Introduction

Media representations of minorities offer people the opportunity to virtually encounter social groups they would seldom meet in everyday life (Joyce & Harwood, 2014; Park, 2012; Ramasubramanian, 2013), serving as a basis for their attitude formation. In general, media representations work as parasocial contact (Ortiz & Harwood, 2007; Riggle et al., 1996), with similar consequences to direct contact (Allport, 1954): depending on whether they are positive or negative, they can reduce or increase prejudice toward social groups. One minority group whose salience in mainstream media discourse has increased rapidly in the last few years—sometimes with positive consequences (e.g., fostering understanding and acceptance), sometimes with negative ones (e.g., turning them into targets of hate)—are trans and gender diverse (e.g., non-binary, gender non-conforming, genderfluid, agender) people. Their identities, roles, and/or presentations do not conform to the gender/sex they were assigned at birth (American Psychological Association, 2015). They comprise 1–2% of the population, reaching up to 3% in younger generations (Boyon, 2021). In the US, less than 40% of the population personally knows at least one TGD individual (Adam, 2015; Frankovic, 2019). When the chances of direct contact with a group are limited, people have to rely on means of indirect contact, such as media representations, to shape their attitudes toward the group (Schemer & Meltzer, 2020).

People’s attitudes toward gender minorities, but also the legal protection these groups experience, vary across countries (Flores & Park, 2018). This raises the question whether the way TGD people are represented in news media correspond with the broader sociocultural context of a given country. Uncovering such correspondences would demonstrate that media representations are involved in (re)producing the social contexts gender minorities live in. So far, TGD media representations have mostly been investigated within a US and a UK context (e.g., US: Billard, 2016; Capuzza, 2014; Mogul et al., 2011; Wood et al., 2019; UK: Baker, 2014; Zottola, 2018; for exceptions see: Sweden; Åkerlund, 2019; Spain; Olveira-Araujo, 2022). Also, past research has focused on TGD people in entertainment media (e.g., films, books; see: Capuzza & Spencer, 2017; McLaren et al., 2021; Ryan, 2009), since their entrance in the world of news media is far more recent—and consequently less explored.

Therefore, the present study investigated the ways TGD people’s news media representations are linked to differences in legal protection and social acceptance of TGD people in three European countries (namely Sweden, the UK, and Italy). To our best knowledge, this is the first study to provide such cross-cultural comparison and also the first to investigate TGD news media representations in Italy.

Media Representations of TGD People

In the past, entertainment and news media representations of TGD people have often been stigmatizing, depicting them as deviants, tricksters, oversexualized, if not outright mentally ill and unstable (e.g., Baker, 2014; Billard, 2016; Mogul et al., 2011). While presenting common themes and patterns across time and space (Birchmore & Kettrey, 2022), such representations also evolve in ways that are unique to the societal, cultural, and legal context they are situated in, simultaneously producing and being reproduced by media (Arendt, 2013; Ramasubramanian, 2007). To compare how media representations of TGD people vary across countries, we have identified five common and recurrent dimensions in the past literature: (1) the valence of media representations, (2) the roles in which TGD people are depicted, (3) specific aspects of TGD people’s gender identities, (4) interest for their bodies, and (5) whether they are represented as individuals or as a group. These dimensions of media representations may reflect and influence evaluations of and attitudes toward TGD people in a given society.

Media representations of minority groups can be neutral, positive, or negative. When a minority group gains sudden salience in media discourse, its initial media representations are often negative—as it was the case for gay men in the 1990s, thrust in the media discourse and initially portrayed as deviant, mentally ill, and HIV-positive (Alwood, 1996; Gross, 1991). Such negative media representations work as negative parasocial contact and can increase prejudice and aversion towards of social groups (Park, 2012; Schemer & Meltzer, 2020). For example, media depictions of minority groups as immoral can promote dehumanization and less favorable attitudes towards them (e.g., Esses et al., 2008). On the other hand, positive media representations of social groups work as positive parasocial contact and can reduce people’s prejudice and improve attitudes towards them. According to Clark (1969), media representations of minorities are considered respectful and equal to the majority’s representations when they include both positive and negative roles, with positive and negative valence. However, past research has shown that, as time moves forward, progress in representation does not necessarily follow: comparing the coverage of TGD-related stories in British press in 2012 and 2018–19, Baker (2014, 2019) found that TGD people were written about in an increasingly negative way.

