Subtle Gender Stereotypes in the News Media and Their Role in Reinforcing a Culture Tolerating Violence Against Women

This paper examines how news media play a crucial role in tolerating genderbased violence by constantly reinforcing ‘subtle’ gender stereotypes. After presenting the institutional background in which media, gender stereotypes and gender-based violence have been studied over the last 30 years, subtle gender stereotypes are defined as those stereotypes that are less evident than the obvious ones and are not usually correlated to gender violence. Nevertheless, they can be detected in the underrepresentation and misrepresentation of women in the news, as measured by the GMMP (Global Media Monitoring Project), which is the longest and most extensive study on the representation of women in news media.


Introduction
I start from a brief presentation of the international background from which the awareness of the relationship between media, gender stereotypes and gender-based violence has arisen in the early 1990s. Considering the main statements issued by international organisations such as the United Nations and the Council of Europe, I focus on the ambivalent power of mass media. Media can strengthen a patriarchal culture that legitimates or tolerates violence against women, or, on the contrary, they can contribute to preventing and contrasting it. International organisations, as well as feminist associations and researchers, have long ago highlighted the relationship between virtual and actual violence: images of violence against women, pornography, sexualisation or objectivation of women, disseminated in and through the media, are charged to be factors contributing to violence against women (United Nations 1993, 1995. In recent years, a new consciousness has emerged: gender stereotypes spread out and built-in mass media contribute to reinforce a 'symbolic' gender inequality that tolerates, or even legitimates, violence against women (Council of Europe 2011, 2016. In this paper I consider the role of subtle gender stereotypes, namely those stereotypes that are less evident than the obvious ones and are not usually correlated to violence against women in the common opinion. After presenting the results of the Global Media Monitoring Project (GMMP), the longest and most extensive study on the representation of women in news media, I will suggest considering the underrepresentation and misrepresentation of women detected by the GMMP in the last decades as indicators of subtle gender stereotypes that strengthen the historically unequal power relations between men and women, on which gender violence is grounded.
Finally, considering the causes of female underrepresentation and misrepresentation in media from different perspectives, I argue that in most cases subtle gender stereotypes are not caused by deliberate mystification of the reality, but are the result of several factors, such as the low progress of gender equity, the woman's status in society, and, last but not least, the journalistic practices. Challenging these practices could be the first step for news media to prevent and contrast gender-based violence. Journalists have the power to represent a more realistic world, where women have made headway in gaining equal rights, and should increase their focus on women in the news. This great responsibility is renewed by the challenge to fight gender-based violence launched by the Council of Europe with the Convention on Preventing and Combating Violence against Women and Domestic Violence (Council of Europe 2011).

2
The Connection between Media, Gender Stereotypes, and Gender-Based Violence Violence against women in society arises as a powerful issue on the international agenda during the 1990s. The Declaration on the Elimination of Violence against Women of the United Nations (1993a) is the first international instrument expressly devoted to the issue. Complementary to The Convention on the Elimination of All Forms of Discrimination against Women (United Nations 1979) and the Vienna Declaration and Programme of Action (United Nations 1993b), it provides the most widely used definition of violence against women, as "a manifestation of historically unequal power relations between men and women" and "one of the crucial social mechanisms by which women are forced into a subordinate position compared with men" (United Nations 1993a, 1). The 1995 UN Beijing Declaration and Platform for Action (henceforth BDPfA) includes violence against women as one of the twelve critical areas of concern to be addressed by governments, the international community and civil society, including nongovernmental organisations and the private sector, to support the advancement of women around the world. It also draws attention to the specific role of media: images in the media of violence against women, in particular those that depict rape or sexual slavery as well as the use of women and girls as sex objects, including pornography, are factors contributing to the continued prevalence of such violence, adversely influencing the community at large, in particular children and young people (United Nations 1995, 49) In the first years after the BDPfA, the issue of the relationship between media and violence against women mainly focuses on two points: (i) the blatant negative representation of women in media content, in its various forms such as pornography, sexualisation, objectification, materialisation, and its contribution to violence against women in society; (ii) the role of news media in presenting violence against women as a human rights violation and stimulating a larger public debate (Gallagher 2000). During the 2000s, the connection between media and violence against women becomes more intricate. Digital media, especially social media, have made it possible to share ideas, thoughts, and information on a global scale, creating new opportunities for the participation of women in communication and media, for the dissemination of information about women, for the contrast of violence against women (for an overview about governance networks on gender and media see e.g. Padovani, Pavan 2015, 2016. However, globalised media are also used to spread worldwide gender stereotypes and imag-es or speeches demeaning to women (Ross 2012(Ross , 2013Citron 2014;Spallaccia 2017).
