Gender and Race as “Both/And” in Partisan Vote Choice

As is revealed in the forthcoming pages, race is a key factor in explaining how and why gender is important in American elections.Footnote ⁵ Instead of beginning with the prior that either gender or race is dominant, the story of women voters is a dynamic tale of race and gender working together to affect the electoral fortunes of candidates for US president. By contrast, and in the immediate wake of the 2016 election, some experts prioritized the impact of gender – and in Clinton’s case, the deficit of being a woman (Reference RestonReston 2016) – by arguing that gender is more polarizing to voters than race. Accounts such as these are incomplete at best, and there can be no more denying that women voters in US presidential elections vary substantially, and that race is a key element of the divergence. Acknowledging this does not mean that gender is unimportant and that only race matters. Instead, gender and race exist in a “both/and” rather than an “either/or” relationship for partisan vote choice in elections. Consider the counterfactual of another Obama running for and winning the American presidency in 2008 and 2012; that it had to be him and not her is a clear example of how race and gender work together rather than separately as “either/or.” Former First Lady Michelle Obama cannot be either a woman or an African American because she is simultaneously “both/and” (Reference Gay and TateGay and Tate 1998).

Simple as this insight may appear to be, the impulse to fold all within gender at the expense of recognizing cross-cutting cleavages with competing categories of marginalization such as race is both strong and undeniable. Being “either/or” and on the sharp end of the stick remains a potent rallying cry to mobilization. Certainly, women have been subjugated, disenfranchised, and at best ignored in US politics, but the form and degree to which this is the case intersects with race and ethnicity. For example, while the 100-year anniversary of the ratification of the Nineteenth Amendment was celebrated, it was only white women who universally gained suffrage in 1920. In contrast, women of color would wait another half-century for federal voting rights legislation to secure their enfranchisement across the nation.

Put more concretely, while Hillary Clinton’s grandmother would have gained access to the ballot box with the ratification of the Nineteenth Amendment, Michelle Obama’s did not. And, of course, Trump’s male ancestors would not have needed constitutional amendment for enfranchisement. The reality of this historical context represents the contours of agency and constraint within the broader structure of US democratic institutions and practices defined and structured by white heteropatriarchy (Reference Strolovitch, Wong and ProctorStrolovitch, Wong, and Proctor 2017). The history of the right to vote for some and its flip side of systematic disenfranchisement for others in the United States create an intersectional dynamic of restriction based on gender and race. Similarly, voter preferences developed within the institutional context of the ideological positions of US political parties reflect the divergence between the two major parties on issues of civil rights and women’s rights.

Thus the context within which voters align with political parties, make decisions about which candidates to support, and are mobilized to turn out are structured by hierarchies of advantage and marginalization based in both gender and race. While political discourse rarely implicates explicit justifications for the enfranchisement of men and whites alone, the legacy of the status of white men in particular as both the “modal voter” and the typical candidate for US president persists. Neither Trump nor Biden had to be referred to with race and gender modifiers in 2020 because both existed within the “default category” of presidential candidates. In contrast, and precisely because Hillary Clinton was an anomaly as the first woman to run on a major party ticket for president, she was often referred to as a female candidate, modified by a category of political marginalization in a similar way that Barack Obama was described as an African American presidential candidate. Conversely, going unnamed by race and gender signifies the privilege of being typical.

While women candidates for president may be anomalous, by contrast, women voters are far from unusual, and it is women, rather than men, who are the modal voters in US presidential elections. Data supporting this fact have been visible in the public record for decades, and the first important correction to conventional wisdom about gender and elections is the recognition that women voters are the largest gendered group in the electorate. A second update to the dynamics of gender and elections is simultaneously obvious and yet ignored. Far from a monolith, the category of women voters is one that represents a heterogeneous group in terms of political interests and partisan support, both of which intersect with race and ethnicity (Reference Huddy, Cassese, Lizotte and WhitakerHuddy, Cassese, and Lizotte 2008). While analysts routinely cite the variation, the structural context and consequences of racial diversity among women voters is nevertheless frequently elided by misleading stereotypes about women voters as Democratic Party supporters.

While the large crowds of white women wearing pussy hats and marching in protests following Trump’s inauguration illustrated their support of Hillary Clinton in 2016, these women are in fact in the minority among white women voters overall. Instead, more white women voted for Trump, in 2016 and again in 2020, a pattern of support for Republican Party candidates set decades ago. At the same time, white women also make up the largest piece of the diverse coalition of Democratic Party supporters in those two elections. Instead of being a contradiction, these facts illustrate the “both/and” nature of gender and race that coexists in the dynamics of US presidential elections. Thus, and from a broader perspective of analyzing American voting behavior, that women are both heterogeneous and the modal voter demonstrates with clarity that gender and race should be front and center in any analysis of US elections.

Obvious as this pattern is, political analysts and scholars of contemporary US politics have identified the fault lines around race and gender, but tend to do so separately in either/or fashion. In contrast, we argue that race and gender should be analyzed together simultaneously with an intersectional approach that considers the variation in agency of choice among voters based on both race and gender. Being white or a person of color and being a man or woman are more than individual-level traits to be “controlled for” in analysis. Instead, these identity markers also signify positionality within the twin and overlapping structures of racial hierarchy and patriarchy as locations with more or less constraint on political behavior and choices at the individual level.

Women Voters Are the Modal Voter

It bears reminding that women have been included in the American electorate for only the past 100 years, and that the right to vote was at first extended universally only to white women (Reference KeyssarKeyssar 2009). Prior to this, for the first century and a half of the history of the United States, women were nearly invisible in politics. The introduction of women voters into the American polity began in earnest after the ratification of the Nineteenth Amendment and resulted in a steady rise in women’s voting participation throughout the twentieth century (Reference Corder and WolbrechtCorder and Wobrecht 2016; Reference Christina and Kevin CorderWolbrecht and Corder 2020). Four decades into their formal enfranchisement, American women voters overtook men in turnout. What is clear from the data in Figure 1 detailing the gender composition of the electorate over the last sixty years is that there have been more women voters than men since the time political scientists began collecting systematic survey data from individuals. Nevertheless, and at that time, the picture in the mind’s eye of the American voter was that of a white man; a representative agent that could be aggregated to draw conclusions about political behavior in the macro polity.

Figure 1 Gender composition of the American electorate, 1964–2020

Source: The data used to create this figure was sourced from the Center for American Women and Politics website: https://cawp.rutgers.edu/facts which presents analysis of the US Census Bureau Current Population Survey P20 reports

This normative position is evident in Campbell, Converse, Miller, and Stokes’s classic work, The American Voter, published in 1960, where they concluded this about women voters in the United States: “Women who tend to ‘leave politics to the menfolks’ even though they are willing to go to the polls register here as having a more impoverished level of political concept formation” (p. 492). Campbell et al. mollified concern that politically unsophisticated women voters would cause instability in the electorate, claiming that wives would follow their husbands.

Among the conclusions to the ten-page section titled “Sex” in The American Voter is the assessment that there is no discernible and gender-related pattern in voting behavior. “In the current era, there is no reason to believe that women as women are differentially attracted to one of the political parties” (emphasis original, p. 493).

The passage of time has not diminished the influence of early scholarship to set assumptions about the relationship between gender and power in the electorate, and therefore the way partisan candidate choice is studied. While the predispositions of these initial observations to frame how to analyze and interpret voting behavior are difficult to dislodge, it is clear that both the conclusions and the analytical practices supporting them are in dire need of correction.

Women Voters Are Diverse

Contrary to this early research from the 1950s and 1960s, patterns of voting behavior from subsequent decades identified a partisan gender gap in candidate and party preference, voting turnout, and policy positions (Reference Sapiro and ConoverSapiro and Conover 1997; Reference Norrander and WhitakerNorrander 2008). Scholarship from the 1980s and onward showed women held different political views from men, and the finding that women favored Democrats became part of a new conventional wisdom on the gender gap (Reference Kaufmann and PetrocikKaufmann and Petrocik 1999; Reference KaufmannKaufmann 2006; Reference Blinder and RolfeBlinder and Rolfe 2018; Reference Deckman and CasseseDeckman and Cassese 2021).

Journalists and political analysts along with scholars continue to toe this line, expecting women to support Democratic Party candidates for US president in particular. While women overall show majority support for Democrats, this aggregated result is a function of the heavily lopsided Democratic preferences of African Americans, Latinas, Asian Americans, and other women of color voters who now make up more than a quarter of women voters in the United States.

Table 1 details the changes in the composition of the women voting population in the United States between 1980 and 2020, based on an analysis of US Census Current Population Survey (CAWP 2022). Several patterns in the data in Table 1 are noteworthy. During the Reagan years, white women made up 86–88% of women voters while women of color (including African American and Latina voters) represented 12–14% of the women electorate.Footnote ⁶ In the elections that followed, the number of minority voters expands rapidly due to the implementation of voting rights legislation and the growth in Latin American and Asian immigration. Women of color voters now make up more than a quarter of the women electorate. Consequently, women voters became much more diverse, and the proportion of women of color voters exceeds a quarter of all women voters, having quadrupled in numbers from 6.1 million in 1980 to 23.5 million in 2020.

The significance of this dynamism in the population of American voters overall and women voters in particular is the effect it has on the composition of supporters of the nominees for US President for each of the two major political parties in the United States. While we detail the distribution by race and gender in the next section, Figure 2 provides a window into the magnitude of importance women of color play in the Democratic Party coalition.

Figure 2 Racial composition of Democratic women voters, 1964–2020

Source: ANES

Over time women of color have grown to make up an increasing share of the Democratic coalition, approaching nearly half of all that party’s women supporters in 2016 and 2020. This persists despite the fact that white women make up 73% of women voters and just over half of Democratic Party supporters. As we will demonstrate in a forthcoming section, the data in Figure 2 presage the results that show patterns of support for Democrats, and indeed the development of the gender gap in partisan candidate choice that is driven by the steadily increasing presence of women of color voters in the electorate who bring with them strong and consistent support of that party’s nominees.

Women Voters Are Increasingly Racially Diverse While Republicans Remain White

The rapid increase in racial diversity among women voters is the result of the reduction of race-based exclusion from voting due to federal civil rights, voting rights, and immigration legislation of the 1960s. The consolidation of African Americans into the Democratic Party took another two decades to complete and occurred at just the time when the electoral consequences of the 1965 Immigration and Nationality Act began to be felt (Reference TateTate 1993; Reference Segura and BowlerSegura and Bowler 2011). Despite the enactment of amendments to the US Constitution guaranteeing that the right to vote not be denied on the basis of sex and race, the inclusion of voters of color – whether in the United States for generations or recently arrived – and their full enfranchisement remains a work in progress. Long-standing disenfranchisement on the basis of US territorial residence continues to keep out millions of potential voters (Reference SparrowSparrow 2006). In addition, contemporary efforts to disqualify voters have multiplied following the US Supreme Court’s decision in Shelby County (2013), particularly in southern states.

Variation by race and ethnicity in candidate choice would not be relevant for analyzing election outcomes unless there was companion variation in partisan candidate choice by race. Figure 3 presents data from the American National Election Study visualized as stacked bar charts of the gender-race composition of voters who cast their ballots for the Democratic versus Republican Party candidate for US president in 1968 and 2020. In the earlier election, non-white voters were mostly African Americans, and women and men voters of color made up 21% of the Democratic Party’s base of support in 1968. As the composition of the population of the nation and the electorate changed, that proportion was more than 40% in 2020. This pattern stands in stark contrast to the data for support of Republican Party candidates among voters of color. For both Republican candidates in those elections – Nixon and Trump – their voter base is heavily and disproportionately composed of white voters, whereas non-white voters tally 2% in 1968 and 11% in 2020.

Figure 3 Gender-race composition of Democratic and Republican voters, 1968 and 2020

Source: ANES

Data on voter demographics are usually not presented in this way, and instead it is more common for analysts to report statistics such as the following: 55–58% of white voters supported Trump in 2020. Useful as these data are, they do not translate to saying that Republican supporters during that election were composed of that proportion of white voters. Instead, the correct answer is that 89% of Trump’s voters are white. Similarly, aggregating all Republican Party voters together reveals that the biggest group inside of his coalition of supporters is white women (45%) followed closely by white men (42%). The same pattern is evident for Nixon in 1968, but by a greater margin where white women represented 55% of his supporters compared to 43% for white men. In contrast, the Democratic Party’s coalition is a diverse group, with white women leading the way in 1968 for Hubert Humphrey (46%), and voters of color for Joe Biden (43%) in 2020.

