Introduction
The 2024 presidential election saw a woman with a long-standing political career lose to President Donald Trump for the second time in the United States. Vice President Kamala Harris’ 2024 campaign was unprecedented in several ways, including her status as the first woman of color running for president under a major political party. Throughout this campaign, Harris faced intense scrutiny and sexism, as is often encountered by women seeking political office. Some of the American public react negatively to women running for office, especially those seeking the presidency (Winter Reference Winter2023; Ditonto Reference Ditonto2019; Paul and Smith Reference Paul and Smith2008; Ditonto and Mattes Reference Ditonto and Mattes2024).
This project examines evaluations of former Vice President Harris relative to President Trump across race and gender. Harris’ campaign offered several communities an opportunity to be represented in the highest office in the United States. Scholars of descriptive representation argue that minorities seek out and support politicians who share their marginalized status (Mansbridge Reference Mansbridge1999) in hopes that these figures will represent their groups’ interests. One might anticipate women and racial minorities, Black (Brown and Slaughter Reference Brown and Slaughter2024) and Asian Americans especially, to support Harris. However, many women, particularly White women, voted for Trump (Sanders Reference Sanders2024).
This paper examines the political consequences of different conceptualizations of sexism, assessing the public response to a presidential candidate with multiple marginalized identities across race, ethnicity, and gender. Harris’ background as a multiracial, multiethnic woman seeking the presidency makes her a distinct case. Such identities equip Harris to uniquely appeal to the growing multiracial and multiethnic population in the United States (Ventura and Flores Reference Ventura and Flores2025). Furthermore, her identities bring disparate communities into conversations and offer a foundation to build electoral coalitions with one another. The effectiveness of these coalitions may be further shaped by the gender of those within these racial communities. Through taking an intersectional lens to understand public opinion formation toward Harris as a woman of color, I offer insights into the feasibility of intersectional candidates’ presidential campaigns as a foundation to coalition building.
I explore the question, to what extent do sexist attitudes influence Harris’s favorability across race and gender? Additionally, how do these attitudes differ upon examining perceptions of Trump? Harris faced sexism and misogynoir throughout her campaign from her political opponents, the media, and the public (Yount and Sharma Reference Yount and Sharma2021). Newspaper coverage included a range of gendered and racial stereotypes about Harris throughout both her presidential bids (Heckman et al., Reference Heckman, Bhargava and Boardman Ndulue2023; Zurbriggen and Vallerga Reference Zurbriggen and Vallerga2022). On social media, the public described Harris as sexually promiscuous and angry (Nee Reference Nee2023), common stereotypes used to shame Black women (Harris-Perry Reference Harris-Perry2011).
Drawing on the 2020 Collaborative Multiracial Post Election Survey (CMPS), I assess the role of sexism in public opinion formation toward Harris and Trump. I find modern sexism is negatively correlated with Harris’ favorability and positively associated with Trump support across race and gender. However, benevolent sexism is associated with increased favorability for Harris among racially diverse men. Contrarily, for White men, benevolent sexism is associated with increased Trump favorability.
I highlight three core contributions this article offers to the field. Firstly, it takes an intersectional approach to understanding public opinion formation toward Harris. Through disaggregating by race and gender, this work reveals racial nuances surrounding sexism and its influence on perceptions of Kamala Harris. A great deal of scholarship on Harris focuses broadly on majority-White samples or specifically on communities of which she is a member. Thus, this article’s second contribution lies in utilizing the CMPS to consider cross-racial dynamics that create distinct perspectives of Harris. As a multiethnic, multiracial woman in a national campaign, Harris’ campaign demands unique attention to racial and ethnic dynamics. People tend to assume Black people in the United States are African American rather than multiethnic (Zou and Cheryan Reference Zou and Cheryan2017) and associate women with Whiteness (Sesko and Biernat Reference Sesko and Biernat2018; Sesko and Biernat Reference Sesko and Biernat2010). Harris’ diverse ethnic identity, strong community ties, and status as a woman defy expectations of group membership in these communities, potentially yielding distinct outcomes in public opinion toward her. Finally, this article invites researchers to question whether women of color candidates can utilize paternalistic attitudes to obtain men’s support among racial minorities. In the following section, I draw on theories of descriptive representation, intersectionality, and sexism to establish a theoretical framework to better account for these findings.
Literature Review
To understand public opinion formation toward the 2024 presidential candidates, I consider distinctions in attitudinal development across the public’s gender and race. Women generally demonstrate more liberal attitudes than their male counterparts (Box-Steffensmeier, Boef, and Lin Reference Box-Steffensmeier, Boef and Lin2004). Women and politics scholars find that the gender gap is largely driven by partisanship, with women more likely to vote for and support Democratic political candidates since the 1960s (Ondercin Reference Ondercin2017). However, recently some argue that women’s racial differences must be considered to understand the partisan gender gap. Upon disaggregating by race, scholars find that the partisan gender gap is a race and gender gap, as the only group of female voters to support Republican candidates in recent elections are White women (Junn and Masuoka Reference Junn and Masuoka2020; Wolf et al. Reference Wolf, Kim, Brisbane and Junn2025). Latinas (Monforti Reference Monforti2017), particularly those with later generational status, were more likely than their male counterparts to support liberal policies and Democratic candidates (Bejarano Reference Bejarano2014). While Asian women are less consistent in these tendencies (Lien Reference Lien1998), they do not consistently support Republican candidates like their White counterparts. Finally, Black American women have the strongest and most consistent relationship with the Democratic party compared to both women of other racial groups (Junn and Masuoka Reference Junn and Masuoka2020) and Black men (Slaughter, Crowder, and Greer Reference Slaughter, Crowder and Greer2024). As racial groups all have different relationships with the Democratic party and their political candidates, we can anticipate distinct reactions to women presidential candidates alongside these dimensions. I now turn to further consider this relationship through theories of intersectionality and descriptive representation.
