women.

Background:

Contemporary shifts in the primary roles (i.e., worker, spouse, parent) women occupy and in the ethnic composition of the U.S. necessitate a re-examination of how roles impact U.S. women’s mental health. Moreover, family member and friend roles are relatively understudied.

Method:

Drawing data from the nationally representative Collaborative Psychiatric Epidemiology Surveys (CPES) (N=7370), ordinary least squares (OLS) regression analysis is used to assess the relationship between role volume, role type, role configuration, and mental health for women across nine ethnic groups. We report ethnicity-stratified models.

Results:

On one hand, role accumulation was psychologically beneficial for Non-Latina White, African American, Puerto Rican, and Chinese women. On the other hand, the psychological benefits of social roles diminished after accumulating three social roles for Cuban, Mexican, and Filipina women. The psychological influence of specific roles and role configurations for women was dependent on ethnicity.

Conclusion:

This study demonstrates the powerful impact of ethnicity on social role engagement as well as the influence of such roles on women’s psychological health.

Two notable changes in the U.S. over the past sixty years include women’s engagement in social roles (e.g., worker, spouse, parent) and growing racial and ethnic diversity. Some roles increased (e.g., employment) and others decreased (e.g., marriage) in participation. Industrialized nations experienced increased rates of marital dissolution, employment among mothers, and single motherhood (McLanahan, 2004). Various societal forces contribute to changes in women’s roles, such as egalitarian gender ideologies (McLanahan, 2004), women’s increasing educational attainment (Livingston & Cohn, 2010), and the Great Recession (Sobotka et al., 2011). In addition, the U.S. has experienced tremendous racial and ethnic diversification. Throughout the text, “race” refers to a socially constructed set of categories historically based on physical features such as phenotype, hair texture, and facial features and is rooted in the subordination of non-European people (D. T. Williams, 2021). “Ethnicity” is group membership adopted by individuals based on similarities in culture or nationality (Jenkins, 1994) and refers to subgroups within broad racialized groups (e.g., Asian American) who share a common national origin (e.g., Vietnamese). Immigration from Latin American, Caribbean, and Asian countries has transformed the U.S. demographic landscape, with Blacks, Latinxs, and Asians representing 13%, 19%, and 6% of U.S. adults, respectively (U.S. Census Bureau, 2019).

Given shifts in women’s role engagement and burgeoning ethnic diversity, how social roles influence psychological well-being is likely contingent upon women’s racial and ethnic identities. Research on the association between roles and mental health among women assumes roles are salubrious for all women. The intersectionality framework, however, suggests gendered and racialized systems of oppression impinge on minoritized women’s ability to occupy societally valued social roles (Cho et al., 2013; Crenshaw, 1991). Using data from the Collaborative Psychiatric Epidemiology Surveys (CPES), the following question motivates this study: Do the effects of women’s social role engagement (i.e., role volume, type, and configuration) on mental health vary by race and ethnicity?

Theoretical Framework

Role theory posits that fulfilling employment, parental, and marital roles require meeting formal, informal, and normative expectations such as adhering to contractual financial agreements or providing caregiving to children (K. Williams et al., 2009). Employee, spouse, and parent are the most widely studied roles and the ability to occupy primary social roles, daily and in the long term, matters for quality of life (Voydanoff & Donnelly, 1999). In addition to primary roles, this study investigates other family and friendship roles in women’s lives.

Women who are not married, parents, or employed, may benefit from less traditional roles (Thoits, 1983). In addition to nuclear familial ties (e.g., children, marital partner), other “family” roles are often critical sources of socioemotional and instrumental support for racially minoritized women (Sarkisian, Gerena, & Gerstel, 2006). In contrast to White Americans, Asian, Latinx, and Black households are more likely to be multigenerational (Marquez-Velarde, 2020) and include extended family members (Kulkarni, 2019). Moreover, friends are often considered family (i.e., fictive kin) in Black family life (Taylor, Chatters, Cross, & Mouzon, 2022), within immigrant communities (Ebaugh & Curry, 2000), and among White Americans (Taylor et al., 2022).

This study integrates role theory and the intersectionality framework, as the latter interrogates how systems of inequality influence women’s role engagement (Cho et al., 2013; Collins, 2019; Crenshaw, 1991). Cho, Crenshaw, and McCall (2013) emphasize, “…[W]hat makes an analysis intersectional…is its adoption of an intersectional way of thinking about the problem of sameness and difference and its relation to power…” (p. 795). If women’s quality of life is impacted by the ability to engage in social roles, then it is important to interrogate power differences. An intersectional analysis attends to which women have the power to marry, to be gainfully employed, or to become a parent, and the conditions under which they have agency to opt into or out of roles. For example, women who immigrate to the U.S. may gain citizenship through marriage, and with citizenship, access employment.

We adopt an inter-categorical intersectional approach by comparing how social roles influence the mental health of nine ethnic groups (McCall, 2005). Though impossible to capture the myriad sociohistorical processes undergirding the experiences of several ethnic groups in a single study, we attempt to “derive meaning from the observed data and interpret individual level data within a larger sociohistorical context of structural inequality that may not be explicit or directly observable in the data” (Few-Demo, 2014, p. 175). Importantly, our within-group analysis moves beyond the Black-White racial dichotomy by examining ethnic heterogeneity within racial group, thereby making a significant step in capturing the dynamic, multifaceted nature of “race” (Few-Demo, 2014).

