Gender stereotypes in print and online media coverage of Slovak presidential candidates in 2009 and 2019

Abstract In Slovakia, women are poorly represented in politics and public life. Yet it is the first country in Central Europe with a female president. By applying a mixed-methods approach to analyzing an original dataset containing media coverage of leading presidential candidates (n = 1492), this study explores how the media covered them and discusses under what conditions gender-stereotypical coverage could be detrimental or beneficial to electoral outcomes. The results show media outlet type was not significantly associated with a gender-stereotypical attribution of communal and agentic traits to candidates. Tabloids and quality press equally perpetuated gender stereotypes. Irrespective of their gender, journalists were more likely to depict women candidates as possessing communal qualities perceived as incompatible with leadership. However, findings from the qualitative analysis suggest that when corruption perception is high, and public trust in institutions is low, communal traits stereotypically attributed to women are appreciated. Novelty also works to women’s advantage. These findings have important implications for women candidates’ campaign strategies.


Introduction
Shortly before midnight on 30 March 2019, the Old Market Hall in Bratislava broke into thunderous applause and euphoria. 'Zuzana, Zuzana!' the crowd chanted as the counted votes indicated that Slovakia had elected its first woman president. Zuzana Čaputová, a civil society activist and newcomer to politics, thanked the crowd. 'We may have thought that politeness is just a sign of weakness, and today we see it as a sign of strength', she said. 1 The values Čaputová represented in the year-long campaign are often stereotypically attributed to women and perceived as disadvantageous in the competitive political arena. Yet, she won the elections against her male opponentskilled diplomat Maroš Šefčovič, in a region where politics is male-dominated and conservative, and women are underrepresented in parliament, government, and other domains.
Gender and politics scholars have produced excellent research, carried out in other national contexts, explaining why women are less likely to hold public office. The reasons include: women having less interest in politics and lower political ambition, subjecting themselves (Fox and Lawless, 2011;Dahl and Nyrup, 2021), and being subjected to stricter qualifications and ability standards (Bauer, 2020). Gender stereotypes perpetuated by the media and voters also discourage women from actively participating in politics and affect their chances of success (Huddy and Terkildsen, 1993;Bligh et al., 2012;Bauer, 2015).
Gendered mediation scholars, that is, scholars who study the implications of gender in political communication and reporting, have predominantly focused on the descriptive representation of women political actors and how much the media cover them compared to their male colleagues (Hooghe et al., 2015;Lühiste and Banducci, 2016;O'Neill et al., 2016;Vu et al., 2018). Those who studied how they are portrayed have concentrated on women's physical appearance and personal lives (Heldman et al., 2005;Conroy et al., 2015). Far less attention has been paid to media coverage of a woman's leadership traits, which speaks more directly to whether a woman politician is portrayed as possessing what it takes to be a political leader. Few studies have examined the substantive aspects of news coverage to determine if gendering is also present in media depictions of leadership abilities, qualifications, or policy ideas (Wagner et al., 2019;Wagner et al., 2021;Trimble et al., 2022). Moreover, research concentrating on gender stereotypes of political actors in Central Europe has been in short supply (Ibroscheva and Raicheva-Stover, 2009;Maďarová, 2011;Ostertágová, 2011;Dan and Iorgoveanu, 2013).
The present study investigating how the media portrayed female and male presidential candidates in Slovakia aims to add to this literature and fill the gap in English language scholarship on this topic. While some noteworthy studies were published in the Slovak language, they either explored candidates' self-presentation strategies to voters (Maďarová, 2011) or, if they studied gendered mediation, they were descriptive (Ostertágová, 2011).
By analyzing the content of commentaries published in news outlets a year ahead of the Slovak presidential elections in 2009 and 2019, this study not only explores the differing gendered nature of media coverage but also explains how the type of news outlet affects the prevalence of gender stereotypes. Building on Fridkin and Kenney's (2014) strategic stereotype theory and research in organization psychology (Brown et al., 2011) and political science (Goetz, 2007;Barnes and Beaulieu, 2014;Funk et al., 2021), it also discusses specific circumstances in which gender stereotypes might benefit women candidates, such as increased corruption perception and plummeted public trust, and how this could have been relevant in the Slovak presidential elections. So far, only studies that debated how women candidates can capitalize on their stereotypical portrayal in the European context have focused on far-right women political actors (Snipes and Mudde, 2020).
The results of this study's quantitative analysis are consistent with previous research on this topic and conclude that news outlets reflect the society in which they operate. In Slovakia, the media tend to perpetuate a gender-stereotypical portrayal of political candidates. Interestingly, the type of outlet did not impact on the presence of gender stereotypes, which occurred in both tabloids and quality press. Similarly, journalist gender was not significantly associated with stereotypical attribution of communal and agentic traits to female and male candidates. Finally, communal traits were mentioned in connection with Čaputová significantly more than any other candidates in the 2009 and 2019 elections. The qualitative analysis of commentaries shows that journalists appreciated communal traits in leadership in 2019, which was characterized by extraordinarily low levels of trust in public institutions, the disintegration of the rule of law, and low social cohesion. Čaputová's novelty was also depicted as a great advantage.
