Despite interest in women political leaders’ handling of the COVID-19 crisis (Henley and Roy Reference Henley and Roy2020; Politics & Gender 2020), historically, the association of masculinity with political leadership has made it difficult for women to be perceived as legitimate actors in the political arena (Duerst-Lahti Reference Duerst-Lahti and O'Connor2010; Duerst-Lahti and Kelly Reference Duerst-Lahti, Kelly, Duerst-Lahti and Kelly1995). Desirable leadership traits have tended to align with the agentic qualities that men are presumed to possess, while the less desirable attributes align with the communal qualities ascribed to women (Eagly and Carli Reference Eagly and Carli2007). Thus, women politicians aspiring to leadership positions encounter a double bind. If they emphasize so-called feminine qualities to meet societal expectations, they are perceived as lacking the appropriate leadership skills. But if women emphasize qualities that are conventionally understood as masculine, they are seen as violating gender roles. “Either way, they may leave the impression that they don't have ‘the right stuff’ for powerful jobs” (Eagly and Carli Reference Eagly and Carli2007, 66).
Media depictions of politicians have historically reinforced the masculine underpinnings of political leadership. The gendered mediation thesis asserts that journalists apply a masculine lens when covering politics (Gidengil and Everitt Reference Gidengil and Everitt2000; Sreberny-Mohammadi and Ross Reference Sreberny-Mohammadi and Ross1996). They do so by treating politics like a sport (Lawrence Reference Lawrence2000) and prioritizing policy issues associated with men, such as the economy and defense (Kahn and Goldenberg Reference Kahn and Goldenberg1991; Meeks Reference Meeks2012). Studies have also found that news coverage is more likely to emphasize allegedly masculine traits than feminine ones when discussing the qualities of individual politicians, regardless of gender (Cantrell and Bachmann Reference Cantrell and Bachmann2008; Harmer, Savigny, and Ward Reference Harmer, Savigny and Ward2017; Meeks Reference Meeks2012). Journalists have tended to maintain the masculine boundaries of politics by trivializing women's candidacies through excessive attention to personal characteristics such as their physical appearance or romantic partnerships (Falk Reference Falk2010; Heldman, Carroll, and Olson Reference Heldman, Carroll and Olson2005; Tolley Reference Tolley2016; Trimble et al. Reference Trimble, Wagner, Sampert, Raphael and Gerrits2013).
This leads us to ask, does the news media similarly disadvantage women by emphasizing leadership qualities they are presumed to lack? Or has the increased presence of women in executive leadership roles eroded gendered assumptions about leadership skills and competencies by positioning women as capable government leaders? To answer these questions, our study draws on the gender stereotypes and political leadership literatures to develop an innovative coding framework that measures eight negative and eight positive qualities, for a total of 16 leadership markers. Our statistical model allows us to investigate the influence of premier gender on media discourses about political leadership.
One of the major challenges to conducting research on women political leaders is the limited number of female heads of government at the national level: just 12 held office in 2020 (Council on Foreign Relations 2020; UN Women 2019). To overcome this barrier, our study turns to the subnational level of government in Australia and Canada, where a sizeable number of women have risen to government leadership roles since 2007. Eleven women—five in Australia and six in Canada—led state or provincial governments during this time period, offering a unique opportunity to investigate how women and men premiers are introduced by the news media to the public as their jurisdiction's new leader. Our study quantitatively analyzes media depictions of the leadership qualities of 11 women premiers and their immediate male predecessors, for a sample of 22 government leaders who rose to office in Australia and Canada between 1998 and 2017. We focus on newspaper coverage in the first seven days after their election, as this is a period when journalists introduce new leaders to the public by identifying their qualities, skills, and experience.
We found a relatively level playing field for evaluations of new leaders. Women and men premiers received approximately the same amount of attention to their leadership skills and qualities. As important, they were assessed based on the same criteria, with a particular focus on experience, strength, and competence. The reporting was gendered in the sense that journalists prioritized “masculine” qualities of experience and strength and downplayed “feminine” qualities of honesty and compassion in coverage. However, as the longitudinal analysis reveals, while the media continue to emphasize “masculine” leadership norms in politics,Footnote 1 they also expect new government leaders to demonstrate competence, a trait increasingly associated in the public's mind with women. Further, leaders are praised when they work collaboratively with others and criticized for acting in an authoritarian manner. The news media, in other words, view the traditionally feminine trait of collaboration as key to effective political leadership. Once they have reached the pinnacle of power, premiers are expected to perform their duties with competence, intelligence, and levelheadedness, regardless of their gender identity. As our study shows, media conceptualizations of political leadership competencies are slowly expanding in ways that make it easier for women to be seen as effective political leaders. As a result, researchers should reevaluate assumptions about the gendering of leadership traits and be attentive to changes in news representations over time.
LITERATURE REVIEW
Gender stereotypes—those attributes, traits, and behaviors that women and men are presumed to possess by virtue of their gender—inform many aspects of political life. Stereotypes are “widely held beliefs about the nature and behavior of different groups and their individual members” (Wagner and Everitt Reference Wagner, Everitt, Wagner and Everitt2019, 9; see also Crawford et al. Reference Crawford, Jussim, Madon, Cain and Stevens2011). Stereotypes are not only ascriptive but prescriptive. Each gender group is presumed to possess the qualities necessary to perform their social roles. The sexual division of labor has historically slotted women and men into different fields of activity: men in the public realm of politics, business, military, and sports, and women in the domestic domain of housework and child-rearing. Men are thus assumed to possess agentic qualities of being assertive, decisive, tough, aggressive, competitive, self-confident, courageous, and independent (Eagly et al. Reference Eagly, Nater, Miller, Kaufmann and Sczesny2020; Prentice and Carranza Reference Prentice and Carranza2002). Women, by contrast, are regarded as possessing communal qualities, such as honesty, compassion, warmth, sensitivity, emotionality, and cautiousness (Eagly et al. Reference Eagly, Nater, Miller, Kaufmann and Sczesny2020; Prentice and Carranza Reference Prentice and Carranza2002). In short, men are seen as leaders and women as helpmates.