Women’s Social Role Engagement And Mental Health

The psychological implications of social roles have been studied in the context of their volume (i.e., count), type (i.e., specific social roles), and repertoire (i.e., combination of roles). Here, we briefly review literature on primary roles and women’s psychological health.

Role Type

The psychological impact of primary roles is well documented (i.e., worker, spouse, and parent). Employment is beneficial to women’s mental health because it provides monetary gain, opportunities to experience success, and an expanded frame of reference (Barnett, 2004). Despite declines in marriage, matrimony remains psychologically beneficial for women, particularly when characterized by high satisfaction and financial stability (K. Williams et al., 2009). The psychological effect of motherhood is more equivocal (Evenson & Simon, 2005; Leupp, 2017; Simon, 1998). Though mothers often express high commitment to parenting (Simon, 1998), role strain limits the psychological benefits of motherhood (Evenson & Simon, 2005). In sum, marital and employee roles tend to be salubrious, yet motherhood appears to elicit distress. Whether family and friend roles are as crucial for women’s mental health remains unclear.

Role Volume

Considerable evidence supports the salubrious effects of holding multiple roles (Black et al., 2009; Davis, Sloan, & Tang, 2011; Sachs-Ericsson & Ciarlo, 2000). Consistent with a role enhancement perspective, multiple role engagement boosts well-being by enhancing resources, self-conception, and life satisfaction (Leupp, 2017). Alternatively, the role strain perspective argues that occupying multiple roles has negative consequences for well-being by increasing incompatible role demands (Marks, 1977). Accordingly, the relationship between role volume and mental health might be curvilinear: occupying too few (i.e., social isolation) and too many (i.e., role strain) social roles can be psychologically damaging (Voydanoff & Donnelly, 1999).

Role Repertoires

Role repertoires are another way to characterize role involvement. We use “repertoire”, “combination”, and “configuration” interchangeably to describe the amalgamation of occupying different sets of primary roles. Prior research documents the psychological consequences of social role combinations (Menaghan, 1989; Sachs-Ericsson & Ciarlo, 2000). Compared to married mothers, single parenthood is associated with higher rates of psychiatric disorder and distress among women (Menaghan, 1989; Sachs-Ericsson & Ciarlo, 2000). Occupying all three primary roles (married, working, mother) is associated with fewer depressive symptoms compared to other primary role repertoires (Baruch & Hyde, 2001).

Ethnic Variation in Women’s Social Roles and Mental Health

Prior research confirms that ethnicity (Jackson & Erving, 2020) and gendered dynamics (Leupp, 2017) influence role experience and mental health. Yet it remains unclear whether minoritized women’s racial and gender disadvantage influences their ability to acquire roles and the extent to which they psychologically benefit from role occupation. We use intersectionality to theorize how diverse women experience social roles and the concomitant psychological effects of role engagement. We provide an intersectional, sociohistorical overview of roles among ethnically diverse women as well as describe the psychological ramifications of role volume, role type, and role configuration by ethnicity.

Intersectional Sociohistorical Considerations: Ethnic Contrasts in Women’s Role Involvement

Anti-Blackness shapes Black women’s social roles. During the antebellum period, Black individuals were barred from biological, nuclear family configurations. Accordingly, social roles such as marriage and motherhood have historically found definition and legitimacy through exclusion of disadvantaged racial groups (D. T. Williams, 2021). The slave economy necessitated Black women’s physical, childbearing, and childrearing labor; moreover, single working motherhood has been a way of life for Black women for centuries (Hartman, 2016). The targeted criminalization and disenfranchisement of Black and Latinx men and women prevents minoritized women from establishing marital and parental roles (D. T. Williams, Simon, & Cardwell, 2019). African American women may experience the work role differently than their non-Black counterparts who have, on average, lower labor force participation rates; work may be more central to their self-concept, as combining parental and employment roles are expected realities for many Black women (Dow, 2016).

Black women from Caribbean countries may experience social roles differently from African Americans. For example, Afro-Caribbean women experience higher marriage rates and satisfaction than their U.S.-born counterparts (Bryant et al., 2008). Also, gendered expectations in parental and marital roles are less egalitarian within Caribbean immigrant families (Baptiste, Jr. et al., 1997). This study disrupts monolithic characterizations of Black women by ascertaining whether social roles have differential psychological effects for Afro-Caribbean and African American women who share a racial identity but have distinct ethnicities.

Each Asian and Latinx ethnic group has a unique history and orientation to the U.S. milieu. Experience of roles is influenced by motivations to emigrate, culture, as well as social and human capital. The selective U.S. migration of upper-class Cubans during the 1960s, in part, explains their high socioeconomic status (SES) among Latinxs (Bishin & Klofstad, 2012). Moreover, Latina mothers have different histories and contemporary experiences. For instance, single motherhood is lower among Cubans (27%) relative to Mexicans (41%) and Puerto Ricans (59%) (National Center for Health Statistics, 2003). Thus, Cuban American women have higher SES than Mexican or Puerto Rican women and greater access to resources.

As with other immigrant groups, recent Asian immigrants face challenges of reconciling their native, culturally embedded roles with the role configurations of their more assimilated counterparts (Walton & Takeuchi, 2010; M. Zhou & Bankston III, 2001). Chinese principles of family interdependency and harmony are distinct from U.S. ideologies of independence in family relations; hence, in Chinese American families, cross-cultural tensions may arise. Foreign-born Asian American women hold more traditional family expectations than U.S.-born Asian women (Walton & Takeuchi, 2010). Intermarriage (primarily to Whites) rates are higher among Filipina and Chinese women compared to Vietnamese women, presenting another variation in marital experience across Asian ethnic groups (Chen & Takeuchi, 2011). A consideration of the unique histories from which ethnically diverse women’s roles emerge reveals that roles are not equally accessible across ethnic groups; in addition, their psychological impact may differ by ethnicity.

