How Women Attack: Candidate Reliance on Feminine and Masculine Traits in Campaign Negativity

Abstract

We develop a framework to identify how women use negative messages in political campaigns. We argue that women will be more likely to use contrast negativity, messages that include a negative message against an opponent and a positive message about the candidate sponsoring the ad, rather than attack negativity, messages that only criticize an opponent. We also identify how feminine and masculine traits emerge in negativity—a strategy we call gendered trait negativity. We analyze gendered trait negativity in television ads from the Wesleyan Media Project (WMP) for House, Senate, and gubernatorial races from 2010 to 2018. Using logistic regression models, we find that women use contrast messages more than attack messages when running against a man. Second, we find that women are more likely to highlight feminine traits over masculine traits in negative messages when their opponent is a man relative to when their opponent is a woman.

Keywords

female candidates negative ads traits all-women races gender stereotypes political communication

In American elections, over 50% of the campaign messages a voter sees will either be an attack ad that criticizes an opponent or a contrast ad that criticizes an opponent while also highlighting the positive qualities of the candidate who sponsored the ad.¹ Of these negative messages, many are likely to be sponsored by a female candidate (Bystrom, 2006; Lau & Pomper, 2004) and female candidates are also likely to be the subject of negative messages from their opponents (Walter, 2013). Negativity is impactful because voters tend to remember negative messages more than positive messages (Franz et al., 2007; Geer, 2006). The use of negativity can, however, be fraught for female candidates as experimental research suggests that women can face a gendered punishment for airing negative messages as negativity violates expectations that women be pleasant, warm, and agreeable (Craig & Rippere, 2016; Krupnikov & Bauer, 2014). At the same time, women are particularly vulnerable to negative messages that use feminine traits to characterize them as lacking the masculine qualities that align most strongly with expectations for political leaders (Cassese & Holman, 2018).

We propose a framework of gendered trait negativity that considers how gender stereotypes affect how women use negative messages. Gender stereotypes characterize women as lacking the masculine qualities needed for political leaders (Eagly & Karau, 2002). Women can overcome this perception by emphasizing masculine qualities (Bauer, 2017) but emphasizing masculinity can lead to a backlash for women for failing to conform to feminine stereotypes (Bauer et al., 2022). We argue that women will manage these competing expectations to be both feminine and masculine in the types of negative messages they deploy. Past work often considers negative messages to include both contrast and attack messages (Geer, 2006), but the distinction between these two different forms of negativity matters for how women develop negative messages. Contrast messages allow women, we argue, to undercut the strengths of their male opponents while appearing to conform to feminine stereotypes of women as agreeable and kind through the positive component of the message. Women will also strategically use masculine traits to bolster their own perceived qualifications and to undercut perceptions of their male opponents. In all-women races, we argue women will air more attack ads than contrast ads because voters will be less likely to see a negative message as a violation of feminine norms when there is no male candidate as a point of comparison against the woman.

We tracked the use of feminine and masculine traits in contrast and attack messages with a dataset of approximately 28,000 campaign ads in U.S. Congressional and gubernatorial races that aired from 2010 through 2018 using data available from the Wesleyan Media Project (WMP).² We merged our data on gendered traits in campaign ads with data on candidate characteristics (i.e., candidate gender, race/ethnicity) and the dynamics of each electoral contest (i.e., mixed-gender contests vs. all-women races, competitiveness). We rely on television ads because the bulk of money that campaigns spend on advertising goes toward television ads as opposed to digital advertising (Wesleyan Media Project, 2018) and televised campaign ads are often seen by large segments of the electorate through incidental media exposure (Fowler et al., 2021). The time-period we rely on, nearly a decade, is particularly instructive about how women manage gendered campaign dynamics as this period saw a steady increase in the numbers of women running for and winning political office (CAWP, 2020).

Examining the dynamics of gendered trait negativity is important because negative messages may be a previously undetected source of bias against female candidates and these messages can increase the barriers to electoral success that women already face in a gendered political campaign environment (Bauer, 2020a; Dittmar, 2015; Jamieson, 1995). Female candidates, on average, run in more competitive races and against higher quality challengers and women win their elections by smaller margins compared to male candidates (Branton et al., 2018; Pearson & McGhee, 2013). The competitiveness of women’s races means that these candidates cannot avoid using negativity as doing so leaves them vulnerable to attacks from their male opponents. Women’s descriptive under-representation in elected political offices reduces the perceived legitimacy of democratic political institutions and reduces a person’s likelihood of participating in the political system (Clayton et al., 2019). These findings have critical implications for research on women’s political underrepresentation and the gendered barriers facing women as they pursue political leadership roles.

Our results show three key findings. First, female candidates are more likely to use contrast messages than male candidates, but female candidates are just as likely to rely on attack messages as male candidates. Second, we show that female and male candidates are both more likely to use feminine traits in contrast and attack ads in mixed-gender elections. Third, we find that women in all-women races use negative feminine traits in attack messages more than women in mixed-gender elections. Our findings hold important insights into the obstacles facing women’s political success in the American political system but our findings also shed light on the barriers facing women in proportional democratic systems where gender biases can affect women (Aaldering & Van Der Pas, 2020; Phillips, 2021; Rohrbach et al., 2020).

Gender Stereotypes and Campaign Negativity

Gender stereotypes play a role in shaping how women develop their campaign strategies (Bauer & Santia, 2023). Feminine stereotypes characterize women as nurturing, compassionate, and more suited for supportive or caregiving social roles (Prentice & Carranza, 2002). Masculine stereotypes characterize men as tough, assertive, and more suited for agentic or power-oriented roles (Holman et al., 2019; Koenig et al., 2011). Gender stereotypes hold political relevance because masculine stereotypes most directly map onto the ways that voters think about and conceptualize political leadership (Eagly & Karau, 2002; Schneider & Bos, 2019). Feminine stereotypes, especially feminine traits, are incongruent with the dominantly masculine stereotypes voters associate with political leadership (Aaldering & Van Der Pas, 2020; Ditonto, 2019). Voters frequently perceive female candidates as lacking masculine traits, and this perception exacerbates gender-role incongruity dynamics (Schneider & Bos, 2014), and emphasizing feminine traits in campaign messages undercuts the perception of women as political leaders (Bauer, 2015).

Women face a gender-role incongruity problem where women who violate feminine expectations by engaging in stereotypically masculine behaviors run the risk of a backlash for not being “nice enough” (Eagly & Karau, 2002; Sweet-Cushman, 2022); but, female candidates need to engage in masculine behaviors to overcome the perception that they lack the masculine traits needed in leaders (Bauer, 2020b). Female candidates often consider how feminine and masculine stereotypes shape the perceptions voters form of them (Dittmar, 2015) and adjust their strategies to minimize the potential for gender bias and to overcome the perception that they lack masculine traits (Bauer & Santia, 2022). Past research offers little insight into how women mitigate the potential for a gendered backlash through campaign negativity.