People with multiple marginalized identities (i.e., women of color) have unique experiences as they simultaneously combat multiple systems of oppression and their intersections (Crenshaw Reference Crenshaw1989, Reference Crenshaw1991). Black feminist theorists find that these positions encourage Black women to create their own perspective on the world (Collins Reference Collins1986; Hill Collins Reference Hill Collins2009), cultivating distinct political ideologies (D. K. King Reference King1988; Simien Reference Simien2006) and engagement (Isoke Reference Isoke2013; Nwakanma Reference Nwakanma2023). Additionally, scholars find similar frameworks among women of color’s political engagements more broadly (García Reference García2020; Mohanty Reference Mohanty2005). These intersectional political perspectives can extend into public opinion formation toward political elites, especially those with similar identities.
The theory of descriptive representation argues people prefer candidates who share their, typically marginalized, identities (Mansbridge Reference Mansbridge1999; Tate Reference Tate2004; Garcia et al. Reference Garcia, Gibson, Mackey, Stout and Tate2025). Women, particularly those who are concerned about women’s interests, are more likely to support female candidates (Rosenthal Reference Rosenthal1995; Simien, Hayes, and Conway Reference Simien, Hayes and Conway2024). They might prefer women candidates, as increased female representation can lead to gains in substantive representation (Mendelberg, Karpowitz, and Goedert Reference Mendelberg, Karpowitz and Goedert2014). Women’s political leadership often leads to positive outcomes for women who become more knowledgeable and care more about their representatives’ policy positions than women represented by men (Jones Reference Jones2014). Additionally, these women demonstrate increased trust in political institutions (High-Pippert and Comer Reference High-Pippert and Comer1998). This positive outlook carries broader implications for women’s lives, with many seeking out new jobs as a means of self-improvement (West and Duell Reference West and Duell2024).
While some scholarship finds supporting a woman candidate bolsters women’s political engagement (Mansbridge Reference Mansbridge1999), recent work finds little change in women’s participation (Wolak 2020) or political attitudes (Dolan Reference Dolan2006; West Reference West2017). One potential explanation for this could be shifting political attitudes across race. While some White women feel a strong sense of gender consciousness (Gurin Reference Gurin1985), others focus more on their racial identity. White people, particularly those who feel a strong sense of White-linked fate, are more likely to prefer White candidates (Jardina Reference Jardina2019; Schildkraut Reference Schildkraut2017). White women face a choice in which identity to prioritize when considering the weight of descriptive representation in political decision-making (Phillips Reference Phillips2018). Alternatively, some women of color, Black women in particular, report both their racial and gender identities as important to them (Simien Reference Simien2006).
The potential for intersectional representation can bolster minorities’ sense of empowerment and desire to engage politically (Bobo and Gilliam Reference Bobo and Gilliam1990; Gleason and Stout Reference Gleason and Stout2014). More specifically, descriptive representation can be invigorating for some women of color, particularly Black women (Stout and Tate Reference Stout and Tate2013). Black women politicians are most likely to be supported by Black women (Philpot and Walton Reference Philpot and Walton2007); however, these are not their sole supporters. Black men describe feeling more represented by Black women than candidates who share neither their race nor gender (Montoya et al. Reference Montoya, Bejarano, Brown and Allen Gershon2021). Feeling represented can influence political behavior, with Black Americans more likely to participate in elections with Black candidates (Whitby Reference Whitby2007).
Scholars find that women across multiple racial groups can respond positively to women of color seeking office. Black women and Latinas tend to demonstrate the most interest in seeing women of color in political office (Matos, Greene, and Sanbonmatsu Reference Matos, Greene and Sanbonmatsu2020; Nguy, Davis, and Chan Reference Nguy, Davis and Chan2025), particularly when framing themselves as women of color, whereas White women are less likely to be interested. The political salience of intersectional feminism is further emphasized as women of color feminists are less likely to support Trump than White feminists (Simien, Hayes, and Conway Reference Simien, Hayes and Conway2024). Meanwhile, Latinos and Black men, specifically those who feel a sense of men of color linked fate, show support for women of color candidates (Matos and Sanbonmatsu Reference Matos and Sanbonmatsu2024). While scholars explored attitudinal development toward Black women, few investigated the public’s response to a biracial and multiethnic Black-presenting woman seeking office. While often discriminated against, these marginalized identities can serve as the foundation for a broader political coalition and open unique tactics to candidates.