Role Volume, Type, and Configuration Influences on Ethnically Diverse Women’s Mental Health

Empirical evaluations of the role enhancement hypothesis across diverse ethnic groups of women yield mixed findings. Though greater social role involvement is associated with lower depression rates for African American and Non-Latina White (hereafter, White) women (Cochran et al., 1999), role volume is unrelated to distress among Chinese women (Tang et al., 2002). Less considered is that the role volume-mental health relationship may be curvilinear rather than cumulative (i.e., linear). Role volume’s influence on Afro-Caribbean, Vietnamese, Filipina, Puerto Rican, Mexican, and Cuban women’s mental health is lacking in the literature.

The psychological effects of specific roles could vary for women with different ethnic identities. Past research reveals African American and White women psychologically benefit from employment (Brown & Cochran, 2003; Cochran et al., 1999; Coleman et al., 1987). Employed Chinese women report lower distress than unemployed peers (Tang et al., 2002). Whether women of other Asian ethnicities experience similar mental health benefits associated with employment remains unclear. In contrast, Mexican American women’s benefits from employment may depend on gender-role orientations, with traditional women experiencing heightened distress when occupying the employee role (Krause & Markides, 1985). There is a lack of research examining psychological benefits for other Latinx and Afro-Caribbean women.

The spousal role is psychologically beneficial for Black women (Brown & Cochran, 2003; Cochran et al., 1999; for an exception, see Coleman et al. 1987). Among African American women, the formerly married have greater risks for psychiatric disorder than married peers (D. R. Williams et al., 1992). Latinas exhibit similar patterns: the formerly married report higher distress than their married counterparts (Darghouth et al., 2015). Among Asian Americans, married women experience a mental health advantage over the never-married (Walton & Takeuchi, 2010). Disaggregating “Black”, “Latina”, and “Asian” groups by ethnicity may reveal greater variation in the marriage-mental health association.

The effect of motherhood on mental health is inconsistent (Nomaguchi & Milkie, 2020). Motherhood has either a null (Coleman et al., 1987; Marcussen & Piatt, 2005) or negative (Waldron & Jacobs, 1989) influence on Black women’s mental health. Equivocal findings could be attributable to the unique stress Black mothers face in an era of mass incarceration and state-sanctioned violence which necessitate intensive protective mothering strategies to ensure their children’s survival (Elliott & Reid, 2019). Though little research engages how motherhood influences Latina and Asian American women’s mental health, that they experience lower parental stain than Black mothers suggests potentially salubrious effects (Nomaguchi & House, 2013). Whether the parental role influences the psychological health of U.S. Latina and Asian women remains unclear.

Beyond primary roles, other familial and friendship roles are consequential for psychological health (Jackson, 1997; Jackson & Erving, 2020). Latinx (Viruell-Fuentes & Schulz, 2009), Asian (Walton & Takeuchi, 2010), and Black (Taylor et al., 2021) Americans have close relationships with family and friends; these social ties facilitate access to social resources and instrumental and socioemotional aid (Taylor et al., 2015). We anticipate family and friendship roles will be psychologically beneficial for non-White women.

With regards to role configuration, women of different ethnic backgrounds have historically held distinct primary role combinations. Relative to White peers, Black women historically worked and reared children simultaneously, often without the aid of a co-residential partner. Recent studies find that Black women psychologically benefit from employment in combination with spousal and parental roles (Christie-Mizell et al., 2019; Patterson et al., 2021). Furthermore, employed, unmarried, childless Black women reap similar psychological benefits as employed, married mothers (Christie-Mizell et al., 2019; Patterson et al., 2021).

Compared to African American women, Latina and Asian American women practice more traditional gendered cultures that heighten spousal and parental expectations (Jones, Buque, & Miville, 2018) which, in turn, induce distress when familial and employee roles are integrated (Darghouth et al., 2015). Early work showed employment was associated with greater distress among Mexican American mothers with young children (Krause & Markides, 1985). As such, flexible gender role expectations may buffer the strain of combining work and family roles for African American women, whereas traditional familial roles in tandem with employment may have negative psychological consequences for Latina and Asian American women. Ethnic heterogeneity among Latina and Asian American women may elicit variegated findings.

Data and Methods

Data are from the Collaborative Psychiatric Epidemiology Surveys (CPES) which consists of three surveys: National Comorbidity Survey Replication (NCS-R), National Survey of American Life (NSAL), and National Latino and Asian American Study (NLAAS) (Pennell et al., 2004). Collected between 2001 and 2003, these data collectively provide national probability samples of understudied U.S. ethnic groups. NCS-R includes English-speaking adults, age 18 and over, residing in households in the coterminous U.S. A total of 9,282 respondents were interviewed. The NCS-R included Non-Latino White, African American, Caribbean Black, Mexican, Other Latino, and Other Asian respondents. NSAL is the first national probability study of Blacks of immediate Caribbean descent and includes a nationally representative sample of African American-identified Black respondents (Jackson et al., 2004). A total of 6,082 interviews were completed among English-speaking Caribbean Blacks, African Americans, and Non-Latino Whites. NLAAS includes adults 18 and over residing in households in the coterminous United States, Alaska, and Hawaii. A total of 4,649 interviews were conducted with Mexican, Puerto Rican, Cuban, Chinese, Filipino, Vietnamese, Other Latino, and Other Asian respondents. Interviews were conducted in English, Spanish, Chinese, Tagalog, and Vietnamese. When combined, CPES includes a sample of 9,111 female-identifying respondents. We include Non-Latina White, Mexican, Puerto Rican, Cuban, African American, Afro-Caribbeans, Chinese, Filipina, and Vietnamese Americans in the analysis. There were 976 respondents for whom a specific ethnic group could not be identified and after dropping cases with missing values, we retain 91% of the sample (N=7,370).