The roles TGD people are represented in may influence the perceptions of their characteristics (see Correspondent Inference Theory; Jones & Davis, 1965). TGD people have often been represented in stereotypical roles with low morality and agency such as criminals (e.g., Baker, 2014; Capuzza & Spencer, 2017) and victims of violence and discrimination (e.g., Billard, 2016; Capuzza, 2014). However, TGD people are also represented as resilient pioneers (Baker, 2014; Wood et al., 2019) and successful professionals in various fields (e.g., Jacomella, 2017; Scipioni, 2022). The depiction of TGD people as ‘gender-deviant’ criminals has a long history in modern media (e.g., Kane, 2012; Mogul et al., 2011); these narratives may be reinforced when paired with other stereotypically illicit roles (e.g., sex worker, illegal immigrant). The role of victim is even more widespread than the criminal one (DeJong et al., 2021; Kane, 2012); it can be caused by subtle transphobia and discrimination or blatant transphobia and violence (Baker, 2014). Still, in recent years, TGD pioneers have managed to gain more rights and increase the number of people who felt safe coming out as gender diverse (Caron, 2018); for example, there has been an increasing number of well-known public figures and celebrities coming out as trans or gender diverse (e.g., Elliot Page, Caitlyn Jenner, Bella Ramsey). This also meant TGD people managed to enter spaces they were previously excluded from, especially in respected professions such as politics (Caron, 2018).

Another important dimension of TGD media representation is the way their sex/gender and gender identities are reported. Data show that the number of trans men and trans women within the TGD population is nearly the same (Leinung & Joseph, 2020); however, trans women are overrepresented in the media compared to trans men and gender diverse people (Åkerlund, 2019; Billard, 2016; Capuzza, 2014). One reason for this overrepresentation may be found in gender essentialist beliefs that see trans women as men (Smiler & Gelman, 2008), whose transitions are perceived as more shocking and intriguing than those of trans men and gender diverse people. Indeed, men’s deviations from masculinity are deprecated and punished more harshly than women’s deviations from femininity (Schilt & Westbrook, 2009; Sirin et al., 2004). As such, the reporting of TGD people’s sex/gender can lead people to perceive certain gender identities (e.g., trans women) as predominant within the TGD community, even when this is not the case.

Related with TGD people’s sex and gender identity are the issues of misrepresenting them (Mocarski et al., 2019). For example, when actor Elliot Page came out as a trans man, many headlines used the actor’s birth name and pronouns (e.g.,“Ellen Page from Juno comes out; theactress considers herself a ‘non-binary’ trans, she will be using ‘he/they’ pronouns from now on”; TGCOM24, Italy, 1/12/20; emphasis added, where only he/they reflect their identity). Such misrepresentations of TGD people’s gender identity—usually in favor of the gender they were assigned at birth—are known as misgendering (Simpson & Dewaele, 2019). Not using a TGD person’s chosen name (i.e., using Ellen instead of Elliot) is a specific form of misgendering known as deadnaming. Since journalists aim to make headlines as clear and attractive as possible, they might use a TGD public figure’s former name to make them immediately recognizable. On the other hand, media guidelines (e.g., Yurinova, 2023) emphasize the need to be respectful and unbiased toward the article’s target and provide suggestions to avoid using a TGD person’s former name, even in the coming out phase. Whether intentional or not, both misgendering and deadnaming can have negative consequences, such as leaving the target feeling misrepresented and stigmatized (McLemore, 2015, 2018).

Media representations often focus on TGD people’s bodies—both in medical and sexual terms. Their bodies tend to attract exaggerated interest to their genitalia and surgeries (or lack thereof). Talking about trans people solely in terms of surgeries and thereby reducing TGD identities to a binary conception of post-operative transsexuality is a common practice known as medicalization (Billard, 2016; Goldbach et al., 2022). Furthermore, TGD people’s sexual behaviors and appearances are also often under scrutiny, leading to sexual objectification (Fredrickson & Roberts, 1997); objectification may increase the likelihood of TGD people experiencing fetishization and being associated with sex work (Anzani et al., 2021; Flores et al., 2018; Goldbach et al., 2022).

Finally, gender minorities can be described as individuals (e.g., “trans celebrity Caitlyn Jenner”) or as a group (e.g., “trans men”, “transgender athletes”) in the media. Generally, media representations of people are personalized and involve a disclosure of personal information (Capuzza & Spencer, 2017). Such personalized information can be helpful, but also harmful: for example, writing about the murder of an individual TGD person without positioning the crime within a larger frame of transphobic violence ignores the systemic issues that TGD people face (Barker-Plummer, 2013), whereas reading about a specific TGD person being successful might serve as positive parasocial contact. On the other hand, representing TGD people as part of a cohesive outgroup can trigger ‘us versus them’ dynamics and increase negative stereotypes (Knowles et al., 2022), but also remind readers that TGD people are not an isolated phenomenon and belong to a supportive community. Thus, both an individual and a group focus can have positive or negative implications.

It is reasonable to assume that these aspects of TGD media representations correspond with legal protection and social acceptance of TGD people. Countries with higher levels of legal protection and social acceptance likely feature more positive valence, more positive roles (i.e., pioneer, professional), an unbiased representation of gender identities, less misgendering, less focus on bodies, and a focus on both individuals and groups compared to countries with lower levels of legal protection and social acceptance.

