The Intersectional Effects of Diverse Elections on Validated Turnout in the 2018 Midterm Elections

Abstract

The 2018 midterm elections in the United States were unprecedented in their gender and racial diversity. Voters across the country, especially younger voters, elected the most diverse U.S. Congress in history. Despite increased electoral diversity along lines of gender, race, and the intersections of both, extant literature has remained siloed, focusing on the effect of either gender or race on turnout but rarely examining both in relation to one another. Using a novel data set of racially diverse young adults that includes demographic information for congressional candidates and vote-validated data, this study investigates how the intersection of race and gender influence voter turnout across diverse electoral contexts. Our study provides important insights for both unpacking the 2018 elections and more generally understanding how race and gender interact to influence youth voter turnout as candidate profiles and electoral contexts continue to diversify.

Keywords

voter turnout political behavior race ethnicity young voters

Introduction

The 2018 midterm elections in the United States were unprecedented in their gender and racial diversity. Voters across the country—particularly young voters—elected record numbers of women and people of color to office, resulting in the most diverse U.S. Congress in history. For the first time, about a quarter of the members of the House of Representatives were people of color, and approximately 23 percent were women (DeSilver 2018; Geiger, Bialik, and Gramlich 2019). Furthermore, the demographic diversity of Congress reached across boundary lines of race or gender. Of the twenty-four people of color in the 2018 freshman class of representatives over half were women, increasing the total number of women of color in the House to a record high of forty-three (Center for American Women and Politics [CAWP] 2018). These gains were in no small part due to increased turnout of younger voters. Voter turnout among young voters jumped from 20 percent in 2014 to 36 percent in 2018, the largest increase for any group (Misra 2019).

Despite a marked improvement in the racial and gender composition of its members, however, the demographics of Congress are still below parity compared with the country overall, which is growing increasingly diverse with every generation (Frey 2018; Medenica 2018; Rouse and Ross 2018). Even with their record gains in 2018, women are underrepresented in Congress, making up 51 percent of the population but only 23 percent of the House of Representativeness. People of color are similarly underrepresented. African Americans, for example, constitute 13 percent of the population and only 9 percent of Congress. Lack of adequate representation is even more exaggerated for Latinxs, who make up roughly 18 percent of the population but only 6 percent of Congress. Nevertheless, the growing number of women and people of color running for office coupled with higher rates of youth turnout raise the salience of important questions regarding voter behavior, particularly at a time when the demographics of the country are also changing substantially. How is one’s race and/or gender related to voter turnout? Do the presence of women and/or minority candidates influence voter turnout? And, if so, is there variation across individual race, gender, and/or region?

Although these questions are not new to political science research, they warrant revisiting. Existing literature on the effects of race and gender on voter turnout has remained relatively siloed, focusing on the effect of either gender or race on political behavior but rarely examining both gender and race in relation to one another (Brown 2014; Junn and Brown 2008; Smooth 2006). What’s more, studies that have focused on the effects of candidate characteristics like gender and race on turnout often focus on one racial or gender group in the electorate and rarely examine effects across groups (Barreto 2007, 2010; Dolan 2006; McElroy and Marsh 2009). This categorical rather than comparative approach to studying the effects of candidate characteristics on individuals often misses important and electorally consequential reactions of voters who belong to the nontarget group of study (McConnaughy et al. 2010; Sapiro and Conover 1997; Tesler 2016). In this study, we bridge related yet parallel work on the effects of race and ethnicity and gender on voter turnout by outlining how and under what electoral conditions race, gender, and the intersections of race and gender affected voter turnout among young adults in the 2018 congressional elections.¹ In doing so, this article contributes to the literatures on the effects of race, gender, and candidate identities on political participation and to the small but growing and critically important body of work on how intersectional race–gender identities shape political behavior.

We address the limitations of previous work by using a novel data set of racially diverse young adults, the demographic future of the U.S. electorate, that includes individual-level vote-validated data for 2018 as well as demographic information for 2018 congressional candidates across congressional districts in the United States. Moreover, instead of “controlling” for race and/or gender as analyses often do, we acknowledge that women and people of color occupy structurally distinct social locations in society and thus take care to disaggregate our analyses by race, gender, and race–gender (Chow, Wilkinson, and Zinn 1996; Dawson and Cohen 2002; Junn 2017; Masuoka and Junn 2013).

When applying this analytical strategy of racial and gender disaggregation, we find that the drivers of youth voter turnout in 2018 differed significantly between whites and people of color, between men and women, and between communities of color. We also find evidence that elections in districts with more candidates of color, particularly women candidates of color, led to increased turnout among people of color, driven primarily by black women. Interestingly, we also found increased turnout among white men in Southern districts featuring candidates of color, perhaps providing further evidence of racialized countermobilization (Jardina 2018; Major, Blodorn, and Blascovich 2016; Reny, Collingwood, and Valenzuela 2019; Schaffner, Macwilliams, and Nteta 2018). We do not find evidence of significant turnout effects of gender-diverse elections among any race, gender, or race–gender group. These findings provide important insights for unpacking the 2018 elections and contribute to the rich literature on the effects of descriptive representation by examining how race and gender interact to influence youth voter turnout, a topic that will only continue to increase in importance as candidate profiles and electoral contexts continue to diversify alongside the population in the United States.

Gender, Race, and Voter Turnout: What We Know and Should Expect

Gender

Much of the existing scholarship on the politics of gender has focused largely on comparisons between men and women, highlighting several important gaps in political attitudes and behavior (Burns, Jardina, and Yadon 2017; Huddy and Cassese 2011). While a substantial amount of this work has focused on differences in policy preferences (Baxter and Lansing 1983; Shapiro and Mahajan 1986) and subsequent partisan identification and vote choice (Box-Steffensmeier, De Boef, and Lin 2004), there has also been a noticeable and growing gap in voter turnout between men and women since 1980 (Carroll 2006; CAWP 2017). For the past several decades, women have shown up to the ballot box, consistently turning out for presidential and midterm elections, including the 2018 midterms, at higher rates than men (Hartig 2019). Puzzlingly, however, scholars have struggled to explain the drivers behind the higher turnout rates exhibited by women, especially as women tend to participate in other forms at rates lower than men (Burns, Scholzman, and Verba 2002; Lizotte and Sidman 2009; Mondak and Anderson 2004).

Nevertheless, as the numbers of women running for elected office have increased in recent years, scholars have investigated the effects of female candidates on voting. Research on whether the presence of a woman candidate on the ballot impacts turnout has been decidedly mixed and conditional on a number of factors. For example, some work has found that women candidates are more likely to campaign on issues of interest to women, which may disproportionately appeal to women voters and impact turnout, particularly in elections where those issues are salient (Dabelko and Herrnson 1997; Dolan 2008; Herrnson, Lay, and Stokes 2003). Support for female candidates is also strongly tied to women’s stronger support of social services and their greater emphasis on social welfare issues in deciding for whom to vote (Huddy, Cassese, and Lizotte 2008).

Perhaps one reason for finding evidence of conditional effects may be that most of the work that examines how people respond to the presence of women candidates has focused on a single election or a small number of elections (Burns, Scholzman, and Verba 2001; Koch 1997; Sapiro and Conover 1997). Yet, when using national data spanning from 1990 to 2004, Dolan (2006) finds little empirical evidence of the presence of women candidates translating into any systematic influence of individual political attitudes or behaviors. The mere presence of a woman candidate on the ballot, it seems, has no significant effect on voter turnout among women or men. In light of these previous findings, we expect to find that gender was a significant predictor of voter turnout in 2018 but that elections featuring women candidates (i.e., gender-diverse contexts) did not have a significant effect on turnout for either women or men.

Race and Ethnicity

The turnout gap between whites and people of color is a staple in American politics. Indeed, recent work suggests that this gap is widening even as the U.S. population continues to become more racially and ethnically diverse (Fraga 2018). Lower rates of voting and participation among people of color have profound implications on representation, especially as their proportion of the population grows (Clifford 2012; Hajnal 2010; Hajnal and Trounstine 2016). Scholars have identified numerous reasons for this gap, including lack of resources (Leighley and Nagler 2013; Rosenstone and Hansen 1993; Wolfinger and Rosenstone 1980), nonexistent or ineffective mobilization by parties and campaigns (García and Michelson 2012; Ramírez 2013), and perceived lack of electoral influence in elections (Fraga 2018).