Recent psychological research, however, suggests that gender stereotypes are changing in response to evolving gender roles (Eagly et al. Reference Eagly, Nater, Miller, Kaufmann and Sczesny2020). Women's increasing participation in the workforce in general, and male-dominated professions in particular, has led to a reappraisal of female stereotypes. Competence is now ascribed more often to women than to men. The American public now rates women as equally creative, inventive, innovative, organized, detail-oriented, and open-minded as men, as well as equally intelligent and smart (Eagly et al. Reference Eagly, Nater, Miller, Kaufmann and Sczesny2020). However, their segregation into communal-type roles, even in male-dominated professions, means that women continue to be associated with communal rather than agentic qualities (Eagly et al. Reference Eagly, Nater, Miller, Kaufmann and Sczesny2020, 9–11). Despite women's advancement in the workforce, other research finds that they continue to be less likely than men to see themselves as possessing the qualities of a good manager (Powell and Butterfield Reference Powell and Butterfield2015). Men thus continue to have an advantage in being taken seriously as political leaders.
Gender stereotypes play a key role in political socialization by closely aligning idealized masculine qualities with those required of political leaders (Duerst-Lahti Reference Duerst-Lahti and O'Connor2010; Johnson Reference Johnson2013). In the past, a political leader was expected to be competent, knowledgeable, honest, trustworthy, experienced in elected office, energetic, and aggressive, but not necessarily warm, humble, or selfless (Kinder et al. Reference Kinder, Peters, Abelson and Fiske1980; Trent et al. Reference Trent, Short-Thompson, Mongeau, Nusz and Trent2001). Men were presumed to possess many, but not all, of these qualities. More recent research finds that journalists also value some qualities associated with women, such as honesty (Aaldering and Vliegenthart Reference Aaldering and Vliegenthart2016), but on the whole, women's traditionally prescribed characteristics remain at odds with the prerequisites for political leadership.
The gendered mediation scholarship is devoted to understanding the role of gender in political reporting. It examines how journalists draw on gender stereotypes when shaping narratives about political actors. Sreberny-Mohammadi and Ross (Reference Sreberny-Mohammadi and Ross1996) have argued that masculine assumptions inform the way journalists cover politics and undermine women as legitimate political actors. The gendered mediation thesis advanced by Gidengil and Everitt (Reference Gidengil and Everitt2000, 105–6) conceptualizes how masculine assumptions work in political reporting:
The notion of gendered mediation implies that voters will be presented with a more filtered picture of women politicians and that the filter itself will be gendered. This filtering means, first, that the behavior of women politicians will be subject to more analysis than their male counterparts, and, second, that the women's coverage will reflect traditional masculine conceptions of politics.
Observing that political journalism's masculine norms had not changed despite increases in the number of women in politics and journalism, Gidengil and Everitt (Reference Gidengil and Everitt2003b, 208) warned that unless the news media eliminated their masculine filter, women would continue to be sidelined in politics.
Scholars have explored several aspects of political reporting to understand how this masculine filter works. First, the news media emphasize overtly masculine behaviors. In a series of studies on Canadian federal elections, Gidengil and Everitt (Reference Gidengil and Everitt2000, Reference Gidengil, Everitt and Cross2002, Reference Gidengil and Everitt2003a) discovered that news coverage highlighted women's physically aggressive behavior when describing what party leaders did during political debates but downplayed similar actions by men. When women conformed to female gender stereotypes and avoided aggressive behavior, however, journalists ignored them. Second, the media use masculine metaphors from the male-dominated worlds of war and sports, reinforcing politics as a male domain (Gerrits et al. Reference Gerrits, Trimble, Wagner, Raphael and Sampert2017; Gidengil and Everitt Reference Gidengil and Everitt2003b; Sampert and Trimble Reference Sampert, Trimble, Tremblay and Trimble2003). Women do not fit within this frame because they have not historically been viewed as warriors or boxers. Depicting women as combatants not only violates gender-role expectations that women be docile but also depicts them as outsiders in politics. Third, news organizations link traditionally “masculine” issues such as defense with men politicians and traditionally “feminine” issues such as reproductive rights with women (Falk Reference Falk2010; Kahn and Goldenberg Reference Kahn and Goldenberg1991; Meeks Reference Meeks2012; Vidal Correa Reference Vidal Correa2020), thereby supporting a gendered division of labor in politics. Fourth, scholars have examined more subtle gendering, observing how journalists treat men politicians more favorably than women politicians, for instance, when assessing their electoral viability (Heldman, Carroll, and Olson Reference Heldman, Carroll and Olson2005; Vidal Correa Reference Vidal Correa2020). The message is that men are winners and women are not.
The gendered mediation literature also explores the gendering of character traits, but with less nuance and specification than it does for personal characteristics. Research on media personalization of women politicians extensively explores depictions of their gender, physical appearance, clothing, age, relationships, offspring, and upbringing (Beail and Longworth Reference Beail and Longworth2013; Murray Reference Murray and Murray2010; Williams Reference Williams2017, Reference Williams2020). Some studies examine media portrayals of a specific type of personalization, such as how Western conceptions of the idealized family inform evaluations of political leadership and legitimacy (Auer et al. Reference Auer, Trimble, Curtin, Wagner and Woodman2020). Yet scholars have tended to take an aggregate approach to quantitative analyses of media representations of leadership and character traits. They draw on the gender stereotypes literature and previous gendered mediation studies to identify the character traits to be examined and the gender group to which they have been ascribed, if they have been ascribed at all. Studies rarely list the specific set of traits included in the coding (for exceptions, see Banwart Reference Banwart2010; Kim Reference Kim2012). Instead, one or two sentences highlight a few traits commonly associated with each gender category. For example, Kahn and Goldenberg's (Reference Kahn and Goldenberg1991, 195) influential study noted that “male” traits are “those traits that are consistently associated with men (e.g., independent, objective, competitive, strong leader, insensitive, aggressive, unemotional, ambitious, and tough), while ‘female’ traits are linked with women (e.g., dependent, non-competitive, passive, gentle, emotional, weak leader, and compassionate).”