Some argue that rather than gender undermining a female candidate’s chances, women seeking office may have an advantage. Women who emphasize their focus on feminine issues experience boosts in their levels of support (Bauer Reference Bauer2020a). Additionally, women office seekers tend to be highly qualified to a greater extent than men (Bauer Reference Bauer2020b). For women of color candidates, some scholars theorize there may be a “double advantage” with their race and gender appealing to broader audiences. Some speculate that as gender weakens the sense of threat associated with racial minorities, White voters demonstrate greater willingness to vote for women of color than their male counterparts (Bejarano Reference Bejarano2013). As voters perceive women of color as experienced and strong leaders, their support increases for these candidates (Gershon and Lavariega Monforti Reference Gershon and Lavariega Monforti2021). Furthermore, though frames of motherhood can harm White women’s perceived electability, such framing can enhance support for Black women (Sweet-Cushman and Bauer Reference Sweet-Cushman and Bauer2024). Harris occasionally leaned into such gendered dynamics, including her Internet series “Cooking with Kamala” (Price Reference Price2024). Harris leveraged this series as a tool emphasizing her gender while shifting her multiracial identity from an exotic racialized threat to a fun curiosity (Parry-Giles et al. Reference Parry-Giles, Aboagye, Choi, Hourigan, Itoh, Robbins, Salzano, Schultz and Sturm2022). More explicit approaches, such as framing herself as a woman of color, lead to no notable shifts in favorability toward Harris among women (Greene, Matos, and Sanbonmatsu Reference Greene, Matos and Sanbonmatsu2022). In the next section, I expand on my theoretical framework to understand the role of sexism in attitudes toward Kamala Harris.
Theoretical Framework
Sexism is crucial for understanding political attitude formation toward women political elites. Women politicians, especially women of color, encounter several structural norms that shape the feasibility of becoming political leaders, including a lack of resources (Sanbonmatsu Reference Sanbonmatsu2015; Shah, Scott, and Gonzalez Juenke Reference Shah, Scott and Gonzalez Juenke2019) and supporters (Paul and Smith Reference Paul and Smith2008). Sexism can undermine support for women candidates while boosting support for their male counterparts even among co-partisans (Utych Reference Utych2021). While some assume women to be women candidates’ largest base of supporters, this is not always the case. Women high in sexism attitudes tend to prefer Republican candidates (Cassese and Barnes Reference Cassese and Barnes2019). Sexism’s influence is unique across partisan lines, weakening Democrat and strengthening Republican political participation (Banda and Cassese Reference Banda and Cassese2021). Given the political salience of sexism attitudes, I anticipate that they will meaningfully shape public perceptions of the 2024 presidential candidates.
Though there are various theories of sexism and means to understand it, this study focuses on modern and benevolent sexism. This study examines the role of modern sexism in public opinion formation toward Trump and Harris, recognizing across the past three elections Trump’s campaigns have espoused modern sexism beliefs. Particularly, the Trump administration’s attacks on DEI, especially in their second administration (Ng et al. Reference Ng, Fitzsimmons, Kulkarni, Ozturk, April, Banerjee and Louise Muhr2025), align Trump with modern sexism’s opposition to gender equity initiatives (Archer and Kam Reference Archer and Kam2020).
While benevolent sexism focuses on reinforcing traditional gender roles, modern sexism emphasizes the justification of dismantling gender equality initiatives. I explore how each elite’s associations with gender result in distinct public opinion outcomes across race and gender. This study also considers the political implications of benevolent sexism, recognizing recent shifts in the discussions of traditional gender roles in the American political environment, particularly through social media. Online communities such as the manosphere and #tradwife content have been on the rise, particularly after the covid-19 pandemic (Elmhirst Reference Elmhirst2024) and recent presidential elections (Dignam and Rohlinger Reference Dignam and Rohlinger2019; Much et al. Reference Much, Rutherford, Greenfield, Tucker and Nagler2025). These communities’ growth is indicative of an increasing public interest in following traditional gender roles (Bower Reference Bower2024), making benevolent sexism an ideal framework through which to explore these dynamics.
Benevolent Sexism
Benevolent sexism is a component derived from the ambivalent sexism inventory (Glick and Fiske Reference Glick and Fiske1997). Ambivalent sexism understands sexism through two underlying concepts: antagonistic (hostile) and paternalistic (benevolent) attitudes toward women. Benevolent sexism emphasizes positive attitudes towards women to justify male dominance in gender roles. While women consistently demonstrate negative attitudes toward hostile sexism, their reactions to benevolent sexism are nuanced (Barreto and Ellemers Reference Barreto and Ellemers2005), in some cases demonstrating favorable attitudes (Fischer Reference Fischer2006). These attitudes are more appealing to women, reinforcing their adherence to traditional gender roles, with some demonstrating increased life satisfaction following such roles (Connelly and Heesacker Reference Connelly and Heesacker2012).
Scholars find types of sexism vary in the outcomes they shape and those affected by them. Upon examining tendencies to hold benevolently sexist attitudes, we uncover distinctions across race and gender. Latinos across gender identity were more likely to perceive benevolent sexism positively (Herrera Hernandez and Oswald Reference Herrera Hernandez and Oswald2024). Similarly, Black women and men have higher levels of benevolent sexism than their White counterparts (Davis, Settles, and Jones Reference Davis, Settles and Jones2022). Some scholars argue people tend to feel more benevolent sexism towards White women rather than Black women (McMahon and Kahn Reference McMahon and Barsamian Kahn2016); however, these studies tend to examine majority-White samples. Expanding this scope finds Black participants demonstrate higher levels of benevolent sexism when thinking of Black women (Campbell, Hudson, and Ratliff Reference Campbell, Hudson and Ratliff2023). Additionally, scholars find that racial minorities tend to be more supportive than White men of paternalistic behaviors when they perceive a legitimate threat to women (Davis, Settles, and Jones Reference Davis, Settles and Jones2022).