Scholars have also identified important ways for addressing the turnout gap between whites and people of color. Chief among these pathways are candidates of color running for office. In contrast to the mixed effects of candidate gender on voter turnout, findings from work on race and ethnic politics suggest that candidates of color running for office are generally associated with increased turnout among people of color (Barreto 2007, 2010) due, in large part, to strong historically forged bonds of racial solidarity (Chong and Rogers 2005; Dawson 1994; Gutierrez et al. 2019; Masuoka 2006; Sanchez and Masuoka 2010; Stokes 2003). Perhaps the most prominent example of the effects of candidate co-ethnicity on voter turnout has been the election of Barack Obama in 2008, which witnessed significant increases in African American turnout compared with previous elections (McKee, Hood, and Hill 2012; Osborn, McClurg, and Knoll 2010; Philpot, Shaw, and McGowen 2009). Indeed, recent election outcomes have underscored the reality that African Americans, especially African American women, are the backbone of the Democratic Party and American left more generally (Dowe 2016).

The increased salience of race in American politics also has consequences for the attitudes and behavior of whites (Gay 2001). Recent scholarship suggests that demographic shifts toward a majority-minority population and the election of America’s first African American president have activated group anxiety among whites, leading them to identify more strongly with their racial group and express more conservative attitudes and preferences (Abrajano and Hajnal 2017; Jardina 2018; Reny, Collingwood, and Valenzuela 2019; Sides, Tesler, and Vavreck 2018; Tesler 2016). This is particularly true in the American south, a region that scholars have long identified as being politically distinct from the rest of the United States (Key 1949) and where the contemporary political attitudes of whites can, at least in part, be traced back to the political legacy and geography of slavery (Archaya, Blackwell, and Sen 2016, 2018).

Given these twin sets of findings on the turnout gap and the effect of candidates of color on voter turnout, our expectations for how race and racial context impacted voter turnout in 2018 differ somewhat from our expectations as they relate to gender. In line with existing trends in voter turnout by race and ethnicity, we expect that whites more likely to vote than people color. Considering extant work on the importance of candidate race and ethnicity to voter attitudes and behavior, we also expect that elections featuring racially and ethnically diverse candidates were associated with higher turnout among people of color and Southern whites.

Political Participation at the Intersection of Race and Gender

Although research on gender and race have advanced political science in necessary and crucial ways, neither gender nor race alone is sufficient for a comprehensive understanding of political attitudes and behavior. Scholarship that focuses on one or the other but does not consider both in relation to one another often ignores systematic intragroup differences and fails to account for how the intersections of identity characteristics uniquely structure individual experiences and shape behavior. Coined and formulated by black feminist scholars and activists (Collins 1990; Crenshaw 1988, 1991; Hancock 2016), an intersectional framework, that is to say an analytical framework that moves beyond essentialist perspectives on categories like race or gender to recognize that social categories overlap in ways that create distinctive experiences and material outcomes, allows scholars to “account for multiple grounds of identity when considering how the social world is constructed” (Crenshaw 1991, 1245). Portraying or understanding groups along one identity dimension without acknowledging intersections with other dimensions of identity flattens complexity and treats groups as homogenous categories.

Categories of gender and race and ethnicity are not monoliths nor are they mutually exclusive (Brown 2014). Instead, both are composed of diverse, heterogeneous, and overlapping subgroups. Women of color, for instance, exist within both racial/ethnic and gender categories concurrently and, as such, experience their race and gender simultaneously (Brown and Gershon 2016; Hancock 2007). Systematic differences by race within gender categories, and by gender within categories of race, were on display in the 2016 presidential election. Within racial categories, majorities of white men and white women voted for Donald Trump while women of color and men of color instead preferred to cast a ballot for Hillary Clinton (Cassese and Barnes 2018; Frasure-Yokley 2018; Junn 2017; Major, Blodorn, and Blascovich 2016; Phillips 2018). Within gender categories, there were significant differences in the level of support for each candidate, with greater numbers of white men than white women voting for Trump and greater numbers of women of color than men of color voting for Hillary Clinton (Junn 2017). What’s more, previous research has similarly demonstrated that African American women consistently vote at higher rates than African American men in presidential elections and often score higher on measures of political participation than both African American men and white women (Baxter and Lansing 1983; Simien 2006).

Intragroup differences are also incredibly important in understanding the political experiences, attitudes, and behaviors of pan-ethnic groups like Latinx and Asian Americans. Scholars of Latinx and Asian American politics are well versed in examining within-group differences and grappling with the complexity of collapsing difference into pan-ethnicity (Beltrán 2010; Lien, Conway, and Wong 2004). In addition to differences by gender (Lavariega Monforti 2017), the attitudes and behaviors of Latinx and Asian American individuals also vary along lines of nativity, nation of origin, and language abilities, just to name a few examples (Jones-Correa 1998; Wong et al. 2011).

Given the historic number of campaigns by women and people of color but those at the intersection of race and gender—the number of women of color—for public office in 2018, attending to within-group differences is especially important for understanding the 2018 midterm elections. As Simien (2015, 4) finds in her work on the pathbreaking campaigns of Shirley Chisolm in 1972, Jesse Jackson in 1984, and Barack Obama and Hillary Clinton in 2008, “the mere presence of a ‘historic first’ [candidate] who mirrors a marginalized group pictorially signals greater access to electoral opportunities and, at the same time, motivates political agency.” According to Simien, the public visibility of historic candidates raises the salience of identities shared by the candidate with segments of the mass public, such as race, ethnicity, and gender. This newfound identity alignment between candidates and voters is “symbolically empowering” and mobilizing for certain populations. For many voters across the country in 2018, the midterm elections were the first time they were able to vote for candidates who are women, candidates of color, and, importantly, candidates who are women of color. Similar to Simien’s analysis of “historic first” presidential candidates, the historic candidacies of 2018 may have symbolically empowered those traditionally understood to stand on the periphery of mainstream electoral politics, namely, young women of color and particularly black women.

In this article, we build off of existing literature that bridges work on race and gender and approach our analyses using an intersectional race–gender framework. We expect that any effects of race and racial context on the probability of people of color casting a ballot in 2016 is largely driven by women of color, especially black women. Conversely, we expect that the effects of racial context among whites will be greater among white men than white women.

Data and Methods

We investigate how the intersection of gender and race and ethnicity influence youth voter turnout across diverse electoral contexts using an original data set compiled from two distinct sources: (1) the 2019 February GenForward Survey at the University of Chicago, and (2) the Daily Kos’s 2018 Congressional Election Candidate Demographics Tracker.

The GenForward Survey at the University of Chicago is an ongoing, nationally representative survey of young adults between the ages of eighteen and thirty-four and is particularly well suited for analyses of youth voter turnout by race and ethnicity because it is one of the few nationally representative surveys that offers validated measures of voter turnout as well as oversamples of African Americans, Latinxs, and Asian Americans. Oversampling communities of color allows researchers to examine not only differences by race and ethnicity, which continues to be a challenge in many surveys, but also how race and ethnicity intersect with gender (Barreto et al. 2018). Having adequate sample sizes for populations of color is especially important for studies of young adults; adults between the ages of eighteen and thirty-four represent the most racially and ethnically diverse segment of the American populations (Frey 2018).

The 2019 February GenForward Survey includes validated vote data for 2,134 respondents, of which 763 racially identified as white (315 men, 448 women), 547 as African American (191 men, 356 women), 515 as Latinx (222 men, 293 women), and 251 as Asian American (131 men, 120 women). These data were collected by the survey research firm NORC,² via both web and telephone modes with in-person follow-up efforts to minimize nonresponse. Surveys were conducted in both English and Spanish and weighted to match population characteristics using appropriate U.S. Census benchmarks.³

The Daily Kos has a long history of compiling detailed data on U.S. elections. The 2018 Congressional Election Candidate Demographics Tracker data set compiled demographic information, including gender and race and ethnicity, for all major congressional House candidates in 2018 across all 435 congressional districts in the United States. In cases where candidate demographic data were missing or incomplete, we hand-coded the missing information so that the final data set included complete gender and race and ethnicity data for the universe of major congressional candidates.⁴

Merging individual-level responses from GenForward to the candidate data set first required matching respondent zip codes to their appropriate congressional districts. As congressional district boundary lines are often irregular and overlap multiple congressional districts, we assigned zip codes to districts using a two-step approach that first identifies the geographic center, or the “centroid,” of the zip code and, second, assigns the zip code to whichever congressional district contains that centroid. Although there is some inevitable error in this matching approach as it is possible that the centroid of a zip code lies in a district that does not contain the majority of the population in a zip code area, previous work has found this error to be minimal, resulting in a 92 percent accuracy rate (Gimpel, Lee, and Kaminski 2006). The final merged data set comprised 2,118 respondents, of which 758 racially identified as white (313 men, 445 women), 542 as African American (189 men, 353 women), 511 as Latinx (220 men, 291 women), and 249 as Asian American (131 men, 118 women), a difference of only sixteen respondents from the original nonmerged data set.