The study advances knowledge on gendered media coverage as well as challenges the traditional narrative in gender research that communal traits are to the detriment of women politicians. By examining the Slovak context, this study adds to the empirical studies outside of the North American and Western European contexts. It demonstrates that certain political circumstances can improve the media coverage of communal traits and women politicians' leadership abilities in executive offices. Finally, the findings also have important implications for women candidates' campaign strategies.

Invisible women: political representation in Slovakia
Before proceeding to the study, I briefly present the political context in which two major female presidential elections took place and their candidates: female candidate Iveta Radičová and male incumbent Ivan Gašparovič in 2009 and female candidate Zuzana Čaputová and male candidate Maroš Šefčovič in 2019. I also discuss how Slovakia fares regarding women's representation and public opinion towards women and their role in society.
Slovakia is a landlocked country with 5.5 million inhabitants located in Central Europe. Many consider the country to be a model of a successful transition from a centrally planned to a freemarket economy. After the 1998 government change, when Mikuláš Dzurinda of the center-right Slovak Democratic and Christian Union (SDKÚ-DS) replaced the authoritarian Vladimír Mečiar in Prime Minister's role, the country experienced one of the highest GDP growth rates in the EU. 2 Two successive SDKÚ-DS-led governments (1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006) adopted several economic reforms, contributing to Slovakia's EU membership in 2004 and the eurozone membership in 2009. However, despite economic growth, the population did not favor some of the liberal economic reforms, especially flat-tax and pension reform. A left-leaning social-democratic party Smer-Social Democracy (Smer-SD) seized public discontent as an opportunity and successfully campaigned in 2006 around topics such as the increase of social welfare programs and the reintroduction of progressive taxation. In the 2009 presidential election, Radičová was nominated by SDKÚ-DS, which was declining in popularity, and Gašparovič by Smer-SD, whose popularity was growing. However, while in 2009, Smer-SD support was not harmful to Gašparovič, it was detrimental to Šefčovič in 2019. Then, the reporting of murdered investigative journalist Ján Kuciak revealed large corruption schemes linked to Smer-SD. His murder encouraged unprecedented public protests. Public trust in institutions plummeted, recording its historically lowest point in 2018 when only 21% of the population trusted the government and parliament (European Commission, 2018b). Čaputová was a political newcomer backed by an extra-parliamentary party (Progressive Slovakia). So, from the perspective of partisan politics, female candidates found themselves in different positions in 2009 and 2019.
That said, conservative attitudes toward women's social roles have always been prevalent in Slovakia. In 2009, women in key decision-making positions in politics and other domains were underrepresented (see Table 1). Slovakia regularly scores poorly in major gender equality rankings and opinion polls on gender issues. In 2015, compared to other EU nations, Slovaks were most likely to say that tackling gender equality should not be an EU priority (European Commission, 2015). They also did not believe that women face prejudices because of preconceived ideas about social roles (European Commission, 2017).
These attitudes are likely to affect voting behavior. At the time of Čaputová's campaign in 2019, the situation had not changed from a gender perspective. Slovakia was still one of the worstperforming EU member states in the Gender-Equality Index 3 and Global Gender Gap Index. 4 Women's representation in politics saw minimal improvement. Slovaks stereotypically believed that women were more likely than men to base their decisions on emotions and perceived the importance of traditionally divided roles (European Commission, 2018a). Overall, ten years apart, politically, the situation was different for presidential candidates. However, from a gender perspective, female presidential candidate Čaputová did not find herself in a more favorable position than her predecessor Radičová.

Studying gendered coverage of political candidates
In this study, I employ several theoretical frameworks from social psychology, political science and communication studies, to investigate how candidates in the 2009 and 2019 Slovak presidential elections were presented in the selected news outlets. First, I define stereotypes and their implications for women's careers and success in high-level politics. Then, I discuss the role of media in reinforcing and dismantling gender stereotypes about women political candidates. Last, I summarize research that suggests that certain sociopolitical circumstances might favor women's leadership and communal traits.

The origins and consequences of gender stereotypes
From getting a slightly different recommendation letter than their equally qualified male classmate (Schmader et al., 2007) to being questioned as a suitable presidential candidate for being overly ambitious against a male candidate with zero political experience (Bracic et al., 2019), women's chances for success are hindered by many hidden obstacles which stem from stereotypes. Stereotypes are an intrinsic part of how we make sense of the reality surrounding us and think of different categories. They operate automatically and are easily activated in the absence of information about individuals. These information gaps are conveniently filled with available group representations (Eagly et al., 2000;Eagly and Karau, 2002).
