Who is Afraid of More Women in Politics, and Why? An Analysis of Public Opinion in 28 European Countries

ABSTRACT We study how individual and country-level variables interact in affecting political gender attitudes in Europe. Based on data from the 2017 Eurobarometer survey, we show that there are high levels of support for more women in politics and legal measures to achieve gender parity across the EU. In fact, more people, and women in particular, put higher trust in female compared to male political representatives than the other way round. We find that – at the individual level – gender, age and education have significant effects on political gender attitudes. Contrary to theory, however, the effect of gender is not mediated by beliefs about the proper role of women in politics and society. We also do not find support for the contextual effects of masculine culture and the religiosity of society, but we do uncover significant gaps in political gender attitudes between post-communist and other countries, especially for men.


Introduction
Despite considerable progress in increasing women's participation in politics over the past decades, women still remain under-represented in political life, especially when it comes to positions of political leadership and power.Even in Europe, where progress has been strong, fewer women than men become politicians and are elected as party leaders and political representatives, appointed as ministers and chosen to lead their countries as prime-ministers and presidents (Goddard 2021).Moreover, there is considerable variation in improving women's participation in politics between the European states, with counties from the Northern, Southern and Eastern regions of the continent having rather different trajectories.
The reasons for women's under-representation in politics are diverse and include institutional, cultural and historical factors.Public attitudes, however, play an important role as well: directly and indirectly by sustaining institutional practices, enacting cultural shared meanings and perpetuating the effects of historical legacies. 1Therefore, understanding the nature of public attitudes toward the participation of women in politics -what we call political gender attitudes -is important for understanding the reasons for the gender gap in politics and for devising ways of reducing it.This task, however, is challenging because of the significant variation in political gender attitudes across individuals, across countries and cultural contexts, and across different dimensions of these attitudes.Most existing research has focused on identifying individual-level determinants of support for more women in politics (i.a., Alexander and Jalalzai 2016;Beauregard 2016;Bridgewater and Nagel 2020).Based on general theories of gender-related prejudice (Bauer 2020;Dassonneville, Quinlan, and McAllister 2021;de Geus et al. 2021;Piazza and Diaz 2020), role congruity (Bauer 2015(Bauer , 2020;;Eagly, Wood, and Diekman 2000) and gender affinity (Alexander 2012;Bridgewater and Nagel 2020;Dolan 2008), we know that females, younger and more educated people are more likely to support wider participation of women in politics.However, how country context moderates these effects and how exactly gender stereotypes mediate them has remained insufficiently studied.Moreover, the strength of these effects can differ depending on the political gender attitude of interest, for example trust in female politicians or support for legal measures to increase women's presence in politics.
In this article, we study how individual and country-level variables interact in affecting three dimensions of political gender attitudes in Europe.Based on data from the 2017 Eurobarometer survey (European Commission 2019), we show that there are high levels of trust in female political representatives, support for more women in politics and for legal measures to achieve gender parity across the EU.In fact, more people, and women in particular, put higher trust in female compared to male political representatives than the other way round.In line with existing knowledge, we find thatat the individual level -gender, age and education have significant effects of political gender attitudes (Alexander and Jalalzai 2016;Bauer 2020;Coffé and Reiser 2023;de Geus et al. 2021;Scarborough, Sin, and Risman 2019).Contrary to the theoretical expectations, however, we do not find evidence that the effect of gender is mediated by beliefs about the proper role of women in politics and society (Batista Pereira and Porto 2020;Eagly, Wood, and Diekman 2000;Garcia-Retamero and López-Zafra 2006).We also do not find univocal support for the contextual effects of masculine culture and the religiosity of society, but we do uncover significant gaps in political gender attitudes between post-communist and other countries, especially for men.This gap is very significant in size and declines only marginally with the age of the respondent.Our study identifies men in post-communist countries as the group least likely to trust female politicians and support gender parity in politics in Europe.
Our main contribution is to show that political gender attitudes are at least partly independent from stereotypes about gender roles.Apparently, strong support for more participation of women in politics can co-exist with stereotypical beliefs that women are less interested in positions of power or that their proper place is at home.Even support for gender quotas and other legal measures for gender parity in politics can coexist with gender stereotypes (cf.Batista Pereira and Porto 2020).The finding about women in more masculine cultures being more likely to have progressive gender attitudes is also novel and important, as it hints at the potential for polarization along gender lines of these attitudes.In methodological terms, our study proposes a new measure of net trust in female politicians that puts political gender attitudes in sharper focus.

Gender prejudice and the gender affinity effect
Research on female political leadership and representation has significantly expanded in the last decades, in parallel with the continuous increase in the election and appointment of women in political positions.Although the existing literature covers a wide variety of research approaches and theories, this review focuses on two major topics, which emerge repeatedly: gender-related prejudice against female leaders (Bauer 2020;Dassonneville, Quinlan, and McAllister 2021;de Geus et al. 2021;Eagly and Karau 2002;Eagly, Wood, and Diekman 2000;Piazza and Diaz 2020) and the gender affinity effect (Alexander 2012;Alexander and Jalalzai 2016;Beauregard 2016;Bridgewater and Nagel 2020;Dolan 2008;Liu and Banaszak 2017).