Informed by previous literature on the politics of race and ethnicity, the politics of gender, as well as the effects of race and gender-diverse elections on voter behavior, we examine the roles of race, gender, and intersecting race–gender identities on voter turnout across election contexts using the analytical strategy of comparative relational analysis (Masuoka and Junn 2013). In applying this strategy to our analyses, we estimate separate models by the identity categories of race, gender, and the intersections of race and gender and then compare results across models, an increasingly popular analytical strategy in studies of marginalized groups. Estimating separate models acknowledges that race, gender, and race–gender identities structure individual experiences in unique and important ways and allows for the estimation of structurally different relationships between explanatory variables, like diverse election contexts, and voter turnout. Specifying separate models is an especially important strategy for intersectional analyses of race and gender due to the distinctive standings of women of color, men of color, white women, and white men in the United States (Brown 2014).

Dependent Variable

The primary dependent variable in our analyses is voter turnout in the 2018 midterm elections. We use data from the GenForward Survey, which is the first nationally representative survey to offer validated voter turnout data for the 2018 midterm elections.

To validate estimates of voter turnout in the GenForward Survey, all cases were sent to the firm Catalist, who then matched the cases against their database of registered voters in the United States. Each individual is coded as 1 if they are successfully matched to the database, and there is a record of them voting in 2018 (n = 963). Individuals are coded as 0 if they are successfully matched to the database, and there is a record of them not voting in 2018 (n = 631). Individuals who were not successfully matched to the database likely mean that they are not a registered voter and very likely did not vote. Because it is highly unlikely a nonmatched individual would have voted in 2018, we code respondents who were not successfully matched to the database as nonvoters (n = 524). After applying the appropriate coding, 963 respondents in the sample are validated voters and 1,155 are nonvoters, resulting in a final weighted turnout rate of 37.8 percent among eighteen to thirty-four-year-olds.⁵

Independent Variables

There are five key independent variables this study: (1) race, (2) gender, (3) the racial diversity of elections, (4) the gender diversity of elections, and (5) region, namely, the south. To disaggregate our analyses by race, we code respondents as 1 if they identify as a member of a particular racial group and 0 if they do not identify as a member of that racial group. Gender is also treated dichotomously, with male respondents coded as 1 and female respondents coded as 2.

Electoral context variables of the race and gender diversity of elections are coded on a three-point scale ranging from “least” to “most” diverse.⁶ For the racial context variable, elections are coded as 1 if there are two white candidates running against each other; 2 if there is one white candidate and one candidate of color running against each other; and 3 if there are two candidates of color running against each other. We use a similar coding strategy for the gender context variable. Roughly 61 (45) percent of respondents were in districts where two white (two male) candidates were running, 33 (47) percent were in districts were one white (one male) candidate and one minority (one female) candidate were running, and 7 (8) percent were in districts where two minority (two female) candidates were running for congressional office.

We approach our analysis in three stages: (1) modeling voter turnout in the absence of election context, (2) modeling voter turnout with the inclusion of election context, and (3) because of several notable high-profile races and the unique racial politics of the south (Archaya, Blackwell, and Sen 2016, 2018), modeling voter turnout in and out of the south. Due to the dichotomous nature of the dependent variable, we use logistic regression to model voter turnout as a function of race, gender, the intersection of race and gender, and electoral context. While we estimate separate models by race, gender, and race-gender categories, we also include interaction terms to further investigate and ultimately uncover key differences within models. The use of interaction terms highlights the importance of understanding intragroup differences among expansive, heterogeneous categories like race and gender (Brown 2014).

Beyond these key independent variables, each model also includes a number of standard individual-level variables that have been shown to impact voter turnout and political participation generally. These include party identification, education, age, income, employment status, marital status, and region.⁷ We further control for differences across geography by including district-level covariates obtained from the 2010 decennial census and the 2012–2017 American Community Survey (ACS). These include percent urban, black, Latinx, and Asian American/Pacific Islander (AAPI). Finally, because we recognize that electoral context varies across districts, we also include measures of election competitiveness⁸ and whether or not there is a concurrent Senate election.

Results

Disaggregating Voter Turnout

We begin by modeling voter turnout in the absence of election context. Table 1 presents results from an aggregate model of turnout (column 1), disaggregated models by race (columns 2 through 5), disaggregated models by gender (columns 6 and 7), and the interaction of race and gender (column 8) devoid of electoral context.

Table 1.

Logit Coefficients for 2018 Validated Turnout.

	Base	White	Black	Latinx	Asian American	Men	Women	Interaction
White	0.418**	—	—	—	—	0.799***	0.014	—
White	(0.146)					(0.217)	(0.189)
Female	−0.007	−0.344	0.806**	0.133	0.537	—	—	—
Female	(0.135)	(0.210)	(0.252)	(0.233)	(0.351)
Party ID	−0.151***	−0.150**	−0.118	−0.135	−0.194	−0.121*	−0.163**	−0.141***
Party ID	(0.040)	(0.055)	(0.102)	(0.086)	(0.125)	(0.060)	(0.055)	(0.041)
Education	0.638***	0.719***	0.425**	0.969***	0.012	0.704***	0.633***	0.658***
Education	(0.090)	(0.146)	(0.162)	(0.146)	(0.313)	(0.133)	(0.114)	(0.093)
Age	−0.013	−0.018	0.032	0.036	−0.147*	−0.043	0.012	−0.014
Age	(0.018)	(0.027)	(0.034)	(0.030)	(0.058)	(0.027)	(0.022)	(0.018)
Income	0.061***	0.099***	0.027	−0.020	0.147***	0.093***	0.038	0.066***
Income	(0.018)	(0.029)	(0.032)	(0.028)	(0.043)	(0.028)	(0.022)	(0.018)
Employed	−0.005	0.010	−0.012	−0.028	−0.245	0.011	0.027	−0.028
Employed	(0.172)	(0.286)	(0.264)	(0.275)	(0.438)	(0.281)	(0.197)	(0.173)
Married	0.136	0.529*	−0.704	−0.016	−0.450	0.265	0.039	0.154
Married	(0.157)	(0.234)	(0.447)	(0.302)	(0.514)	(0.254)	(0.198)	(0.160)
South	0.063	−0.020	0.289	0.149	1.060*	0.126	−0.081	0.041
South	(0.166)	(0.261)	(0.332)	(0.296)	(0.475)	(0.247)	(0.209)	(0.166)
Competitive election	−0.001	−0.001	−0.004	0.003	−0.002	0.000	−0.003	−0.001
Competitive election	(0.003)	(0.005)	(0.005)	(0.005)	(0.008)	(0.005)	(0.004)	(0.003)
Concurrent senate	0.410*	0.650**	0.149	−0.062	−0.073	0.288	0.531*	0.445**
Concurrent senate	(0.168)	(0.233)	(0.284)	(0.372)	(0.598)	(0.253)	(0.212)	(0.169)
Percent urban	0.007	0.006	0.023*	0.002	0.024	0.011	−0.000	0.006
Percent urban	(0.005)	(0.008)	(0.011)	(0.012)	(0.017)	(0.008)	(0.007)	(0.005)
Percent black	0.003	−0.003	0.005	−0.008	−0.006	0.002	0.007	0.002
Percent black	(0.006)	(0.011)	(0.009)	(0.013)	(0.021)	(0.010)	(0.008)	(0.006)
Percent Latinx	−0.001	−0.016	0.001	0.009	0.012	−0.002	0.004	−0.001
Percent Latinx	(0.005)	(0.009)	(0.014)	(0.007)	(0.015)	(0.007)	(0.006)	(0.005)
Percent AAPI	−0.012	0.007	−0.022	−0.010	0.001	−0.017	−0.010	−0.009
Percent AAPI	(0.012)	(0.023)	(0.038)	(0.020)	(0.023)	(0.017)	(0.018)	(0.012)
Female	—	—	—	—	—	—	—	−0.357
Female								(0.201)
Black	—	—	—	—	—	—	—	−0.851**
Black								(0.279)
Latinx	—	—	—	—	—	—	—	−0.676**
Latinx								(0.246)
Asian	—	—	—	—	—	—	—	−1.173***
Asian								(0.300)
Female black	—	—	—	—	—	—	—	1.249***
Female black								(0.343)
Female Latinx	—	—	—	—	—	—	—	0.560
Female Latinx								(0.311)
Female Asian	—	—	—	—	—	—	—	0.867*
Female Asian								(0.434)
Constant	−3.102***	−2.819**	−5.726***	−4.361***	−0.604	−3.488***	−2.870***	−2.663***
Constant	(0.657)	(1.000)	(1.435)	(1.219)	(2.365)	(0.919)	(0.762)	(0.654)
Observations	2,036	750	536	506	244	846	1,190	2,036
Pseudo R²	.102	.148	.0985	.136	.188	.130	.0987	.112
AIC	1.229	1.843	0.730	0.994	0.764	1.474	1.050	1.221

AAPI = Asian American/Pacific Islander; AIC = Akaike information criterion.

p < .05. **p < .01. ***p < .001.

