Scholars express deep concern over the rise of affective polarization in the United States and the subsequent threats it poses to American democracy. While much of this work focuses on the feelings among mass publics toward political parties and elites, less attention has been given to the communication strategies of political elites who may amplify or mitigate partisan divides. This project examines how candidate gender shapes the tone of campaign communication, asking whether women’s participation in elections contributes to greater civility in political discourse.
Prior research shows that women are generally perceived as more collaborative, empathetic, and norm-abiding in political contexts. At the same time, gendered expectations constrain women’s rhetorical range: they face steeper penalties for incivility or aggression because such behaviors are incongruent with gendered stereotypes. These cross-pressures suggest that women may navigate campaign messaging differently from men.
Despite substantial interest in negative campaigning, limited work has explored gendered difference in campaign rhetoric. Existing studies vary widely in context and often employ inconsistent definitions of negativity, leaving little theoretical consensus. To address this gap, I identify the three forms of attack ads highlighted in the literature—negative, personal, and uncivil—and evaluate whether men and women differ in their use of these rhetorical strategies. In doing so, this project connects the literature on affective polarization with that on gendered political communication.
Empirically, this project draws on AdImpact’s comprehensive archive of all campaign advertisements from the 2024 election cycle, merged with candidate-level data from the Database on Ideology, Money in Politics, and Elections (DIME). This linked dataset captures over 300,000 ad-week observations across federal, statewide, and local races. I plan to assess campaign tone across the three attack categories using a combination of human-coded metadata and LLM text analysis, allowing for an evaluation of how candidate gender predicts both the use and targeting of negative appeals. I hypothesize that women are less likely to sponsor, and more likely to be targeted by, uncivil or personal attacks. These patterns would offer new evidence that greater gender diversity among political elites may indirectly temper partisan hostility in American politics.
