Gender Stereotypes and Presidential Candidate Evaluations: Evidence from the 2024 New Hampshire Primary Election

There is a significant gender gap among Republican candidates, with far fewer Republican women elected officials as compared to men. Do gender stereotypes play a role in this process? More specifically, to what extent does the double bind, or the perceived tradeoff between likeability and competence for women candidates, continue to influence Republican voters’ evaluations of presidential candidates? We explore this question, focusing on the ways in which Republican voters in an early presidential primary state evaluated a diverse range of potential candidates for the 2024 election, including white men, white women, and women of color.

Our key and consistent finding is that the double bind persists for Republican women presidential candidates. Compared to their male counterparts, Republican women candidates are generally more likely to be seen as likeable than as competent or as a potential commander in chief. Moreover, Republican women have a larger gap between their likeability rating and their competence rating, and this gap grows even larger when comparing likeability to ratings as potential commander in chief. 

Research Motivation and Questions

The so-called double-bind faced by women candidates (Jamieson 1995) reflects the tradeoff between a candidate’s likeability and their perceived competence or electoral strength. In other words, when women candidates are perceived as more likeable, they are often portrayed as less competent, and vice versa. This is consistent with the Stereotype Content Model (SCM), which maps traits onto two dimensions, competence and warmth (Cuddy et al. 2009).

This tradeoff is thought to be strongest when conservative voters evaluate women candidates, at least in part because of the ways in which stereotypes about femininity may clash with more masculine stereotypes around politicians (Schneider and Bos 2014; Bauer 2020). The disconnect may be felt most strongly among conservative voters, given norms around traditional gender roles. In recent research, Ditonto and Andersen (2025) found that women candidates are rated as equally likeable compared to men candidates, though Democrats tend to rate women candidates as more likeable, while Republicans rate women candidates as less likeable.

In our study, we were interested in understanding whether Republican women presidential candidates are disproportionately evaluated through the lens of the double bind. Would likeable women candidates also be seen as strong leaders? Or, would women candidates continue to have a stronger “likeability penalty,” as compared to Republican men? 

Our focus on presidential politics pushed us to consider how masculine expectations relate to perceptions of candidate competence and likability. While competence is associated with “expertise masculinity,” the presidency also uniquely taps into ideas about “dominance masculinity” (Duerst-Lahti 2006), particularly through the role of the head of the military (commander in chief). In our analysis, we test whether likeability offsets perceptions of either of these masculine characteristics, especially among women candidates.

Research Method 

In March 2023, we worked in collaboration with the New Hampshire Institute of Politics (NHIOP) to field two original surveys designed to evaluate Republican attitudes about potential candidates early in New Hampshire’s first-in-the-nation presidential primary. Respondents were selected from NHIOP’s sample of likely Republican primary voters and matched registered voter patterns on key demographic variables. Our first survey had a sample size of 499, while the second survey had a sample size of 814.

Our research design allowed us to move beyond studies of hypothetical candidates. Instead, we took advantage of a primary election with a diverse potential candidate pool to look for broad trends in the ways in which gender stereotypes might impact evaluations of real-world political figures. This approach allows us to avoid some of the questions that arise with experimental research around external validity, or whether hypothetical results would translate into real-world settings with actual candidates. Voters might have more information about candidates that shape their evaluations than they would in experimental treatments, which can reduce voter reliance on identity-based stereotypes. However, because we conducted our surveys early in the presidential primary process, we were able to capture attitudes about candidates at a time when fewer opinions about their individual candidacies were likely to have been formed, and thus stereotypes around race and gender were likely to be most impactful (Crowder-Meyer et al. 2020). 

President Trump announced his candidacy for re-election on November 15, 2022. At the time we conducted our surveys, Trump remained the Republican frontrunner, but the primary was not uncontested. On February 14, 2023, Nikki Haley announced her primary candidacy.1 Though he had not yet announced his campaign at the time of our survey, Florida Governor Ron DeSantis was viewed by many as the strongest alternative to Trump. Because the candidate field remained relatively open at the time of our surveys, we also included the names of several other high-profile Republican figures (some of whom were being discussed as potential candidates, and others who were not) to better capture diversity within the Republican Party.

In both surveys, we asked respondents their opinions about a range of potential Republican presidential primary candidates. Specifically, we asked how they would rank candidates on characteristics including “competent,” “likeable,” and “would be a strong commander in chief.” Responses ranged from 1 (extremely well) to 5 (not well at all). 

We included candidates with whom respondents were likely to be familiar to varying degrees. Of course, we cannot discount the possibility that prior knowledge of and perceptions about specific candidates might influence candidate evaluations. However, by looking at evaluations of multiple candidates, we are able to highlight gendered patterns that are consistent across candidates and surveys. 

