BSA 42 | Gender Identity

Gender Identity

This research investigates the complexities of gender identity and its impact on social attitudes and policy preferences in the UK. Our own sense of gender identity, and how feminine or masculine we feel can influence what we think. But what influences our sense of gender and how do different gender identities influence support for gender equality and LGBTQ+ policies? In this report, we examine the nuanced understanding of gender identity, providing valuable insights into the evolving landscape of gender roles and their implications for public policy.

Traditional gender norms may dominate, but they are not polarised

• On average, self-reported gender identity is more important for women than men – 66% of women say that being a woman is extremely or very important to how they think of themselves, compared with 52% of men who feel similarly about being a man.
• Only 36% of women can be classed as ‘hyper-feminine’ (high on femininity and low on masculinity) and only 35% of men are ‘hyper-masculine’ (high on masculinity and low on femininity).

Being man/woman extremely or very important to how respondents think of themselves

See Table 1 for unweighted bases.

Education and youth are key drivers of gender identity becoming less polarised

• Younger people report less polarised gender identities: 35% of women aged under 35 feel ‘very’ feminine, compared with 46% of aged 35 and over. Similarly, 30% of men aged under 35 feel very masculine compared with 41% of men aged 35 and over.
• Gender identity is also correlated with level of education: 36% of women with a degree feel ‘very feminine’ compared with 49% of women with no educational qualifications. Meanwhile, 31% of men with a degree feel ‘very masculine’ while 44% of men with no educational qualifications do so.

Perceptions of gender identity influence views on public policy

• Support for gender equality and LGBTQ+ policies is associated with how people feel about their gender identity.
• Those who are hyper-masculine are less likely (50%) to support equal parental leave than those with androgynous identities (57%) and are also less likely to support fining companies for having a gender pay gap (65% vs 81%) or rights for same-sex couples to adopt children (57% vs 89%). 

Introduction

Since the British Social Attitudes (BSA) survey began in 1983, researchers have paid increasing attention to how gender shapes political attitudes and behaviours. Much of this focus has centred on differences between men and women—commonly referred to as the "gender gap"—in areas such as party support, electoral turnout, and policy preferences. For example, women were historically more likely than men to support the Conservative Party in Britain, but this trend has reversed over the past two decades (Campbell and Shorrocks, 2024), and the latest BSA data show little difference in the proportion of men (19%) and women (20%) voting for the Conservatives at the last election.  

However, while such analyses offer valuable insight, they rely on a binary understanding of gender as male or female. As gender roles shift alongside broader social changes—such as increased female labour force participation—understanding how individuals relate to gender identity, whether aligning with traditional or more progressive conceptions, is key to explaining variation in political and social attitudes. These identities shape interests and values, which in turn influence support for gender equality policies, left–right ideological positions, and views on issues such as same-sex relationships. This report examines how this understanding of gender may be equally—if not more—significant than traditional binaries in explaining these political and social attitudes.

While research on the "gender gap" has shown differences between men and women in political preferences and attitudes, there are also considerable variations in attitudes that draw our attention toward gender identity as a more complex and socially embedded concept. Rather than assuming gender fits neatly into binary categories, scholars are increasingly exploring how people identify along spectrums of masculinity, femininity, and non-binary identities (Gidengil and Stolle, 2021). This reflects broader societal changes in how gender is understood—no longer as a simple extension of biological sex, but as a multidimensional aspect of personal and social identity. These gender identities are not evenly distributed across the population; they vary with factors such as age, education, and social context, and are closely linked to political and cultural attitudes (Hatemi et al., 2012; McDermott, 2016). Understanding the diversity of gender identities and how they intersect with other social and demographic characteristics offers a more nuanced picture of how individuals navigate political and social life (Wängnerud, Solevid, and Djerf-Pierre, 2019). This report examines these patterns in the United Kingdom using a new set of survey questions designed to better capture gender identity and expression, and how they relate to factors such as age, education, and social attitudes.

There is already strong evidence that a more nuanced conception of gender identity may help to explain social attitudes across a range of issues. Despite a move towards more progressive views on gender equality, a gender gap on egalitarian attitudes persists with women consistently reporting more progressive attitudes than men (Shu and Meagher, 2025) and this highlights the need to extend our analysis of gender to include assessments of femininity and masculinity. First, while we are usually drawn to a “gap” between men and women, there is also considerable overlap and both men and women have become more egalitarian over time. This overlap though will mask differences in attitudes that may be driven by gender identity and variations in how firmly one adheres to self-perceptions of femininity/masculinity. For example, femininity is a strong predictor of progressive attitudes about homosexuality (Falomir-Pichastor, Berent, and Anderson, 2019; Whitley, 2001) and masculinity is a strong predictor of support for radical right parties (Coffé, 2019).   Second, education and ageing are expected to shape these self-assessments (Shu and Meagher, 2025). Education is a key driver of progressive gender attitudes, with higher levels of education consistently linked to more egalitarian views on gender and sexuality (Alexander, Bolzendahl, and Öhberg, 2021). Education increases exposure to diverse ideas and liberal values, which can encourage rejection of binary gender norms and reinforce the desire for recognition of diverse gender identities. Older generations were socialised during childhood into rigid gender norms (Inglehart and Norris, 2003), while younger generations may feel less pressure to conform to binary gender identities. We also may expect that as individuals age, accumulating resources and taking on new responsibilities, that their views become more traditional (Shu and Meagher, 2025).

