Attitudinal Shifts towards Homosexuality among Secondary School Students in Hong Kong over a Decade: A Multiple Group Latent Class Analysis

In recent decades, favorable attitudes toward homosexuality have been on the rise in many parts of the world (Roberts, 2019). Despite this progress, sexual minority groups continue to experience persistent social stigma, as well as discriminatory laws, policies, and practices (Badgett & Crehan, 2016). These challenges contribute to elevated physical and mental health issues within these populations (National Academies of Sciences, Engineering, and Medicine, 2020; Russell & Fish, 2016).

Public attitudes toward homosexuality are complex and multifaceted (Adamczyk & Liao, 2019; Adolfsen et al., 2010), with existing research addressing a broad range of related topics. These include public displays of affection (Doan et al., 2014), civil rights and liberties (Lewis, 2003; Loftus, 2001), same-sex marriage (Baunach, 2011, 2012), adoption (Rye & Meaney, 2010), and biological basis of homosexuality (Lewis, 2009; Whitehead & Baker, 2012). However, it is crucial to recognize that public attitudes can vary substantially across these different dimensions. Between 1973 and 1998, Americans became increasingly supportive of gay and lesbian civil liberties, however, it was not until the 1990s that their views regarding the morality of homosexuality became more liberal (Loftus, 2001). Another study found inconsistent attitudes toward formal rights (e.g., partnership benefits) and informal privileges (e.g., public displays of affection) of same-sex couples, with variations depending on the respondents’ sexual orientation. Specifically, heterosexual respondents showed no preference for unmarried heterosexual couples over same-sex couples regarding formal rights but expressed less approval for informal privileges for same-sex couples; in contrast, lesbian and gay respondents exhibited a significant in-group preference for formal rights, but not for informal privileges (Doan et al., 2014). Additionally, previous studies have indicated differing attitudes toward gay men and lesbian women, with more unfavorable attitudes toward gay men (Einbinder et al., 2012; Herek, 2002; Stoever & Morera, 2007).

These findings have prompted growing interests in exploring population heterogeneity in attitudes toward homosexuality or sexual minority, though research in this area remains relatively limited. A few studies have examined population heterogeneity based on four dimensions of sexual orientation beliefs: 1) naturalness (e.g., “Sexual orientation is innate”), 2) discreteness (e.g., “Sexual orientation is a category with clear boundaries: a person is either gay/lesbian, bisexual, or heterosexual”), 3) entitativity (e.g., “Individuals with the same sexual orientation seem to be connected to one another by some invisible link”), and 4) personal and social importance of identity (e.g., “It is possible to know about many aspects of a person once you know her or his sexual orientation”) (Grzanka et al., 2016; Tierney et al., 2021). Latent profile analyses consistently identified three belief profiles: 1) Naturalness-only, 2) Multidimensional essentialism, and 3) High discreteness, entitativity, and importance (Grzanka et al., 2016; Tierney et al., 2021). Similarly, another study classified university students’ attitudes toward gays and lesbians into three attitudinal groups: unbiased, moderately biased, and highly biased (Rutledge et al., 2012). While these studies attempted to differentiate between various dimensions of attitudes and explore population heterogeneity, no research to date has investigated the potential temporal shift in heterogeneity of attitudes.

Various factors have played a role in shaping public attitudes toward homosexuality. Previous studies consistently reported a sex difference, with males typically holding more negative attitudes toward homosexuality than females (e.g., Donaldson et al., 2017). This pattern may be attributed to males’ tendency to hold more traditional gender beliefs (Kite & Whitley 1996; Verweij et al., 2008). Investigations into social changes have found that younger generations were generally more egalitarian (Donaldson et al., 2017; Lee & Mutz, 2019), which is believed to be a driving force behind shifts in public attitudes. Studies of adolescents and young adults found mixed results regarding the relationship between age and attitudes (Feng et al., 2012; Horn, 2006; Jenkins et al., 2009), which may be due to the differences in examined dimensions. Other factors that support favorable attitudes toward homosexuality include non-heterosexual orientation (Gulevich et al., 2023; Verweij et al., 2008), higher education levels (Ohlander et al., 2005; Takács et al., 2016; Van den Akker et al., 2013), and lower levels of religiosity (Dotti Sani & Quaranta, 2020; Jäckle & Wenzelburger, 2015; Van den Akker et al., 2013). In addition to socio-demographic variables, studies indicated that increased contacts with sexual minority groups (Lewis, 2011; Skipworth et al., 2010), and greater sexual knowledge (Feng et al., 2012) can help mitigate negative stereotypes of homosexuality.