Gender prejudice toward female leaders
A first body of theories of public opinion toward female political leadership is largely focused on gender-related prejudice.These are rooted in social role theory, which is a theoretical framework that seeks to explain the differences and similarities between men and women in social behavior (Eagly, Wood, and Diekman 2000).According to social role theory, each gender is associated with specific expectations or norms about what a social group can and should do, which in their turn shape the gender role.As a consequence, men and women accommodate to gender roles by pursuing rolerelated skills.
This process gives rise to a distinction between communal and agentic characteristics.Women are associated with communal behavior focused on the wellbeing of other people, stemming from the domestic female role.Women are thus seen as having strong interpersonal and communication skills and as being more empathetic, compassionate, honest and trustworthy (Eagly and Karau 2002;Piazza and Diaz 2020).Conversely, men's adaptation to the employment role favors a pattern of association with agentic behaviors, including being more assertive and independent.What emerges from this psychological process of association is the creation of gender stereotypes related to gender-typical occupational roles (Eagly and Karau 2002;Eagly, Wood, and Diekman 2000).
Moreover, "sex segregation of the labor force" sets in motion a mechanism through which perceivers associate a set of typical skills to each gender from observing the type of occupation that is most commonly undertaken by either men or women (Eagly, Wood, and Diekman 2000, 467).Thus, the fact that more men are employed in jobs that involve agentic behavior leads to the incorporation of the stereotype into male gender roles.This in turn creates expectations, which act as normative pressure and foster behavior coherent with gender-typical work roles (Eagly, Wood, and Diekman 2000).
Based on these assumptions, the role congruity theory of prejudice toward female leaders (Eagly and Karau 2002) explains why people are inclined to view women as less qualified than men for leadership positions.It argues that leadership roles are commonly perceived in masculine agentic terms.Thus, since women are associated with communal characteristics, people tend to view them as less suitable than men to become leaders.In other words, there is an incongruity between the social expectations of the leadership role and the feminine gender role (Koenig et al. 2011).Roles related to political representation are very much related to leadership positions, as they require that the politicians actively steer and stand for the interests of the represented.
In addition to the role-typicality standards, individuals are influenced by gender stereotypes when they evaluate if a candidate can fulfill a leadership role.Gender stereotypes influence the perception of voters who evaluate political representatives based on the expected competencies and skills inherent to their gender.According to the feminine stereotype, women are perceived as less qualified than men to perform roles requiring agentic traits.This influences voters who refuse to view women candidates as having the required skills and policy expertise to become political leaders due to a crucial mismatch between the stereotypical female characteristics and the typical skills attributed to a politician.Hence, such role incongruity inevitably creates a disadvantage for female political candidates (Bauer 2015(Bauer , 2020;;Bridgewater and Nagel 2020;Eagly and Karau 2002;Lawless 2004;Simon and Hoyt 2008).Note, however, that the stereotypes about male agency do not have a strong effect on how voters judge the governing performance of men and women, with the latter actually receiving less blame and as much credit for the performance (de Geus et al. 2021).
Based on the role congruity theory of prejudice toward female leaders, Garcia-Retamero and López-Zafra (2006) found that female candidates are generally considered less qualified than men in masculine and gender-neutral working environments.Yet, females are considered to be as effective as male candidates in feminine environments which are more compatible with the feminine gender role.Recent and innovative studies found that voters judge female party leaders in a more positive way than male ones (Bridgewater and Nagel 2020;Cormack and Karl 2022;Dassonneville, Quinlan, and McAllister 2021).Moreover, at the individual-level, respondents from both sexes tended to view female leaders more favorably than male leaders.This implies that, given the gendered nature of the political environment, voters acknowledge the differences based on gender of political candidates, but there is no evidence regarding voters' preferences for male candidates (Bridgewater and Nagel 2020).
The media plays an important role in promoting or hindering certain gender stereotypical beliefs.Based on a meta-analysis of 90 studies, Van der Pas et al. (2014) found a gender bias in the way women politicians were covered in the media, with significantly more attention to appearance and personal life and more stereotypical issues and trait coverage.The amount of news coverage itself that women politicians received was significantly lower only in proportional electoral systems, but not in majoritarian ones.Jungblut and Haim (2023) also report differences in the way female and male politicians are visually depicted that are related to emotional gender stereotypes.
It should be acknowledged that more recently scholars have discovered that, while gender stereotypes can give rise to a female disadvantage, the effect does not occur in all contexts.Expectations about leadership and gender are evolving (Karl and Cormack 2023), and female politicians might constitute a particular subtype, with characteristics that are not clearly defined, but which puts female politicians at a disadvantage (Schneider and Bos 2014).

Gender affinity effect
The most common individual-level factor used to investigate public opinion toward female leaders is the gender of respondents.Several studies which include this variable found that female respondents evaluate female leaders more positively than their male counterparts (Alexander 2012;Alexander and Jalalzai 2016;Beauregard 2016;Bridgewater and Nagel 2020).In other words, female respondents have a preference for female candidates (Dolan 2008).This phenomenon is recurrent and it is referred to as the "gender affinity effect."The literature highlights three main reasons that can explain this effect.First, a feeling of group solidarity may lead women to support female candidates because of a feeling of demographic identification with a candidate of the same gender.Second, since women are still underrepresented in the political arena, female voters may seek descriptive representations by supporting female candidates.Third, recalling social role theory, voters tend to see female candidates as more suited to deal with issues such as child-care, abortion and welfare.Thus, since women voters attribute greater importance to these issues, they will vote for female candidates (Dolan 2008;Herrnson, Celeste Lay, and Kai Stokes 2003).
Other studies focused on the effect of women's leadership positions on political attitudes claim that the presence of female political leaders has a stronger positive impact on women rather than on male respondents (Alexander 2012;Alexander and Jalalzai 2016;Beauregard 2016;Liu and Banaszak 2017).These scholars argue that women's descriptive representation in legislatures (Alexander 2012;Beauregard 2016;O'Brien 2019), as well in executives (Alexander and Jalalzai 2016) has a symbolic effect especially on women as it increases women's trust in women's ability to govern.