Regardless of whether scholars code for traits individually or by gender group, they typically combine traits into three main variables—“male/masculine” traits, “female/feminine” traits, and neutral traits—when presenting the empirical results. We thus learn that journalists give greater credence to “masculine” qualities than “feminine” ones when evaluating the character and capabilities of political leaders (Cantrell and Bachmann Reference Cantrell and Bachmann2008; Meeks Reference Meeks2012), but we do not know which individual traits are noted most often. Do journalists emphasize strength and compassion but downplay objectivity and weakness? Qualitative studies employing discourse analysis and other techniques provide some answers, but these are heavily dependent upon the character, context, and communication strategies of the woman politician receiving media attention (cf. Ejaz Reference Ejaz2018).
A parsimonious approach makes statistical analysis tidier, especially when a study examines several aspects of political reporting, but it ignores the possibility that not all “masculine” traits are seen as important and not all “feminine” traits are regarded as irrelevant. As a result, we cannot determine which traits journalists consider to be vital to effective political leadership, nor can we track possible changes in media discourses over time. An alternative approach would recognize the possibility that journalists might have reinterpreted political leadership traits in ways that include the “feminine” alongside the “masculine” ones, thus creating a discursive opening for women politicians to be fully accepted as legitimate political leaders.
Studies analyzing reporting about distinct leadership competencies reveal important nuances in media coverage of political leadership. Examining coverage of party leadership contests in Canada between 1975 and 2012, Wagner, Trimble, and Sampert (Reference Wagner, Trimble and Sampert2019) discovered that language skills, communication skills, intelligence, and experience were valorized, but women candidates were generally subjected to higher levels of, and more negative, mediation in these areas than their male counterparts. Journalists were more likely to highlight men's intellect, knowledge, and expertise and downplay women's experience in elected office and cabinet positions. Only fluency in Canada's two official languages—English and French—was evaluated the same way for women and men. By examining individual traits such as intelligence and experience, these authors identified which particular attributes the media consider most important for political leadership.
Understanding media discourses about political leadership is important considering that research on public perceptions of women politicians has found evidence of trait expansion (Brooks Reference Brooks2013; Schneider and Bos Reference Schneider and Bos2014). Women's habit of incorporating both “feminine” and “masculine” traits into their political personas is helping to broaden the range of traits that a political leader can possess. Deborah Jordan Brooks (Reference Brooks2013, 169) argues that women politicians have become a subtype of women with their own set of gender stereotypes: “[H]aving seen many examples of strong, tough, empathetic female political leaders in their lifetimes, most people probably have available a mental category—a subtype—available for political women.” Generally speaking, a subtype “is characterized as a new category with its own unique stereotypical characteristics, while a subgroup shares many characteristics with the larger stereotype category” (Aaldering and Van Der Pas Reference Aaldering and Van Der Pas2020, 915). Women politicians are a subtype because their unique collection of “feminine” and “masculine” traits does not neatly overlap with those of women in general. Men politicians, meanwhile, are presumed to be a subgroup of men because they are perceived to possess the same set of “masculine” traits as men in general. Since political leadership has traditionally been defined according to “masculine” standards (Duerst-Lahti and Kelly Reference Duerst-Lahti, Kelly, Duerst-Lahti and Kelly1995), men are automatically assumed to have the relevant traits to be political leaders.
What has not been considered in the gender stereotypes literature is that male politicians might also be, or be in the process of becoming, a subtype of their gender group. The political communication literature has shown that men routinely point to their family roles and experiences to demonstrate that they understand regular people and their problems and are committed to resolving those issues (Langer Reference Langer2010; Shaw and Watson Reference Shaw, Watson, Watson and Shaw2011). A family strategy is designed to humanize politicians (Trimble Reference Trimble2017), but it also allows men to claim traditionally feminine traits such as compassion, emotional intelligence, and likability that are increasingly important for political leaders. Australian prime minister Bob Hawke used this strategy in 1984 when he shared a story of his daughter's heroin addiction. The media was sympathetic while characterizing his behavior as a stark contrast to his usual “larrikin” or boisterous style (Paine and Graham Reference Paine and Graham2019). That male politicians might form a subtype of their gender group could help explain why political leadership is being reconceptualized to incorporate traditionally feminine traits.
Another limitation of the gendered mediation literature is its primary focus on election campaigns. Much of what we know regarding the nature of political reporting and its detrimental effects on depictions of women politicians is derived from studies that examine election coverage (cf. Beail and Longworth Reference Beail and Longworth2013; Falk Reference Falk2010; Gerrits et al. Reference Gerrits, Trimble, Wagner, Raphael and Sampert2017; Gidengil and Everitt Reference Gidengil and Everitt2000, Reference Gidengil, Everitt and Cross2002, 2003a, 2003b; Heldman, Carroll, and Olson Reference Heldman, Carroll and Olson2005; Vida; Correa Reference Vidal Correa2020; Wagner, Trimble, and Sampert Reference Wagner, Trimble and Sampert2019; Williams Reference Williams2017). With some exceptions (e.g., Hayes and Lawless Reference Hayes and Lawless2015, Reference Hayes and Lawless2016), these studies have found consistency over time in the trivialization of women's candidacies (Falk Reference Falk2010; Heldman, Conroy, and Ackerman Reference Heldman, Conroy and Ackerman2018).