Gender stereotypes emerge from the observed social roles traditionally attributed to women and men. According to the social role theory, the dissimilarity of these roles produced gender differences in social behavior (Bakan, 1966;Bem, 1981;Eagly and Steffen, 1984;Eagly et al., 2000). While men historically occupied the public sphere and participated in the labor force, women were associated with the roles of homemakers confined to the private domain. As these roles require different characteristics and skills, women and men have become associated with traits on opposite sides of the agency-communality scale (Bakan, 1966;Bem, 1981). Traditionally viewed as homemakers, women continue to be viewed as more selfless, compassionate, cooperative, and willing to compromise than men. These traits are perceived as high in communality. In contrast, the characteristics associated with men are seen as deficient in communality but high in agency. Their roles are viewed to require rationality, competitiveness, ambition, and the ability to withstand pressure. The continuous experience of women performing some social roles, but not others, nourishes stereotypes and affects performance expectations placed on them (Correll and Ridgeway, 2003). Eagly et al. (2000) showed that people automatically think of a woman when a profession perceived to require communal traits is mentioned. However, once a task necessitates agentic traits, women are at a disadvantage. As most high-status positions, including political careers, are largely viewed as requiring agency, women do not have equal chances to succeed in them, as they are seen as unfit. As a consequence, stereotypes preserve 'power and prestige structures' (Correll and Ridgeway 2003: 29). This grounding literature on gender stereotypes in the USA has been widely applied by scholars in different geographical and cultural contexts, largely coming to the same conclusions (López-Sáez et al., 2008;Steinmetz et al., 2014;Froehlich et al., 2020).

Explaining gendered media coverage
Citizens in democratic countries turn to media daily to search for information (Zaller, 1992(Zaller, , 2003. Serious journalism in the high-quality press or public service media that provides in-depth analyses helps the public to keep informed (de Vreese and Boomgaarden, 2006;Fraile and Iyengar, 2014) and enables them to participate in public life (Schudson, 1998). Media frame political actors, events, and policies in a way that emphasizes 'a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described' (Entman 1993: 52). Such framing has a profound influence on public opinion and voting behavior (Schlehofer et al., 2011: 70). Unavoidably, it is not free from stereotyping.
As mentioned, gendered mediation research has predominantly focused on the descriptive representation of women political actors, that is, how much voice is given to them. Research looking at how women political actors are substantively represented in the media indicates that this is centered around women's physical appearance and personal lives (Heldman et al., 2005;Conroy et al., 2015). Only a few studies have looked at the stereotypical depiction of women and men political actors in the media with regard to personal or leadership traits. Evidence from Canada, tracing back the portrayal of candidates for political party leaders since 1975, confirmed gender incongruence with women candidates being subjected to more negative and gendered evaluations of their communication skills, intelligence, and political experience, compared to their male challengers (Wagner et al., 2019).
Differences in media coverage of women and men can be explained by the type of outlet and societal-level structural and occupational gender inequalities in a country (Shor et al., 2015;Humprecht and Esser, 2017). In other words, women's low political representation translates into the underrepresentation of women's voices in the media. As for the outlet type, evidence suggests that compared to men, women appear in tabloids to a greater extent than in the quality press due to the soft news angle that tabloids adopt in their reporting (Dan and Iorgoveanu, 2013;Humprecht and Esser, 2017). At the same time, this soft news angle contributes to greater stereotyping.
However, while newer evidence from Canada and Australia suggests that journalists continue to emphasize agentic traits, such as experience and strength, as necessary for leadership positions and downplay communal traits, such as honesty and compassion (Wagner et al., 2021), the traditionally 'feminine' trait of collaboration has gained prominence as a crucial leadership ability (Wagner et al., 2021;Trimble et al., 2022).
Building on this literature, I propose the following hypotheses. First, considering the prevalence of conservative attitudes towards women's social roles and their low political representation in Slovakia, I still expect the media to perpetuate the gender stereotypes present in the society: Hypothesis 1a. The news outlets stereotypically depicted women candidates in the Slovak presidential elections in 2009 and 2019 as more communal than their male counterparts.
Hypothesis 1b. The news outlets stereotypically depicted male candidates in the Slovak presidential elections in 2009 and 2019 as more agentic than their female counterparts.
Second, I propose the hypotheses linked to the type of outlet: Hypothesis 2a. Tabloid news outlets were more likely than quality news outlets to depict women presidential candidates stereotypically as possessing communal qualities.
Hypothesis 2b. Tabloid news outlets were more likely than quality news outlets to depict men presidential candidates stereotypically as possessing agentic qualities.
Can gender stereotypes benefit women candidates?
If the media stereotypically portray women political candidates, are these stereotypes always detrimental to them? Bligh et al. (2012) have shown that how the media depict women affects their likability and chances to win. According to the role congruity theory of prejudice toward women leaders, women's communal portrayal is incongruent with how we perceive leadership roles and what we expect from those who occupy them (Eagly and Karau, 2002). This discrepancy produces a prejudice whereby women are deemed to have less leadership ability. So, if women are portrayed in a gender-stereotypical way, their electoral chances might lessen. Cassese and Holman (2018) found that campaign attacks, centered around stereotypical traits, harmed mostly women candidates who were criticized for not being 'strong enough' to deliver on the promises if elected. In Slovakia, these findings echo those of Ostertágová (2011). However, breaking away from gender norms triggers disapproval (Correll and Ridgeway, 2003). For example, although Hillary Clinton used masculine language more over time, it did not help her achieve victory (Jones, 2016). She was seen as cold and unlikable. If women candidates adopt stereotypically agentic traits, for example, their campaigns are hostile towards other candidates, their public perception becomes negative (Krupnikov and Bauer, 2014). At the same time, Fridkin and Kenney (2014) argued in their strategic stereotype theory that not all communal traits stereotypically attributed to women harm them. For example, stereotypes that present women candidates positively as warm, compassionate, and caring might benefit their campaigns, and women candidates can capitalize on them. Also, the ability to collaborate, which used to be perceived as a traditionally 'feminine' trait, is now increasingly viewed as necessary for leadership (Wagner et al., 2021;Trimble et al., 2022).