The full complement of American National Election Study (ANES) data between 1948 and 2020 puts a finer point on the conclusion that the gender-race composition of Democratic Party supporters has mirrored the diversity of the voting population over time, while the Republican Party has remained heavily composed of white voters. Figure 4 displays the composition of voters who cast their ballots for Democratic Party nominees. As the electorate itself has changed from more than 90% white in the pre–Voting Rights Act era, to around three-quarters white in 2020, Democratic candidates count 43% of their support from voters of color. Nevertheless, the single largest gender-race group of Democratic supporters in 2016 was white women, who made up 32% of Clinton’s coalition and 29% of Biden’s voters in 2020. While a smaller proportion, white men made up the second largest race-gendered group at 27% for Biden, with women of color following closely behind as the third largest group of Democratic Party supporters.

Figure 4 Gender-race composition of Democratic Party voters, 1948–2020

Source: ANES

In contrast, the composition of Republican Party supporters shows a starkly different profile, and the data are displayed in Figure 5. Republican voters have been composed 90% and more of white Americans. Despite the substantial diversification of the US population overall to more than one-third non-white (and roughly a quarter of voters), Republican Party supporters remain heavily dominated by white Americans. A final observation is that the largest group of voters in the Republican coalition in all but one presidential election across the time series (1996) is white women. While white women are more Democratic when compared to white men, they nevertheless still compose the largest portion of the Republican electorate.

Figure 5 Gender-race composition of Republican Party voters, 1948–2020

Source: ANES

These data showing the gender, racial, and ethnic composition of voters who supported Democratic Party candidates versus Republican Party nominees for US president over the last half century and more reveal a stark difference. The coalition of American voters who back Democrats is racially diverse, while those who support Republicans are relatively stable and white. This distinction is a fitting divergence for Democratic Party nominees who have run as the party of change versus Republican Party candidates that conversely have prioritized conserving tradition. Taken together with the data presented at the beginning of this section on the modal voter status of women voters and with their increasing racial diversity, this analysis of compositional change in the US electorate highlights the political power of women voters in choosing the president of the United States.

Taking the lead from these empirical realities, conclusions about women voters should no longer be made in comparison to the default category of men and their party and candidate preferences. Instead, the variation within women voters can be better observed and the patterns that lie within can be marshalled to better explain election dynamics and outcomes. Continuing to ignore the variation in partisan candidate choice among women voters has the consequence of inaccurately stereotyping women as Democratic Party supporters, because while there are plenty of white women voters who support Democrats, there are more who support Republican nominees.

The Story of Women Voters: Roadmap

In the sections to come, a set of important corrections for how we should understand the relationship between gender, race, and electoral outcomes at the individual level are revealed. Grounding this analysis in Section 2 is a conceptual articulation of the context of this heterogeneous behavior that is based in the unequal privileges and constraints associated with race and gender. This intersectional approach begins with the theoretical proposition that Americans are constrained in unequal and systematic ways by their position in the combined structures of patriarchy and the US racial hierarchy (Reference Patricia HillCollins 2008; Reference HancockHancock 2015; Reference CrenshawCrenshaw 2017). Both gender and race at the individual level matter and must therefore be analyzed together, following the logic of the salience of social groups and structural inequality (Reference YoungYoung 1990).

Section 3 examines the partisan candidate preferences of women of color voters, demonstrating their stalwart support of Democratic Party nominees for US president. Data from the Collaborative Multiracial Post-Election Survey (CMPS) detail the variation between the larger groups of women of color who support Democrats versus the much smaller proportion of African American, Latina, and Asian American women who vote for Republican Party candidates for president. We uncover variation across racial and ethnic minority groups, with Black women voters the strongest and most consistent supporters of Democrats, while recent data collections from 2016 and 2020 evinced strong support among Latinas and Asian American women voters for Hillary Clinton and Joseph Biden.

Switching the focus to white women, Section 4 delves deeper into the diverse partisan politics of white women voters. Using ANES data for analysis, we show that there is substantial partisan diversity among white women and clear variation in the factors that lead white women voters to support Republican Party nominees for president. The analysis in this section reveals the partisan candidate choices of white women since 1988 and highlights their Republican-leaning preferences. The analysis illuminates when factors such as regional location, religion, education, income, and marital status have been significant predictors of white women’s preference for Republicans. Wide diversity of partisanship preference, social context, and political attitudes among white women, combined with broader agency to choose sides, results in this group of voters spanning the distance between feminists, career moms, suburban warriors, and handmaidens to patriarchy.

The fifth section reviews the evidence on the gender gap in partisan vote choice in presidential elections based on the conventional wisdom that women voters are Democratic voters. We show that the gender gap in voting in the United States is the result the steady increase of women of color into the electorate who bring their strong and consistent support of Democratic candidates into the voting booth, thereby canceling out majority Republican support among white women. The results provide a comprehensive view of the gender gap within racial groups, as well as a demonstration of the dynamics of the gender gap as a consequence of compositional change in the US electorate over time.

In the concluding section, we reiterate that both race and gender matter in American elections and argue that ignoring one or the other produces faulty predictions and incomplete conclusions. At the same time, while accounting for gender and race together is more complicated than the standard and more conventional analyses political commentators have become accustomed to, an intersectional approach accounting for multiple marginalization, structural and interpersonal, is “a mess worth making” (Reference SmoothSmooth 2006). Dynamism in the American electorate both in partisanship and in compositional change that is the result of immigration requires analysts to update their methodological strategies for analysis to better understand how and why US presidential elections are won and lost, and the vital role American women voters play in these contests.

The Right to Vote: The Contemporary Legacies of Systemic Disenfranchisement

The legacy of disenfranchisement and the positions of the major political parties on either enhancing or limiting voting rights must be acknowledged and accounted for when analyzing candidate choice among voters in the contemporary context. Significant barriers to inclusion have been based in race, gender, youthful age, class, and citizenship status. The path to voting for women and people of color was both long and arduous in comparison to the exception of the speedy and relatively uncontroversial ratification of the Twenty-Sixth Amendment in 1971, which enfranchised a population of previously excluded voters (ages 18–20) who are not defined in marginalization by either race or gender.

At the founding of the United States it was unnecessary to make arguments defending the extension of the right to vote only to propertied white men, not least of which because they were the ones with standing under the law. Class, in the form of property ownership posed an early hurdle but was eliminated for white males by the mid-nineteenth century. Chattel slavery and the dehumanization of Blacks along with the exclusion of the native inhabitants of the territory were written into and protected by the US Constitution. Protections for white voters continued with passage of the Naturalization Act of 1790 which limited immigrant naturalization to only whites, and being a US citizen remains a necessary condition for casting a ballot in federal elections (Reference HaydukHayduk 2006).

Women won the right to vote first among the systematically excluded but not without fierce opposition. Justifications and rhetoric against woman suffrage were based on the inferior quality of women’s minds as well as their natural unsuitability to the violent and unseemly world of politics (Reference KeyssarKeyssar 2009; Reference GoodieGoodie 2012). The argument that women were paragons of moral virtue was simultaneously used by suffragists, doing double duty in claims that the presence of women voters would help to purify politics while juxtaposing the virtues of women against the presumed scourge of Black men. It is precisely the iconic suffragists – Susan B. Anthony and Elizabeth Cady Stanton – who not only privileged the enfranchisement of white women over African Americans, but also actively weaponized their white womanhood to demonize Blacks as unfit for inclusion (Reference Barkley Brown, Gordon, Collier-Thomas, Bracey, Avakian and BerkmanBarkley Brown 1997; Reference Terborg-PennTerborg-Penn 1998). While coalitions between abolitionists and suffragists were present, the right to vote for white women was in the end won in closer alliance with white supremacists, a broad and deep constituency at the time (Reference DuddenDudden 2011). Among the most potent legacies of the Nineteenth Amendment is the reification of white women in superiority to Black people and other people of color, rendering the latter unworthy of inclusion as voters in the American polity. The right to vote for women secured by the Nineteenth Amendment was thus not universal, but instead a privilege granted to white women.

African Americans and people of color would wait another half-century for the right to vote across the United States. The practice of Jim Crow and the active participation of white women in supporting racial segregation both during slavery and through the Civil Rights era and into the present help to explain why Blacks, Latinos, and Asian Americans were denied the franchise despite the Fifteenth Amendment, which was ratified a century prior to the 1965 Voting Rights Act (Reference LewisLewis 2006; Reference McRaeMcRae 2018; Reference GilmoreGilmore 2019). Political conservatism and anti-Black racial animus among supporters of the modern Republican Party – no longer the party of Abraham Lincoln – fuel renewed efforts at voter disqualification, and the bitter aftertaste of battles over inclusion in the American polity persists (Reference Schlozman and RosenfeldSchlozman and Rosenfeld 2024).

This toxic amalgamation thwarting egalitarian voting practices in the United States is not simply an accident of history. Rather, the combination of domination on the basis of both gender and race is a case in point of the pairing of the twin impulses of patriarchy and racial supremacy, two ideologies deeply embedded in the history of the United States (see, e.g., Reference Parker and BarretoParker and Barreto 2013; Reference SchmidtSchmidt 2021). Together the elitist impulse for domination by race and gender intersect to create a rapid current of racial hierarchy and patriarchy in the United States that became institutionalized, and as a result is difficult to swim against. In this structural system, men and whites are at the top, Blacks and Native Americans at the bottom for most of the history of the nation, and other groups in between beneath whites (Reference KimKim 1999; Reference Masuoka and JunnMasuoka and Junn 2013). Women are not treated equally to men, with women of color distinguished from white women in placement below both men and whites.

How political parties developed over time and into the modern era to either nurture or oppose voting rights for all Americans, and the extent to which ordinary voters would react to the divergence, are the next part of the story of the context of unequal agency and constrained choice by gender and race in the American electorate.

The Parties Diverge: Civil Rights, Women’s Rights and the Republican Embrace of Hierarchy

Political parties in the United States have both been altered by and been important actors in the transformation of politics around voting enfranchisement (Reference Marisa and HajnalAbrajano and Hajnal 2015). While the degree of variation between the Republican Party and the Democratic Party in terms of formal platform policies and ideological positions has waxed and waned over time, the two major US political parties since the U.S. Civil Rights era now present distinctive choices for American voters. Eric Schickler’s analysis of racial realignment following the Great Depression through 1965 details the stark divide voters and party leaders pressed forward, molding the modern Democratic Party to become relatively more progressive on both civil rights and gender equity, in contrast to the Republican Party’s embrace of “racial conservatism” and adherence to traditional gender roles rooted in “social conservatism” (Reference Carmines and StimsonCarmines and Stimson 1989; Reference SchicklerSchickler 2016).

“Racial conservatism” is a euphemism sometimes used by political scientists to describe Republican Party opposition to voting rights in addition to their early support of racial segregation and accompanying opposition to civil rights. Partisan realignment in the south among white voters took shape in the mid-twentieth century starting with their abandonment of the Democratic Party to support racial segregationist candidates for president, including Strom Thurmond (who ran for the Dixiecrat Party in 1948), Harry F. Byrd (unpledged electors in 1960), and George Wallace (who ran for the American Independent Party in 1968). These contests saw white voters in southern states award their Electoral College votes to candidates who supported the continuation of racial segregation and the disenfranchisement of African American citizens. Eventually, and with Nixon’s “southern strategy” in action, white southerners coalesced behind the modern Republican Party, and it was only fifteen years after the passage of the VRA that Ronald Reagan kicked off his 1980 presidential campaign with his “states’ rights” speech at the Neshoba County Fairgrounds near Philadelphia, Mississippi.

Republican Party politicians in contemporary politics continue to diminish voting rights, with among the most consequential moves the federal lawsuit brought by Shelby County, Alabama, challenging the constitutionality of crucial provisions of the Voting Rights Act (Shelby County v. Holder 2013). Since then, states and localities across the nation, and not just in the American South, passed new legislation regulating voting, making it more difficult to register, to qualify to vote, to vote by mail, and to submit one’s ballot (Reference Grose and Brinson BellGrose and Bell 2023). Justifying the policies with allegations of voter fraud, Republican state legislatures and politicians alike took a page from past rhetoric on eliminating what they deemed as unqualified and undesirable voters from participating. Efforts at legislative gerrymandering continue this pattern of behavior.

Regarding gender hierarchy, the embrace of “social conservatism” within the Republican Party is most often used to describe opposition to policies on equal pay for women, abortion rights, and LGBT rights, among other issues that challenge traditional gender roles. The record provides ample evidence of Republican Party positions at odds with the policies aimed at attenuating structural inequalities based in race and gender (Reference MilbankMilbank 2022; Reference MuribMurib 2023). Opposition to reproductive rights for women, affirmative action for underrepresented minorities, support of restrictive immigration policies, and the embrace of punitive crime policy are among the clearest contemporary issues on which Republican and Democratic Party candidates for office differ (Reference LowndesLowndes 2008). The Dobbs (2022) decision by the US Supreme Court overturning the precedent set in Roe (1973) that was made possible by the appointment of conservative jurists by a Republican president is the latest example of the divide separating the parties and their candidates on gender equity.