Recent studies of postelection periods suggest that gendered mediation remains pronounced. In her longitudinal study of Australia, Canada, and New Zealand, Trimble (Reference Trimble2017) discovered that the gender stereotypes employed by the news media persisted when covering female prime ministers after an election, only lessening once the woman had been in office for a long time. Thomas and colleagues (Reference Thomas, Harell, Rijkhoff and Gosselin2020) similarly found that Canadian women premiers received less, and more gendered, coverage in their first year in office than did their male counterparts. Examining coverage of Dutch cabinet ministers, Aaldering and Van Der Pas (Reference Aaldering and Van Der Pas2020, 924) determined that the media discussed men's political leadership abilities significantly more often than they did women's after the election, but they found no such gender difference in election coverage. One reason for this difference might be that during election campaigns, “journalists are extremely conscious of the norm of fair reporting and aim to avoid any semblance of bias” (Aaldering and Van Der Pas Reference Aaldering and Van Der Pas2020, 922). For example, in their study of media coverage of male and female candidates in U.S. House of Representative elections, Hayes and Lawless (Reference Hayes and Lawless2015) found that sex bore little relationship to newspaper coverage or to the reporting of traits such as competence, leadership, integrity, and empathy. The only statistically significant difference discovered was that women candidates received slightly more attention than did men on the positive competence dimension of leadership trait coverage (Hayes and Lawless Reference Hayes and Lawless2016, 71). They argue this is partly because the perceived “novelty” of women candidates has diminished, but also because campaign reporting is expected to be unbiased and focused on the contest. That said, once the election has passed, the media may be less attentive to this requirement of gender neutrality, when politics becomes “business as usual.” In total, then, this scholarship to date reveals that while gender stereotypes and the representation of personal traits may be shifting over time, damaging gendered mediation of candidates and leaders in election campaigns persists. Whether this holds in a postelectoral context, however, remains understudied, and it is this gap that our research seeks to fill.
DATA AND METHODS
Our study advances knowledge about media representations of political leadership by analyzing the gendered aspects of news coverage of the leadership qualities of 22 Australian and Canadian premiers. The women premiers held office between 2007 and 2017, while their male predecessors held office between 1998 and 2017. These two countries are often compared because of similarities in their Westminster-style parliamentary institutions and federal systems of government (Chappell Reference Chappell2002). Canada features 10 constitutionally autonomous subnational units, called provinces, while Australia has six states, thereby offering opportunities for women to rise to powerful and highly visible government leadership roles. Fifteen women have held the premiership in the two countries, 11 within the same decade (2007–17). Over this period, six women attained the premiership in Canada and five in Australia. Notably, four women, two in each country, were the second in their jurisdiction to hold the office of premier. Comparing news coverage of 11 women and their immediate male predecessors offers a large sample of similarly situated government leaders. Appendix A in the supplementary material online provides details regarding the premiers, their political affiliations, and terms in office. We excluded premiers of territorial governments to maximize comparability. While the federal constitutional arrangements differ somewhat between Australia and Canada, the countries are similar in that states and provinces hold significantly more power and autonomy over law and policy than do their territorial counterparts. As a result, territorial premiers have less political and media prominence than do their more senior subnational counterparts.
To understand the gendered dimensions of media discourses about political leadership, it is necessary to examine news coverage of politicians at different stages of their careers. As already noted, much of the gendered mediation research focuses on elections, whether it be primaries, general elections, or leadership campaigns. Because few studies explore nonelectoral periods (Van der Pas and Aaldering Reference Van der Pas and Aaldering2020), we know little of the extent to which reporting of leadership skills might vary across the political life cycle. While media discourses can change over a politician's career (Trimble Reference Trimble2017), our study examines first impressions of a newly selected government leader. If gendered mediation does occur in nonelectoral periods, we would expect it to be reflected in the days following the ascension of a new head of government, when journalists are most interested in leadership potential. Of course, comparisons of pre- and postselection news coverage would best illuminate variations in gendered mediation of leadership abilities, but differences in party leadership selection processes between the two countries make an equitable comparison difficult. Our study thus focuses on coverage of the first seven days after a person became premier, when journalists in Australia and Canada evaluate their ability to effectively steer the government and achieve their policy goals. Introducing new political leaders to the general public is a central task of journalists during this phase of the governing cycle. The first week of reporting captures critical first impressions of a political leader, which tend to be recirculated in media coverage through the person's political career (Trimble Reference Trimble2017). Because of variation in the timing of swearing-in practices, which range from a few hours to several weeks after a new premier is selected, we made a methodological decision to start collecting news coverage when each premier was publicly recognized as the state or province's leader. When there was a change in government, the announcement of the election winner sparked news coverage of the new premier. For those whose party was in government, data collection began when the party's choice of leader was confirmed.
News stories were drawn from the top national newspaper in the premier's country and a high-circulating newspaper in the largest media market in the premier's state or province.Footnote 2 The national newspapers are the Australian and the Globe and Mail, while the regional newspapers are Brisbane's Courier Mail, Hobart's Mercury, and the Sydney Morning Herald in Australia, and the Edmonton Journal, Montreal Gazette, Toronto Star, and Vancouver Sun in Canada. The two national newspapers were selected because of their agenda-setting power in their respective countries, within both political and media circles (Taras Reference Taras1999; Tiffen Reference Tiffen, Cunningham and Turnbull2014). The regional newspapers hold a similar position within their respective states and provinces. The Canadian newspapers are all broadsheets, while the two Australian regional papers are tabloid in format. Therefore, our study provides a mix of potentially sensationalist and serious reporting of political leadership, but the limited number of tabloid newspapers makes a statistical analysis less fruitful. Future research should examine a larger number of tabloid newspapers to assess the role of newspaper format in news coverage.
