Women of color are running for office at higher rates in recent U.S. elections, and they are using a variety of strategic political messages. One approach to building support among voters is for women of color to use emotional appeals in their campaign messages. This strategy could help make them more relatable and could create a channel of communication between the candidates and groups who might otherwise feel disconnected from them. We examine how women of color use emotions in their campaign messages and how voters react to these messages through two empirical investigations. Specifically, we examined televised campaign advertisements across multiple electoral years to track women’s use of emotions; then we complemented this data with a survey experiment testing whether positive and negative emotional appeals from women of color shape voters’ evaluations of the fitness and electability of these candidates.
Data
Our dataset is a compilation of advertising data that aired on television during the 2010, 2012, 2014, 2016, and 2018 congressional elections. This data comes from the Wesleyan Media Project (WMP), which tracks all political advertising aired by candidates in the 200 U.S. television media markets. We analyzed whether women candidates developed strategic campaign messages using emotional cues that appealed to fear, anger, sadness, enthusiasm, pride, and humor. We integrated additional information on campaign ads’ features, campaign dynamics, and candidates’ characteristics using several online data repositories, including the Center for American Women and Politics.
The final dataset contains a total of 27,495 ads from 15,798 unique candidates for both the U.S. House and the U.S. Senate, including 579 women of color. We use the term “women of color” to distinguish between the experiences of women from a racial/ethnic minority background relative to white women. However, we also acknowledge that women from different racial/ethnic backgrounds have unique experiences, such that the experiences of Latinas may differ from the experiences of Black women relative to Asian women and other women.
Results
We use this data to test whether women of color will display more positive emotions, like enthusiasm, and fewer negative emotions, like anger, relative to their white female counterparts to deter voters’ backlash and negative media coverage. Extant scholarship on political campaign advertising suggests that political candidates make extensive use of emotional appeals (Kühne & Schemer, 2015; Weber, 2012), but it remains unclear whether women of color’s emotional appeals are largely positive or negative. Employing emotional appeals could prove advantageous for women of color, as they may be able to reach a broad array of voters and build substantial voting coalitions. Positive emotional appeals, in particular, may excite voters and generate enthusiasm for these candidates.
Our results show that women of all races/ethnicities, and especially Asian women, tend to employ ads appealing to positive emotions, namely enthusiasm (see figure below). Cueing enthusiasm might serve to get voters excited about the election and the respective candidate. White women are more likely to use anger in their messages while Black women are the least likely to use appeals to anger. The latter is consistent with the idea that Black women tend to avoid the stereotype of the “angry Black woman” (Brown, 2014) that portrays them as irrational. Instead, Black women were more likely to use positive emotional appeals, such as enthusiastic appeals, in their campaign messages. Together, these patterns seem to suggest that women of color candidates largely avoid appealing to negative emotions, chiefly anger, relative to white women candidates.
Experiment
To examine the potential effects of these emotional appeals on voters’ appraisals of candidates, we complement our analyses of the advertising data with an online survey experiment conducted in April 2022. We recruited a convenience sample of 1,125 U.S. adults (female = 58.04%) using an online platform (i.e., Prolific Academic), and although our sample skewed highly White/Caucasian (72.80%) and Democrat (57.3%), we argue that this sample is still appropriate for our purposes because women of color tend to be perceived as Democrats and often run as such (Fraga, Shah, & Juenke, 2020). We showed participants the written text of a campaign message from a fictional woman candidate (a white female candidate, a Black woman candidate, an Asian woman candidate, and a Latina candidate) alongside her picture. The message was modeled after the style and overall tone of campaign ads employed by actual political candidates and used fictional names to cue the candidate’s intended race/ethnicity. Participants were randomly exposed to either a campaign message appealing to positivity and hope, or a negative message associated with anger and danger. Participants then answered the following question, “If this candidate were running in your state, how likely are you to vote for her?” Responses were measured on a three-point scale where a higher value indicated a more favorable preference.
Experimental Results
We compared voting preferences for positive and negative messages separately for all candidates. Although we did not find significant differences between positive and negative messages for the Latina candidate, the Asian woman candidate, and the white woman candidate, our results suggest that participants were 7% more likely to express a preference for the Black woman candidate employing positive messages compared to negative messages, and this difference was significant, p=0.079 (see figure below). This is an affirmation of the strategies real Black women candidates tend to employ in their avoidance of negative emotional appeals; Black women who adopt negative appeals may risk voter backlash.
When looking specifically at differences among participants, we found that women of color who read the positive emotional message from the Latina candidate were 24% less inclined to support her compared to white female participants, and this difference was highly significant, p=0.003 (see Figure 3). Non-significant results, however, were found for the Asian woman and the Black woman among women of color and white female participants. Overall, our analyses suggest that women of color candidates are not necessarily seen more favorably among women of color voters compared to white female voters, and Latinas may be punished among women of color when they employ positive emotional appeals. Past research shows that Latinos may be more receptive to angry appeals and disagreement with issues that hurt their community, including restrictive immigration policies (Gutierrez, Ocampo, Barreto, & Segura, 2019; Pantoja, 2018), and therefore Latinas may be able to leverage anger to mobilize Latinos.
Implications
The main takeaways from our study are:
- Women of color employ different emotional appeals on the congressional campaign trail and are rated differently by voters.
- Women of color, and especially Black women, are less likely to employ negative emotional appeals compared to white female candidates.
- Voters are more favorable of positive messages, as opposed to negative messages, from the Black woman candidate.
- Latina candidates may want to avoid positive emotions, such as enthusiasm, to engage and mobilize women of color voters.
Future studies in this area can further explain the nuance of these differences by taking into consideration the intersectional identities of women of color that render them unique candidates navigating the current gendered and racialized political terrain.
References
Brown, N. E. (2014). Sisters in the statehouse: Black women and legislative decision making. New York, NY: Oxford University Press.
Fraga, B. L., Shah, P., & Juenke, E. G. (2020). Did women and candidates of color lead or ride the democratic wave in 2018? PS: Political Science & Politics, 53(3), 435-439.
Gutierrez, A., Ocampo, A. X., Barreto, M. A., & Segura, G. (2019). Somos más: How racial threat and anger mobilized Latino voters in the Trump era. Political Research Quarterly, 72(4), 960–975.
Kühne, R., & Schemer, C. (2015). The emotional effects of news frames on information processing and opinion formation. Communication Research, 42(3), 387-407.
Pantoja, A. D. (2018). Latino voters will turn anger into action in the 2018 congressional midterm elections. Latino Decisions. Retrieved from www.latinodecisions.com/blog/2018/07/30/ latinos-voters-will-turn-anger-into-action-in-the-2018-congressional-midtermelections/
Weber, C. (2012). Emotions, campaigns, and political participation. Political Research Quarterly, 66(2), 414-428.