Individual and country-level factors affecting public opinion
Many studies explore the influence of individual-level variables because individual variation in social perceivers can influence the perception of female leaders (Alexander and Jalalzai 2016;Bauer 2020;Coffé and Reiser 2023;de Geus et al. 2021).But the literatures on the influence of gender stereotypes on the perception of political leaders sends an ambiguous message.Existing female leadership theories emphasize the role of gender stereotypes which give rise to a female disadvantage in politics (Bauer 2015(Bauer , 2020;;Eagly and Karau 2002;Lawless 2004;Simon and Hoyt 2008), while more recent research indicates that there is no evidence regarding the existence of a strong voters' preferences for male candidates (Bridgewater and Nagel 2020;Cormack and Karl 2022;Dassonneville, Quinlan, and McAllister 2021).
Still, contextual factors, such as culture, are likely to moderate individual public opinion toward female political leaders.Some existing studies compare cross-country attitudes toward women's political leadership and shed light on the strong impact of political culture on the public opinion toward female leaders (Alexander 2012;Banaszak 2006;Norris and Inglehart 2000).Yet, the majority of the above-mentioned works direct little to no attention to the interactions between individual-level and country-level factors in explaining the variation in trust toward male and female political representatives.While they apply a macro-level approach to study public opinion on female political roles and leadership, how the macro level moderates the micro-level forces of public opinion remains relatively understudied.
Regarding the scope of the empirical analysis, most studies focus on the United States or on a single country, while cross-national analysis is relatively lacking.Finally, as the majority of studies focus exclusively on female leaders, there is a gap in the comparison of public opinion for both male and female political representatives, fostering calls in the literature for more comprehensive analysis.
Our contribution to these literatures is to offer a comparison of public opinion toward female political representatives that considers opinions on male political representatives by using a new variable called "net female trust", which measures the respondents' difference in the level of trust in female and male political representatives.In addition, to address the identified literature gaps, our research investigates the factors which account for the variation in public opinion toward female political leadership by conducting a cross-national analysis of 28 European countries.In particular, we focus on how gender stereotypes mediate the influence of individual-level predictors and how national context -feminine/masculine culture, religion and communist legacy -moderates their effects.

Prejudice and country-level context: mediating and moderating effects
In this section of the article, we derive hypotheses based on the theories and literature reviewed, which we later examine empirically.We formulate the theoretical expectations with regard to the concept of "political gender attitudes," which we employ as an umbrella concept that captures attitudes and beliefs about the role of women in politics.As we explain below, in the empirical part of the analysis we employ three different variables that capture different aspects of political gender attitudes, namely (a) net trust in female political representatives, (b) beliefs that more women should be in politics, and (c) support for legal measures for gender parity in politics.
At the individual level, we expect that the gender, age, education level, profession, and left-right positions of the respondent will affect political gender attitudes. 2These expectations are based on existing literature and are -at least when it comes to gender (Alexander and Jalalzai 2016;Eagly and Karau 2002;Simon and Hoyt 2008;Dolan 2008; but also see; Bridgewater and Nagel 2020;Cormack and Karl 2022;Dassonneville, Quinlan, and McAllister 2021;Garcia-Retamero and López-Zafra 2006), age (Alexander 2012;Alexander and Jalalzai 2016;Garcia-Retamero and López-Zafra 2006;Inglehart and Norris 2003;Scarborough, Sin, and Risman 2019) and education (Alexander 2012;Coffé and Reiser 2023) -very well documented.With regard to professions, the study conducted by Eagly and Karau (2002) shows that male managers perceive their female counterparts as successful as male managers are.
With regard to political positions, we expect that people who consider themselves to the right on the political left-right spectrum will be less likely to support political gender equality in its various aspects.This is because right-wing political parties tend to espouse more traditionalist and authoritarian viewpoints, which do not allow for much space for women in politics (Bauer 2018).Such parties aim to decrease the salience of the issue of gender equality in politics and to dismiss calls for wider participation of women in political life, also in positions of leadership, while it is left parties are traditionally champions of such issues and positions. 3 Now we turn to the mediation hypotheses about the factors that translate the influence of the demographic factors.According to theory, at least some of the individual-level effects discussed above should be mediated by the beliefs people have about the proper role of women in society and politics and their stereotypes about which tasks and professions women are interested in and suitable for (de Geus et al. 2021;Eagly and Karau 2002;Eagly, Wood, and Diekman 2000).If indeed the effects of the individual-level variables discussed above are exercised through changes in such beliefs, we would expect that once we account for these beliefs in the empirical models, the individual-level effects should disappear or at least get significantly reduced in size.In other words, we would expect that net of their beliefs about the proper role of women in society and politics, men and women, old and young, highly educated and less educated, managers and other professions would not exhibit significant differences in their gender political attitudes anymore.
There are two related but conceptually distinct relevant gender prejudices that could play such a mediating role.The first one is the belief that the most important role of women is to take care of her home and family.This captures general attitudes toward the proper role of women in society and personal life.The second one is more specific to the domain of politics and relates to the belief that women have less interest in political affairs.Such beliefs can be projected to rationalize why people want to limit the presence of women in politics.

