Intersectionality and labor market outcomes

Elevator pitch

The privilege or disadvantage of individuals is not determined by a single social identity. The sexual division of labor affects women’s and men’s labor supply, the industries and occupations they enter, their earnings and progress. However, being a racial/ethnic or sexual minority (along with class and age) also influences job opportunities. Evidence shows that minority women’s experiences are unique and jointly determined by their gender and minority status, although the gender effect appears to precede that of minority.

Key findings

Author's main message

In many countries, women’s average wages are lower than men’s before and after adjusting for characteristics associated with earnings (such as education and experience). However, women and men are not homogenous groups. The experiences of racial/ethnic or sexual minority women are not the same as those of majority women or minority men, either because they face different levels of discrimination/privilege or because their labor-related characteristics are different. Age and social class also influence their outcomes. An effective policy response requires accounting for diversity within groups.

Motivation

Chances of hiring and promotion, occupations, wages, training opportunities, and working conditions in general are not the same for women and men or for Black and White people. Ethnicity, nativity, age, sexual orientation, class, religious beliefs, and physical ability are sociodemographic categories that also shape individuals’ outcomes in the labor market. However, each of these social identities alone is not enough to explain the opportunities that individuals face. Racial, ethnic, and nativity disparities are not necessarily of the same magnitude for women and men. Likewise, gender disparities within a racial/ethnic minority or among immigrants may differ from those within the majority. The intersection of gender with other sources of domination gives rise to new categories with their own identities and positions in society. It is at the intersection of these various identities that labor inequalities are generated because some identities cause privileges and other hindrances.

Discussion of pros and cons

Intersectionality and cross-country comparability

The term intersectionality, coined in 1989 in a legal context, was originally introduced to highlight the multiple forms of discrimination against Black women in the US, although the experiences of oppression of Black women had previously entered the feminist debate, both in the US and Latin America. This view was later extended to deal with other axes of exclusion and domination, such as ethnicity, nativity, sexual orientation, age, and class. Intersectionality means putting the lens on the interaction of the various identities that individuals possess because the intersection creates unique experiences of advantage and disadvantage.

Empirical literature on labor market outcomes and intersectionality is quite recent and has been centered on the intersection of gender and race/ethnicity in the US context, although newer research also includes other identities (nativity, age, sexual orientation, and social class) and geographies (Australia, Brazil, Canada, South Africa, and Europe), especially in studies of wage inequality.

The categories to which race and ethnicity refer vary considerably across geographies, making cross-national comparisons difficult. In the US, for example, ethnicity refers to Hispanic versus non-Hispanic status, and most studies distinguish among non-Hispanic individuals of the major racial groups (White, Black, and Asian people, and less frequently American Indian/Alaska Native people) and Hispanic individuals of any race [1], [2]. In the European context, however, there is some reluctance to use the term race, and most labor market data sets do not provide this information. Despite this, some studies explore Black and Asian populations who were born (or whose parents were born) outside of Europe, identifying them by their country/region of origin [3]. In general, European-based studies tend to identify ethnicity with immigrant status, although Roma are an ethnic group also studied [4], [5]. In Latin America, information availability on race/ethnicity is also limited in data sets. For some countries there is no information, while for others the definition of minority varies, ranging from being part of an indigenous group to skin color and mother tongue [6]. In Australia, studies often distinguish between Indigenous and non-Indigenous people [7].

Gender, race/ethnicity, and labor outcomes

Intersectional studies add new insights to the literature that explores race/ethnicity and gender separately (but it offers limited evidence on labor outcomes other than wages). First, the gender wage gap is not limited to a specific group, although it is unclear whether the gap is higher for the majority group [1], [2], [3], [4], [5], [8]. In the US, the gender wage gap is higher for White people than it is for Black people—not because Black women have a racial advantage but because of the racial disadvantage of Black men—and other racial minorities [1], [2]. In Australia, there is evidence that the gender wage gap is larger for the non-Indigenous population than it is for the Indigenous population [7]. In the UK, Spain, and Germany, the gender wage gap is larger for natives than it is for immigrants [3], [5], [8]. However, in France the ranking is the opposite [3]. In Canada, immigrants landing as adults also experience a larger gender wage gap than the native born. In some Central and South-Eastern European countries, the gender wage gap is larger for the Roma population than it is for the non-Roma [4].