In 2011, some years after the Beijing Conference, the Convention for Preventing and Combating Violence against Women and Domestic Violence focuses once again on media, introducing art. 17, which draws attention to the relation between media, gender stereotypes and violence against women and recalls the media sector to its public responsibility (Council of Europe 2011). To facilitate the participation of the media sector in fighting gender-based violence, the Council of Europe has recently published a document entitled Encouraging the Participation of the Private Sector and the Media in the Prevention of Violence against Women and Domestic Violence. Article 17 of the Istanbul Convention (Council of Europe 2016). Starting from the assumption that violence against women is a manifestation of historically unequal power relations between men and women, this document highlights the connection between gender stereotypes in media contents -regarding specific and blatant forms of stereotyping, such as sexism, the trivialisation of violence or hypersexualisation of women -and gender-based violence: Media can reinforce the status quo by perpetuating gender stereotypes and attitudes that condone violence against women. […] Ideas about the inferiority of women, notions of male entitlement, or preconceived views of female or male sexuality significantly influence behaviour and justify men's use of violence against women as a form of control. Furthermore, the media play a central role in the production and dissemination of images of women and men, which significantly influence public perceptions of both sexes and gender roles. (Council of Europe 2016, 13) The results of the GMMP show that news media, despite the progress of women in society, continue to portray asymmetrical and hierarchical gender identities and roles. This portrayal is not a blatant stereotyped representation, it needs to appropriate tools to be analysed and recognised, nevertheless it contributes to reinforcing the patriarchal culture that tolerates violence against women.

3
The Global Media Monitoring Project After the Bangkok conference, MediaWatch contacted hundreds of organisations and individuals to run the ambitious project of a media monitoring day all over the world. Erin Research, a Canadian research firm, founded in 1981 4 with a long history in client satisfaction research and media analysis, was in charge of designing the research, preparing a news monitoring guide and research instruments, and analysing the data. The aim was and remains to map the representation and portrayal of women in the world's news media, develop a grassroots research instrument, build solidarity among gender and communication groups worldwide, create media awareness, and develop media monitoring skills on an international level. 5 The focus of the research was and still is on news media because they are the major source of information, facts, ideas and opinions for people throughout the world. They are the famous "window on the world", from which we are expecting to be told "what we want to know, we need to know, we should know" (Tuchman 1978b, 1). However, how Gaye Tuchman stated in 1978: like any frame that delineates a world, the news frame may be considered problematic. The view through a window depends upon whether the window is large or small, has many panes or few, whether the glass is opaque or clear, whether the window faces a street or a backyard. (Tuchman 1978b, 1) The GMMP research is exactly designed to analyse the 'size', the 'shape', the 'cleanliness' and the 'point of view' on the world of the national and regional news across various countries and regions worldwide.
WACC assigns to each participating country a minimum and a maximum number of newspapers, radio and television newscasts and, since 2015, online news sites and Twitter feeds, so that the media monitoring is based on the national media density. Each national coordi-nator is responsible for building the sample, following the criteria of density and diversity -audience, ownership, and language -in their country. In each country, research teams analyse the most important television and radio newscasts of the day, 12 to 14 stories appearing on newspapers and websites, and 15-20 tweets posted no later than 6:30 p.m. on the media monitoring day. Monitoring involves a quantitative and qualitative analysis. The quantitative approach captures statistical data on news topics, women and men in the news -such as journalists, people interviewed or people about whom the news is -the types of news stories in which they appear, their function in the news and some of their personal and social data, such as age, occupation or social position. The qualitative analysis examines the patterns of gender in the news, more in-depth, considering the role of story angle and perspective, language and visual representations in constructing and sustaining or challenging gender stereotypes.