For our analysis, we draw on three measures of these feminine and masculine identities: a) importance of one’s gender to oneself; b) self-assessment of overall masculinity and femininity, and c) self-assessment of typical masculine and feminine traits. Self-assessment of masculinity and femininity is captured by asking respondents to place themselves on two scales. Traits associated with masculine and feminine identities—such as dominance or compassion—have been measured in the tradition of the Bem Sex Role Inventory (BSRI) created by the American psychologist Sandra Bem (Donnelly and Twenge, 2017), who developed a framework categorising personality traits according to gendered social expectations. The meaning of these traits has changed over time with feminine traits becoming more weakly associated with women (Eagly et al., 2020).  Below we discuss these measures in greater detail and show how gender identity can influence individual behaviour, social attitudes, and policy preferences, making it a key consideration in both social research and policy analysis.

In our analysis of the influence of gender identity on social attitudes and policy preferences, we will pay particular attention to attitudes on gender equality policies and homosexuality. This is for several reasons. First, we know gender is important in determining egalitarian attitudes, but we suggest it will give us a better understanding of where gender gaps occur if we account for gender identity. Second, resistance to support for gender equality and homosexual rights can be seen as affirmations of traditional gender roles (Falomir-Pichastor et al., 2019). Finally, people’s attitudes towards gender equality overlap with their attitudes towards freedom of sexuality (Inglehart and Norris, 2003) and liberalisation of attitudes about sexuality coincided with liberalising attitudes about gender (Clements and Field, 2014). In research exploring the relationship between gender and support for various social policies, gender-role attitudes are often theorised to mediate this relationship. Specifically, individuals who hold traditional notions of gender roles typically express less support for policies aimed at facilitating women's increased participation in the labour force, such as childcare provision and parental leave schemes. Conversely, individuals with less traditional or more flexible gender-role attitudes tend to be more supportive of such equality-enhancing policies. Similarly, policies that directly challenge traditional conceptions of masculinity—for instance, strict enforcement of workplace harassment regulations—are more likely to find support among those whose understanding of gender identities and roles is more progressive and fluid.

Our data were collected across the United Kingdom by the European Social Survey (ESS) in 2023 and they come from a new module on contemporary attitudes about gender and gender equality, fielded as part of Round 11 of the European Social Survey European Research Infrastructure (ESS ERIC) 2024. Beginning first by describing the measures of gender salience, we outline the set of questions from this module with a special focus on these measures and self-assessed femininity and masculinity. We next examine how the femininity and masculinity scale are related to traits that have been linked to femininity and masculinity. These traits are seen as stereotyped expressions of gender that represent ways in which society has traditionally expected men and women to behave. We then move to an examination of how femininity and masculinity and the salience of gender identity shape policy preferences and attitudes about gender equality and sexual freedoms.

Gender salience: non-binary measures of femininity and masculinity

Gender identity refers to an individual's internal sense of self in relation to gender, drawing on concepts from social psychology and sociology. Unlike biological sex, which is typically assigned at birth based on physical characteristics, gender identity is constructed through social and cultural experiences and reflects how one perceives oneself in relation to notions of masculinity and femininity. Rather than treating masculinity and femininity as opposite ends of a single spectrum, both measures recognise that individuals may identify with both, neither, or varying combinations of these traits. In other words, being high on a femininity score does not mean one is low on masculinity.  Some may see themselves as possessing both masculine and feminine characteristics, while others may not feel connected to either. For some, gender identity is a central part of self-conception, while for others it may play a minimal role. These identities are shaped through socialisation, cultural norms, and expectations about appropriate behaviour, often influenced by traditional gender roles and stereotypes. In this section we cover self-reported gender, the importance of this gender identity and an alternative way of capturing gender through femininity and masculinity scales that allows for a less rigid categorisation of gender identities.

Self-reported gender and gender identity salience

For the first time in its latest survey, the ESS asked for self-reported gender identity, and allowed for gender identities other than “man” or “woman”, but at the same time repeated the regular practice of the interviewer assigning the sex of the respondent. ESS has always relied on interviewer assigned sex. The self-report of gender with multiple categories of responses and the possibility for an open-ended description allows for people not to identify as cis-gender – that is, as someone whose gender identity conforms to their sex assigned at birth.  There is a high degree of overlap between the self and interviewer reported measures. Among the respondents that were assigned as male by interviewers (n = 824), 13 self-identified as women. In contrast, of the 860 interviewer assigned females, none self-identified as a man. There were three respondents who identified as ‘other’ and these were the only responses apart from “man” or “woman”. The wording of our question on the salience of gender identity drew on this self-reported measure: those who identified as men were asked about the importance of being a man while women were asked about the importance of being a woman:

Many things can influence how people think about themselves. These may include where they are born, their profession, their age, or whether they are a man or a woman. Using this card, how important is being a man/woman to the way you think about yourself? 

Respondents rated themselves on a scale of 0 (not at all important) to 6 (extremely important).

Given (i) the large overlap between the new self-reported gender question on the survey and interviewer-coded sex, (ii) that our measure of the importance of gender identity relies on the self-reported measure, and (iii) that interviewer-coded sex is not necessarily a more reliable measure than that which the respondent reports, we will rely on self-described identification as a man or woman throughout this analysis.

Overall, as shown in Table 1,  self-reported gender identity appears more important for women than for men, with 66% of women indicating that being a woman is extremely or very important to how they think of themselves, compared with 52% of men who feel similarly about being a man. Education shapes the importance placed on gender identity among both genders. Men without qualifications are most likely to view their gender identity as highly important (63%), compared with 41% of degree holders. Similarly, women without qualifications attribute the greatest importance to their gender identity (71%), with a somewhat lower proportion among those with qualifications but no degree (67%) and a lower figure again among those holding a degree (60%). 

Note: The denominator used to calculate percentages includes don’t knows/refusals/NA to ‘how important being man/woman to how one thinks of themselves’. Respondents who did not identify as man or woman were not asked about importance and are thus not included in the analysis or denominator. The 42 cases missing a value on education are not included. Age is grouped into two categories – younger (under 35 years) and older (35 years and older).