Hypothesis 1a:
Beliefs about the general role of women in society mediate the effects of the individual-level factors on gender political attitudes.

Hypothesis 1b:
Beliefs about the specific interest of women in politics mediate the effects of the individual-level factors on gender political attitudes.
In other words, we expect that once we include measures of beliefs about the role of women in society and their interest in politics in the statistical models, the effects of the individual-level variables (age, gender, education, etc.) should be substantially reduced or altogether disappear.It is important that we identify potential mediators of gender political attitudes, because such mediators can shed light on why people hold the attitudes they do and to suggest targets for intervention (e.g., increasing support for more women in politics by reducing stereotypes about their interest in politics).
We also study how two country-level variables -masculine culture and communist legacy moderate or condition the effects of individual-level predictors.We derive the variable masculine culture from Hofstede's cultural dimensions theory (Hofstede 2011).According to Geert Hofstede, Gert Jan Hofstede and Michael Minkov (2010), culture is about "the collective programming of the mind distinguishing the members of one group or category of people from others."The dimension "masculinity/femininity" is a societal rather than individual characteristic, which is related to the division of roles and values between men and women that are considered appropriate.In countries with feminine cultures, there is a "minimum social role differentiation between the genders" as they both have the same values (Hofstede 2011, 12).On the contrary, in masculine countries, there is a strong social role differentiation as men are perceived and expected to be assertive and ambitious while women should be modest and caring.This variable is not commonly used in the empirical research on female leadership, 4 however, we include it in the study because it allows to detect whether countries with stronger descriptive and injunctive norms on gender roles will have a lower level of net female trust (Eagly and Karau 2002;Eagly, Wood, and Diekman 2000).In the Supplementary Material we report tests using different measures related to masculinity culture.
We expect that the effect of gender differs in countries with a masculine culture.The gender affinity effect (Dolan 2008) might have different weight across cultures and countries, depending on the prevailing gender norms in the country.Our hypothesis is that in masculine cultures women try to compensate the strong social norms about the gender role (Hofstede 2011) by endorsing stronger and more positive attitudes toward women in politics.Thus, we expect countries with a masculine culture to exhibit a stronger gender affinity effect (Dolan 2008), and we therefore propose the following crosslevel interaction hypotheses: Hypothesis 2: In countries with more masculine cultures, the negative effects of gender and age on gender political attitudes are stronger.
The second country-level variable we focus on is communist legacy.The literature shows that Eastern European communist countries were leading examples of gender equality as they encouraged women to enter the workforce to rapidly achieve industrialization.Also, these countries showed high numbers of women in parliaments (Alexander 2012;Banaszak 2006).Yet, after the fall of communism, they experienced the rise of an antifeminist political culture with the end of quotas for women and a dramatic fall in the percentage of female members of parliament (Alexander 2012).Furthermore, post-communist countries display more traditional attitudes toward female leadership (Inglehart and Norris 2003;Norris and Inglehart 2000).
According to Norris and Inglehart (2000), in post-industrial countries younger generations have far more egalitarian attitudes than older generations.However, this difference is minimal or absent in post-communist and developing countries (Inglehart and Norris 2003;Norris and Inglehart 2000).These considerations lead us to expect that communist legacy moderates the effects of age and gender.We assume that the general display of more traditional attitudes toward female leadership in postcommunist countries (Inglehart and Norris 2003;Norris and Inglehart 2000) reinforces the social role theory's assumption that men will trust female representatives less than women do because of a more masculine and agentic perception of leadership (Alexander and Jalalzai 2016;Eagly and Karau 2002;Simon and Hoyt 2008).
Hypothesis 3: In post-communist countries, the negative effects of gender and age on gender political attitudes are stronger.
Finally, we include the country-level variable religiosity as a control.Whether countries have a religious or secular culture is possibly relevant to the level of net trust in women as political leaders.A broad literature suggests that "religion is a primary agent of gender role socialization" (Alexander and Jalalzai 2016, 8) and exerts a major influence over people's attitudes and perceptions of gender roles (Alexander 2012;Inglehart and Norris 2003).In particular, evidence shows that many religious faiths associate women with traditional and subordinate roles, thus having a negative implication for trust in female leaders (Alexander 2012;Alexander and Jalalzai 2016).On the other hand, secularization is associated with a rising tendency of gender equality in politics (Inglehart and Norris 2003;Norris and Inglehart 2000).
It is important to study the moderating effects of country-level variables, because this provides information about how demographic and other individual-level factors are related to gender political attitudes in different contexts.Masculine culture, communist past or high levels of religiosity might affect not only absolute levels of trust in female politicians, but they might also lead to different causal structures shaping these attitudes.Knowing how individual effects are conditioned by culture, legacies and institutions, not only provides more nuanced understanding, but can support tailored policy interventions as well, e.g., targeting young males in post-communist contexts, but old women in masculine cultures.

Research design, data and method of analysis
To explore the individual and country-level factors which account for a variation in public opinion toward female political leadership across Europe, we rely upon the Eurobarometer 87.4 survey of public opinion (European Commission 2019).The survey was conducted in June 2017 in the 28 Member States of the European Union (as of 2017).Data were collected from 28,093 respondents.We supplement this dataset with data on masculine culture, religion and communist legacy retrieved from different sources: respectively, the Hofstede Insights indicators (https://www.hofstede-insights.com); the Special Eurobarometer 341 (European Commission 2010); and a comparative dataset provided by Armingeon and Careja (2004).
We focus on Europe for a combination of research design-related and practical reasons.From a research design perspective, we want the keep the sample of countries relatively homogenous.If we show that national context moderates the effects of individual-level variables in this set of countries, it is likely that these effects will be even greater on a global scale, where there is more variation in institutional and cultural context.Furthermore, given the relatively high level of support for female political rights in Europe, this continent provides a hard test for finding significant differences at the individual level.
From a practical point of view, the focus on European countries allows us to use a single source of data for public opinion with consistent operationalizations of the relevant attitudes.

