Introduction

The audiovisual plane is a “hegemonic space” (term used by ^{García Canclini (1984)} according to the approach by Gramsci and Bourdieu) that is projected and influences the social behavior of audiences by reproducing pre-existing models or, conversely, by transforming them. How different problems today are addressed is therefore a never-ending circle: predominant ideas are represented, and these models and attitudes are continuously perpetuated by reproducing them on the screen (^{Binimelis, 2016, p. 10}). Someone who has never met a LGTBI+ person can thus use their own imagination to construct what being a part of the LGTBI+ community means just by what they have perceived on the screen (^{Marcos Ramos & González de Garay, 2019, p. 1}). It is important to focus on the image offered by this type of media in order to analyze and observe possible implications in society (^{van Dijk, 1997}, ²⁰⁰³).

This study analyzes Spanish fiction series on VOD platforms in order to contribute to the scarce literature on LBTI+ women in the media, as well as increase knowledge on a growing, predominant form of audiovisual consumption. The aim is to ascertain whether female LBTI+ characters share the stereotypical features of how of cis-hetero women are represented (under-representation, predominance of personal/ private sphere over work/public, etc.). This first requires a review of the main academic publications on the subject of study, presenting a content analysis of a sample of 749 studies, and finally conclusions and a discussion of prior literature, concluding that, although there has been some progress, LBTI+ characters continue to suffer bias.

Theoretical framework

Television series have taken on a significance previously attributed to film productions as they have become the most widely-consumed audiovisual products (^{Bermejo & Núñez, 2008}; ^{Moreno & Rodríguez, 2016}) thanks to the expansion of the Internet, ICT development, new information technologies and, above all, the recent rise in video on demand platforms (VOD). How LTBI+ women are represented in these productions is therefore an important issue as the image they shape and broadcast has a series of repercussions or effects on consumer attitudes towards this part of the collective in real life (^{Bond & Compton, 2015}; ^{Bonds-Raacke et al., 2007}; ^{Calzo & Ward, 2009}; ^{Moroni, 2014}; ^{Schiappa et al., 2006}). These representations also affect members of the LGTBI+ collective, especially younger members, by promoting models and feelings related to belonging to the community (^{Bond, 2015}; ^{Gomillion & Giuliano, 2011}; ^{Meyer, 2003}).

However, despite information available based on this research, the representation of LGTBI+ characters is still an unexplored field, especially regarding LTBI+ women due to their notable under-representation, compared to male members of the collective. Several studies continue to call for new research related to obtaining data on attitudes towards fictional LGTBI+ people and characters (^{Calzo & Ward, 2009}; ^{Gomillion & Giuliano, 2011}; ^{McLaughlin & Rodriguez, 2016}; ^{Madžarević & Soto-Sanfiel, 2018}, ²⁰¹⁹).

A report by the ^{USC Annenberg Inclusion Initiative (2020)}, after analyzing the 100 most popular films in the US in 2019, found a total of 1.4% (N=61) of characters belonged to the LGBT+ collective -of which 10 were lesbians, 45 gay, 3 bisexual and 3 transgender-, but more than half of the roles (N = 32) were classified as “insignificant” to the plot (2020, p. 5). The study also reveals that around 80% of these LGTBI+ characters are male. US television has also been analyzed for this same purpose, finding that trans men and gender queer characters are still invisible (^{Capuzza & Spencer, 2017}), trans characters continue to be stereotyped (^{Sandercock, 2015}) and anchored to plots focusing on their gender identity (^{Reitz, 2017}), and gay couples reproduce heteronormative domestic patterns such as stability or exclusivity (^{Dhaenens, 2012}).

Research on LGBTI+ representation in Spain is scarce. However, there is a solid corpus of literature on the social construction of women in Spanish television series. For example, ^{Galán (2007)}, ^{García-Muñoz et al. (2012)}, ^{Lacalle and Gómez (2016)}, ^{Chicharro Merayo (2018)} or ^{Hidalgo-Marí (2017)}, note a paradigm shift in how women are represented in series towards a professional, yet still stereotyped, woman. Other research studies focus on a specific period, such as ^{González de Garay and Alfeo (2017)}, who analyze the image portrayed during the Spanish Franco dictatorship, from which we have inherited a very stereotypical portrait full of concepts such as guilt, marginalization or ridicule.