The Present Study

This research examined how TGD people are represented in news media in three European countries that differ in their levels of legal protection and social acceptance of gender minorities. We selected countries to reflect high (Sweden), medium (UK), and low (Italy) levels. Regarding legal protection, Sweden ranks as granting more (6th out of 49 European countries, with 68% of LGBTI human rights achieved) than the UK (14th, 53.4%) and Italy (33rd, 24.7%; ILGA Europe, 2022). More specifically, in Sweden the inclusion and protection of TGD people became part of the Constitution in 2018, and discrimination against transgender identities and expression was deemed a hate crime (Carlson, 2021). Furthermore, TGD people are also protected in the workplace and public spheres by the Swedish Discrimination Act (Diskrimineringsombudsmannen, 2008). In the UK, the 2010 Gender Recognition Act offers protection against discrimination for binary trans people who are undergoing or have undergone gender reassignment processes; the definition is exclusionary of gender diverse people (e.g., non-binary, genderfluid) and binary trans people who do not plan to undergo medical transition (Dunne, 2021). In Italy, in contrast to Sweden and the UK, there is no explicit law protecting TGD people against discrimination based on gender identity or intersex characteristics, neither at a general level nor in the workplace (Spinelli et al., 2022). Italy also recently voted against extending its law against hate crimes to include gender identity and sexual orientation (Tondo, 2021), and does not legally recognize non-binary gender identities. Regarding social acceptance, TGD people reported less transphobic discrimination in Sweden (57%) and the UK (56%) than in Italy (66%, the highest in Europe; European Commission, 2019). Likewise, 90% of Swedish citizens and 89% of UK citizens said they would feel comfortable with having a trans coworker, while in Italy only 50% did (European Commission, 2019). When it comes to legal protection and equal rights, Sweden exhibits significantly more inclusive legislations than the other two countries; regarding laypeople’s attitudes, on the other hand, both Sweden and the UK appear to be more tolerant and accepting than Italy.

While TGD media representations have been investigated in full-text newspaper articles (e.g., Baker, 2014; Billard, 2016; Capuzza, 2014), no research has so far focused on short newspaper headlines. We decided to investigate headlines because they are read more often than full-text articles (Gabielkov et al., 2016), represent essential information about the article’s content, attract attention (Metila, 2013), and can inform people’s perceptions and attitudes towards the article’s content (Reis et al., 2015). For this purpose, we analyzed online newspaper headlines from each country’s most read news providers. In correspondence with the countries’ societal contexts, we expected Sweden to feature more progressive representations of TGD people than the UK and Italy. We also expected the UK to feature more progressive representations of TGD people than Italy.

Method

Sample

The headlines were downloaded from the most read online news providers (see Reuters Report; Newman et al., 2021) in Sweden, the UK, and Italy that offer free access to text articles and a search function. The search terms were: trans* (e.g., transgender, transsexual, transphobia/transphobic, transition, transvestite), non-binary (or non binary or nonbinary), genderfluid, genderqueer; they were translated in Italian and Swedish (see Table 1; Fig. 1). We excluded headlines that contained the term trans with other meanings (e.g., transmission, or transportation); duplicates were removed.

Table 1 Keywords used to select relevant headlines in each country
Full size table
Fig. 1
figure 1

Frequencies of search terms found in headlines by countries

Full size image

We aimed to collect 100 headlines from three news providers of each country (i.e., 300 headlines per country and 900 headlines in total), starting with articles published on the 31st of August 2021 and going backwards in time. This way, we collected 300 headlines from Sweden, 300 headlines from the UK, and 230 from Italy (since one news provider featured only 30 articles that matched the search terms). The final dataset thus consisted of 830 headlines. Table 2 displays detailed information about the headline sampling. The UK headlines covered a shorter time span because British news providers published a larger number of TGD-related articles than Sweden and Italy—which is likely related to the UK being the country with the highest output of entertainment and news media in Europe (Newman et al., 2021). For each headline we documented country, news provider, search term, date of publication, and article link.

Table 2 Information about headline sampling
Full size table

Coding Scheme and Procedure

We conducted a content analysis following the procedure outlined by Neuendorf (2011). Based on past research about TGD people’s representations in entertainment and news media, we created a coding scheme (Table 3) with five dimensions and multiple subdimensions: valence, roles, gender aspects, body aspects, and focus.

Table 3 Coding scheme details and headline examples
Full size table

Valence referred to the article’s positive, neutral or negative representation of TGD people. It was coded as positive, negative, and neutral/mixed. Positive valence occurred in headlines with a positive meaning for TGD people; examples included: TGD people achieving something and/or being the first to do something; social inclusion of TGD people; TGD people gaining rights, getting medical treatment and gender confirmation surgery; people standing up for the TGD community; criticism/row/debate/protest against transphobia and discrimination; calls for positive action. Neutral/mixed valence occurred in headlines without clear indicators of positive or negative valence, or with indicators for both; examples included: a TGD person coming out; information; criticism/row/debate/protest when it was not possible to assess whether they were pro or anti-TGD. Negative valence occurred in headlines with a negative meaning for TGD people; examples included: TGD people being attacked, discriminated, harassed, harmed; social exclusion of TGD people; rights being taken away and treatments being denied for TGD people; people standing against the TGD community; criticism/row/debate/protest against TGD people; questioning TGD legitimacy and minimizing TGD issues.

