When Politics Is a Woman’s Game

Abstract

Research on the interplay of gender and political party in voters’ candidate evaluations has long focused on all-male elections and more recently on mixed-gender elections. This study takes the next theoretical step and focuses on woman-versus-woman elections. Specifically, we examine political party- and gender-based “ownerships” of political issues and character traits in the context of female-only elections. With an experimental design, adult participants were randomly assigned to read news articles that presented either two Republican or two Democratic women competing for Governor. Candidates were presented as “owning” stereotypically masculine or feminine issues and traits. Findings show that self-identified Democrats and Republicans eschewed the so-called masculine candidate, and preferred instead a partisan woman who created a gender balance of masculinity and femininity.

Keywords

women candidates gender ownership party ownership electability candidate evaluation

When women first started running for political office, they were often told they should “run like a man,” because a women’s place for much of American history has been seen as outside the political arena. Specifically, women have been located in the private domain, keepers of home and hearth, while men have been seen as the guardians of the public domain, including politics (Braden, 1996; Sanbonmatsu, 2004). This history has fostered a male-centric cultural conception of politics in which masculine values and political values are often interchangeable (e.g., Braden, 1996; Herrnson, Lay, & Stokes, 2003; Huddy & Terkildsen, 1993a). In the words of masculinity scholar Raewyn Connell (2005), “public politics on almost any definition is men’s politics” (p. 204). This traditional understanding of politics still has a foothold in contemporary society, and brings with it a set of crucial expectations. For example, women candidates are routinely asked how they will manage being a mother and a politician—a question that suggests these two roles do not easily mix. Gender, however, does not exist in isolation. Most women in U.S. politics run as Democrat or Republican, which brings a set of partisan expectations—some of which align with their gender and some of which conflict. These converging and diverging expectations present a nuanced obstacle for women candidates. Dolan (2014) and King and Matland (2003) have shown that a woman’s party identification can shape the impact of gender on candidate evaluations. Such work indicates the importance of examining the intertwining of gender and political party affiliation in voters’ evaluative processes.

Scholarship at the crossroads of these areas provides some insight. Most research devoted to examining party and voter expectations has focused on all-male elections (e.g., Ansolabehere & Iyengar, 1994; Bélanger & Meguid, 2008; Petrocik, 1996), but a growing body of work has begun to explore how party and gender interact to affect evaluations in mixed-gender elections—those with at least one female and one male candidate (e.g., Banwart, 2010; Dolan, 2014; Hayes, 2011). The increasing prevalence of female candidates makes this work vital, and we take another step by examining gender impressions, partisan identities, and voter expectations in the context of woman-versus-woman elections. Specifically, we conducted an experiment in which news articles reported on two women competing for a gubernatorial office. The design consisted of parallel Republican and Democratic primaries—thereby holding constant both gender and party affiliation. The women candidates were described as emphasizing and embodying two key elements of voter assessment: political issues and character traits. These elements serve as the basis for scholarship on party- and gender-based ownership, and dictate what voters expect from certain political identities. With women increasingly confronting and overcoming the long-standing tradition of prototypical male politicians, this study attempts to move scholarship forward in examining how women and their partisan standings fit and expand popular notions of electability.

Woman-Versus-Woman Elections

We focus on woman-versus-woman elections for three reasons. First, our aim is to expand on existing scholarship on the interplay of gender and party identification in varying electoral settings: This has focused overwhelmingly on all-male elections due to their history and abundance in the United States, and a growing body of work has examined mixed-gender elections.¹ Much of this research will be explored in-depth in the following sections, and we seek to thoroughly examine these elements in this relatively new electoral setting. Second, all-female elections are on the rise and this trend is likely to continue. The first all-female congressional or gubernatorial general election occurred in 1944, with Clare Booth Luce competing against Margaret Connor for a U.S. House of Representatives seat in Connecticut (The Center for American Women and Politics [CAWP], 2012a). In recent years especially, all-female primary and general elections have increased, and several have been historic (CAWP, 2012a). Recent examples include all-female general elections in 2010 in Oklahoma and New Mexico, with both states electing their first female Governor, and in 2012 in New Hampshire, where Maggie Hassan beat fellow Democrat Jacalyn Cilley in the primary and went on to become the second woman to hold this gubernatorial office. Furthermore, Palmer and Simon (2005) analyzed over six decades of elections and found that as women ran for the U.S. House of Representatives, one result was a general increase in the participation of female candidates in their own party as well as the opposition party. It seems plausible, therefore, that as more women run for office, they will increasingly be competing against another woman.

Third, woman-versus-woman elections are rich in conceptual intrigue. Most notably, these elections create a new dynamic for women because the “political novelty” of their gender becomes shared with their opponent, which requires journalists and voters to move beyond simplified schema of women seeking to enter a highly masculine arena. This shared novelty also means that both women have access to being portrayed as and perceived as change-agents. Because women tend to be seen as novel additions to politics, they are often also viewed as outsiders in the political arena (Braden, 1996; Heldman, Carroll, & Olson, 2005). This status can be beneficial if voters are looking for a change from the status quo. For example, in the 2010 gubernatorial elections, Republican and Democratic women were viewed as more likely to bring about change than male counterparts (Barbara Lee Family Foundation, 2010). Furthermore, news media and voters must consider the viability of not just women candidates generally, but ones with distinct partisan affiliations. This perspective is particularly interesting in light of an additional component. Namely, women politicians, regardless of party, are typically considered more liberal than men (Hayes, 2011; Huddy & Terkildsen, 1993b; McDermott, 1997). As a result, Republican women candidates are likely to be perceived as more liberal than their male Grand Old Party (GOP) counterparts, and Democratic women candidates are likely to be perceived as especially liberal. This perception may explain why in the woman-versus-woman gubernatorial elections in New Mexico and Oklahoma in 2010, for instance, Republican candidates outperformed the party identification margin—meaning they won more votes than their party holds in voter registration (Barbara Lee Family Foundation, 2010). Taken together, the rise of all-female elections and the complexity of such elections create a dynamic space for investigating how gender and party collide and affect contemporary notions of electability and voter evaluations.

With all of this in mind, we decided to focus on woman-versus-woman primaries for three reasons as well. First, most electoral studies focus on general elections, but candidates almost always have to win a respective party primary first. Thus, this study examines a crucial, yet under-examined hurdle for women seeking elected office. Second, research has shown that Republican and Democratic women have different rates of primary victory. CAWP (2013) analyzed partisan primaries for the U.S. House of Representatives from 1994 to 2012, finding that Democratic women had a higher win rate than Republican women, especially in 2010 and 2012. Therefore, looking across partisan identities in primaries seems important. Finally, controlling for both candidate gender and candidate party, which is only possible in a primary setting, enables us to zero in on the impact of political ownerships of issues and traits. In short, a focus on primaries is a valuable theoretical space.