Estimating a single model that pools respondents, as we do in column 1, confirms long-standing findings on the importance of race, education, and income in predicting turnout. Being white, college educated, and earning a higher income were associated with an increased predicted probability of voting in 2018 of 10, 15, and 2 percentage points, respectively.⁹ Interestingly, and perhaps evidence in support of the “Blue Wave” narrative that defined much of the 2018 election coverage, we also find that partisanship was significantly associated with turnout; Republicans were significantly less likely to vote than Democrats in 2018.¹⁰ Gender was not significantly associated with voter turnout in any way.

Lumping respondents into a single model obscures the possibility of observing different sets of relationships between variables that arise as a function of one’s race or gender. Indeed, when we disaggregate respondents and estimate separate models by race, as we do in columns 2 through 5, we find that gender is significantly and positively associated with turnout among black Americans only. Black women were approximately 19 percentage points more likely to vote than black men in 2018. An individual’s probability of voting in 2018 did not significantly differ by gender among any other racial group.

A similar pattern of asymmetrical effects emerges when we disaggregate respondents by gender instead of race, as we do in columns 6 and 7. Race does not have the same effect on turnout among men and women. Instead, race is only significantly associated with turnout among men; white men were approximately 19 percentage points more likely to vote than men of color. Race was not associated with turnout among women. Interestingly, partisanship did not affect men’s probability of voting. Male voters had a statistically equivalent probability of voting in 2018 regardless of whether they identified as Democrat or Republican. This was not the case among women. Identifying as a strong Republican was associated with a decreased predicted probability of voting of roughly 21 percentage points compared with women who identified as strong Democrats. Although these results lend support and credibility to media narratives of women being integral to the outcome of the midterm elections, and perhaps a reaction to the #MeToo movement and actions of the Trump administration, they cannot adjudicate between Republican women staying at home and not voting versus increased turnout among women who identify as Democrats.

Disaggregating our analyses by race and gender demonstrates that the drivers of voter turnout in 2018 varied significantly between groups. A related modeling strategy for empirically uncovering key differences within groups is the use of interaction terms. Interacting two independent variables, such as race and gender, in a single model allows us to test whether relationships between independent and dependent variables are conditional on a third explanatory variable (Brambor et al. 2006; Friedrich 1982). In this case, we examine whether the effect of gender on turnout in 2018 is conditional on race. Setting up our analysis in this way offers a stricter test of any conditional relationships because all other explanatory variables in the model are held equal instead of being allowed to vary by group.

Results of the interaction model are presented in column 8 of Table 1. When interacting gender and race using this framework, we again find that black women had a significantly higher probability of voting in 2018 when compared with black men. We see similar differences between Asian American women and Asian American men. There were no such differences between men and women among white or Latinx Americans.

Investigating the Effects of Electoral Context on Voter Turnout

Given the diversity of the 2018 congressional elections (CAWP 2018; Geiger, Bialik, and Gramlich 2019) and extant literature on the effects of candidate characteristics on political behavior (Barreto 2007, 2010; Dolan 2006), we extend our analysis by including variables related to an election’s race and gender context. As previously, Table 2 presents results from an aggregate model of voter turnout (column 1), disaggregated models by race (columns 2 through 5), and disaggregated models by gender (columns 6 and 7). We also estimate models by intersecting race–gender identities in Table 3.

Table 2.

Logit Coefficients of 2018 Validated Turnout with Election Context (by Race & Gender).

	Base	White	Black	Latinx	Asian American	Men	Women
Racial Context	0.163	0.098	0.497	−0.079	0.194	0.204	0.180
Racial Context	(0.159)	(0.267)	(0.273)	(0.265)	(0.382)	(0.248)	(0.198)
Gender Context	−0.107	−0.211	−0.124	0.207	0.152	−0.001	−0.235
Gender Context	(0.117)	(0.181)	(0.238)	(0.220)	(0.287)	(0.182)	(0.146)
White	0.343*	—	—	—	—	0.728**	−0.056
White	(0.156)					(0.235)	(0.203)
Female	0.012	−0.318	0.861**	0.120	0.786*	—	—
Female	(0.147)	(0.218)	(0.276)	(0.267)	(0.385)
Party ID	−0.150***	−0.116*	−0.141	−0.191*	−0.323*	−0.107	−0.166**
Party ID	(0.043)	(0.058)	(0.106)	(0.087)	(0.137)	(0.064)	(0.058)
Education	0.662***	0.737***	0.440*	1.128***	0.001	0.765***	0.628***
Education	(0.097)	(0.149)	(0.173)	(0.170)	(0.365)	(0.143)	(0.119)
Age	−0.016	−0.016	0.027	0.012	−0.116	−0.057*	0.021
Age	(0.019)	(0.027)	(0.035)	(0.035)	(0.062)	(0.029)	(0.023)
Income	0.074***	0.111***	0.028	0.001	0.158**	0.105***	0.057*
Income	(0.019)	(0.030)	(0.034)	(0.032)	(0.049)	(0.030)	(0.023)
Employed	0.008	0.033	−0.036	−0.142	−0.011	0.089	0.009
Employed	(0.189)	(0.296)	(0.288)	(0.319)	(0.482)	(0.312)	(0.211)
Married	0.161	0.441	−0.667	0.135	−0.511	0.249	0.102
Married	(0.168)	(0.238)	(0.469)	(0.335)	(0.510)	(0.275)	(0.214)
South	−0.041	−0.063	0.323	−0.095	0.735	0.104	−0.254
South	(0.177)	(0.272)	(0.340)	(0.325)	(0.487)	(0.266)	(0.218)
Competitive election	−0.003	−0.009	0.012	0.003	−0.008	−0.006	−0.001
Competitive election	(0.004)	(0.007)	(0.009)	(0.007)	(0.011)	(0.006)	(0.005)
Concurrent senate election	0.407*	0.655**	0.156	−0.255	−0.176	0.150	0.660**
Concurrent senate election	(0.179)	(0.240)	(0.318)	(0.431)	(0.609)	(0.266)	(0.227)
Percent urban	0.004	0.006	0.016	−0.006	0.023	0.008	−0.003
Percent urban	(0.006)	(0.008)	(0.011)	(0.013)	(0.017)	(0.009)	(0.007)
Percent black	0.002	−0.002	−0.019	−0.004	−0.009	0.004	0.006
Percent black	(0.008)	(0.013)	(0.013)	(0.020)	(0.026)	(0.013)	(0.010)
Percent Latinx	−0.001	−0.014	−0.012	0.016	0.015	0.002	−0.001
Percent Latinx	(0.006)	(0.011)	(0.015)	(0.011)	(0.019)	(0.010)	(0.008)
Percent AAPI	−0.014	−0.010	−0.005	0.001	0.005	−0.032	0.008
Percent AAPI	(0.014)	(0.033)	(0.049)	(0.022)	(0.026)	(0.021)	(0.020)
Constant	−2.966***	−2.769*	−5.604***	−3.574*	−1.858	−3.320**	−2.923***
Constant	(0.743)	(1.098)	(1.516)	(1.506)	(2.460)	(1.054)	(0.857)
Observations	1,768	703	439	417	209	723	1,045
Pseudo R²	.109	.151	.113	.169	.198	.135	.120
AIC	1.252	1.858	0.736	0.965	0.772	1.513	1.059

AAPI = Asian American/Pacific Islander; AIC = Akaike information criterion.

p < .05. **p < .01. ***p < .001.

Table 3.

Logit Coefficients of 2018 Validated Turnout with Election Context (by Race–Gender).

