During the past three decades, political science research has uncovered substantial evidence that race and gender influence representation in the United States. Historically, various institutionalized race and gender biases have worked not only to limit the number of women and minorities running for office but also to channel and confine their opportunities to certain majority-minority or “women-friendly” jurisdictions (Arceneaux Reference Arceneaux2001; Darcy, Welch, and Clark Reference Darcy, Welch and Clark1994; Davidson and Grofman Reference Davidson and Grofman1994; Lublin Reference Lublin1997; Lublin et al. Reference Lublin, Brunell, Grofman and Handley2009; Palmer and Simon Reference Palmer and Simon2012; Preuhs and Juenke Reference Preuhs and Juenke2011; Sanbonmatsu Reference Sanbonmatsu2006). Once in public office, African Americans are more likely than others to focus on interests and issues particularly relevant to African Americans and Latinx legislators are more likely to do the same on behalf of Latinx interests; similarly, women are generally more likely than men to focus their representational activity on women’s interests and issues (Bratton and Haynie Reference Bratton and Haynie1999; Canon Reference Canon1999; Casellas Reference Casellas2011; Griffin and Newman Reference Griffin and Newman2008; Grose Reference Grose2011; Haynie Reference Haynie2001; Minta Reference Minta2011; Osborn Reference Osborn2012; Reingold Reference Reingold2000; Rouse Reference Rouse2013; Swers Reference Swers2002, Reference Swers2013; Thomas Reference Thomas1994; Wilson Reference Wilson2017). Yet, it is still the case that little research has examined whether and how race and gender together simultaneously influence who our elected officials are (i.e., “descriptive representation”) and what they do in office (i.e., “substantive representation”) (Pitkin Reference Pitkin1967).
When studying race, gender, and representation, political scientists often assume that there are no gender differences among minority representatives and no racial differences among female representatives. More often than not, attention has been given only to what factors influence the descriptive and substantive representation of women or to what factors influence the descriptive and substantive representation of African Americans and/or Latinxs. These one-at-a-time, “single-axis” approaches (Crenshaw Reference Crenshaw1989) to the study of representation are clearly overly simplistic. As Hawkesworth (Reference Hawkesworth2003) and others implore, we must think about and study representation as not only raced and not only gendered but as “raced-gendered” (Brown Reference Brown2014; Fraga et al. Reference Fraga, Martinez-Ebers, Lopez, Ramírez and Reingold2008; Hardy-Fanta et al. Reference Hardy-Fanta, Lien, Pinderhughes and Sierra2016; Reingold Reference Reingold, Wolbrecht, Beckwith and Baldez2008; Smooth Reference Smooth2006, Reference Smooth2011).
In Race, Gender, and Political Representation, Kerry Haynie, Kirsten Widner, and I take up that call and examine how and to what extent political representation is simultaneously raced and gendered—in the context of late-twentieth- and early-twenty-first-century US state legislatures (Reingold, Haynie, and Widner Reference Reingold, Haynie and Widner2021). Instead of investigating which conditions are conducive for “minority” representation or for “women’s” representation and rather than ponder whether “women” and “minorities” in office are more likely to advocate on behalf of other women and minorities, respectively, we ask: How do gender and race interact to affect the election, behavior, and impact of all individuals—raced women and gendered minorities alike? Addressing this question, we argue, requires a more intersectional approach to the study of legislative representation. Indeed, our analysis demonstrates the power of intersectionality—as a critical research paradigm—for understanding the many complex ways that race and gender together shape democratic institutions and the representational opportunities and challenges they present.
What exactly does an intersectional approach like ours entail? Much of the work of intersectionality is accomplished by simply classifying legislators in terms of their race and gender identity and comparing their election, behavior, and impact in office. Most of our data analysis, for example, examines the presence, behavior, and impact of Black women, Latinas, white women, Black men, Latinos, and white men in state legislatures. But an intersectional approach is much more than that.
What exactly does an intersectional approach like ours entail?
Importantly, our intersectional approach places women of color at the center of the analysis—precisely because they (and others who stand at the intersections of multiple systems of disadvantage) are the ones whose experiences and accomplishments are most likely to be neglected and obscured by dominant single-axis approaches. In this way, our analysis often questions whether what we know about “minority” or “women’s” representation is reflective of only the dominant categories within—namely, men of color and white women. Centering women of color means privileging and building on what little existing research there is about them, especially as candidates and public officials. However, an intersectional approach also requires us to critically reread the single-axis literature in both race and ethnic politics and gender politics. As we reread, we look not only for the similarities and differences revealed but also for the intersectional, race-gender implications we can infer. This rereading provides a wealth of both single-axis and intersectional theories about descriptive and substantive representation to test, but it also prompts us to critically reexamine how we test those theories. Most notably, our intersectional approach to the study of representation closely examines issues of measurement, particularly how we define and operationalize key theoretical concepts such as group interests. Throughout the design and implementation of our research, intersectionality also compels us to critically evaluate our own analytic categories, checking to see whether the generalizations we make about women and men of color, for example, are valid.