There are also clear differences in response to this question by age. Among men, older individuals attribute greater importance to their gender identity (55%) than younger men (43%), but there is no statistically significant difference between older (66%) and younger (68%) women. One interpretation would be that these age-based differences highlight that gender identity becomes more significant to men's self-conception as they age, whereas for women, getting older makes no difference to their gender identity. However, given we are only examining one point in time, we cannot demonstrate how these generational dynamics play out over time and across generations, and if male identity becomes more important for people as they age or whether male identity is simply more important for older generations.

Femininity and masculinity

We next turn to self-assessments of femininity and masculinity. All respondents, irrespective of their gender identity, rated themselves on a scale from 0 (not at all) to 6 (very much) for both femininity and masculinity:

Regardless of whether they were born male or female, people sometimes differ in how masculine or feminine they feel.  Overall, how [masculine/feminine] would you say you feel?

Where responses are more evenly distributed we would suggest this is evidence of more flexible gender identities, while distributions that are skewed toward the extremes (either not at all or very masculine for men and feminine for women) show a more rigid gender identity and an adherence to traditional conceptions of gender.

Note: The denominator used to calculate percentages includes don’t knows/refusals/NA to femininity scale. Respondents who did not identify as man or woman (n=3) are not included in the analysis or in the denominator. Age is grouped into two categories – younger (under 35 years) and older (35 years and older). We use three education categories – no qualifications reported, qualifications but no report of a university degree or equivalent, university degree or higher. The 42 cases missing a value on education are not included.

As shown in Table 2, femininity self-ratings are notably polarised: 31% say they ‘do not feel feminine at all’ and 22% say they ‘feel very feminine’ while the intermediate points on the scale (1–5) each attract between 6% and 13% of respondents. Clear gender differences emerged. Men overwhelmingly reported low levels of femininity, with 61% selecting the lowest category (do not feel feminine at all) and less than 1% selecting the highest (feel very feminine). Women, by contrast, largely identified with higher levels of femininity: 44% selected the highest level, and only 1% selected the lowest. Nevertheless, respondents did place themselves at points other than the extremes: 39% of men did not opt for ‘not feminine at all’ and 56% of women opted for other than ‘very feminine’.

Age differences highlight potential generational trends or the effects of ageing on this issue. Although both were unlikely to feel feminine, older men were substantially more likely (67%) than younger men (47%) to feel ‘not at all’ feminine. ‘Meanwhile, older women (46%) were more likely than younger women (35%) to say they felt ‘very feminine’. Young women also reported high femininity, though slightly less than older women, with 35% selecting the highest level.

Note: The denominator used to calculate percentages includes don’t knows/refusals/NA to masculinity scale. Respondents who did not identify as man or woman (n=3) are not included in the analysis or in the denominator. Age is grouped into two categories – younger (under 35 years) and older (35 years and older). We use three education categories – no qualifications reported, qualifications but no report of a university degree or equivalent, university degree or higher. The 42 cases missing a value on education are not included.

The responses for masculinity, shown in Table 3, demonstrate a somewhat similar pattern, with men identifying strongly with masculinity and women tending to disidentify. Men strongly associated themselves with masculinity: 39% selected the highest level, while less than 1% chose the lowest. In contrast, among all women, 57% did not feel masculine at all while 1% reported feeling very masculine.  Even though there is polarisation along gender, as with femininity, we do note that respondents opted for other than the extreme ends of the scale: 61% of men opted for other than ‘very masculine’ and 43% of women opted for other than ‘not at all masculine’. When comparing this to feminine identity, a slightly higher proportion of women feel ‘very feminine’ (44%, see Table 2) than men feel ‘very masculine’ (39%).

As with femininity, there are differences by age within gender. Older men were the most likely to feel very masculine (41% at highest level), while younger men showed more flexibility, though 30% still said they felt very masculine. Older women were the least likely to say they felt masculine, with 61% indicating they did not feel masculine at all. Younger women reported somewhat higher masculinity than older women, though a substantial proportion (41%) still selected the lowest level.

The response patterns reveal strong alignment with traditional gender norms, with 39% of men feeling very masculine and 44% of women as very feminine. The direction of the relationships with these demographics is similar to that seen for the importance of masculine/feminine identity. Older individuals show more polarised identities, whereas younger groups—particularly younger women—are slightly more fluid in how they perceive their gender. However, while we note that young men are less likely to feel ‘very masculine’ (30%) than young women are to feel ‘very feminine’ (35%) and a similar pattern is evident for older respondents, where men are less likely to feel ‘very masculine’ (41%) than women are to feel ‘very feminine’ (46%). Thus, the influence of traditional gender roles remains prominent across age and gender groups, but somewhat more so for women.

In combination, Tables 2 and 3 also show the relationship between people’s response to the femininity and masculinity scales and their educational qualifications. Men with no qualifications are least likely to feel feminine (66% do not feel feminine at all). However, for men with a degree or higher, only 51% strongly reject femininity, though they still largely do not identify as feminine. Women with no qualifications are most likely to strongly identify as feminine (49% feel ‘very feminine’), compared with 36% of women with degrees who do so. However, higher-educated women still mostly give themselves a relatively high score on the femininity scale.

Men with no qualifications have the strongest identification with masculinity: 44% feel ‘very masculine’, the highest of any educational group, compared with 31% of those with a degree. Among women, educational attainment is also associated with identification with masculinity. Women with no qualifications overwhelmingly reject masculinity, with 66% reporting they ‘do not feel masculine at all’, compared with 44% of women with degrees who feel this way.