Ideologically speaking, seven of the newspapers are center-right while another two are center-left.Footnote 3 Newspaper ownership varies across the sample but not as much within each country: the Murdoch-owned NewsCorp owns three of the four Australian newspapers, while PressMedia owns three of the five Canadian newspapers. Limited variation in newspaper ideology, format, and ownership means this study cannot control for these factors in the overall or country-specific regressions. All stories featuring substantive discussion of the premiers—those mentioning them three or more times—were included in the analysis, for a total of 842 news items. Items not written by journalists or guest columnists, such as letters to the editor, were excluded.
We use content analysis to identify themes in news coverage of incoming premiers. Based on a review of the gender stereotypes, political leadership, and political psychology literatures, the study explores 16 leadership qualities that represent positive and negative versions of traditionally recognized masculine, feminine, and neutral traits. The “masculine” qualities are rationality, strength, and experience, while the “feminine” qualities are compassion, likeability, honesty, emotionality, and weakness. Eagly and colleagues (2020) found that competence and intelligence, long associated with men, are increasingly associated with women, leading us to characterize these two traits as neutral. We developed the following trait pairings to capture positive and opposite negative leadership qualities: Competence/incompetence, strength/weakness, intelligence/unintelligence, experience/inexperience, compassionate/lack of compassion, likability/unlikability, honesty/dishonesty, and rationality/emotionality. Appendix B outlines the 16 qualities and how they were measured, including key descriptors and evaluators. For example, the emotion variable captured any instance in which a premier was described as expressing an emotion, such as anger, fear, or sadness, or performing an emotion, such as crying, while rational mentions depicted the premier as being composed, measured, levelheaded, or keeping their nerve.
Coding was a three-stage process. The first stage involved developing a codebook that identified a potential corpus of key words and phrases for each of the leadership qualities. The entire research team discussed what each variable was intended to capture before initial coding began. The second stage involved a preliminary discourse analysis, whereby the three coders highlighted all mentions of leadership traits in a news story but only assigned codes to those that fit into one of the variables. The third stage involved employing data analysis software to identify overlooked mentions. Using MAXQDA, lexical searches found and coded all instances of words and phrases associated with each leadership quality. Every story was reviewed by coders to detect any further qualities not identified in the initial hand-coding or computer-coding exercises. Excel files listing all mentions for each leadership variable were then produced.
Using the online utility ReCal to calculate results (Freelon Reference Freelon2010), intercoder reliability tests on a random sample of stories generally found strong levels of agreement among coders (Cohen Reference Cohen1960), with the notable exceptions of the competence, likability, and strength variables. To resolve this disagreement and as a further check on reliability, the three coders independently assessed the appropriateness of each mention for all 16 variables—including competence, likability, and strength—and then met to resolve disagreements. A final round of assessments was conducted to double check the results, and the codebook was updated to reflect the coding decisions. This rigorous approach resulted in a comprehensive corpus of leadership quality mentions in news coverage about the 22 premiers.
The quantitative analysis conducted for this study featured a series of binary logistic regressions with the 16 individual leadership qualities as the dependent variables. The statistical model investigates the influence of premier gender on media discourses about political leadership skills. The gendered mediation thesis, which posits that women politicians will experience more extensive evaluation of their leadership abilities than men politicians, is directly tested by a dichotomous premier gender variable, with women premiers coded as 1 and men premiers coded as 0. A goodness-of-fit test revealed that the model for two of the 16 leadership qualities—unintelligent and lacking compassion—failed to achieve statistical significance because of infrequent use and thus cannot capture the reasons why journalists did, or did not, attach these two traits to political leaders. Our analysis focuses on the remaining 14 leadership qualities.
The model also considers alternative explanations for media discourses, including gender novelty, pathway to office, sex of reporter, and country effects. Journalists routinely note historic developments in politics, including politicians who are the first of their social group to hold a political position. The novelty of a woman becoming a jurisdiction's first premier could result in more extensive media assessments of their suitability for office than is the case for non-novel women premiers.
Politicians who become premier by winning the leadership of the governing party might see more media discussion of their leadership abilities than those politicians who become premier by leading their party to victory in a general election. In Australia, leadership contests differ across the two major parties and across states—historically most have taken place within the parliamentary caucus and, as a consequence, have been relatively brief, meaning that journalists have little opportunity to evaluate the leadership abilities of prospective candidates. Leadership contests in Canada take longer but often receive limited news coverage, with the notable exception of those occurring within governing parties (Sampert et al. Reference Sampert, Trimble, Wagner and Gerrits2014). Just three of the 10 Australian premiers came to power through a general election (one woman and two men) while five of the 12 Canadian premiers did so (two women and three men). Selection processes aside, Australia and Canada might also differ in terms of the journalistic cultures shaping political news coverage. We included a country variable (Canada = 1) to account for potential differences between the two media cultures in regressions on the whole data set.
Finally, we included a series of controls: type of story (opinion = 1, other types of news stories = 0), newspaper type (national newspaper = 1, regional newspaper = 0), and word count (number of words in each story). Type of story might shape emphasis on leadership skills. While editorial writers are expected to provide opinions about politics, reporters writing regular news stories are not as likely to offer overt evaluations of leadership qualities. National newspapers might limit their coverage of subnational leadership transitions to information of greatest interest to a national audience. Readers of regional newspapers, in contrast, expect more fulsome and detailed reporting of newly elected leaders. Longer stories give journalists more space in which to offer leadership assessments than do short news pieces.