Roles referred to the social roles TGD people are often represented in in the media, and it included four non-mutually exclusive sub-dimensions: criminal, victim, pioneer, and professional. A criminal role was coded as present in headlines that reported TGD people breaking the law; otherwise it was coded as absent. Examples of its presence included: TGD people being violent, TGD people engaging in illegal sex work or drug-dealing. A victim role was coded as present in headlines that reported TGD people being discriminated, attacked, harassed, victimized, excluded, and further distinguished in victim of discrimination and victim of violence; otherwise it was coded as absent. Examples of its presence included: TGD people being discriminated/physically attacked at an individual or systemic level. Examples of its absence included: discrimination that was corrected, condemned, abandoned; legal battles that were won. A pioneer role was coded as present in headlines that reported TGD people pioneering the way, and further distinguished in being first, coming out, gaining rights; otherwise it was coded as absent. Examples of its presence included: TGD people achieving something for the first time; TGD people being trailblazers; TGD people coming out; TGD people winning legal battles and fights for rights. A professional role was coded as present in headlines that explicitly described TGD people as professionals (e.g., athlete, politician); otherwise it was coded as absent. The professional field was further specified as open text by the coder and later grouped by the authors (i.e., politics, military, sports, showbusiness, or other).

The gender aspects dimension referred to specific aspects related to the TGD person/people’s gender and included two non-mutually exclusive sub-dimensions: target’s gender and misgendering. The target’s gender was coded in headlines that explicitly stated the TGD person/people’s gender identity as trans man, trans woman, gender diverse; otherwise it was coded as not applicable/unspecified (NA/U). Misgendering was coded as present in headlines that explicitly referred to TGD people with the wrong pronouns, terms, and/or first name (deadnaming); otherwise it was coded as absent.

The body aspects dimension was coded as present in headlines that included mentions of TGD people’s bodies, and further distinguished in medicalization and objectification; otherwise it was coded as absent. Examples of its presence included: reports of TGD people’s surgeries, hormonal treatments, gender confirmation processes; (hyper)sexualization of TGD people. Examples of its absence included: a TGD person working in a body-focused profession or in situations that could be associated with objectification (e.g., beauty pageants).

The focus dimension was coded as present in headlines that described TGD people at an individual or group level, and further distinguished in individual and group; otherwise it was coded as not applicable/unspecified (NA/U).

For each country, two native speakers coded the headlines. All coders were researchers in Psychology (two of them were authors); all countries had an equal balance of cisgender and TGD coders, so that the interrater agreement involved both cis and TGD coders’ perspectives. The coders were instructed to focus on explicit content; for example, a headline such as “Tommy Dorfman comes out as a trans” was neither coded into the professional role nor the gender dimensions, since Dorfman’s professional status or gender are not made explicit.

We report Cohen’s kappa (Cκ) and percentage agreement in Table 4. Two sub-dimensions, criminal and misgendering, presented high percentage agreement but low occurrences and low Cκ; therefore they were excluded from statistical analyses. Apart from these, Cκ were substantial (McHugh, 2012; Neuendorf, 2017), ranging from 0.65 to 0.95. The average intercoder reliabilities across countries were Mκ = 0.81, Magreement = 89.6%. All disagreements were recoded after joint discussion.

Table 4 Inter-rater reliability for each coding dimension
Full size table

Analyses

The content analyses included two steps: first, calculating the frequencies of each coding dimensions; then, statistically comparing the frequencies between countries to establish which dimensions statistically differed between countries. For country comparisons, we employed 5 multinomial logistic regressions (when the criterion had more than two levels; e.g., trans man, trans woman, gender diverse person, NA/U) and 2 binomial logistic regressions (when the criterion had two levels; i.e., presence or absence) with country as predictor and the coded dimensions as criteria.

Results

First, we present and describe the outcomes of the coding in terms of frequencies. Table 5 shows the frequencies and percentages of the examined dimensions, their sub-dimensions and codes for each country. Second, we report the statistical analyses of the cross-cultural comparisons; Table 6 displays the results of the multinomial and binomial logistic regression analyses. Since sample size guidelines for multinomial logistic regression indicate a minimum of 10 cases per independent variable (Schwab, 2002), codes with less than 10 occurrences in at least two countries (marked by * in Table 5) were not included in further analyses. We reported odds ratios (ORs); in case of ORs < 1 we reported reciprocal odds ratios (1/OR), because odds ratios smaller than 1 are difficult to interpret (Petrucci, 2009). Throughout the paper, the conventional p < .05 is the criterion of statistical significance.