Voter Evaluations

At the most basic level, the voter evaluative process is a matter of encountering information and deciding what to do with it. McGraw (2003) called this impression formation, which is a process in which individuals construct an impression of another person based on knowledge and inferences. Scholarship on the evaluative process is wide-ranging, spanning questions about what information is taken in and how it is stored, recalled, and acted on. For the purpose of this study, we focus on how a candidate’s gender and party interact in voters’ impressions when they are provided with two types of information about the candidates: political issues and character traits. Many initial forays in scholarship concerning how citizens should arrive at political decisions positively cast the ideal democratic voter as someone who relies solely on political issues in their evaluative process (for overview, see Popkin, 1991). This research focused on the politician as someone who pushes policies, and therefore, citizens should cast their votes based on issue affinity. Other works, though, have suggested that voters should use character traits or candidate persona in making political decisions. Barber (1972), for example, decreed that for voters the

First need is to see the man whole—not as some abstract embodiment of civic virtue, some scorecard of issue stands, or some reflection of a faction, but as a human being like the rest of us, a person trying to cope with a difficult environment. To that task he brings his own character. (pp. 3-4)

Glass (1985) added that issue stances and priorities can change, especially as politicians take office, and that an investment in character assessment is a “rational response of intelligent political observers to a political system in which policy positions are often cloudy and candidate attributes are far more concrete” (p. 531). A number of studies have professed a similar call for the importance of traits (e.g., Funk, 1999; McGraw, 2003; McLeod, Glynn, & McDonald, 1983). These studies do not necessarily indicate a lack of cohesion in this scholarly domain; rather, these works suggest that issues and traits are important in the impression formation process of voters, to varying degrees.

Ultimately, therefore, the voter evaluation process is complex, but issues and traits provide a solid starting point. In an extensive review of scholarship on voter assessment, Louden and McCauliff (2004) identified two key evaluative categories that align with these components. The first focuses on character traits and personal qualities, and includes assessments of how personable, genuine, and honest—that is, how likeable—the candidate appears to be. The next category is issue handling, and includes assessments of how capable, qualified, and experienced—that is, how competent—the candidate seems to be. Recent research has found that there is a correlation between being competent and being likeable for women candidates, a finding not present for men (Barbara Lee Family Foundation, 2012). Thus, a focus on the effects of issue competency and trait portrayal on electability may be especially important for women candidates. This study, therefore, uses the issue/trait dynamic as a springboard to examine the interaction of party and gender on voters’ evaluations of candidate electability.

Political Ownership of Issues and Traits

In the political arena Republicans and Democrats, men and women have developed political reputations—sometimes intentionally, sometimes not—for “owning” certain issues and character traits. These images provide a useful framework for examining how candidates and news media construct political identities, and how voters evaluate them. The idea of ownership emerges from work by Budge and Farlie (1983) and Petrocik (1996), which suggests that particular political parties are perceived to be better able to “handle” certain matters. “Handling” is the ability to address and resolve problems, and is derived from a “history of attention, initiative, and innovation toward these problems, which leads voters to believe that one of the parties (and its candidates) is more sincere and committed to doing something about them” (Petrocik 1996, p. 826). The idea of ownership offers valuable insight about some of the challenges and opportunities faced by women politicians, particularly in a partisan system, when seeking office.

In particular, political issues and character traits have divided along both party and gender lines, and candidates seek to employ these to their advantage when possible. Petrocik’s (1996) analysis of multiple national surveys showed that Republicans were viewed as more successful at handling issues of crime, foreign policy, defense, and economic matters, and Democrats were perceived as better able to handle social welfare issues, such as improving education and health care, protecting social security, and helping the elderly and the poor. A 2014 Gallup survey of self-identified partisans largely upheld these findings. With these perceptions in mind, candidates seek to highlight their party’s issue strengths while skirting their opponent’s issue assets (Petrocik, 1996; Petrocik, Benoit, & Hansen, 2003; cf. Ansolabehere & Iyengar, 1994; Bélanger & Meguid, 2008). While candidates cross party lines to discuss important issues, there are electoral advantages when they discuss party-congruent issues. For example, Benoit (2007) analyzed presidential TV ads from 1952 to 2004 and found that Democratic and Republican winners stressed party-owned issues more than Democratic and Republican losers. Furthermore, these ownerships have been magnified by news content (Petrocik et al., 2003). This combination has created a self-reinforcing loop for party-based issue ownership: Partisan candidates emphasize certain issues, news tends to echo these ownerships, voters then associate certain parties with certain issues and expect those parties to be better at handling such issues, and electoral outcomes commonly complete the circle.

Men and women are also perceived as owning certain political issues. Based on political ownership, Herrnson et al. (2003) developed the concept of “gender ownership” of issues, with men and women seen as differentially better at handling certain issues. For example, as supposed caretakers, female candidates are culturally viewed as better at handling “compassion issues,” such as health care and education, while, as supposed protectors and breadwinners, male candidates are viewed as better equipped to handle national security and the economy (Herrnson et al., 2003; Lawless, 2004; Sanbonmatsu & Dolan, 2009). This dynamic has produced the political shorthand of “masculine issues” and “feminine issues,” to suggest which issues are thought to be more easily claimed by male and female candidates, respectively. One may imagine that as women have increasingly moved outside of the household and into the workforce that such perceptions have changed, but recent research shows that stereotypes persist. For example, Lawless (2004) found that citizens view male candidates as better equipped to handle matters of national security and military crises, and perceive women candidates as better at handling issues such as aiding the poor and abortion. Sanbonmatsu and Dolan (2009) found similar results with women having a perceived advantage on education and men having a perceived advantage on crime. As with party ownership, news coverage often amplifies these gendered dynamics (e.g., Jalalzai, 2006; Kahn, 1994; Major & Coleman, 2008). Candidates do break away from these stereotypes, with men discussing stereotypically feminine issues and women discussing stereotypically masculine issues (e.g., Banwart & Kelly, 2013; Banwart & McKinney, 2005; Meeks, 2013), but research across various political offices, timeframes, and modes has shown that candidates often favor their perceived issue assets: Men tend to emphasize stereotypically masculine issues more than women, and women focus on feminine issues more than men (e.g., Bystrom, 2006; Iyengar, Valentino, Ansolabehere, & Simon, 1997; Kahn, 1993; Niven & Zilber, 2001). Gender-based issue ownerships, therefore, are embedded deeply in the U.S. political and news systems.

It is important to highlight that party and gender ownerships of issues intertwine in patterned ways. Republican issue ownerships align with stereotypical masculinity, while Democratic issue ownerships align with traditional notions of femininity. These pairings are more than coincidence; rather, they are rooted in history and culture. Since the 1970s, Democrats have generally supported the Equal Rights Amendment and women’s rights, creating an alignment among organized feminism, women voters, and Democrats, and thus supporting the association of Democrats with issues typically perceived as feminine (Sanbonmatsu, 2004). This coupling of support may explain in part why women are more likely to identify as Democrats, and since 1980 they have been more likely, than men, to vote for Democratic presidential candidates (CAWP, 2012b). Conversely, beginning in the late 1970s, Republicans switched from a decades-long pattern of support for the Equal Rights Amendment(ERA) to endorse a more traditional view of women and femininity (Domke & Coe, 2008; Sanbonmatsu, 2004). These dynamics on the conservative side of the partisan aisle have led the Republicans to be seen as the owners of stereotypically masculine matters.