Longitudinal research suggests journalistic assessments are generally consistent over time. Falk (Reference Falk2010) found few variations in coverage of women seeking the American presidency over a 136-year period, from 1872 to 2008. Journalists consistently described female hopefuls as unfit and incapable of being president. Trimble (Reference Trimble2017) found a similar consistency in coverage of female prime ministers in Australia, Canada, and New Zealand, regardless of whether the woman became prime minister in 1993 or in 2010. These doubts only disappeared for the lone long-serving female prime minister, who had time to demonstrate her leadership qualities. Examining coverage of new premiers between 1998 and 2017 provides an opportunity to assess whether media representations of subnational political leaders have changed as more women rose to positions of power and became highly visible in these roles. The model would ideally control for the year in which a story was published, but analyses discovered this variable had high multicollinearity with premier novelty. To preserve our ability to assess potential differences in how the media cover groundbreaking politicians, the regressions do not control for year. However, we do track potential changes in media discourses over time through one-way ANOVAs.
RESULTS
Descriptive and bivariate results, presented in Figure 1, indicate that journalists emphasize a mix of “masculine” and neutral traits when discussing the leadership qualities of new government leaders. A premier's prior experience in politics, business, and other domains of life is noted in 39.3% of all news stories, followed by strength (30.4%) and competence (22%). Journalists clearly care about whether premiers understand their job and have the will and ability to do it well. Less emphasized in news stories are a premier's likability, honesty, intelligence, and compassion. These results provide initial support for assertions that political journalism is inherently masculine in outlook (Sreberny-Mohammadi and Ross Reference Sreberny-Mohammadi and Ross1996). Journalists pay close attention to the ways in which premiers embody leadership qualities associated with men, such as experience and strength, but less so to those traits associated with women, such as likability, honesty, and compassion. Because political leaders are expected to be rational, journalists also keep a close eye on emotional displays (mentioned in 14.3% of news stories) but waste little space discussing a premier's composure (3.9%).
Figure 1. Percentage of stories mentioning different types of leadership qualities of premiers, by gender and overall.
The lone result that challenges traditional assumptions about the masculine underpinnings of political leadership relates to competence. Not surprisingly, journalists believe an effective government leader must know how to do his or her job, and recent research indicates that this trait is increasingly associated more with women than with men (Eagly et al. Reference Eagly, Nater, Miller, Kaufmann and Sczesny2020). Changing public perceptions of women's abilities have thus created an opening for women to be taken more seriously by journalists as credible political leaders. The COVID-19 crisis is a recent example. Media assessments of how global leaders have responded to the pandemic suggest that women's leadership skills are gaining new respect (Henley and Roy Reference Henley and Roy2020), enabling women politicians to disrupt the binary nature of stereotypes, but early scholarly research on COVID-19 reporting suggests this might not be a uniform response (Aldrich and Lotito Reference Aldrich and Lotito2020; Smith Reference Smith2020).
Our study reveals that journalists give equal attention to the competencies of women and men premiers across time. A series of one-way ANOVAs examining media depictions of competence during three periods found no gender difference in the average number of competence mentions per story for premiers. Between 2003 and 2007, women premiers received an average of 0.23 mentions per story compared to 0.26 for men premiers, while both groups received 0.17 mentions per story between 2008 and 2012. Women experienced a slight advantage between 2013 and 2017, when they received 0.30 mentions per story compared to 0.29 mentions for men. While our results do not support a conclusion that competence is increasingly associated with women, they do demonstrate that women are seen as every bit as competent in the political realm as their male counterparts.
In fact, Figure 1 reveals that journalists evaluate women and men government leaders according to the same set of leadership characteristics. More than one-third of stories about women (38%) and men (40.5%) discussed their cabinet experience, elected offices, party involvement, union activities, careers, education, and volunteer experience.Footnote 4 A similar pattern holds for strength, which was noted in 30.8% of stories about men, compared to 30% for women premiers. One in five stories about women (22.3%) discuss their competence, a figure that is slightly higher than for men (21.6%). Interestingly, journalists describe men (15.3%) in more emotional terms than women (13.2%) but this result was not statistically significant in our crosstabulations.
The only major difference in media representations of leadership traits related to unlikability: men (8%) were almost four times more likely than women (2.5%) to be described as abrasive, sharp-tongued, autocratic, and not well liked by others. This result, however, was largely driven by media depictions of the period between 2003 and 2007, when men dominated the premiership roles. Only one woman was premier during this period (Anna Bligh was elected in September 2007, and the 2003–07 female figures reflect her coverage).Footnote 5 Male leaders received an average of 0.13 unlikable mentions per story compared to women's 0.03. Men's unlikable mentions dropped dramatically after that time period—to the point that gender differences largely disappeared between 2008 and 2017. One possible explanation is that women's increasing presence in government leadership positions after 2007 forced men to become more likable to appeal to more voters and, thus, party selectors. If this is indeed the case, men's adoption of a “feminine” trait to position themselves as electorally competitive is evidence of men politicians becoming a subtype of their gender group.
Other comparisons found journalistic mentions of individual leadership traits generally fluctuated over time, both overall and for each gender group. However, two notable exceptions emerged from the series of one-way ANOVAs that examined results across four periods. First, journalists increasingly highlight women premiers’ intelligence, from an average of 0.03 mentions per story between 2003 and 2007 to 0.12 mentions between 2013 and 2017. Men's intelligence mentions were largely consistent during the same time period, hovering slightly above or below 0.07 mentions per story. These results align with recent psychology research that suggests intelligence is increasingly regarded as a female trait (Eagly et al. Reference Eagly, Nater, Miller, Kaufmann and Sczesny2020). Second, mentions of the compassionate qualities of men premiers grew over time. No mentions occurred between 1998 and 2002, then 0.03 mentions per story between 2003 and 2007, and from 2008 and beyond we see 0.05 mentions in 2008. Women's compassion mentions were largely consistent across time periods, hovering slightly above or below 0.04 mentions per story. Men's compassion depictions are neither plentiful nor the change over time dramatic, but they could be indicative of an incremental move toward “feminine” qualities being attributed to male politicians (Langer Reference Langer2010).