On the flip side and following the southern racial realignment, the modern Democratic Party has supported civil rights and voting rights legislation and policy, and also been at the more supportive of equal rights and gender equity for American women and people of all genders. The imperative to confront and reduce structural inequalities based in race and gender meant that a broader coalition of voters across race, including progressive whites as well as voters of color, could be forged, and for this reason, the composition of supporters of Democratic Party candidates for US president has become racially diverse. As detailed in the previous section, this diversity among Democratic voters exists in stark contrast to the composition of supporters of Republican Party presidential candidates, the latter of whom remain almost 90% white. More than any other indicator, the patterns of support among American voters by race reflects the divergence in these policy positions by the two major political parties.

Yet the fact that white women are stronger supporters of Republican Party candidates than Democratic Party nominees, even when considering the relatively anti-egalitarian policies of the former, demonstrates just how important position in and protection of the racial hierarchy remains in the contemporary context for some white women voters. Republican and conservative white women respond to patriarchy’s constraint by embracing it in a trade for racial superiority, and the data and a growing literature on conservative women document the tradeoffs contemporary Republican white women voters make between race and gender (Reference SchreiberSchreiber 2008; Reference JunnJunn 2017; Reference Malinga and ShamesOch and Shames 2018).

These are the “deep roots” of the intersectional politics of anti-Blackness and white race supremacy that are intimately intertwined with the subjugation of African Americans and Black women in particular, with the active participation of white women (Reference BleeBlee 2002). The construction of “white womanhood” is forged not only by explicitly racist propaganda but has long been nurtured by traditional gender roles and the “social conservatism” represented in contemporary Republican Party platform policies. The role of white women in particular as bearers of children is an important part of the ideology of the protection racket, which requires an exchange for being put on a pedestal and for taking up residence in a gilded cage (Reference de Beauvoirde Beauvoir 2011; Reference Sjoberg and PeetSjoberg and Peet 2011; Reference ManneManne 2017). That some American women choose the velvet glove despite the knowledge that iron fist of patriarchy lies underneath makes sense within the logic of the protection racket (Reference JackmanJackman 1996; Reference Frasure-YokleyFrasure-Yokley 2018).

In contrast, and while the effects of patriarchal structures, institutions, and values influence people of all races and genders, women of color do not have the same agency to choose occupancy in the gilded cage. They are not welcome precisely because of their intersectional location based in gender and race, reflecting the words of Sojourner Truth: “Nobody ever helps me into carriages, or over mud-puddles, or gives me any best place! And ain’t I a woman?”

As we will show in the sections to come, the support of Republican Party candidates for president that is dominated by white voters is distinct from patterns of support analyzed separately by race and gender. White women are more evenly split between the two parties, though there is a consistent majority of white women voters supporting Republican Party candidates over time. In contrast, women of color heavily and consistently back Democrats. This pattern in the data – in the face of stronger support of gender egalitarian policies by the Democratic Party – indicates that race and gender are not easily separable. Instead, and as we argue next, intersectional location in the overlapping hierarchies of race and gender acts together to constrain the agency of choice.

An Alternative Explanation of Candidate Choice: Intersectionality and Agency of Choice

Conventional approaches to analyzing voting behavior have for the most part assumed that individuals have equal agency when choosing between candidates running for office. In contrast, we argue that intersectional positionality within the overlapping structures of patriarchy and racial hierarchy constrains unequally, and therefore, analyzing partisan candidate vote choice in US presidential elections requires a distinctive intersectional approach to analysis.

When analyzing large-N survey data of American voters, an empirical intersectional approach is distinct from traditional strategies of including all respondents and specifying “dummy variables” for categories of interest and interaction terms where systematic variation is hypothesized to exist. Instead, an intersectional analysis of empirical data on partisan vote choice should begin by disaggregating the population of voters into separate categories of race and gender, after which statistical models can be estimated with the data. This analytical strategy allows for the observation of distinct patterns within the context of the overlapping structures of racial hierarchy and patriarchy that unequally constrain American voters.

Among women voters, white women make the choice between anti-egalitarian positions on gender and race but can choose superiority in the latter. Alternatively, and despite the fact that they have the same set of candidates on their ballots for US president, women of color have more limited agency in deciding who to support because choosing the party that embraces “racial conservatism” and the traditional gender roles of “social conservatism” means agreeing to second place or lower in both hierarchies defining gender and race. Agency is a loaded term precisely because it covers so much ground, and to be clear, we are not taking the position that women of color have less capacity to act in politics. Black women are among the most active participants in the US political system, taking part in elections and campaigns in myriad ways, working in their communities, contacting officials, and voting for candidates (Reference SimienSimien 2006; Reference Farris and HolmanFarris and Holman 2014). Rather, agency in the context of choosing between the candidates for US president is constrained for women of color – and African American women in particular – because one of the alternatives presents an existential threat to many of these voters (Reference BrownBrown 2014).

The heavily white voter composition of supporters of Republican Party candidates and its consistency over time provides the first clue that the two parties provide a rational choice based in ideology for white voters only. However, issues implicating “racial conservatism” and “social conservatism” are only part of the policy platform of Republican Party candidates for president, and white voters can choose between the two major parties if their affinities on fiscal and economic policy or foreign policy are more consistent with Republican candidates even if they disagree on inegalitarian policies based in gender and race. They can because they are less immediately affected in a negative way by Republican Party policies of racial and social conservatism. White women voters – still majority supporters of Republican candidates for president – are nevertheless less supportive than their white male counterparts because socially conservative policies on gender equity affect them more negatively than for men. Among white voters, therefore, gender is more constraining for women than it is for men, and the recent dismantling of the precedent on reproductive rights is one example.

In contrast, while minority voters are still free to choose Republican Party candidates over Democrats, they do so with far lower frequency because of the Republican Party’s positions on policies supporting the maintenance of unequal positionality on the basis of race. Williams theorized this “constraint of race” for African Americans more broadly in politics as a legacy of white privilege (Reference WilliamsWilliams 2003). Among minority voters, women of color are the most constrained, due to both their race and their gendered status as women. Overall, minority voters could exercise their choice to vote and to support, for example, Republican Donald Trump in 2016 and 2020, but relatively few did so. This underscores the fact that voters of color evaluate the policies and candidates of the Republican Party as providing a less hospitable political location for African Americans, Latinos, and Asian Americans (Reference TanenhausTanenhaus 2013). At the same time, we see variation across racial minority groups in their partisan vote choice. Minority voters exercise agency related to their racial position on the hierarchy, with African American voters in particular “captured” by the Democratic Party (Reference TateTate 1993; Reference DawsonDawson 1994; Reference FrymerFrymer 1999; Reference PhilpotPhilpot 2017; Reference White and LairdWhite and Laird 2020).

Thus, it is incorrect to identify either race or only gender alone as influencing variation in partisan preferences and voting behavior among Americans. To the extent that one political party signals a philosophy and policies supporting white racial supremacy and patriarchal dominance of men over women, voters can choose to align or not align with this party. The act of voting for candidates of a party and indeed the decision to turn out to vote are rational responses to signals political parties send to voters. At the same time, and while white women have greater agency than women of color to sort based on their positions on other issues – choosing liberal progressive candidates of the Democratic Party or conservative candidates nominated by the Republican Party – they nevertheless remain constrained by gender and traditional roles assigned to women.

Discussion

Taking account of the legacy of systemic disenfranchisement on the basis of gender and race, and the role of political parties, provides a wider window through which to theorize the conditions under which voters make decisions about which candidate to support in US presidential elections. We advanced the argument that voters do not have equal agency in choice, and that analysts cannot look only to race or only to gender differences to discern the dynamics of candidate choice and voting in US presidential elections.

The next two sections enact this empirical intersectional analytical strategy, beginning with a focus on women of color voters and analyzing separately Black, Latina, and Asian American women in their candidate preferences for US president. We move next to consider the contours and antecedents to the diverse partisan politics of white women voters. This approach to the analysis illustrates the utility of an intersectional analytical approach to revealing how and why diverse women voters are the most influential voters in US presidential elections.

Both/And: Organizing for Progressive Change But Constrained to Support Democrats

The distinctive politics of women of color and their political activism has received relatively scant attention despite the fact that Black, Latina, and Asian American women have been central to organized efforts seeking to challenge racial and gendered oppression. Anti-slavery movements and securing the right to vote were among the first issues of political action among African American women, beginning with Sojourner Truth and progressing through the US Civil Rights movement with icons such as Rosa Parks and Fannie Lou Hamer (Reference BlainBlain 2021; Reference LarsonLarson 2021). Black women’s activism has thus been a constant in US politics and is well documented in the historical record (Reference Bay, Griffin, Jones and SavageBay et al. 2015; Reference GilmoreGilmore 2019). Their activity has continued, unabated, despite substantial barriers, to rise above those hurdles. Some Black women leaders target institutional change by entering elected office such as Shirley Chisholm, who was one of the first women to run for US president in 1972, while others focus their efforts outside of electoral politics by leading political movements for social justice, including Black Lives Matter co-founders Patrice Cullors, Alicia Garza, and Opal Tometi.

Similarly, Latina and Asian American women have advocated for progressive social change on a diversity of issues including labor organizing, environmental justice, immigration, prison reform, and voting rights. These women have been on the forefront of progressive politics in the United States for decades, though their names are less well known and many of their stories are yet to be told. A more focused examination of these activists demonstrates a commitment to social change but at the same time diversity across the content and character of their activism. Among the icons of the American labor movement is Dolores Huerta, a major figure who co-founded the National Farmworkers Association with Cesar Chavez. Citizen activist groups such as the Mothers of East Los Angeles were organizing for environmental justice in the 1980s, holding state government entities and private corporations accountable for polluting local communities (Reference PardoPardo 1998; Reference GilmoreGilmore 2007; Reference JaramilloJaramillo 2010; Reference Milkman and TerriquezMilkman and Terriquez 2012; Reference Montoya and Galvez SeminarioMontoya and Seminario 2020).

Asian American women’s activism ranges from an embrace of radical politics to a more incrementalist approach. This is exemplified by their efforts within social justice movements. Civil rights activist Yuri Kochiyama embraced the radical politics of the Black Panther Party, while Pasty Takemoto Mink, the first Asian American woman elected to the US House of Representatives, sponsored and helped pass Title IX into law (Reference FujinoFujino 2002; Reference JudyWu 2022). Activists such as Helen Zia organized pan-ethnic movements by bringing together different Asian national origin groups to see a shared racialized experience (Reference ZiaZia 2001). Researchers Michi Nishiura Weglyn and Aiko Herzig-Yoshinaga brought to bear evidence used for redress for Japanese American interment (Reference WeglynWeglyn 1966). Grace Lee Boggs and, more recently, Ai-Jen Poo are on the front lines for workers’ and tenants’ rights (Reference ChoyChoy 2005; Reference BaoBao 2006; Reference Fujiwara and RoshanravanFujiwara and Roshanravan 2018), while other Asian American women such as Cecelia Chung advocate for LGBTQ rights.

Given that women of color embrace progressive political agendas, their support for Democratic candidates in the voting booth since the last partisan realignment makes intuitive sense. But their support of Democratic candidates should also be understood in terms of the constraints they experience in the US two-party system rather than simply characterizing them as Democratic Party loyalists. The overwhelming support among women of color for Democrats belies their diverse political opinions and visions for change in challenges to the status quo. Their political goals often did not, or could not, align with mainstream political parties, and their alignment as a core constituency within the Democratic Party was neither easy nor without controversy.

Support in kind from political parties has often not matched the disproportionate votes from women of color. For example, the Democratic Party chose to seat an all-white delegation from Mississippi at its 1964 national convention. A challenge was led by Fannie Lou Hamer, who described in vivid detail the murderous violence against Blacks attempting to register to vote. Hamer’s words describing the arrests, incarceration, and violent beating she endured are preserved and ends with the statement:

Hamer and fellow delegates from the Mississippi Freedom Democratic Party implored national party leaders to seat them at the convention, but they were denied, and media coverage of her powerful speech was interrupted when President Lyndon Johnson gave a press conference at exactly the same time (Reference LarsonLarson 2021).