	White men	White women	Black men	Black women	Latinx men	Latinx women	Asian American men	Asian American women
Racial context	0.356	0.033	0.426	0.741*	−0.425	0.014	0.211	0.302
Racial context	(0.406)	(0.331)	(0.512)	(0.335)	(0.423)	(0.326)	(0.637)	(0.555)
Gender context	−0.100	−0.398	−0.372	−0.039	0.727*	−0.004	0.833	−0.615
Gender context	(0.270)	(0.224)	(0.399)	(0.309)	(0.342)	(0.289)	(0.555)	(0.430)
Party ID	−0.052	−0.130	0.059	−0.257*	−0.420**	−0.061	−0.324	−0.722*
Party ID	(0.084)	(0.077)	(0.188)	(0.115)	(0.142)	(0.113)	(0.201)	(0.291)
Education	0.858***	0.726***	0.496	0.332	1.260***	1.094***	0.090	−0.294
Education	(0.222)	(0.191)	(0.281)	(0.208)	(0.247)	(0.269)	(0.517)	(0.472)
Age	−0.083*	0.048	0.011	0.080	0.080	−0.040	−0.267*	−0.068
Age	(0.040)	(0.035)	(0.061)	(0.043)	(0.052)	(0.048)	(0.123)	(0.086)
Income	0.141**	0.096**	−0.021	0.062	0.071	−0.045	0.264***	0.112
Income	(0.044)	(0.037)	(0.061)	(0.040)	(0.056)	(0.042)	(0.076)	(0.066)
Employed	0.347	−0.133	0.255	−0.103	−0.160	0.006	−1.842*	1.498*
Employed	(0.474)	(0.333)	(0.511)	(0.338)	(0.442)	(0.422)	(0.746)	(0.662)
Married	0.548	0.360	−0.856	−0.799	−0.359	0.261	−0.044	−0.401
Married	(0.376)	(0.312)	(1.113)	(0.522)	(0.562)	(0.414)	(0.990)	(0.756)
South	0.204	−0.502	1.330*	−0.318	−0.360	0.060	−0.086	1.242
South	(0.389)	(0.362)	(0.628)	(0.385)	(0.494)	(0.424)	(0.705)	(1.051)
Competitive election	−0.013	−0.007	0.024	0.001	−0.016	0.011	−0.003	−0.006
Competitive election	(0.010)	(0.008)	(0.018)	(0.011)	(0.012)	(0.009)	(0.016)	(0.016)
Concurrent senate election	0.533	0.684*	−0.901	0.667	−1.616	0.917	−0.980	0.306
Concurrent senate election	(0.343)	(0.322)	(0.568)	(0.350)	(0.855)	(0.564)	(0.853)	(0.870)
Percent urban	0.009	−0.001	0.058**	−0.002	−0.000	−0.012	0.015	0.045*
Percent urban	(0.012)	(0.011)	(0.021)	(0.013)	(0.025)	(0.017)	(0.026)	(0.022)
Percent black	0.005	−0.004	−0.040	−0.007	0.007	0.011	−0.034	0.027
Percent black	(0.021)	(0.018)	(0.026)	(0.014)	(0.040)	(0.021)	(0.034)	(0.044)
Percent Latinx	−0.014	−0.007	−0.031	−0.009	0.038*	0.007	0.002	0.039
Percent Latinx	(0.018)	(0.014)	(0.030)	(0.016)	(0.019)	(0.012)	(0.028)	(0.030)
Percent AAPI	−0.040	0.042	−0.069	0.049	0.019	0.021	0.023	−0.024
Percent AAPI	(0.041)	(0.065)	(0.080)	(0.057)	(0.036)	(0.029)	(0.046)	(0.032)
Constant	−3.198*	−3.925**	−7.185**	−4.030*	−5.245	−2.709	3.424	−2.116
Constant	(1.536)	(1.399)	(2.634)	(1.663)	(2.870)	(1.771)	(3.405)	(2.937)
Observations	287	416	150	289	179	238	107	102
Pseudo R²	.158	.177	.158	.140	.285	.159	.322	.296
AIC	2.345	1.518	0.965	0.652	1.088	0.895	0.776	0.863

AAPI = Asian American/Pacific Islander; AIC = Akaike information criterion.

p < .05. **p < .01. ***p < .001.

Once again, we find strong evidence that race, partisanship, education, and income were strongly associated with voter turnout when pooling respondents (column 1). There is virtually no evidence, however, that the race and gender context of elections had any effect on voter turnout. Based on the results of this model, we would conclude that the historic diversity of candidates in the 2018 midterm elections had no effect on turnout among young adults. Indeed, this conclusion seems to hold when we estimate separate models by race (columns 2 through 5), or gender (columns 6 and 7).

Nevertheless, it may be the case that the effects of racial or gender context may operate in opposite directions for men and women within racial groups and thus wash out. In columns 6 through 9, we therefore disaggregate a step further and estimate separate models at the intersection of both race and gender. Notably, neither race nor gender context is associated with voter turnout for most race–gender groups. Importantly, however, we observe a positive and significant effect of race context for black women, even when accounting for individual- and district-level covariates. The marginal effect of having a candidate of color on the ballot on voter turnout was approximately 18 percentage points among black women. No effects were observed for the gender context of elections among any race–gender group with the exception of Latinx men. Having at least one female candidate on the ballot was associated with a 16 percentage point increase in the predicted probability of voting in 2018, all else equal.

To dive deeper into how the dynamics of racially diverse elections impact women of color, we disaggregate the electoral context variables related to candidate race and gender diversity into separate dummy variables for election match-ups by candidate race and gender.¹¹ As illustrated in Table 4, when grouped together (column 4), women of color were generally more likely to vote in elections where a candidate of color, regardless of gender, was running against a white male candidate but not necessarily against a white female candidate. Once disaggregated by race and ethnicity (columns 1 through 3),¹² however, we see that the aggregate effect is largely driven primarily by black women. We also find that black women, in contrast to the aggregate results, were also more likely to turn out when there was a female candidate of color running against a white female candidate. In races that featured white male candidates, black women were 29 percentage points more likely to vote if a male candidate of color was running and 28 percentage points more likely to vote if a female candidate of color was running against the white male candidate. In races that featured white female candidates, black women were 49 percentage points more likely to vote if a female candidate of color was running against the white female candidate but were no more likely to vote if a male candidate of color was running against the white female candidate. In other words, black women were significantly more likely to vote in elections that featured female candidates of color running against a white candidate regardless of that candidate’s gender.

Table 4.

Effect of Election Match-Ups on 2018 Validated Turnout among WOC.

	Black women	Latinx women	Asian American women	All WOC
WM vs. WW	0.038	−0.047	−0.827	0.047
WM vs. WW	(0.532)	(0.584)	(0.792)	(0.344)
WM vs. MOC	1.193⁺	0.584	0.663	0.807*
WM vs. MOC	(0.613)	(0.615)	(0.864)	(0.375)
WM vs. WOC	1.092*	0.300	1.220	0.653
WM vs. WOC	(0.534)	(0.647)	(1.451)	(0.368)
WW vs. WW	0.568	0.620	−0.221	−0.007
WW vs. WW	(1.136)	(1.305)	(1.125)	(0.744)
WW vs. MOC	0.956	0.000	−0.226	−0.028
WW vs. MOC	(1.232)	(.)	(1.317)	(0.718)
WW vs. WOC	2.006*	0.855	−1.954	0.986
WW vs. WOC	(0.926)	(0.874)	(1.294)	(0.562)
MOC vs. MOC	1.331	−0.157	0.000	0.413
MOC vs. MOC	(0.831)	(0.974)	(.)	(0.565)
MOC vs. WOC	1.424	−0.207	−1.675	0.560
MOC vs. WOC	(0.822)	(0.753)	(1.425)	(0.511)
WOC vs. WOC	−1.967	0.000	0.000	−2.808*
WOC vs. WOC	(1.366)	(.)	(.)	(1.210)
Party ID	−0.226*	−0.048	−0.753*	−0.216**
Party ID	(0.114)	(0.117)	(0.314)	(0.078)
Education	0.342	1.157***	−0.488	0.619***
Education	(0.209)	(0.270)	(0.534)	(0.146)
Age	0.099*	−0.034	−0.038	0.001
Age	(0.043)	(0.049)	(0.092)	(0.030)
Income	0.046	−0.047	0.119	0.018
Income	(0.040)	(0.043)	(0.067)	(0.027)
Employed	−0.039	0.009	1.508*	0.091
Employed	(0.346)	(0.429)	(0.735)	(0.257)
Married	−0.948	0.186	−0.676	−0.554
Married	(0.505)	(0.432)	(0.856)	(0.307)
South	−0.099	0.198	0.956	0.248
South	(0.397)	(0.437)	(1.060)	(0.274)
Competitive election	0.006	0.013	−0.011	0.007
Competitive election	(0.012)	(0.009)	(0.017)	(0.006)
Concurrent senate election	0.812*	0.975	0.237	0.796**
Concurrent senate election	(0.361)	(0.615)	(0.959)	(0.270)
Percent urban	−0.008	−0.018	0.054*	−0.002
Percent urban	(0.013)	(0.018)	(0.025)	(0.010)
Percent black	−0.003	0.017	0.045	0.005
Percent black	(0.016)	(0.023)	(0.042)	(0.011)
Percent Latinx	−0.004	0.007	0.034	−0.001
Percent Latinx	(0.018)	(0.013)	(0.032)	(0.008)
Percent AAPI	0.075	0.034	−0.029	0.004
Percent AAPI	(0.060)	(0.035)	(0.031)	(0.022)
Constant	−4.229**	−2.908	−3.217	−2.890*
Constant	(1.587)	(2.003)	(3.327)	(1.126)
Observations	289	230	102	629
Pseudo R²	.179	.178	.332	.127
AIC	0.676	0.944	0.933	0.752

WOC = Women of Color; WM = White Men; WW = White Women; MOC = Men of Color; AAPI = Asian American/Pacific Islander; AIC = Akaike information criterion.

p = .05. *p < .05. **p < .01. ***p < .001.