Women and men rely on negativity at equitable rates (Walter, 2013). Women who attack their opponents for being weak on issues can benefit by expanding the set of issues that voters perceive them to be experts on (Herrnson & Lucas, 2006) while undercutting the perceived issue strengths of their opponents (H. K. Evans & Clark, 2016; Windett, 2014). The issue expansion effect is particularly beneficial for women as voters tend to see women as having a narrow set of expertise on stereotypically feminine issues (e.g., education or health care) (Schneider, 2014) and see women as lacking expertise on stereotypically masculine issues which are often considered the highest priority issues in an election (e.g., terrorism and defense spending) (Holman et al., 2019). Experimental research suggests that women who sponsor negative messages face a gendered backlash for violating feminine expectations that women should be kind and agreeable (Craig & Rippere, 2016; Krupnikov & Bauer, 2014). Despite the risks of a backlash effect, we know that women rely on campaign negativity (Walter, 2013), but it is not clear how women engage in campaign negativity to limit the potential for a backlash effect while also allowing them to fill the perceived “masculine deficit” that comes from the incongruity between their gender and the masculine stereotypes of leadership.

Negativity poses risks for male candidates who can face a backlash because voters do not like campaign negativity (Garramone et al., 1990; Skaperdas & Grofman, 1995). This type of punishment for negativity differs from the punishment facing women because men are not being punished for violating stereotypic expectations. Men still use campaign negativity to attack their female opponents on stereotypically masculine issues (Gordon et al., 2003; Melich, 2005; Windett, 2014), a strategy that can further reinforce the perception that women lack the masculine qualities needed to hold political office. The extent to which men benefit from attacking their female opponents is not entirely clear. Fridkin et al. (2009), through experimental research, find that attacks by male candidates against women are often ineffective because “the presence of gender stereotypes appears to soften the blow of negative attacks, leading people to discount attacks on women candidates, compared to identical attacks on male candidates” (p. 70). At the same time, Cassese and Holman (2018) find, also using an experiment, that female candidates are especially vulnerable to feminine trait attacks made by their male opponents, especially Democratic female candidates, because such attacks draw attention to the incongruity between being a woman and being a political leader. Experimental evidence suggests that women can be vulnerable to gendered attacks from their male opponents, but it is not clear how frequently male candidates use feminine or masculine traits in campaign negativity that targets their female opponents. Our manuscript fills this gap.

Our study makes several key contributions to the literature on gender and campaign negativity. First, we delineate between the use of contrast and attack ads as a strategy for women to use negativity while mitigating the potential for a stereotypic punishment. Most accepted definitions of negativity rely on a variation of Geer (2006) who defined negativity as “any criticism leveled by one candidate against another during a campaign” (p. 23). This definition implicitly assumes that candidates who use negativity will use contrast and attack messages equitably. We argue that the distinction between contrast and attack ads is critical for understanding how women use campaign negativity in ways that minimize the potential of a gendered backlash. Second, we also consider how women use feminine and masculine traits in contrast and attack messages to establish their masculine credentials while also undermining their opponent’s qualifications. Third, we shed light on how male candidates use gendered traits to leverage role congruity dynamics to their advantage.

Gendered Trait Negativity

We argue that the gendered traits that women and men emphasize in their contrast and attack ads are shaped by two critical factors: gender-role incongruity dynamics and the gender of an opponent. We describe how candidate gender will condition the type of negative messages a female candidate uses (contrast or attack) in a mixed-gender race, how candidates will use feminine and masculine traits in negativity when women and men run against one another, and then we highlight how these dynamics will differ in the context of all-women electoral contests.

Gender Stereotypes and Reliance on Contrast Ads

The relationship between gender stereotypes and candidate gender will affect the types of negative messages women use, whether it is a contrast message or an attack message. Airing a straightforward attack ad that only criticizes a male opponent can lead to a backlash effect for a female candidate because she is likely to be perceived as breaking with feminine stereotypes demanding that women should be agreeable and kind (Krupnikov & Bauer, 2014). We argue that the potential of a gendered backlash for negativity will be most likely to occur in mixed-gender races, when a woman runs against a man. Social psychology research shows that the way people use gender stereotypes to evaluate women depends on the comparative context of the evaluation, or whether people are primed to compare a woman and a man to one another (Biernat & Manis, 1994). A mixed-gender comparative context in an election will make people more likely to fall back on feminine stereotypes to evaluate the woman and masculine stereotypes to evaluate the man (Bauer, 2020b; Meeks & Domke, 2016; Mueller, 1986). The presence of a male opponent, by itself, invites voters to use gender stereotypes to differentiate between the two candidates (Bauer, 2020b). The mixed-gender context of a race means that women need to find a way to criticize their opponents without appearing to violate norms of feminine “agreeability.” We argue that female candidates will overcome this challenge by relying on contrast negativity more than attack negativity. The inclusion of both positive and negative content in a single contrast message means that female candidates will see these messages as a safe way to criticize an opponent compared to a straight-forward attack message.

Male candidates do not have to worry about making the same gendered calculations as women around violating feminine expectations. Men run the risk of a backlash effect for airing an attack ad against a woman (Fridkin et al., 2009) but we think that male candidates are more willing to take this risk compared to female candidates. Men in politics are more risk-acceptant compared to women (Kanthak & Woon, 2015; Preece & Stoddard, 2015; Sweet-Cushman, 2016) and this can provide men with a stronger motivation to use attack ads compared to women. Moreover, voters tend to see male candidates as having the masculine traits that align most strongly with leadership roles (Schneider & Bos, 2014) and voters do not generally care if male candidates lack feminine traits (Bauer, 2017). Our first set of hypotheses argues that women will be more likely to use contrast messages over attack messages while men will use attack messages.

H1a: Female candidates will be more likely to use contrast messages than male candidates.

H1b: Female candidates will be less likely to use attack messages than male candidates.

Managing Feminine and Masculine Traits

We start with outlining the incentives for women to rely on masculine traits in campaign negativity and then we turn to outlining the incentives for male candidates to use feminine traits in contrast and attack messages. Voters do not see female candidates as having the masculine traits needed to be a political leader (Schneider & Bos, 2014), and this perception creates a masculinity deficit for women. Voters assume that male candidates have the masculine traits needed for political office (Bauer, 2020a) because of the congruity between being a man, masculine traits, and being a political leader (Eagly & Karau, 2002). In this section, we consider the use of gendered traits in both contrast and attack ads. While H1a predicts that women will use contrast more than attack messages, we do not have a theoretical reason to think that women will use different gendered traits across contrast and attack messages.

Masculine traits in negative messages, both contrast and attack, can help female candidates in two ways. First, a woman can use masculine traits to frame herself as having positive masculine qualities that align with the expectations voters hold for political leaders and, in a contrast message, women can characterize a male opponent as lacking these positive masculine qualities. By positive masculine qualities, we refer to the masculine traits that are normatively positive in political leaders (Aaldering & Van Der Pas, 2020). Masculine traits such as being strong and assertive are generally seen as positive and valuable qualities for political leaders while masculine traits such as physical aggression or violence are seen as normatively negative qualities in political leaders (Miller et al., 1986; Schneider & Bos, 2014). We expect female candidates to use negative masculine traits in attack ads to portray her male opponent as having the wrong masculine qualities for political leadership while positive masculine traits will bolster the women’s credentials in contrast ads. In other words, the goals of these messages are to lead voters to see a man as too aggressive but to see a woman as more authoritative. Masculine traits in negativity can overcome the masculinity deficit that female candidates face.