Study 1: Persistence of the Double Bind

In our first survey, we selected potential candidates to include through a process of matching prominent Republican women/men with similar professional backgrounds:

• National security leaders: former South Carolina Governor and U.N. Ambassador Nikki Haley and former Secretary of State/former Director of the CIA Mike Pompeo;
• Former Republican governors: Kristi Noem (South Dakota) and Asa Hutchinson (Arkansas);
• Executive leadership roles: Elaine Chao is both former secretary of transportation and former secretary of labor, while Mick Mulvaney’s roles include former White House chief of staff and former director of the Office of Management and Budget.

This lineup of candidates included three white men, one white woman, and two Asian women (Nikki Haley is Indian American while Elaine Chao is Taiwanese American).2 We did not ask respondents to similarly evaluate President Trump, who is often perceived as a hypermasculine candidate (Boatright and Sperling 2019). For our first survey, we provided names and photos. For Haley and Pompeo, we mentioned their specific leadership roles; for Noem and Hutchinson, we shared the state for which they were Governor; and for Chao and Mulvaney, we mentioned that both were former Trump Cabinet secretaries.3

Our results from the first survey are presented in Figure 1. Consistent with gender stereotypes, we found that in general, looking at each individual candidate's ranking across the three characteristics, the potential women candidates were rated more highly on likeability as compared to the masculine characteristics. On the other hand, men scored more highly on competence and on being a strong commander in chief. We note two small exceptions to these results. Our third woman candidate, Elaine Chao, ranked slightly higher on competence (18%) than likeability (17%). We are reluctant to overgeneralize based on such a small distinction from only one candidate. However, we note that being viewed as more competent than likeable seems to be consistent with stereotypes around Asian American women (Rosette et al. 2016), and scholars believe such stereotypes might be more often applied toward individuals from East Asia (for instance, China, Korea, Japan, or Taiwan) than from South or South East Asia (Goh and McCue 2021). Based on this research, if voter evaluations of potential candidates are driven at least in part by both gender and racial/ethnic stereotypes, it is not surprising that we might see different patterns for Chao as compared to Noem or even Haley.

Among the men, both Pompeo and Mulvaney scored higher on competence than likeability, whereas Hutchison scored slightly higher (22%) on likeability than on competence (21%). Notably, Hutchison positioned himself as a prominent Trump critic in this contest. Given that Trump is typically viewed as less likeable compared to other candidates, among respondents who were familiar with Hutchison’s positions, it is perhaps less surprising that Trump’s opponent might be evaluated inversely.

As noted above, because we are seeking to understand attitudes about the presidency, we believe it is important to also capture evaluations for the role of commander in chief. This measure taps into perceptions of dominance masculinity instead of expertise masculinity, which is measured in evaluations of competence. One striking finding from our survey is the gendered pattern in the differing evaluations of candidates for competence and commander in chief. Across all cases, the drop in scores between these categories was much greater for women (between 8-12 percentage points) than for men (between 2-5 points). This distinction suggests that voters are much more likely to attribute expertise masculinity to women than they are to attribute dominance masculinity. 

Looking specifically at the examples of Nikki Haley and Mike Pomepeo further illustrates this point. Both candidates have similar backgrounds in national security. Yet, while there was an 11 percentage point difference in evaluations of Haley as competent vs. commander in chief, that number was just 5 percentage points for Pompeo. But, Pompeo’s rating on competence was 12 percentage points higher than his likeability.

Figure 1.  Overall Candidate Ratings on Competency, Likeability, and Commander in Chief (Percent Responding Extremely/Very; Study 1)

Study 2: Confirming the Double Bind

Our second survey, conducted in the same month as the first (March 2023), asked another group of Republican respondents about their attitudes toward a slightly different pool of candidates. For the men, we included Pompeo and added both Donald Trump (then the front runner) and Florida Governor Ron DeSantis (often seen as an alternative to Trump). For the women, instead of Elaine Chao, we added former New Mexico Governor Susana Martinez, who was the first Latina governor in the U.S. (2011-2018) and also served as the head of the Republican Governors Association. Given that Martinez is another prominent Republican woman of color, we were interested to see whether respondents would evaluate her candidacy distinctly. In this survey, respondents were only provided with the names of the candidates and were given no further information.

Figure 2 replicates the analysis from our first study, looking at ratings on likeability, competency, and commander in chief for all candidates evaluated in the second survey. Notably, respondents in this survey rated the candidates somewhat more positively than in Study 1. Across all candidates in Study 2, the gendered patterns appear even more definitive: looking at the evaluations for each candidate separately, all of the women scored highest on likeability as compared to the two masculine characteristics, while all of the men scored lowest on likeability compared to competence and commander in chief.