Table 5 Frequencies of examined dimensions and their percentages in relation to the total number of headlines in each country
Full size table
Table 6 Multinomial (valence, victim, pioneer, target’s gender, focus) and binomial (professional, body aspects) regression analyses of country as predictor of coding scheme dimensions
Full size table

For transparency, given the different time spans covered by the headlines, we ran additional analyses using only headlines from 2020 to 2021 across all countries. The results are described in the Supplementary Information (https://su.drive.sunet.se/index.php/s/x27RRBznLK2xi6N).

Regarding valence (Fig. 2), headlines with positive valence showed a similar occurrence across all countries, averaging around 30–35% (N = 76–104). While approximately 40% (N = 121–124) of Swedish and British headlines were neutral/mixed, a comparable percentage in Italy was negative. While 25–28% (N = 75–85) of Swedish and British headlines were negative, a comparable percentage (N = 59) in Italy was neutral/mixed. A multinomial logistic regression (see Table 4) indicated that country significantly predicted valence, χ2 (4) = 23.70, p < .001. Italian headlines were 2.60 times more likely than Swedish headlines and 2.35 times more likely than British headlines to have negative rather than neutral/mixed valence; the difference between Swedish and British headlines was not significant. Italian headlines were 1.76 times more likely than British headlines to have positive rather than neutral/mixed valence; the difference between Swedish and Italian headlines was not significant, as well as the difference between Swedish and British headlines.

Fig. 2
figure 2

Percentage distribution of valence in headlines across countries

Full size image

Regarding roles (Fig. 3), almost half (49.3%, N = 148) of Swedish headlines featured TGD people in at least one of the coded roles; the percentage was up in British headlines (62.0%, N = 186) and in Italian headlines (85.7%, N = 197). The criminal role was present in less than 5% (N = 1–13) of headlines (highest frequency in the UK, lowest frequency in Sweden), and was therefore not analyzed further.

Fig. 3
figure 3

Percentage distribution of roles in headlines across countries

Full size image

The role of victim was present in almost half of the Italian headlines (48.3%, N = 111), followed by British headlines (25.2%, N = 76) and Swedish headlines (16.6%, N = 50). There were also differences in how often the target was described as a victim of violence or discrimination. In Sweden and UK, being a victim of violence was less common than being a victim of discrimination, while in Italy violence was more common than discrimination. A multinomial logistic regression indicated that country significantly predicted victim role, χ2 (4) = 94.70, p < .001. Italian headlines were 2.97 times more likely and British headlines were 2.04 times more likely than Swedish headlines to represent TGD people as victim of discrimination; the difference between Italian and British headlines was not significant. Italian headlines were 8.27 times more likely than Swedish headlines and 8.47 times more likely than British headlines to represent TGD people as victims of violence; the difference between Swedish and British headlines was not significant.

The role of pioneer was present more often in Italian headlines (43.1%, N = 99) than British and Swedish headlines (25.6–26.7%, N = 77–79). The largest difference is the high number of TGD people being first in Italian headlines (30.0%, N = 69) as compared to British headlines (14.0%, N = 41) and Swedish headlines (11.4%, N = 32). The other two pioneer codes (coming out and gaining rights) were more equally distributed among the three countries, averaging around 5–7% (N = 14–24). A multinomial logistic regression indicated that country significantly predicted pioneer role, χ2 (6) = 37.70, p < .001. Italian headlines were 3.64 times more likely than Swedish headlines and 2.86 times more likely than British headlines to represent TGD people as pioneers by being first; the difference between Swedish and British headlines was not statistically significant. No statistically significant differences were found across countries for pioneers by coming out and pioneers by gaining rights.

The role of professional was present in 21–33% (N = 62–99) of headlines, with a peak in the UK. In Swedish and British headlines, the most frequent professional field was sports, while in Italy it was showbusiness. A binomial logistic regression indicated that country significantly predicted professional role, χ2 (2) = 11.00, p = .004. British headlines were 1.85 times more likely than Swedish headlines to represent TGD people as professionals. The difference between British and Italian headlines was not significant, as well as the difference between Swedish and Italian headlines.

Regarding gender aspects (Fig. 4), more than half (58.2%, N = 134) of Italian headlines mentioned it explicitly, while approximately 33–36% (N = 101–107) of Swedish and British headlines did. All countries represented trans women far more frequently than trans men and gender diverse people (19–46%, N = 59–106). While Swedish and British headlines represented an equally distributed about of trans men and gender diverse people (around 7%, N = 20–22, in Sweden and 3%, N = 9–10, in the UK), almost 12% (N = 27) of Italian headlines featured trans men and less than 0.5% (N = 1) featured a gender diverse person. While misgendering was present in less than 2.5% (N = 3–7) of Swedish and British headlines, it occurred in almost 9% (N = 20) of Italian headlines; still, since it was present in less than 5% of headlines across two countries, misgendering was not analyzed further. Out of misgendering occurrences, 26.7% (N = 8) included deadnaming, in reference to actor Elliot Page (7 in Italy, 1 in Sweden) and influencer Tone Sekelius (1 in Sweden). A multinomial logistic regression indicated that country significantly predicted target’s gender, χ2 (6) = 76.40, p < .001. Italian headlines were 2.54 times more likely than Swedish headlines and 5.43 times more likely than British headlines to represent trans men; the difference between Swedish and British headlines was not significant. Italian headlines were 3.72 times more likely than Swedish headlines and 2.42 times more likely than British headlines to represent trans women; in turn, British headlines were 1.54 times more likely than Swedish ones to represent trans women.