This research studies Spanish subscription streaming fiction series given that previous works by researchers González de Garay and Marcos Ramos verified less representation of LGTBI+ characters on freeview tv compared to platforms (^{Marcos Ramos et al., 2019}, ²⁰²⁰; ^{Marcos Ramos & González de Garay, 2021}). This, therefore, suggests a need to focus on products offered by video on demand services in Spain as the new, growing form of consuming content that has to date not been the subject of much study in the academic field.

Methodology

This study used a quantitative social research technique known as content analysis, which “consists of special procedures for processing scientific data” (^{Krippendorff, 1990, p. 28}) in order to quantify data obtained and provide objective conclusions based on quantifiable results capable of representing real phenomena. According to ^{Igartua (2006)}, this method allows us to “scientifically approach the analysis of messages (whatever their nature), understand their genesis, obtain precise descriptions of their structure and components” (p. 180). Using content analysis as a research technique is therefore highly useful for Social Science studies as this enables “formulating, based on certain data, reproducible and valid inferences that can be applied to its context” (^{Krippendorff, 1990, p. 28}).

Sample

The sample comprised LTBI+ woman found in Spanish series broadcast on video on demand platforms (VOD) -Netflix, HBO, Movistar, ATRESplayer Premium, Disney+ and Amazon Prime Video- between 2020 and the first quarter of 2021 (up to April 2^nd) in order to determine whether they present an unequal image with regard to their gender (men vs women), their orientation (heterosexual vs homosexual) or gender identity (cis vs trans or non-binary). The characters were the basic unit of analysis and were only considered if they met two requirements: they must appear visually and have at least one phrase of dialogue with another character on screen (^{Koeman et al., 2007}). A total sample of 38 series premiered between 2020 and the first quarter of 2021 was obtained (Table 1). These were then coded and a random episode of each series was selected for coding, given that they usually include a fixed cast of main characters.

An analysis of the sample detected a total of 749 characters, 39 of which were LGTBI+. The sample was reduced to 9 homosexual female characters and 10 transgender women (Table 2) as this study focuses on analyzing LTBI+ women.

Hypothesis and research questions

The code book has been tested in several previous studies (^{Barrios et al., 2021}; ^{Marcos Ramos et al., 2020}; ^{Marcos Ramos & González de Garay, 2021}) and was adapted to the objectives of this study by basing analysis variables on the type of character, narrative level and their social sphere. Variables were grouped into ten major blocks: basic identification details, general details of the character, narrative level of the character, social sphere of the character, violent behavior by the character, victim of violent behavior, health behaviors demonstrated by the character, topics of conversation, personality traits, and functional diversity traits. After reviewing previous studies and defining the research objectives, the following hypothesis and research questions were proposed:

Results

As previously mentioned, in the 38 series of the sample, 749 characters were identified: 59.8% (N=447) were coded as men and 40.2% (N=301) as women -38.9% (N=291) cis-gender women while 1.3% (N=10) were transgender- with no characters identified outside the male/female binomial. Firstly, these data reveal that fiction series appear to follow traditional standardization in terms of gender identity with a lack of non-binary and male transgender characters. Although the appearance of transgender women in these stories is noteworthy, their presence is also very localized as most appear in one series: Veneno (ATRESplayer, 2020). Secondly, there is a clear under-representation of female characters in general as the percentage is far from the 50.9% (N=24 144 815) of women among the Spanish population (^{Instituto Nacional de Estadística [INE], 2021}), a difference of 10.7%.

While 41.3% (N=309) of the characters were identified as heterosexual, 54.7% (N=410) could not be associated with any type of sexual, orientation; 3.7% (N=28) were homosexual, 0.1% (N=1) plurisexual and the remaining 0.1% (N=1) were classified with another type of sexual orientation. By observing these data along with the data mentioned in relation to transgender representation, we can conclude that sexual orientations other than heterosexual -89.58% heterosexual characters and 10.42% non-heterosexual- are slightly under-represented. The same is true of non-binary or male transgender identities. The LGTBI+ community is therefore slightly under-represented as an estimated 11% of the Spanish population identify as non cis-heterosexual (^{Barómetro Control, 2017}). Furthermore, according to the ^{Ipsos survey (2021)}, worldwide 80% of people identify as heterosexual; 3% as gay, lesbian or homosexual; 4% as bisexual; 1% as pansexual or omnisexual; 1% as asexual; 1% as “other”; and finally, 11% don’t know or won’t say. Add to this that 1% of adults identify as “transgender”, “non-binary/non-conforming/gender-fluid” and “in another way rather than as male or female”.