Second, intersectionality shows that the racial/ethnic gap differs across gender groups. In the US, Australia, Germany, and Spain, the majority-minority wage gap is higher for men than it is for women [1], [2], [3], [5], [7], [8], perhaps because the latter occupy lower positions in the economic ladder, i.e., the gender effect seems to come before that of minority. In terms of other labor outcomes, this pattern differs [3].

Third, gender and race/ethnicity interact in complex ways. The effect of holding two disadvantageous identities simultaneously is not necessarily equal to the sum of the penalties associated with each of them, as there may be interaction between identities. In the US, several studies show that the wage gap between Black women and White men is not the sum of the gender and racial gaps [1], [2]. Non-additivity has also been documented in Europe when looking at nativity and gender [5]. This implies that to determine the wage gap between minority women and majority men, one should compare these two groups, rather than looking at gender and minority-majority gaps of the whole population and adding them up.

Fourth, using an intersectionality approach allows to explore whether the reasons behind the disadvantages of minorities are the same for women and men. Although the literature shows that part of the disadvantage of many racial/ethnic minorities in Europe and the US arises from human capital-related factors, minority women in these regions tend to have higher educational attainments than same-minority men. In fact, research in the US context shows that education plays a larger role to explain the low wages of Black and Hispanic men than those of their female peers [2].

Fifth, intersectionality brings new light to the disadvantages faced by women because it opens the door to comparisons of men and women who do not necessarily belong to the same group. Many studies document a larger gap between majority men and minority women (except for some Asian groups in some Anglo-Saxon countries) than they do between majority men and women, even after adjusting for characteristics associated with wages [2], [5], [6]. The US literature also suggests that White, Black, Hispanic, and Native American women have a wage penalty not only when compared to similar men (i.e., those with the same basic characteristics) of the same group, but also to similar men of other groups, which shows the extent of the gender effect [2]. 

Occupational segregation and wages

Occupation (and sector) helps explain an important part of the wage gaps of women and racial/ethnic minorities. Part of the concentration of some minority women and men, but not majority women, in low-paid occupations seems to arise from human capital-related factors, but the segregation appears to go beyond that.

Figure 1 displays the hourly wage of various gender-race groups in the US, before and after controlling for characteristics, decomposed in two terms. The between component, if negative, reflects the wage loss the group has as a result of working mainly in low-paid occupations. If positive, the group works mainly in highly paid occupations. The within term embodies the wage advantage (if positive) or disadvantage (if negative) that the group has within occupations compared to other groups. The sum of the two terms, which are expressed as proportions of the national average wage, equals the wages shown in the graphical abstract. 

White women derive most of their unadjusted wage disadvantage from getting wages below average within occupations. (The light blue bar is below zero while the dark blue bar is almost zero). However, if all groups had the same attributes as White men, White women would get a wage disadvantage also from their occupational sorting (the dark blue bar would be below zero), although to a lower extent than Black, Hispanic, and Native American women would. In other words, these female groups are concentrated in low-paid occupations beyond what one would expect based on their characteristics. Unlike them, Asian women are concentrated in highly paid occupations when looking at raw numbers, but their occupational sorting is no longer beneficial after controlling for characteristics. On the other extreme there are White and Asian men, whose clustering in highly paid occupations seems to exceed by far their composition. Black men stand out among male groups because they tend to be concentrated in low-paid occupations even after adjusting for characteristics (although less so than Black, Hispanic, and Native American women are), a pattern that does not occur in the case of Hispanic or Native American men.

The literature suggests that occupational sorting helps explain a large part of the wage disadvantage of Black women and/or Black men not only in the US but also in Brazil and South Africa [1], [6], [9].

In Europe, immigrants tend to be concentrated in low-paid occupations, but the wage disadvantage that immigrant women get from their occupational sorting is larger than immigrant men’s is (Figure 2). The composition effect (i.e., differences in characteristics) seems to explain most of the cross-country disparities for male immigrants, but not for female immigrants, who tend to work in low-paid occupations regardless of their education levels and additional characteristics.