Thus, disproportionate votes for Democratic over Republican candidates among women of color do not necessarily imply that they feel strong positive attachment or loyalty to the Democratic Party, or that the party is responsive or representative of their political interests. Instead, the divergence between the two major American political parties today provides little choice for women of color voters because candidates representing the Republican Party today have aligned with positions inconsistent with enhancing egalitarian outcomes in politics that benefit women of color. These voters face constraint at the ballot box because the only plausible candidates to represent them are Democrats, even though these politicians may not fully represent the political interests of women of color.

Women of Color Democratic Voters Reflect the Diverse Women of Color Electorate

Our argument builds on the observation that “captured voters” have more limited options when one of the parties in a two-party system embraces anti-egalitarian policies on race. Under these circumstances, there is only one viable option for African American voters to support, other than to abstain from voting (Reference FrymerFrymer 1999). The situation for women of color voters is similar, if not more exaggerated, given the contemporary Republican Party’s positions on issues of gender equity. Thus, the solidly Democratic Party preferences of women of color in part reflect the lack of options they have in the voting booth, and as a result, Latina, Black, and Asian American women voters experience less agency of choice in voting relative to white women voters.

Voting studies have long described partisan vote choice as the result of individuals’ assessment of which party best represents their interests (Hajnal and Lee 2013). In the context of polarization, the two major US political parties have staked out clear and opposing issue positions that allow for sorting of groups into each of the parties (Reference LevenduskyLevendusky 2009). But this sorting occurs primarily among white voters because political parties have been the most responsive to their political interests and policy positions. For women voters, the expectation that sorting will occur along demographic characteristics – urban, educated, and unmarried for Democrats and evangelical, working class, and married for Republicans – applies to explanations of white women voters, and less so for women of color. For example, white evangelicals are among the strongest loyalists in the Republican Party (Reference GeoffryLayman 2001). But these same indicators do not have the same sorting effect for women of color. Instead, scholars have shown that religious conservativism does not translate into strong Republican preference for Asian Americans or Latinos (Reference WhiteWong 2018). Similarly, while narratives emphasizing crime motivate white suburban women to embrace Republican Party candidates (Reference MendelbergMendelberg 2001; Reference RiismandelRiismandel 2020), these messages do not sort Black and Latino voters to Republican support (Reference WhiteWhite 2007; Reference RamirezRamirez 2015). Instead, women of color vote heavily Democratic despite their diversity in terms of those individual-level characteristics deemed consequential for partisan choice.

One useful metric of agency in partisan voting is the degree to which women voters can sort based on their individual level traits. Among women of color, we should not expect there to be as many individual-level factors that predict support of Democrats versus Republicans because of their intersectional position and status as captured voters in a two-party system. In contrast, white women have greater agency in choice because they can vote for the partisan candidate that best represents their identities or interests, and a number of individual-level factors systematically predict vote choice for Republicans versus Democrats. That said, the constraint of intersectionality is not uniform among women of color. Building on our previous work (see Reference Masuoka and JunnMasuoka and Junn 2013) we argue that race constrains life chances for Asian American and Latina women differently from that of Black women. In terms of the US racial order, Blacks are at the bottom of the hierarchy, with Latinos the next lowest in rank, followed by Asian Americans. This order is important to recognizing the structural location from which women of color exercise agency in choosing candidates for US president.

Among women of color voters, Black women have the least agency in partisan candidate choice, given that these voters experience the heaviest social and economic constraints attributed to their race and gender. This translates into near-unanimous support for Democratic Party candidates (Reference TateTate 1993; Reference DawsonDawson 1994; Reference SimienSimien 2006). Because nearly all Black women vote for Democrats, we hypothesize that the empirical analysis will reveal that there are few, if any, individual-level predictors of their partisan vote choice. Among other women of color, we anticipate that proportionally more Asian American women voters believe that they hold agency in partisan choice. Although the historical and present racialized treatment of Asian Americans leads them to demonstrate greater support for Democrats over Republicans, Asian Americans do not always see themselves as a marginalized racial minority group. Studies on Asian American racial attitudes show that a good share of Asian Americans do not see their race to impact their life chances to the same degree as voters from other racial minority groups (Leung, Kim and Masuoka, forthcoming). Furthermore, Asian American women’s greater integration with whites as evidenced by their high intermarriage with white men (Reference Qian and LichterQian and Lichter 2007), residential patterns in white neighborhoods (Reference Iceland and WilkesIceland and Wilkes 2006), and employment in professional fields (Reference DhingraDhingra 2007) also leads to varied political preferences within the group. Given this, we expect to find that while Asian American women vote in the majority for Democrats, there will be more individual-level variation explaining the partisan vote choice among Asian American women compared to Black women voters.

We expect the pattern for Latina voters to fall between that of Black and of Asian American women voters. Like African American women, many Latinas today experience severe marginalization as a group, and so politically are constrained to support Democratic candidates. Latinas are a more heavily immigrant group and conduct the invisible labor needed to support a growing economy which create conditions of vulnerability to abuse and unequal treatment (Reference Hardy-FantaHardy Fanta 1993; Reference Hondagneu-SoteloHondagneu-Sotelo 1994). These shared conditions lead Latinas to support Democratic candidates, in addition to anti-immigration Republican Party positions that have been increasingly racialized as anti-Latino (Reference ChavezChavez 2013; Reference RamírezRamirez 2013; Reference Marisa and HajnalAbrajano and Hajnal 2015). At the same time, Latinas are a more politically heterogeneous group compared to Black women because Latinas as a group are diverse in terms of national origin and immigration status (Reference BeltránBeltran 2010). US relations with Latin American countries also means that certain Latino national origin groups have been treated differently, which leads to varied experiences of marginalization by national origin group (Reference Alejandro and BachPortes and Bach 1985). For example, research shows that some national origin groups such as Cubans are strong Republican supporters while others support the Democratic Party (Reference Fraga, Garcia and HeroFraga et al. 2011). Taken together, we expect that because immigration is a core source of variation among Latinas, we will likely find that Latina vote choice systematically varies along immigration-related characteristics.

Data and Methods

In these next sections, we primarily analyze data collected by the Collaborative Multiracial Post-election Survey (CMPS) which fields nationally representative post-election surveys with sizeable oversamples of Black, Latino, and Asian American respondents (Reference Frasure, Wong, Barreto and VargasFrasure et al. 2016; Reference Frasure, Wong, Barreto and VargasFrasure et al. 2020; see Online Appendix Table A3.1). As a study targeting racial minority populations, the CMPS is conducted in multiple languages and aims to collect a full cross-section of minority populations in the United States. CMPS is one of the few political surveys available that offer the opportunity to compare across racial minority groups. Moreover, given the large sample size of each racial minority group, CMPS enables the opportunity to compare Democratic and Republican minority voters. In contrast, national surveys like ANES collect relatively small samples of minority respondents, thereby offering less statistical leverage to explore the internal variation of any one group of minority voters.Footnote ¹⁰ Given that we aim to analyze three women of color groups across multiple elections, there are many comparisons to be drawn, and therefore to accomplish a more straightforward analysis, we analyze data from the two most recent presidential elections of 2016 and 2020.

Our objective in this section is to understand and identify the types of women of color voters who support Democratic candidates and observe if there are clear differences compared to women of color voters who support Republican candidates. For each election year, we calculate the demographic and other individual-level traits of Democratic and Republican women of color voters.Footnote ¹¹ We consider those factors established in the literature as informing partisan preferences for women including socioeconomic status, marital and family status, religiosity, regional location, and sexuality (Reference Campbell, Converse, Miller and StokesCampbell et al. 1960; Reference Hajnal and LeeHajnal and Lee 2011). Given our focus on Latina and Asian American women voters, we also consider factors related to immigration such as nativity and national origin.

Black Women Voters

For Black women voters, the first question is, who is the average Black woman Democratic voter and who is the average Black woman Republican voter? In other words, what kinds of factors influence Black women to be Democratic voters compared to Republican voters? In the CMPS data, nearly all (93%) Black women voted for the Democratic candidate in both 2016 and 2020 (see Figure 6). The overwhelming share of Black women who vote Democratic is consistent across election cycles demonstrating that their support is solidly Democratic.

Since nearly all African American women vote Democratic, the first important pattern to note is that Black women Democratic voters generally represent the population of Black women voters overall. Therefore, many demographic characteristics that define Black women Democratic voters are those factors that predict being a voter. Black women Democratic voters on average are more educated and older than the general population of Black women, which is consistent with our understanding that those with more resources are more likely to participate in politics (Reference Sidney, Lehman Schlozman and BradyVerba, Schlozman, and Brady 1995) (see Table 2). But on other characteristics not necessarily related to voting participation, we find that Black women Democratic voters are representative of the Black women population. For example, Black women Democratic voters are largely from the South, which is reflective of the residential patterns of the African American population as a whole (see Online Appendix Table A.3.2 for full results).

Figure 6 Vote choice among Black, Latina, and Asian American Women Voters in 2016 and 2020 CMPS

Sources: 2016 and 2020 CMPS

Of course, there is a small minority of Black women Republican voters in every election. Is there a systematic difference that differentiates Black women Republican voters from the Black women Democratic majority? To answer this, we first examined the general demographic makeup of Democratic and Republican Black women voters in 2016 and 2020 using CMPS. In Table 3, the percentages report the demographic characteristics of Black women Democratic voters and of Black Republican voters for each election. We then calculated a difference of proportion to understand if there are statistically significant differences in the characteristics between Black women Democratic and Republican voters. If the differences are statistically significant, the numbers are highlighted in bold in the table.

Comparing the breakdown of Black women Democratic and Republican voters, we find that while it might appear at first glance that there are raw differences across the characteristics of age, socioeconomic status, and family formation, most of these differences are not statistically significant. To further confirm if these differences are significant, we specified a logistic regression to predict support for the Democratic candidate among Black women voters (see Online Appendix Table A3.3 for full model results). In 2016, we find that only two individual-level factors reach statistical significance for predicting vote choice: marital status, and evangelical religious affiliation. In 2016, we find that holding all other factors constant, Black women who were single and unmarried relative to Black women with all other martial statuses were more likely to support the Republican candidate. Those who identified as evangelical were more likely to support the Republican candidate. In 2020, we also find two factors to systematically predict vote choice: age and evangelical affiliation. As in 2016, evangelical Black women were again more likely to support the Republican candidate in 2020, holding all other factors constant. In contrast, older Black women voters were more likely to support the Democratic candidate compared to younger voters. This comparison across election years shows that the types of Black women who vote for Republican candidates vary only slightly across candidate contests. In both election years, only two factors differentiated Black women Democratic voters from Republican voters. The lack of distinguishing features of Black women Republican voters is because this is such a small subgroup relative to Black women Democratic voters.

Another important insight from this analysis is that factors often associated with Republican support among white women voters, such as living in the South or socioeconomic status, have no statistically significant impact on Black women’s vote choice. For example, the majority of Black voters live in the South and nearly all Black women vote Democratic. Similarly, low-income voters make up the majority shares of both Black women Democratic and Republican voters. A systematic sorting of identity and interest groups into the two political parties is not occurring among Black women voters. This lack of systematic relationships between individual-level traits and vote choice for Black women voters demonstrates the constraints they have as a captured voting bloc within the Democratic Party.

While the demographic sorting into the two political parties is one way of understanding the constraint Black women experience at the voting booth, we also examine other indicators that reflect the relationship between constraint and partisan vote choice. Since we argue that the experiences of marginalization associated with both race and gender lead women of color to see the Democratic party as their only viable option at the voting booth, we expect that at the individual level, perceptions of racial and gendered marginalization correspond with stronger support for Democratic candidates.

To measure perceptions of racial and gendered marginalization, we can use survey items that ask respondents to rate the level of discrimination faced by African Americans and by women in society. By asking Black women respondents to rate discrimination against racial groups or against women, we can understand the extent to which they believe there exists systemic mistreatment due to race or gender. We anticipate that those Black women respondents who report facing significant discrimination are those who see race and gender to constrain one’s life chances. Therefore, we anticipate that there should be a positive relationship between the belief that there is a lot of discrimination and vote for Democratic candidates.

Using data from the 2020 election, we find that the overwhelming majority (88%) of Black women believe there to be “a lot” of discrimination facing Black Americans while 47% of Black women strongly disagreed with a statement that discrimination against women is no longer a problem (see Figures 7 and 8). Only a small share of Black women claim there to be no discrimination toward either group (4% believe there to be no discrimination against Blacks and 5% strongly agreed that discrimination against women is no longer a problem). These data show how Black women as a group understand that race and gender constrains their life chances.