Male candidates will use feminine traits in negativity against their female opponents. Because feminine traits are not congruent with the traits associated with political leaders, male candidates will have incentives to use these qualities to undermine their female opponents. Male candidates can use feminine traits in two ways. First, male candidates can use feminine traits that are generally seen as positive to characterize themselves as being caring and compassionate and to frame their female opponents as lacking these qualities. While feminine traits like care and compassion are generally seen as incongruent with masculine leadership roles, these traits are often seen as a net benefit in male candidates (Bishin et al., 2006). Men described with feminine traits are still assumed to have masculine traits; women with feminine traits are seen as lacking masculine traits (Bauer, 2015). We expect to see these positive feminine traits in contrast messages that can both bolster a man’s trait qualities while undercutting a woman’s trait qualities. Second, feminine traits can also be used in attack messages. We expect men to use feminine traits that do not align with leadership roles to characterize women as weak, which is generally seen as a feminine quality, or as lacking feminine qualities such as honesty. Traits such as being weak or soft are normatively negative feminine traits that are seen as bad for women and bad for political leaders to have (Schneider & Bos, 2014), and they represent an opportunity for male candidates to undermine their female opponents.

While we expect differences across gender in whether candidates use contrast and attack messages, we expect the patterns of trait use we map out here to be similar in contrast and attack messages. Thus, we expect women to use masculine traits in both contrast and attack messages, though women will use more contrast messages than attack messages. We expect men to use feminine traits in both contrast and attack messages though we expect men to use more attack messages than contrast messages. Because we do not distinguish here between contrast and attack messages, we simply refer to both types of messages as negative messages. Our second set of hypotheses outlines how candidates will use feminine and masculine traits in negativity.

H2a: Female candidates will be more likely to use masculine traits in negative messages relative to male candidates.

H2b: Male candidates will be more likely to use feminine traits in negative messages relative to female candidates.

Gendered Trait Negativity in All-Women Races

Research on how women communicate in all-women races finds that women are more likely to emphasize feminine stereotypes compared to women in mixed-gender races (H. Evans, 2016; Wagner et al., 2017) and voters reward women who emphasize femininity in all-women races (Meeks & Domke, 2016). We use this work to develop predictions about how women will make different calculations about the use of contrast and attack messages against a woman opponent relative to a man, and the types of traits they will use in negativity. We expect women to worry less about violating feminine norms of “playing nice” in an all-women race. In a mixed-gender race, women expect voters to see an attack ad as a violation of feminine norms because voters will be more likely to hold women to feminine standards of behavior when the opponent is a man (Dittmar, 2015). When there are two women running against one another, gender stereotypes are not helpful informational tools because the same feminine stereotype is likely to be applied to both candidates. The all-women context means that voters may use different candidate cues to evaluate women include the race or ethnicity of the woman and/or partisanship of the women among other characteristics (Brown & Lemi, 2021; Cargile, 2021). The decreased salience of gender stereotypes decreases the likelihood of a backlash for violating feminine norms, and “frees” women to use attack ads.

The shift from a male opponent to a female opponent will also lead women to use more feminine traits in gendered negativity. Voters will be less likely to use gender stereotypes to differentiate between a woman and another woman because stereotypes will just not be helpful differentiation tools in this context. The all-women context will reduce the pressure women feel to conform to masculinity (Meeks & Domke, 2016). Thus, we argue that when it comes to gendered traits in all-women races, women will rely on feminine traits to both bolster their own qualifications and criticize the qualifications of their opponents. We do not expect there to be differences in whether women use feminine traits in attack messages or contrast messages, though we do expect more attack messages in all-women races. We outline our expectations for the use of gendered traits in negative messages in all-women races in H3a and H3b below.

H3a: In all-women races, candidates will be more likely to rely on attack messages relative to female candidates in mixed-gender races.

H3b: In all-women races, female candidates will be more likely to rely on feminine traits in negative messages relative to female candidates running in mixed-gender races.

Data Collection and Key Variables

Campaign Advertising Data

We use data from the Wesleyan Media Project (WMP) to measure how female candidates use gendered traits in contrast and attack messages. We use data from 2010 through 2018 at the gubernatorial, U.S. Senate, and U.S. House levels (Fowler et al., 2014, 2015, 2017, 2019, 2020). The WMP records data on the tone of the ad (contrast, attack, positive), the sponsor of the ad, whether a candidate appears in an ad, advertising cost, and other relevant characteristics of ads.

We use data on televised campaign ads for several reasons. One is that the ads are available. Beyond the convenience factor we use television ads because they offer us a picture of the back-and-forth exchanges of negativity candidates are likely to engage in during a campaign. Second, campaign ads allow us to see how candidates develop negative messages that leverage gendered traits in a strategically planned communication environment. Campaign ads require some money, time, and planning (Franz et al., 2007). The use of a gendered trait to criticize an opponent is more likely to represent a strategically planned message as opposed to a tweet which may be a planned message but might also be an unplanned “off the cuff” message. Third, televised campaign ads are more likely to be seen by a broad segment of the electorate whereas ads through social media are seen by a more targeted, narrow audience (Fowler et al., 2021). While candidates are turning to social media and digital advertising in more recent elections, campaigns still spend the bulk of their funds on television ads (Wesleyan Media Project, 2018). Negativity through campaign messages has the potential to do the most damage in terms of undermining vote support for a woman. Of course, the use of television ads means that our data is limited to those candidates who had the resources to air campaign ads. However, given that negative messages will likely arise in the most competitive races, and competitive races are more likely to have the funds for campaign ads, this is not necessarily a limiting factor. We can track gendered trait negativity in the types of races where these messages can have a very large impact. Fourth, the overarching themes in negative messages embedded in televised ads are likely to represent core themes that emerge in other forms of communication, such as the digital ads increasingly used in later election cycles (Fowler et al., 2021; Kang et al., 2018).

Content Analysis of Traits in Campaign Ads

We produced transcripts of the campaign ads from the video files provided by the Wesleyan Media Project (WMP), and we use these transcripts to record the gendered traits candidates use in contrast and attack messages. Each of the ads was watched and transcribed by hand. The ads average 30 s in length and take approximately 90 s to 2 min to transcribe. We had a random sample of 10% of the campaign ads transcribed by a second researcher to ensure reliability and consistency in producing the initial transcripts. We used a hand transcription process because automated transcription tools produced inaccurate texts for many of the campaign ads. There were just over 28,000 unique ads that aired across all the races and years in our data. Using the transcripts, we recorded whether a set of gendered traits were used to describe the candidates in the ads. Table 1 lists the full set of traits we coded in the ads as well as the classification of a trait as feminine or masculine, and those traits that are positive or negative in political leaders. We instructed our researchers to record whether feminine or masculine traits were used to describe a candidate in the election, and this could mean the trait was used to describe the candidate sponsoring the ad or the opposing candidate. The traits do not necessarily need to be used to characterize both the opponent and the candidate sponsoring the ad. Because contrast ads include an explicit mention of both candidates, if the trait is used just to characterize one of the candidates there will still be an implicit comparison to the other candidate that voters will make because both candidates are explicitly mentioned in the message.

Table 1.

Trait Classifications.

Positive feminine traits	Negative feminine traits
Caring	Weak
Mom/Dad	Distrustful
Compassionate	Liar
Trust	Untruthful
Compromiser	Dishonest
Husband/Wife	Negative
Honest	Misleading
Bipartisan	Soft/Weak
Collaborator
Positive masculine traits	Negative masculine traits
Tough	Corrupt
Leader	Dangerous
Experienced	Hypocrite/hypocritical/hypocrisy
Strong	Insider
Bold	Reckless
Fighter	Career politician(s), out of touch
Hard worker
Stands up to. . .