Other scholars showed that in times of crisis, when social cohesion is threatened, and trust is low, stereotypical attribution of communal traits to women might encourage a preference for women leaders. Brown et al. (2011) demonstrated on the organizational level that extraordinary times create a window of opportunity for women's leadership. When companies perform poorly, women are more likely to get offers for high-level managerial positions. In times of widespread discontent, increasing corruption perception, and lack of trust, women are called to the rescue also on a political level (Funk et al., 2021). Parties are more likely to nominate them than in normal times.
Preference for women in these specific circumstances can be partly explained by the link between greater women's representation and lower corruption. This link is particularly robust in electoral democracies with functional mechanisms for holding political actors accountable (Esarey and Schwindt-Bayer, 2018). Compared to men, women are stereotypically viewed as more honest and, thus, less corruptible. As a result, some countries where political corruption is a major problem, staffed public-facing roles with more women (Goetz, 2007) and adopted electoral gender quotas .
While some scholars relied on stereotypes in explaining the correlation between the greater share of women in politics and lower corruption (Dollar et al., 2001), others disentangled the causal logic of this relationship and demonstrated that women are perceived as less corruptible due to their risk-aversion (Schulze and Frank, 2003) and marginalization within politics (Barnes et al., 2018; Barnes and Beaulieu, 2019). Women are locked out of positions of power and are generally less connected with fewer economic resources than men (Bjarnegård, 2013(Bjarnegård, , 2018Rothstein, 2018). Yet, access to power, resources, and networks are necessary for corrupt behavior. As women candidates are perceived as (1) risk-averse and thus, less likely to engage in corruption than men and as (2) political outsiders, and, thus having fewer opportunities to engage in corruption than men, they are also viewed as less corruptible (Barnes and Beaulieu, 2019). Wängnerud (2012Wängnerud ( , 2015 argued that the career path of candidates before reaching political positions also makes a difference in how they are perceived in relation to corruption and how they act on it when they are in power. According to her research in Mexico, women candidates often come from civil society. Engaging in corrupt behavior would mean alienating their support base in civil society.
I argue that all these scholarly findings are highly relevant and applicable to the Slovak case study. The 2019 presidential election took place in extraordinary sociopolitical conditions, in which, contrary to the traditional narrative in feminist research, the female presidential candidate Čaputová could have benefited from her stereotypical portrayal and her newness to politics.
When the female candidate Radičová, an SDKÚ-DS nominee and a former Minister of Labor, announced her candidacy in 2007, Smer-SD had been in the government for a year and enjoyed high levels of public trust and support. Although the 2008 economic crisis slowed economic growth, most Slovaks (54%) were optimistic and believed that the national economy was in good shape. In 2008, trust in national institutions was substantially higher compared to the EU27 average. Forty-six percent of Slovaks had confidence in the national government, and 41% trusted the national parliament. On the contrary, the reputation of the opposition party SDKÚ-DS and its leader Dzurinda was at its lowest point. Radičová was disadvantaged in three ways: as an SDKÚ-DS nominee, a former Labor Minister serving the SDKÚ-DS-led government, and a woman.
In 2019, the situation was different. The murder of investigative journalist Kuciak and his fiancée Kušnírová a year earlier led to unprecedented public protests, as his reporting revealed large corruption schemes linked to Smer-SD. As a result, public trust in institutions plummeted, recording its historically lowest point in 2018 when only 21% of the population stated they trusted the government and parliament (European Commission, 2018b). In these sociopolitical conditions, having stereotypically 'feminine' traits such as honesty and compassion and an image of an incorruptible leader could have benefited women candidates. Barnes and Beaulieu (2019) also found a link between being perceived as incorruptible and being marginalized in politics. As the data from Table 1 shows, women in Slovakia are underrepresented in positions of power in politics, public administration, and media. However, there was a significant difference between Radičová in 2009 and Čaputová in 2019. While Radičová was an experienced politician linked to, at the time, unpopular SDKÚ-DS, Čaputová was a political novice coming to politics from a civil society organization. Although she was not an independent candidate, she was nominated by a new extra-parliamentary party (PS), free from major scandals. Her clean political past was a significant vantage against her opponentdiplomat Šefčovičwho, despite being perceived as highly skilled for the job, was discarded for his ties with the governing party Smer-SD.