Fig. 4
figure 4

Percentage distribution of gender aspects in headlines across countries

Full size image

Regarding body aspects (Fig. 5), medicalization was present in almost 30% (N = 82) of Swedish headlines, while it was down to 7.6% (N = 23) in British ones and to 2.2% (N = 5) in Italian ones. Objectification, most frequent in Italy, was present in less than 5% of headlines, and was not included in further analyses. A binomial logistic regression indicated that country significantly predicted medicalization, χ2 (4) = 90.10, p < .001). Swedish headlines were 16.67 times more likely than Italian headlines and 4.55 times more likely than British ones to represent medicalization; in turn, British headlines were 3.70 times more likely than Italian headlines to represent medicalization.

Fig. 5
figure 5

Percentage distribution of body aspects in headlines across countries

Full size image

Regarding focus (Fig. 6), Italian headlines specified an individual or group focus more often than the other countries (93.0%, N = 214, versus 70–80%, N = 210–239). While all countries represented TGD people as individuals more often than as a group, approximately 43–47% (N = 131–141) of Swedish and British did; in Italy the percentage was up to almost 70 (N = 159). Swedish headlines had the highest occurrences of TGD people represented as a group (32.7%, N = 98). A multinomial logistic regression indicated that country significantly predicted focus, χ2 (4) = 60.60, p < .001. Italian headlines were 6.83 times more likely than British headlines and 4.30 times more likely than Swedish headlines to represent TGD people as individuals; in turn, Swedish headlines were 1.59 times more likely than British headlines to represent TGD people as individuals. Italian headlines were 3.92 times more likely than British headlines and 2.14 times more likely than Swedish headlines to represent TGD people as a group; in turn, Swedish headlines were 1.82 times more likely than British headlines to represent TGD people as a group.

Fig. 6
figure 6

Percentage Distribution of Focus in Headlines Across Countries

Full size image

Discussion

The present study examined the correspondence of media representations of TGD people, which may (re)produce sociocultural contexts, and legal protection and social acceptance of gender minorities in the European countries Sweden (high levels), the UK (medium levels), and Italy (lower levels). It is the first known study that compared TGD media representations between countries. We expected progressive representations of TGD people—meaning containing more positive valence, more positive roles (i.e., pioneer, professional), an unbiased representation of gender identities, less misgendering, less focus on bodies, and a focus on both individuals and groups—to be more frequent in Sweden than in the UK and Italy, and more frequent in the UK than in Italy.

Italian headlines were more likely than Swedish and British ones to have negative valence, but surprisingly also more likely than British ones to have positive valence. Thus, low levels of negative valence did not equate high levels of positive one. Both Sweden and the UK featured a majority of neutral/mixed valence. It seems that the more gender-egalitarian countries are reaching the ‘respect phase’ (Clark, 1969): a combination of positive, negative, and neutral representations that may indeed provide a nuanced and accurate reflection of the diversity of TGD experiences and issues. Italy, on the other hand, showed polarized headlines that represented most TGD people negatively, and a few positively.

Based on past literature, we coded four roles in which TGD people are often represented: criminal (e.g., Baker, 2014; Mogul et al., 2011); victim (e.g., Baker, 2014; Billard, 2016; Wood et al., 2019); pioneer (e.g., Baker, 2014; Wood et al., 2019); professional (e.g., Baker, 2014). In contrast to past findings (e.g., Åkerlund, 2019; Baker, 2014), representations of TGD people as criminals were almost absent from online news headlines in all three countries. In the present sample, it seems that TGD media representations more often include a victim rather than a criminal perspective. Almost half of Italian headlines represented TGD people as victims and more often as victims of violence, as can be expected from the country with the highest rate of transphobic violence in Europe (Trans Murder Monitoring, 2021). Headlines reporting discrimination were more equally distributed between countries. Both types of victimhood reflect the vulnerable position that many TGD people find themselves in; still, it is important to remember that some negative media representations are necessary for coverage of salient news stories and can help raise awareness of hardships endured by marginalized groups. For example, if a TGD person is victim of a transphobic crime, reporting it can signal that hate crimes are unacceptable (see the Nordic paradox of violence against women; Gracia & Merlo, 2016). At the same time, however, it could also decrease perceptions of TGD people’s agency and reinforce their associations with discrimination and violence.