Figure 7 Black women’s attitudes toward discrimination against Blacks in 2020

Source: 2020 CMPS

Figure 8 Black women’s attitudes toward discrimination against women in 2020

Source: 2020 CMPS

When we examine the relationship between perceived discrimination and Democratic vote choice in Figure 9, we see that among those Black women voters who perceive a lot of discrimination against Blacks, 95% voted for the Democratic candidate in 2020. In contrast, 83% of those who perceive no discrimination against Blacks voted for the Democratic candidate. Similarly, among those Black women respondents who strongly disagreed that discrimination against women is no longer a problem, 96% voted for the Democratic candidate. In contrast, 86% of those strongly agreed that discrimination against women is a problem for the Democratic candidate.Footnote ¹² We do find variation in Democratic vote choice by discrimination attitudes.

Figure 9 Relationship between attitudes toward discrimination and Democratic vote for Black women voters in 2020

Source: 2020 CMPS

This analysis shows that although Black women are a diverse population in terms of individual level demographics, they are nevertheless overwhelmingly Democratic voters. In this way, their experiences rooted in their intersectional race and gender position are of primary importance to their partisan candidate vote choices. Our analysis shows that Black women do not sort into two parties based on their individual-level traits or interests because they are constrained in their agency of vote choice. Black women also generally hold attitudes that reflect their awareness and belief that their race and gender pose severe constraints on their life chances, which also correspond strongly with their partisan vote choice.

Latina Women Voters

In 2016 82% and in 2020 75% of Latina voters cast a ballot for the Democratic candidate (see Figure 6). So while strong majorities of Latinas support for Democratic candidates, more of these voters than Black women cast a ballot for Trump in 2016 and 2020. To test the extent to which there is systematic sorting by identity and interest groups that characterizes Latina voting patterns, we provide data on the demographic breakdown of Democratic and Republican Latina women voters in these two elections in Table 4. Bolded numbers reflect statistically significant differences. The data do not show a clear consistent pattern in the demographic makeup of Latina women Democratic and Republican, and there are fewer differences across individual-level factors in 2016 compared to 2020.

Using logistic regression analysis, we further confirm which factors differentiate Latina Democratic from Republican voters (for full results see Online Appendix Table A3.5). In 2016, we find that immigration-related factors and religion differentiated Latina Democratic and Republican voters. In contrast, in 2020 there were differences across age, income, national origin, immigration status, and religiosity between Latina Democratic and Republican voters. In 2020, those of higher income, of Cuban national origin, evangelical, and the highly religious were more likely to vote Republican. In contrast, those who identify as LGBTQ were more likely to vote for the Democratic candidate. It is notable that there are distinctive differences in individual-level traits among Democratic and Republican Latina voters in 2020 but less so in 2016. It is possible that the 2020 election represented a more typical American national election characterized as a contest between two white male candidates, despite the unique context of the pandemic. When voting in 2020, Latina voters’ individual interests may have played a stronger role in determining their vote choice. Alternatively, it could be that in 2016, Latinas perceived more constraints on their vote choice given that gender was a more relevant factor with Clinton on the ballot.

One interesting pattern we find in this analysis are the findings for immigrant generation status. Although immigration from Latin American countries was a core issue discussed by candidates in 2016 and 2020 (Reference Sophia Jordán and ChrisWallace and Zepeda Millan 2020), we find that there is no significant difference in the share of immigrant voters who make up Democratic and Republican Latina voters. In contrast, we find a distinct pattern for the second-generation (children of two immigrant parents) voters and those in their third generation (at least one parent is born in the United States) and beyond. Latinas in their third generation or beyond made up the majority of Republican Latina voters in 2016. Second-generation voters made up a larger share of Democratic Latina voters compared to that found among Republican Latina voters in both 2016 and 2020. This suggests that being a child of immigrants leads Latinas toward greater support of Democratic Party candidates, while Latinas whose families have been in the United States for more generations are more favorable to Republican Party candidates.

While we find that Latina voters display greater agency of choice in partisan candidate choice compared to Black women, it is still the case that those who perceive strong constraints of race and gender support Democratic candidates. Turning to attitudes on racial and gender discrimination, we find that a sizeable share, 42% of Latina respondents in 2020, perceived there to be a lot of discrimination against Latinos and 28% strongly disagreed with the idea that discrimination against women is no longer a problem (see Figures 10 and 11). As a whole, a large share of Latinas agree that race and gender pose severe constraints, but at the same time Latinas are less likely to perceive discrimination against their ethnic group or against women compared to Black women voters.

Figure 10 Latina attitudes toward discrimination against Latinos in 2020

Source: 2020 CMPS

Figure 11 Latina attitudes toward discrimination against women in 2020

Source: 2020 CMPS

In addition, Latinas who perceive a lot of discrimination against Latinos or against women are more likely to vote for the Democratic candidate than those who perceive no discrimination.Footnote ¹³ As seen in Figure 12, there is a distinct contrast between the share of Biden voters among Latinas who perceive discrimination against Latinos or against women compared to those who do not perceive discrimination. Strikingly, an overwhelming majority (89%) of those who perceive a lot of discrimination against Latinos voted for Biden, whereas only 30% of those who perceive no discrimination against Latinos did so. Compared to what we found for Black women respondents, there is a stronger correspondence between individual attitudes about discrimination and vote choice. Perceptions of the constraints imposed by race or gender thus represent important predictors for Democratic vote choice among Latinas.

Figure 12 Relationship between attitudes toward discrimination and vote choice among Latina voters in 2020

Source: 2020 CMPS

We find that the US presidential elections in this analysis mobilized distinct Latina electorates, and this variation suggests that Latina voters do perceive that they hold agency in vote choice. Candidates and political parties can emphasize certain issues or agendas that generate support from different subgroups of Latina voters. Thus, Latina voters are less of a captured vote in the Democratic Party relative to Black women voters, though they are more constrained in their agency to choose than are white women voters.

Asian American Women Voters

Like Latinas, Asian American women demonstrate a stronger preference for Democratic candidates but at the same time reflect more heterogeneous attitudes compared to African American women. In 2016 79% and in 2020 72% of Asian American women voted for the Democratic candidate (see Figure 6). Relative to the other two women of color groups, Asian Americans demonstrate the widest variety in partisan vote choice all while being strongly supportive of Democrats.

Table 5 reports the demographic makeup of Asian American women voters who supported Democrats and Republicans in the two most recent presidential elections. In this analysis, we examine the same individual traits as for Latinas but include one additional variable, non-Christian religious affiliation, given that there is greater religious diversity within Asian Americans. We find there to be systematic variation in the individual-level traits between these two groups in both elections. We also find different patterns for 2016 than we do for 2020, demonstrating that each presidential race has mobilized different groups of Asian American voters. These results show that rather than demonstrating a stark pattern of captured voters, different subgroups of Asian American voters chose to cast their ballot for the candidate who best addresses their identities and interests in each election.

Results from multivariate analyses show that in 2016, Asian American women who were single, US-born, and Korean were more likely to support the Democratic Party candidate, while those who identified as evangelical were more likely to support the Republican candidate. In 2020, those who were older, lived in the southern states, and identified as evangelical were more likely vote Republican, whereas those who identified as lesbian, gay, or bisexual were more likely to vote Democratic (see Online Appendix Table A3.7).

One factor noteworthy about this analysis of Asian American women is the lack of national origin variation explaining partisan vote choice. Our analysis of national origin variation uses Chinese Americans as the comparison group (which represents the largest national origin group among Asian Americans). Chinese Americans are also the group least likely to identify as a strong partisan, thus making it an appropriate comparison group for analysis (Reference Wong, Karthick Ramakrishnan, Lee and JunnWong et al. 2011; Reference Masuoka, Han, Leung and Quan ZhengMasuoka et al. 2018). Analyses that include all Asian Americans find that Vietnamese Americans are more likely to identify as Republican relative to other groups (Reference Lien, Margaret Conway and WongLien, Conway, and Wong 2004; Reference Wong, Karthick Ramakrishnan, Lee and JunnWong et al. 2011). But this analysis focusing only on Asian American women does not show this same pattern. In contrast, we find that in 2016 there is only one significant difference between Korean women (relative to Chinese women). Furthermore, there were no significant national origin differences in 2020 among South Asian Indians despite the fact that Kamala Harris ran as Vice President on the Democratic Party ticket (but see Lemi, Arora, and Sadhwani 2020). It is possible that national origin differences are driven by Asian American men or within the broader Asian American population that includes non-voters.

Turning to perceptions of discrimination among Asian American women, 29% believe there to be a lot of discrimination against Asian Americans, but an even larger share (41%) of Asian American women report there to be “some” discrimination against Asian Americans in 2020 (see Figure 13). Similarly on the gender discrimination question, we find that a sizeable share of Asian American women respondents hold more moderate attitudes compared to African American women voters. Thirty percent strongly disagree that discrimination against women is no longer a problem, but 28% reported to neither agree nor disagree (see Figure 14). Among Asian American women voters, those with more moderate attitudes on discrimination make up a large share of the group. It is possible that these voters perceive discrimination consistent with the societal racialization of Asian American women where their racial group of Asian is seen as more privileged compared to Blacks and Latinos. Finally, a relatively smaller share of Asian American women voters perceive there to be significant discrimination compared to what we found for Black women respondents, and the pattern perceptions of gender discrimination attitudes are similar to Latina voters.

Figure 13 Asian American women’s attitudes about discrimination against Asian Americans in 2020

Source: 2020 CMPS

Figure 14 Asian American women’s attitudes about discrimination against women in 2020

Source: 2020 CMPS

There is a correspondence between discrimination attitudes and Democratic vote choice in 2020 for Asian American women (see Figure 15). Among Asian American women who perceive there to be a lot of discrimination against Asian Americans, 86% voted for Biden, while for those who perceive there to be no discrimination against Asian Americans, the rate was 46%. We find a similar pattern on attitudes toward discrimination against women.Footnote ¹⁴

Figure 15 Relationship between attitudes toward discrimination and vote choice among Asian American voters in 2020

Source: 2020 CMPS

The variation in individual-level predictors of partisan vote choice across elections demonstrates that Asian American women perceive that they hold more agency in vote choice compared to African American women voters who experience the strongest intersectional constraint. The racial tropes that stereotype Asian Americans have both positive and negative valances (i.e., “model minority” and “foreign outsider”), and as such, Asian American women may perceive racial discrimination as less of a problem than other minority women. Therefore, consistent with their racialization, Asian American women demonstrate more variation in their vote choice compared to Black and Latina women voters.

Discussion

The descriptor “women of color” is an all-encompassing term of relatively recent vintage, with much of its meaning relevant to differentiating women by race and ethnicity. By asserting this distinction, women of color emphasize their intersectional experience as being simultaneously racialized as non-white and distinct from the “default category” political subject as women. The analysis in this section demonstrates that this intersectional experience leads women of color to embrace progressive political positions and be stronger political supporters of candidates representing the Democratic Party. Women of color remain the stalwart of the modern Democratic Party and support their candidates by wide margins in US presidential politics.

Nevertheless, women of color voters have been taken for granted and ignored by political parties and analysts alike, and are not ordinarily the subject of systematic or sustained study. This needs to change, and the empirical analysis presented in this section provides the basis for a clearer picture of the politically relevant traits of Black, Latina, and Asian American women voters. While characteristics such as educational attainment and marital status have often been marshalled by political commentators to explain why women voters supported Trump in 2016 and 2020, our analysis shows that these arguments are less applicable to the partisan candidate choices of women of color. As argued at the outset of the section, all women of color are not the same, and instead, the analysis reveals a different degree of constraint by race, with Black women voters the most “captured” by the Democratic Party, followed by Latinas and then Asian Americans.

Taken together, this analysis and these results instantiate an intersectional analytical approach to explaining partisan vote choice in US presidential elections. While women of color cast ballots heavily in favor of Democratic Party candidates, this degree of support is based on both the issue positions of voters and more progressive candidates while simultaneously being a reflection of the intersectional position of non-white women voters as “captured” voters. To the extent that one of the two major political parties espouses policies to support the status quo marginalization of women and minorities, women of color have constrained agency in partisan candidate choice.

Diverse White Women Voters: Feminists, Working Moms, and (Still) Suburban Warriors

In addition to the presence of the gender gap, the abiding impression that white women voters are more Democratic than Republican is aided by two trends: the increasing national visibility of Democratic women elected officials and the transition of feminist beliefs into the political mainstream. The year 1992 was called the “year of the woman” in politics, a moniker intended to herald the election of a record number of women US senators. This feat enhanced the previous gender balance from 98 men and 2 women, to 93 men and 7 women seated in the Senate chamber of the 103rd Congress. Progress no doubt, but the pronouncement that a woman’s place was “in the House and the Senate” remained a hopeful feminist slogan rather than a reality. Five out of six of those newly elected women senators were Democrats, helping to solidify the perception that women were politically Democratic. Equally important, of the seven women US senators in 1993, all but one was white, and gender politics at the national level was still dominated by white women.