An intertwining of party and gender issue ownerships carries several important implications. Most notably, research indicates that voters see stereotypically masculine issues as more important than so-called feminine issues, and voters perceive news coverage of masculine issues as positively contributing to a candidate’s viability (Huddy & Terkildsen, 1993a; Kahn, 1992; Meeks, 2012). As a result, women and Democratic candidates own issues that are generally seen as less important on the national stage, and therefore these candidates often have to “crossover” during elections to address issues seen as more important. This is the case for both Democratic men and women candidates, but an issue-based crossover is especially difficult for women because it cuts against the cultural notions of gender identity tied to the issues. At the same time, research has shown that women candidates can gain an advantage when they emphasize stereotypically feminine issues (Dolan, 1998; Herrnson et al., 2003; Paolino, 1995). As a result, Republican women candidates may seek to cross the gendered partisan lines on some issues. The hurdle for Republican women is to avoid being seen as more comfortable with issues considered more feminine, which are perceived to be less important to their party and the general public. These, then, are some of the dynamics faced by female candidates competing in partisan milieus.

Character traits are also thought to be owned by particular parties and genders, and create strong distinctions for voters regarding candidate personas. Building on the theory of issue ownership, Hayes (2005) developed a theory of “trait ownership” which proposed that issues and behaviors could be the basis for trait assessments, via an “issue-to-image transformation” (Louden & McCauliff, 2004). For example, because Republican politicians campaign heavily on taxes, business, and national defense, they carve out images of individualism, toughness, and strength; conversely, because Democratic politicians focus on social welfare issues and show a “willingness to reach out to those in need,” they create images of compassion and empathy (Hayes, 2005, p. 211). Both parties experienced a jump in the association between their party and traits in the 1980s, and this association has remained over the years (Winter, 2010). Specifically, Winter (2010) found that stereotypical masculine traits were roughly 7 times more likely to be mentioned as a reason to like Republicans than as a reason to like Democrats, and traditional feminine traits were almost 6 times more likely to be mentioned as a reason to like Democrats than as a reason to like Republicans. This work suggests a strong, enduring perceived link between parties and traits.

Furthermore, party and gender ownerships continue to overlap: this time on character traits, with Republicans and male candidates perceived as owning strength and assertiveness, while Democrats and female candidates are perceived as owning warmth and compassion (Banwart, 2010; Banwart & McKinney, 2005; Connell, 2005; Hayes, 2011). In addition, news coverage has highlighted these congruencies by giving men more stereotypically masculine trait coverage and women more stereotypically feminine trait coverage (Kahn, 1994). This trait-based alignment creates a potential disadvantage for women because voters tend to prefer stereotypical masculine traits in political leaders (Hayes, 2011; Huddy & Terkildsen, 1993a, 1993b). Overall, across issues and traits, Democratic women are in alignment based on gender and party expectations but face a cultural impression of “their” issues and traits as less important, whereas Republican women have to handle the incongruences of their identity because their party “owns” so-called masculine issues and traits, which are seen as more important in the political arena, but their gender culturally aligns with femininity.

Hypotheses and Research Questions

To examine these dynamics, we focused our experimental study on partisan primary contexts to see how self-identified Republican or Democratic participants evaluated female candidates of the same partisan stripes. To be completely clear, then, these expectations are for self-identified Republican participants’ perceptions of Republican candidates, and for self-identified Democratic participants’ perceptions of Democratic candidates. The first set of expectations compares perceptions among partisans of a candidate who focuses only on stereotypically masculine issues and traits versus one of three opponents: (a) a candidate who focuses only on stereotypical feminine issues and traits, (b) a gender-balanced candidate who emphasizes masculine issues and feminine traits, or (c) a gender-balanced candidate who emphasizes feminine issues and masculine traits. The second set of expectations focuses exclusively on the three experimental conditions, comparing perceptions among either self-identified Republicans or Democrats of the feminine-only candidate versus the gender-balanced candidates. This design allows us to examine a series of hypotheses and research questions.

We start with Democrats and then move to Republicans. Among voters who self-identify as Democrats and who were exposed to Democratic candidates, we expect an alignment of gender- and party-based ownerships that prompt two hypotheses: Self-identified Democratic participants will be more likely to favor the feminine-only candidate (Hypothesis 1a [H1a]) or either of the gender-balanced candidates (Hypothesis 1b [H1b]) over the masculine-only candidate as the best representative of their party and as more likeable, competent, and electorally successful. These two hypotheses predict that among Democratic participants, a candidate with at least some feminine political ties will trump the party- and gender-incongruent masculine-only candidate. But when we turn to the perceptions of the feminine-only versus gender-balanced candidates, the scenario among Democrats is tricky. The feminine-only candidate represents complete congruency between gender and party, yet the gender-balanced candidate stresses partial masculinity via issues or traits, which is important because voters commonly perceive masculine elements as more important. Given these complexities, we pose,

Research Question 1 (RQ1): Among self-identified Democrats, who will be considered the best representative of the party and as more likeable, competent, and electorally successful: the feminine-only or the gender-balanced candidate?

Among Republicans, the electoral dynamics quickly get complicated because gender- and party-based ownerships do not align for Republican female politicians. That is, if Republican women candidates seek to go fully feminine or fully masculine, they enter a risky terrain: seek perceptions of culturally affirmed gender authenticity while completely eschewing party expectations, or align wholly with party while risking being seen as gender-inauthentic. Therefore, rather than a hypothesis that would parallel H1a, we pose,

Research Question 2 (RQ2): Among self-identified Republican participants, who will be considered the best representative of the party and as more likeable, competent, and electorally successful: the feminine-only or the masculine-only candidate?

Once the gender-balanced candidate enters the electoral environment, we are on stronger theoretical ground, because now Republican voters will find a female candidate who at least partly meets both gender- and party-based expectations. Therefore, in a prediction that parallels H1b in focus but not in expected outcome, we predict that Republican participants will be more likely to view the gender-balanced candidate than either the masculine-only candidate (Hypothesis 2a [H2a]) or feminine-only candidate (Hypothesis 2b [H2b]) as the best representative of their party and as more likeable, competent, and electorally successful.

The last hypothesis compares perceptions of the candidates’ gender and party intertwinings between partisans. Scholarship makes clear that Democrats tether themselves more closely with traditional femininity and that Republicans typically favor masculinity, as we have argued so far. For our final hypothesis, we are interested in whether these cross-party differences in perceptions of female candidates are a small gap or resemble something more like a significant gulf. We suggest the latter, due to the hyperpolarized partisan divisions that exist in contemporary American politics (see Pew Research Center, 2014). We expect, therefore, that party-bound ownerships of gender will heavily diverge for Democrats and Republicans. Specifically, we predict that Democrats will be significantly more likely than Republicans to consider a candidate expressing femininity, either wholly or partly, as the best representative of their party and as more likeable, competent, and electorally successful (Hypothesis 3 [H3]).

Research Design and Methodology

To explore these expectations, we conducted an experiment with a sample of U.S. adults. Data were collected over 1 month in spring 2011. The experiment was conducted via Survey Monkey—a web-based platform that conducts surveys, polls, and experiments on a daily basis across 30 million people. Participants were also obtained via Survey Monkey. Specifically, the sample for the experiment was obtained from a pool of opt-in, adult participants. Once participants completed a survey from one of Survey Monkey’s clients, they were asked if they would be willing to answer additional questions. If they agreed, they were directed to our study and randomly assigned to a condition. We then focused on participants who, after random assignment, were exposed to a partisan candidate who matched their partisan identification—self-identified Republicans exposed to Republican candidates and likewise for Democrats—resulting in N = 368.² The use of Survey Monkey allowed us to create a “virtual lab” experiment, and to gain access to participants with greater demographic variance than experiments conducted with college-student populations.³ Average age was approximately 41 years (SD = 14.882), women accounted for 66% of the sample, and on political party identification 43% identified as “Strong” or “Lean” Republican and 57% as “Strong” or “Lean” Democrat. Political ideology was measured on a 7-point scale, from extremely liberal (1) to moderate (4) to extremely conservative (7) (M = 3.80, SD = 1.659). The sample identified heavily as White (79.5%), followed by Black (9.3%), Asian (4.6%), Hispanic/Latino/a (3.8%), Native American (0.8%), and finally, Other (1.9%). Education was measured on a 6-point scale, ranging from did not complete high school (1) to completed graduate degree (6) (Mdn = 4, completed college degree). Total family income was measured on a 7-point scale, ranging from less than $10,000 (1) to over $250,000 (7) (Mdn = 4, $50,000-$74,999).