I argue that in the specific sociopolitical circumstances of the 2019 election, the media emphasized stereotypically 'feminine' traits more than ten years previously when describing Radičová. I propose the following hypothesis along those lines: Hypothesis 3. Gender stereotypes about women presidential candidates were more present in selected news outlets during the 2019 election campaign than in the 2009 election campaign.
As Čaputová was the only newcomer to politics among other candidates, the role of her marginalization within politics cannot be quantitatively tested against other candidates. However, as this factor might have played a crucial role, I have explored how the media portrayed her novelty separately in the qualitative analysis.

Data and methods
To test these hypotheses, I adopted a mixed-methods approach combining quantitative and qualitative content analysis of commentaries of the main presidential candidates (n = 1492) in selected news outlets. Compared to other EU member states, traditional media 5 in Slovakia are relatively trusted and have broad audiences. 6 How the press 7 writes about public personas, events of public importance, or policies, influences how social reality and social roles are constructed and how they are represented and contested. I selected the news genre of commentary, as previous research suggests that women appear in commentaries more often than in other news genres (Humprecht and Esser, 2017). Due to its subjective viewpoint and interpretation of facts, it is intrinsically more likely to carry gender stereotypes than news, which values facts and objectivity. The news sections in the Slovak outlets are usually very brief, factual, and devoid of journalists' views. They are unlikely to carry gender stereotypes and, thus, are not an appropriate subject to study. Commentary as a genre is well-placed for the selected research subject and type of analysis.

Data
I selected media outlets based on several criteria: (1) audience size and popularity, (2) nationwide coverage, (3) high frequency of publication, and the ability to capture different types of (4) media outlets, (5) audiences, and (6) ideological orientation. The outlets included in the analysis are five dailies with the largest audience: Nový čas, Plus 1 deň, Pravda, SME, and Denník N (which is only present in the 2019 election analysis due to its launch in 2015). To account for outlet type, I selected both tabloids and high-quality press. While dailies Pravda, SME, Denník N represent high-quality press, that is, broadsheets, as they are issue-oriented and provide in-depth analysis of public interest events based on reliable sources, Nový čas and Plus 1 deň can be categorized as tabloids. To account for ideological bias, I added an ultra-conservative online portal Postoj.sk. The sociodemographic characteristics of the audiences of these selected outlets are diverse (See Table 2). Following Lacy et al.'s recommendations (2015), I retrieved commentaries mentioning studied presidential candidates directly from newspapers' online archives from the candidacy announcement date until two weeks post-election. For the 2009 election, this was from April 4, 2008, until April 18, 2009, and for the 2019 election, from May 29, 2018, until April 13, 2019.
The dataset consists of 1090 unique commentaries of varying length (from 23 to 3404 words, μ = 431 words), published concerning the main presidential candidates in 2009 (Radičová and Gašparovič) and 2019 (Čaputová and Šefčovič) in selected news outlets. As a single commentary could have mentioned both competing candidates, the unit of analysis is not any commentary but a commentary coded for a single candidate (n = 1492). Therefore, if both competing candidates were mentioned in a commentary, this commentary was counted and coded twice for each candidate separately. Each unit of analysis was paired with the following information: date of publication, name of outlet, type of outlet, journalist name, journalist gender, candidate, candidate gender, and election year . For breakdown of candidates' mentions by the outlet type, see Table 3. 5 According to the 2018 Eurobarometer, 51% of Slovaks expressed having confidence in the press, 61% in radio, and 58% in television. See https://europa.eu/eurobarometer/surveys/detail/2215. 6 According to the Reuters Institute Digital News Report 2019, the print versions of selected newspapers reached up to 26% of respondents, and the online versions reached up to 27% of respondents (weekly usage). The sample (n = 2045) was assembled using nationally representative age, gender, region, and education quotas. See https://reutersinstitute.politics.ox.ac.uk/ sites/default/files/2019-06/DNR_2019_FINAL_0.pdf. 7 It is important to note that all selected newspapers have their online versions and some of the commentaries assessed also appear as podcasts, which have a significant reach. For instance, according to the Reuters Institute Digital News Report 2019, Denník N's weekday podcast Newsfilter had almost 10,000 downloads per day by February 2019, more than double print circulation. See https://reutersinstitute.politics.ox.ac.uk/sites/default/files/2019-06/DNR_2019_FINAL_0.pdf. Note. The data about the sold copies are available at http://www.abcsr.sk/aktualne-vysledky/archiv-vysledkov/. The data about online users is from August 18, 2022 and is available at https://www.iabslovakia.sk/ iabmonitor/data-iabmonitor-online/. The sociodemographic data are from 2021 to 2022.

Quantitative analysis
The codebook was derived from previous research on gender stereotypes, especially on the agency-communality scale (Bakan, 1966;Bem, 1981), which is well-established and widely used in contemporary gender studies, psychological and organizational research (Everitt, Best and Gaudet, 2016;Hentschel et al., 2019). Each commentary mentioning the candidate was coded for communal and agentic traits (0 if absent, 1 if present in the commentary). Based on the scale, communal traits predominantly associated with women are: being sensitive, emotional, compassionate, affectionate, honest, patient, polite, unselfish, devoted to others, gentle, helpful, kind, empathetic, warm, cooperative, and supportive. On the opposite side of the scale, there are agentic traits, such as being decisive, courageous, ambitious, hardworking, independent, confident, active, competitive, standing up well under pressure, and calm in emergencies, but also aggressive, arrogant, stubborn, proud, selfish, demanding, and opportunistic. The detailed codebook is available in the online Supporting Information.