Outcome variables
We use three outcome variables to analyze political gender attitudes: net trust in female politicians, belief in the need for more women in politics and support for legal measures for gender parity in politics (European Commission 2019).These are retrieved from the Eurobarometer 87.4 survey questions as indicated in Table 1.Net female trust measures the difference between trust in female and male leaders on a scale from −3 to + 3, where positive values indicate more trust in female leaders and 0 means equal trust in male and female politicians.The variable more women in politics measures on a scale from 1 to 3 whether people think there should be more or fewer women in politics.Finally, legal measure for gender parity indicates support (or lack of it) for legal measures to ensure parity between men and women in politics (European Commission 2019).

Individual-level variables
We select five individual-level variables from the demographic and social characteristics included in the Eurobarometer 87.4 survey: gender, age, education, profession and left-right self-placement.First, gender of respondents is a binary variable, and it is coded 1 for male and 0 for female (European  10), which we rescale between −4 and 5.

Mediating variables
To analyze possible mediation effects on gender political attitudes, we use two variables: gender stereotypes (general) and gender stereotypes (politics).They are retrieved from the survey questions you agree with the statement: the most important role of a woman is to take care of her home and family?" and "Do you agree with the statement: Women are less interested than men in positions of responsibility in politics?"Responses are coded so that higher values indicate a more favorable gender attitude.

Country-level variables
The Eurobarometer dataset offers a cross-sectional sample of the 28 Member States of the European Union, which allows for a country-level analysis (European Commission 2017).Our empirical model includes three country-level variables, namely masculine culture, religion and communist legacy.To measure masculine culture, defined as a societal characteristic related to the division of roles and values between sexes (Hofstede 2011), we rely upon the indicators of the website Hofstede Insights, which provides a percentage of the estimated rate of masculinity -defined as "a preference in society for achievement, heroism, assertiveness, and material rewards for success," for each country in the world.
The values range from 1 to 100.If the value is smaller than 50, it means the country has a prevalent feminine culture; while if it is larger than 50, then the country has a dominant masculine culture.
The variable religion is defined as the percentage of citizens per country who believe in God (European Commission 2010; Sägesser et al. 2018).It is retrieved from the Special Eurobarometer 341 on Biotechnology, which presents national percentages of religious belief.Countries with a percentage above 50% are considered to have a religious culture; on the other hand, countries with a percentage that does not exceed 50% are considered to have a secular culture (European Commission 2010).Finally, the variable communist legacy measures whether countries are post-communist or not (Alexander 2012).It is interpreted as a binary variable where countries with a communist legacy are coded 1 and all others 0. This information is retrieved from a comparative dataset for 28 postcommunist countries in Europe (Armingeon and Careja 2004).

Empirical results
Before we proceed to testing the hypotheses identified above, we discuss some of the descriptive patterns we find in the data.Details about the average values of all variables per country are provided in the Supplementary Material, Table A5.

Descriptive analysis
Altogether, in the 28 European countries included in the survey, trust in females as political representatives is rather high.Looking at the bottom-left panel of Figure 1, which plots the (weighted) average of trust in female political representatives per country, we can see that in all countries people tend to trust that a female can represent their political interests.Unsurprisingly, trust is highest in the Nordic countries -Sweden and Denmark, in the Netherlands and Luxembourg, but also in some of the countries in Southern Europe -Spain and Portugal.Trust is lowest, on average, in Slovakia, Italy, Cyprus, and Romania.This measure of trust, however, is confounded by general trust in political elites, as manifested in the bottom-right panel of Figure 1, which shows the (weighted) country means for trust in male political representatives.The ranking of countries is quite similar.Therefore, to isolate trust in women in particular, we calculate the difference between trusting a female and a male political representative for each respondent.Taking the country-level average of this new variable results in a different picture, shown in the top panel of Figure 1.Now, the top three countries are Cyprus, France and Italy, while the Eastern European countries are at the other end of the scale.It is noteworthy that only in Latvia, Bulgaria and Romania net trust in female political representatives is below 0, meaning that the population tends to trust male political representatives slightly more than female ones.In all other countries, the pattern is reversed, with women perceived as (slightly) more trustworthy for being political representatives.
While most respondents (76%) have the same level of trust (or distrust) in male and female political representatives, 16% place more trust (or less distrust) in female politicians, while only 8% place more trust in male politicians.In other words, twice as many respondents have higher trust in a female to represent their political interests (rather than a male) than the other way round.
Disaggregating these results by the gender of the respondent, 26% of women trust a female political representative more than a male one, while only 7% of men do.A small percentage of women (4.5%) trust a man more than a woman as a political representative, while 12.5% of men do so.All in all, most people think that a male and a female can equally represent their political interests, but from the ones that do not, more people tend to put higher trust in women as political representatives, especially if they are women themselves.The tendency of men to put higher political trust in their own gender is much less pronounced: women are more than twice as likely to have a preference for their own gender as men.
Looking at the other two outcome variables of interest, overall in the combined European sample there is also clear support for the presence of more women in politics.The average value for the full sample is 0.53 (on a scale between −1 and 1), with support being highest in Malta and lowest in Romania (see Table A1).There is also widespread support for legal measures to achieve parity between men and women in politics (average of 3 on the 1-to-4 scale) (cf.Coffé and Reiser 2023 who find low support for positive action measures, such as legal quotas, in Germany).
At the same time, gender stereotypes are widespread as well.The average support across the full sample for the statement that the most important role of women is to take care of her home and family is 2.53 and for the statement that women are less interested than men in positions of responsibility in politics is 2.27 (both measured on 1-to-4 scales with "full support" at 4).The highest prevalence of these stereotypes is in Bulgaria and other countries in Eastern Europe, while the lowest prevalence is in Sweden and Denmark.
The descriptive analysis already hints that support for more participation of women in politics can coexist with negative gender stereotypes and that there is significant cross-country variation in these attitudes.These results could be affected by social desirability bias in survey responses, however.For example, Streb et al. (2008) showed using list experiments that standard public opinion survey might have overestimated support for a female president in the US (see also Benson, Merolla, and Geer 2011).Krupnikov, Piston, and Bauer (2016) also concluded that existing empirical research in the US exaggerated voter support for female political candidates.However, re-deploying the list experiment conducted by Streb et al. (2008) in 2016, Burden, Ono, and Yamada (2017) found that resistance to a female president in the US had all but disappeared among Democratic voters.This shows that the extent of social desirability bias is contingent on the social and political context in a country and can decline relatively fast.Since all of these studies concern the US, it is hard to say to what extent the desire to conform with societal norms influences the responses of Europeans.But even if present, we have no reasons to believe that this bias operates differently across European countries or across demographic groups.Hence, the relative rankings of the average responses from different countries should not be affected.
The next section of the analysis presents individual-level and multilevel multivariate statistical models that try to account for some of this variation and to disentangle the relationships between socio-economic variables, beliefs in stereotypes and political gender attitudes.