Each candidate’s individual ratings on competency and commander in chief are somewhat similar, with candidates generally scoring higher on competence. Interestingly, despite her strong national security credentials, this gap was largest for Haley, whose score for competence was seven percentage points higher than for commander in chief. The gap was smallest for DeSantis and Martinez (3 percentage points each), though likely for different reasons; DeSantis was the most popular of the candidates, with a large majority of our respondents ranking him very favorably across all categories. On the other hand, Martinez had much lower name recognition (just 5 percent, compared to over 85 percent for Trump, DeSantis, Haley, and Pompeo). Likewise, her rankings for each characteristic were significantly lower in comparison to candidates with whom our sample was more familiar. Given her low name recognition, it is also possible that Martinez’ name activated separate stereotypes associated with being a Latina candidate (Cargile 2023). Unfortunately, our data does not allow us to determine how much of her low evaluations were the result of stereotypes versus low name recognition.

The most striking exception to the pattern is former President Trump, who scored higher on commander in chief than overall competency. Given Trump’s particularly masculine and militaristic language (Boatright and Sperling 2019), as well as the fact that he already served as commander in chief, it is perhaps unsurprising that his evaluations would be distinct from other candidates. Notably, Trump’s likeability scores were the lowest of all candidates aside from Martinez, making his gap between likeability and the more masculine characteristics particularly large.

Figure 2. Overall Candidate Ratings on Competency, Likeability, and Commander in Chief (Percent Responding Extremely/Very; Study 2)

Figure 3. Difference between Commander in Chief Ratings and Likeability 

(Study 2)

Note: Bars represent difference between ratings on commander in chief and likeability (CinC – Likeability); percent responding trait describes candidate "Well."

Figure 3 plots the differences between the candidates’ likeability rating and their ratings for commander in chief. The positive values indicate candidates who rate higher on their potential to be commander in chief over their likeability; the negative values are candidates who rate higher on likeability. While we cannot know from our research design that gender is the only driver of these differences, Figure 3 demonstrates clearly that all men rate higher on commander in chief than on likeability. Once again, we see Trump as a substantial outlier, largely as a result of his low likeability score. In contrast, all of the women candidates rate higher on likeability than they do as commander in chief. Taken together, Figures 2 and 3 clearly demonstrate the persistence of the double bind for Republican women candidates, while also confirming the overall patterns found in Figure 1. 

Implications

Our research tests whether the double bind between likeability and competence persists in evaluations of women presidential candidates, using real-world political figures. In addition to the traditional double bind, we also include questions about the role of commander in chief, allowing us to tap into both the expertise and dominance aspects of masculinity. Across our two surveys, we found strong evidence of the persistence of a double bind for Republican women candidates. Comparing individual ratings across the three characteristics, women typically scored highest on likeability, and lowest on commander in chief. In contrast, men tended to score highest in competence and have a smaller gap in the dropoff between commander in chief and the other two characteristics. 

These findings have implications for women candidates, especially in future presidential elections.

• Our research demonstrates that Republican women candidates, at least in presidential contests, must navigate the stereotypes that underlie the double bind. In making their case to voters, they should be cognizant of the persistence of gendered perceptions about their competence and capacity to be commander in chief, both traits that are especially important in presidential contests. Notably, our sample consists of likely Republican voters in New Hampshire, who are generally more moderate than Republicans nationwide. Given evidence that traditional gender stereotypes tend to be strongest among more conservative voters, if anything, we would expect these dynamics to be even more pronounced in a nationally representative sample of Republican voters.
• In addition, our research confirms the delicate balance that future Republican women candidates for the White House need to navigate: establishing that they possess sufficient masculine traits associated with presidential leadership, while also continuing to present as likeable to forestall any backlash that might come from challenging traditional gender expectations. 

Our research also suggests areas for future research.

• The anomalies we found around certain women of color political figures (Elaine Chao, Susana Martinez) underscore the need for future research to better understand the power of intersectional stereotypes (Cassese 2019). 

• We also see that Trump’s hypermasculine candidacy is an outlier, with rankings reflecting what might be considered an inverted double bind: very strong dominance masculinity at the expense of likeability. More research is needed to better understand if and to what degree this matters for men or women candidates of either major political party.

   

Works Cited

ABC News. 2008. “Video Barack Obama Tells Hillary Clinton She’s ‘Likeable Enough.’” ABC News, January 5. https://abcnews.com/video/34428886/.

Bauer, Nichole M. 2020. The Qualifications Gap: Why Women Must Be Better than Men to Win Political Office. Cambridge University Pess.

Bejarano, Christina E. 2013. The Latina Advantage: Gender, Race, and Political Success. University of Texas Press.

Boatright, Robert G., and Valerie Sperling. 2019. Trumping Politics as Usual. Oxford University Press.