Alongside the growth of white women elected officials, there were also visible white feminist activists of the “second wave” generation who advocated for equality in politics, at home, in society, and in the workplace (Reference KleinKlein 1984). Almost two decades after the Equal Rights Amendment (ERA) failed to be ratified as a constitutional amendment, the feminist movement of mid-century had grown into the mainstream by the “year of the woman.” Younger women of a new generation had substantially more reproductive freedom than their mothers and grandmothers, and were able to access higher education and, concomitantly, a broader set of employment opportunities. Together these developments provided the context that explained why some white women voters cast their support for Democrats and are credited with creating the gender gap in preference for Democrats. But, of course, this is only part of the story.

Despite progressive change supporting greater equality for women in American society, conservative white women who believe in traditional gender roles and support Republican Party policies and candidates remain ever present though with a smaller imprint on the collective imagination compared to progressive and feminist politicians. The common – and erroneous – portrayal of white women voters as majority supporters of Democrats is a reflection of greater public attention to those white women who choose to reject patriarchal constraints, compared to white women conservatives who abide the normalization of traditional gender roles. Derisive characterizations of liberal white women are commonplace, from candidate Trump labeling his opponent a “nasty woman” to Republican US Senate Majority Leader Mitch McConnell shaming Senator Elizabeth Warren for continuing to speak (“nevertheless she persisted”). Within conservative political rhetoric, those who challenge a patriarchal worldview of appropriate behavior for women receive more attention than those who toe the line of traditional gendered behavior. In this area of the culture wars around sex and gender, analysts have overlooked the substantial political diversity and indeed the dominance of conservative and Republican Party supporters among white women voters in the United States.

Conservative women in politics have been present and politically influential since at least the time white women were granted the franchise (see also Reference JacobsJacobs 2014). Take, for example, Phyllis Schlafly, the conservative activist whose opposition to the ERA – marketed with catchy red octagonal signs emblazoned with “S.T.O.P.” (stop taking our privileges) – remains deeply rooted among white women voters (Reference CritchlowCritchlow 2008). Schlafly’s reference to privileges included being protected and cherished, and as such required the embrace of marriage and motherhood. But Schlafly is not the only example of influential white women in conservative politics. Scholars have detailed political leadership among Republican white women at different times of the conservative movement, and in her book Suburban Warriors, historian Lisa McGirr identified conservative white women activists in Orange County, California, in the early 1960s as among the originators of the “new American right” by cultivating politics that fused traditional Western liberalism with xenophobia and Christian fundamentalism (McGirr 2011). Indeed, Schlafly and other conservative white women suburban warriors of this era helped to forge the rhetoric and the playbook of modern conservative women, their legacy is clearly audible in contemporary efforts such as reproductive rights and advocacy for charter schools (Reference RymphRymph 2006). Just as importantly, conservative white women did not limit their politics to only gendered issues and have been involved in early efforts opposing universal suffrage (Reference GoodieGoodie 2012) and upholding formal practices of Jim Crow (Reference LoveLove 2016; Reference McRaeMcRae 2018; Reference DarbyDarby 2020).

White women voters overall reflect a diverse electorate that, while stereotyped as liberal and Democratic in their politics, are instead more conservative and Republican.

A Dynamic Electorate: Agency in Vote Choice for White Women Voters

Analyzing the diversity of the white women electorate in US presidential elections yields the fundamental fact that despite majority Republican Party candidate support among women, there remains a strong subset of this electorate that votes Democratic. Indeed, and because white women are the biggest slice of the pie within the American electorate, they are also the largest group of supporters for both Republican and Democratic Party candidates. Looking at 2016, for example, white women represented 32% of all Democratic voters and 45% of all Republican voters. Moreover, white women outnumber white men in both parties. White men represented only 26% of Democratic voters and 42% of Republican voters in 2016. Given that women are more likely to turn out to vote compared to men, white women represent a critical voting bloc for both Democrats and Republicans. Understanding the dynamics of white women voters is therefore crucial to unraveling the factors and reasons for candidate choice among this powerful group in the electorate.

At the same time, we point to the power and privilege white women voters hold in American elections. We argue that white women are understood overall as a powerful voting bloc and a core electoral base of both the Democratic and Republican parties. Unlike women of color who experience party capture and vote primarily for Democrats, white women are seen as important electorates to be cultivated by candidates and sometimes can be potentially lured to cross party lines. White women voters cast their ballots for parties and candidates who run on platforms that appeal to their interests and stances on political issues, and this is no different from what we expect of white men voters.

In this way, white women represent the group of women voters who hold the highest degree of agency in vote choice during presidential elections. Because white women voters can cast their ballot for the candidate or party that best addresses their identities and interests, we expect there to be systematic variation in the demographic makeup of white women Democratic voters compared to that of their Republican counterparts in each election. At the same time, because social, economic, and political norms are dynamic, we expect that the variation that exists among white women voters will vary across time and election context. We thus also expect that the factors predicting either Democratic or Republican vote choice for white women voters will vary in each election. The fact that it can change is a further reflection of their agency in vote choice.

Data and Methods

For this analysis, we utilize data from the 2008, 2012, 2016, and 2020 ANES post-election data. The ANES represents the ideal survey data for this analysis of white women voters over multiple elections given that the dataset includes a consistent set of survey items for every presidential election since 1948. While surveys like the CMPS offer large oversamples of racial minority respondents to identify variation within each minority group, the ANES provides a nationally representative sample of white voters. We caution that these data do not allow us to analyze at the individual level if voters change their opinions across elections since it is not a panel study design. Rather, it provides information about the traits among white women electorates supporting Democratic or Republican Party candidates for each election. Through a comparison of elections we can draw useful insights to show if, for example, Republican candidates across elections attract similar types of white women voters across elections or if they vary by election. The analysis provides evidence as to the extent to which white women voters are strong partisans who vote for the same party across elections or if they shift their support in response to particular candidates and platforms.Footnote ¹⁵

We selected the most recent four elections (2008–2020) given that within this set there were a number of historic firsts such as the opportunity to elect the first Black president, the first woman Republican Party vice president, the first woman president, and the first Black and Asian vice president. In addition, both 2008 and 2016 were considered relatively competitive contests since the party nominees did not include an incumbent and research shows that more voters are more likely to be engaged when the contest is competitive (Reference Groffman, Collet and GriffithGrofman, Collet, and Griffin 1998). As in in the previous section, we present factors commonly identified as being related to partisan candidate choice.

Who Are White Women Democratic Voters and Who Are White Women Republican Voters?

Table 6 presents the characteristics of white women Democratic and Republican electorates in each presidential election between 2008 and 2020.Footnote ¹⁶ The percentages report the demographic characteristics of white women Democratic voters and Republican voters as well as a difference of proportion. Statistically significant differences are highlighted in bold in the table. For example, in 2008, 35% of white women Obama voters held no more than a high school level of education and 37% of white women Obama voters were college educated. However, there are no significant differences in the distribution of education among white women Democrat versus Republican voters in 2008. In contrast, in 2016 and 2020, white women Democratic voters were more likely to be college-educated compared to Republican voters. We present a visualization of these differences in Figure 16.

Figure 16 Educational makeup of white women Democratic and Republican voters 2008, 2012, 2016, and 2020

Source: ANES 2008, 2012, 2016, and 2020 Time Series

There is also consistency in the religious affiliation, sexual orientation, and regional makeup of white women Democratic and Republican voters across elections. In terms of regional makeup, Republican supporters are a third or more composed of voters from southern states. For example, white women Republican supporters in 2008 were heavily represented by southern women, with 55% that election year living in that region. This is in contrast to the 30% of the sample of white women who voted for Obama in 2008 that lived in the south. We note that the data take an interesting turn over time, however, with much smaller proportions of both supporters of Republicans and Democrats residing in the south. This suggests that overall, white women electorates are becoming more dispersed in terms of residence across the United States.

In terms of religion, the data show that at least a third of Republican white women voters are highly religious (as measured by church attendance rates) and a good share identify as evangelical or born-again Christian. We acknowledge that the distinct difference between the share of religious voters in 2008 compared to other election years might be due to differences in question wording across surveys rather than changes in the makeup of electorates over time.Footnote ¹⁷

A third characteristic of which we find there to be statistically significant differences across Democratic and Republican white women voters is the makeup in terms of sexual orientation. Those who identify as lesbian, bisexual, or other make up a small share of the overall population, but they comprise a larger share of white women Democratic voters compared to white women Republican voters. This finding on is important as it demonstrates how white women voters can be responsive to party cues that speak to a marginalized status. Like the political stances made on behalf of racial and gender equality, the Democratic Party has over recent election cycles taken a clear political stance supporting gay marriage and LGBT+ rights.

While there were some consistent patterns across elections, more notable are the differences across electorates on the demographic factors analyzed. In terms of age, white women Democratic voters on average are a few years younger on than their Republican counterparts, though the difference is not large. When looking at specific age groups, in 2008 and 2020, younger white women voters made up a larger share of Democratic voters than their makeup of Republican voters. In 2016, older voters (those 65 years or older) made up 30% of Trump voters compared to the 23% of Clinton voters. This shows that some age groups were more mobilized to support specific presidential candidates in certain contests.

The most striking finding, however, is the absence of a strong pattern of Republican Party or Democratic Party candidate support by socioeconomic status across election years. In particular, each shows different compositions of income categories among voters for each party’s candidates. Using household income as a measure, Democratic-supporting white women voters were composed more of lower income voters in 2008. By 2016, however, it is Republican white women who have a larger representation among lower-income voters within Trump’s coalition that year. For instance, in 2008, an estimated 54% of Democratic women voters had a household income of less than $50,000 a year, while in 2016 only 30% of Hillary Clinton’s white women voters were in this income bracket. Interestingly, while there are significant differences on makeup by household income for 2008, 2012, and 2016, in 2020 there are not distinctive differences in income level between Democratic and Republican voters.

Similarly, and in terms of educational attainment, data from 2008 show no significant differences between Democratic and Republican white women voters. An almost identical proportion of Obama supporters had high school as their highest attainment, as did Republican white women voters. Similarly, there is a near-perfect match in proportion with a college degree in 2008 for both the Democratic and Republican candidate coalitions. But since 2012, college-educated voters make up a larger share of Democratic voters compared to Republican voters. In 2016 and 2020, there are big differences between Clinton and Trump voters, with over half of Democratic voters holding a college degree (57% in 2016 and 56% in 2020).

Finally, and in terms of family structure, married women make up a large share of both parties but represent a larger share of Republican white women voters across the presidential elections between 2008 and 2020. In contrast, women who have never married make up over one fifth of Democratic voters across all four elections. We do not find any significant differences in the makeup of women with children in the household across the two parties.

Predicting Republican Support across the Last Four Presidential Elections

The analysis in the previous section helped to form a picture of the white women electorate and what differentiates Democratic versus Republican voters. However, it is possible that demographic factors may be correlated with one another. For example, it could be the case that southern voters are more likely to be older or married and so we would need to account for the possibility that regional location has a distinct impact on partisan vote choice from that of age and marital status. To ascertain if there are factors that have a systematic impact on partisan candidate choice we turned to multivariate analyses.

Since more white women voters prefer Republican candidates, for this analysis we specified a logistic regression model predicting vote choice for the Republican candidate (compared to vote for a Democratic candidate).Footnote ¹⁸ For independent variables, we tested for the same demographic characteristics analyzed in the previously. Given that we use a logistic regression, the coefficients for each independent variable are not easily comparable across one another. To compare the magnitude of each coefficient with that of each of the others in the model, we also calculate log odds for each independent variable in the model. For log odds, values greater than 1 demonstrate that the independent variable has a positive effect on the dependent variable with larger numbers reporting there to be a stronger magnitude. Values between 0 and 1 demonstrate that the independent variable has a negative effect on the dependent variable, with values closer to 0 reflecting a stronger (negative) magnitude of an effect. For ease of interpretation we only present significant results in Table 7.Footnote ¹⁹

The results from the multivariate analyses show that there are some factors that have a unique impact predicting Republican candidate choice among white women. Yet, just like the demographic analysis, we find that there are not consistent predictors when we compare results across presidential elections between 2008 and 2020. This tells us that white women who vote Republican are not necessarily comprised of the same population across elections. However, we do find some factors that are significant predictors across elections, for example, being an evangelical Christian and living in the South both predict Republican vote choice among white women. Identifying as LGBT+ is the one variable that consistently predicts Democratic vote choice among white women.