We classified traits as positive or negative and feminine or masculine using a twofold approach. The negative feminine and masculine trait classifications designate traits that are normatively negative and do not fit into political leadership stereotypes (Aaldering & Van Der Pas, 2020), such as being dishonest or dangerous, but still have associations with one gender over the other. The positive feminine and masculine trait categories capture traits that could be normatively positive depending on the gender of the candidate they are attached to, such as being caring or being a hard worker, and represent some of the more subtle criticisms that can undermine women’s candidacies (Cassese & Holman, 2018). We identified feminine traits that, in general, are not necessarily a deficit in political leaders, such as being honest or caring (Bishin et al., 2006; Miller et al., 1986), but are traits that, for women, can evoke the role incongruity between being a woman and being a political leader. We classified familial gender roles (such as being a mother) as feminine traits because this is a role that leads to further feminine trait associations for women (Bauer, 2015; Deason et al., 2015). We pre-tested our list of traits with a short survey that asked respondents whether they thought a woman, or a man was more or less likely to display a specific quality. We used a threshold of 60% for designating a trait as belonging to a gender stereotypic category and we borrow this threshold from work on partisan stereotyping (Petrocik, 1996). Our trait pre-test is in Web Appendix B.

Dependent Variables

Our dependent variables are the negative tone of an ad as a contrast or an attack ad. Our contrast ad variable is coded 0 if an ad is not a contrast ad and 1 if the ad is classified as a contrast ad. Our attack ad variable is coded 0 if an ad is not an attack ad, and 1 if the ad is classified as an attack ad.

Key Independent Variables

Using our trait data, we created four variables: positive feminine traits, positive masculine traits, negative feminine traits, and negative masculine traits. Each variable is dichotomous and coded as 1 to indicate that an ad used at least one trait to describe a candidate. We created two types of feminine trait variables and two types of masculine trait variables to distinguish between the types of traits that are considered positive in political leaders such as being authoritative as a masculine trait voters look for in a leader (Aaldering & Van Der Pas, 2020) and the negative trait variables represent traits that do not fit with the qualities voters look for in political leaders, such as being weak (Schneider & Bos, 2014). We expect the negative trait variables will be used more in attack messages, but the positive trait variables will be used in contrast messages. Our four trait variables allow us to capture the different ways that candidates will use feminine and masculine traits in campaign negativity. We use our trait variables as independent variables to predict whether there are gender differences in the incorporation of these traits in contrast and attack ads. H2a predicts that women will be more likely to use masculine traits relative to male candidates while H2b argues that men will be more likely to use feminine traits relative to their female opponents. In all-women races, H3b predicts that women will use feminine traits in contrast and attack messages.

Examples of gendered trait ads include a 2018 ad for the Senate race in Washington state from challenger Susan Hutchinson (R) against Senator Maria Cantwell (D) using feminine traits in an attack ad declaring her the “second worst Senator to work for, how can we trust Cantwell to work for us?” This ad uses feminine traits to attack Cantwell using the trait of “trust” which is a quality that female politicians and leaders are thought to embody more than male politicians and leaders (Barnes & Beaulieu, 2019). Another example that uses the masculine trait of being corrupt to attack a candidate is a 2016 ad for Mark Holbrook against Chellie Pingree that said, “Chellie Pingree is a member of the corrupt Washington ruling class who tells the rest of us what to do.” We include a full set of ad examples for all types of ads and traits in Web Appendix B.

Additional Independent Variables

We collected additional data on the characteristics of each candidate, the dynamics of each electoral contest, and the characteristics of campaign ads to account for other factors that affect whether candidates air campaign ads at all. The candidate characteristics we collected data for include candidate gender, incumbency, candidate party, and race/ethnicity. We relied on the Center for American Women in Politics for the data on candidate gender. We coded the candidate gender variable as 1 if the candidate was a woman and 0 otherwise. We collected information on candidate race/ethnicity through congressional biographies at the Library of Congress, Who Leads America, and the Center for American Women and Politics. We include incumbency status because we expect incumbents to be less likely to air negative ads, either contrast or attack ads, unless they are in a competitive race (Kahn & Kenney, 2004).

At the level of the electoral contest, we recorded whether the election was mixed-gender or all-women. All the models include a dichotomous variable for whether the race was a House race or a Senate race, leaving gubernatorial races as the excluded category. We additionally recorded whether the race was competitive using data from the Cook Political Report and whether the race was an open seat. At the level of the campaign ad, we included the length of the ad and the average cost of campaign ads by media markets. Because candidates who run attack ads are likely to abstain from appearing in their own ads to avoid potential backlash, we include whether a candidate appeared in or narrated their ad as a control variable (Banda & Windett, 2021). Finally, we recorded the feminine and masculine issues that candidates referenced in their ads, and we include a dichotomous variable coded as 1 if a candidate mentioned a stereotypically feminine issue and a dichotomous variable coded as 1 if a candidate mentioned a stereotypically masculine issue. Stereotypically feminine issues are issues that disproportionately affect women, such as reproductive rights, or issues that connect to feminine traits, such as education which connects to being caring (Lizotte, 2020). Masculine issues are those that reinforce masculine traits such as defense policy connecting to strength and authority (Holman et al., 2019). Web Appendix B, Table A6 includes a full discussion of the gendered issue classifications. We clustered all our standard errors at the level of the media market.

Intercoder Reliability

To ensure a high level of reliability within the coding of traits from the transcripts, 10% of the ads were coded by a second independent coder who did not know the purpose of the trait coding project but was asked to look for the list of traits used to describe the favored candidate, the candidate the ad is designed to support, or the opposing candidate, the candidate the ad criticizes. Only traits used to describe a candidate in the race were coded. We conducted this reliability test because the use of traits in ads about the candidates requires making some subjective judgments. We analyzed the reliability in the coding of our four key trait variables broken down by office and by election year. We found a few cases where reliability thresholds fall just below the 0.667 threshold for Krippendorf’s alpha with the negative masculine traits in 2010 House races, positive feminine traits in 2012 governor’s races, negative feminine traits in the 2016 governor’s races, positive feminine traits in the 2012 House races, and positive feminine traits in the 2014 Senate races. Further investigations of these low reliability levels suggest these are cases where there were just very few mentions of the relevant traits and missing one or two cases can produce very low reliability levels. Out of the 60 tests of reliability we ran, only these five failed to meet a minimum level of reliability. We include more information about our process in Web Appendix B.

We also conducted an intercoder reliability test with the coding of the tone of the campaign ads. We wanted to be sure that the tone coding provided by the WMP data researchers matches how we considered tone in our study. A student researcher blind to the purpose of this exact study was asked to consider whether an ad was a contrast ad, an attack ad, or a positive ad. We provided conceptual definitions of ad tone. We told the researcher that contrast ads include both positive content about one candidate and negative content about the opponent; attack ads only include negative content about an opponent; positive ads only include positive content about a candidate. We compared the student coding of tone to the WMP coding of tone. Table A4 includes the results of these analyses. We find a high level of consistency in the coding of ad tone with a minimum level of reliability using Krippendorf’s alpha greater than 0.75.

Empirical Results

Contrast versus Attack Messages, Mixed-Gender Races

We predict in H1a that female candidates will use contrast messages more than male candidates, and we argue that this is a strategic choice designed to manage the potential of facing a gendered backlash for violating feminine norms. H1b argues that female candidates will be less likely to air attack ads, but male candidates will be more likely to air attack ads. We expect male candidates to worry less about facing a backlash as voters do not hold them to gendered expectations about “niceness.”³ We estimated two separate logistic regression models. The first model predicts the likelihood of a candidate airing a contrast ad and the second model predicts the likelihood of a candidate airing an attack ad. Our contrast and attack ad outcome variables are each coded as 0 or 1 with a value of 1 indicating that an ad is either a contrast or an attack ad, respectively. We display the full model results in Web Appendix C, and we display the predicted probabilities of candidate gender on airing a contrast or an attack ad in Figure 1.