Another way to examine feminine and masculine identities is to combine the scales to identify patterns of ‘hyper-gender’ and ‘androgynous’ identities. A hyper-feminine identity is one where there is strong attachment to femininity and a corresponding weak attachment to masculinity. Similarly, a hyper-masculine identity corresponds to having an extreme masculine identity and a weak or non-existent feminine identity. Androgynous identities are those where respondents have placed themselves in the middle of both the femininity and masculinity scales indicating that they do not strongly identify with either.  In Table 4 we show the proportion of respondents who display hyper-feminine and hyper-masculine identities. Following on Gidengil and Stolle (2021), for hyper-feminine identities we have taken respondents who scored a 6 (fits extremely well) on the femininity scale and a 0 (not at all) on masculinity. We take the opposite ends of the two scales to define hyper-masculinity.  To designate androgyny, respondents who gave scores between 2 and 4 on both scales are coded as androgynous in their gender identity. We have not reported in Table 4 the gender identities that are weak (e.g. a response of 2 on one scale and 5 on the opposing scale) or atypical (e.g. high values for both femininity and masculinity, or low values for both). We also only report variations in the level of hyper-masculinity among men and that of hyper-femininity among women. ¹

Note: The denominator used to calculate percentages includes don’t knows/refusals/NA to the femininity/masculinity scale. Respondents who did not identify as man or woman (n=3) are not included in the analysis or in the denominator. Age is grouped into two categories – younger (under 35 years) and older (35 years and older). We use three education categories – no qualifications reported, qualifications but no report of a university degree or equivalent, university degree or higher. The 42 cases missing a value on education are not included.

Table 4 shows that only just over a third (35%) of men are hyper-masculine, while a similarly low proportion of women (36%) are hyper-feminine. Meanwhile 13% of men and the same proportion of women can be classified as androgynous. These distributions suggest a landscape of gender identity expression in Britain where many individuals do not conform strictly to either hyper-masculine or hyper-feminine conceptions of gender.

The prevalence of hyper gender identities is strongly associated with educational attainment. Among men, hyper-masculine identities are highest among those with no qualifications (40%) and those whose highest qualification is less than a degree (39%). The proportion is much lower among those with a degree (25%). Conversely, hyper-feminine identities among women are most prevalent among those with no qualifications (39%) and least common among women holding a degree (25%). The pattern of response is similar when we asked about femininity and masculinity separately (Tables 2 and 3), where higher educational attainment appears to be associated with less polarised identities among both genders. Conversely, in both genders, those with a degree are somewhat more likely than those without a degree to have an androgynous sense of identity.

Age also influences attitudes on this issue. Hyper-masculinity is considerably higher among older (38%) than younger men (25%). Among women, older respondents are considerably more likely to have hyper-feminine identities (38%) compared to younger women (26%). Also, androgynous identities are more likely to be found among younger age groups. These results either reflect generational shifts towards more flexible or nuanced gender identities among younger individuals, or a move to less flexible conceptions of gender among people as they age. Unfortunately, as we only have a cross-section of data, it is not possible to tell whether the former or latter explains the relationship between hyper-gender identities and age.

This evidence is, however, consistent with the view that gender identity in Britain is perhaps becoming increasingly diverse and less anchored in traditional binaries, particularly among the young and highly educated. The trends are visible in both men and women. When combined with the evidence that gender identities are less important for the young and highly educated, generally, our results point to a weakening of the association of identity with the traditional binary categories of gender. We find it particularly interesting that similar trends are evident for women and men. Given that traditional masculinity is seen to be under a greater amount of wider scrutiny, making deviations from it are more significant markers of identity change; it may be surprising that a similar proportion of women identify as ‘hyper-feminine’ (36%)  as men do ‘hyper-masculine’ (35%). However, the cross-sectional nature of the data leaves open the question of whether current patterns among younger respondents will persist or evolve as they age.

Perceptions of gender role traits

Another way to evaluate gender identity, as noted in the introduction, is based on the Bem Sex-Role Inventory (BSRI), a psychological construct developed by Sandra Bem. Similar to self-expressions of femininity and masculinity, rather than defining gender as binary and fixed identities, the BSRI measures people’s level of identification with traits that are culturally stereotyped as masculine or feminine, such as assertiveness (masculine) or compassion (feminine). These associations of traits with gender are not static — as social roles evolve, so too can perceptions of which traits are considered masculine or feminine. For instance, leadership may increasingly be seen as a feminine trait as more women take on leadership roles, while compassion may become associated with masculinity as men engage more in caregiving.

In this context, men who identify strongly with masculinity may emphasise traits like risk-taking and emotional restraint, whereas others with more flexible gender identities might still identify as masculine while also embracing sensitivity and compassion. Comparing people’s self-assessment of their gender to the BSRI allows us to explore how individuals relate to evolving gender norms and the extent to which their identities reflect either traditional or more fluid conceptions of gender roles (Coffé, 2019).