The first edition of the GMMP took place on 18 January 1995. The results were published in a report entitled Global Media Monitoring Project. Women's Participation in the News (MediaWatch 1995) and were released during the UN Fourth World Conference on Women convened in Beijing in September that same year. The BDPfA adopted at the end of the Beijing Conference included mass media as one of the twelve critical areas for the empowerment and advancement of women all over the world. According to the BDPfA, media are not only responsible for the images of violence against women which adversely contribute to the continued prevalence of such violence, but they also play a wider and more powerful role in facilitating "a global communication that transcends national boundaries and has an impact on public policy, private attitudes and behaviour" (United Nations 1995, 9). On the basis of this statement, the UN and its Member States established two strategic objectives: first, "increase the participation and access of women to expression and decision making in and through the media and new technologies of communication" (ob. j1); second, "promote a balanced and non-stereotyped portrayal of women in the media" (ob. j2). The objectives of the BDPfA "Area J", namely "Women and the media", became, and remain, the touchstones for the GMMP. The findings of GMMP have become a benchmark for evaluating the extent to which the aims of Area J have been achieved since 1995.
The subsequent GMMP, which took place on 1st February 2000, was coordinated by the World Association for Christian Communication and involved more organisations in the research and made the study more contextual. The monitoring day created great enthusiasm and solidarity among the hundreds of monitoring groups in 70 countries which provided over 50,000 data recorded from more than 16,000 news stories. The preliminary findings were released in time for the five-year review of the implementation of the BDPfA (Beijing Subtle Gender Stereotypes in the News Media Quaderni del Comitato Unico di Garanzia dell'Università Ca' Foscari Venezia 1 203 Language,Gender and Hate Speech,[197][198][199][200][201][202][203][204][205][206][207][208][209][210][211][212] + 5) 6 and the final report was published providing an extensive analysis of gender representation and portrayal in the world's news media. The research found that women's visibility was uniformly extremely low, as it was in the 1995 report. In no medium, region or news topic did the female-male ratio approach parity (Spears et al. 2000).
The third GMMP in 2005 saw the participation of 76 countries. Some progress in women's presence was evident: 21% of news subjects -people interviewed or whom the news is about -were female vs. 17% of 1995 and 18% of 2000. However, the main and substantive outcome was women's near invisibility in the news: for every woman appearing in the news there were five men. Men as newsmakers in every major news topic outnumbered women. Just less than 10% of all stories were focussed specifically on women. Women were rarely central in the most relevant stories covered by the news agenda. The authoritative and prestigious role of the expert was played by men in 83% of cases vs. 17% of female cases. And women were under-represented in all professional categories. Overall, news stories were twice as likely to reinforce gender stereotypes rather than challenging them. News stories on gender (in)equality were almost nonexistent. Thus, the third GMMP revealed that the world reported in the news is mostly male (WACC 2005).
The fourth GMMP (WACC 2010) witnessed an explosion in participation: 1281 newspapers, television and radio stations were monitored in 108 countries in Africa (26), Asia (13), the Caribbean (11), Europe (32), the Middle East (6), the Pacific region (5), Latin America (13) and North America (2). The collected data amounted to 16,734 news items, 20,769 news personnel announcers, presenters or reporters, and 35,543 total news subjects (people interviewed as experts, spokespersons, vox populi or eyewitnesses, and individuals whom the news is about). The research found some improvement in comparison to 1995 when only 17% of the people in the news were women, in 2010 these had risen to 24%. But an increase of 7 points of percentage meant that if conditions remained unchanged and the rate of progress was maintained it would take more than 40 years to reach parity (WACC 2010).
At the end of the GMMP 2010, WACC planned a roadmap to accelerate the progress in the portrayal and representation of women in the news, proposing a plan of action to be carried out in partnership with editors, media regulatory bodies, journalist associations, training institutions, governments, the communication and information society, and the civil society. The roadmap included several actions, such as building a list of women experts covering diverse promoted thematic areas, creating gender-and-media curricula in schools, journalism training institutes and centres, adopting and applying policies on gender parity in the media (WACC 2010).
Notwithstanding the efforts made by WACC and its international networks in 2010-15, the results of the GMMP 2015 were disappointing: even though women made up about 50% of the world population, only 24% of the people heard, read about or seen in newspaper, television and radio news of all the world (114 participating countries) were women, exactly as it was in 2010. Women's relative invisibility in traditional news media had also crossed over into digital news delivery platforms. Only 26% of the people in Internet news stories and media news tweets combined were women. Once again, GMMP showed that news media were (are) a narrow window on a male-centred world.

