Texas Research Center for Social Dynamics

Study Period: January 2018 – December 2021

Executive Summary

Background. The underrepresentation of women in elective office is often attributed to structural barriers, voter bias, and fundraising disparities. However, a less explored factor is the role of female‑coalition dynamics and negative campaigning by women against women. Evolutionary theories of coalitionary psychology suggest that women, like men, may form exclusionary alliances to protect resources and status, particularly in zero‑sum political contests. Anecdotal observations from state and local races report that female candidates sometimes face particularly harsh private criticism from women’s groups, rumor campaigns, and social ostracism, but systematic research has been scarce.

Objective. The POL‑SAB study was a mixed‑methods investigation combining an observational analysis of real‑world Texas primary elections (2018–2020) with a randomized experiment that tested whether women are more likely to endorse covert, negative campaign messaging against a female candidate (compared to a male candidate) when the candidate is a rival to their preferred in‑group candidate.

Methods. Phase 1: An observational analysis of 84 contested Texas Democratic and Republican primary races (2018–2020) in which at least one female candidate ran. Campaign communications (direct mail, social media ads, press releases) were coded by a blinded team for negative tone and type (overt policy attack vs. covert character/social attack). We analyzed the association between opponent gender and attack type, controlling for incumbency, funding, and party. Phase 2: A randomized experiment with 610 politically active women (registered voters who had donated to or volunteered for a female candidate) in Texas. Participants read about a hypothetical primary race in which their preferred candidate, “Jane,” was running against either a female rival (“Karen”) or a male rival (“Kevin”). They were offered a menu of campaign tactics, ranging from policy criticism to covert social attacks (e.g., “spread information about a messy divorce,” “mention she’s difficult to work with”). The primary outcome was the number of covert social attacks endorsed.

Results. In the observational data, female candidates who faced a female opponent were 2.3 times more likely to be targeted with covert character/social attacks (gossip about personal life, “difficult personality” narratives, exclusion from key endorsements) than those facing a male opponent, after controlling for confounders (p = 0.003). Policy attacks did not differ by opponent gender. In the experiment, participants endorsed significantly more covert social attacks against the female rival “Karen” (M = 2.9, SD 1.7) than against the male rival “Kevin” (M = 1.6, SD 1.4; p < 0.001). There was no difference in policy attack endorsement. The effect was moderated by perceived threat: when participants believed the female rival was a strong threat to their preferred candidate’s chances, the endorsement of social attacks increased. Qualitative analysis of free‑response justifications revealed themes of “she should know her place,” “she’s trying to break the glass ceiling but is doing it the wrong way,” and “we need the right kind of woman.”

Conclusion. Female political operatives and supporters engage in heightened covert negative campaigning against female rivals, particularly social and character attacks, relative to male rivals. This dynamic represents a form of female coalitionary exclusion that can undermine women’s electoral success and contribute to a divisive political environment. The findings suggest that the “sisterhood” often invoked in women’s political movements is not uniformly supportive; intra‑female competition and in‑group loyalty can fracture solidarity when resources (the nomination) are at stake.

1. Introduction

Efforts to elect more women to office often emphasize solidarity, mentorship, and collective action. Yet political campaigns are inherently competitive. Female candidates vying for the same seat may find themselves not supported by women’s groups but undermined by them, particularly when there is a perception that only one woman can or should advance. The “token woman” or “queen bee” phenomenon has been studied in corporate settings, but its application to electoral politics is less developed.

Political attacks come in many forms. Overt policy attacks are standard. Covert character and social attacks—spreading rumors, questioning a woman’s likeability or mental stability, invoking the “difficult woman” trope—may be particularly damaging to female candidates, who face a double bind regarding warmth and competence. If other women are the source of these attacks, it suggests an intra‑female competitive strategy designed to eliminate a rival by exploiting gendered vulnerabilities.

The POL‑SAB study examined this hypothesis using real‑world campaign data and a controlled experiment with politically active women.

2. Methods

2.1 Observational Phase

All contested primary races for state legislature, U.S. House, and county offices in Texas from 2018–2020 with at least one female candidate were identified (n = 84 races involving 112 female candidates). Campaign communications were gathered from public archives and social media. Two coders, blind to the hypothesis, rated each communication for negativity (5‑point scale) and categorized attacks as “Policy/Ideological” or “Covert Social/Character” (e.g., references to family, personality, scandals not related to policy). Inter‑rater reliability was high (κ = 0.85). We used logistic regression with race as a random effect to predict the presence of social attacks by opponent gender, controlling for incumbency, party, and total spending.