Multivariate statistical models
We first present the results from the individual-level analysis.Table 2 summarizes six models: for each of the three outcomes variables, we first run a model with the individual-level explanatory factors of interest, and then we add the hypothesized mediators.
As expected, the gender of the respondent has significant negative effects on all three attitudes we study.Being male is associated with a significantly lower likelihood of trusting a female politician more than a male one, of considering that more women in politics are needed, and of supporting legal measures for achieving gender parity in politics.The effects are substantial in size: being male leads to approximately half of a standard deviation change in "net female trust," one-third of a standard deviation change in "more women in politics" and one-fifth of a standard deviation change in "legal measures for parity." The effects of education are positive with regard to the first two outcome variables, but there is no effect on support for legal measures for gender parity in politics.The effects of age are, as expected, negative on all three variables, but they are quite small in size (to remind, in Table 2 the coefficients for age and education are given for a change of 10 years).
Being a manager increases the probability that one would trust a female political representative (more than a male one), but has no significant associations with the other two outcome variables.Conversely, the left-right position of the respondent is associated only with the idea that there should be more women in politics and that legal measures should be used to achieve that, but not with net trust.Unsurprisingly, gender stereotypes have large and significant effects on gender political attitudes.More relevant for our purposes, however, is what happens to the effects of the other individual-level variables, once they are included in the models.Contrary to our expectations, the effect of gender does not change much.In other words, even net of general and politics-specific gender stereotypes, men are much less likely to support all aspects of gender political attitudes.
This is not what we would expect based on theory, which posits that ideas about the role of women in society and politics play a crucial mediating role in the formation of gender political attitudes (cf.Eagly and Karau 2002).
In line with these theoretical expectations, the effects of age and education indeed disappear completely or get significantly reduced once gender prejudices are included in the models.The effect of being a manager on net trust in female politicians persists, while the effect of left-right position evaporates.However, when it comes to support for more women in politics and legal measures to achieve parity, gender stereotypes do not mediate complete the effect of left-right position.
In Table 3 we report the analyses of the country-level variables and the interactions with the individual-level predictors.Masculine culture is not associated significantly with lower support for gender political attitudes.The average level of religiosity of society even increases support for legal measures to achieve gender parity in politics. 5Post-communist legacy has the predicted negative effects.When it comes to the interactions, the effects of gender (being male) are much stronger in more masculine cultures and in post-communist countries when it comes to net trust in female politicians and preferences for more women in politics.However, looking at support for legal measures for parity, the negative effect of being male does not vary significantly with masculinity and post-communist legacy.
The interactions with age, however, do not conform to the hypotheses.Older people in masculine cultures are not less likely to have negative gender political preferences than older people in feminine cultures.If anything, when it comes to support for legal measures to achieve, the negative effect of age in masculine cultures is significantly smaller.Similarly, in post-communist countries the negative effects of age are significantly smaller, for two out of the three outcome variables.
Figure 2 illustrates the interaction effects.The left panel plots the predicted probability of net trust in females as political representatives as age varies from 20 to 80 years, based on the estimates from Model 1 in Table 3.As we can see from the figure, net trust declines with age for men (blue lines) as well as for women (red lines), in post-communist (dotted lines) and in other countries (solid lines).The gap between men and women is much greater in the post-communist countries, however, and it remains similar across the values of age.There is a relatively small difference in the gap between the The gap between the net trust of males in post-communist and other countries is huge and remains stable throughout age.In fact, the only group that has on average more trust in male than in female political representatives (negative values on the y-axis) is men in post-communist countries, for all ages.
The right panel of Figure 2 shows the predicted probabilities of support for legal measures to achieve gender parity in politics as a function of age, according to the estimates of Model 3 from Table 3.The patterns here are quite different.In post-communist countries, support for such legal measures increases slightly with age, both for men and women.In the other countries in Western and Southern Europe, however, support for such measures declines significantly with age, again both for men and women.
To what extent are these results dependent on the particular operationalization of masculinity culture that we use?To answer this question, we re-estimated the models from Table 3 with alternative measures: (a) the Gender Equality Index, developed by the European Institute for Gender Equality, (b) the percentage of ever-partnered women who ever suffered intimate partner physical and/or sexual violence (compiled from various sources and distributed from the OECD), and (c) the share of people that agree with the statement that when jobs are scarce, men have more right to a job than women (from the European Values Survey 2017).The first measure is a composite indicator that synthesizes gender-relevant information related to work, money, knowledge, time, power, health and violence).The second measure is a behavioral one but its collection is dependent on reporting practices, which themselves can be strongly influenced by national culture and institutions).The third measure captures one particular aspect of gender culture.
The Gender Equality Index turns out to be significantly negatively associated with net trust in female politicians.This is a puzzling result, which could be due to the fact that the index already incorporates information about the place of women in politics.The interactions with individual's gender and age are similar to the ones reported in Table 3 (see for details Table A1 in the Supplementary Material).
The share of women suffering violence from partners is significantly negatively associated with support for more women in politics and legal measures for parity, but not with net trust.There are some negative interactions with individual's gender (being male), which suggests that the negative effects are relevant for women in particular.Conversely, the "scarce jobs for men' attitude is negative associated with the same outcome variables, but the interactions with "male" are negative, so the effects are even more negative for men (see Tables A2 and A3).Altogether, we have to conclude that different aspects of masculinity culture have different associations with political gender attitudes in Europe, but the net trust in female politicians, in particular, is rarely affected.

