Written evidence submitted by Cathy Devine [GRA1160]

Introduction

1.1 I am an Independent Researcher. Previously, I worked for 23 years as Senior Lecturer in Sport at the University of Cumbria, five years at the National Coaching Foundation with responsibility for women in coaching, and two years as Head of Sport at the National Federation of Women’s Institutes. My research can be accessed via Researchgate: (16) Cathy Devine (researchgate.net) and Google Scholar: Cathy Devine‬ - Google Scholar‬

1.2 My research is in human rights, equality and inclusion for girls and women in sport. Currently, I am investigating the impact of transgender inclusion policies on the fair inclusion of girls and women in single sex sport categories both nationally and internationally. My work with British Olympians was reported in the Guardian in 2019: British Olympians call for IOC to shelve ‘unfair’ transgender guidelines | International Olympic Committee | The Guardian. My current work focuses on the fair inclusion of girls and women in sports participation in England and the impact of self-identification of gender on this.

Summary

2.1 Single sex sport is legislated for in both the 2010 Equality Act and the 2004 Gender Recognition Act for reasons of fairness and safety in ‘gender (meaning sex) affected activities’.

2.2 Most sports are ‘gender affected’ since from puberty onwards, at all levels from participation to excellence, they involve competitive activity in which the physical strength, stamina or physique of average persons of one sex would put them at a disadvantage compared to average persons of the other sex.

2.3 Even with a GRC transgender participants can or must be legally excluded from single sex sport categories of the opposite biological sex for reasons of fairness and/or safety, primarily to avoid indirect discrimination against females.

2.4 This provision is widely misunderstood or ignored and sport organisations often advise that eligibility into single sex categories should be on a case by case basis according to ‘gender identity’ rather than on a category basis, that of biological sex, even in ‘gender affected’ sports. This may contravene the EA by indirectly discriminating against females with the protected characteristic of sex.

2.5 Sport England data show females are significantly underrepresented in most sports. In team sports 77% of participants are male and 33% female. Football, cricket, rugby union and golf participants are 90%, 89%, 84%, and 84% male; and 10%, 11%, 16%, 16% female, respectively.

2.6 Academic research shows the transsexual population is an extremely small proportion (maybe 1%) of the much larger transgender population which is estimated by Stonewall to be, 1% of the total population. Further, two thirds of the adult transgender population are biological males.

2.7 Because of the much larger pool of males in most traditional sports, using ‘gender identity’ rather than biological sex eligibility criteria would have a significant and disproportionate impact on the % and numbers of transgender people eligible to participate in female sport.

2.8 For all competitive team sports, the % of transgender (biologically male) participants eligible for female sport categories would range from 1% (7 373) to 4.4% (32 932), whereas the % of transgender (biologically female) participants eligible for male sport categories would range from 0.1% (2 174) to 0.2% (4 784).

2.9 For example, the % of transgender (biologically male) footballers eligible for female categories would range from 2.5% (5 528) to 10.5% (24 690), whereas the % of transgender (biologically female) footballers eligible for male sport categories would range from 0.03% (638) to 0.1% (1 404).

2.10 Consequently, despite being clearly laid out in both Acts, the single sex sport provisions seem to be routinely disregarded which appears then to indirectly discriminate against females with the protected characteristic of sex.

Why is the number of people applying for GRCs so low compared to the number of people identifying as transgender?

3.1 We do not have the research to answer this but we do know that the numbers identifying as transgender without undergoing any medical transition (via hormones or surgery) are much larger that the transsexual population who do undergo a medical transition of some kind. Collin et al1 state ‘the empirical literature on the prevalence of transgender highlights the importance of adhering to specific case definitions because the results can range by orders of magnitude. Standardized and routine collection of data on transgender status and gender identity is recommended.’ They find ‘whereas in most studies estimating the prevalence of surgical or hormonal gender affirmation therapy or transgender-related diagnoses, the prevalence estimates generally ranged between 1 and 30 per 100,000 individuals, self-reported transgender identity was orders of magnitude higher ranging from 100 to 700 per 100,000 or 0.1%-0.7%.’ They also find the prevalence of transgender identity in adults is almost twice as high in biological males identifying as women as compared with biological females identifying as men.

3.2 The Government Equalities Office (GEO) acknowledges ‘No robust data on the UK trans population exists. We tentatively estimate there are approximately 200,000-500,000 trans people in the UK2 which equates to a prevalence of 0.3-0.8%. Stonewall3 estimates 1% of the population fall under the ‘transgender umbrella’, not just transsexuals but many other groups including cross dressers.

3.3 This is likely to be the main reason for the disparity. The 5000 GRCs awarded to date probably relate primarily to transsexuals who make up maybe 1% of the much bigger transgender population, the majority of whom socially transition only. Further, many transsexuals appear unhappy about the conflation of the two categories as witnessed by 17 transsexuals who state: ‘Replacing the evidenced-based process for obtaining a gender recognition certificate with an over-the-counter style self-declaration blurs the distinction between us and transgender people who remain physically intact’4.