In this next section, we examine how the ascription of these gendered traits vary across men and women by age and education.  Social role theory predicts that gender differences in these stereotyped traits should decline over time as gender equality increases. However, recent evidence suggests a more complex pattern. Eagly et al. (2020) found that women’s relative advantage in feminine traits has grown, while men’s advantage in masculine traits has remained unchanged. Donnelley and Twenge (2017) similarly show little to no shift in masculine traits among men but a decline in feminine traits among women. Either way, there are divergent trends rather than a uniform decline for both masculine and feminine traits. We measured these feminine and masculine traits in the survey by presenting respondents with four traits: two reflecting stereotypically feminine qualities (‘sensitive to others’ needs’ and ‘acting compassionately towards others’) and two representing stereotypically masculine qualities (‘liking to take risks’ and ‘liking to be a leader’). Responses were measured on a scale from 0 (not at all) to 6 (completely). In Table 5 we have grouped the responses into three categories: Not at all (0) to 2, meaning the trait does not fit well; the midpoint score 3 and 4 to completely (6), indicating the trait fits them well. The questions read as follows:

Please tell me to what extent the following statements describe you: 

• I like to take risks
• I like to be a leader
• I am sensitive to others’ needs
• I act compassionately towards others

The data in Table 5 show that, across the board, the feminine-coded traits of compassion and sensitivity to others’ needs are more widely endorsed by the public than the masculine-coded traits of risk-taking and leadership. Around nine in ten (89%) people say that being compassionate fits them well or completely (options 4-6), while 81% say the same for being ‘sensitive to others’ needs’. By contrast, less than half (47%) strongly identify with leadership, and 37% feel that ‘likes to take risks’ describes them completely. This suggests that the population is perhaps more ready to align with traits associated with emotional attunement and care, even though these have traditionally been gendered as ‘feminine’.

Note: This table displays the responses to the four questions on whether instrumental and expressive traits fit the respondent. Responses are grouped into three categories 0 – 2 indicating trait does not fit well, a middle category of 3 and 4 to completely 6 indicating the trait fits them well.

When we break this down by gender, patterns of gender stereotyping can be seen but there is also considerable similarity in how men and women feel these traits fit them, in contrast to the clear gender split for the questions on self-identification and the importance of gender. True, men are more likely than women to describe themselves as liking to take risks (42% vs. 33%) and as enjoying leadership roles (51% vs. 43%). In contrast, women are more likely to see themselves as compassionate (93% vs. 85%) and sensitive to others’ needs (86% vs. 75%). However, these differences represent only a slight indication that these traits follow traditional gender patterns. The high levels of endorsement across all four traits suggest that many individuals, regardless of gender, identify with both assertive and caring characteristics.

Tables 6 and 7 report the proportion of respondents who feel the traits fit them well or completely (values 4, 5 and 6) by age and education. As previously, we have grouped these by whether respondents identify with being a man or a woman. The data show clear differences in how men and women of different ages perceive themselves in terms of gender role traits. Younger women and men are equally likely to report a tendency to take risks (56%) though this trait is less prevalent among older people of both genders — and especially so among women (24%) than men (38%). When it comes to leadership, younger people of both genders are more likely to see themselves as leaders than their older counterparts, with men rating themselves slightly higher than women in both age groups. Interestingly, these two findings mean younger respondents are more likely to ascribe to themselves typically masculine traits even though they are less likely to see themselves as hyper-masculine. We expect that this indicates that younger respondents may be less likely to see these traits in gender stereotypical ways.

Compassion is a trait where women, regardless of age, rate themselves more highly than men. However, men’s self-reported compassion is higher among older people. A similar pattern is observed in sensitivity to others’ needs, where women again score higher than men, and men show greater sensitivity as they age. For men, feelings of compassion and sensitivity are positively associated with age, which interestingly is the opposite of the relationship shown in Table 4, where hyper-masculine identities are more common among older people. 

Note: This table displays the % of respondents who say one of the four questions on whether instrumental and expressive traits fit the respondent fits them well or completely. Younger respondents include those up to the age of 34 and older respondents are 35 years of age or older. Don’t knows/refusals are also shown and represented in the base number of resondents.

Table 7 summarises gender role traits by education level, showing the percentage of men and women who agree that certain traits fit them well or completely (values 4,5 and 6). In all except one instance, the proportion of respondents agreeing the trait fits them well is positively associated with level of education for both men and women. The one exception is ‘sensitive to others’ needs’ among women where those in the middle category of education are slightly more likely to agree with the trait than women with a degree (91% compared to 88%).  As we have already seen, men are more likely than women to say they ‘like to take risks’, but the gap between men and women is bigger among those with higher levels of education and education is associated with a great increase in this trait for men than for women. For example, 35% of men with no qualifications agree, compared with 50% of those with degrees, a gap of 15 percentage points.  Among women, the proportion is also higher among those with a degree (36%) than it is among those with no qualifications (27%) but the increase is smaller. In the case of  leadership, the reverse is true; the gap for women (29 points) is greater than that for men (22 points). Agreement with the trait of compassion is high amongst all respondents so there is no evident relationship with education. On sensitivity, there is little difference by education among women but among men those with a higher level of education are more likely to ascribe themselves the trait. Overall, traditional gender differences in self-perceived traits are irrespective of education, though in the case of ‘likes to be a leader’ and ‘sensitive to others’ needs’ the gap is smaller among those with a degree. 

Note: The denominator used to calculate percentages includes don’t knows/refusals/NA. We use three education categories – no qualifications reported, qualifications but no report of a university degree or equivalent, university degree or higher. The 42 cases missing a value on education are not included.

Gender, age and education all shape attitudes toward traditional gender-role traits. With younger generations who are more likely to have gone to university, some of the age versus education effects may be difficult to disentangle. Nevertheless, there is some evidence that younger women may be less traditional in their greater likelihood to ascribe themselves masculine traits than older women. On the other hand, young men are more likely to ascribe themselves masculine traits and less likely to ascribe themselves feminine traits than older men. The experience of higher education makes both men and women more likely to acknowledge both feminine and masculine traits. Thus, the patterns are not consistent with less stereotyped perceptions at higher education levels across all traits, but rather they show a convergence of all traits among the more highly educated and a divergence into stereotyped perceptions among those without a degree.