Conclusion
In this article we analyzed the political gender attitudes of European citizens.Based on existing literature and theories, we formulated a number of hypotheses about individual-level effects on these attitudes and their mediators.We also proposed a number of country-level hypotheses and interaction between individual-and country-level factors.Our empirical results only partially support the theoretical expectations.
First, we should emphasize that we find very high levels of trust in women as political representatives, as well as support for more women in politics and legal measures to achieve gender parity.In fact, more people would trust a female as a political representative rather than a male than the reverse, especially among women themselves.These high levels of trust and support hold for the EU-28 sample as a whole, but also in the vast majority of individual countries, with very few exceptions in Eastern Europe.This is consistent with recent findings from the US that the public evaluates women politicians more highly than men (Cormack and Karl 2022), but the mechanism is likely to be different than this article proposes -partisanship (see also Bridgewater and Nagel 2020;Dassonneville, Quinlan, and McAllister 2021).
Second, we do find the expected effects of gender, age, and education on political gender attitudes (Alexander and Jalalzai 2016;Bauer 2020;Coffé and Reiser 2023;de Geus et al. 2021;Scarborough, Sin, and Risman 2019).The effects of professional background and political ideology are more variable across different outcome variables and not always significant (cf.Bauer 2017).
Third, and most importantly, however, the effect of the respondent's gender in particular persist even after we include measures of general and politics-specific gender-stereotypical beliefs in the models.In other words, men are significantly less likely to trust females as political representatives, agree that we need more women in politics and support legal measures to achieve parity even after controlling for their beliefs about the proper role of women in society and their (dis)interest in positions of political leadership.This result is unexpected in terms of the theories that predict that political gender attitudes are based on such perceptions about what women want and should be doing (Bauer 2020;Eagly and Karau 2002).
Fourth, our results about the incomplete mediation of political gender attitudes by gender stereotypes is consistent with the "benevolent sexism" idea, which can be traced back to the Ambivalent Sexism Inventory (ASI) developed by Glick and Fiske (1996).The ASI identifies two dimensions of sexism that are correlated but separate: hostile sexism, which is based on antipathy, and benevolent sexism, which is subjectively positive from men's point of view, but it is nevertheless based on and perpetuates biases and stereotypes.More recently, Batista Pereira and Porto (2020, 887) apply the idea in the context of Brazil.They posit that "benevolent sexists tend to reject gender equality based on views about gender complementarity and stereotypes about women's purity," but at the same time they can support electoral gender quotas in order to foster such values.So, public support for policy measures, such as quotas, is actually reinforced by paternalistic views about the role of women in politics.
Fifth, from the country level variables that we examine, masculine culture and the level of religiosity do not have the expected negative main effects (cf.Inglehart and Norris 2003).This could be due to the fact that in such countries -for example in Spain and Portugal -women have mobilized against traditional discourses rooted in the national culture and religion that restrict their roles in politics.So, the effects of culture and religion are counteracted by the backlash of political mobilization in favor of more women in politics and in positions of political power.
The lack of evidence for effects of these two country-level variables might be due to the fact that they operate in more complex interactive ways.For example, Neundorf and Shorrocks (2022) argue that both societal modernization and women's political empowerment need to be present to increase support for women in politics.More generally, the literature suggests that it is difficult to anticipate the local effects of factors such as the presence of female political role models in East and Southeast Asia (Liu 2018) or affirmative action measures in Africa: not only do such factors fail to increase support for women in politics, but they might be generating backlash as well.
Our study finds that the legacy of communist continues to have a very strong effect on political gender attitudes (Alexander 2012;Inglehart and Norris 2003), especially among men living in the post-communist countries in Eastern Europe.To some extent this is quite surprising, not only because more than 30 years have now passed since the fall of communism in Europe, but also because actual levels of participation of women in politics and business life are not too low, even if they fall short of the highest levels observed in the Nordic countries.Many of these countries have had female primeministers, deputy prime-ministers, ministers, chairs of the legislatures, and party leaders (This is consistent with the result of Costa and Wallace 2021 that women's presence does not influence significantly their perceived political efficacy or the likelihood to vote for female candidates.).Still, men in post-communist countries, in particular, exhibit low trust in females as political representatives and do not think that more women in politics are needed, let alone legal measures to achieve that.
It is also noteworthy that the gap between respondents in post-communist and other countries does not disappear with age (Inglehart and Norris 2003;Norris and Inglehart 2000), again especially for men.While younger men have more favorable political gender attitudes, the effect of age is relatively small and does very little to close the gap with Western and Southern Europe, and the gap between men and women within Eastern Europe as well.This result has an important policy implication: active measures are needed to increase trust in female politicians and support for gender equality in politics in the post-communist countries; these changes would not come simply as new generations come and older generations retire.
Another policy implication of our work is that correcting negative gender stereotypes (such as women are not interested in positions of political power or that their place is at home) would not close the gap in political gender attitudes between men and women.As the mediation analyses show, the gap remains just as wide even after gender stereotypes are taken into account.
One important limitation of our work is that it only taps into attitudes and opinions, and we cannot say much about how such attitudes and opinions translate into actual relevant behaviors, such as voting for female political candidates or taking political actions for gender parity in politics.We also only look at one snapshot in time, and we miss how the relationships we study develop over time, both in terms of descriptive trends and in terms of causal effects.We should also remind of the possibility of social desirability bias affecting the estimates of absolute levels of public support (Burden, Ono, and Yamada 2017;Krupnikov, Piston, and Bauer 2016;Streb et al. 2008).
Future research can address these limitations by linking opinion data to behavioral data on voting in particular, but also on various forms of political action, such as participation in demonstrations, petitions, and other forms of political activism.The research should also be extended to other countries in Europe and beyond where baseline levels of support for political gender equality and trust in female politicians are much lower than observed in the 28 EU members that we study.In these different contexts, the determinants of gender political attitudes might be quite different, but their substantive importance could be even greater.On a global scale, political gender attitudes could depend even more strongly on national culture and institutions.