This carries political consequences, as paternalistic ideologies can encourage men to act against sexism (Good, Sanchez, and Moss-Racusin Reference Good, Sanchez and Moss-Racusin2018), which could extend into supporting a woman politician. Research on the relationship between benevolent sexism and public opinion of women candidates is somewhat mixed. Some scholars argue benevolent sexism is positively associated with support for female candidates, specifically White women candidates rather than their male counterparts (Spencer Reference Spencer2021). Though it seems most participants support Republican candidates across gender identity. This relationship can be particularly influential in the perceived electability for White women candidates (Britzman and Mehić-Parker Reference Britzman and Mehić-Parker2023). While this is advantageous for White women politicians, we know less about if benevolent sexism is helpful for women of color political elites.
Other scholars, particularly those focused on the 2016 election, find a negative association between benevolent sexism and support for women politicians. Winter finds in the 2016 election that benevolent sexism predicted opposition to Clinton (2022). This opposition remains in place despite women high in benevolent sexism perceiving Clinton as more competent (Cassidy and Krendl 2019). Notably, Trump voters demonstrate higher levels of benevolent sexism attitudes than Americans who voted for Clinton (Ratliff et al. 2017). Despite holding some positive attitudes towards women, it may be the case that these benevolent sexist attitudes drive behavior that discourages women’s political ambitions. This research investigates how participants’ racial identity may account for these contrasting findings in benevolent sexism’s role in public opinion formation of women candidates.
Generally, men’s benevolent sexism tends to be reserved for women of their same racial group (Glick and Raberg Reference Glick and Raberg2018). Given this tendency, we might anticipate that White men may feel less favorable to a woman of color political elite. Additionally, people high in benevolent sexism prefer candidates with masculine leadership styles regardless of the candidate’s actual gender identity (Winter Reference Winter2020). This may benefit Black women politicians, as society frequently strips Black women of their femininity, associating Blackness with masculinity (Goff et al., Reference Goff, Thomas and Christian Jackson2008; Sesko and Biernat Reference Sesko and Biernat2018). Therein, we may anticipate that Black men high in benevolent sexism will positively regard Harris as a political candidate. Additionally, her multiracial identity may incline men of color more broadly to regard her favorably.
Drawing on this scholarship, it seems men of color demonstrate stronger benevolent sexism attitudes than White men. Such attitudes can encourage men of color to act when women are facing a threat. Politically, this may present as these men feeling more supportive of women politicians. Given that these attitudes are intensified by women of the same race, we might anticipate Black and Asian men to be particularly supportive of Harris. However, as a multiracial political candidate, Harris may be able to push beyond the bounds of her racial groups to garner support from benevolently sexist men of color more broadly. From the above scholarship I develop the study’s first hypothesis,
H1: Men of color with high levels of benevolent sexism will be more likely to favor Harris than their women counterparts.
As I compare these attitudes between Harris and Trump, I also consider the role of benevolent sexism in Trump evaluations. Research focused on the 2016 election initially argued benevolent sexism tended to be unrelated to favorability of Trump (Glick Reference Glick2019). More recent scholarship finds that benevolent sexism can improve evaluations of Trump, with people high in benevolent sexism less likely to feel negative emotions toward Trump (Winter Reference Winter2023). This shift may be attributed to Trump’s increased benevolent sexism messaging on social media (Scotto di Carlo Reference Scotto di Carlo2021). These messages would be most effective for men who are more likely to be following Trump on social media (Wang, Luo, and Zhang Reference Wang, Luo, Zhang, Ciampaglia, Mashhadi and Yasseri2017). Additionally, people with high levels of sexist attitudes are more likely to favor White male candidates over women of color (Nelson 2021). However, these studies rarely disentangle the role of audience race and gender, allowing for little exploration of such dynamics. Upon considering the nuances of race and gender, we may anticipate benevolent sexism’s influence on Trump attitudes to be distinct. While benevolent sexism can drive protective attitudes in men, these attitudes tend to be limited to women of their own racial group. Subsequently, I would expect that any positive attitudes that benevolent sexism might offer Harris would be constrained by her racial identity. Alternatively, Trump’s benevolently sexist messaging might appeal to White men, leading me to the second hypothesis.
H2: White men high in benevolent sexism will be more likely to favor Trump than Harris.
Modern Sexism
I also examine the influence of modern sexism on Americans’ perceptions of former Vice President Harris across race and gender. Unlike benevolent sexism, women do not demonstrate a tendency to perceive modern sexists or such attitudes favorably (Swim et al. Reference Swim, Aikin, Hall and Hunter1995). Modern sexism is a construct that emphasizes subtle sexism built into social structures and norms (Swim and Cohen Reference Swim and Cohen1997). Rather than its more overt alternatives, modern sexism captures a refusal to perceive sexism as a contemporary issue. As attitudes of old-fashioned sexism (e.g., opposition to women working) declined, people began expressing more resentment toward gender equity policies (Swim et al. Reference Swim, Aikin, Hall and Hunter1995). Modern sexism is distinct from benevolent sexism’s focus on seemingly positive traits of women within traditional gender norms, though these beliefs are ultimately paternalistic. Modern sexism instead tends to consider the matter of gender inequality resolved with the victories of the second-wave feminism movement, specifically increased access to employment and education.