The results from these analyses reveal that if we want to characterize the partisan candidate choices of white women voters, one way to describe them is as swing voters. Systematically examining the factors that predict partisan candidate choice across four presidential elections shows that while there are some demographic differences in the makeup of Democratic and Republican white women voters, this absence of consistent patterns of partisan support – with the exception of living in the south, religiosity, and sexual orientation – demonstrates changing electorates in each election. One conclusion from the findings of the absence of level of education or marital status or even income having consistent effects over time among white women in their partisan candidate choices is that white women voters are both dynamic and heterogeneous. Differences in the context, candidates and other time-specific factors found in each election affect the vote choice of white women voters, giving credence to the truism that every election has a distinct electorate.

Discussion

The fact that white women voters are more evenly split in their partisan candidate choices for US president underscores both the broader agency of white women to choose between the political parties and their electoral power in choosing the President of the United States. Constrained as they are by their gender, in terms of the positions the two major political parties take on traditional gender roles and accompanying policies on gender equity, white women nevertheless can and do choose between the two parties in a more evenly distributed fashion especially compared to women of color. In contrast, Black, Latina, Asian American, and other women of color give heavily lopsided support for Democratic Party candidates. Women of color do not have the same degree of agency to choose, for the policies of one of the parties constrain their realistic choices at the ballot. Together these patterns of partisan candidate choice among highly heterogeneous white women voters and constrained agency for women of color lay the groundwork for revealing the dynamics of the gender gap in the next section.

Identifying and Explaining the Gender Gap

The narrative about women voters being stronger Democratic Party supporters compared to men is based on a decades-long literature in political science scholarship (Reference Box-Steffensmeier, DeBoef and LinBox-Steffensmeier, DeBoef, and Lin 2004; Reference WhitakerWhitaker 2008; Reference Fox and CarrollFox and Carroll 2013). The longitudinal data presented in Figure 17 document the gender gap over time for the entire series of ANES data between 1948 and 2020. While there is variation over the nearly seventy years of US presidential election data, the gender gap in the early years is negative, meaning women voters were less supportive of Democrats compared to men prior to 1964. Since then, the gender gap is positive and modest, usually in single digits with the exception of the US presidential election years in the 1990s when Democrats Bill Clinton and Al Gore were more popular with white women voters.

Figure 17 Gender gap difference between women’s and men’s support for Democratic candidates, 1948–2020

Source: ANES Cumulative File, 1948–2020

The elections of 1992, 1996, and 2000 were, if not unique in the history of modern presidential contests, nevertheless different from other years since the 1980s because of the inclusion of third-party candidates who drew substantial votes from the two major party candidates. Ross Perot ran in both 1992 and 1996, siphoning support from Republican Party nominees George H.W. Bush and Robert Dole, and therefore inflating the size of the gender gap in those two elections. In 2000, the candidacy of Ralph Nader drew voters from the opposite end of the political spectrum. Overall, and taken together, the gender gap in partisan candidate choice is visible and persistent.

Another way to look at the data is to track the preferred party candidate of women and of men over time. Figure 18 presents the same data used to calculate the gender gap in Figure 17, but instead of highlighting the difference between men and women, it shows the share of women and of men who supported the Democratic Party candidate in each presidential election. Until 1980, men voters supported Democrats at a higher rate than women voters in some elections and for others the reverse is true, and there is no discernible gender and party phenomenon in the earlier years. In contrast, and starting in the 1980s, the data show consistent majority support for Democratic Party nominees among women. In sum, the conventional wisdom since the 1980s is that women voters appear to be majority Democratic in their support of US presidential candidates, and also more Democratic than men voters.

Figure 18 Percent support of Democratic Party candidates by gender, 1948–2020

Source: ANES Cumulative File, 1948–2020

Since the gender gap was identified, researchers have attempted to explain the reasons for the variation. Academic studies narrowed in on gender differences on policy issues such as social welfare (Reference Howell and DayHowell and Day 2000), foreign policy (Reference Chaney, Michael Alvarez and NaglerChaney, Alvarez, and Nagler 1998), abortion (Reference MansbridgeMansbridge 1986; Reference Cook, Jelen and WilcoxCook, Jelen, and Wilcox 1993), economic attitudes (Reference Welch and HibbingWelch and Hibbing 1992; Reference Seltzer, Newman and Voorhees LeightonSeltzer, Newman, and Leighton 1997), and feminism (Reference ConoverConover 1988; Reference Cook and WilcoxCook and Wilcox 1991) that were more consistent with progressive and Democratic Party positions. The interpretation that developed into conventional wisdom is that women are more politically liberal on issues compared to men due to their primary role as caregivers and the experience of gendered stereotypes leading to a stronger feminist consciousness. Other scholars have argued that increased economic independence of women (Reference Manza and BrooksManza and Brooks 1998), the declining significance of traditional marriage (Reference Iversen and RosenbluthIversen and Rosenbluth 2006), and increased educational opportunities for women (Reference Gillion, Ladd and MeredithGillion, Ladd, and Meredith 2020) explain the conditions under which the gender gap developed.

While most of the literature focuses on the distinctiveness of women’s attitudes and the implications of those opinions for candidate choice, other scholars have pointed out that women’s candidate choices and partisan identities have remained relatively stable over time. Instead, they interpret the gender gap as a reflection of the diminution of support among men for Democratic Party candidates for president (Reference Kaufmann and PetrocikKaufmann and Petrocik 1999). Alternatively, Reference Box-Steffensmeier, DeBoef and LinBox-Steffensmeier, De Boef, and Lin (2004) argue that fluctuations in the size of the gender gap are due to larger, macro-level factors such as the political climate (like public support for conservatism) and economic conditions (such as unemployment and inflation) since these factors influence the individual lives of men and women in systematically different ways (Reference OndercinOndercin 2017).

Taken together, this research points to the reaction of women to support a politics that challenges, rather than enhances, patriarchal dominance and inequality on the basis of gender. Important as this is for the study of electoral politics and candidate choice, the fact remains that while women voters have shown stronger support for Democratic Party candidates compared to men voters, there is far from uniformity among women.

The Gender Gap Is a Race Gap

The persistence of the gender gap in election results alongside the robust academic literature on the concept reinforces the widespread characterization of American women as Democratic Party supporters (Reference DittmarDittmar 2017). But the data do not support the argument that this is the modal position among all American women voters. Instead, the reliance on defining women’s choices in comparison to those of men – as the default category from which all else deviates – is a flawed analytical position because it obscures the variation within women voters.

As the dataFootnote ²¹ in Figure 19 demonstrate, the most striking difference in support for Democratic Party candidates within the population of women voters in the United States is race. White women voters are the least supportive of Democrats compared to African American women and Latina voters. While the data for Latinas in the ANES surveys are sparse for this fast-growing group of voters and most reliable from the election of 2008 and onward, the basic pattern among Latinas is instructive.Footnote ²² Black women are the strongest and most consistent supporters of Democratic Party candidates, followed by Latinas who also vote for Democratic nominees by a strong majority. In contrast, white women show the lowest support for Democratic Party candidates, consistently less than 50%, with only two exceptions in the years of 1964 and 1996 in majority support for Democrats. Rather than comparing women to men to generate the conventional gender gap number, comparing partisan candidate choice by race and ethnicity within the population of women voters reveals that not all women voters in US presidential elections are Democratic supporters. Instead, the gender gap is created by a race gap, with women of color leading the way.

Figure 19 Percent support of Democratic Party presidential candidates by race, 1948–2020

Source: ANES Cumulative File, 1948–2020

Democratic Party support among African American women surpasses 90% in 1964 and then essentially never wavers. Black women consistently and overwhelmingly support Democratic candidates, and these voters are the stalwart of the Democratic Party in US presidential elections (Reference Jordan-ZacheryJordan-Zachary 2007). Support for Democratic candidates among Latinas is less consistent compared to African American women, particularly in the early elections between 1988 and 2000.Footnote ²³ But since 2000, the data reveal a steady increase in Democratic Party candidate support among Latinas, never dropping below 70% in the last four elections. Although there are less data on Asian American women, surveys focusing on this population of voters show that Asian American women support Democratic candidates at a level similar to Latina voters. Explanations for the strong Democratic support among women of color have been discussed in a growing body of research and reflect the intersectional position of women of color as marginalized by both gender and race (Garcia Reference Bedolla and ScolaBedolla and Scola 2006; Reference Hardy-Fanta, Lien, Pinderhughes and Marie SierraHardy-Fanta et al. 2016).

Despite the wide variation between white women and women of color, there remains a partisan gender gap by race. Table 8 details the proportions who supported Democratic Party candidates for US president by gender and race/ethnicity during the last four elections.Footnote ^24,Footnote ²⁵ Looking at the gender gap calculations, we find that there are differences in the size of the gap between women’s preferences for Democratic Party candidates compared to men for all racial groups in all US presidential elections between 2008 and 2020.Footnote ²⁶ The relatively modest sample sizes for women of color in the American National Election Study data caution one from drawing strong conclusions about the observed differences. Nevertheless, the data show that women of all races and ethnicities are more supportive of Democratic Party nominees for US president than men within their group across the last four elections.

These findings are consistent with the small, but growing, literature on the gender gap among minority voters. Political science research has identified a consistent gender gap in support for Democratic Party candidates among Black voters, though studies analyzing Asian American and Latino partisan candidate choice have been less conclusive. Some elections reveal a gender gap among Asian American or Latino voters, while others show only a small or insignificant difference between candidate preferences (Reference LienLien 1998; Reference Bedolla, Lisa and PantojaGarcia Bedolla, Lavareiga Monforti, and Pantoja 2007; Reference Bejarano, Manzano and MontoyaBejarano, Manzano, and Montoya 2011; Reference Lizotte and CareyLizotte and Carey 2021).

Given the position of the two major political parties in the United States on gender and racial equality, it can be expected that we would find important variation both by race and by gender of the voter. But while both matter, the data here show that in terms of predicting partisan vote choice, there are greater differences by racial group than there are by gender.

The Significance of Race in Support for Democrats among Women Voters

In this section, we analyze the significance of race in vote choice for Democratic Party candidates among women voters in the United States between 1988, beginning around the time the gender gap was first noted, and through the most recent election of 2020, again utilizing data from ANES. In the previous section we presented the proportion of voters who supported Democratic Party candidates by race and gender. However, it remains a question whether or not race – and specifically being white – is a primary factor differentiating presidential vote choice among women voters. Not only does each election offer different candidates for voters to choose from, but every election also comes with a distinct electorate, the composition of voters changing with time and population replacement and also as a function of who decides to turn out to vote. Thus, variation in the quality of candidates for US president, their communications and mobilization strategies, and the reactions of voters all matter for the outcome of the elections.

In this analysis, we identify the key individual-level traits that predict Democratic vote choice in each election since 1988 using multivariate analysis. By comparing results across elections, we can ascertain whether or not the predictors of support for Democratic Party candidates vary by election, or if there are some traits that consistently predict partisan candidate choice across elections. If we find race to be a factor that consistently predicts vote choice, the evidence that the gender gap is driven by a race gap is more persuasive.

We specify statistical models that account for factors that prior scholarship has identified as being relevant to partisan candidate choice, including age, income, educational level, race, being from an immigrant family,Footnote ²⁷ and religiosityFootnote ²⁸ among the measures relevant to supporting Democratic Party candidates (Reference Campbell, Converse, Miller and StokesCampbell et al. 1960).Footnote ²⁹ The main difference between the model specified here and established practice is that we include respondent race as white as a dummy variable rather than the reference group. This model specification is consistent with traditional analyses of political behavior that have isolated particular groups into dummy variables in order to identify whether or not they differ from the reference group, which in this case is women of color. This strategy is distinct from the intersectional approach we took in earlier sections, and is deployed here to demonstrate that it is white women voters who are the outlier category among US women voters overall.

In addition to individual-level traits, and because we are predicting partisan candidate choice instead of turnout, the model is specified to include factors that account for whether or not the voter supports the Democratic nominee in that presidential election year. Therefore, we also include characteristics relevant to a voter’s position in the economy (being employed and being a union member) that has been argued to play a role in vote choice. We include controls for the location of the voter such as living in the SouthFootnote ³⁰ and urbanicityFootnote ³¹ given the long-standing geographic factors that influence partisan politics. Because the model is estimated for women voters only, past literature has indicated the importance of family status for women voters, and thus marital status is also included.