The procedure of the experiment was as follows. Each participant was randomly assigned to read one of six news articles, and then asked to complete a short questionnaire. The news articles featured either two Republican or two Democratic women launching their campaigns for an open-seat primary race for Governor of Missouri for the 2012 elections. In all of the stimuli, the articles presented Linda Knowles, the control candidate, as holding the stereotypical masculine traits of independence and ambition and as focusing on the perceived masculine issue of crime/law and order. A second candidate Beth Rollins, the variant candidate, was reported as one of the following combinations: (a) holding the stereotypical feminine traits of warmth and honesty and focusing on the perceived feminine issue of education, (b) holding the stereotypical masculine traits of confidence and toughness and focusing on the perceived feminine issue of education, or (c) holding the stereotypical feminine traits of warmth and honesty and focusing on the perceived masculine issue of the economy (see the appendix for the full stimuli). Issues and trait emphases, as well as party identification, were the only components that varied between each stimuli; the remaining content was constant. Thus, we employed a 3 × 2 design: variant candidate’s expression of femininity (feminine issues and traits, feminine issues and masculine traits, and masculine issues and feminine traits) by control and variant candidates’ party affiliation (Republican, Democratic). In featuring two Republican or two Democratic women, the design held constant party affiliation and gender, and focused on voters’ evaluation of issue and trait emphases.⁴

To confirm that the “gendering” of issues and traits in the news articles were effective, we included two manipulation checks.⁶ We asked participants to choose either Knowles, who was always reported as accentuating masculine issues and traits, or Rollins, who was always reported as emphasizing at least some aspect of femininity, be it issues and/or traits, for the following questions: “Which candidate seems to have more masculine qualities?” (76.1% said Knowles was more masculine), and “Which candidate seems to have more feminine qualities?” (78.5% said Rollins was more feminine). Therefore, participants were much more likely to rate the candidate characterized as all-masculine—control candidate Knowles—as having more masculine qualities, and were much more likely to rate the candidate characterized as partially or fully feminine—variant candidate Rollins—as having more feminine qualities. These manipulation checks indicate that both participants were attentive to the content and processed it in a manner consistent with gender stereotypes. In short, the stereotypically gendered presentation of candidate issues and traits in the news articles was sufficiently recognized and recalled by participants.

For each dependent variable, respondents were asked to choose a candidate on a 6-point, Likert-type scale, with level of support for the control candidate Knowles represented as 1 to 3, ranging from strongly (1) to slightly (3), and level of support for the variant candidate Rollins represented as 4 to 6, ranging from slightly (4) to strongly (6). Each of the measures, therefore, assessed participants’ perceptions of the control, all-masculine candidate relative to perceptions of the variant candidate.

First, participants were asked to assess how well the candidates represented a political party. The question asked was, “Which candidate seems to be a better representative of their political party?” (M = 3.94, SD = 1.939). Next, three categories of questions tapped electability. The first category asked participants about personable aspects of the candidates, and included two questions: “Which candidate seems more likeable?” (M = 4.37, SD = 1.702), and “more authentic?” (M = 4.04, SD = 1.873). These variables were combined to create a mean scale of Likeability (r = .56, p < .001; α = .712), which ranged from 1 (strongly control candidate Knowles) to 6 (strongly variant candidate Rollins) (M = 4.22, SD = 1.532). The next category asked participants to assess the capability of the candidates, and included three questions: “Which candidate seems more competent?” (M = 3.85, SD = 1.801), “more capable of handling important issues?” (M = 3.78, SD = 1.839), and “more qualified for the office of Governor?” (M = 3.90, SD = 1.810). These variables were combined to create a mean scale of Competence (α = .832), with the same range as above (M = 3.78, SD = 1.560). A third category targeted electoral success, and asked participants to assess their vote intention: “If the election were held today which candidate would you vote for?” (M = 4.27, SD = 1.863), and the candidate’s viability: “Which candidate do you think is more likely to win this election?” (M = 4.13, SD = 1.670). These variables were combined to create a mean scale of Electoral Success (r = .60, p < .001; α = .745), with the same range as above (M = 4.23, SD = 1.608).⁵

Results

We begin with perceptions among partisan voters. The first set of hypotheses focused on self-identified Democrats exposed to a Democratic primary. We predicted that Democratic participants would be more likely to favor the feminine-only candidate (H1a) or either of the gender-balanced candidates (H1b) over the masculine-only candidate as the best representative of their party and as more likeable, competent, and electorally successful. To test these hypotheses, we conducted within-condition comparisons. First, we calculated the percentage of participants who favored either the all-masculine candidate (represented as 1-3 on the dependent measures) or the variant candidate (represented as 4-6) for each dimension of a variable. Next, we conducted Difference-of-Proportions tests to assess whether the percentages significantly differed. Last, we calculated perception gaps by either (a) subtracting the percent preferring the all-masculine candidate from the percent preferring the all-feminine candidate or (b) subtracting the percent preferring the all-masculine candidate from the percent preferring the gender-balanced candidate. A positive gap indicates that participants preferred the all-feminine or gender-balanced candidate over the all-masculine candidate. These analyses both test and visually facilitate examination of whether a candidate had an advantage on each dimension (see Table 1).

Table 1.

Perception Gaps for Democratic Participants Exposed to Democratic Primaries.

	Candidate emphases on issues and traits
	All feminine (n = 64)	Feminine issues/masculine traits(n = 78)	Masculine issues/feminine traits (n = 68)
Representative of party (%)	+37.4**	+51.2**	+23.6**
Likeability (%)
Likeable	+50.0**	+48.8**	+67.6**
Authentic	+40.6**	+41.0**	+20.6*
Competency (%)
Competent	+12.4	+25.6**	+23.6**
Capable	+15.6†	+12.8	+14.8†
Qualified	+6.4	+20.6*	+32.4**
Electoral success (%)
Vote intention	+43.8**	+59.0**	+35.2**
Viability	+31.2**	+33.4**	+58.8**

Note. Significance based on Difference of Proportions tests comparing the percent of participants who selected the all-masculine candidate to those who selected the all- or partial-feminine candidate. Perception gaps were created by subtracting those who preferred the all-masculine candidate from those who preferred the all- or partial-feminine candidate.

†

p < .10. *p < .05. **p < .01.

The positive perception gaps in the first column of percentages in Table 1 indicate that in every comparison, Democratic participants preferred the completely congruent all-feminine Democratic candidate to the completely incongruent all-masculine candidate, thereby supporting H1a. Perception gaps ranged from a 6.4% advantage for the all-feminine candidate regarding which candidate is more qualified for office to a 50% advantage for the all-feminine candidate concerning which candidate is more likeable. Notably, the gaps are smallest when measuring the individual variables that comprise competency—a finding that aligns with the idea that competency is more typically associated with masculinity. Nonetheless, all of the gaps are positive and the results in Table 1 indicate that Democratic participants strongly preferred a woman who emphasized party- and gender-congruent feminine elements to a candidate who focused solely on masculinity.