Typically, men are more likely to endorse modern sexism beliefs than women (Swim et al. Reference Swim, Aikin, Hall and Hunter1995). However, this finding is inconsistent across racial groups, as women of color’s endorsement of modern sexism beliefs is like their male counterparts’. Broadly, racial minorities’ modern sexism attitudes tend to be lower than Whites (Hayes and Swim Reference Hayes and Swim2013), with African Americans the least likely to express these beliefs. Modern sexist attitudes are associated with more favorable perceptions of men than women in society (Swim and Cohen Reference Swim and Cohen1997); subsequently, this preference is likely present in political opinions as well.
Across multiple presidential elections, support for Republican candidates is correlated with modern sexism, most prominently in the 2016 election (Valentino, Wayne, and Oceno Reference Valentino, Wayne and Oceno2018). Regarding politicians’ gender identity, modern sexism is negatively correlated with attitudes toward women in politics (Ditonto Reference Ditonto2019; Swim et al. Reference Swim, Aikin, Hall and Hunter1995). During the 2016 election, modern sexism was correlated with positive attitudes toward Trump and negative perceptions of Clinton (Godbole, Malvar, and Valian Reference Godbole, Malvar and Valian2019; Knuckey Reference Knuckey2019; Long, Dawe, and Suhay Reference Long, Dawe and Suhay2022). Endorsement of modern sexism influenced White men’s and women’s vote choices similarly in 2016, with increased likelihood to vote for Trump (Hanley Reference Hanley2021). Given the role of modern sexism in boosting support for Trump, it’s reasonable to anticipate a similar response among White respondents in this study. Additionally, Latinos, Asian Americans, and Black Americans high in modern sexism are more likely to vote for Trump (Geiger and Reny Reference Geiger and Reny2024; Hickel Jr. and Deckman Reference Hickel and Deckman2022). Rarely has such scholarship taken an intersectional approach to presidential candidate attitudes. This project assesses whether modern sexism will similarly weaken support for Harris and strengthen Trump support among women of color relative to White women.
The little research that explores the effect of modern sexism on non-White women candidates finds that modern sexism is negatively correlated with evaluations of Black women elites (Knuckey and Mathews Reference Knuckey and Mathews2024), particularly when interacting with racial resentment attitudes (Carr Reference Carr2025). White women high in hostile sexism attitudes were more likely to vote for Trump than their women of color counterparts (Frasure-Yokley Reference Frasure-Yokley2018; Junn Reference Junn2017; Merolla Reference Merolla2018). However, few scholars have explored the influence of modern sexism specifically on women’s attitudes across racial groups. I anticipate that White women’s relationship to modern sexism and Harris evaluations will be distinct from women of color. Women of color are more likely to be supportive of racially diverse women seeking office and in positions of political leadership (Matos, Greene, and Sanbonmatsu Reference Matos, Greene and Sanbonmatsu2020; Philpot and Walton Reference Philpot and Walton2007). I anticipate this to be especially true for Black and Asian women, as Harris is a member of both their racial groups. I further suspect that holding attitudes of sexism will intensify tendencies of White women to perceive women of color elites poorly. This leads to my third hypothesis,
H3: White women with high levels of modern sexism will be less likely to favor Kamala Harris than their Black or Asian counterparts.
Methods and Data
To test these hypotheses, I utilized the 2020 CMPS to assess the influence of these types of sexism on attitudes toward Kamala Harris and Donald Trump. The CMPS is distinct in offering substantive samples of racial minority groups, allowing for cross-group comparisons between gender and racial groups. The CMPS was fielded from April 2021 to August 2021, online, and is weighted to the 2019 American Communities Survey for gender, ethnicity, age, nativity, and education (Frasure-Yokley et al. Reference Frasure-Yokley, Wong, Vargas and Barreto2020). I subset this analysis by race and gender, with the women sample consisting of White women (n = 1789), Black women (n = 2079), Asian American women (n = 1703), and Latinas (n = 1795). The sample of men includes White men (n = 1,508), Black men (n = 1,647), Asian American men (1,420), and Latinos (1,226).
The CMPS is an ideal data set given its large sample sizes, multiple measures of sexism, and its function as one of the few data sets that assesses attitudes toward multiple political elites, including both former Vice President Harris and President Trump. The key dependent variable of this study is a favorability measure requesting participants indicate their overall opinion of several political figures. The response options range from very favorable to very unfavorable, with the option for having no opinion. I code these responses on a scale ranging from 0 (very unfavorable) to 1 (very favorable) in intervals of .25.
The independent variables of interest are benevolent and modern sexism. Benevolent sexism is part of the eight-item ambivalent sexism inventory (Glick & Fiske, Reference Glick and Fiske1997), with four items measuring benevolent sexism. Respondents were asked how strongly they agreed or disagreed with multiple statements assessing their attitudes toward women. The 2020 CMPS includes the four popularized items (see Appendix), allowing me to utilize the validated measure in this study. I combined these four items into a scale with a Cronbach’s alpha score of .67; the final variable ranges from 0 to 1 in .25 intervals. A score of 0 demonstrates low benevolent sexism, increasing to 1, which indicates highly benevolent sexist attitudes. Tables of benevolent sexism’s weighted mean scores by race and gender are included in the appendix.