While occurrences of coming out and gaining rights did not differ across countries, Italy was far more likely to represent TGD people being the first in a specific situation (e.g., to hold a public office, to be featured on television). A possible explanation for this is that Italy is now covering many ‘first times’ and mundane news stories (e.g., news of the first trans bride in the country) that have already happened or that would not warrant news coverage in countries with higher levels of inclusion of TGD people.

Describing TGD people as professionals was expected to be more common in countries that warrant them more legal protection in the workplace, preventing them from having to seek out money from underground economy (e.g., sex work, drug dealing). British headlines were significantly more likely than Swedish ones to represent TGD people as professionals. The more common professional fields were sports and showbusiness, likely influenced by current newsworthy events (e.g., the 2021 Olympic Games, celebrities such as Elliot Page, Tommy Dorfman, and Ezra Furman coming out). Positive representations of TGD people as professionals can also be beneficial for TGD people themselves—for example, by providing more real-life TGD role models to look up to (Bryant & Zimmerman, 2003). Having a role model can help minority youths develop a positive sense of identity (Yancey et al., 2002); however, transgender teenagers are less likely than LGB teenagers to have any role models because of the lack of TGD representation in mainstream media (Bird et al., 2012). Thus, seeing successful TGD professionals can be a source of hope for TGD youth.

Despite equal numbers of trans men and trans women in society, trans women were featured at least twice as often as trans men across all countries, in line with past literature (Åkerlund, 2019; Billard, 2016; Capuzza, 2014). This might be a consequence of biological essentialist beliefs that see trans women as men, and their transgressions of gender norms as more newsworthy (Schilt & Westbrook, 2009; Sirin et al., 2004); indeed, this would be in line with cisgender men dominating the news compared to cisgender women (Gustafsson Sendén et al., 2015b). Swedish headlines featured more gender diverse people than the other countries; this co-occurs with the recent inclusion of the gender-neutral pronoun hen in the Swedish dictionary (Gustafsson Sendén et al., 2015a), with increased media salience of both non-binary pronouns and non-binary people (Vergoossen et al., 2020). While Italian headlines featured the lowest frequency of gender diverse people, they had the highest of trans men. This could be connected to the large media coverage given to a trans men who was victim of a violent hate crime that took place during the timeframe the headlines were collected in (Valia, 2020).

As expected, misgendering was less frequent in Sweden and the UK than in Italy. Misgendering might be associated both with legal, social, and cultural context, as well as with grammatical aspects of the language. While Swedish and English are natural gender languages, Italian is a grammatical gendered language; this means that both personal and inanimate nouns are grammatically coded for gender (Gygax et al., 2019), increasing the chances of misgendering and deadnaming. Indeed—unlike in natural gender languages—all nouns, adjectives and verbs referring to a person need to be adjusted to their gender identity, not only the pronouns. Regarding deadnaming, from a journalistic perspective, using a public figure’s former name could be justified in the name of making the headline easily understandable by a larger audience. However, there are many alternative strategies available in such situations to avoid misgendering a TGD person (e.g., Armelli, 2020; Yurinova, 2023).

Similarly to the criminal role, and in contrast with previous literature (e.g., Billard, 2016; Mogul et al., 2011), representations of TGD people as sexualized were almost absent from the current dataset. Only in Italy objectification was present in more than 5% of headlines; given that (sexual) objectification leads to fetishization and, consequently, dehumanization, a connection could be hypothesized between high frequencies of representations of TGD people as victims of violence and as objectified. As TGD people are already at a greater risk of experiencing sexual harassment and violence than cisgender people (European Union Agency for Fundamental Rights, 2014), it is vital to ensure that media representations do not fuel these forms of discrimination further. Still, the low occurrences of objectification in Sweden and the UK can be interpreted as an improvement in TGD media representations when compared to previous research (e.g., Åkerlund, 2019; Capuzza & Spencer, 2017).

Regarding the individual or group focus, all countries represented TGD people as individuals more often than as a group. Moreover, Italy represented TGD people as either individuals or groups in the vast majority of headlines, while the percentage was lower for Sweden and the UK. This could be due to the high level of personalization that news generally has (Capuzza & Spencer, 2017), and show that Italian headlines refer to TGD people (e.g., “a non-binary celebrity”, “trans people in the army”) more often than to TGD issues (e.g., “trans rights”, “non-binary issues”). While parasocial contact with specific individuals can be effective in reducing prejudice and improving attitudes, provide role models for the youth, and long-lastingly impact attitudes (Bryant & Zimmerman, 2003; Ortiz & Harwood, 2007), depicting TGD people in a highly personalized way only may lead the audience to not perceive them as representative of the minority they are part of, dispelling the positive effects of parasocial contact. As such, representing TGD people as a group can reinforce their perceived agency and cohesiveness, but also trigger negative dynamics of intergroup conflict (Knowles et al., 2022).