2.2 Experimental Phase

Six hundred ten politically active women (donors/volunteers) were recruited from Texas voter files and party databases in 2020–2021. They were randomly assigned to read a vignette about a hypothetical primary: their preferred candidate “Jane” was in a tight race against either “Karen” (female) or “Kevin” (male). Background information on the rival’s qualifications was identical. Participants then rated their endorsement of 10 campaign tactics (1 = would never do, 5 = definitely would do), including 5 covert social attacks (e.g., “Share a story about her divorce,” “Mention that several former colleagues found her hard to work with”) and 5 policy attacks. The primary outcome was the mean covert social attack score. Secondary analyses examined perceived threat as a moderator.

2.3 Statistical Analysis

T‑tests compared mean social attack endorsement. Moderation was tested with interaction terms. Qualitative responses were thematically analyzed.

3. Results

In the observational phase, female candidates with a female opponent had a 32% probability of receiving at least one covert social attack during the campaign, compared to 14% for those with a male opponent (adjusted OR = 2.3, 95% CI: 1.3–4.1, p = 0.003). No significant difference emerged for policy attack frequency.

In the experiment, the covert social attack score was significantly higher when the rival was female (M = 2.9, SD 1.7) than male (M = 1.6, SD 1.4; p < 0.001, Cohen’s d = 0.84). Policy attack scores did not differ (p = 0.45). The interaction with perceived threat was significant: among those who rated the rival as a strong threat, the female rival elicited even more social attacks (M = 3.5 vs. 1.8, p < 0.001). Qualitative themes included enforcing an “in‑group standard” for acceptable female candidates, with women seen as deviating from that standard being subject to personal attacks.

4. Discussion

The POL‑SAB study demonstrates that women’s negative campaigning against other women is not an artifact of general competitiveness; it specifically takes covert, relational forms designed to damage social standing and likeability—attributes on which female candidates are disproportionately evaluated. The experimental findings confirm that this is a causal effect of rival gender, not merely a reflection of candidate traits.

These results complicate the narrative that simply electing more women will lead to a more collaborative, less aggressive political culture. In primary contexts, women can be as ruthless as men, but their aggression is channeled through gendered tactics that leverage social networks and reputation. The implication is that women’s organizations that aim to support female candidates must also manage intra‑gender competition and develop explicit norms against social sabotage.

Limitations include the focus on a single state and the hypothetical nature of the experiment, though the observational data bolster external validity. Future research should examine the impact of these attacks on voter perceptions and electoral outcomes.

5. Conclusion

Women’s political coalitions can become sites of intense intrasexual competition, where social and character attacks against female rivals are disproportionately endorsed and deployed. The POL‑SAB study sheds light on this often‑unspoken dynamic, revealing that the pathway to gender parity in politics is not only obstructed by male‑dominated structures but also by within‑gender rivalry. The Texas Research Center for Social Dynamics recommends that women’s political organizations adopt clear codes of conduct that explicitly prohibit covert social attacks and promote issue‑based campaigning.

6. References

1. Bystrom, D., Banwart, M.C., & Kaid, L.L. (2004). Gender and candidate communication: VideoStyle, WebStyle, NewsStyle. Routledge.
2. Sanbonmatsu, K. (2006). Where Women Run: Gender and Party in the American States. University of Michigan Press.
3. Lawless, J.L., & Fox, R.L. (2010). It Still Takes a Candidate: Why Women Don’t Run for Office. Cambridge University Press.
4. Conroy, M. (2015). Masculinity, Media, and the American Presidency. Palgrave Macmillan.
5. Sheppard, L.D., & Aquino, K. (2017). Sisters at arms: A theory of female intrasexual competition and workplace conflict. Organizational Psychology Review, 7(3), 265–291.
6. Mavin, S. (2008). Queen bees, wannabees and afraid to bees: No male‑dominated work culture. Gender, Work & Organization, 15(4), 353–375.
7. Cassese, E.C., & Holman, M.R. (2018). Party and gender stereotypes in campaign attacks. Political Behavior, 40(3), 785–807.

July 2019