Overall Analysis
Table 1 presents the binary logistic results for 14 of the 16 types of leadership mentions. The overall regressions reveal that, when controlling for a range of alternative explanations for news coverage, women premiers were not usually subjected to greater mediation of their leadership abilities. Length of the story was a factor, with journalists discussing a premier's leadership abilities when they had more space to do so. Pathway to office also mattered: when the premier's party was already in government when they were selected, news stories gave more attention to their leadership traits. Australian journalists were also significantly more likely than their Canadian counterparts to note a new government leader's strength, weakness, inexperience, and dishonesty. Canadian journalists were more preoccupied with a premier's capacity to execute the role competently and rationally. For their part, columnists and editorial writers placed greater emphasis on a premier's competence and intelligence than did reporters writing regular news stories. These overall regression results provide further evidence that journalists, for the most part, did not place differing emphasis on the leadership abilities of women and men premiers.
Table 1. Binomial logistic regression results for different types of leadership qualities mentioned in newspaper coverage of Australian and Canadian premiers
Notes: N = 842. The table reports the B, followed by the SE in parentheses. Models for the following variables did not achieve goodness-of-fit and thus were not analyzed: unintelligent and lack compassion.
* p < .05; ** p < .01; *** p < .001.
While multivariate results demonstrate that women premiers can generally expect similar media treatment as their male counterparts, exceptions in news coverage suggest that both men and women have their own challenges in being seen as credible government leaders. First, Table 1 reveals that journalists devoted significantly more space to discussing men's unlikability and emotions. The latter finding is especially problematic for male leaders because this trait contradicts gender stereotypes that men are rational. As women have learned, gender-role incongruity runs the risk of triggering negative evaluations (Gidengil and Everitt Reference Gidengil and Everitt2003b). The former result might not be as harmful because men are not expected to be likable; the media normally direct this expectation at women (Cauterucci Reference Cauterucci2019). The likability result might therefore be a product of women politicians working harder to portray themselves as likable to avoid this censure.
The second exception to gender equality in media representation of leadership abilities relates to novelty. Journalists were significantly more likely to discuss the emotions of women premiers who were the first in their province or state to hold that position compared to women who were the second to be premier. Politicians are expected to be careful about showing emotions for fear of being seen as mentally unstable and thus unfit for the job. Women politicians have been especially vulnerable to this criticism (Falk Reference Falk2010). After delivering a scathing indictment of sexism in Australian politics in 2012, Prime Minister Julia Gillard was depicted by the news media as “having lost control of her emotions” and being “incapable of handling the pressure of political office and leadership because she is a woman” (Wright and Holland Reference Wright and Holland2014, 463; see also Curtin Reference Curtin2015; Johnson Reference Johnson and Feldman2020; Williams Reference Williams2017). Journalists who are not used to a woman in a powerful political position might be hyper-attentive to emotional displays. Yet, as the regression reveals, this surveillance disappears once journalists see other women become premier, suggesting the normalization of women in executive office might mitigate gendered mediation.
Country Comparisons
The overall analysis reveals few gendered differences in how the news media cover the leadership abilities of newly selected government leaders in Australia and Canada. A similar pattern emerges when we examine media behavior within each country. Table 2 presents the binary logistic results for Australia, while Table 3 does so for Canada. To ensure comparability, the tables only present results for the eight leadership abilities for which the model achieved goodness-of-fit in both countries: competence, incompetence, strength, intelligence, experience, compassion, likability, and emotionality. As the tables show, Australian and Canadian journalists treated women and men leaders in largely the same manner. Scribes in both countries paid the same amount of attention to women's and men's strength, intelligence, compassion, likability, emotions, and incompetence. These findings indicate that women leaders in Australia and Canada are not disadvantaged when it comes to quantitative representations of their leadership abilities, both good and bad traits.
Table 2. Binomial logistic regression results for different types of leadership qualities mentioned in newspaper coverage of Australian premiers
Notes: N = 351. The table reports the B, followed by the SE in parentheses.
* p < .05; ** p < .01; *** p < .001.
Table 3. Binomial logistic regression results for different types of leadership qualities mentioned in newspaper coverage of Canadian premiers
Notes: N = 491. The table reports the B, followed by the SE in parentheses.
* p < .05; ** p < .01; *** p < .001.
Where the two countries differ are in discussions of competence and experience. Table 2 shows that Australian journalists devoted significantly more space to highlighting the competence of women leaders than of men leaders. No gender differences were found in mentions of a new leader's prior experience, such as in politics or business. These results indicate that newly selected female premiers are presented to the Australian public as having the skills and experience to perform well in their new role. The opposite is true in Canada. As Table 3 shows, Canadian journalists placed much greater emphasis on men's competence and experience than they did for women. These findings are problematic considering that journalists rate experience as the most important quality of a government leader. By paying greater attention to men premiers’ experience and competencies, Canadian journalists send the message that men are more qualified than women to be government leaders in that country.