Figure 1.

The effect of candidate gender on the use of negativity.

Female Candidates and Contrast Ads

We find, in accordance with our prediction in H1a, that female candidates are more likely to air contrast ads relative to male candidates, and this is a statistically significant effect (β = .125, SE = 0.064, p = .052). The predicted probability of a female candidate airing a contrast ad is 0.215 (or roughly 22%) while the predicted probability of a male candidate airing a contrast ad is 0.197 (or roughly 20%).

Male Candidates and Attack Ads

Next, we turn to attack ads. We expected men to be more willing than women to absorb the risks of running attack ads in H1b. We find no support for H1b. Women and men are just as likely to air attack messages (β = −.166, SE = 0.114, p = .150).⁴

Gendered Trait Negativity, Mixed-Gender Races

H2a argues that female candidates will use masculine traits in contrast and attack ads to target their male opponents. H2b argues that male candidates will use feminine traits in contrast and attack messages to target their female opponents. We estimated separate models predicting whether the candidate aired a contrast ad and an attack ad, and we included an interaction between candidate gender and our four trait variables (positive feminine, negative feminine, positive masculine, negative masculine). We use the predicted probabilities to test our second set of hypotheses, and we display the predicted probabilities in Figure 2 with the full tables in Web Appendix C. We organize our discussion of the results in the following order: First, we discuss the use of masculine traits in contrast and then attack messages across candidate gender. Next, we turn to the use of feminine traits in contrast and then attack messages for female and male candidates.

Figure 2.

The use of feminine and masculine traits in contrast and attack messages.

Female Candidates and Masculine Traits

We predicted in H2a that female candidates, more than male candidates, would sponsor ads that use positive and negative masculine traits more in contrast and attack messages. The top left panel of Figure 2 displays the predicted probabilities for positive masculine traits. The predicted probability of a female candidate using positive masculine traits that match good leadership traits in contrast messages is 0.21 (SE = .01) while male candidates have a slightly lower probability at .19 (SE = .01) of positive masculine traits, but this difference is not significant, p = .100. The probability of women using positive masculine traits in attack messages is 0.25 (SE = .01) and the probability for a male candidate is 0.27 (SE = .01), and these are statistically significant differences, p = .044, but are in the opposite direction we expected in H2a. Male candidates are slightly more likely to use positive masculine traits in attack messages against women.

The top right panel in Figure 2 displays the predicted probabilities of using negative masculine traits. These are masculine traits that do not align with traits that are considered good leadership qualities (e.g., being corrupt). We find no differences across candidate gender in the use of negative masculine traits in contrast ads, p = .425 or attack ads, p = .467. Thus far, we find no support for H2a.

To summarize, we do not find that women are more likely to use masculine traits in contrast or attack messages contrary to our prediction in H2a. The only significant difference we find is that male candidates are somewhat more likely to use positive masculine traits in attack messages.

Male Candidates and Feminine Traits

H2b predicted that male candidates would use positive and negative feminine traits in contrast and attack messages against female candidates. We start with the use of feminine traits that are generally seen as positive traits but can provoke a role-incongruity effect. These results are in the bottom left panel of Figure 2. The probability of female candidates using a positive feminine trait in a contrast message is 0.19 (SE = .01) and the probability of male candidates using positive feminine traits is 0.16 (SE = .01), a difference that does not reach statistical significance using p < .05 as the threshold, p = .091. We find that women have a higher probability, at .31 (SE = .01), of using positive feminine traits in attack messages while men have a lower probability at .28 (SE = .01), p = .010. These differences show that women use positive feminine traits more than men, and this is the opposite of what we expected in H2b.

With negative feminine traits in contrast ads, shown in the bottom right panel of Figure 2, we find that women’s predicted probability of a negative feminine trait message in a contrast ad is 0.50 (SE = .03) while for men the predicted probability is 0.44 (SE = .01), though the difference falls short of a p < .05 significance threshold at p = .061. We find no differences across candidate gender in the use of negative feminine traits in attack ads, p = .924.

Our findings on the use of feminine traits reveal no support for H2b. We find that when there are differences in the use of positive or negative feminine traits it is women who are incorporating these traits in contrast messages more than men. In attack messages, we find no differences across candidate gender in the use of positive or negative feminine traits. Contrary to our prediction, we do not find that men use feminine traits in attack and contrast messages more than women.

All-Women Races

We argue that the gender of a woman’s opponent will affect how women use negativity and the types of traits that women use in negative messages. We expect all-women races to use more attack ads relative to mixed-gender races. Our logic is that women will worry less about voters using feminine stereotypes to evaluate their behavior and this should free them to use attack messages (H3a). We also expect women to use more feminine traits in negative messages (H3b). We walk through our tests of each of these predictions in the following sections.

Attack Ads and All-Women Races

Hypothesis 3a argued all-women races will have more attack messages as women are less constrained by feminine norms without a male opponent. We estimated logistic regression models predicting the use of contrast and attack messages based on the gender composition of a race to test H3a. The full results are in Web Appendix C, see Table A8. We find no differences across the gendered composition of an election, in the use of contrast messages, p = .556, or attack messages, p = .98. These findings suggest no support for H3a. The gender composition of an election does not change the type of negative message a woman sponsors.

Feminine Traits and All-Women Races

Hypothesis 3b argued that women running in all-women races will use feminine traits in contrast and attack messages more than women in mixed-gender races. We estimated logistic models predicting the use of feminine traits in contrast and attack messages. We used the same set of controls from the previous analyses, and we included an interaction term between the election composition variable (mixed-gender vs. all-women), candidate gender, and the gendered strategy used in the campaign ad. We start with the results showing differences in the use of positive feminine traits in contrast and attack messages and then the use of negative feminine traits in contrast and attack messages.

Figure 3 shows the results for the use of gendered traits in contrast ads. The bottom two panels show the results for feminine traits across the gender composition of the election. The bottom left panel of Figure 3 shows the results for positive feminine traits. There are no differences in the use of positive feminine traits in contrast ads based on the gender composition of the race, p = .544. Turning to positive feminine traits in attack ads, we find that women running in all-women contests have a 0.20 (SE = .02) predicted probability of relying on positive feminine traits in attack ads while mixed-gender contests have a .25 (SE = .01) predicted probability of using these same traits, and this difference is statistically significant, p = .029. These findings do not support our expectation that women in all-women races would use feminine traits more than candidates in mixed-gender races. We find an increased used of positive feminine traits by women in attack messages when the opponent is a man, the opposite of our prediction.

Figure 3.

Gendered traits in attack ads, mixed-gender versus all-women races.

The bottom right panel of Figure 3 shows the results for the use of negative feminine traits. In all-women contests, the predicted probability of candidates using a negative feminine trait in a contrast ad is 0.43 (SE = .05) and the predicted probability for candidates using these same traits in mixed-gender contests is 0.55 (SE = .03), and this difference is statistically significant, p = .0035. However, we find that women in all-women races use negative feminine traits in attack messages, with a 0.43 (SE = .03) predicted probability more than women in mixed-gender elections who have a .33 (SE = .02) predicted probability of using feminine traits against a male opponent, p = .009. These findings provide partial support for H3b.