The cultural context of Chinese society, deeply influenced by Confucian values, is markedly distinct from that of Western societies. It is characterized by a paternalistic culture that emphasizes deference to hierarchical authority, places significant importance on family bonds, and prioritizes the needs of the community over individual interests (Adamczyk & Cheng, 2015). These values have further reinforced consensus and order in the society (Adamczyk & Cheng, 2015). A study based on four waves of Chinese General Social Survey (CGSS) found that while the percentage of people disapproving of same-sex sexual behaviors decreased by 5% during 2010–2017, a vast majority (79.9%) still held such views in 2017 (Lin & Wang, 2021). Several studies have highlighted the profound influence of cultural factors specific to Chinese society, such as filial piety, on individuals’ attitudes toward homosexuality (Feng et al., 2012; Lin, 2023; Lin et al., 2016). These findings underscore the importance of adopting culturally sensitive approaches when developing strategies to promote inclusion of sexual minority groups within the community.

Hong Kong, a special administrative region of China since 1997, has a unique socio-cultural environment that reflects its Chinese heritage, while also being shaped by its history as a former British colony. In Hong Kong, faith-based organizations play a significant role in the operation of social services, with approximately half of primary and secondary schools having religious affiliations, predominantly Christian (Jackson, 2015). In other words, Hong Kong faces sexual prejudices toward homosexuality stemming from both Confucian values, similar to the Mainland (Kwok & Wu, 2015), and Western Christian values (Wang et al., 2009; Zhou, 2006). At present, Hong Kong’s anti-discrimination ordinances do not extend to protections for sexual orientation or gender identity. In educational settings, policies and initiatives to combat homophobic and transphobic bullying are either absent or insufficiently implemented (Kwok, 2023). Sexual minority groups in Hong Kong face discrimination across various domains of life, including employment, housing, interactions with teachers and peers, and even within their own families (Pang et al., 2022; Suen & Wong, 2017). Beyond overt discrimination, diverse forms of microaggressions also persist (Kwok & Kwok, 2022). For instance, heteronormative assumptions and omissions of content related to LGBTQ + issues were pervasive in sexuality education curricula, workshops, and programs in Hong Kong (Chan, 2022). Such microaggressions have been associated with an increased risk of mental health problems, substance use, and suicidal behaviors among sexual minority populations (DeSon & Andover, 2024).

Public attitudes toward homosexuality in Hong Kong have undergone progressive changes during the past decades. A 2006 survey of 2,024 Hong Kong residents revealed that over 40% of the respondents did not consider homosexuality to be “psychologically normal”, and half (51.4%) were unaccepting of having lesbian/gay family members (Home Affairs Bureau, The Government of the Hong Kong Special Administrative Region, 2006). However, data from a 2013 telephone survey showed a contrasting picture, with only 16% of the 410 respondents claimed that an individual’s sexual orientation would affect their acceptance of them, and 83% believed that parents should love their children regardless of their sexual orientation (Loper et al., 2014). A recent report in 2023 based on three waves of telephone surveys found that 60% of respondents supported same-sex marriage, which was a notable increase from 50.4% in 2017 and 38% in 2013 (Lau et al., 2023). Furthermore, 71% of respondents in 2023 believed Hong Kong should introduce laws to protect against sexual orientation-based discrimination, up from 69% in 2017 and 58% in 2013 (Lau et al., 2023).

In summary, the unique cultural characteristics of Chinese societies have shaped public attitudes toward homosexuality in ways that differ from those observed in Western societies. While trend analyses of these attitudes have been conducted in the Chinese settings, research specifically focusing on adolescents remains limited. This focus is crucial, as adolescents not only play a pivotal role in shaping peer relationships and the overall school climate through their attitudes but also serve as key drivers of social change and reliable indicators of a society’s future direction. Furthermore, existing studies in both Hong Kong and broader Chinese contexts have rarely considered the multidimensional nature of public attitudes toward homosexuality when examining temporal changes, nor have they explored the potential heterogeneity within the population. To fill these gaps, this study aims to achieve three objectives through a repeated cross-sectional survey conducted between 2011 and 2021: 1) to provide an updated understanding of current attitudes toward homosexuality among secondary school students in Hong Kong, while also tracking any temporal changes over a decade; 2) to discover potential heterogeneity among adolescents based on their responses to different dimensions of attitudes and examine how sex, age, parental responses to sexuality-related questions, and online sexual knowledge exposure are related to the heterogeneity; 3) to investigate whether and how these population heterogeneity had shifted between 2011 and 2021.

Drawing on a territory-wide survey, our study provides the first results on attitudes toward homosexuality among secondary school students in Hong Kong. It is also a pioneering effort within the Chinese context to examine both the population heterogeneity in these attitudes and their temporal changes. Between 2011 and 2021, our findings revealed a growing acceptance of homosexuality among Hong Kong adolescents. This aligns with the Global Acceptance Index (GAI) trend for Hong Kong which indicates increasing acceptance of sexual minority populations (Flores, 2021). Despite the increase in global average level of acceptance, the GAI for Asia has shown minimal change over the past decades (Flores, 2021). Moreover, global divides and polarization persist, with 56 out of 175 countries and regions experiencing increases in acceptance of sexual and gender minority groups between 1981 and 2020, while 57 reporting a decline (Flores, 2021). These patterns highlight the importance of region- and culture-specific analyses, such as our study.