The final gender difference related to premier novelty. Australian journalists were significantly less likely to note the competence of women who were the first in their jurisdiction to become premier than of the women who came after them. In other words, trailblazing politicians were not presented as able leaders to the Australian public. Subsequent cohorts do not experience this disadvantage. As with competence and experience, Canadian journalists moved in the opposite direction as Australia, showing an inclination to highlight the competence of novel premiers more so than of non-novel ones but this result did not achieve statistical significance. Canadian journalists were more emphatic in their emphasis of a novel premier's strength. Perhaps these variations reflect the impact of gender novelty at the national level in the two countries. Canada's first woman prime minister secured the job in 1993, at a time when few women had held leadership roles at either level of government. While Kim Campbell's tenure was brief, it drew considerable media attention to the gender angle (Trimble Reference Trimble2017) and might have conditioned the media to see women as experienced and capable leaders. In Australia, Julia Gillard's ascent to the prime minister's office occurred in 2010, and also attracted much gendered media analysis (Holland and Wright Reference Holland and Wright2017; Williams Reference Williams2017; see also Curtin Reference Curtin2015). While five women had served, or were serving, as state premiers prior to Gillard's promotion, it might be that the novelty of the first woman prime minister influenced the gendered stereotypes and assumptions represented in the Australian media.
Differences between the two countries, while not overwhelming, point to the importance of comparative analysis in understanding gendered mediation. Political reporting in Western liberal democracies might share many features, but even subtle differences between countries could have a major impact on the media environment encountered by women political leaders.
DISCUSSION AND CONCLUSION
By examining the role of gender stereotypes in media representations of leadership abilities, this study demonstrates that journalists evaluate women and men premiers to the same degree and based on the same set of leadership qualities. All political leaders are expected to perform their duties with competence, intelligence, and levelheadedness. Further, those qualities are not exclusively seen as masculine. Newspaper coverage continues to prioritize the traditionally “masculine” qualities of experience and strength and downplay the putatively “feminine” qualities of honesty and compassion. These findings lend support to the gendered mediation thesis, which asserts that the news media use a masculine filter when reporting on politics in general and on women politicians in particular. Yet the media expect new government leaders to demonstrate competence and be able to work collaboratively with others, qualities associated with women (Eagly et al. Reference Eagly, Nater, Miller, Kaufmann and Sczesny2020; Gerrits et al. Reference Gerrits, Trimble, Wagner, Raphael and Sampert2017). Journalists are also slowly increasing their depictions of men politicians as compassionate and women as intelligent. These results suggest media conceptualizations of political leadership are slowly challenging traditional associations between gender and particular leadership traits. This evolution is significant insofar as it could be creating a discursive opening for women to be seen as effective political leaders. A gender-balanced understanding of political leadership might also lead men politicians to, one day, become a subtype of their gender group as they try to claim “feminine” qualities deemed desirable in a leader. We believe that scholars must continue examining the political communication strategies of male politicians to understand when, and why, men make such appeals when highlighting their leadership capabilities and suitability.
While media discourses might be changing, the media's continuing emphasis on traditionally masculine leadership qualities remains evident in their monitoring of these premiers’ emotional states, but not their rationality. Journalists expect rationality in all political leaders, and so we can assume they see no need to emphasize this quality. Instead, they monitor emotional transgressions seen to violate the presumption of rationality. That women and men premiers received equal attention to their emotional displays, however, is contrary to previous findings in the gendered mediation literature (Falk Reference Falk2010). One explanation could be the timing of the leadership assessments. In an election context, journalists might reassert traditional masculine boundaries of political leadership by emphasizing the emotions of women politicians vying to become government leader, subtly suggesting they are not suitable for the position. But a certain amount of emotionality expressed when an individual achieves the top job might be regarded as appropriate to women and men alike.
What this study cannot determine is how journalists discussed each leadership quality. For example, did journalists highlight different kinds of leadership competencies for women and men? And how were these skills evaluated in terms of women's and men's ability to be effective premiers? Future research should go beyond a quantitative analysis to identify the latent and manifest messages circulating in media discourses about political leadership. Women and men leaders might receive equal amounts of coverage for their qualifications, but those qualifications might be discussed in ways that either continue to valorize men as political leaders or challenge traditional conceptions of leadership in ways that create space for women. A qualitative analysis is needed to assess potential gender differences in how journalists discuss individual leadership skills such as competence and strength. For example, media discourses about strength might depict men as possessing physical prowess and determination but depict women as unnaturally aggressive. Media discussions of other traits might advantage women. We found evidence of the latter when it came to competence. Journalists, politicians, and voters have identified the ability to work well with others as an important quality for political leaders, which is contrary to the traditional expectation that a leader exercise power over, rather than power with, others (Gerrits et al. Reference Gerrits, Trimble, Wagner, Raphael and Sampert2017). A qualitative analysis found that collaboration was read the same for both women and men, resulting in women being portrayed as having an important leadership skill.
Finally, this examination of media behavior in Australian and Canada makes a number of contributions to understanding of mediation, gender stereotypes, and political leadership. First, by examining leadership depictions of newly selected premiers, we establish that women politicians have the same opportunities as their male counterparts to present themselves as capable leaders and that the media are destabilizing long-held distinctions between “masculine” and “feminine” leadership abilities by utilizing a wider range of abilities to evaluate political leadership for both women and men. What we cannot determine is the impact of this substantive coverage on how voters evaluate political leaders or how it shapes levels of support. Experimental research focusing on evaluations of politicians’ leadership abilities, rather than their personal qualities, would help identify the media's influence on public assessments in this regard. Second, by reviewing the first seven days of news coverage after a premier has been selected, this study advances understanding of gendered mediation during crucial nonelectoral periods. As our findings suggest, once a new leader has been chosen, journalists shift focus from campaign strategy and electoral viability to leadership capability and suitability for a governance role. Gendered mediation remains present in that “masculine” qualities remain the primary focus of news coverage, yet the inclusion of not only “feminine” interpretations of these qualities but also an equitable amount of coverage indicates that women and men premiers are largely portrayed in the same manner. A change in media expectations thus produces a different kind of gendered mediation, one that gives women a fairer shot at being taken seriously as political leaders.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/S1743923X21000131.