Measures

We assessed psychological distress using the Kessler 6-item scale (K6) (Kessler et al., 2010) which asked respondents how often over the past month they felt nervous, hopeless, restless or fidgety, so depressed nothing could cheer them up, everything was an effort, and worthless. Response options were none of the time (0), a little of the time (1), some of the time (2), most of the time (3), and all of the time (4). K6 is an unweighted summary scale (range: 0–24), with higher scores reflecting higher distress.

Role Occupancy.

We develop three measures of social role occupancy: role volume, role type, and role configuration. For role volume, we include a count of five roles: worker, spouse, parent, family member, and friend. Respondents are considered workers if they reported being employed. Spouses refer to currently married individuals and those who are cohabiting with a romantic partner. We distinguish between formerly married (i.e., divorced/separated/widowed) and never married respondents in the role type analysis (see Table 3). Parents refer to respondents with coresident children under the age of 18. Those occupying the family member role reported having contact (i.e., see, write, or talk on the phone) with non-residential family members at least once a month. Similarly, respondents who occupy the friend role reported having contact with friends at least once a month. Dummy variables were created with 0 reflecting the absence and 1 indicating the possession of the role. The number of roles increased by 1 for each possessed social role. Because prior research indicates a potential curvilinear association between social role occupation and mental health, we also analyzed a squared version of role number (number of roles²) (Jackson, 1997). To measure role type, binary variables are used for the five social roles with a score of 1 assigned to those who hold the role and a score of 0 to those who did not. Two dummy variables were created to distinguish between never married and formerly married (i.e., divorced, separated, or widowed) respondents. Married/cohabiting respondents serve as the comparison category. The role configuration analysis focuses on three primary social roles: worker, spouse, and parent (Jackson, 1997; Menaghan, 1989). Those occupying all three roles (employed, married parent) are the referent. In the role configuration analysis, we also control for family member and friend roles.

Table 3.

Weighted Coefficients for Effects of Specific Roles on Psychological Distress for Women, by Ethnicity

	Non-Latina White	Mexican	Cuban	Puerto Rican	African American	Afro-Caribbean	Chinese	Filipina	Vietnamese
Role
Worker	−.40	−.40+	−2.04**	−1.03*	−.88***	.58	−.64	−.03	.11
Spouse (reference)
Formerly Married	1.47***	1.35**	.18	.98	−.18	−.57	.76	−.14	1.33+
Never Married	.44	.85+	−1.07+	.79	.34	.99+	.75	.62	1.61+
Parent	.23	−.32	−.25	.34	−.32	.35	−.68	−.38	.03
Family Member	−.35	−1.41**	−2.24+	−.96	−1.14*	−1.83+	−1.53**	−.15	−.12
Friend	−.65	−.11	−1.80*	−2.23**	−.40	−2.53*	−.89	.15	−.23
R2	.13	.11	.30	.20	.17	.17	.15	.13	.19

Source: Collaborative Psychiatric Epidemiology Surveys, 2001–2003

Note: Each regression model controls for age, education, household income, nativity, region, and health problems that limit mobility in past month.

⁺p<.10

^*p<.05

^**p<.01

^***p<.001

Controls.

All models adjust for factors associated with mental health in the broader literature. Age is a four-category measure: 18–34 (reference), 35–49, 50–64, and 65 years and older. Education includes less than high school (reference), high school diploma/some college, and college degree. Annual household income ranges from $0 to $200,000 (top-coded) and is measured in $10,000 increments. We also control for nativity (U.S.-born = 1). Region of residence included South (reference), Northeast, Midwest, and West. To account for selection into roles, we control for difficulty with mobility due to health problems (yes = 1).

Analytic Strategy

We first report means and standard deviations of psychological distress as well as the distribution of role volume, role type, and role configuration by ethnicity (Table 1). Descriptive statistics for control measures are available in a supplemental file. Given our interest in ethnicity-specific nuances in how social role engagement influences mental health, we present ethnicity-stratified models. This methodological decision is consistent with an intersectional approach that provokes exploration of intra-categorical diversity and complexity (Few-Demo, 2014; McCall, 2005). To assess whether the role occupancy-mental health association differed across ethnic groups, we estimated separate OLS regression equations for each ethnic group. Table 2 includes the role volume analysis. We used the suest post-estimation STATA command to conduct Wald tests to assess whether the effect of role volume significantly differed across groups. Table 3 includes the role type analysis. In Table 4, we include an analysis for the effect of role configurations on mental health by ethnicity. Significance levels were set to p<.10 to accommodate small samples of some groups (e.g., Vietnamese women). Analyses were conducted with STATA 14.2. All analyses were weighted to adjust for sample design effects.

Table 1.