A random sample of mentions (n = 175), generated by a random number generator, was coded by a second independent coder hired for this task. The sub-sample size (12% of the sample) is standard for this type of study (see McGregor et al., 2017;Vu et al., 2018). The second coder was a native Slovak with a graduate degree in public policy, so she was familiar with the country's political context. She had undergone several rounds of coder training to safeguard the consistent application of a developed codebook. Krippendorff's alpha (α), a reliability coefficient developed to measure the agreement among coders, was 0.768 (95% confidence intervals 0.6519-0.8695) for the outcome variable communal traits, which indicates substantial agreement. For the outcome variable agentic traits, Krippendorff's alpha was 0.8644 (95% confidence intervals 0.7782-0.9384), which suggests near-perfect agreement.
Given high inter-coder reliability, I performed main analyses for each of the categorical outcomes of interest: (1) communal traits and (2) agentic traits. Categorical treatment variables -(1) candidate (Radičová, Gašparovič, Čaputová, Šefčovič) and (2) candidate gender (female, male) identified based on prior studies, and a priori hypotheses were investigated for association with the outcomes of interest. Control variables include election year (2009, 2019), outlet type (tabloid, quality press), and journalist gender (female, male). A detailed description of variables is available in the online Supporting Information. The potential effect of a journalist was accounted for by including a journalist as a random effect (intercept), as there were multiple observations from some journalists. The final models' fit was evaluated using the Hosmer-Lemeshow goodness-of-fit test and by examining the residuals. The statistical analysis was conducted in R, version 3.6.2. Statistical significance was set at a P-value of <0.05.

Qualitative analysis
The quantitative analysis can adequately test the proposed hypotheses and show if women candidates were stereotypically attributed communal traits and men candidates were assigned agentic traits and how stereotyping differed by outlet type and year of election. However, the results of Note. Percentages in columns Tabloids and Quality press are calculated from the total counts for each candidate. Percentages in the column Total for a candidate are calculated from the total number of mentions in the sample (n = 1492).
these analyses cannot satisfactorily show if stereotyping was equally harmful to all candidates and explain the role of other factors, such as corruption perception, candidates' partisan affiliation, or Čaputová's marginalization within politics. Therefore, I conducted a deductive thematic analysis (Braun et al., 2019;Braun and Clarke, 2020) using the same population of mentions to understand better how media stereotyping worked during both campaigns and how other factors could have influenced electoral outcomes. I selected this type of qualitative content analysis due to its suitability for studying the selected genrecommentary. One of the key features of thematic analysis is a strong link between theory and themes, which do not emerge from the data but are purposefully selected prior to the analysis (ibid). I identified the following themes based on the role congruity theory and strategic stereotype theory, literature on gender stereotypes and corruption, and marginalization within politics: (1) double standards and the gender double bind, (2) corruption and public discontent, (3) partisan politics, and (4) political experience.
In addition to coding communal and agentic traits, which were the basis for the quantitative analysis, I coded all candidates' mentions for these themes. To capture double standards and gender double bind, I analyzed when agentic traits of women candidates were described negatively. As corruption and public discontent represent factors that might encourage positive public perceptions of some communal traits, I coded their presence or absence in each mention. I also coded if the partisan affiliation of presidential candidates was mentioned, as partisan politics could have played an important role in both elections. Finally, I coded if their political experience or absence thereof was brought up and, if so, in positive or negative terms. See Table 4.
When I collected a subset of mentions for each of the themes, I carefully read and analyzed all mentions within these subsets and reflected on them in light of the presented theoretical frameworks. Eventually, I merged the theme of partisan politics with the theme of corruption and public discontent, as they were interconnected. The results for both quantitative and qualitative analyses are reported below.

News outlets nurture stereotypical portrayals of candidates
The overall results from the quantitative analysis show that women candidates are more often portrayed as communal than their male opponents, who are, on the contrary, more associated with agentic traits. Čaputová, the winner of the 2019 election, was associated with communal traits most of all the candidates. 8 Outlet type had no significant association with the presence or absence of communal or agentic traits. Table 5 presents the results of the final multivariable logistic regression models for the outcome communal traits. Model 1 has two independent variables significantly associated with the outcome. Candidate gender was associated with mentions of communal traits: women candidates were seven times more likely to be associated with mentions of communal traits compared to men candidates. Second, the election year was also statistically significant. Compared to the election year 2009, in 2019, communal traits were more than two times more likely to be mentioned in the commentaries. Variables outlet type and journalist gender were not statistically significant. Model 2 explored the relationship between individual candidates and the mentions of their communal traits. The candidate was significantly associated with mentions of communal traits: compared to Gašparovič (2009), Radičová (2009) was seven times more likely to be associated with mentions of communal traits in the commentaries. Šefčovič (2019) was two times more likely, and Čaputová (2019) was sixteen times more likely to be associated with mentions of her communal traits in commentaries published in the selected newspapers during the election campaign. Variables outlet type and journalist gender were not statistically significant. See Figures  1 and 2 for predicted probabilities of attributed communal traits for both models. Table 6 presents the results of the final multivariable models for the outcome agentic traits. Model 3 has only one independent variable significantly associated with the outcome. Candidate gender was associated with mentions of agentic traits: women candidates were found to have a lower likelihood of being associated with mentions of agentic traits compared to men candidates, with an odds ratio of 0.23.