Throughout the three waves of data, the respondents exhibited varying degrees of approval with respect to different topics of homosexuality. On one hand, almost half of secondary school students had expressed their support for same-sex marriage and same-sex parenting by 2011, which may be ascribed to the increasing representations of sexual minority groups in the media and the progress of LGBT + rights globally. On the other hand, it was only as recent as 2021, the acceptance of public displays of affection between people of the same sex reached comparable levels. Given that the majority of the respondents were heterosexual, a possible explanation for this inconsistency is that as the gradual elimination of institutional discrimination, social interaction may become a more important site where dominant groups maintain their subtle advantages and informal privileges over minority groups (Doan et al., 2014; Harris, 2001; Schwalbe et al., 2000). Additionally, we found that male homosexuality was less accepted than female homosexuality among secondary school students, consistent with previous studies (Bettinsoli et al., 2020; J. Wang et al., 2022). The discourse of heteronormativity acknowledges only two genders that align with biological sex and considers sexual attraction between opposite genders as the only acceptable norm (Kitzinger, 2005); hence, homosexuality is socially constructed as a violation of gender roles (Herek, 2002). Although both sexes face social pressure from gender non-conforming behaviors, pressure for males is usually more intense (Roberts et al., 2013; Sirin et al., 2004; Spivey et al., 2018; Vandello et al., 2008). Furthermore, males often have fewer socially acceptable options for emotional and physical expression compared to females, leading society to perceive displays of affection or physical contact between males as less appropriate or acceptable (Lim, 2002).

As the results of the LCAs indicated, the patterns of attitudes toward homosexuality among secondary school students have undergone temporal changes over the years. In 2011, three classes (intolerant, neutral, and inclusive) were identified. In both 2016 and 2021, a four-class model was optimal, consisting of three classes similar to those observed in 2011 and a new “partially inclusive” class. Notably, members of the inclusive class in 2011 were less likely to express support than those in the later years, primarily manifested as a lower level of tolerance toward public displays of affection between people of the same sex. In 2011, non-accepting attitudes toward public displays of affection were mainstream among respondents; hence, an inclusive class that consistently showed support across aspects of attitudes may have been too small in size to emerge as a representative class. Over time, with the attitudes toward homosexuality becoming more positive, a partially inclusive class emerged in 2016 and 2021 samples. The inconsistency reflected in this class highlighted the less positive attitudes toward public visibility of homosexuality. To better interpret the changes across the three waves, we built a multiple group LCA with partial measurement invariance. The findings from this analysis further demonstrated that the secondary school students were becoming more inclusive: a “consistent” inclusive class emerged in 2016, and almost half of the respondents in 2021 belong to this class, while the intolerant class shrank by 25% over the decade. Interestingly, the proportion of neutral class remained stable across the three waves, suggesting the presence of a subgroup of secondary school students who had either not yet developed an awareness of social issues or were simply uninterested in them.

The covariate analyses showed that females, older respondents, and non-heterosexual respondents had a higher probability in classes with more favorable attitudes. According to the theory of hegemonic masculinity, males employ various means to establish and legitimize their dominant positions by promoting hegemonic masculinity (Connell, 1993; Connell & Messerschmidt, 2005). Consequently, males are more prone to resist behaviors that deviate from traditional gender roles and sexual patterns to maintain their societal advantages (Chi & Hawk, 2016). Age also emerged as a significant predictor of the attitude patterns in our study. During adolescence, individuals embark on a journey toward greater autonomy and independence and their parents’ influence starts to wane. They are exposed to a wider range of information sources and their cognitive abilities continue to develop, leading to a better understanding of their surroundings and a constantly evolving worldview. As a result, older adolescents tend to exhibit a greater degree of acceptance toward homosexuality, relative to their younger peers. It is unsurprising that non-heterosexual respondents showed more favorable attitudes toward homosexuality. However, it is worth noting that a small proportion of this population was classified into latent classes with non-accepting attitudes. This phenomenon may be attributed to internalized sexual stigma, wherein sexual minority individuals adopt negative societal attitudes toward their community as a part of their own value system (Herek et al., 2009). Such self-directed stigma may increase risks of mental health problems among sexual minority groups (Berg et al., 2015; Xu et al., 2017). In comparisons between heterosexual adolescents and those uncertain about their sexual orientation, the latter exhibited less positive attitudes toward homosexuality in 2016 and 2021, with a tendency toward neutrality observed in 2011. These differences may be attributed to the fact that the attitude classes identified in 2011 differed from those generated in the subsequent survey waves. One possible explanation for the greater tendency of uncertain adolescents to express negative/neutral views is that they may be in a process of questioning and exploring their sexual identity. Research has shown that questioning youth often experience higher levels of internalized stigma than other sexual minority youth (Watson et al., 2019). Alternatively, some of the uncertain adolescents may simply be less attuned to experiences involving sexual/romantic attraction/intimate relationship in general.