Next, we assessed the gender-balanced candidate in comparison with the all-masculine candidate. The two right-hand columns of Table 1 show the perception gaps for this comparison. As with the initial hypothesis, the consistently positive gaps in every comparison indicate that Democratic participants preferred the gender-balanced candidate to the all-masculine candidate. Specifically, for the feminine issues/masculine traits candidate the preferences ranged from a 12.8% advantage for capability to a 59% advantage for vote intention over the all-masculine candidate. Similarly, for the masculine issues/feminine traits candidate, the preferences ranged from a 14.8% advantage for capability to a 67.6% advantage for being likeable. Therefore H1b was supported because Democratic participants favored the partially congruent candidate who offered some feminine elements to align with gender- and party-based expectations over the incongruent all-masculine candidate.

As a next step, we moved to across-condition comparisons, to examine RQ1: Among self-identified Democrats, who will be considered the best representative of the party and as more likeable, competent, and electorally successful: the feminine-only or the gender-balanced candidate? To explore this question, we first combined the conditions featuring a mix of femininity and masculinity, and because this analysis is across conditions, we were able to run a MANCOVA among self-identified Democratic participants, with participant’s age and gender as covariates. These covariates were selected due to unequal distribution of these characteristics across the treatment groups. The gender-balanced candidate was favored in three of the four comparisons, but the initial MANCOVA was not significant. An additional MANCOVA comparing all three treatment conditions—feminine-only, feminine issues/masculine traits, and masculine issues/feminine traits—was also run and yielded insignificant results. Thus, Democratic participants’ evaluations of a candidate who completely aligned with gender and party ownerships and another who partially aligned with such ownerships were similar across the board. When combined with the results from Table 1, these findings reveal that Democrats prefer a candidate who embraces some form of femininity, and the type and extent of that feminine embrace is not particularly consequential.

The next set of research questions and hypotheses explored self-identified Republican participants’ perceptions of candidates in a Republican primary. Because of the conflict between gender- and party-based ownerships, we started with RQ2: Among self-identified Republican participants, who will be considered the best representative of the party and as more likeable, competent, and electorally successful: the gender-congruent feminine-only candidate or the party-congruent masculine-only candidate? To explore this question, we repeated the perception gap analysis (see Table 2).

Table 2.

Perception Gaps for Republican Participants Exposed to Republican Primaries.

	Candidate emphases on issues and traits
	All feminine (n = 55)	Feminine issues/masculine traits (n = 55)	Masculine issues/feminine traits (n = 48)
Representative of party (%)	−27.2**	+5.4	+16.6
Likeability (%)
Likeable	+52.8**	+38.2**	+25.0*
Authentic	+5.4	+31.0**	+4.2
Competency (%)
Competent	−20.0*	+20.0*	+29.2**
Capable	−27.2**	+23.6*	+33.4**
Qualified	−9.0	+23.6*	+33.4**
Electoral success (%)
Vote intention	−1.8	+31.0**	+33.4**
Viability	+9.0	+34.6**	+29.2**

Note. Significance based on Difference of Proportions tests comparing the percent of participants who selected the all-masculine candidate to those who selected the all- or partial-feminine candidate. Perception gaps were created by subtracting those who preferred the all-masculine candidate from those who preferred the all- or partial-feminine candidate.

p < .05. **p < .01.

The mix of positive and negative gaps in the first column of percentages in Table 2 indicates that Republican participants varied in their preferences depending on the question at hand. The negative gaps for five of eight comparisons indicate that they generally preferred the all-masculine candidate to the all-feminine candidate, ranging from a relatively small advantage of 1.8% for vote intention to an advantage of 27.2% for both best party representative and capability. Conversely, the largest gap across all eight comparisons was a 52.8% advantage for the all-feminine candidate for being perceived as more likeable. Overall, the mix of findings and varying levels of significance represents the conflict between preferring a Republican woman candidate who either completely aligns with party or who completely aligns with gender.

With these party and gender conflicts expected, we then predicted that Republican participants would be more likely to view the gender-balanced candidate than either the masculine-only candidate (H2a) or feminine-only candidate (H2b) as the best representative of their party and as more likeable, competent, and electorally successful.

Data to test H2a are presented in the two right-hand columns of Table 2. The consistently positive gaps indicate that when Republicans were able to pick between a female candidate who mixed party and gender ownerships and a female candidate who was completely congruent with party ownership, they overwhelmingly preferred the more balanced persona. Specifically, for the feminine issues/masculine traits candidate, the preferences ranged from a 5.4% advantage over the all-masculine candidate for who is the best representative of the party to a 38.2% advantage for being more likeable. Similarly, for the masculine issues/feminine traits candidate, the preferences over the all-masculine candidate ranged from a 4.2% advantage for authenticity to a 33.4% advantage for the three variables of being more capable and qualified, as well as vote intention. Overall, H2a was supported: Self-identified Republicans favored the candidate who balanced gender and party over the all-masculine candidate.

To test H2b, we conducted across-condition analysis. Specifically, we combined the conditions containing candidates emphasizing a mix of femininity and masculinity and then ran a MANCOVA, with participant gender and age as covariates, among self-identified Republican participants to compare perceptions of all-feminine and gender-balanced candidates. The MANCOVA yielded significant results, Wilks’s Λ = .920, F(4, 126) = 2.732, p = .032, $η_{p}^{2}$ = .080. Follow-up ANCOVAs revealed two significant differences wherein Republican participants evaluating female candidates in a Republican primary preferred the gender-balanced candidate who exhibited both masculinity and femininity. Specifically, Republicans viewed the gender-balanced candidate as more competent (M = 3.86, SD = 1.606) than the feminine-only candidate (M = 3.23, SD = 1.496), F(1, 129) = 4.911, p = .028, $η_{p}^{2}$ = .037, and as more electorally successful (M = 4.17, SD = 1.782) than the feminine-only candidate (M = 3.58, SD = 1.789), F(1, 129) = 2.918, p = .090, $η_{p}^{2}$ = .022. This trend was continued with Republicans perceiving the gender-balanced candidate as a better representative of the party (M = 3.51, SD = 1.835) than the feminine-only candidate (M = 3.00, SD = 1.961), but this ANCOVA was insignificant, F(1, 129) = 2.283, ns, $η_{p}^{2}$ = .017. Notably, though, in evaluating likeability, Republican participants saw the feminine-only candidate (M = 4.08, SD = 1.616) as holding her own with the candidate exhibiting both masculinity and femininity (M = 3.95, SD = 1.572), F(1, 129) = .230, ns, $η_{p}^{2}$ = .002—a finding that aligns with scholarship indicating a strong correlation between femininity and likeability for women candidates. Overall, we found partial support for H2b: Republican participants viewed the feminine-only candidate—the woman who is most at odds with her party ownership—as less competent and electorally successful than the gender-balanced woman, and a less ideal representative of their party, yet nonetheless equally likeable.