By using this intersectional approach, we can learn much more about the complexities of race, gender, and representation in state legislatures. Our extensive analysis of bill sponsorship as a form of policy leadership and substantive representation (Reingold, Haynie, and Widner, ch. 3 and 4) illustrates our approach to intersectionality and its empirical utility especially well. We began by reexamining foundational, single-axis conceptions of marginalized-group interests and issues that lie at the heart of any analysis of legislative activity on behalf of women and/or minorities. To gauge the links between descriptive and substantive representation, political scientists have used various definitions of group interests, often distinguishing between more narrowly defined “racial” or “women-specific” issues and broader issues, such as health and education, that also are salient but less explicitly or directly tied to gender, race, or ethnicity alone. To what extent, then, do our definitions of group interests affect who is or appears to be more or less willing to act for African Americans, Latinxs, and women?
Intersectionality cautions against generalizing about representation across differences in race and gender, suggesting that any single-axis conception of marginalized-group interests risks concealing or distorting the representational advocacy provided by women of color while privileging that which is provided by white women and men of color. The more narrow and single-group specific definitions of women’s/Black/Latinx issues may be particularly problematic, especially compared to broader issue areas such as health and education, which are salient to women, African Americans, and Latinxs alike (Smooth Reference Smooth2011).
To test this proposition, we examined the agenda-setting policy leadership (i.e., bill sponsorship) of Democratic (and, to a lesser extent, Republican) state legislators in 15 state houses, in 1997 and 2005, across multiple definitions of group issues and interests.Footnote 1 We found that no matter which definition of group interests we used, Democratic women of color never appear any less committed to providing substantive representation than anyone else. They sponsor as many women-specific bills as their white female colleagues, as many Black- and/or Latinx-specific bills as their minority male colleagues, and more health and education bills than anyone else. Among Democrats and Republicans, women of color are the leading sponsors of health and education bills targeted to address the interests of women or racial and ethnic minorities in particular. Thus, we concluded that research that relies only on narrowly defined, group-specific conceptions of policy interests will overlook and underestimate the truly distinct representational leadership of women of color—especially as they work to shape healthcare and education policy to address the interests of women and people of color.
Next, we looked beyond single-axis conceptions of group interests and substantive representation to explore what we call “race-gender policy leadership.” To what extent and how do representatives address both race and gender in their policy-making initiatives? Who is more or less likely to do so? Recognizing that various group interests (i.e., Black, Latinx, and women’s) can be—or at least appear to be—distinct, overlapping, or intersecting (Brown and Banks Reference Brown and Banks2014; Minta and Brown Reference Minta and Brown2014), we distinguished and measured multiple approaches to race-gender policy leadership among the same set of Democratic (and Republican) lawmakers serving in 15 US state houses in 1997 and 2005. Specifically, we tested hypotheses that legislative women of color are more likely than others (including white women and men of color) to sponsor (1) both narrowly targeted, group-specific women-interest and minority-interest bills, one-at-a-time; (2) bills that simultaneously address multiple group-specific interests (e.g., standard antidiscrimination and affirmative-action measures); and (3) bills that address the particular interests of disadvantaged subgroups of women and/or minorities, such as poor women of color (Strolovitch Reference Strolovitch2007).
Although our conceptions of race-gender policy making are fairly inclusive, our data show that relatively few legislators engage in this type of leadership, even among Democrats. Nonetheless, women of color (in both parties) often assume important leading roles in addressing the policy needs of multiple and multiply disadvantaged groups. Depending on the measure of race-gender policy leadership, either Black women or Latinas stand out from their peers, sponsoring more race-gender legislation than their minority male or white female counterparts. Among Democrats, Latinas stand out as the most likely to sponsor at least one women-specific bill and one Latinx- and/or Black-specific bill. Black women are more likely than any other group of Democrats to sponsor at least one welfare/poverty bill addressing the interests of intersectionally disadvantaged subgroups of women and people of color. Among Republicans, Latinas are the most likely to do the same. The only type of race-gender policy leadership in which women of color do not stand out is the sponsorship of bills that simultaneously address multiple forms of discrimination and inequality. Black men are the leading sponsors of these measures (although not by wide margins), almost all of which offer traditional civil rights approaches to remedying discrimination on the basis of sex/gender, race/color, or national origin/ethnicity.
These are only a few of the race-gender similarities and differences evident in the experiences, activities, and accomplishments of the state legislators that we studied. Again and again, we saw how the continued shortage of elected women of color undermines the effective substantive representation of multiple and multiply disadvantaged groups—and how intersectionality matters. Few if any of our conclusions about race, gender, and representation would have been revealed had we not taken a more intersectional approach to examine a multitude of race-gender similarities and differences among minorities and women, as well as among women and men of color. Without intersectionality, we would be bereft of the critical race-gender questions that prompted our research and the race-gender analytic tools with which to address them.
Nonetheless, our study is only one of many “first steps” toward a more intersectional approach to the study of race, gender, and political representation. To claim otherwise—that our work should stand alone or be considered the final word—would undermine the integrity and power of intersectionality as a critical and productive research paradigm (Davis Reference Davis2008). As scholars and citizens, we must continue to ask and pursue difficult, “messy” (Smooth Reference Smooth2006), and “unsettling” (May Reference May2015) questions about complex intersections of multiple axes of identity and power as they relate to political representation and our ability to understand it. Intersectional research always must be (considered) provisional, tentative, and partial because we always can move toward a more—or different—intersectional approach (Carbado et al. Reference Carbado, Crenshaw, Mays and Tomlinson2013; May Reference May2015, 251; McCall Reference McCall2005). Race and gender, the central foci of our inquiry, are important, especially given the history of American politics and its study. They are a good starting point. However, there undoubtedly are multiple layers of intersectional complexity that warrant further investigation.