Overall, the largest differences in news media representations of TGD people were found between Italy (low levels of legal protection and social acceptance of TGD people) compared with Sweden (high levels) and the UK (medium levels), and across the coding dimensions valence, victim role, target’s gender, medicalization, and focus. British and Swedish representations appeared more similar, and presented the largest differences across the coding dimensions professional role, medicalization, and focus.

Limitations and Future Research Directions

We identified the present study’s limitations and questions that future research should address. First, the time frames of the headines did not overlap, with time-specific events (e.g., the 2021 Olympic games) influencing the headlines’ content. To control for these differences, we ran additional analyses dividing the headlines by year of publication (see Supplementary Information, https://su.drive.sunet.se/index.php/s/x27RRBznLK2xi6N). Future research could benefit from not only examining TGD media representations within or between countries, but within and between countries.

Another potential limitation is the linguistic and discursive limitations that come with analyzing headlines only. For example, due to length constraints, headlines are less likely to feature descriptive words that enrich the content, such as adjectives and adverbs. This likely reduced the occurrence of TGD-specific discursive strategies that have been found in past research, such as ‘wrong body discourse’ and ‘containment of gender nonconformity’ (e.g., Åkerlund, 2019; Barker-Plummer, 2013). It also likely reduces chances of misgendering, which can increase with a repeated use of pronouns and first names. The overall picture of media representations of TGD people could be enriched by future studies investigating full-text articles across countries. However, in line with our research question on how legislation and cultural values might affect media discourse, the focus on headlines was deemed appropriate as it allowed for quantitative comparisons of the headlines and investigated a power tool of attitude formation.

The implications of explicit coding should also be noted: for celebrities and public figures, we only coded target’s gender and profession if explicitly stated. Thus, for readers with knowledge about the person in question, more background information about their genders, professions, and roles are available. Likewise, assuming an individual’s gender identity from their name is complex; for example, headlines reporting “Tommy Dorfman comes out as a trans woman” were not misgendering the actress, since she decided to keep her (typically masculine) birth name (Lockhart et al., 2023). In fact, the present study made us aware of how people’s gender is often implicitly—and incorrectly—assumed from first names. While sex/gender is still the primary determinant of first name choices in Western countries (Pilcher, 2017), we conclude that names are not a reliable proxy to ascribe gender, for both cisgender and TGD people. This limitation could also be overcome by analyzing full-text articles, which usually provide more details about the targets’ gender and profession.

Future studies should expand their reach to other regions of the world and more gender diverse samples, for example multi-gender societies such as the Hawaiian Mahu and the Swahili Mashoga (Nanda, 2014). Moreover, comparing languages with different gender systems (Gygax et al., 2019) presented some issues: for example, Swedish uses the same word for ‘sex’ and ‘gender’ (both kön). The higher presence of issues of medicalization in Swedish headlines could be an artifact of the search words we used: the Swedish word for transition is könsbekräftande, which is more often used in medical context and therefore differs from the English transition and the Italian transizione (see Table 1). Also, since Italian is a grammatical gender language, the target’s gender is made explicit more often than in Swedish and English; this also increases the chances of misgendering, which are much higher in Italian headlines. Future research could compare languages with the same gender systems to control for differences attributable to different linguistic structures.

Practice Implications

This research has practical implications for future research, policy makers, and minority communities. Our findings demonstrate that media representations of TGD people and a country’s social, cultural, and legislative context are intertwined. By presenting a cross-cultural comparison that goes beyond the Anglo-Saxon context, the current study has practical implications for research. The coding scheme we created based on previous literature and adapted to the current panorama of TGD representations could be employed by other researchers to carry out analyses in more diverse contexts, and by media policy officers to expand media guidelines on how to respectfully write about TGD people. Additionally, the current study informs community outreach by highlighting areas of negative representations that could be the target of interventions to improve media representations of TGD people. Although gaining visibility for minority groups is crucial (e.g., Baker, 2014; Bryant & Zimmerman, 2003), the negative media representations of TGD people found in the present study highlight the need for TGD-supportive media policies and guidelines. While such guidelines exist in some contexts such as journalism (e.g., Armelli, 2020, in Italian), media studies (e.g., RFSL, 2021, in Swedish), and psychology (e.g., American Psychological Association, 2015, in English), there is still room for improvement, particularly in countries with lower levels of legal protection and social acceptance of TGD people. It is also important to notice that the practical applications of these findings go beyond the TGD community and can also inform guidelines on gender-fair language and unbiased journalism about inclusive communication strategies.

Conclusion

The present study revealed correspondences between the way TGD people are represented in news media headlines and a country’s levels of legal protection and social acceptance of gender minorities. Overall, we found that countries with higher levels of legal protection and social acceptance featured more progressive representations of TGD people (more neutral than negative valence, less representations of victims of discrimination and violence, more representations of gender diverse people, less misgendering, and less objectification). These correspondences indicate that media play a role in (re)producing the societal contexts TGD people live in, highlighting the importance of interventions aimed at improving media representations of gender minorities.