There was no significant relationship between parental responses to sexuality-related questions and the attitude patterns among secondary school students. In Chinese contexts, where traditional Confucian beliefs prevail, sexual orientation remains a sensitive topic between parents and children. Discussions related to sexual orientation may raise concerns regarding an individual’s ability to fulfill family responsibilities, such as marriage and having children (Wang & Zheng, 2021). Furthermore, stigma toward sexual minorities may extend to their parents, thereby causing shame and disgrace to their family (Choi & Luo, 2016). As a result, adolescents may find it difficult to discuss issues related to sexual orientation with their parents.

There appears to be a conflicting trend in the relationship between exposure to online sexual knowledge and the attitude patterns. Since we did not measure the exact content that respondents were exposed to, we can only provide a potential explanation for this phenomenon. On one hand, increased exposure to online sexual knowledge means more opportunities to learn about different sexual orientations, which may lead to more positive attitudes toward homosexuality. On the other hand, most current sex education programs are still framed from a heteronormative perspective, which may unconsciously convey a message that homosexuality is a deviation from the norm. As a result, some respondents with greater exposure to online sexual knowledge might fall into the intolerant class, rather than the neutral class.

Overall, our study provides updated insights into public attitudes toward homosexuality in Hong Kong, offering valuable guidance for policymakers and practitioners to evaluate whether their policies and practices align with the evolving societal views. While we observed increasing acceptance among Hong Kong adolescents, a notable portion still holds intolerant attitudes, underscoring the need for continued progress. To promote broader societal acceptance, legal efforts, beginning with anti-discrimination legislation for sexual minority groups, are crucial, as they lay the foundation for further institutional support. Moreover, as our study highlighted, modern prejudice can manifest in various forms. It is, therefore, essential not only to emphasize the formal rights of sexual minority groups but also to be mindful of more subtle yet significant forms of discrimination (Doan et al., 2014), such as microaggressions.

This study has several limitations. First, the measurement of attitudes toward homosexuality was not exhaustive. Although scales have been developed to capture the multifaceted nature of the attitudes (Adolfsen et al., 2010), they have not yet been validated for use in the Chinese context. Future studies may benefit from developing or validating a multidimensional scale specifically tailored to the Chinese context, incorporating a wider range of dimensions for comparison and pattern analyses. Second, our analysis was restricted to a limited set of covariates. Future studies should investigate the influence of factors such as family socioeconomic status, parental attitudes toward homosexuality, and contact with sexual minority groups on the attitude patterns. Moreover, the way Chinese society perceives and approaches homosexuality issues is shaped by the importance of family kinship and the social boundaries between family members and outsiders (Yeo & Chu, 2018). A notable disparity in acceptance levels toward homosexuality was observed across different types of relationships: acceptance rates were substantially higher for co-workers (83.57%) and friends (76.92%) compared to relatives (40.19%) (Yeo & Chu, 2018). As such, future research could explore the potential impact of Chinese traditional values, such as emphasis on conformity and family harmony, on societal attitudes toward homosexuality in the Chinese context. Third, although we aligned the sex and age composition across years to enhance comparability, the changes in sampling methods may have introduced biases beyond our control. Nevertheless, to the best of our knowledge, our dataset represents the largest repeated cross-sectional survey investigating the changes in attitudes toward homosexuality among adolescents in Hong Kong. Fourth, our study may not be representative of the adolescent population in Hong Kong as students from non-Chinese speaking schools and evening schools were not included. However, no more than 6% of the day school students enrolled in non-Chinese speaking schools during 2010 and 2020 (Education Bureau, The Government of the Hong Kong Special Administrative Region, 2024), and the number was even lower for those attending evening schools. Fifth, given that our sample was limited to secondary school students in Hong Kong, our findings may not generalize to other population groups or socio-cultural contexts. Future research should extend to other age groups such as young adults or working adults across diverse occupations for a more comprehensive understanding of attitudes toward homosexuality in the general public. Investigating diverse socio-cultural contexts and age cohorts can help inform targeted strategies to reduce discrimination. Sixth, our study adopted a repeated cross-sectional design which limits our ability to examine within-person changes in attitudes over time. Future research could employ a longitudinal design and latent transition analysis to track individual-level changes. Finally, although the present study emphasized respondent anonymity and confidentiality and minimized teacher involvement during the survey process, the sensitive nature of the topic could induce social desirability bias in the responses and our results.