The Results of the GMMP Exercise in Italy
If we take the Italian context into consideration, in the old media (radio, TV and newspapers), which have been monitored since 1995, some progress had been made: the visibility of women as sources or subjects of news has increased from 7% to 21% (Azzalini, Padovani 2015a, 2015b, 2016. However, Italy still scores lower than the European and global average (respectively 25% and 24%) and still shows shortcomings as far as equal opportunity for women in news media. The main asymmetries reported in 1995 tend to persist in 2015 and demonstrate a gap between the advancement of women in society and the image of women (and men) reflected by news media. For example, despite the increased participation of women in Italian political life and institutions (in 2015 there were 30% of women in Parliament), only 15% of politicians in the Italian news were female (vs. 85% of men; total 231) and only 15% of subjects and sources of political news were women (men 85%; total 123). Overall, women are under-represented in the most prestigious and prominent news topics e.g. politics/government (15%) and the economy (10%). When we look at occupation or social position (e.g. homemaker, parent, child, student, villager) of persons, 77% of individuals coded as homemakers are female, 67% of students or pupils are female and 43% of villagers or residents. This means that women are more represented in general 'social groups' than in more specific 'professional groups'. Among the professional categories, women are more represented as health, social and childcare workers (67%), celebrity, artist, actor, writer, singer, TV personality (51%), activist or worker in civil society organisation or NGOs, trade union (45%) and doctor, den-Subtle Gender Stereotypes in the News Media Quaderni del Comitato Unico di Garanzia dell'Università Ca' Foscari Venezia 1 205 Language,Gender and Hate Speech,[197][198][199][200][201][202][203][204][205][206][207][208][209][210][211][212] tist, health specialist (44%). The high visibility of female celebrities is linked to the statement made by the famous actress Angelina Jolie and reported by the majority of media outlets on 25 March 2015 (date of the GMMP day sample) concerning her choice to have the ovaries removed to prevent herself from getting the same form of cancer that affected many women in her family. The high visibility of female worker in health and medical professional groups -both at low level, such as health, social and childcare workers and high level, such as doctors, dentists, health specialists -as well as workers of the so-called 'third sector' is maybe a positive, although weak, signal of a change in the Italian news media towards a more realistic representation of women. Contrarily, the high percentage of women whose profession or social position is not made explicit (41%) is a clear sign of a permanent marginalisation of women. The high percentage of women whose profession or social position is not made explicit has been a recurring result in the previous editions of the GMMP (Spears et al. 2000;WACC 2005WACC , 2010, indicating that women have easier visibility when they are viewed 'anonymously', independent of their social role or professional skills. Women only represent 18% of the people recorded by the GMMP 2015 as sources of expertise or qualified competence invited to explain or comment an event or a specific issue and are under-represented in occupations with high social status such as academic experts (25%), businesspersons (24%), lawyers, judges, magistrates, legal advocates (21%). Women are more likely to make the news as victims in a ratio of 1:4 vs. the male ratio of 1:10.
Considering digital media analysed by the GMMP for the first time in 2015, the Internet is by far the most inclusive new medium for women in Italy, while Twitter is the most exclusive. On the pages of the online newspapers monitored on 25 March 2015, women reached 29% (out of a total of 347 people); whilst the visibility of women on Twitter was only 17%. In any case, the scenario provided by the digital media is not unlike the one provided by the old media: a world in which women are virtually absent and only play marginal roles. For example, the role of expert is infrequent on the Internet, unlike what happens in traditional media, but as in traditional media it remains a male prerogative; only in 16% of cases is it held by a woman. In one case out of four, women make the news for reasons presumably unrelated to either specific skills, or to political, institutional, social roles, having neither a professional role nor a social position made explicit. Women for which the profession is known or can be deduced from the news in 16% of cases are politicians that have a high degree of visibility in digital media, however less than their male colleagues (24%). As in the traditional media, women in Internet news are more likely to be presented as victims than men (24% vs. 13%) and are more often described on the basis of some family relations such as mother, daughter or wife than men (15% vs. 11%). Moreover, women are central in 22% of Internet news. This confirms the wellknown fact that the Internet media are more accessible to women, but this is not enough. Regarding stereotypes, little news both on the Internet and Twitter contains elements of challenge (respectively 6% of the total of 99 news and 4% of 144 tweets); whilst almost all Internet news (84%) reinforces gender stereotypes vs. 26% of Twitter.