Using the aforementioned factors as independent variables in the model specification, we estimated nine logit models predicting voting for the Democratic Party candidate for each presidential election between 1988 and 2020 among women voters. The results from the logistic regression model estimations are detailed in Table 9. Because the equations estimated here are logit models, the data in the table show the log odds for those factors that reached statistical significance (p<0.05). Log odds allow us to present a standardized reporting for each independent variable in the model and allow us to understand the magnitude of the effect of that variable. To interpret log odds results, as the number rises above 1, it reflects a stronger positive relationship (and those that predict Democratic support). Numbers between 0 and 1 reflect a negative relationship (and those that predict Republican support) with numbers closer to 0 reflecting a strong negative relationship.Footnote ³²

Of those factors that reach statistical significance, most of the effects are consistent with expectations from the literature. In the models for the elections between 1992 and 2020, we find that those who attend church more regularly are more likely to be Republican. However, while the directionality of each factor is consistent with what we would expect, the other important finding we learn from the results of the models is that few variables consistently predict partisan vote choice over every election. Some factors play a stronger role in explaining partisan vote choice in some election years but not in others. This tells us that there are many features related to the specific characteristics of each election such as the types of candidates on the ballot and level of electoral competitiveness that influence which voters choose to turn out in any given election, and how those contours of the electorate affect the outcome.

Strikingly there is only one variable that reaches statistical significance for every election between 1988 and 2020, and it is whether the woman voter is white. In every election, white women are significantly less likely to vote Democratic compared to women of color, all other conditions held constant. The effect of race does not appear to be influenced by choices in statistical modeling since we find that even in model specifications that include other relevant variables such as partisan identification, race continues to have a constant and significant effect on partisan vote choice.Footnote ³³ This analysis demonstrates that race is less subject to election-specific variation, and instead shows that being white is central to the formation of an individual’s partisan candidate choice compared to other demographic factors. In terms of studying women voters, the implication is that race must first be factored into the analysis before we can estimate the role of gender. This conclusion is consistent with suggestions made by scholars who have studied women voters, including Reference Huddy, Cassese, Lizotte and WhitakerHuddy, Cassese, and Lizotte (2008), who argue that the partisan variation that occurs across subgroups of women is evidence that women cannot be framed as one coherent voting bloc but are rather characterized by political “disunity.”

Taken together with the data demonstrating the systematic variation between white women and women of color in support for Democratic Party candidates – with the former consistently voting majority Republican and the latter consistently voting strongly majority Democratic – it is clear that race is a crucial source of variation in how women voters express their preferences in presidential elections. That said, the data presented also demonstrate that gender remains important as well, and, in particular, that women of all races and ethnicities are more supportive of Democratic Party nominees compared to men.

Combined with the compositional over time data presented in the first section of this Element, the gender gap will continue to persist as long as women of color remain strong supporters of Democrats, and as African American, Latina, Asian American, and other women of color join the electorate. Compositional change in the American electorate has occurred over time as a function of enhanced voter enfranchisement of minority Americans along with the increase in immigrant populations as they naturalize to citizenship and bring their U.S.-born children into the polity. The future direction of women voters in terms of their partisan candidate preferences, however, is subject to change, and it is the political parties themselves that have the most to gain and alternatively the most to lose in a political context of dynamism, dealignment, and realignment. The implications of divergent patterns of partisan candidate support among women of color and white women voters in the United States, along with the companion set of results showing the gender gap is a race gap, are taken up next in the final section.

Compositional Change Informs Partisan Formation and Coalitions

Compositional change in the population as the result of immigration and increased racial diversity has generated anxiety among those who comprise the majority of the American population (Reference Perez, Vicuna and RamosPerez et al. 2023). Policy on immigration in the United States has most often been interpreted through a racial and ethnic lens, a pattern dating from the British colonial period, though who was defined as an outsider in these terms has changed over time. Xenophobic sentiment from negative tropes leveled against German newcomers (Reference FranklinFranklin 1755), to Asian exclusion and the federal laws that followed mirrored in important ways the dynamics of the racialization of the Irish, Italians, southern Europeans, Jews, Slavs, and others as “less than white” that culminated in the 1924 National Origins Act (Reference IgnatievIgnatiev 1996; Reference JacobsonJacobson 1999). More recently, policy barring immigrants from Latin America and predominantly Muslim nations is a stark reminder that the perceived desirability of newcomers remains an important modifier of a conditional welcome in US immigration politics.

In the contemporary context, compositional change has increased the racial and ethnic diversity of the population and, in so doing, incentivized political parties and elected officials to make immigration and race politics a political issue. While immigration policy has become a divisive partisan issue in recent decades, there have been periods of time when the parties were not in conflict with one another and where both of the two major political parties advocated for restriction (Reference GyoryGyory 1998; Reference TichenorTichenor 2002). Similarly, the 1965 passage of the landmark immigration legislation, the Immigration and Nationality Act, was supported by a large bipartisan majority, for few expected this law to have the effect it did on racial diversity (Reference Masuoka and JunnMasuoka and Junn 2013). In the two decades that followed, the positions of the Republican and Democratic parties on immigration policy have evolved from what Daniel Tichenor has described as one of “strange bedfellows” – where it was Republican President Ronald Reagan who signed a law that provided a path to citizenship for previously unauthorized immigrants and where it was Democratic President Bill Clinton who a decade later signed a punitive law that criminalized unauthorized immigration – into a politics of clear “dividing lines” between Republicans and Democrats.Footnote ³⁴ Republicans are now seen as staunch restrictionists whose nominees for US president have pursued policies to “build a wall” and separate migrant children from their parents at the southern border, while Democratic Party presidential candidates are relatively progressive in their immigration policy positions, indicating a more welcoming stance for newcomers who are disproportionately non-white minorities in the United States.

Thus, immigration and, by extension, the politics of race and ethnicity have become politicized issues, with the two major political parties at opposing ends of a continuum representing restrictionist to more progressive positions. Parties and elites drive this policy development together with their voters, and on an issue as politicized and framed by racial and ethnic tropes as is the politics of immigration, the distinction between the Republican Party and the Democratic Party presents a stark choice for voters for whom immigration is an important issue and importantly for those whose are classified by race and ethnicity as something other than white. Thus, as the parties have moved toward opposite ends of the immigration policy continuum, voters who are less constrained by their race and ethnicity can choose on the basis of which party they agree with more and then sort accordingly. This degree of agency, however, is constrained for voters who themselves and their families may be the target of restrictionist immigration policy. In this way, compositional change informs how individual voters develop attachments to a political party, and partisan candidate choice for president is guided not only by ideological beliefs about the role of government or positions on economic policy issues, for example, but also by how one is affected by compositional change.

Compositional change in the population in terms of race and ethnicity is thus a source of dynamism in elections precisely because it informs not only the positions political parties take on immigration policy but also the contours of coalitions among voters. The full extent to which American voters are influenced by this population change and the immigration policy stances of parties in their partisan formation and loyalty to a political party remains to be seen, particularly for the most recent newcomers in the American electorate. But as the data on strong support for Democratic Party candidates among Latina and Asian American women – two groups with the highest proportion of immigrant and second-generation voters – attest, persistent Republican Party restrictionist positions on immigration and policies favoring criminalization and incarceration of migrants present an unwelcoming position and a constraint for minority voters to support Republican Party candidates.

Partisan Dealignment and Realignment Are Happening Now

The Republican Party and the Democratic Party are visibly divergent on policies related to the maintenance or attenuation of hierarchies based in both race and gender. Immigration policies are one prominent area of variation, as are the politics of reproductive rights. In contrast, the Democratic Party has marketed itself and pursued policies supporting gender equity, racial justice, and relatively progressive positions on sexual orientation and gender identity, while Republican Party policy at the national level has become dominated by candidates and platforms that have instead doubled-down on maintaining hierarchies and supporting “racially conservative” and “socially conservative” policy. While the internal dynamics of political parties are beyond the scope of this Element, the symbiotic relationship between parties and voters is evident in the data we presented on partisan candidate choice among women voters.

The extent of the polarization between the two major political parties and within American voters has been identified as a source of concern for the health of democratic politics. The fact that the electorate is dynamic due to immigration, compositional change as well as generational replacement, and that voters are open to conversion, are constants. There are periods when the ideological positions of the parties are changing in order to attract certain blocs of voters into their coalition and we argue is the present condition in US politics. This is an indicator of how parties are responding to what is happening in the electorate. While most of the attention among analysts has focused on political polarization, an equally important dynamic is the persistent dealignment of voters away from both parties and the accompanying rise in the proportion of American voters who call themselves independent of either party. This is the elephant in the room, and as evident as dealignment has been, it can only be ignored because elections are still, for the most part, contests between candidates nominated by the two major parties that voters must choose from. In tight elections, however, third parties can make a difference, but, more importantly, mobilizing independent and swing voters is key to winning.

Our analysis of the partisan candidate choice of women voters provides a window into the extent to which the positions taken by the parties on race and gender are increasingly leading to a systematic sorting of voters. The data presented earlier in this Element showed that the makeup of the Democratic and Republican parties is becoming more distinct over time. with the Democratic Party attracting greater race-gender diversity while the Republican Party remains comprised of white voters. The fractures in the political party system that are evident in the splintering of the Republican Party before and after the election of Donald Trump in 2016 and now into the present are an important indicator of partisan dealignment and realignment. Given the modal voter status of women in the electorate combined with the growth of the population of minority voters, the gender and racial composition of politicians will also change accordingly, though we might expect the path to elected office will remain more difficult for women overall and for Republican women (Reference Holman, Merolla and ZechmeisterHolman, Merolla, and Zeichmeister 2022).

To understand the partisan dealignment and realignment that is occurring now, analysts will need to go beyond traditional New Deal realignment explanations based on voter conversion, mobilization, and generational replacement models, and instead incorporate additional explanations. In the present political moment, the population itself is changing due to immigration, and if not fully parallel to the creation of a Democratic majority in 1928–1936, that period and its dynamic is an important starting point for analysis (Reference AndersenAndersen 1979). The significance of race and ethnicity in the modern context should also overlay the analysis (Reference SchicklerSchickler 2016). Embracing dynamic and intersectional approaches to analysis will contribute to a better understanding of the dynamics of voter dealignment from the two major US political parties, and help to chart the path of a new partisan realignment in the future.

Identity Politics Will Continue to Be Relevant

Identity politics has been derided most frequently by conservatives and members of the Republican Party when referring to equality and rights-based claims by women and people of color in particular. Responses to political movements such as Black Lives Matter, #MeToo, and the 2017 Women’s March have denied the importance of identity categories in politics, while simultaneously reinforcing the race and gender hierarchies that make those categories politically relevant. Agency and constraint continue to be in tension, and among the most important imperatives of an “identity” politics is to assert power from a marginalized position whether based in a racial or ethnic category, or a broader minority identity as “people of color” (Reference PerezPerez 2021; Reference Perez, Vicuna and RamosPerez, Vicuna, and Ramos 2023).

When Barack Obama ran as the Democratic Party nominee for US president in 2008 he was often described as the Black candidate for president. When Hillary Clinton ran at the top of the Democratic Party ticket in 2016 she was described as the women candidate for president. But in those two elections, it was not necessary for John McCain to be described as a candidate with a racial modifier, nor was it necessary for Donald Trump to be labeled a white male presidential candidate, because they represented the default categories of race and gender in American political leadership. Nevertheless, and in an environment of deep political polarization and in the presence of a politics of grievance – some whose roots can be traced to the rapid racial and ethnic diversification of the population and the rise of women in politics – a variety of new angles of “identity politics” have emerged.

White Republican voters in particular have voiced their own version of identity politics, highlighting their aggrieved status perceived to be on the receiving end of political correctness and critical race studies, for example. White racial identity is the default racial category in American politics to the point where it is a status that until recently did not need to be named. The impact of strong identification with being a white American has been shown to have important effects on partisan candidate choice and voter mobilization (Reference Weller and JunnWeller and Junn 2018; Reference JardinaJardina 2019). A companion set of claims for upholding the virtues of traditional masculinity is among the newest set of claims in identity politics, with US Senators and other politicians from the Republican Party heralding the need for prioritizing manhood. Thus, identity politics is no longer the purview of traditionally marginalized groups of Americans, and now includes an updated set of categories vying for recognition.

Identity politics – whether rooted in a long history of discriminatory treatment based in gender and race or activated by a sense of grievance at what has been perceived to be lost – are the result of compositional change, policy innovation, and macro-structural conditions within and outside of US politics. Groups of Americans who have long been at the forefront of progressive politics advocating for social justice and policy change aimed at enhancing civil rights, voting rights, LGBT+ rights, and women’s rights remain at the heart of identity politics precisely because patriarchy and racial hierarchy remain in place. All the same, the emergence of groups of voters representing new identities is very much a part of the dynamism in American politics we should expect, and these updated contours of identity politics will continue to be relevant to electoral politics and partisan candidate choice. Throughout these dynamisms, the significance of women voters in US elections remains constant.