Other axes of domination: age, sexual orientation, and social class

Age

There is evidence for Europe and the US of race and/or gender interplaying with age during the hiring process [10], [11]. Based on experiments, this literature suggests that discrimination by race and gender shapes applicants’ access to vacancies and wages in a differentiated way along the life cycle. However, the relationship between racial discrimination and age does not increase or decrease constantly along the life cycle. There is some evidence for the US of racial discrimination against Black applicants to diminish when moving from prime age into middle age, although it raises again for older applicants. There is also evidence for the UK of discrimination against older White and Black applicants, both females and males, compared to younger White men, with Black women experiencing the highest adverse employment outcomes. Some studies also suggest that age discrimination in hiring is stronger for women than it is for men.

Sexual orientation

Most studies document a wage penalty for gay men (or men living in same-sex couples), compared to heterosexual men (or men living in different-sex couples) with similar education, experience, and geographic location. For lesbians (or women living in same-sex couples), the results are less conclusive. In some countries, they have a wage premium compared to similar heterosexual women (or women living in different-sex couples), while in others they have a wage penalty.
 
The literature that looks at the intersection of gender, race/ethnicity, and sexual orientation is scarce, mainly due to data availability, and is focused on the US. There is some evidence of heterogeneity in the wage premium among lesbians in this country, with the sexual orientation wage premium of Black lesbians being negligible, and the premium for White lesbians being lower than that of Hispanic lesbians [12]. This suggests that lesbians’ premium is racialized. Part of the sexual orientation wage premium of White and Hispanic lesbians seems to arise from having a more evenly distribution across occupations than their heterosexual peers. There are also significant differences in the penalties faced by gay men of different races/ethnicities in the US. The sexual orientation wage penalty seems to be greater for White people. Additionally, there are differences in the racial wage penalty among sexual-orientation male groups, the gap seems to be higher for heterosexual men. 

Class

To operationalize class, sociologists often use individuals’ position in the wage distribution, occupational status, or educational achievements. This means that class is determined at the individual level and is intertwined with either educational or labor outcomes. Unlike them, economists often employ family’s income and wealth as class indicators, which makes it clearer that social class transcends the individual. This approach allows to explore the connection between the social class of the family to which one belongs and the labor outcomes of each of its members. However, this approach has been barely used to address intergroup labor inequalities. An exception is a recent study that operationalizes social class using family income to show that the adjusted hourly wages of White, Black, and Hispanic women in the US are lower than those of any male group of the same social class [13]. This suggests that the gender effect is pervasive: it crosses all social classes (and races/ethnicities). This study also shows that the occupational barriers faced by Black women affect all social classes; it is detected even in the upper class. There is also evidence that White women do not thrive in the labor market as much as one could expect given their privileged positions in terms of class and race. One possible explanation is that the assortative mating that allows White women to reach a good position in the class structure may limit their progress in the wage distribution. However, more studies are needed to find out whether these patterns extend to other countries, and also whether they would remain if family wealth were used instead.

Limitations and gaps

Comparative studies based on race/ethnicity are more difficult to conduct than gender studies as different regions encompass different groups. In higher-income European countries, race/ethnicity is often identified with an immigrant background, while in Central and South-Eastern Europe, Roma is a major ethnic group. In North and South America other categorizations exist, often distinguishing among White, Black, and Indigenous people, inter alia. Moreover, a group’s position in the labor market is not independent of the socioeconomic, institutional, and historical context and, therefore, it is also subject to changes across regions and over time.

The genderization and racialization of work shape the opportunities that individuals face, including information access to jobs openings, probability of being hired in certain types of occupations, working conditions, promotion, and wages. However, most empirical intersectional studies focus on wages, mainly in the US, and there is limited evidence for other labor outcomes and countries. Additionally, intersectional studies do not necessarily address the same research questions (e.g., they often compare racial/ethnic groups of a single gender), which makes generalization difficult.