Descriptive Statistics on Role Occupancy by Ethnicity (N=7370)

	Non-Latina White	Mexican	Cuban	Puerto Rican	African American	Afro-Caribbean	Chinese	Filipina	Vietnamese
N	2193	655	298	278	2337	759	312	268	270
K6 Scale (0–24)	4.87(4.93)	3.32(4.37)*	3.74(5.16)*	3.84(4.55)*	4.36(4.62)*	3.42(3.80)*	3.13(3.38)*	2.05(2.45)*	1.96(3.24)*
A. Number of Roles Enacted
0	.004	.005	.01	.00	.006	.00	.003	.01	.02
1	.04	.05	.06	.03	.05	.04	.04	.02	.09
2	.20	.20	.19	.29	.22	.18	.25	.15	.25
3	.40	.35	.36	.34	.36	.39	.29	.35	.31
4	.29	.32	.28	.26	.28	.28	.33	.31	.26
5	.07	.08	.10	.09	.09	.11	.09	.15	.08
Mean Number of Roles Enacted	3.16	3.18	3.13	3.08	3.12	3.23	3.17	3.36*	2.94*
B. Type of Role
Worker	.60	.49*	.52*	.53	.62	.72*	.57	.63	.52
Spouse	.58	.63	.62	.52	.35*	.40*	.67*	.65*	.77*
Formerly Married	.25	.17*	.28	.24	.32*	.27	.16*	.14*	.09*
Never Married	.17	.20	.10*	.23*	.33*	.33*	.17	.21	.14
Parent	.18	.44*	.24*	.31*	.36*	.29*	.25*	.31*	.33*
Family Member	.90	.87	.91	.91	.94*	.92	.82*	.90	.72*
Friend	.91	.74*	.85*	.81*	.86*	.91	.85	.86*	.60*
C. Combinations of Primary Roles
None	.16	.11	.20	.21	.20*	.16	.17	.13	.09*
Employed Only	.22	.13*	.11*	.13*	.24	.32*	.15*	.16	.11*
Married Only	.18	.16	.19	.14	.07*	.06*	.16	.15	.23
Parent Only	.01	.05*	.05*	.05*	.08*	.03	.004*	.01	.02
Employed and Married	.26	.16*	.26	.21	.13*	.19*	.27	.25	.24
Employed Parent	.03	.07*	.03	.08*	.14*	.10*	.01*	.05	.01*
Married Parent	.05	.18*	.05	.07	.03	.04	.10*	.08*	.13*
Employed, Married, Parent	.09	.13	.12*	.11	.11	.13	.14*	.18*	.17*

Source: Collaborative Psychiatric Epidemiology Surveys, 2001–2003

Weighted means and proportions are shown in the table.

^*Indicates a significant difference compared to Non-Latina White women.

Table 2.

Weighted Coefficients for the Effect of Number of Roles on Psychological Distress among Women, by Ethnicity

	b	SE	R2
Non-Latina White
Model 1: Number of Roles	−.44*	.16	.12
Model 2
Number of Roles	−.39	.77	.12
Number of Roles squared	−.01	.11
Mexican
Model 1: Number of Roles	−.63***	.11	.10
Model 2
Number of Roles	−2.35**	.78	.11
Number of Roles squared	.28*	.12
Cuban
Model 1: Number of Roles	−1.03+	.51	.27
Model 2
Number of Roles	−3.39*	1.48	.28
Number of Roles squared	.40+	.20
Puerto Rican
Model 1: Number of Roles	−.86**	.28	.17
Model 2
Number of Roles	−2.84+	1.42	.18
Number of Roles squared	.31	.22
African American
Model 1: Number of Roles	−.50***	.10	.16
Model 2
Number of Roles	−.44	.53	.16
Number of Roles squared	−.01	.08
Afro-Caribbean
Model 1: Number of Roles	−.40	.35	.11
Model 2
Number of Roles	−.03	.99	.11
Number of Roles squared	−.06	.14
Chinese
Model 1: Number of Roles	−.87**	.23	.14
Model 2
Number of Roles	−2.02+	1.04	.15
Number of Roles squared	.19	.15
Filipina
Model 1: Number of Roles	−.22	.13	.12
Model 2
Number of Roles	−1.47*	.66	.13
Number of Roles squared	.20+	.11
Vietnamese
Model 1: Number of Roles	−.31	.19	.17
Model 2
Number of Roles	.30	.81	.17
Number of Roles squared	−.11	.14

Source: Collaborative Psychiatric Epidemiology Surveys, 2001–2003

Note: Each regression model controls for age, education, household income, nativity, region, and health problems that limit mobility in past month.

⁺p<.10

^*p<.05

^**p<.01

^***p<.001

Table 4.

Weighted Coefficients for Effects of Social Role Combinations on Psychological Distress for Women, by Ethnicity

	Non-Latina White	Mexican	Cuban	Puerto Rican	African American	Afro-Caribbean	Chinese	Filipina	Vietnamese
No roles	.97+	2.07*	1.88	1.30	1.52**	−.78	2.28**	.32	1.09
Employed only	.88+	.75	−.29	−.09	.78**	.98	.80	.29	1.10
Married only	.21	.21	1.75	−1.17+	1.54*	.35	.91	.62	−.14
Parent only	2.15*	.78	2.84+	2.02	1.83***	2.53**	−.36	−1.51**	.88
Employed and married	−.35	−.03	.21	−.68	1.12*	.66	.56	−.31	−.52
Employed and parent	1.45+	.72	−2.09*	−1.31	.53	.83	.55	−.08	2.90*
Married and parent	.42	.004	1.63	1.06	1.39*	1.01	.20	−1.24*	−.83
R2	.13	.11	.31	.23	.17	.18	.16	.15	.19

Source: Collaborative Psychiatric Epidemiology Surveys, 2001–2003

Note: Each regression model controls for age, education, household income, nativity, region, health problems that limit mobility in past month, and the family member and friend roles.