As authors of the GMMP Italy Report wrote "Twitter gives us mostly reality pills" (Azzalini, Padovani 2016, 283). "Pills" of a "strange" world from which women are left out. Even if most of the news tweeted, as well as most of the news covered by newspapers, radio and television, does not contain blatant stereotypes, the world that they describe is a fake reality. The strength of the GMMP approach consists precisely in revealing not only the blatant but also the 'subtle' gender stereotypes. Women in the news do not outnumber 30% (depending on the context) of the persons seen, heard or read about, despite accounting for half of the world population. Furthermore, they have feeble identities, being socially and professionally 'anonymous' in most cases. In one case out of four, they are presented as victims. More than men they are brought back to family relationships (mother of-, daughter of-, wife of-). If they only play marginal roles in news media, despite their advancement in all aspects of social life, this means that news media continue to reflect the historically unequal power relations between men and women.

5
The Role of Journalist Practices Early research on media and gender representation (Tuchman 1978a), based on George Gerbner's theory of cultivation (DeFleur, Ball-Rockeach 1995), stated that the so-called "symbolic annihilation" 7 of women was due to the social role of the mass media (Gerbner 1972). Starting from the hypothesis that media 'cultivate the resistance' of the value system of the socially dominant group (the male one), these studies focussed on the resistance of media to represent social changes, including, among others, those related to gender identities, roles and relationships. Media were charged to be structurally conservative and unilaterally powerful (Van Zoonen 1994; Ross, Sreberny-Mohammadi 1997). This thesis was subsequently -although not totally -overcome by more complex methodologies and approaches more attentive to the change in the media system.

Monia Azzalini Subtle Gender Stereotypes in the News Media
Since the second half of the 1990s, several studies have investigated the factors related to the underrepresentation and misrepresentation of women in the news media, finding somewhat different evidence. Lynn M. Zoch and Judy van Slyke Turk (1998), for instance, demonstrate, through the analysis of three United States newspapers between 1986 and 1996, that female journalists are more inclusive than male journalists: they are more likely to quote female sources, but also middle management sources, of either gender, while their male colleagues only quote those sources considered top managers, male in most cases. This evidence founded the optimistic prediction that by increasing the number of female journalists, women's representation in the news media could improve. But even if the number of female journalists has risen, gender balance remains far from the reality. Thomas Hanitzsch and Hanush Folker (2012) deepen the matter, analysing the role concepts of female and male journalists in 18 countries. Their research reports a substantial gender similarity in the perception and exercise of the profession, independent from either the individual level or in newsrooms dominated by women, or in sociocultural contexts where women have achieved a high level of empowerment. This would explain why women continue to be underrepresented and misrepresented in the news media, despite the increase of female journalists.
Other recent studies report contrasting evidence. According to Shor et al. (2015), the persistence of gender differences in media coverage is more connected to the progress of equal opportunities in a country than to culture and journalistic practices. Using longitudinal data (13 American newspapers, between 1983 and2008), the authors empirically isolate media-level factors and examine their effects on women's coverage rates in hundreds of newspapers, finding that societal-level inequalities are the dominant determinants of continued gender differences in news media coverage. Since media focus nearly exclusively on the highest strata of occupational and social hierarchies, in which women's representation has remained poor, women continue to be underrepresented.