Model 4 explored the relationship between individual candidates and the mentions of their agentic traits. The candidate was associated with mentions of agentic traits: compared to Gašparovič (2009), both Radičová (2009) and Čaputová (2019) were at reduced odds of being associated with mentions of agentic traits. On the contrary, Šefčovič (2019) was one time more likely than Gašparovič to be associated with the mentions of agentic traits in commentaries published in the selected newspapers. Other variables were not statistically significant. Note. Other covers mixed-gender author teams (n = 18) and unknown authors (n = 599).  These results are consistent with hypothesis 1a and hypothesis 1b, that is, women candidates in the 2009 and 2019 Slovak presidential elections were portrayed in selected news outlets as more communal than their male opponents, who were attributed with agentic traits. Overall, the media nurtured gender stereotypes about all the main candidates. I did not find support for hypotheses 2a and 2b, that is, tabloids were not more likely to portray presidential candidates in a more stereotypical way than the quality press. For none of the models, was the outlet type statistically significant. I did not find a statistically significant association between journalist gender and the stereotypical portrayal of female and male candidates. A potential explanation for null results here is that conservative attitudes toward women's social roles are pervasive and stereotypes are equally held by both men and women. See Figures 3 and 4 for predicted probabilities of attributed agentic traits for both models.
The results also provide support for hypothesis 3. The association of communal traits with Čaputová in 2019 was statistically significant and was substantially larger for her than any other candidate in both elections.
Women candidates: penalized for agency, rarely rewarded for communality The results of the quantitative analysis show that compared to men candidates, women candidates in both election years were significantly more likely to be associated with communal traits. However, analyzing the content qualitatively, suggests that Radičová and Čaputová were not appreciated for communality equally. The thematic analysis of commentaries under the theme of double standards and the gender double bind provides evidence towards this end. Note. Other covers mixed-gender author teams (n = 18) and unknown authors (n = 599).
Several commentaries concluded that Radičová is honest, polite, and authentic. Her political and professional past was not marked by scandals and controversies, contrary to her opponent Gašparovič who was described as opportunistic, mostly due to his Communist Party membership in authoritarian Czechoslovakia, his role as the Speaker of the Slovak National Parliament under the controversial Mečiar-led government in the mid-1990s and his close ties to then Prime Minister Fico. However, the presence of communal traits that are normally viewed as positive was not perceived as such in the case of Radičová. While some commentators saw her politeness and unwillingness to become hostile toward her opponent as her strength, others perceived it as a lack of attitude and an effort to please everyone [March 18, 2009, Postoj.sk]. They argued that it is not possible to 'get along with everyone' [March 30, 2009, SME] and suggested that to win the election, Radičová had to be more 'confrontational ' [March 22, 2009, SME], 'more decisive and tougher at times' [March 30, 2009, SME], and they point to the controversial past of her opponent. The commentators did not receive the conciliatory tone of her campaign well and occasionally questioned her communality as incompatible with the leadership position.
At the same time, being described in agentic terms was also at her disadvantage, which confirms the role congruity theory. In some commentaries, Radičová was referred to as a 'hyper-ambitious lady' [March 22, 2009, SME] or 'the best-prepared candidate, but : : : not a good candidate' [March 22, 2009, SME], which suggests that she was assessed against stricter standards than her male opponent. Even when her competencies for the presidential position were evaluated as high, they did not seem to suffice [March 30, 2009, SME]. It became most visible once she lost against Gašparovič in the second round and was debated as a champion for the SDKÚ-DS leadership. At that point, those who supported her as a presidential candidate questioned her capabilities to become a party leader [April 5, 2009, Nový čas;April 8, 2009, Nový čas]. Her party colleague was quoted: ' : : : I can't imagine her at all in the position of the SDKÚ-DS chairperson. Being popular, being able to inspire people, and being a party leader are two different things'. [April 5, 2009, Pravda] While men candidates were appreciated for having prior political experience (Gašparovič in 2009) and their professionalism (Šefčovič in 2019), Radičová was penalized for her expertise. She was criticized for using expert language too much while the same commentators were complaining that the campaign lacked focus on important policy issues [January 30, 2009, Pravda].
In turbulent times, communal traits are appreciated, but partisan politics still matters The thematic analysis of commentaries under the theme of corruption and public discontent, and partisan politics explains why communal stereotypes could have benefited Čaputová in 2019.