As before, an additional MANCOVA comparing all three treatment conditions was run and suggested marginally significant differences for Republican participants, Wilks’s Λ = .898, F(8, 250) = 1.724, p = .093, $η_{p}^{2}$ = .052. Follow-up ANCOVAs produced one marginally significant difference regarding competency, F(2, 128) = 2.466, p = .089, $η_{p}^{2}$ = .037. Republicans viewed the masculine issues/feminine traits candidate (M = 3.95, SD = 1.698) and the feminine issues/masculine traits candidate (M = 3.79, SD = 1.550) as more competent than the feminine-only candidate (M = 3.23, SD = 1.496). This finding, then, mimics the previous analysis and showcases that Republican participants perceive a gender-balanced Republican candidate as more competent than a feminine-only candidate.

Before moving on to the final hypothesis that compares inter-party perceptions, we wish to summarize the results so far for our intra-party comparisons. For Democratic female candidates, party and gender ownership align, and the feminine-only candidate represents complete congruency and the masculine-only candidate portrays complete incongruency. The consistently positive perception gaps of Table 1 revealed that self-identified Democratic participants favored the candidate in complete alignment with gender- and party-based ownerships. In addition, all of the perception gaps for both variations of the gender-balanced candidate were also positive, suggesting that Democrats fully rejected the gender- and party-incongruent candidate. If we move across the columns in Table 1, the introduction of some masculinity did appear to create larger gaps regarding perceptions of being competent and qualified, and the feminine issues/masculine traits candidate had a marked advantage over the other variant candidates for best representative and vote intention. Thus, the MANCOVA analysis sought to compare perceptions of the all-feminine and gender-balanced candidates to see whether Democratic participants’ perceptions varied based on the level of congruency. There were no significant differences between the three candidate types, allowing us to conclude that Democratic participants generally favored any candidate who was fully or partially congruent.

Analysis of Republicans told a slightly different story. The completely party-congruent candidate for Republicans was the all-masculine candidate, and the completely gender-congruent candidate was the all-feminine candidate. The negative perception gaps in the first data column of Table 2 indicate that Republican participants more often favored the party-congruent candidate. However, across the columns in Table 2, we saw a different trend: a Republican candidate partially congruent with her gender and party not only surpassed the all-masculine candidate, but she gained ground on being the best representative, competency, and electoral success. In other words, when a Republican woman moves away from being completely gender-congruent and introduces some masculinity into her persona, she gains considerable strength as a candidate. The biggest swings in perception from the all-feminine candidate to the gender-balanced candidate were for the masculine issues/feminine traits candidate. The smallest swing was for viability, in which the all-feminine candidate had a 9-point lead on the all-masculine candidate to a 29.2-point lead for the masculine issues/feminine traits candidate, creating a 20.2-point swing. The largest movement was for perceptions of capability, with the all-feminine candidate having a 27.2-point disadvantage compared with the all-masculine candidate to a 33.4-point advantage for the masculine issues/feminine traits candidate, creating a 60.6-point swing. These findings suggest that Republican participants favor gender-balanced candidates who depart from solely party- or gender-congruency.

The final hypothesis compared perceptions across Democratic and Republican participants. For H3, we predicted that Democrats would be significantly more likely than Republicans to consider a candidate expressing femininity, either wholly or partly, as the best representative of their party and as more likeable, competent, and electorally successful. To test this hypothesis, we ran a MANCOVA, comparing Republicans exposed to Republican primaries and Democrats exposed to Democratic primaries. The initial MANCOVA was significant, Wilks’s Λ = .922, F(4, 312) = 6.573, p = .000, $η_{p}^{2}$ = .078—as were all of the ANCOVAs. Note that for all comparisons, higher means indicate greater preference for the candidate expressing full- or partial-femininity.

Two findings were apparent in the follow-up analysis. First, Democrats perceived the more feminine candidate as a significantly better representative (M = 4.34, SD = 1.827) than Republicans (M = 3.31, SD = 1.896), F(1, 315) = 24.269, p = .001, $η_{p}^{2}$ = .072. Therefore, Democrats were more likely to perceive the woman expressing some party-aligned femininity as a better partisan representative. Second, Democratic participants found the woman candidate who expressed some femininity as more electable. Specifically, Democrats perceived the more feminine candidate as significantly more likeable (M = 4.41, SD = 1.456) than Republicans (M = 4.00, SD = 1.585), F(1, 315) = 5.081, p = .025, $η_{p}^{2}$ = .016; as slightly more competent (M = 3.94, SD = 1.490) than Republicans (M = 3.61, SD = 1.588), F(1, 315) = 3.786, p = .053, $η_{p}^{2}$ = .012; and as significantly more electorally successful (M = 4.44, SD = 1.390) than Republicans (M = 3.94, SD = 1.80), F(1, 315) = 9.106, p = .003, $η_{p}^{2}$ = .028. In sum, when partisans of the major political parties were presented with two women candidates in a primary election context, self-identified Democrats were significantly more likely than self-identified Republicans to consider the candidate portrayed as more feminine to be a better partisan representative and to be more electable.

Discussion

Political communication scholarship has examined the intertwining of party and voter expectations in the long-common setting of all-male elections, and has more recently started to examine mixed-gender elections. Such research has been valuable in understanding how politicians navigate and capitalize on these expectations. However, as more women enter politics, they are increasingly finding their competitor to be a fellow woman, which demands a next step in scholarship on ownerships in politics and culture. With this in mind, we focused on woman-versus-woman elections—thereby allowing us to examine how the current scholarship applies to women. Across this analysis, two patterns emerged.

First, an emphasis solely on stereotypical masculinity rarely pays off in a woman-versus-woman election. This finding was present both among-partisan and across-partisan condition comparisons. Among partisans, there were a total of 48 perception gaps calculated comparing those who preferred the all-masculine candidate to the variant candidate. Of these 48, 43 showed that participants favored the woman who portrayed traditional full- or partial-femininity. In only five instances did the all-masculine candidate have an advantage, and all of those were present in the all-masculine versus all-feminine comparison among Republicans. For across-condition comparisons, the means show a favoring of the variant candidate. Specifically, the means across all of the ANCOVAs surpassed the midway points of the variables’ scales, indicating a stronger preference for the variant candidate who expressed full- or partial-femininity. This pattern indicates that voters preferred their female gubernatorial candidates to embrace at least some stereotypical femininity. We would expect this outcome for Democrats, but its presence for Republicans is noteworthy because it conflicts with party ownership. These findings suggest that in a woman-versus-woman primary election, “running like a (stereotypical) man” is not a winning strategy.

Second, Democratic and Republican women had different paths to success in our woman-versus-woman elections. Specifically, self-identified Democrats considered the female candidate who emphasized typical feminine issues and/or traits to be a better representative of the Democratic Party and more electable overall—thereby supporting the alignment between Democrats and femininity. It appears then, that by simultaneously fulfilling party and gender ownership expectations, Democratic women candidates can carve out an electable image in a primary environment. Research on party ownership has found similar findings within all-male electoral contexts (e.g., Ansolabehere & Iyengar, 1994; Bélanger & Meguid, 2008; Petrocik, 1996), and research examining partisan candidates in mixed-gender elections has also found an alignment between candidates’ party and voter expectations (Banwart, 2010). The findings here thus extend scholarship to a relatively unexamined electoral setting. Our findings are also suggestive of a broader trend: The lack of significant difference between perceptions of the all-feminine and gender-balanced Democratic candidates suggests that Democratic women candidates have considerable leeway in a female-only primary election because they can be stereotypically all-feminine or trespass party/gender lines and incorporate some traditional masculinity and still be considered electable.