However, Short et al.'s hypothesis does not explain why the representation of women in top positions, such as female politicians, is lower than their real representation in society, as demonstrated by the results of the GMMP. For instance, as mentioned above, the latest edition of the GMMP reported that 15% of politicians represented in the Italian news were women, compared to the actual 30% in the Italian parliament that year (2015). In 2005, in Rwanda -which had the highest proportion of female politicians in the world (49%) -only 13% of politicians in the news were women (WACC 2005). A hypothesis to explain this phenomenon is that there is a positive correlation between the media representation of women and the status they enjoy as a social group in a society: In most societies, women are still assumed to have lower status than men. Hence their views are regarded as less important. These cultural assumptions have an impact on journalistic practices in privileging the male. Sometimes it seems that women's very existence has been forgotten or overlooked in the selection of guests or interviewees. (Portraying Politics 2006, 13) This is a statement that suggests a research hypothesis recently explored, for instance, by Humprecht and Esser (2017). By analysing the political news of 48 online newspapers, of 6 different types (popular, public, private, corporate, independent and web-based), of 6 western countries (France, Germany, Great Britain, Italy, Switzerland and the United States), for 6 weeks from October 2012 to February 2013, the authors investigated the correlation between journalistic practices, a country's gender culture and the level of advancement of women. The authors correlated a cultural indicator -the masculinity index in the Hofstede classification (2001) -with an indicator of gender equality -the Global Gender Gap Index of the World Economic Forum (2015) -with the results of a content analysis of the 48 selected online newspapers. The conclusion they reached is that journalistic routines seem highly influenced both by the level of masculinity of a country (the higher the level, the fewer women present in the media) and the progress of women in that country (the more advancement in gender equality in a country, the more women are included in online political news).
As we have seen, research on gender representation and news media reports somewhat different and contrasting evidence, possibly due to the complexity of the matter. In any case, no research perspective supposes that the media misrepresentation of women is caused by deliberate mystification of reality; it is the result of several factors, including, as mentioned above, journalistic practices. But journalistic practice can be changed. As Margaret Gallagher states, gender-fair journalism is nothing more than quality journalism and the GMMP shows that if the journalistic practices remain unquestioned, they often produce news that reinforces gender stereotypes (Gallagher 2006). Therefore, the last edition of the GMMP has developed an Action Plan with 5 targets that involve journalistic practice in order to end news media sexism by 2020. First: create conditions for newsrooms that support gender equality (100% of national public media and 40% private media in each country), through media ethics codes, reporting guidelines, training, mentoring, peer support and "other strategies to transform gender-biased news media practice and newsroom cultures" (WACC 2015, 99). Second: increase in women's presence as subjects and voices in the news (overall global presence: 50%) with a systematic institutionalization of a gender-sensitive journalistic cul-ture. Third: "increase in the proportion of content that challenges gender stereotypes" (WACC 2015, 99) (global average: 30%), in order to counteract and eliminate implicit stereotypes underlying genderbased discrimination. Fourth: increase the proportion of news that highlights issues of gender equality and inequality (30%). Fifth: increase the percentage of news reporting that is anchored in a critical human rights perspective (30%).
In addition to 5 targets, the GMMP 2015 Action Plan provides "specific priority actions" involving media regulatory agencies, media houses, civil society, media professional training institutions, funding agencies and researchers. All of these stakeholders are asked to take steps to end media sexism by 2020 in an integrated way. Particularly, they are called to support a change of journalistic practices in their country with several requirement, such as adoption and enforcement of gender policies and guidelines for gender-sensitive reporting (media regulatory agencies), development of gender aware ethics and practice codes (media houses), lobbying editors to apply gender policies (civil society), lending support to journalists and media engaged for gender equality (civil society), incorporating gender dimension in all modules of journalism training curricula (media professional training institutions), supporting media development work that emphasises gender equality in content production (funding agencies), working with unions to carry out surveys addressed to journalists about their experiences of workplace environments (researchers).

Conclusions
The GMMP provides effective tools of research to measure and explain the features of the world pictured by the news. Its results clearly show that 'who' or 'what' appears in the media and 'how' or 'why' this is done affect gender equity in the news but also affect society, through the news. According to the argument of the 2016 Council of Europe document, the GMMP could be considered as a tool to measure how mass media can reinforce rather than change a culture that tolerates or even legitimates violence against women. News media are the major and most influential source of information, ideas and opinion for most people around the world. Over the last few years, the quick advancement of ICTs has increased and expanded global access to information. However, as the findings of the GMMP show, the news content has not changed as fast as the technologies. Journalists have the power to depict a more realistic world where women have made great headway. Nowadays, faced with the renewed challenge launched by the Council of Europe to fight gender-based violence, the journalist's role to provide gender-fair news is a great power and a great responsibility.
With its Action Plan the GMMP also provides specific priority actions to be taken by several stakeholders to support the strategic role of journalists in every country of the world.