Cargile, Ivy A. M. 2023. “Stereotyping Latinas: Candidate Gender and Ethnicity on the Political Stage.” Politics, Groups, and Identities 11 (2): 207–25. https://doi.org/10.1080/21565503.2021.1946097.

Cassese, Erin. 2019. “Intersectional Stereotyping in Political Decision Making.” In Oxford Research Encyclopedia of Politics. https://doi.org/10.1093/acrefore/9780190228637.013.773.

Crowder-Meyer, Melody, Shana Kushner Gadarian, and Jessica Trounstine. 2020. “Voting Can Be Hard, Information Helps.” Urban Affairs Review 56 (1): 124–53. https://doi.org/10.1177/1078087419831074.

Cuddy, Amy J. C., Susan T. Fiske, Virginia S. Y. Kwan, et al. 2009. “Stereotype Content Model across Cultures: Towards Universal Similarities and Some Differences.” British Journal of Social Psychology 48 (1): 1–33. https://doi.org/10.1348/014466608X314935.

Ditonto, Tessa, and David Andersen. 2025. Gender and Candidate Likability. Center for American Women and Politics, Eagleton Institute of Politics, Rutgers University, New Brunswick, NJ. https://cawp.rutgers.edu/research-analysis/cawp-research-grants/gender-….

Dolan, Kathleen. 2014. When Does Gender Matter? Women Candidates and Gender Stereotypes in American Elections. Oxford University Press.

Duerst-Lahti, Georgia. 2006. “Presidential Elections: Gender Space and the Case of 2004.” In Gender and Elections: Shaping the Future of American Politics, edited by Susan J. Carroll and Richard L. Fox. Cambridge University Press.

Goh, Jin X., and Jameson McCue. 2021. “Perceived Prototypicality of Asian Subgroups in the United States and the United Kingdom.” Tqwu3_v1. Preprint, PsyArXiv, February 9. https://doi.org/10.31234/osf.io/tqwu3.

Jamieson, Kathleen Hall. 1995. Beyond the Double Bind: Women and Leadership. Oxford University Press.

King, David C., and Richard E. Matland. 2003. “Sex and the Grand Old Party: An Experimental Investigation of the Effect of Candidate Sex on Support for a Republican Candidate.” American Politics Research 31 (6): 595–612. https://doi.org/10.1177/1532673X03255286.

Newton-Small, Jay. 2016. “Hillary Clinton Sexism: Is She Likable Enough?” TIME. https://time.com/4347962/hillary-clinton-donald-trump-likability/.

Rosenwasser, Shirley M., and Norma G. Dean. 1989. “Gender Role and Political Office: Effects of Perceived Masculinity/Femininity of Candidate and Political Office.” Psychology of Women Quarterly 13 (1): 77–85. https://doi.org/10.1111/j.1471-6402.1989.tb00986.x.

Rosette, Ashleigh Shelby, Christy Zhou Koval, Anyi Ma, and Robert Livingston. 2016. “Race Matters for Women Leaders: Intersectional Effects on Agentic Deficiencies and Penalties.” The Leadership Quarterly, Special Issue: Gender and Leadership, vol. 27 (3): 429–45. https://doi.org/10.1016/j.leaqua.2016.01.008.

Schneider, Monica C., and Angela L. Bos. 2014. “Measuring Stereotypes of Female Politicians.” Political Psychology 35 (2): 245–66. https://doi.org/10.1111/pops.12040.

Schwarz, Susanne, and Alexander Coppock. 2022. “What Have We Learned about Gender from Candidate Choice Experiments? A Meta-Analysis of Sixty-Seven Factorial Survey Experiments.” The Journal of Politics 84 (2): 655–68. https://doi.org/10.1086/716290.

Teele, Dawn Langan, Joshua Kalla, and Frances Rosenbluth. 2018. “The Ties That Double Bind: Social Roles and Women’s Underrepresentation in Politics.” American Political Science Review 112 (3): 525–41. https://doi.org/10.1017/S0003055418000217.

• 1

  A week after Haley announced her candidacy, Vivek Ramaswamy also entered the Republican primary race. As this was just before our survey, and we did not anticipate his candidacy, he was not included in our list of potential candidates.

• 2

  While Chao was not being discussed as a potential primary candidate, we included her name in our survey to try and capture different stereotypes that respondents may have to another example of a senior Asian American woman. Because Chao was born in Taiwan, she would technically not have been eligible to serve as president. But, her Taiwanese background allows us to explore the potential for varied stereotypes around Asian American women from different national origins and with distinct presentations.

• 3

  In presenting this information, we experimentally manipulated 1-2 sentences describing these past experiences using either masculine or feminine language, randomizing which respondent read which description for each candidate. For the purposes of this analysis, we combine responses across our treatment conditions. Differences across treatment groups are generally minimal, and overall trends are consistent with the analysis presented in this memo.