Regional Differences in the Drivers of Voter Turnout

Due to the unique racial politics of whites in the south (Archaya, Blackwell, and Sen 2018; Kousser 2010) and several high-profile races that garnered national attention, we also modeled voter turnout among whites in and out of the south. As demonstrated by columns 1 and 2 in Table 5, we find important differences in the effects of gender and race context between Southern and non-Southern whites, respectively. We estimate separate models for Southern and non-Southern whites and include interaction terms within models to examine whether the effect of race and gender context on turnout in 2018 is conditional on gender, which is, to say, different for men and women. Using this analytical approach allows us to effectively compare between two theoretically distinct groups, in this case Southern and non-Southern whites, while also testing for important differences within distinct groups.

Table 5.

Effect of Election Context on 2018 Validated Voter Turnout among Whites.

	South	Non-South
Female	3.820**	−0.078
Female	(1.446)	(0.875)
Racial context	1.779*	−0.066
Racial context	(0.711)	(0.391)
Female racial context	−2.215**	0.223
Female racial context	(0.857)	(0.503)
Gender context	0.217	−0.073
Gender context	(0.558)	(0.300)
Female gender context	−1.324	−0.152
Female gender context	(0.698)	(0.431)
Party ID	−0.161	−0.118
Party ID	(0.107)	(0.069)
Education	0.678*	0.868***
Education	(0.265)	(0.189)
Age	−0.027	−0.017
Age	(0.051)	(0.034)
Income	0.098	0.111**
Income	(0.052)	(0.034)
Employed	−1.144*	0.483
Employed	(0.565)	(0.355)
Married	1.304**	0.196
Married	(0.454)	(0.307)
Competitive election	0.005	−0.021*
Competitive election	(0.013)	(0.009)
Concurrent senate election	2.090***	0.048
Concurrent senate election	(0.487)	(0.308)
Percent urban	−0.016	0.003
Percent urban	(0.016)	(0.010)
Percent black	0.020	0.009
Percent black	(0.024)	(0.020)
Percent Latinx	−0.002	−0.015
Percent Latinx	(0.019)	(0.014)
Percent AAPI	0.273*	−0.017
Percent AAPI	(0.118)	(0.037)
Constant	−5.028*	−2.966*
Constant	(2.467)	(1.289)
Observations	211	492
Pseudo R²	.321	.169
AIC	1.823	1.736

AAPI = Asian American/Pacific Islander; AIC = Akaike information criterion.

p < .05. **p < .01. ***p < .001.

Comparing Southern with non-Southern whites (columns 1 and 2), we find that one’s gender identity and the racial diversity of elections are associated with increased turnout among Southern whites only. Moreover, the effect of race context on turnout among white Southerners was conditional on one’s gender; white men in the south were significantly more likely to cast a ballot in 2018 if the candidates in their congressional district were racially diverse by approximately 49 percentage points. The predicted probability of white Southern women voting in 2018 was not affected by the racial diversity of candidates in their district. Similarly, the gender diversity of congressional candidates had no effect on turnout among Southern or non-Southern whites even when interacted with the respondent’s own gender identity. Taken together, these results provide preliminary evidence for the mobilization of Southern white men in reaction to increased racial candidate diversity but not increased gender diversity. We do not find any evidence of similar reactive mobilization among white women in the south or among whites of any gender outside of the south.

Discussion and Conclusion

This analysis bridges two critical yet often parallel literatures on race and gender, respectively, and in doing so provides a more nuanced understanding of how and under what conditions race, gender, and the intersections of both shape voter turnout among young people, a demographic group with traditionally low levels of voter turnout but that exhibited the biggest increase in voter turnout in 2018. Existing work on race and gender has traditionally focused on each identity category in isolation rather than in tandem while scholarship on the intersections of identity has largely been left to the margins of the discipline. This remains true even in the wake of the 2016 election, which brought renewed attention to the critical importance of intersectional, or at the very least multidimensional, understandings of political attitudes and behavior. This study contributes to this important yet often ignored literature on political behavior and finds that the drivers of voter turnout are not uniform across categories of race, gender, or race–gender, instead reflecting structural and systematic variation both across and within groups.

In examining the effects of race, gender, and diverse electoral contexts on youth voter turnout in 2018, a number of findings from our analysis stand out. First, in line with expectations from previous work, our analysis suggests that race was significant predictor of voter turnout among voters ages eighteen to thirty-four. Whites were substantially more likely to have voted in 2018 than people of color. Second, gender was an important driver of voter turnout, primarily among black voters, with black women significantly more likely to have voted in 2018 than black men. Similarly, when disaggregating by gender, race was associated with increased turnout among men but not women. Third, although we find no evidence that the race or gender contexts of elections had any effect on voter turnout in the aggregate or across groups of race or gender, we do find differential effects of electoral context once we disaggregate our analysis by categories of race–gender, highlighting the importance of an intersectional approach for uncovering a more complete understanding of political behavior. Racially diverse elections led to increased turnout among black women and white men in the south, suggesting that race and gender, along with the politics of race and gender, shape both women and men in consequential ways (Weldon 2006). Despite observable effects among both black women and white men in the south, the increased participation among each respective group is likely driven by distinct, and perhaps even opposing, motivations that warrant further investigation. Fourth, in line with previous work (Dolan 2006), we find limited evidence of any turnout effects of gender-diverse elections. Finally, by disaggregating respondents beyond binary categories of white and “non-white,” our findings underscore existing cautions against pooling respondents of color into a since category and highlight the importance of attending to both between- and within-group distinctiveness among marginalized groups.

The nuances uncovered by this study in explaining voter turnout among racially diverse young adults in 2018 echo and reaffirm the growing chorus of voices advocating for an intersectional approach to investigating political attitudes and behavior, particularly along lines of race and gender but also other dimensions of identity, including age and generation. Identifying what drove young adults, in particular, to the polls in 2018 is a crucial step in understanding the 2018 midterm elections given the high level of youth turnout. Despite being known for relatively low turnout (Wolfinger and Rosenstone 1980), young voters played a central role in 2018 and have the potential to become a decisive voting bloc in future elections. Indeed, young adults below the age of thirty-five years constitute both the largest living generation and the largest share of the voting-eligible population (Fry 2018). Examining the factors that may lead to greater electoral participation in light of traditionally low voter turnout, or put differently, the factors that translate the electoral potential of young adults into electoral impact, is especially important for contemporary politics given this demographic shift.

Although this study contributes to the literature on youth turnout, and our findings align with extant research on the intersectional effects of candidate diversity (e.g., Philpot and Walton 2007; Simien 2015; Stokes-Brown and Dolan 2010), our sample limits us from making inferences or conclusions about voter turnout among the general population in 2018 or beyond. Thus, we encourage scholars to revisit this line of inquiry in future research to test whether the findings presented here apply to understanding voter turnout more generally. This includes both expanding the sample to include adults of all ages as well as considering midterm elections beyond 2018 to address concerns about the potential uniqueness of the 2018 midterms with respect to race, gender, and race–gender considerations. Although gender and race have long been prominent features in national politics,¹³ the election of Donald Trump in the wake of the Obama presidency may have impacted voter turnout in 2018 in distinctive ways. Although we attempt to address this concern by testing the robustness of our findings against Trump approval (see Supplemental Appendix), additional research accounting for multiple elections is needed. Future research should also seek to test why these inter- and intragroup differences are observed.

Although our results are situated in the specific of context of young voter turnout in 2018, they nonetheless provide insight into the connection between candidate characteristics and voter turnout. The act of voting is foundational to the functioning of a representative democracy, which is perhaps why scholars often focus on understanding voter turnout. Nevertheless, to examine voter turnout in the aggregate or by race and gender alone is to examine only part of a larger story. This study represents an important step in understanding the larger story of the 2018 midterm elections.

Supplemental Material

Online_Appendix_PDF – Supplemental material for The Intersectional Effects of Diverse Elections on Validated Turnout in the 2018 Midterm Elections

Supplemental material, Online_Appendix_PDF for The Intersectional Effects of Diverse Elections on Validated Turnout in the 2018 Midterm Elections by Vladimir E. Medenica and Matthew Fowler in Political Research Quarterly

Research Data

medenica_and_fowler_prq_replication_data – Research Data for The Intersectional Effects of Diverse Elections on Validated Turnout in the 2018 Midterm Elections

Research Data, medenica_and_fowler_prq_replication_data for The Intersectional Effects of Diverse Elections on Validated Turnout in the 2018 Midterm Elections by Vladimir E. Medenica and Matthew Fowler in Political Research Quarterly

This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Vladimir E. Medenica

Supplemental Material

Supplemental materials and replication materials for this article are available with the manuscript on the Political Research Quarterly (PRQ) website.

Notes

References

Abrajano

Marisa

Hajnal

Zoltan

. 2017. White Backlash: Immigration, Race, and American Politics. Princeton: Princeton University Press.

Archaya

Avidit

Blackwell

Matthew

Sen

Maya

. 2016. “The Political Legacy of American Slavery.” The Journal of Politics 78 (3): 621–41.