There is therefore a greater representation of GTBI+ men who account for 67.9% (N=19) of homosexual characters and 100% in the plurisexual (N=1) and other sexual orientations (N=1) categories, while LTBI+ only represent 32.1% (N=9) of homosexuals and are not represented in the other categories. Considering gender identity, 100% (N=10) of transgender characters are female and, therefore, there are no trans male characters in the sample.

In relation to RQ2, statistically significant differences were found between gender and job variables: χ2 (32, N=561) = 142.712, p<0.000. Interesting issues can thus be observed. For example, jobs involving the highest academic and professional training (senior management) are mostly male, with 10.1% of men compared to 8.3% of women. On the other hand, based on the middle management category, there is a significant difference as this field is led by women with 30.2%. This, together with the fact that women hold more office jobs (3.4% compared to 0.9%), unqualified jobs (7.8% compared to 4.9%) and are the only ones to carry out unpaid domestic work (1.5%), highlights the effect known as the “glass ceiling”.

The fact that women have lower status jobs than men is not because the latter have higher education, as there are more female student characters than male (13.7% compared to 5.2%). These data are not far from reality as in 2018, the percentage of women graduating higher education was 53.6% compared to 46.4% of men (INE, 2019). Even so, women are paid 9% less than men in Spain, even if they have the same qualifications (^{Ministerio de Educación y Formación Profesional, 2020}). In 2019, 76.3% of working women were employees -with bosses but no subordinates- while on 0.4% were executives at large or medium-sized enterprises (^{INE, 2020}). Therefore, TV series are a reflection of the traditional stereotypes associated with the role of men and women in the workplace.

Job inequality among male and female characters is quite visible, but a more detailed study of female LTBI+ characters shows that this reality is even more noticeable: despite a representation of 18.8% (N=3) in occupations such as student or 12.5% (N=2) in retail, the “others” category has more female LTBI+ character representation (37.5%, N=7), mostly trans characters who work in prostitution. The few LTBI+ women portrayed therefore have lower-classed or lower-qualified jobs with practically no characters, or a very small minority, in higher-qualified jobs

Another statistically significant difference shows that there are more women (61% compared to 48.7% of men) who pursue personal objectives compared to men (χ2 (4, N=579) = 11.946, p<0.018). The results also show a greater presence of men (44% of women compared to 54.7% of men) pursuing work objectives as opposed to women (χ2 (2, N=580) = 6.203, p<0.045). Women are therefore still more associated with the personal sphere and men with work, perpetuating the image of the woman in charge of the home and family (^{Lacalle & Sánchez, 2015}; ^{Marcos Ramos et al., 2019}). This situation is more pronounced in LTBI+ women: 63.6% pursue no employment objectives, but 90.9% do pursue personal objectives (8.4% and 30.9% more than cis-heterosexual women, respectively). The percentage is observed to increase when compared to cis-heterosexual women, which could be related to an attempt to integrate their orientation or identity into cannons commonly associated with women (^{González de Garay, 2013}).

An analysis of the topics of conversation attributed to each group of characters yields the following results. Cis-heterosexual women talk more about three topics: health (χ2 [3, N=748] = 26.835, p<0.000) with 28.1%; education (χ2 [3, N=748] = 9.607, p<0.022) with 17.1% compared to 9.8% of men; and family (χ2 [3, N=748] = 16.020, p<0.001), with 50.2%, unlike 21% of men. The topics of conversation statistically related with cis-heterosexual men are: violence (χ2 [3, N=748] = 10.295, p<0.016) with 32.2% compared to 24.9% in women; and immigration (χ2 [3, N=748] = 9.607, p<0.017) 7.9% compared to 7.1% of women. Finally, in the case of LTBI+ women, there are four predominant topics of conversation: love (χ2 [6, N=748] = 67.173, p<0.000) with 45.6% compared to 37.7% of other women; friendship (χ2 [3, N=747] = 21.424, p<0.000) with 47.6%; sex (χ2 [3, N=748] = 10.546, p<0.014) with 28.6%; and, finally, empowerment (χ2 [3, N=748] = 14.664, p<0.002) with 14.3% compared to 10% of other women.