Another difficulty is that the number of observations for some minorities in many data sets is small, which limits their study. This is especially the case when trying to explain the sources of the labor market disparities, since this involves accounting for several characteristics (e.g., education, age, or geographical patterns) simultaneously. Sample size also restrains research dealing with more than two social identities. Other limitations arise from the variables included in data sets with respect to race/ethnicity, nativity, or sexual orientation. For example, in the European Union, the Structure of Earnings Survey does not include information about nativity, although this information is provided by the national statistical offices of some countries, thus limiting comparative studies on wage gaps by nativity (the European Labor Force survey includes birthplace but not individual wages). With respect to sexual orientation, many studies identify LGBTQ+ individuals based on same-sex partnership to have large samples, although recent large individual-level samples suggest that results based on couple-based samples do not necessarily hold for non-partnered workers.

Finally, many studies undertake racial/ethnicity or nativity analyses separately for women and men, or they compare women to men of the same group. Analogously, studies on the sexual orientation wage gap tend to compare lesbian to heterosexual women, and gay to heterosexual men. However, to have a complete picture of the situation of privilege and disadvantage, a full comparison of the incumbent groups should be conducted. The wage disadvantage of minority women (based on either race/ethnicity or sexual orientation) cannot be fully determined by comparing them to majority women or minority men. And although the gender wage gap of Black women in some countries is lower than that of White women, this does not mean that Black women fare better than their White peers when compared to White men. In the same vein, despite lesbians having a wage premium with respect to heterosexual women in some countries, and gay men having a wage penalty with respect to heterosexual men, lesbians’ wages are lower than those of either gay or heterosexual men (before and after adjusting for characteristics). Moreover, to fully understand intergroup disparities, one should look not only at the hindrances of disadvantaged groups but also at the advantages of the privileged ones.

Summary and policy advice

The reasons that explain the positions of the various groups of women (and men) in the labor market are not the same. For some racial/ethnic minorities (and immigrant groups), years of education play a significant role in explaining their wage disadvantages, especially those of men. This is the case of Hispanic men in the US and, to some extent, that of Hispanic women. This means that if these groups could increase their educational levels, a large part of their wage disadvantage would disappear. However, for other groups, years of education do not explain their positions on the wage distribution. In the US, the hourly wages of White and Asian women would fall significantly if they had the same education level as White men do (which is lower than theirs). Part of the hourly-wage disadvantage of these two female groups seems to stem from working part time. Therefore, policies that promote part-time work may halt the wage progress of these women. 

The literature also suggests that the concentration of immigrant women, sexual minority women, most racial minority women, and majority women in low-paid occupations goes beyond what one would expect based on their education levels and experience. This is so due not only to their underrepresentation in highly paid occupations but also from feminized occupations receiving lower wages than masculinized occupations with similar skill requirements. Combating occupational segregation would help reduce the wage gaps. But disadvantaged minorities also require social protection to help them during transition to better jobs given that they are more susceptible to labor exploitation, especially women of color and immigrant women.

Unlike their female peers, the concentration of many racial minority men and immigrant men in low-paying occupations would vanish if they had higher educational levels and, therefore, education attainment alone would help them move up the wage ladder. An exception among minority men is Black men in the US, who tend to be concentrated in low-paid occupations after accounting for years of education (and other basic characteristics). The clustering of majority men (and some minority men) in highly paid occupations seems to exceed their characteristics.

Although many countries have closed or even reversed the gender gap in years of education, gender disparities persist in field of study, which translates into wage gaps. Differences in the type of university that students attend could also play a role in explaining the racial/ethnic wage gaps. To reduce these gaps, policies should include awareness campaigns to motivate both girls and minority boys to pursue college studies associated with highly paid occupations in which they are underrepresented, together with campaigns to motivate majority boys to pursue traditionally feminized studies. This should be complemented with financial support for students whose options are limited by their family resources, which affects some minorities with especial intensity. It is also essential to fight against school segregation by class and race/ethnicity at earlier stages of education by providing public schools with the necessary resources to offer quality education.

Acknowledgments

The authors thank the anonymous referee(s) and the IZA World of Labor editors for many helpful suggestions on earlier drafts. This work was supported by grants PID2020-113440GB-100 and PID2022-137352NB-C42 funded by MCIN/AEI/ 10.13039/501100011033 and by “ERDF A way of making Europe.”

Competing interests

The IZA World of Labor project is committed to the IZA Guiding Principles of Research Integrity. The author declares to have observed these principles.

© Olga Alonso-Villar and Coral del Río Otero