Archaya

Avidit

Blackwell

Matthew

Sen

Maya

. 2018. Deep Roots: How Slavery Still Shapes Southern Politics. Princeton: Princeton University Press.

Barreto

Matt A.

2007. “Sí Se Puede! Latino Candidates and the Mobilization of Latino Voters.” American Political Science Review 101 (3): 425–41.

Barreto

Matt A.

2010. Ethnic Cues: The Role of Shared Ethnicity in Latino Political Participation. Ann Arbor: University of Michigan Press.

Barreto

Matt A.

Frasure-Yokley

Lorrie

Vargas

Edward

Wong

Janelle

. 2018. “Best Practices in Collecting Online Data with Asian, Black, Latino, and White Respondents: Evidence from the 2016 Collaborative Multiracial Post-election Survey.” Politics, Groups & Identities 6 (1): 171–80.

Baxter

Sandra

Lansing

Marjorie

. 1983. Women and Politics: The Invisible Majority. Ann Arbor: University of Michigan Press.

Beltrán

Cristina

. 2010. The Trouble with Unity: Latino Politics and the Creation of Identity. New York: Oxford University Press.

Box-Steffensmeier

Janet M.

De Boef

Suzana

Lin

Tse-Min

. 2004. “The Dynamics of the Partisan Gender Gap.” American Political Science Review 98 (3): 515–28.

10.

Brambor

Thomas

Clark

William Roberts

Golder

Matt

. 2006. “Understanding Interaction Models: Improving Empirical Analyses.” Political Analysis 14 (1): 63–82.

11.

Brown

Nadia E.

2014. “Political Participation of Women of Color: An Intersectional Analysis.” Journal of Women, Politics & Policy 35 (4): 315–48.

12.

Brown

Nadia E.

Gershon

Sarah Allen

. 2016. Distinct Identities: Minority Women in U.S. Politics. New York: Routledge.

13.

Burns

Nancy

Jardina

Ashley

Yadon

Nicole

. 2017. “Women as a Force in Electoral Politics.” In The Oxford Handbook of U.S. Women’s Social Movement Activism, edited by McCammon

Holly J.

Taylor

Verta

Reger

Einwohner

Rachel L.

, 507–21. New York: Oxford University Press.

14.

Burns

Nancy

Scholzman

Kay L.

Verba

Sidney

. 2002. The Private Roots of Public Action Gender, Equality, and Political Participation. Cambridge, MA: Harvard University Press.

15.

Carroll

Susan J.

2006. “Moms Who Swing, or Why the Promise of the Gender Gap Remains Unfulfilled.” Politics & Gender 2 (3): 362–74.

16.

Cassese

Erin C.

Barnes

Tiffany D.

2018. “Reconciling Sexism and Women’s Support for Republican Candidates: A Look at Gender, Class, and Whiteness in the 2012 and 2016 Presidential Races.” Political Behavior 41:677–700.

17.

Center for American Women and Politics. 2017. Gender Differences in Voter Turnout. New Brunswick: CAWP.

18.

Center for American Women and Politics. 2018. Results: Women Candidates in the 2018 Elections. New Brunswick: CAWP.

19.

Chong

Dennis

Rogers

Reuel

. 2005. “Racial Solidarity and Political Participation.” Political Behavior 27 (4): 347–74.

20.

Chow

Esther Ngan-ling

Wilkinson

Doris

Zinn

Maxine Baca

. 1996. Race, Class, and Gender: Common Bonds, Different Voices. Thousand Oaks: SAGE.

21.

Clifford

Scott

. 2012. “Reassessing the Unequal Representation of Latinos and African Americans.” The Journal of Politics 74 (3): 903–16.

22.

Collins

Patricia Hill

. 1990. Black Feminist Thought: Knowledge, Consciousness and the Politics of Empowerment. New York: Routledge.

23.

Crenshaw

Kimberle

. 1988. “Race, Reform and Retrenchment: Transformation and Legitimation in Anti Discrimination Law.” Harvard Law Review 101 (7): 1331–87.

24.

Crenshaw

Kimberle

. 1991. “Mapping the Margins: Intersectionality, Identity Politics, and Violence against Women of Color.” Stanford Law Review 43 (6): 1241–99.

25.

Dabelko

Kristen La Cour

Herrnson

Paul S.

1997. “Women’s and Men’s Campaigns for the U.S. House of Representatives.” Political Research Quarterly 50 (1): 121–35.

26.

Dawson

Michael

. 1994. Behind the Mule: Race and Class in African-American Politics. Princeton: Princeton University Press.

27.

Dawson

Michael

Cohen

Cathy J.

2002. “Problems in the Study of the Politics of Race.” In Political Science: The State of the Discipline, edited by Katznelson

Ira

Milner

Helen V.

, 488–510. New York: W.W. Norton.

28.

DeSilver

Drew

. 2018. A Record Number of Women Will Be Serving in the New Congress. Washington, DC: Pew Research Center.

29.

Dolan

Kathleen

. 2006. “Symbolic Mobilization? The Impact of Candidate Sex in American Elections.” American Politics Research 34 (6): 687–704.

30.

Dolan

Kathleen

. 2008. “Is There a ‘Gender Affinity Effect‘ in American Politics? Information, Affect, and Candidate Sex in U.S. House Elections.” Political Research Quarterly 61 (1): 79–89.

31.

Dowe

Pearl K. F.

2016. “African American Women: Leading Ladies of Liberal Politics.” In Distinct Identities: Minority Women in U.S. Politics, edited by Brown

Nadia E.

Gershon

Sarah Allen

, 49–62. New York: Routledge.

32.

Fraga

Bernard L.

2018. The Turnout Gap: Race, Ethnicity, and Political Inequality in A Diversifying America. New York: Cambridge University Press.

33.

Frasure-Yokley

Lorrie

. 2018. “Choosing the Velvet Glove: Women Voters, Ambivalent Sexism, and Vote Choice in 2016.” Journal of Race, Ethnicity, and Politics 3 (1): 3–25.

34.

Frey

William H.

2018. Diversity Explosion: How New Racial Demographics Are Remaking America. Washington, DC: Brookings Institution Press.

35.

Friedrich

Robert J.

1982. “In Defense of Multiplicative Terms in Multiple Regression Equations.” American Journal of Political Science 26 (4): 797–833.

36.

Fry

Richard

. 2018. Millennials Projected to Overtake Baby Boomers as America’s Largest Generation. Washington, DC: Pew Research Center.

37.

García

Bedolla

Michelson

Lisa Melissa R.

2012. Mobilizing Inclusion: Transforming the Electorate through Get-out-the-vote Campaigns. New Haven: Yale University Press.

38.

Gay

Claudine

. 2001. “The Effect of Black Congressional Representation on Political Participation.” American Political Science Review 95 (3): 589–602.

39.

Geiger

Abigail W.

Bialik

Kristen

Gramlich

John

. 2019. The Changing Face of Congress in 6 Charts. Washington, DC: Pew Research Center.

40.

Gimpel

James G.

Lee

Frances E.

Kaminski

Joshua

. 2006. “The Political Geography of Campaign Contributions.” The Journal of Politics 68 (3): 626–639.

41.

Gutierrez

Angela

Ocampo

Angela X.

Barreto

Matt A.

Segura

Gary

. 2019. “Somos Más: How Racial Threat and Anger Mobilized Latino Voters in the Trump Era.” Political Research Quarterly 72:960–75.

42.

Hajnal

Zoltan

. 2010. America’s Uneven Democracy: Race, Turnout, and Representation in City Politics. New York: Cambridge University Press.

43.

Hajnal

Zoltan

Trounstine

Jessica

. 2016. “Injustice in the City: Race and Class Inequalities in Local Politics.” In The Double Bind: The Politics of Racial And Class Inequalites in the Americas, edited by Juliet Hooker and Alvin B. Tillery, Jr., 139–155. Washington, DC: Report of the APSA Task Force on Racial and Social Class Inequalities in the Americas.

44.

Hancock

Ange-Marie

. 2007. “When Multiplication Doesn’t Equal Quick Addition: Examining Intersectionality as a Research Paradigm.” Perspectives on Politics 5 (1): 63–79.

45.

Hancock

Ange-Marie

. 2016. Intersectionality: An Intellectual History. New York: Oxford University Press.

46.

Hartig

Hannah

. 2019. In Year of Record Midterm Turnout, Women Continued to Vote at Higher Rates than Men. Washington, DC: Pew Research Center.

47.

Herrnson

Paul

Lay

J. Celeste

Stokes

Atiya Kai

. 2003. “Women Running “as Women”: Candidate Gender, Campaign Issues, and Voter-targeting Strategies.” The Journal of Politics 65 (1): 244–55.

48.