Notes
1.There is a huge literature on the effects of public opinion on public policies, which in turn influence outcomes in society, such as women's political under-representation.For example, Lax and Phillips (2009) study the responsiveness of gay rights in the American states to public opinion.More general studies include Stimson (2015) for the US and Rasmussen, Reher, and Toshkov (2019) for the case of Europe.The book by an interval Inglehart and Norris (2003) is a classic survey of gender equality around the world, which examines public opinion in detail, but has relatively little on the role of public policies.Lomazzi and Crespi (2019) is a study of gender mainstreaming policies in Europe that pays attention to the role of public opinion as well.2. In the Supplementary Material, we also report models that include the marital status of respondents as an individual-level predictor.3. Note, however, that the meaning of political left and political right is highly context specific and differs from country to country (e.g., Wojcik, Cislak, and Schmidt 2021).In some, it has mostly socio-economic connotations, while in others it evokes sentiments related to cultural issues, moral permissiveness and nativism.Therefore, it a combined sample with respondents from different European countries, the effect of left-right self-placement can be diluted due to the different interpretations that respondents have of these terms.Political positions could also be partly endogenous to the other socio-demographic variables that we consider.4. The scale with which this dimension of culture is operationalized is a continuous one, so countries are not classified as either masculine or feminine, but receive a score between 0 and 100.The scores aggregate information about the values of the population of a country collected from a Values Survey Module.For an evaluation of the external validity and internal consistency of the scales, see Gerlach and Eriksson (2021).5.This significant positive association is robust to alternative measures of religiosity (the percentage of people who say religion is very or quite important and the percentage of people who attend church regularly, both variables from the European Values Survey 2017, missing values for Greece and Malta) (European Values Survey 2022).Church attendance is significantly positively associated with support for more women in politics as well.

Figure 1 .
Figure 1.Trust in females (bottom left) and males (bottom right) as political representatives in 28 European countries, measured on a scale between 1 and 4. The top panel shows the country-level averages of net trust in female political representatives, calculated at the individual level as the difference between trust in female and trust in male.All country averages are weighted by the survey poststratification weights.

Figure 2 .
Figure 2. Predicted probabilities of net trust in females as political representatives (left panel) and support for legal measures for gender parity in politics (right panel) as a function of age.Blue lines show the predicted probabilities for men and red lines for women.Dotted lines are for respondents in countries with a communist legacy, and solid lines in countries Without.The predicted probabilities are based on models 1 and 3 from Table 3 (Other Covariates Fixed at Their Means).

Table 1 .
Operationalization of Variables.Commission 2019).Second, age of respondents is measured in years since birth.Third, to assess the educational background, respondents were asked at what age they stopped full-time education (European Commission 2019).The respondents' profession is operationalized as a binary variable and coded as 1 for managers and 0 for other professions (European Commission 2019).Left-right selfplacement is measured on a scale between left (1) and right (

Table 2 .
Results from Individual-Level Multivariate Linear Regression Models.

Table 3 .
Results from Multilevel Multivariate Linear Regression Models.
net trust for young women in post-communist and in other countries, which disappears almost complete with age.