The reference category of social role combinations is employed, married, parent.

⁺p<.10

^*p<.05

^**p<.01

^***p<.001

Results

Descriptive Statistics

Table 1 includes mean psychological distress and proportions for role occupancy by ethnicity. Relative to ethnic minority women, distress levels were significantly higher among White women (M = 4.87; SD = 4.93). Filipina (M = 2.05; SD = 2.45) and Vietnamese (M = 1.96; SD = 3.24) women had the lowest psychological distress. The modal number of roles was three; the one exception was Chinese women who had four modal number of roles. Filipina women had a significantly higher mean number of roles (3.36) compared to White women (3.16), though Vietnamese women reported a significantly lower mean number (2.94).

We found variegated patterns by ethnicity when examining specific role types. First, though 60% of Whites were employed, significantly fewer Mexicans (49%) and Cubans (52%) occupied the work role. Afro-Caribbeans (72%) had the highest employment. Compared to Whites (58%), Asian women were more likely to occupy the spousal role: 77% of Vietnamese, 67% of Chinese, and 65% of Filipina women were married or cohabiting. African Americans (35%) and Afro-Caribbeans (40%) were the least likely to occupy the spousal role. Thirty-two percent of African Americans were formerly married, a proportion significantly higher than White women (25%). Asian and Mexican women were less likely to be formerly married. Compared to White women, African American, Afro-Caribbean, and Puerto Rican women were significantly more likely and Cuban women were less likely to be never married. In terms of the parental role, all ethnic minority groups were significantly more likely to have coresident children relative to White women. Most ethnic groups occupied the family member role; however, this role was especially high among African Americans (94%), yet lower among Chinese (82%) and Vietnamese (72%) women. Other groups fell in the middle, ranging from 87% to 92%. Last, 91% of White and Afro-Caribbean women occupied the friendship role, while Vietnamese women had the lowest proportion (60%).

Primary role combinations revealed greater complexity across groups. African American women (20%) were more, yet Vietnamese women (9%) were less, likely to occupy no primary roles (worker, spouse, parent) compared to White women (16%). Afro-Caribbean women were most likely to be “employed only” (32%). Nearly one-quarter of Vietnamese women were “married only”, followed by Cuban (19%) and White (18%) women. Compared to White women (1%), Mexican (5%), Puerto Rican (5%), Cuban (5%), and African American (8%) women were more likely to occupy the “parent only” role repertoire. “Employed and married” was the most common role configuration for six groups: 27% of Chinese, 26% of White and Cuban, 25% of Filipina, 24% of Vietnamese, and 21% of Puerto Rican women occupied this role combination. Within the “employed parent” category, African Americans (14%) and Afro-Caribbeans (10%) had the highest proportion represented compared to only 3% of White women. For “married parents,” the highest proportion was Mexican women (18%), followed by Vietnamese (13%), Chinese (10%), and Filipina (8%) women. Occupying all three primary roles was highest for Filipina (18%), Vietnamese (17%), Chinese (14%), and Cuban (12%) women who had significantly higher proportional representation than White women (9%).

Regression Analysis

Role Volume

Coefficients from regression analyses assessing the effect of role volume (and social roles²) on psychological distress are reported in Table 2. Results from Model 1 indicate occupying a greater number of social roles was associated with lower distress for most groups. Three exceptions were Afro-Caribbean, Filipina, and Vietnamese women; however, coefficients were in the expected direction. In Model 2, we included social roles² to ascertain whether there was a curvilinear association between role number and distress. We found evidence of a curvilinear association for Mexican, Cuban, and Filipina women, suggesting diminishing returns to accumulating social roles. Of note, the group with the highest R² was Cuban women: 28% of their variation in distress was explained by social role number (and the controls).

To provide a visual depiction of the analysis, Figure 1 includes predicted values of significant linear effects for White, African American, Afro-Caribbean, and Chinese women. Figure 1 shows a clear inverse association between role number and distress for these groups. The steepest declines in psychological distress with each added social role were observed for Chinese and Puerto Rican women. Figure 2 includes predicted values of significant curvilinear effects for Mexican, Cuban, and Filipina women. Occupying additional roles beyond three did not offer added psychological benefit for Mexican, Cuban, and Filipina women.

FIGURE 1.

PREDICTED VALUES OF PSYCHOLOGICAL DISTRESS BY NUMBER OF SOCIAL ROLES AND ETHNICITY

FIGURE 2:

PREDICTED VALUES OF PSYCHOLOGICAL DISTRESS BY NUMBER OF SOCIAL ROLES² AND ETHNICITY

Role Type

Table 3 presents the effects of specific roles on psychological distress by ethnicity. No particular social role was universally psychologically beneficial. In fact, occupying the parental role was not significantly associated with mental health for any ethnic group. Given variation in the findings, we review effects of specific social roles by ethnic group. Among White women, the formerly married experienced higher distress relative to those who occupied the spouse role (b = 1.47, p <.001). No other social roles were associated with distress. For Mexican women, the employee, spousal, and family member roles were psychologically beneficial.