We can therefore conclude that the predominant topics among cis-heterosexual women are personal (health, education and family) as with LTBI+ women (love, friendship, sex and empowerment), while men mostly talk about non-personal topics (immigration and violence).

By comparing the topics of conversation of male GTBI+ characters and female LTBI+ characters the situation changes significantly. Except for empowerment, where women in the collective represent 14.3% compared to GTBI+ men with 0%, the other topics highlighted as predominant among female LTBI+ characters become characteristic of the male part of the collective: love is spoken about by 66.7% of men compared to 47.6% of women; friendship by 50% compared to 47.6%; and, finally, sex by 38.9% compared to 28.6% of female characters. This is in line with the study by ^{Fisher et al. (2007)} who detected, after a content analysis of 2 700 series episodes, that around 15% of the programs contained at least one case of sexual behavior or sex was discussed in relation to homosexuals, lesbians or bisexuals. ^{Netzley (2010)} reached a similar conclusion with her content analysis of 98 entertainment programs, which verified that homosexual characters on television were more likely to be shown in sexual situations than heterosexual characters, and women were more likely to be shown in sexual situations among people of the same sex than men. A richer and more varied conversation was also detected as they were also predominant in the topics of health (38.9%) and immigration (27.8%).

Conversely, no statistically significant differences were observed between the sexual orientation of characters and the type of role (in terms of narrative weight) they play in the series (χ2 [6, N=339] = 4.477, p<0.622). Furthermore, there were no leading roles played by plurisexual characters or other orientations. Plurisexual characters only appear in supporting roles (100%, N=1) and “others” only in background roles (100%, N=1).

On the other hand, statistically significant differences were found between gender identity and type of character (χ2 [4, N=747] = 14.168, p<0.007). Cis-gender women are predominant in leading roles (23%) ahead of cis-gender men (15%), who stand out in supporting (29.8% compared to 23.7% of cis-women), and background roles (55.2% compared to 53.3%). Regarding transgender identity, only female transgender stand out as no characters were coded as male transgender. Trans-female characters play background roles (90%) except in the series Veneno, where they take on leading roles. Therefore, we can conclude that fiction series continue to relegate transgender characters (female) to mainly background roles with barely any narrative weight in the plots, being limited to bringing a touch of “color” to the series.

To the question of whether female LTBI+ characters would tend to present or develop an attitude considered more as a “victim”, no statistically significant differences were found in most variables and, therefore, there is no relationship between LTBI+ women and being represented as victims of violent behaviors or attitudes. Even so, analyzing the items individually reveals that categories such as “suffers minor physical attacks” (χ2 [3, N=748] = 14.523, p<0.002) or “suffers hostile comments” (χ2 [3, N=748] = 7.884, p<0.048) show a high affirmative percentage compared to other categories such as male GTBI+ or female cis-heterosexual characters.

Finally, in relation to the last research question, three personality traits are statistically linked to female LTBI+ characters: friendly (χ2 [9, N=607] = 21.771, p<0.010); extroverted (χ2 [9, N=614] = 21.115, p<0.012); and good (χ2 [9, N=544] = 33.586, p<0.000). The predominant traits of female cis-heterosexual characters are: good (χ2 [9, N=544] = 33.586, p<0.000) with 31.1% compared to 18.8% of female LTBI+ characters; intelligent (χ2 [9, N=550] = 31.762, p<0.000), 38.2% with 0% among women from the collective; grateful (χ2 [9, N=478] = 46.420, p<0.000) in 19.7% compared to 0% of LTBI+ women; and finally, maternal (χ2 [9, N=518] = 34.751, p<0.000), 19.9% of cis-heterosexual women while the figure among LTBI+ women is 0%.