The measure of modern sexism is a proxy for the original scale, including three items, with response options ranging from strongly agree to strongly disagree. The full modern sexism scale includes eight items, with most political science researchers focusing on four. While this study includes one of the original items, “Discrimination against women is no longer a problem in the U.S.,” the other two have been modified. They focus on gender equality in politics specifically rather than in society at large. The specific wording of these items can be found in the Appendix. These items combine into a variable I utilize as a proxy for modern sexism with a Cronbach’s alpha of .58. Tables of the weighted mean scores of modern sexism by race and gender are included in the appendix.
Upon operationalizing these variables, I control for standard social demographics, age, education, income, marital status, nativity, and ethnicity. I also account for standard political attitude covariates, including partisan identification and political ideology. Beyond these demographics, I also control for racial group linked fate, originally conceptualized around Black Americans’ political behavior (Dawson Reference Dawson1994), with scholars finding racial group linked fate can influence candidate favorability beyond the Black community (Campi and Junn Reference Campi and Junn2019; Jardina Reference Jardina2019; Masuoka, Ramanathan, and Junn Reference Masuoka, Ramanathan and Junn2019; Sanchez Reference Sanchez2006). Gendered linked fate can shape political attitudes, particularly among Latinas and White women (Stout, Kretschmer, and Ruppanner Reference Stout, Kretschmer and Ruppanner2017). Similarly, intersectional linked fate can shape public evaluations of political candidates and the feasibility of cross-racial candidate support (Bejarano et al. Reference Bejarano, Brown, Gershon and Montoya2021; Gershon et al. Reference Gershon, Montoya, Bejarano and Brown2019; Matos and Sanbonmatsu Reference Matos and Sanbonmatsu2024). Herein, I control for a sense of linked fate with men or women of color for their respective genders.
Both marital status and, subsequently, one’s partner’s politics are included in the model, as they can shape public opinion formation and political engagement (García-Castañon Reference García-Castañon2018; Stout, Kretschmer, and Ruppanner Reference Stout, Kretschmer and Ruppanner2017). Racial resentment can shape attitudes towards candidates across race, gender, and partisanship (Lucas and Silber Mohamed 2021). Racially resentful attitudes are correlated with decreased support for Black candidates (Hale Reference Hale2020; Stout Reference Stout2015) and boosts for White candidates (Knuckey Reference Knuckey2011; Hooghe and Dassonneville Reference Hooghe and Dassonneville2018). Thus, motivating my inclusion of racial resentment as a control across all models.
Given Trump’s success in appealing to White evangelical communities, I account for religious attitudes as well (Margolis Reference Margolis2020; Whitehead, Perry, and Baker Reference Whitehead, Perry and Baker2018). For Harris, I consider the potential influence of her Indian and Jamaican identities alongside research on Indian American Harris attitudes (Lemi, Arora, and Sadhwani Reference Lemi, Arora and Sadhwani2022). I control for such ethnic identities among Asian and Black respondents, respectively. In this analysis, I conduct an OLS linear regression to analyze the relationship between sexism attitudes and favorability of 2024 presidential candidates. Afterwards I tested the models for multicollinearity, examining their variance inflation factors, all having a score no higher than 1.8, indicating little concern regarding correlations within the models.
Results
In this section, I explore determinants in the favorability of Harris and Trump across gender and racial groups, discussing each group’s results in relation to the hypotheses. Below, Figure 1’s predicted probability graph demonstrates the findings regarding benevolent sexism and Harris’s favorability among men with confidence intervals at 95%.
Predicted Probabilities of Attitudes Toward Harris—Benevolent Sexism.
Black Men’s Harris Attitudes
Briefly, examining Harris’s attitudes in general, Black men’s Harris favorability lies around 76%, the highest among men across race. I find benevolent sexism is positively correlated with Black men’s attitudes toward Harris (.22, p < .001), supporting H1. Moving benevolent sexism attitudes from their lowest to their highest is associated with a 33-percentage-point increase in favoring Harris among Black men. As Figure 1 demonstrates, this is the largest shift in attitudes among men across benevolent sexism attitudes.
I find that Black men high in modern sexism are less likely to perceive Harris favorably.
Notable covariates include Democrat partisanship, liberal ideology, age, and religious centrality being positively correlated with Black men’s perception of Harris. Additionally, gender-linked fate and racial resentment are negatively correlated with Black men’s perceptions of Harris.
Asian American Men’s Harris Attitudes
Asian American men’s mean Harris favorability is at 61%. In examining the role of benevolent sexism in Asian men’s evaluations of Harris, I find Asian men high in benevolent sexism are more likely to favor Harris (.11, p < .05), supporting H1. As benevolent sexism attitudes increase, Asian American men demonstrate a 16-percentage-point increase in Harris’s favorability. I find modern sexism is negatively correlated with Asian American men’s perception of Harris (-.31, p < .001). Looking at the covariate’s liberal ideology, Democrat partisanship, and partner’s Democratic partisanship, are all positively correlated with favoring Harris. Education is negatively associated with Harris’s favorability among Asian men.
Latinos Harris Attitudes
Latinos’ Harris favorability averages at 65%. Latino men high in benevolent sexism feel less favorably toward Harris than those lower in benevolent sexism (.15, p < .001), supporting H1. As benevolent sexism attitudes increase, there is a 26-percentage-point increase in favorable Harris attitudes among Latinos. In examining the role of modern sexism, I find Latinos high in modern sexism perceive Harris more negatively (-.21, p < .01). Examining relevant covariates, I observe liberal ideology, Democrat partisanship, and age are positively associated with Harris attitudes. Meanwhile, racial resentment is negatively associated with favorable Harris attitudes.