Being employed, a family member, and friend were associated with lower distress for Cuban women. Moreover, the never married reported lower distress compared to their married counterparts, suggesting the spousal role was not psychologically beneficial for Cuban women. For Puerto Rican women, employee and friend roles were associated with lower distress. Among African American women, being employed and a family member were mental health protective. For Afro-Caribbean women, those who were never married experienced relatively higher distress compared to the married; family member and friendship roles were associated with better mental health. For Chinese women, being a family member was associated with lower distress. No specific roles were associated with Filipina women’s mental health. Among Vietnamese women, only occupying the spousal role was psychologically beneficial: the formerly married and never married reported significantly higher distress.

Role Configuration

We assessed the psychological effects of primary role repertoires in Table 4. The reference category was employed married parents. For White women, parent only and employed parent were the most psychologically compromising role combinations. No roles and employed only were associated with higher distress than occupying all three primary roles. Any role combination involving marriage was more psychologically beneficial than not for White women.

Findings for African American women provided strongest evidence in support of the hypothesis that occupying all three social roles is the most psychologically beneficial. However, distress levels of employed parents and employed married parents did not significantly differ among African Americans. Among Afro-Caribbean women, parent only was associated with higher distress than being an employed married parent.

For Mexican and Chinese women, occupying no primary social roles was most distressing. In fact, compared to occupying all three roles, having no primary social roles was associated with an average increase of 2.07 and 2.28 in distress for Mexican and Chinese women, respectively. For Vietnamese women, employed parents had more distress than employed married parents. Unexpectedly, for three ethnic groups, occupying one or two primary social roles appeared more psychologically beneficial than being a married employed parent. First, among Filipinas, parents only or married parents were less distressed than married employed mothers. Second, for Puerto Ricans, the married only were less distressed than married employed mothers. Third, although parent only was associated with higher distress among Cuban women, employed mothers had lower distress relative to employed married mothers.

Discussion

This study used role theory, an intersectional framework, and nationally representative data to examine the social role-mental health association among nine ethnic groups. Study results revealed heterogenous psychological effects of role engagement. Consistent with the role accumulation hypothesis, occupying a greater number of social roles elicited less distress for White, Puerto Rican, African American, and Chinese women. Yet, for Mexican, Cuban, and Filipina women, occupying more than three social roles yielded diminishing psychological returns. Perhaps for Mexican, Cuban, and Filipina women, expectations associated with occupying all five roles induce role strain or conflict. On the other hand, occupying no roles is associated with social isolation and poor mental health for all groups.

Examining the influence of specific role types on mental health revealed complex, unique ethnic patterns. Despite the literature pointing to the psychological benefits of marriage for women (K. Williams et al., 2009), the spousal role was not psychologically consequential for Puerto Rican, African American, Chinese, nor Filipina women. Never married Cubans were at a psychological advantage relative to the married. In contrast, White and Vietnamese women benefitted only from the spousal role. Perhaps marital quality and satisfaction could explain why marriage is psychologically beneficial for White and Vietnamese women, yet neutral to detrimental for the well-being of women from other ethnic backgrounds. We note the marriage role, as measured in our study, excludes unmarried, non-cohabiting women in long-term romantic relationships, a dynamic more common among African Americans (Lincoln, Taylor, & Jackson, 2008). Therefore, African American women in committed, intimate relationships may benefit similarly to those who are married or living with a partner.

Employment was associated with lower distress among African American women who have historically had high workforce participation (Toossi & Morisi, 2017). The work role was also psychologically beneficial for Latina women (regardless of ethnicity). Despite having relatively lower workforce participation relative to Black women, recent trends suggest an upward trajectory among Latina women (Toossi & Morisi, 2017); thus, it is insightful that employment was psychologically beneficial. Cuban women especially benefitted from being employed. Perhaps employment offers financial independence and self-direction not available to those who do not participate in the workforce.

The parental role was unrelated to mental health which may reflect the equivocal role of parenthood for women’s psychological health (Nomaguchi & Milkie, 2020). Though parenthood can be personally rewarding, it is often accompanied by stressors, especially early in the child’s life (Leupp, 2017). Perhaps rewards and strains of motherhood level out, producing a null effect on mental health. Some parenting strains are also racialized: Black mothers must consider potential dangers their Black children may encounter outside the confines of their home (Dow, 2016; Elliott & Reid, 2019). Among Vietnamese mothers, tensions between personal cultural traditions and their children’s desire for independence could make the parental role less salubrious (M. Zhou & Bankston III, 2001).

Healthy extended family and friendship dynamics contribute to psychological health in ways underemphasized in the roles literature. Study findings challenge dominant theoretical perspectives that dismiss extended family and friends as primary roles. Despite a null association for White women, familial and friendship roles were psychologically beneficial for most racial and ethnic minority women. For Cuban and Afro-Caribbean women, family and friendship roles were associated with lower distress. Friendship was associated with lower distress for Puerto Rican women, and family roles were protective for African American and Mexican women. For Chinese women, family member was the only beneficial role. Findings underscore the importance of social roles beyond “primary” roles. Filipina and Vietnamese women, however, were the exceptions. In sum, we uncover the intersectional complexity of minority women’s social lives by de-centering the White Eurocentric standard of the “nuclear family” as the lynchpin of psychological health. For some women, extended familial and friendship ties are even more consequential for mental health than primary roles (e.g., Afro-Caribbean women).