Republican female candidates seemingly have a more complex road to success. Scholarship on party ownership would have predicted that Republicans would favor the Republican woman who emphasized solely masculinity via issue emphases and traits. If this were so, we would have found negative perception gaps and mean score evaluations that were consistently below the midway point on our dependent variables. We saw the opposite in most comparisons. We found in general that in women-only elections, Republican women candidates need to cultivate an image that effectively balances both party and gender ownership. In particular, the gender-balanced women candidates—who combined stereotypical femininity and masculinity via issues and traits—were seen as more electable than the traditionally all-masculine candidate or all-feminine candidate. This finding was present when comparing the gender-balanced candidate with the all-masculine candidate and in the gender-balanced versus feminine-only comparisons. Republicans, as a party, tend to have a more traditional view of a woman’s role in society than Democrats (Sanbonmatsu, 2004). This may explain why Republicans prefer a woman candidate who talks about typical masculine matters, while still putting forth a traditionally feminine image through an embrace of some femininity. In the big picture, this need to project a “both/and” gendered identity means Republican women candidates almost certainly have a trickier political line to walk than their male counterparts because they need to balance Republican ownership and female ownership, even though they have been found to culturally conflict, to be electable.

But these data also suggest that Republican voters in women-only elections may no longer hold Republican women candidates to Jamieson’s (1995) femininity/competence double bind, which stipulated that women can meet societal expectations of femininity at the cost of being perceived as incompetent, or meet professional standards of competency and risk being perceived as not womanly enough or as gender-inauthentic. Jamieson observed this double bind for women across a number of fields and industries in America, which were typically highly masculine and often put women at a disadvantage. It created an “either/or” situation for women that was ultimately a lose/lose. But our finding that Republican participants preferred a more gender-balanced female candidate, and indeed saw her as more electable than a solely masculine or solely feminine female candidate, suggests a potentially more-expansive approach: Republican women candidates appear to have more movement along the gender continuum when they are squaring off against other women, and are not tethered to either anchor point and the constraints attached to those points. This evidence does not suggest that the double bind is no longer a cultural bulwark in America, but it does suggest that prototypical masculinity has its limits in this electoral context. Future studies should continue to examine how party- and gender-based ownerships intersect to create new and expanding opportunities for partisan women in women-only elections as well as other electoral contexts.

Our study not only examined a number of perceptions but also came with some limitations. First, as with all experiments, there was some artificiality in the study design. For instance, the candidates were hypothetical and the participants were asked to assess candidates based on the information in one article. However, it is important to note that the news articles used in this study were drawn from actual news content of candidates launching their campaigns for the office of Governor. The order of content, the presentation of biographical information, and the focus on the candidate’s political issues and character traits of interest were all consistent with contemporary campaign news. Second, the control and variant candidates were ascribed issue or traits from a collection of stereotypical masculine and feminine options, and it is possible, for example, that the issue or traits assigned to the control candidate were less desirable than the issue or traits ascribed to the variant candidate. However, the control candidate was kept constant in all conditions, and a large intent of the study was to compare the variant candidates across conditions to further examine electability at the crossroads of gender and party. And again, the issue of positions and trait emphases were the ones often present in the U.S. political arena. Third, it is possible that having non Missouri citizens evaluate a Missouri race may affect the findings. However, using Survey Monkey to gather a more geographically diverse sample aids the external validity of this study, and this design—where citizens evaluate an out-of-state candidate—is common in other scholarship focused on the impact of gender and party on candidate evaluations (e.g., Banwart, 2010; Fridkin, Kenney, & Woodall, 2009). Overall, we sought to minimize these weaknesses as much as possible.

Finally, we focused on woman-versus-woman primaries for a number of reasons, but this focus also limits generalizations to other electoral contexts. A study by Kahn (1992) two decades ago, though, is interesting. In it, participants read a news article about one male or female Senate candidate, and the article included gender-stereotypic issue and trait coverage of the candidate. Kahn found that for both males and females, when candidates’ news coverage featured them stressing masculine issues and traits, they were considered more electorally viable than candidates whose news coverage focused on feminine issues and the more gender-neutral trait of being effective. Therefore, all candidates were considered more viable when news coverage portrayed them as all-masculine versus more gender-balanced. In our study, the all-masculine candidate did not fare even remotely close to as well; instead, participants consistently favored all-feminine and gender-balanced candidates. The parallels to our study are not exact, but it does seem plausible that the cultural preferences for candidate emphases and qualities have changed over time or that the woman-versus-women campaign setting—or any direct competition, which Kahn did not have—is much different than assessments of singular candidates. Ultimately, it would be a fruitful endeavor for future work to examine party- and gender-ownerships across all-male, mixed-gender, and all-female elections to assess contemporary notions of electability.

In considering the interactions of party, gender, and voter expectations, the once-simple schema of who owns what in male-only elections has become more complicated with more women seeking and gaining political offices. In this research, we focused on partisan women candidates who face particular challenges, who compel the news media to move beyond conveyance of mutually exclusive gender identities, and who prompt citizens to grapple with multifaceted candidate identities when casting their votes. Ultimately, both Democratic and Republican women face advantages and disadvantages at the intersection of party and gender. Democratic women benefit from strong congruence between party and gender, and are rewarded by partisan voters when they fulfill these aligned ownerships. However, they too are bound by these expectations. Masculine issues such as the economy and defense are omnipresent factors in today’s society. Thus, Democratic women need to carefully blend a predominant focus on feminine elements while maintaining an image that does not ignore contemporary issues of import. Republican women also need to express masculinity and femininity to satisfy party and gender expectations, and when they do so, they appear to break constrictive double binds. This blended identity affords them more room to maneuver in crafting their self-presentation, but more room in a campaign setting can sometimes open up space for straying too far in any given direction—creating more opportunities for political missteps—or for moving around too much during the campaign, and thus failing to create a unified self-presentation. Woman-versus-woman elections ultimately create a seemingly paradoxical space: Their presence breaks the masculine norms of politics and creates a new range of movement for candidates, especially Republican women, but all the while they still tether women to certain areas of ownership and expectations. Future research should continue to examine this theoretically rich space to explore what new and old challenges and opportunities it presents for women candidates, the news media, and voters.

Footnotes

Appendix

Condition 1: Republican variant candidate emphasized feminine issues and feminine traits.

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.

Notes

Author Biographies

Lindsey Meeks is an assistant professor in the Department of Communication at the University of Oklahoma. Her research areas include political communication, gender, and media, with a focus on the intersection of these areas within elections.

David Domke is a professor in the Department of Communication at University of Washington. His research focuses on political discourse, journalism, and public opinion, with a particular focus on racial, religious, and gender emphases.

References

Ansolabehere

Iyengar

(1994). Riding the wave and claiming ownership over issues: The joint effects of advertising and news coverage in campaigns. Public Opinion Quarterly, 58, 335-357.

Banwart

(2010). Gender and candidate communication: Effects of stereotypes in the 2008 election. American Behavioral Scientist, 54, 265-283.

Banwart

Kelly

(2013). Running on the web: Online self-presentation strategies in mixed-gender races. Social Science Computer Review, 31, 614-624.

Banwart

McKinney

(2005). A gendered influence in campaign debates? Analysis of mixed-gender United States senate and gubernatorial debates. Communication Studies, 56, 353-373.