White Men’s Harris Attitudes
Finally, White men’s Harris favorability is at 41%, the lowest of all groups. Benevolent and modern sexism are statistically insignificant in White men’s attitude formation toward Harris. This is in alignment with my expectations in H2, as Figure 2 demonstrates below. Covariates positively associated with Harris favorability include partner Democrat partisanship, Democrat partisanship, and liberal ideology. Racial resentment is negatively correlated with Harris’s favorability among White men.
Predicted Probabilities of Attitudes Toward Harris & Trump—White Men.
White Men’s Trump Attitudes
White men’s Trump favorability is at 49%, the highest of men considered in this analysis. White men with a strong sense of benevolent sexism are more likely to perceive Trump favorably (.15, p < .01). Moving benevolent sexism attitudes from low to the highest, there is a 31-percentage-point increase in Trump favorability among White men. While the direction of the relationship is similar for both elites, the steeper correlation line and narrower confidence bands in Figure 2 demonstrate the statistical significance of this result for Trump favorability rather than Harris. This supports H2, wherein I posit that White men high in benevolent sexism will be more likely to support Trump, a significant relationship, than Harris. While benevolent sexism is a predictor of Trump support, it is not a determinant of Harris support among White men.
Similarly, the endorsement of modern sexism is a significant and positive variable in attitude formation toward Trump. White men who strongly endorse modern sexist beliefs are more likely to feel favorably toward Trump (.38, p < .001). Briefly considering additional covariates, people high in education, age, White linked fate, liberal ideology, and Democrat partisanship, and men with Democrat partners are negatively associated with Trump support. Positive covariates include racial resentment, religious centrality, high income, and mixed-race identity. I now turn the focus towards testing H3 and estimating the effect of modern sexism on women’s perception of Harris shown in Figure 3 below.
Predicted Probabilities of Attitudes Toward Harris—Modern Sexism.
Black Women’s Harris Attitudes
I note Black women generally hold favorable attitudes toward Harris, with a weighted mean of 76% Harris favorability. Covariates that are significant and positive include Democratic partisanship, liberal ideology, partner’s partisanship, age, and religious centrality. Turning my focus toward the role of sexist attitudes in Harris’s evaluations, these results elucidate which types of sexism are politically influential for Black women.
Benevolent sexism positively correlates with perceptions of Harris, though this effect is statistically insignificant, offering support for H1. Black women who endorse modern sexism are significantly less likely to favor Harris (−. 15, p < .01), in contrast to my expectations for H3. As modern sexism attitudes increase, favorability for Harris decreases 25 percentage points among Black women. I further consider the reason for this outcome in the conclusion.
Asian American Women’s Harris Attitudes
Asian American women generally hold positive Harris attitudes, with a weighted mean of Harris favorability at 61%. Notable covariates include Democrat partisanship, liberal ideology, and a partner’s democratic identification as positively associated with Harris’s favorability. Racial resentment is negatively associated with favoring Harris for Asian American women. Testing my hypotheses, I find both benevolent and modern sexism attitudes are not determinants for Asian American women’s assessments of Harris. Benevolent leans positively with Harris’s favorability, though it is insignificant in shaping Harris’s evaluations supporting H1. Modern sexism is negatively correlated and statistically insignificant for Asian American women’s Harris favorability, offering partial support for H3. I explore this outcome more thoroughly in the conclusion section.
Latinas Harris Attitudes
Latinas’ Harris favorability lies at 65%. Across the controls, Democrat partisanship, liberal ideology, partner’s partisanship, religious centrality, and age are positively associated with Harris favorability. Racial resentment and nativity are negatively correlated with attitudes toward Harris. Endorsement of modern sexism beliefs is negatively correlated with favorability for Harris (−.26, p < .001). As modern sexism attitudes increased, Latinas’ support for Harris decreased by 38 percentage points. In testing my first hypothesis, I find that there is no significant effect on favorability toward Harris for Latinas high in benevolent sexism, unlike their male counterparts, supporting H1.
White Women’s Harris Attitudes
White women, similar to their male counterparts, have the lowest Harris favorability, with a mean score of 41%. Herein I focus my discussion of results for the White women sample on the relationship between modern sexism and Harris favorability. This offers some support for H3, as White women high in modern sexism were significantly less likely to favor Harris (−.17, p < .01). As modern sexism beliefs increased, Harris’s favorability decreased by 51 percentage points. Additionally, benevolent sexism is statistically insignificant in White women’s attitude formation toward Harris. Briefly examining the covariates, I observe Democratic partisanship, liberal ideology, a Democrat partner, age, and multiracial identity are positively associated with Harris attitudes. Racial resentment is negatively associated with favorable attitudes for Harris.
Conclusion
This project clarifies the role of sexism in shaping evaluations toward former Vice President Harris and compares its effect relative to President Trump’s attitudes. I find that benevolent sexism is positively associated with Harris’ favorability for Latinos, Black, and Asian American men. As expected, this relationship does not hold for women across racial groups. Furthermore, benevolent sexism is a significant determinant factor solely in Trump favorability among White men, supporting my hypotheses. White women high in modern sexism are significantly less likely to demonstrate favorable Harris attitudes. However, contrary to expectations, Black women high in modern sexism are significantly less likely to favor Harris, unlike Asian women.