Literature on role configurations suggest occupying all three primary social roles is psychologically ideal. Nevertheless, the disproportionate burden employed, married, mothers experience because of gendered expectations calls into question this assumption (Leupp, 2017; Sachs-Ericsson & Ciarlo, 2000). We found strongest evidence in support of the hypothesis that employed married mothers enjoy optimal mental health among African Americans (Christie-Mizell et al., 2019). African American and Cuban women also psychologically thrive as unmarried employed parents. This novel finding reveals marriage is not a prerequisite for their mental health. In essence, occupying all three primary roles is not the sole pathway to positive mental health.

For other groups, marriage only or in combination with other social roles appeared ideal. For White women, any social role combination involving marriage was relatively more psychologically beneficial than not occupying the spousal role. Among Puerto Ricans, the married only were less distressed than their married employed parent counterparts. This is somewhat consistent with a perspective that combining marital roles with employment and parenthood may induce distress, particularly for women who hold more traditional gender role orientations (Krause & Markides, 1985). For Afro-Caribbean women, unemployed unmarried mothers experienced the most psychological distress. This status is rare among Afro-Caribbean women (3%), which may make it especially stigmatizing.

For Filipinas, unmarried unemployed mothers and married mothers were less distressed than married employed mothers. This suggests motherhood is rewarding for Filipina women, but could be stressful when time-intensive expectations associated with employment and marriage are involved (Leupp, 2017). Given research on social role occupation and mental health among Filipina women is rare, these findings merit additional empirical attention, as they insinuate ethnic heterogeneity in Asian American women’s role experiences.

Though we utilize individual-level data in our analysis, study findings implicate the structural nature of gendered-racialized disadvantage that marginalizes non-White women. Structural intersectional interpretations of our analysis necessitate changes in cultural norms and social policies (Crenshaw, 1991). Similarly, as noted by Risman (2004), gendered theorizing must be accompanied by praxis, offering interventions at the interactional, cultural, and structural levels. For Cuban, Puerto Rican, and Filipina women, occupying all three primary roles is psychologically precarious, as other role repertoires are more psychologically beneficial. Moreover, for Cuban, Mexican, and Filipina women, occupying more than three roles offers little to no psychological benefits. These findings suggest unique intersectional vulnerabilities for Latina and Filipina women, and addressing systemic issues is necessary to alleviate the strains associated with multiple role occupancy. One such change pertains to the challenges women encounter while working and parenting young children. U.S. parental leave policies are infamously less generous than other industrialized nations (Livingston & Thomas, 2019), presenting barriers to integrating work and parental roles. Parental leave policies are limited or non-existent in industries where Filipina, Mexican, and Puerto Rican women are concentrated like the domestic and service sectors (Bucknor 2016; Filipino Migrant Center 2021). Generous parental leave policies may allow women in these industries to integrate parental and work roles more effectively.

With regards to ethnic groups with a significant immigration population (e.g., Vietnamese and Chinese women), a structural-level change would entail a re-imagining of immigration policy. Family reunification in the context of immigration policy is limited to marital partners and immediate family members (Gubernskaya & Dreby, 2017). Accordingly, policy-level decisions have implications for immigrant women’s abilities to acquire certain roles and shapes conditions under which they adopt particular roles. Our results suggest the nuclear family structure advantages White women but may be less psychologically advantageous for ethnic minority women who may rely on other family and friends for emotional and instrumental support. Accordingly, changes in policies that expand the definition of “family” could allow for U.S. entry of extended family members to aid in providing tangible household support.

Limitations

First, given how varied the effects of specific roles on mental health were across ethnic groups, role quality and strain may explain patterns discovered here (Davis et al., 2011). For example, work-family conflict could be useful to assess ethnic differences in role strain among working married mothers. Second, gender ideology affects how women experience social roles. Future work should assess gender attitudes among ethnically diverse women and how ideologies influence role experiences and psychological health. Third, U.S.-born and foreign-born women may navigate social roles in distinct ways, which warrant further attention. Relatedly, challenges with the operationalization of “ethnicity” could have influenced study findings. Last, operationalizations of social roles are open to interpretation, varying across studies. In our analysis, parental status was limited to those with young coresident children. Nevertheless, adult children with stressful life conditions may cause distress for their parents. Another role measurement issue involves family and friend role measures, which are qualitatively and quantitatively different from the three primary role measures. We used past research as a guide (e.g., see Jackson, 1997) to develop a measure capturing sustained contact with relatives and friends, which attempts to more closely approximate commitment to primary roles.

Conclusions

This study demonstrates the powerful influence of race, ethnicity, and gender – as social structures – on how social roles affect psychological health. Though we unveil nuanced patterns that complicate the extant literature, explanations for these findings deserve further investigation. For example, why is marriage psychologically problematic for Cuban women? Why is employment not psychologically beneficial for Asian American women? These questions and others grant opportunity for future research to interrogate how unique experiences of role engagement among ethnically diverse women translate into distinct psychological patterns.

Study results suggest roles can have some psychological benefits, yet for some women, role accumulation has limits. As opposed to simply focusing on women obtaining the “ideal” role configuration, investment in structural (e.g., expanding the definition of “family reunification” immigration policy) and commitment to cultural (e.g., achieving gender equity in household division of labor) transformation are steps toward achieving social justice for ethnic minority women. Such efforts will also reduce strains often accompanying multiple role integration and, hence, improve the mental health of women embedded within interlocking systems of gender, racial, and economic oppression.