Male GTBI+ characters display the following statistically significant characteristics: friendly (χ2 [9, N=607] = 21.771, p<0.010) with 38.9%, a 16.7% difference compared to female LTBI+ characters; extroverted (χ2 [9, N=614] = 21.115, p<0.012) with 38.9%; and irresponsible (χ2 [9, N=514] = 18.207, p<0.033), 13.3% compared to 9.1% of LTBI+ women.

These results reveal that, while the most characteristic traits of female LTBI+ characters can be classed as positive, the fact that these characters do not represent the majority percentage in these categories must be taken into account. There is a lack of traits or characteristics that clearly define these characters so, despite being represented in the series, we must ask whether the image they project on the screen is appropriate, sufficient and, above all, complete so that fiction series have LTBI+ characters with a well-defined personality.

Discussion and conclusions

Having analyzed the results, women (cis and trans) are under-represented compared to sociodemographic indexes in Spain and the image portrayed is still quite stereotypical, with women who stay at home or have lower qualified jobs (^{Guash, 1991}; ^{Viñuales, 1999}) -although, according to the data available, female characters study more-, are more concerned about home and family and, consequently, their objectives and topics of conversation are more associated with the personal sphere than men who have more conversations related to work and the professional sphere. However, female characters have more leading roles with greater narrative weight than men, in line with the findings of previous studies (^{Marcos Ramos et al., 2019}, ²⁰²⁰; ^{Marcos Ramos & González de Garay, 2021}).

According to the study, traits traditionally associated with female characters are accentuated in how LTBI+ women are portrayed, with less representation in audiovisual fiction than GTBI+ men, and who predominately talk about love, friendship, sex and empowerment.

Overall, non-heterosexual characters are slightly under-represented in these series. Despite the percentage of “non-heterosexual” orientations increasing or, at least, equaling previous studies (^{Marcos Ramos et al., 2019}, ²⁰²⁰), no transgender men or non-binary individuals are portrayed and the presence of plurisexual characters is still scarce. Nevertheless, the figures do appear to be much closer to the demographic figures of non-heterosexual population and also seem to be even more optimistic and promising when compared with freeview television (^{Marcos Ramos et al., 2019}), as there is an increase in the number of non cis-heterosexual characters overall (^{Marcos Ramos et al., 2020}). However, by putting the data into perspective, we can see that the series Veneno has had a major influence in the rise in LTBI+ characters and their variety (including several trans women); and also the importance of #Luimelia (ATRESplayer, 2020-), a lesbian spin-off of Amar es para siempre (Antena 3, 2013-), produced thanks to pressure on social media by the couple’s fandom (^{Acosta, 2020}). These characters are represented in some key series and, therefore, inclusion has not reached all the series analyzed.

To conclude this study, we must highlight a trend towards gender equality and also a greater (and better) representation of LTBI+ women in Spanish fiction series even though some aspects still prevent true equivalence. For example, clear under-representation is still a problem: themes such as family, care-giving and the home are still strongly associated with women, and more highly-qualified jobs are closely related with men, perpetuating certain stereotypes linked to the masculinization of qualified jobs with greater status.

We must recall that audiovisual fiction “can become a vehicle for education and entertainment for training in equality” (^{Padilla Castillo & Sosa Sánchez, 2018, pp. 91-92}) in order to promote a reduction in stereotypes associated with LTBI+ women with female characters portraying behaviors and characteristics much closer to reality given that “viewers will imitate the representations of the world offered in their favorite series and will use the archetypes, values and how we relate to others” (^{Padilla Castillo & Sosa Sánchez, 2018, pp. 91-92}). Social cognitive theory (^{Bandura, 1986}) indicates that television influences viewers by providing vicarious experiences on which to model beliefs, attitudes and behaviors with real life experiences are more limited. For this reason, it is important to provide the viewer who in real life will not have contact with people from this group with a true, non-stereotypical image. Special attention must also be paid to formats that target young people as fiction, according to ^{Mateos-Pérez, (2021)} “tries to respond through its characters and their values -friendship, tolerance, generosity- in an inclusive way, attempting to make the viewer aware of the problems and offering them resources to address and overcome them” (p. 151).

The purpose of this study is to complete previous research on the topic that has traditionally been more qualitative. It therefore aims to provide quantitative data in order to analyze how LTBI+ women are represented in Spanish original subscription streaming TV series. Despite its limited sample, this paper aims to be the source of future research, both quantitative and qualitative.