Comparing the results presented in Table 6 and 7 to the results for expressions of femininity and masculinity in Tables 3 and 4 raises questions about the comparability of the two types of measures of gender identity. There were clearer patterns on self-expressions of femininity and masculinity that younger and higher educated individuals had less rigid identities. As we indicated in the introduction, these two ways of capturing gender identity were once seen as complementary but more contemporary research has shown that the ascription of traits has shifted over time (Donnelly and Twenge, 2017; Eagly et al., 2020).  We next examine the extent to which these gender role traits are aligned or not with people’s self-assessment of femininity and masculinity. In order to examine how closely the traits are related to feelings of masculinity and femininity, we calculate a measure of association. The closer the measure of association is to 1, the more closely the two are related in a positive for the gender trait. A negative value indicates that as femininity or masculinity increases, the likelihood of strongly ascribing to one of the gender traits declines. When the measure is 0, or close to 0, this means there is no association or a very weak association between femininity/masculinity and the gender trait.

Figure 1 Association between gender role traits and femininity/masculinity

Note: This heat map displays measures of association (Pearson correlation coefficients) between selected gender-related traits—risk-taking, leadership, compassion, sensitivity and self-assessments of femininity and masculinity—separately for men and women. The values reflect how strongly the traits are associated with femininity and masculinity. Negative associations are shaded yellow and darker shades represent stronger negative relationships. Positive associations are shaded blue and darker shades of blue represent stronger positive relationships. Where there is not association this is shaded green.

We have calculated these measures of association between all traits and femininity and masculinity, separately for men and women, and have displayed them in Figure 1. When comparing the associations between gender-role traits, femininity, and masculinity for men and women, several notable differences emerge. Among women, self-assessed masculinity is moderately and positively correlated with willingness to take risks (0.3) and preference for leadership (0.3). ² However, for men, masculinity has none to a very weak correlation with these traits (risk-taking 0.0; leadership 0.1). ³  This indicates that self-perceived masculinity among women is meaningfully linked with traditionally masculine traits such as leadership and risk-taking, while for men, masculinity does not strongly predict these attributes. Both men and women show very weak correlations between femininity and traits like compassion and helping others (women: compassion 0.1, sensitive to others 0.1; men: compassion 0.1, sensitive to others 0.1). ⁴  These generally weak relationships suggest that contemporary understandings of masculinity and femininity no longer map neatly onto specific personality traits. This mismatch could reflect a loosening of traditional gender-role expectations, where the identification with a gendered identity label (masculine or feminine) does not necessarily imply endorsement of the stereotypical traits that have been historically associated with that label.

Gender equality and sexual freedoms: attitudes and policy preferences

In this section we examine how hyper-femininity, hyper-masculinity and androgyny are related to social and political orientations that are relevant for contemporary political debates. Here we will focus on attitudes toward gender equality policies and freedoms for gay and lesbian people. The ESS asks respondents whether they agree or disagree with measures promoting equal rights and opportunities for men and women. It also asks about freedoms for sexual minorities. Previous research on values and value change shows that attitudes towards gender equality reflect whether people back conformity to traditional gender roles or are open to diversity and flexibility in gender roles – and thus in turn their sense of gender identity (Inglehart and Norris, 2003). Consequently, individuals who hold polarised conceptions of their own masculine and feminine traits—such as hyper-masculinity or hyper-femininity (Gidengil and Stolle, 2021) — may therefore be resistant to equality-oriented policies. Conversely, individuals with more fluid or balanced conceptions of their own masculine and feminine traits may be more supportive of gender equality policies, diversity, and inclusive social values.

Note: Don’t knows/NA included in the denominator. The proportion of respondents who strongly or somewhat support each policy are shown in the table. Hyper-feminine: 6 femininity, 0 masculinity. Hyper-masculine: 6 masculinity, 0 femininity. Androgynous: scores 2–4 on both scales. Hyper-feminine includes women only and hyper-masculine category includes men only.

In Table 8 we show how masculine and feminine identities are related to support for gender equality policies. We ask about four policy areas. ⁵  Two are directly related to gender issues in the workplace: firing employees who make harassing comments directed at women in the workplace and fining companies where there is a gender pay gap. A third policy is related to the private sphere and the role of parenting, asking respondents whether they would favour requiring both parents of a new baby to take equal periods of parental leave. This policy would represent a significant role change for women, who are the primary providers of childcare in the UK (Andrew et al., 2024) and who typically take longer periods of statutory parental leave than new fathers. The fourth policy area asks about gender equality in the political sphere and whether respondents favour distributing parliamentary seats equally between men and women. This latter policy also addresses a sphere where men continue to hold more power. However, if seats were allocated equally, and not based on electoral outcomes, this may be perceived as achieving equality in a manner that is contrary to democratic practices.  

The results in Table 8 reveal clear patterns in how gender identity shapes support for gender equality policies, though the extent to which it does so varies depending on whether the policies challenge traditional gender roles in the private or public spheres. Policies directly related to workplace gender equality, such as firing employees for harassment and fining companies for gender pay gaps, enjoy relatively high support overall (73% and 75%, respectively). Although there is no difference between men and women in their attitudes to firing for harassing behaviour, there is an 11% gap between men and women in their view on fining companies for a gender pay gap. Meanwhile, analysis by gender identity reveals that hyper-masculine men express somewhat lower support for firing harassers (67%). Conversely, respondents with weak attachments to femininity and masculinity (that is, are androgynous) strongly favour firing employees for harassment (76%), showing that those who tend to reject strong gender identities endorse strong interventions against gender-based discrimination at work.

The policy that challenges traditional gender roles in the private sphere—requiring equal parental leave—has notably lower overall support (53%). Men (49%) and those showing hyper-masculinity (50%) are the least likely to support this, likely reflecting a perceived threat to traditional gender roles regarding parenting and caregiving responsibilities. In contrast, women show the highest level of support (61%), but hyper-feminine respondents show lower levels of support (56%). Support for political gender quotas (which challenge traditional gender roles in the public sphere as they involve redistributing political power traditionally held by men) has the lowest overall support (40%). Resistance is particularly pronounced among hyper-masculine respondents (31%), highlighting discomfort with directly redistributing power from men to women.