We did not have predictions about the use of masculine traits in all-women races; but we conducted these analyses displayed in the top two panels of Figure 3 to provide an instructive point of comparison across the gender composition of an election. We found no statistically significant differences based on the gender composition of the race in the use of positive or negative masculine traits.

Robustness Checks

We find some differences in the effects of party and incumbency on the differences between women and men airing contrast and attack ads with, overall, incumbents and Democratic women being less likely to air these types of ads (full results in Web Appendix D). We also compared by level of office, House, Senate, or gubernatorial race (see Table A2 for a full set of comparisons by level of office). We find some small differences in the way that women use positive and negative masculine traits for Senate races and gubernatorial races relative to House races, but these patterns are not always consistent. We also examined the use of both feminine and masculine traits in a single ad. We found these mixed-trait ads to be relatively rare for attack ads. Mixed-trait strategies are more common in contrast ads with about 20% emphasizing feminine and masculine traits in a single message, but we found female candidates were no less likely to air these types of ads relative to male candidates.

Discussion and Conclusion

We find that female candidates are more likely than male candidates to air contrast ads and just as likely to air attack ads (H1a and H1b). We do not find that women are more likely to use masculine traits relative to male candidates (no support for H2a) and we find the women and men are just as likely to use feminine traits in contrast and attack messages (no support for H2b). Finally, H3a argued that all-women races will use more attack messages relative to mixed-gender races, but we do not find this effect (no support for H3a). We find partial support for H3b where we find that all-women races use more negative feminine traits than mixed-gender races but women in mixed-gender races use positive feminine traits more in contrast and attack messages. In this final section, we document some of the limitations of our research, potential pathways for future scholarship, and the implications of our research.

Missing from our research is whether contrast messages are more effective than attack messages in bolstering perceptions that women have the masculine qualities that align most strongly with political leadership. Research on how voters respond to campaign negativity suggest that voters dislike any kind of negative campaign message and this might apply to contrast ads as well (Kahn & Kenney, 2004). Scholarship comparing voter responses to contrast and attack ads finds support for our argument that voters see contrast ads more favorably than attack ads (Meirick, 2002). Future work should follow up on our analyses with experimental tests documenting differences in how people respond to the use of feminine and masculine traits in contrast and attack messages when those messages are sponsored by a female candidate or a male candidate.

Our tests of gendered trait negativity centered on data from American elections. It is possible that the dynamics of gendered trait negativity will affect how women launch negative messages and the types of negative messages used against women in other countries with competitive electoral systems. Indeed, Van der Pas and Aaldering (2020) found that countries using proportional systems of representation may be more likely to exhibit gendered biases toward women compared to the majoritarian system used in the U.S. Research suggests that many of the gendered dynamics related to gender role incongruity affect how women in political leadership roles are seen in other countries (Bast et al., 2022; Boussalis et al., 2021; Holman et al., 2022). Analyses of negativity in Britain, the Netherlands, and Germany suggests that female party leaders may be more likely to use negativity relative to male party leaders (Walter, 2013). Examining how female party leaders use gendered traits in negative messages in country contexts outside the U.S. is an important next step for identifying how women develop messages that maximize their chance for electoral success around the globe.

Our study only examined negativity communicated through televised campaign ads. Campaign negativity can be communicated through a variety of other platforms including social media (H. K. Evans et al., 2017; Phillips, 2021) or at in person political rallies or televised debates (Maier & Renner, 2018). Future work should extend our research to examine whether the patterns of negativity uncovered in television ads replicates on other platforms with our findings here. Fowler et al. (2021) find important differences in the content of messages from candidates who use television ads relative to digital advertising. The use of negativity by women may also vary across platforms based on the affordances provided by a particular medium (social media vs. television vs. in-person rallies). Moreover, we only examined negative messages sponsored by the candidate. Research increasingly shows that candidates outsource negativity to political action committees (or PACs)⁵ (Dowling & Wichowsky, 2015) but the differences in the use of PACs on behalf of or against female and male candidates are not entirely clear. This is an area for future work to expand on gendered dynamics of negative campaigning.

Our research finds that all-women races do not see candidates using more feminine traits in contrast or attack messages. Important to note is that all-women races are a relatively recent phenomenon with 11 all-women races in 2010 jumping to 33 in 2018, even a significant increase from 17 in 2016 to 33 in 2018 in just 2 years. It is important to extend this research to better clarify how the strategic use or avoidance of gender stereotypes shifts in response to the gender of their opponent. Other contextual factors that could shift the use of gendered traits in attack or contrast messages included the race or ethnicity of an opponent or the sponsor of the ad (Brown & Lemi, 2021; Cargile, 2021). Future work should examine how the gendered and racialized context of a campaign shifts the strategic communication choices made by candidates.

What do our findings mean for how female candidates should campaign to maximize their chances for electoral victory? Female candidates facing the tough electoral challenges they often face (Branton et al., 2018) must behave strategically in their campaign messages and our research suggests women might be best positioned for electoral success with contrast over attack ads. Of course, our research also suggests that women need to pay attention to their strategic context, namely, whether they are in an all-women or mixed-gender race. Our findings also hold implications for the gendered dynamics of politics. The use of feminine traits to criticize a woman reinforces that there is little room for feminine traits and qualities in politics (Sweet-Cushman, 2022), and that masculine traits are the most valuable in political leaders. Feminine traits should carry more currency in politics as voters routinely express a desire for more compromise and consensus, a stereotypically feminine behavior (Eagly & Carli, 2007). The lack of value of feminine traits reinforces the perception of politics as masculine.

Women’s political representation is at its highest level ever in the U.S., and in many other Western, democratic countries. In the American context, the ability for women to win elections often depends on their ability to manage gendered expectations about the tone of message that is most appropriate for women to use in their advertising and gendered perceptions about the feminine and masculine traits that women do and do not have. Male candidates do not have to worry about facing a gendered backlash for airing negativity and male candidates do not have to worry about whether voters see them as having masculine traits. Women’s success at managing gendered trait negativity is critical to women’s success at the ballot and increasing women’s political representation.

Supplemental Material

sj-docx-1-crx-10.1177_00936502231179290 – Supplemental material for How Women Attack: Candidate Reliance on Feminine and Masculine Traits in Campaign Negativity

Supplemental material, sj-docx-1-crx-10.1177_00936502231179290 for How Women Attack: Candidate Reliance on Feminine and Masculine Traits in Campaign Negativity by Nichole M. Bauer, Caley Hewitt and Pamela Labbe in Communication Research

Footnotes

Acknowledgements

The authors are indebted to the student researchers in the Gender & Politics Research Lab who helped with the coding of the gendered traits in campaign ads necessary for this project.

Contributions

The research question and plan was conceptualized by Nichole Bauer. Caley Hewitt developed the section about negativity in all-women race, and all the authors worked on the analyses, results, and the concluding discussion.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Nichole M. Bauer

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biographies

Nichole M. Bauer is an associate professor of political communication in the Department of Political Science and the Manship School of Mass Communication at Louisiana State University. Her research investigates how gender stereotypes affect voter decision-making and the communication strategies female candidates develop.

Caley Hewitt is a doctoral candidate at the Manship School of Mass Communication at Louisiana State University. Her research examines how female candidates develop strategic messages in all-women electoral contests.

Pamela Labbe received her PhD from the Manship School of Mass Communication at Louisiana State University in 2022. Pamela’s research investigates the conditions under which expressions of partisan identity shift and how individuals use social media to express partisan identities and engage in political discussion.