Barbara Lee Family Foundation. (2010). Turning point. Available from http://www.barbaraleefoundation.org/

Barbara Lee Family Foundation. (2012). Pitch perfect. Available from http://www.barbaraleefoundation.org/

Barber

(1972). The presidential character. Englewood Cliffs, NJ: Prentice Hall.

Bélanger

É.

Meguid

B. M.

(2008). Issue salience, issue ownership, and issue-based vote choice. Electoral Studies, 27, 477-491.

Benoit

W. L.

(2007). Own party issue ownership emphasis in presidential television spots. Communication Reports, 20, 42-50.

10.

Braden

(1996). Women politicians and the media. Lexington: University Press of Kentucky.

11.

Budge

Farlie

(1983). Explaining and predicting elections: Issue effects and party strategies in twenty-three democracies. London, England: Allen & Unwin.

12.

Bystrom

(2006). Gender and campaign communication: TV ads, web sites, and media coverage. Berkeley: Institute of governmental studies, University of California, Berkeley.

13.

The Center for American Women and Politics. (2012a). Woman versus woman: Congressional and gubernatorial races 1944-2012. Available from http://www.cawp.rutgers.edu/

14.

The Center for American Women and Politics. (2012b). Gender differences in voter turnout. Available from http://www.cawp.rutgers.edu/

15.

The Center for American Women and Politics. (2013). Primary problems: Women candidates in U.S. House primaries. Available from http://www.cawp.rutgers.edu/

16.

Connell

(2005). Masculinities. Berkley: University of California Press.

17.

Dolan

(1998). Voting for women in the “year of the woman.” American Journal of Political Science, 42, 272-293.

18.

Dolan

(2014). Gender stereotypes, candidate evaluations, and voting for women candidates: What really matters? Political Research Quarterly, 67, 96-107, doi:10.1177/1065912913487949

19.

Domke

Coe

(2008). The God strategy: How religion became a political weapon in America. New York, NY: Oxford University Press.

20.

Fridkin

Kenney

Woodall

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

21.

Funk

(1999). Bringing the candidate into models of candidate evaluation. The Journal of Politics, 61, 700-720.

22.

Gallup. (2014). Democrats’ and Republicans’ confidence in their party to handle key issues. Available from http://pollingmatters.gallup.com/

23.

Glass

(1985). Evaluating presidential candidates: Who focuses on their personal attributes? Public Opinion Quarterly, 49, 517-534.

24.

Hayes

(2005). Candidate qualities through a partisan lens: A theory of trait ownership. American Journal of Political Science, 49, 908-923.

25.

Hayes

(2011). When gender and party collide: Stereotyping in candidate trait attribution. Politics & Gender, 7, 133-165.

26.

Heldman

Carroll

Olson

(2005). “She brought only a skirt”: Print media coverage of Elizabeth Dole’s bid for the Republican presidential nomination. Political Communication, 22, 315-335.

27.

Herrnson

Lay

Stokes

(2003). Women running “as women”: Candidate gender, campaign issues, and voter-targeting strategies. The Journal of Politics, 65, 244-255.

28.

Huddy

Terkildsen

(1993a). Gender stereotypes and the perception of male and female candidates. American Journal of Political Science, 37, 119-147.

29.

Huddy

Terkildsen

(1993b). The consequences of gender stereotypes for women candidates at different levels and types of office. Political Research Quarterly, 46, 503-525.

30.

Iyengar

Valentino

Ansolabehere

Simon

(1997). Running as a woman: Gender stereotyping in political campaigns. In Norris

(Ed.), Women, media, and politics (pp. 77-98). New York, NY: Oxford University Press.

31.

Jalalzai

(2006). Women candidates and the media: 1992-2000 elections. Politics & Policy, 34, 606-633.

32.

Jamieson

(1995). Beyond the double bind: Women and leadership. New York, NY: Oxford University Press.

33.

Kahn

(1992). Does being male help? An investigation of the effects of candidate gender and campaign coverage on evaluations of US Senate candidates. The Journal of Politics, 54, 497-517.

34.

Kahn

(1993). Gender differences in campaign messages: The political advertisements of men and women candidates for US Senate. Political Research Quarterly, 46, 481-502.

35.

Kahn

(1994). The distorted mirror: Press coverage of women candidates for statewide office. The Journal of Politics, 56, 154-173.

36.

King

Matland

(2003). Sex and the Grand Old Party: An experimental investigation of the effect of candidate sex on support for a Republican candidate. American Politics Research, 31, 595-612.

37.

Lawless

(2004). Women, war, and winning elections: Gender stereotyping in the post-September 11th era. Political Research Quarterly, 57, 479-490.

38.

Louden

McCauliff

(2004). The “authentic candidate”: Extending candidate image assessment. In Hacker

(Ed.), Presidential candidate images (pp. 85-103). Lanham, MD: Rowman & Littlefield.

39.

Major

Coleman

(2008). The intersection of race and gender in election coverage: What happens when the candidates don’t fit the stereotypes? Howard Journal of Communications, 19, 315-333.

40.

McDermott

(1997). Voting cues in low-information elections: Candidate gender as a social information variable in contemporary United States elections. American Journal of PoliticaSl Science, 41, 270-283.

41.

McGraw

(2003). Political impressions: Formation and management. In Sears

D. O.

Huddy

Jervis

(Eds.), Oxford handbook of political psychology (pp. 394-432). Oxford, UK: Oxford University Press.

42.

McLeod

J. J.

Glynn

McDonald

(1983). Issues and images: The influence of media reliance in voting decisions. Communication Research, 10, 37-58.

43.

Meeks

(2012). Meeting expectations: Issues, traits, party, and gender in a woman-versus-woman election. Presented in the Public Opinion Division at the American Association of Public Opinion Research, Orlando, FL.

44.

Meeks

(2013). He tweets, she tweets: Examining gendered presentations of self in 2012 Senate elections. Presented in the Political Communication Division at the National Communication Association, Washington, DC.

45.

Niven

Zilber

(2001). Do women and men in congress cultivate different images? Evidence from congressional web sites. Political Communication, 18, 395-405.

46.

Palmer

Simon

(2005).When women run against women: The hidden influence of female incumbents in elections to the U.S. House of Representatives, 1956–2002, Politics & Gender, 1, 39-63.

47.

Paolino

(1995). Group-salient issues and group representation: Support for women candidates in the 1992 Senate elections. American Journal of Political Science, 39, 294-313.

48.

Petrocik

(1996). Issue ownership in presidential elections, with a 1980 case study. American Journal of Political Science, 40, 825-850.

49.

Petrocik

Benoit

Hansen

(2003). Issue ownership and presidential campaigning, 1952-2000. Political Science Quarterly, 118, 599-626.

50.

Pew Research Center. (2014). Political polarization in the American public. Available from http://www.people-press.org/

51.

Popkin

(1991). The reasoning voter: Communication and persuasion in presidential campaigns. Chicago, IL: University of Chicago Press.

52.

Sanbonmatsu

(2004). Democrats, Republicans, and the politics of women’s place. Ann Arbor: University of Michigan Press.

53.

Sanbonmatsu

Dolan

(2009). Do gender stereotypes transcend party? Political Research Quarterly, 62, 485-494.

54.

Winter

(2010). Masculine republicans and feminine democrats: Gender and Americans’ explicit and implicit images of the political parties. Political Behavior, 32, 587-618.