Huddy

Leonie

Cassese

Erin

. 2011. “On the Complex and Varied Political Effects of Gender.” In The Oxford Handbook of American Public Opinion and the Media, edited by Edwards III

George C.

Jacobs

Lawrence R.

Shapiro

Robert Y.

, 471–87. New York: Oxford University Press.

49.

Huddy

Leonie

Cassese

Erin

Lizotte

Mary-Kate

. 2008. “Gender, Public Opinion, and Political Reasoning.” In Political Women and American Democracy, edited by Wolbrecht

Christina

Beckwith

Karen

Baldez

Lisa

, 31–49. New York: Cambridge University Press.

50.

Jardina

Ashley

. 2018. White Identity Politics. New York: Cambridge University Press.

51.

Jones-Correa

Michael

. 1998. “Different Paths: Gender, Immigration, and Political Participation.” International Migration Review 32 (2): 326–49.

52.

Junn

Jane

. 2017. “The Trump Majority: White Womanhood and the Making of Female Voters in the U.S.” Politics, Groups and Identities 5 (2): 343–52.

53.

Junn

Jane

Brown

Nadia E.

2008. “What Revolution? Incorporating Intersectionality in Women and Politics.” In Political Women and American Democracy, edited by Wolbrecht

Christina

Beckwith

Karen

Baldez

Lisa

, 64–78. New York: Cambridge University Press.

54.

Key

V. O.

1949. Southern Politics in State and Nation. Knoxville: University of Tennessee Press.

55.

Koch

Jeffrey

. 1997. “Candidate Gender and Women’s Psychological Engagement in Politics.” American Politics Quarterly 25 (1): 118–33.

56.

Kousser

Morgan J.

2010. “The Immutability of Categories and theReshaping of Southern Politics.” Annual Review of Political Science 13 (1): 365-383.

57.

Lavariega Monforti

Jessica

. 2017. “The Latina/o Gender Gap in the 2016 Election.” Aztlán: A Journal of Chicano Studies 42 (2): 229–48.

58.

Leighley

Jan E.

Nagler

Jonathan

. 2013. Who Votes Now? Demographics, Issues, Inequality, and Turnout in the United States. Princeton: Princeton University Press.

59.

Lien

Pei-te

Conway

M. Margaret

Wong

Janelle

. 2004. The Politics of Asian Americans: Diversity and Community. New York: Routledge.

60.

Lizotte

Mary-Kate

Sidman

Andrew H.

2009. “Explaining the Gender Gap in Political Knowledge.” Politics & Gender 5 (2): 127–51.

61.

Major

Brenda

Blodorn

Alison

Blascovich

Gregory J.

2016. “The Threat of Increasing Diversity: Why many White Americans Support Trump in the 2016 Presidential Election.” Group Processes & Intergroup Relations 21 (6): 931–40.

62.

Masuoka

Natalie

. 2006. “Together They Become One: Examining the Predictors of Panethnic Group Consciousness among Asian Americans and Latinos.” Social Science Quarterly 87 (5): 993–1011.

63.

Masuoka

Natalie

Junn

Jane

. 2013. The Politics of Belonging: Race, Public Opinion, and Immigration. Chicago: University of Chicago Press.

64.

McConnaughy

Corrine M.

White

Ismail K.

Leal

David

Casellas

Jason

. 2010. “A Latino on the Ballot: Explaining Co-ethnic Voting among Latinos and White Americans’ Response.” The Journal of Politics 72 (4): 1199–21.

65.

McElroy

Gail

Marsh

Michael

. 2009. “Candidate Gender and Voter Choice: Analysis from a Multimember Preferential Voting System.” Political Research Quarterly 63 (4): 822–33.

66.

McKee

Seth

Hood

M. V.

III Hill

David

. 2012. “Achieving Validation: Barack Obama and Black Turnout in 2008.” State Politics and Policy Quarterly 12 (1): 3–22.

67.

Medenica

Vladimir E.

2018. “Millennials and Race in the 2016 Election.” Journal of Race, Ethnicity, and Politics 3 (1): 55–76.

68.

Misra

Jordan

. 2019. “Behind the 2018 U.S. Midterm Election Turnout: Voter Turnout Rates among All Voting Age and Major Racial and Ethnic Groups Were Higher than in 2014.” U.S. Census Bureau. https://www.census.gov/library/stories/2019/04/behind-2018-united-states-midterm-election-turnout.html.

69.

Mondak

Jeffrey J.

Anderson

Mary R.

2004. “The Knowledge Gap: A Reexamination of Gender-based Differences in Political Knowledge.” The Journal of Politics 66:492–512.

70.

Osborn

Tracy

McClurg

Scott D.

Knoll

Benjamin

. 2010. “Voter Mobilization and the Obama Victory.” American Politics Research 38 (2): 211–32.

71.

Phillips

Christian

. 2018. “Wanting, and Weighting: White Women and Descriptive Representation in the 2016 Presidential Election.” Journal of Race, Ethnicity, and Politics 3 (1): 29–51.

72.

Philpot

Tasha S.

Shaw

Daron R.

McGowen

Ernest B.

2009. “Winning the Race: Black Voter Turnout in the 2008 Presidential Election.” Public Opinion Quarterly 73 (5): 995–1022.

73.

Philpot

Tasha S.

Walton

Hanes

Jr.

2007. “One of Our Own: Black Female Candidates and the Voters Who Support Them.” American Journal of Political Science 51 (1): 49–62.

74.

Ramírez

Ricardo

. 2013. Mobilizing Opportunities: The Evolving Latino Electorate and the Future of American Politics. Charlottesville: University of Virginia Press.

75.

Reny

Tyler T.

Collingwood

Loren

Valenzuela

Ali A.

2019. “Vote Switching in the 2016 Election: How Racial and Immigration Attitudes, Not Economics, Explain Shifts in White Voting.” Public Opinion Quarterly 83 (1): 91–113.

76.

Rosenstone

Steven J.

Hansen

John Mark

. 1993. Mobilization, Participation, and Democracy in America. New York: MacMillan Publishing Company.

77.

Rouse

Stella M.

Ross

Ashley D.

2018. The Politics of Millennials: Political Beliefs and Policy Preferences of America’s Most Diverse Generation. Ann Arbor: University of Michigan Press.

78.

Sanchez

Gabriel R.

Masuoka

Natalie

. 2010. “Brown-utility Heuristic? The Presence and Contributing Factors of Latino Linked Fate.” Hispanic Journal of Behavioral Sciences 32 (4): 519–31.

79.

Sapiro

Virginia

Conover

Pamela Johnston

. 1997. “The Variable Gender Basis of Electoral Politics: Gender and Context in the 1992 US Election.” British Journal of Political Science 27 (1): 497–523.

80.

Schaffner

Brian F.

Macwilliams

Matthew

Nteta

Tatishe

. 2018. “Understanding White Polarization in the 2016 Vote for President: The Sobering Role of Racism and Sexism.” Political Science Quarterly 133 (1): 9–34.

81.

Shapiro

Robert Y.

Mahajan

Harpreet

. 1986. “Gender Differences in Policy Preferences: A Summary of Trends from the 1960s to the 1980s.” Public Opinion Quarterly 50 (1): 42–61.

82.

Sides

John

Tesler

Michael

Vavreck

Lynn

. 2018. Identity Crisis: The 2016 Presidential Campaign and the Battle for the Meaning of America. Princeton: Princeton University Press.

83.

Simien

Evelyn M.

2006. Black Feminist Voices in Politics. Albany: State University of New York Press.

84.

Simien

Evelyn M.

2015. Historic Firsts: Historic Firsts: How Symbolic Empowerment Changes U.S. Politics. New York: Oxford University Press.

85.

Smooth

Wendy

. 2006. “Intersectionality in Electoral Politics: A Mess worth Making.” Politics & Gender 2 (31): 400–14.

86.

Stokes

Atiya Kai

. 2003. “Latino Group Consciousness and Political Participation.” American Politics Research 31 (4): 361–78.

87.

Stokes-Brown

Atiya Kai

Dolan

Kathleen

. 2010. “Race, Gender, and Symbolic Representation: African American Female Candidates as Mobilizing Agents.” Journal of Elections, Public Opinion and Parties 20 (4): 473–94.

88.

Tesler

Michael

. 2016. Post-racial or Most-racial? Race and Politics in the Obama Era. Chicago: University of Chicago Press.

89.

Weldon

S. Laurel

. 2006. “The Structure of Intersectionality: A Comparative Politics of Gender.” Politics & Gender 2 (2): 235–48.

90.

Wolfinger

Raymond E.

Rosenstone

Steven J.

1980. Who Votes?. New Haven: Yale University Press.

91.

Wong

Janelle

Ramakrishnan

S. Karthick

Lee

Taeku

Junn

Jane

. 2011. Asian American Political Participation: Emerging Constituents and Their Political Identities. New York: Russell Sage Foundation.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.39 MB

3.25 MB