Though I anticipated Black and Asian women to be distinct from their White counterparts, modern sexism is a predictor of negative Harris attitudes for women across racial groups, excluding Asian women. While I theorized Asian and Black women’s attitudes would primarily be driven by Harris evoking a sense of descriptive representation, such shared identification does not undermine the influence of modern sexism among Black women. A great deal of the literature on modern sexism and attitudes toward women candidates focuses on opinion through vote choice (Ditonto Reference Ditonto2019; Spencer Reference Spencer2021). Utilizing vote choice as the dependent variable, scholars’ findings are more in alignment with my expectations, with Black women’s support for women and Democratic candidates among the most consistent (Kim and Junn Reference Kim and Junn2024) and sexism having no notable effect. These attitudes are driven in part by the longstanding relationship between Black people, women in particular (Slaughter, Crowder, and Greer Reference Slaughter, Crowder and Greer2024), and the Democratic party (Frymer Reference Frymer2010; Wamble et al. Reference Wamble, Laird, McConnaughy and White2022; White and Laird Reference White and Laird2020). Though Black women’s voting behavior is restricted, their favorability of candidates is more open-ended, with Harris facing criticism in the Black community throughout her 2020 presidential primary election (M. King Reference King2019). The effectiveness of modern sexism may be indicative of Harris’ struggles to overcome these critiques among Black women, which she would improve in her 2024 campaign (PBS News Desk 2024).
While these findings are fascinating, it’s crucial to consider improvements for researchers to explore in future work. One such improvement is the dependent variable assessing attitudes toward Harris and Trump. The 2020 CMPS includes a battery of items assessing favorability of prominent political elites. Favorability is an often-used measure in political science; however, additional measures may offer a broader understanding of attitudes toward Harris. Future work may incorporate additional measures such as competence and trustworthiness, often utilized in assessing gender stereotypes of political leaders.
Additionally, the dataset I utilize to examine these dynamics was collected in 2021 prior to Harris’ recent presidential campaign. While more recent data could offer insightful reflections on the outcome of the 2024 presidential election, this data provides a foundation for understanding Harris’ perceptions and the role of sexism in attitude formation toward her prior to Harris’ recent efforts to manipulate such perceptions. Scholars might utilize post-2024 election data to explore new forms of analysis, such as examining how the 2024 campaign shifted attitudes toward Harris and considering the role of the campaign’s late start. For instance, 2024 data would facilitate analysis that more directly assesses the difference in perceptions between Trump and Harris rather than examining these attitudes independent of one another. Such analysis would favor 2024 data, as at the time of data collection, the American public were thinking of each figure in relation to one another rather than focusing on their attitudes toward Trump and Biden as in the 2020 election. Additionally, though here we observe benevolent sexism having a positive effect on perceptions of Harris, preliminary data seems to imply a shift among men of color toward voting for Trump (Gamboa et al. Reference Gamboa, Acevedo, Silva, Sesin and Hampton2024; Geiger and Reny Reference Geiger and Reny2024; Robertson and Tesler Reference Robertson and Tesler2024; Yam Reference Yam2024). More recent data further confirming this relationship may bring to light additional weaknesses in the Harris-Walz campaign for future women of color candidates to be mindful of.
This research highlights the different struggles women of color candidates and their white counterparts face in seeking electoral office. A great deal of the scholarship on sexism and women candidates focuses on White candidates or majority-White respondents. This study leverages the substantial sample sizes to capture nuances and distinctions of women’s attitude formation across racial groups. This analysis expands our understanding of impediments to women of color candidates fostering cross-racial support among women. Additionally, it emphasizes the complex balancing act women of color candidates are expected to manage; appealing to men’s paternalistic attitudes while assuring women of their ability to represent their interests can often push these women into contradictory behaviors. This study offers crucial insights into understanding how women politicians with multiple marginalized identities might appeal to their constituents and factors that may shape their evaluations.
This research carries crucial substantive implications for American politics. As women become more frequent candidates for high offices in American politics, this work offers insights for women candidates to consider which communities are most swayed by sexist attitudes and develop strategies to adapt. Given the distinct role of sexism across race and gender, these findings invite politicians to consider that rather than sexism weakening men of color’s Harris support, there may be deeper issues with the Democratic party’s relationship with these communities than the gender of their candidates. Though benevolent sexism is positively correlated with men of color’s Harris attitudes, the Harris campaign typically avoided emphasizing her gender. Additionally, as increasingly popular online communities seek to persuade men and women to adhere to traditional gender roles, politicians need nuanced approaches to gender in such social contexts. Given these platforms’ popularity, some politicians seek to use such appeals to draw in voters (Much et al. Reference Much, Rutherford, Greenfield, Tucker and Nagler2025); however, this research cautions political actors to be mindful of the differing implications of gender roles across race, particularly among men.
As the United States barrels forward toward becoming a majority-minority nation with a growing population of biracial and multiracial electorate, it’s crucial to understand how the public evaluates political elites of diverse backgrounds. This project considers one piece of the puzzle that is multiracial candidates, women especially, developing their political strategies. Given the growing racial diversity and political ambition among the marginalized in the United States, understanding public opinion formation of candidates across their ethnic, racial, and gender diversity is essential. Through studying public opinion across race and gender of candidates distinct from many politicians, we can uncover new insights into understanding American politics.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/rep.2025.10036.
Competing interests
The author declares none.
Open access