Commentators described Čaputová in similar terms as they did Radičová ten years previously The thematic analysis of commentaries reveals that specific sociopolitical conditions, preceding the 2019 election, created a demand for communal traits in leadership. The murder of an investigative journalist, the corruption scandals he reported on, which were linked to political elites in governing parties, were reflected in the commentaries. Journalists labeled the governing SMER-SD party as 'a criminal organization ' [January 18, 2019, Pravda], which embodies all the malaises of the country, such as 'corruption, arrogance, the disintegration of the rule of law, mafia's influence in public institutions : : : .' [March 26, 2019, Denník N]. Although Šefčovič's career was free from scandals, his ties with the governing party, Smer-SD, discarded him as a candidate. He was portrayed as an 'extended arm' of the party and its politics [March 18 and 31, 2019, Pravda;March 27, 2019 Being marginalized within politics as an advantage As mentioned above, partisan politics played an essential role in the coverage of both elections. However, one aspect of it deserves special attentionnovelty and a lack of political experience. In the past elections, Slovak voters have shifted from new to even newer parties, strengthening tendencies towards party fragmentation and high electoral volatility (Haughton et al., 2021).
The thematic analysis of commentaries shows that Čaputová benefited from her marginal position in politics. As a newcomer to politics, she did not have to carry the weight of corruption scandals, which was crucial at the time. Journalists strongly emphasized her novelty. They described her as

Discussion and conclusions
Scholars criticize media trivialization and personalization of women politicians, arguing it diverts audiences from substantive news coverage that could portray women as serious political candidates. Yet few studies engage with the substantive aspects of news coverage to determine if gendering is also present in media depictions of leadership abilities, qualifications, or policy ideas. This study, exploring the gendered nature of news media coverage of Slovak presidential candidates in 2009 and 2019, advances our limited knowledge of leadership traits.
Findings from the quantitative analysis show that candidate gender was significantly associated with mentions of communal traits. Compared to men candidates, women candidates were several times more likely to be associated with communal traits in both election years. In contrast, the odds of men candidates being associated with the mentions of their agentic traits were higher than the odds of women candidates in both election years. These results add to empirical studies of gendered media coverage and are consistent with previous research concluding that media nourish gender-stereotypical portrayal.
In addition, stereotypes were present in both tabloids and quality press. The outlet type was not statistically significant, which is surprising, as available research suggests that the type of outlet affects gendered coverage, and tabloids are more prone to spread gender stereotypes (Dan and Iorgoveanu, 2013;Humprecht and Esser, 2017). Journalist gender and stereotypical portrayal of female and male candidates were also not significantly associated.
As expected, communal traits were mentioned in connection with Čaputová significantly more than any other candidates in both elections. The qualitative analysis suggests why, contrary to the role congruity theory, Čaputová's communal traits were appreciated, but for Radičová, they represented an obstacle. Commentaries pointed to a specific sociopolitical situation in Slovakia during the 2019 election campaign, characterized by low levels of public trust in institutions and pervasive corruption perception, which created an appreciation for communal traits in leadership. Honesty, politeness, and trustworthiness are often seen as necessary for tackling the roots of the above problems. Moreover, being a newcomer, marginalized within politics, and not having political networks, was also seen as an advantage in the 2019 election.
So, what are the implications, and where do they apply? As previously demonstrated, genderstereotypical portrayal can negatively affect women candidates' self-perceptions (Fox and Lawless, 2011), political ambition, and participation (Fox and Lawless, 2011;Haraldsson and Wängnerud, 2019;Dahl and Nyrup, 2021), as well as their electoral chances. However, this study shows that, in politically turbulent times, communal traits stereotypically attributed to women become appreciated. It makes a theoretical contribution to the international literature by demonstrating the importance of political context on media (re)assessments of desirable leadership abilities. The findings are largely applicable to democracies and may be instructive for successful campaign strategies but should be used with caution, as conforming strategically to the media's stereotypical portrayal and establishing campaigns on communal traits does not come without risk. First, by doing so, women candidates risk perpetuating stereotypes about communality further. Second, when they secure leadership positions in a dire political and social climate, they might find themselves on a glass cliff. If they cease to be perceived as successful in their leadership roles, they may be judged against stricter standards that might discourage other women from competing for leadership roles.
This study does not come without limitations. First, its sample is limited to four presidential candidates at two different points in time only. However, this is a broader structural limitation that comes with studying female political candidates in countries with poor women's representation. Gradually increasing the share of women in politics over time will hopefully allow for studying larger samples quantitatively. Second, while I measure associations of candidate gender and individual candidates with communal and agentic traits, I cannot disentangle these traits as portrayed in the news outlets from candidates' inherent personalities and abilities. For example, was Čaputová portrayed as communal only as a result of journalists being prone to stereotype? Is she inherently more communal than other candidates, or at least she depicted herself as such in the campaign? Future research would benefit from studies linking news outlets' coverage to candidates' campaigns and testing the scale of journalists relying on politicians' campaigns and press releases. Last, the research considers only news outlets' coverage. Studying content aired on television would substantially enrich research on gender stereotypes in media portrayal.