References

Aaldering

Van Der Pas

D. J.

(2020). Political leadership in the media: Gender bias in leader stereotypes during campaign and routine times. British Journal of Political Science, 50(3), 911–931. https://doi.org/10.1017/s0007123417000795

Banda

K. K.

Windett

J. H.

(2021). Candidate appearance in campaign advertisements. Electoral Studies, 70, 102275. https://doi.org/10.1016/j.electstud.2020.102275

Barnes

T. D.

Beaulieu

(2019). Women politicians, institutions, and perceptions of corruption. Comparative Political Studies, 52(1), 134–167. https://doi.org/10.1177/0010414018774355

Bast

Oschatz

Renner

A.-M.

(2022). Successfully overcoming the “Double Bind”? A mixed-method analysis of the self-presentation of female right-wing populists on Instagram and the impact on voter attitudes. Political Communication, 39, 358–382. https://doi.org/10.1080/10584609.2021.2007190

Bauer

N. M.

(2015). Emotional, sensitive, and unfit for office? Gender stereotype activation and support female candidates. Political Psychology, 36(6), 691–708. https://doi.org/10.1111/pops.12186

Bauer

N. M.

(2017). The effects of counter-stereotypic gender strategies on candidate evaluations. Political Psychology, 38(2), 279–295. https://doi.org/10.1111/pops.12351

Bauer

N. M.

(2020a). The Qualification gap: Why Women must be More qualified than men to win political office. Cambridge University Press.

Bauer

N. M.

(2020b). Shifting Standards: How voters evaluate the qualifications of female and male candidates. The Journal of Politics, 82(1), 1–12. https://doi.org/10.1086/705817

Bauer

N. M.

Kalmoe

N. P.

Russell

E. B.

(2022). Candidate aggression and gendered voter evaluations. Political Psychology, 43(1), 23–43.

10.

Bauer

N. M.

Santia

(2022). Going feminine: Identifying how and when female candidates emphasize feminine and masculine traits on the campaign trail. Political Research Quarterly, 75(3), 691–705. https://doi.org/10.1177/10659129211020257

11.

Bauer

N. M.

Santia

(2023). Gendered times: How gendered contexts shape campaign messages of female candidates. Journal of Communication. Advance online publication. https://doi.org/10.1093/joc/jqac052

12.

Biernat

Manis

(1994). Shifting Standards and stereotype-based judgments. Journal of Personality and Social Psychology, 66(1), 5–20.

13.

Bishin

B. G.

Stevens

Wilson

(2006). Character counts?: Honesty and fairness in election 2000. Public Opinion Quarterly, 70(2), 235–248.

14.

Boussalis

Coan

T. G.

Holman

Muller

(2021). Gender, candidate emotional expression, and voter reactions during televised debates. American Political Science Review, 115(4), 1242–1257. https://doi.org/10.1017/s0003055421000666

15.

Branton

English

Pettey

Barnes

T. D.

(2018). The impact of gender and quality opposition on the relative assessment of candidate competency. Electoral Studies, 54, 35–43. https://doi.org/10.1016/j.electstud.2018.04.002

16.

Brown

Lemi

D. C.

(2021). Sister Style: The Politics of appearance for black women political elites. Oxford University Press.

17.

Bystrom

D. G.

(2006). Advertising, websites, and media coverage: Gender and communication along the campaign trail. In Carroll

S. J.

Fox

R. L.

(Eds.), Gender and elections, shaping the future of American politic (pp. 168–188). Cambridge University Press.

18.

Cargile

I. A. M.

(2021). Stereotyping Latinas: Candidate gender and ethnicity on the political stage. Politics, Groups, and Identities. Advance online publication. https://doi.org/10.1080/21565503.2021.1946097

19.

Cassese

E. C.

Holman

M. R.

(2018). Party and gender stereotypes in campaign attacks. Political Behavior, 40(3), 785–807. https://doi.org/10.1007/s11109-017-9423-7

20.

CAWP. (2020). Women in Elective Office 2020. Author.

21.

Clayton

O’Brien

D. Z.

Piscopo

J. M.

(2019). All male panels? Representation and democratic legitimacy. American Journal of Political Science, 63(1), 113–129. https://doi.org/10.1111/ajps.12391

22.

Craig

S. C.

Rippere

P. S.

(2016). He said, she said: The impact of candidate gender in negative campaigns. Politics and Gender, 12, 391–414.

23.

Deason

Greenlee

J. S.

Langner

C. A.

(2015). Mothers on the campaign trail: Implications of politicized motherhood for women in Politics. Politics, Groups, and Identities, 3(1), 133–148.

24.

Ditonto

(2019). Direct and indirect effects of prejudice: Sexism, information, and voting behavior in political campaigns. Politics, Groups, and Identities, 7, 590–609. https://doi.org/10.1080/21565503.2019.1632065

25.

Dittmar

(2015). Navigating gendered terrain: Stereotypes and strategy in political campaigns. Temple University Press.

26.

Dowling

C. M.

Wichowsky

(2015). Attacks without consequence? Candidates, parties, groups, and the changing face of negative advertising. American Journal of Political Science, 59(1), 19–36.

27.

Eagly

A. H.

Carli

L. L.

(2007). Through the Labyrinth: The truth about how women become leaders. Harvard Business School Press.

28.

Eagly

A. H.

Karau

S. J.

(2002). Role congruity theory of prejudice toward female leaders. Psychological Review, 109(3), 573–598. https://doi.org/10.1037/0033-295x.109.3.573

29.

Evans

(2016). Do Women only Talk about "Female Issues"? Gender and issue discussion on Twitter. Online Information Review, 40(5), 660–672. https://doi.org/10.1108/oir-10-2015-0338

30.

Evans

H. K.

Clark

J. H.

(2016). "You tweet like a girl!" How female candidates campaign on Twitter. American Politics Research, 44(2), 326–352. https://doi.org/10.1177/1532673x15597747

31.

Evans

H. K.

Smith

Gonzales

Strouse

(2017). Mudslinging on Twitter during the 2014 election. Social Media + Society, 3, 1–9. https://doi.org/10.1177/2056305117704408

32.

Fowler

E. F.

Franz

M. M.

Martin

G. J.

Peskowitz

Ridout

T. N.

(2021). Political Advertising Online and offline. American Political Science Review, 115(1), 130–149. https://doi.org/10.1017/s0003055420000696

33.

Fowler

E. F.

Franz

M. M.

Ridout

T. N.

(2014). Political Advertising in 2010 (Version 1.2). Wesleyan Media Project.

34.

Fowler

E. F.

Franz

M. M.

Ridout

T. N.

(2015). Political Advertising in 2012 (Version 1.1). Wesleyan Media Project.

35.

Fowler

E. F.

Franz

M. M.

Ridout

T. N.

(2017). Political Advertising in 2014 (Version 1.0 [dataset]). Wesleyan Media Project.

36.

Fowler

E. F.

Franz

M. M.

Ridout

T. N.

Baum

L. M.

(2019). Political Advertising in 2016 (Version 1.0 [dataset]). Wesleyan Media Project.

37.

Fowler

E. F.

Franz

M. M.

Ridout

T. N.

Baum

L. M.

(2020). Political Advertising in 2018 (Version 1.0 [dataset]). Wesleyan Media Project.

38.

Franz

M. M.

Freedman

Goldstein

K. M.

Ridout

T. N.

(2007). Campaign Advertising and American Democracy. Temple University Press.

39.