Men who say gender is important to their identity show lower support for gender equality policies than other men, especially on measures such as fines for the gender pay gap (64%) and political gender quotas (36%). In contrast, women for whom gender is important remain highly supportive of equality policies, even slightly above the average for all women—suggesting that gender salience reinforces traditional views among men but strengthens egalitarian commitments among women.

Overall, these findings illustrate that those policies that may be perceived as posing the greatest threat to traditional gender roles—particularly those involving the private sphere or direct redistribution of power — appear to generate greater resistance from individuals with traditional or hyper-traditional gender identities.

Note: Don’t knows/NA included in the denominator. The proportion of respondents who strongly or somewhat support each policy are shown in the table. Hyper-feminine: 6 femininity, 0 masculinity. Hyper-masculine: 6 masculinity, 0 femininity. Androgynous: scores 2–4 on both scales. Hyper-feminine includes women only and hyper-masculine category includes men only.

The data in Table 9 show responses to a set of three questions about rights for gay and lesbian people. ⁶ The distribution of responses suggests that gender identity is an important factor in shaping attitudes toward sexual freedoms, with more flexible or less stereotypical gender identities generally associated with higher levels of support for gay and lesbian rights. Overall, support is relatively high across the sample: 88% agree that gay and lesbian people should be free to live as they wish, 84% report no shame if a close family member is gay or lesbian, and 71% support adoption rights for same-sex couples. There is broadly similar support among men and women across all items, except for adoption rights where support among women is higher (76% vs. 68%). The variation becomes more pronounced when analysed by gender identity type. Those who identify as androgynous—who acknowledge both masculinity and femininity — report the highest level of support across all measures, with 94% supporting freedom to live as they wish, 92% reporting no shame, and 89% supporting adoption. In contrast, those with more rigid gender identities, such as hyper-masculine and hyper-feminine individuals, show lower levels of support, especially among the hyper-masculine group, where only 57% support adoption rights. These findings support the idea that less stereotyped gender identities are associated with greater openness to sexual diversity and stronger support for the rights of gay and lesbian individuals.

While women overall are slightly more supportive than men of gay and lesbian rights, the differences become clearer when comparing those for whom gender is personally significant. Among men who say gender is important to their identity, support is notably lower: only 83% agree that gay and lesbian people should be free to live as they wish (vs. 89% of all men), and 76% express no shame if a close family member is gay or lesbian (vs. 83% of all men). Among this group of men, similar to those with hyper-masculine identities, support for adoption by gay and lesbian couples is lowest at 57%. Among women for whom gender is important, support levels also dip slightly compared to all women—particularly in terms of tolerance (82% vs. 84% report no shame)—but remain high overall.

Conclusion

What does this tell us about the state of gender identity in the UK today?  The findings paint a picture of a society that is still somewhat shaped by traditional gender norms, but with apparent signs of gradual change—especially among younger and more educated groups. Most men continue to identify as strongly masculine and not feminine, while most women identify as strongly feminine and not masculine. However, most respondents do identify, even weakly, with the other, and a sizeable minority do not consider their gender important to how they think of themselves. Femininity is only weakly associated with traditionally feminine traits like compassion and helping others. This suggests that either people do not strongly associate these traits with their feminine identity, or that femininity is being defined in broader or more varied ways than traditional norms suggest. We also see that masculinity means different things for men and women. For women, identifying as masculine is meaningfully tied to traditionally masculine traits like risk-taking and leadership, but this is not the case for men.

Gender identities are strongly shaped by age and education. Because we only have a snapshot in time, we cannot draw any firm conclusions about changes in gendered traits and perceptions of gender identity over time. However, our results indicate that there are differences between how younger and older age groups see themselves in relation to their gender identity. Younger men and women have more flexible perceptions of their gender identity and consequently they are less polarised in how they see themselves along masculine and feminine lines. These age differences may be due to younger respondents being socialised at a time where women’s role in the workplace is firmly embedded and where men are taking on more responsibilities in the home. Thus, there is a less clear demarcation between expected roles for men and women. However, what is less clear is whether these more flexible notions of gender will change as the younger group ages: as they enter into relationships, have children etc., will these life events have an effect on how they perceive their gender identity? If this happens, they may then begin to look more like our older respondents and have more rigid or polarised notions of gender identity. Beyond age, education emerges as a key factor in shaping gender identity. Higher levels of education are associated with a softening of rigid gender roles, with both men and women expressing more moderate and less stereotypical gender self-concepts. This pattern could point to the role of socialisation and exposure to more diverse perspectives on gender that higher education affords as an important force in reshaping gender identity.

What are the implications for public policy, especially around gender equality and support for LGBTQ+ rights? While before there has been a focus on differences between the attitudes of men and women, gender identity plays a crucial role in understanding these differences and is an important force in itself. For most policies we find majority support, except for equal parliamentary representation for men and women, which shows a degree of consensus across gender and gender identities. The persistence of the normative gender binary continues to shape how policy gaps are framed and understood—most often in terms of “men versus women.” Nevertheless, our analysis of the responses to the gender equality policy questions in the ESS clearly shows that some attitudes towards these policies are shaped by how individuals attach meaning to their gender. In general, these policies have more chance of gaining support in a society where people hold less rigid conceptions of gender identity. In future, therefore, the success or otherwise of these policies in Britain will in no small part be influenced by how conceptions of gender identity develop. Policy-makers on both sides of the debate would be well-advised to pay attention to these trends when assessing whether policies on gender equality or LGBTQ+ rights will garner public support or opposition.

Acknowledgements

Susan Banducci’s work on this report is supported by the European Research Council (ERC) under the EU’s Horizon Europe 2020 research and innovation programme (grant agreement no. 101019284).

The European Social Survey (ESS) is a European Research Infrastructure Consortium (ERIC). Participating countries contribute to the central coordination costs of the ESS ERIC as well as covering the costs of their own fieldwork and national coordination. In the UK these funds are provided by the Economic and Social Research Council (ESRC). The views expressed are the authors alone. 

References

Alexander, Amy C., Catherine Bolzendahl, and Patrik Öhberg. 2021. ‘Gender, Socio-Political Cleavages and the Co-Constitution of Gender Identities: A Multidimensional Analysis of Self-Assessed Masculine and Feminine Characteristics’. doi:10.1332/251510820X16062293641763.

Andrew, Alison, Oriana Bandiera, Monica Costa Dias, and Camille Landais. 2024. ‘Women and Men at Work’. Oxford Open Economics 3(Supplement_1):i294–322. doi:10.1093/ooec/odad034.

Clements, Ben, and Clive D. Field. 2014. ‘Public Opinion toward Homosexuality and Gay Rights in Great Britain’. Public Opinion Quarterly 78(2):523–47. doi:10.1093/poq/nfu018.

Coffé, Hilde. 2019. ‘Gender, Gendered Personality Traits and Radical Right Populist Voting’. Politics 39(2):170–85. doi:10.1177/0263395717745476.

Donnelly, Kristin, and Jean M. Twenge. 2017. ‘Masculine and Feminine Traits on the Bem Sex-Role Inventory, 1993–2012: A Cross-Temporal Meta-Analysis’. Sex Roles 76(9):556–65. doi:10.1007/s11199-016-0625-y.

Eagly, Alice H., Christa Nater, David I. Miller, Michèle Kaufmann, and Sabine Sczesny. 2020. ‘Gender Stereotypes Have Changed: A Cross-Temporal Meta-Analysis of U.S. Public Opinion Polls from 1946 to 2018.’ American Psychologist 75(3):301–15. doi:10.1037/amp0000494.

European Social Survey European Research Infrastructure (ESS ERIC). 2024. ‘ESS11 - Integrated File, Edition 2.0’. Sikt – Norwegian Agency for Shared Services in Education and Research. doi:10.21338/ESS11E02_0.

Falomir-Pichastor, Juan Manuel, Jacques Berent, and Joel Anderson. 2019. ‘Perceived Men’s Feminization and Attitudes Toward Homosexuality: Heterosexual Men’s Reactions to the Decline of the Anti-Femininity Norm of Masculinity’. Sex Roles 81(3):208–22. doi:10.1007/s11199-018-0985-6.

Gidengil, Elisabeth, and Dietlind Stolle. 2021. ‘Beyond the Gender Gap: The Role of Gender Identity’. The Journal of Politics 83(4):1818–22. doi:10.1086/711406.

Hatemi, Peter K., Rose McDermott, J. Michael Bailey, and Nicholas G. Martin. 2012. ‘The Different Effects of Gender and Sex on Vote Choice’. Political Research Quarterly 65(1):76–92. doi:10.1177/1065912910391475.

Inglehart, Ronald, and Pippa Norris. 2003. Rising Tide Gender Equality and Cultural Change Around the World. Cambridge: Cambridge University Press.

McDermott, Monika L. 2016. Masculinity, Femininity, and American Political Behavior. Oxford: Oxford University Press USA - OSO.

Shu, Xiaoling, and Kelsey D. Meagher. 2025. ‘Mind the Gap: Gender, Racial, and Educational Differences in American Gender Attitudes from 1977 to 2018’. Sex Roles 91(2):1–20. doi:10.1007/s11199-025-01566-6.

Wängnerud, Lena, Maria Solevid, and Monika Djerf-Pierre. 2019. ‘Moving beyond Categorical Gender in Studies of Risk Aversion and Anxiety’. Politics & Gender 15(4):826–50.

Whitley, Bernard E. 2001. ‘Gender-Role Variables and Attitudes Toward Homosexuality’. Sex Roles 45(11):691–721. doi:10.1023/A:1015640318045.

Appendix

Full text for questions included in Table 8

To what extent are you in favour or against firing employees who make insulting comments to women in the workplace?

Imagine a couple who both work full-time and earn roughly the same amount, and now have a newborn child. Both are eligible for paid leave if they stop working for some time to care for their child. Still using this card, to what extent are you in favour or against a legal measure that requires both parents to take equal periods of paid leave to care for their child?

Still using this card, to what extent are you in favour or against making businesses pay a fine when they pay men more than women for doing the same work?

To what extent are you in favour or against a legal measure that requires dividing the number of seats in Parliament equally between women and men?

Full text for questions included in Table 9

‘Gay men and lesbians should be free to live their own life as they wish’

‘If a close family member was a gay man or a lesbian, I would feel ashamed’

‘Gay male and lesbian couples should have the same rights to adopt children as straight couples’

Publication details

Clery, E., Curtice, J. and Jessop, C. (eds.) (2025)
British Social Attitudes: The 42nd Report. 
London: National Centre for Social Research  

© National Centre for Social Research 2025

First published 2025

You may print out, download and save this publication for your non-commercial use. Otherwise, and apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, this publication may be reproduced, stored or transmitted in any form, or by any means, only with the prior permission in writing of the publishers, or in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the National Centre for Social Research.

National Centre for Social Research 
35 Northampton Square 
London  
EC1V 0AX
info@natcen.ac.uk 
natcen.ac.uk