Fridkin

K. L.

Kenney

P. J.

Woodall

G. S.

(2009). Bad for men, better for women: The impact of stereotypes during negative campaigns. Political Behavior, 31, 53–77.

40.

Garramone

G. M.

Atkin

C. K.

Pinkleton

B. E.

Cole

R. T.

(1990). Effects of negative political advertising on the political process. Journal of Broadcasting & Electronic Media, 34(3), 299–311.

41.

Geer

J. G.

(2006). In defense of negativity. University of Chicago Press.

42.

Gordon

Shafie

D. M.

Crigler

A. N.

(2003). Is negative advertising effective for female candidates?: An experiment in voters’ use of gender stereotypes. The Harvard International Journal of Press/Politics, 8, 35–53.

43.

Herrnson

P. S.

Lucas

J. C.

(2006). The fairer sex? Gender and negative campaigning in U.S. elections. American Politics Research, 34(1), 69–94.

44.

Holman

M. R.

Merolla

J. L.

Zechmeister

E. J.

(2022). The curious case of Theresa May and the public that did not rally: Gendered reactions to terrorist attacks can cause slumps not bumps. American Political Science Review, 116(1), 249–264. https://doi.org/10.1017/s0003055421000861

45.

Holman

M. R.

Merolla

J. L.

Zechmeister

E. J.

Wang

(2019). Terrorism, gender, and the 2016 U.S. Presidential election. Electoral Studies, 61, 1102033–1102038. https://doi.org/10.1016/j.electstud.2019.03.009

46.

Jamieson

K. H.

(1995). Beyond the double bind: Women and leadership. Oxford University Press.

47.

Kahn

K. F.

Kenney

P. J.

(2004). No holds barred: Negativity in US Senate Campaigns. Pearson Education Inc.

48.

Kang

Fowler

E. F.

Franz

M. M.

Ridout

T. N.

(2018). Issue Consistency? Comparing television advertising, tweets, and E-mail in the 2014 Senate Campaigns. Political Communication, 35(1), 32–49. https://doi.org/10.1080/10584609.2017.1334729

49.

Kanthak

Woon

(2015). Women don’t run? Election aversion and candidate entry. American Journal of Political Science, 59(3), 595–612.

50.

Koenig

A. M.

Eagly

A. H.

Mitchell

A. A.

Ristikari

(2011). Are leader stereotypes masculine? A meta-analysis of three research paradigms. Psychological Bulletin, 137(4), 616–642.

51.

Krupnikov

Bauer

N. M.

(2014). The relationship between campaign negativity, gender and campaign context. Political Behavior, 36(1), 167–188. https://doi.org/10.1007/s11109-013-9221-9

52.

Lau

R. R.

Pomper

G. M.

(2004). Negative campaigning: An analysis of US Senate elections. Rowman & Littlefield Publishers, Inc.

53.

Lizotte

M.-K.

(2020). Gender differences in public opinion: Values and political consequences. Temple University Press.

54.

Maier

Renner

A.-M.

(2018). When a man meets a woman: Comparing the use of negativity of male candidates in single- and mixed-gender televised debates. Political Communication, 35(3), 433–449. https://doi.org/10.1080/10584609.2017.1411998

55.

Meeks

Domke

(2016). When politics is a woman’s Game: Party and gender ownership in woman-versus-woman elections. Communication Research, 43(7), 895–921. https://doi.org/10.1177/0093650215581369

56.

Meirick

(2002). Cognitive responses to negative and comparative political advertising. Journal of Advertising, 31(1), 49–62. https://www-jstor-org.web.bisu.edu.cn/stable/4189207

57.

Melich

(2005). From the trenches: Attacking first-time women candidates for congress. Journal of Women, Politics & Policy, 27(1-2), 85–107. https://doi.org/10.1300/j501v27n01_06

58.

Miller

A. H.

Wattenberg

M. P.

Malanchuk

(1986). Schematic assessments of presidential candidates. American Political Science Review, 80(2), 521–540.

59.

Mueller

C. M.

(1986). Nurturance and mastery: Competing qualifications for women’s access to high public office. In Moore

Spitze

G. D.

(Eds.), Women and politics: Activism, attitudes, and office-holding (pp. 211–232). JAI Press.

60.

Pearson

McGhee

(2013). What it takes to win: Questioning "gender neutral" outcomes in U.S. house elections. Politics and Gender, 9, 439–462.

61.

Petrocik

J. R.

(1996). Issue ownership in presidential elections, with a 1980 Case Study. American Journal of Political Science, 40(3), 825–850. https://doi.org/10.2307/2111797

62.

Phillips

J. B.

(2021). Negative political communication on social media and the gender gap: A Study of men’s and women’s reactions to presidential candidate attacks on Facebook in 2012 and 2016. Politics and Gender, 17, 454–482.

63.

Preece

Stoddard

(2015). Why women don’t run: Experimental evidence on gender differences in political competition aversion. Journal of Economic Behavior & Organization, 117, 296–308.

64.

Prentice

D. A.

Carranza

(2002). What women and men should Be, shouldn’t be, are allowed to be, and don’t have to be: The contents of prescriptive gender stereotypes. Psychology of Women Quarterly, 26(4), 269–281. https://doi.org/10.1111/1471-6402.t01-1-00066

65.

Rohrbach

Fiechtner

Schönhagen

Puppis

(2020). More than just gender: Exploring contextual influences on media bias of political candidates. The International Journal of Press/Politics, 25, 692–711.

66.

Schneider

M. C.

(2014). The effects of gender-bending on candidate evaluations. Journal of Women, Politics & Policy, 35, 55–77. https://doi.org/10.1080/1554477x.2014.863697

67.

Schneider

M. C.

Bos

A. L.

(2014). Measuring stereotypes of female politicians. Political Psychology, 35(2), 245–266. https://doi.org/10.1111/pops.12040

68.

Schneider

M. C.

Bos

A. L.

(2019). The application of social role theory to the study of gender in politics. Political Psychology, 40(S1), 173–213. https://doi.org/10.1111/pops.12573

69.

Skaperdas

Grofman

(1995). Modeling negative campaigning. American Political Science Review, 89(1), 49–61.

70.

Sweet-Cushman

(2016). Gender, risk assessment, and political ambition. Politics and the Life Sciences, 35(2), 1–17. https://doi.org/10.1017/pls.2016.13

71.

Sweet-Cushman

(2022). Legislative vs. executive political offices: How gender stereotypes can disadvantage women in either office. Political Behavior, 44, 411–434. https://doi.org/10.1007/s11109-021-09721-x

72.

Van der Pas

D. J.

Aaldering

. (2020). Gender differences in political media coverage: A meta-analysis. Journal of Communication, 70(1), 114–143. https://doi.org/10.1093/joc/jqz046

73.

Wagner

K. M.

Gainous

Holman

M. R.

(2017). I am woman, hear me tweet! Gender differences in Twitter use among congressional candidates. Journal of Women, Politics & Policy, 38, 430–455. https://doi.org/10.1080/1554477x.2016.1268871

74.

Walter

A. S.

(2013). Women on the battleground: Does gender condition the use of negative campaigning? Journal of Elections Public Opinion and Parties, 23(2), 154–176.

75.

Wesleyan Media Project. (2018). Digital spending in senate races varies. Author.

76.

Windett

J. H.

(2014). Gendered campaign strategies in U.S. elections. American Politics Research, 42(4), 628–655. https://doi.org/10.1177/1532673x13507101

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.13 MB