Recommendation:

Government agencies including the Census and the ONS collect baseline data for both transsexual and transgender prevalence since the two are often conflated. The two distinct categories are of central importance for policy and planning purposes, not least in sport.

Are there challenges in the way the Gender Recognition Act 2004 and the Equality Act 2010 interact? For example, in terms of the different language and terminology used across both pieces of legislation.

4.1 There are language problems throughout the legislation and to avoid confusion and misrepresentation it is essential for this to be clarified in Government guidelines.

4.2 There are three main conceptions of gender in common usage:

‘Gender’ as a synonym for, and conflated with, sex. This is unhelpful, particularly in legislation which requires precision. Biological sex refers to human sexual dimorphism, determined at conception, immutable and unproblematically classified in 99.98% of humans5.
‘Gender’ often means socially constructed historically and culturally contextualised feminine and masculine stereotypes relating to roles ascribed to the female and male sexes within societies7. For example, historically women have been considered emotional rather than rational and therefore suited to domestic rather than public life. Further, women were historically excluded from sport because of concerns it would damage their reproductive capabilities. Boxing for women has only been included in the Olympic Games since 2012 and rugby since 20166. Gender stereotypes change over time.
‘Gender identity’ is a more recent concept meaning an unverifiable subjective inner sense of being feminine or masculine, not necessarily attached to biological sex7.

4.3 Both Acts conflate biological ‘sex’ and ‘gender’ in sport in that they refer to ‘gender affected activities’ when they mean activities affected by biological sex. Further, the Acts are not aligned. The Equality Act correctly defines them as activities in which ‘the physical strength, stamina or physique of average persons of one sex would put them at a disadvantage compared to average persons of the other sex’9 whereas the GRA inaccurately defines them as activities in which ‘the physical strength, stamina or physique of average persons of one gender would put them at a disadvantage to average persons of the other gender’10. Further, the conflation of ‘gender’ with ‘sex’ in the Acts often results in a conflation of ‘gender’ with ‘gender reassignment’ which is a totally separate protected characteristic with a different meaning, and even with ‘gender identity’ which is not a protected characteristic at all, by agencies attempting to comply with both Acts (see for example the Sports Council’s UK Equality Standard8 p4 which replaces sex with gender).

4.4 In essence, both Acts distinguish between biological sex and legal sex in relation to sport in Section 195 (2)9 of the EA and Section 1910 of the GRA given that even with a GRC transsexual people can be excluded from sport categories of the opposite sex ‘if it is necessary to do so to secure in relation to the activity—(a) fair competition, or (b)the safety of competitors’. This provision protects female sport given the widely documented differences between the two sexes due to reproductive biology, and knock-on effects for fairness and safety (Bermon et al 201911; Handelsman Hirschberg and Bermon 201912; Hilton and Lundberg 202013; Knox Anderson and Heather 201814; Thibault et al, 201015).

Recommendations:

Government to issue guidance clarifying the following:

the difference between biological sex, and legal sex based on a GRC;
that in sport it is biological sex rather than legal sex which is relevant for ‘gender affected activities’ and/or to update both Acts to refer accurately to ‘sex affected activities’;
the difference between biological sex and socially constructed gender;
the difference between the binary biological sex and the many subjectively experienced ‘gender identities’;
the difference between ‘gender reassignment’ which is a protected characteristic, and ‘gender identity’ which is not;
the difference between transsexual (relating to a diagnosis of gender dysphoria and medical transition) and transgender (relating to a subjective ‘gender identity’ of which there are many)

Government to require relevant agencies to comply with the law and correctly document the protected characteristics as listed in the Equality Act rather than interpret these using alternative confusing terminology.

Are the provisions in the Equality Act for the provision of single-sex and separate-sex spaces and facilities in some circumstances clear and useable for service providers and service users? If not, is reform or further guidance needed?

5.1 Leaving aside problems with terminology, provisions in the Acts for single sex sport in ‘gender affected activities’ for the purposes of fairness and safety primarily for females are reasonably clear, but widely misinterpreted or ignored.

5.2 The EA clearly states ‘A person does not contravene this Act, so far as relating to sex, only by doing anything in relation to the participation of another as a competitor in a gender-affected activity’9. Therefore, in a ‘gender affected activity’ (meaning ‘sex affected activity’) single sex sport provision ‘in relation to the participation of another as a competitor’ is lawful and presumably required in order to avoid indirect discrimination against females. ‘Gender (meaning sex) affected activities’ are defined as ‘a sport, game or other activity of a competitive nature in circumstances in which the physical strength, stamina or physique of average persons of one sex would put them at a disadvantage compared to average persons of the other sex as competitors in events involving the activity’.

5.3 This provision quite rightly refers to the whole sex category (female or male) and not to individuals within categories. Given the average male in most sports will outperform the average female because of the performance advantages of male puberty, the provision does not refer to individuals. This is important given averages, by definition, capture the whole range in a data set meaning that at any given centile, males will outperform females, so males at centile 4 in a male performance hierarchy will outperform females at centile 4 in the female performance hierarchy (for example in under 16’s football). In this way averages determine statistically significant differences between overlapping categories, in this case between males and females from the onset of puberty, at all levels from participation to excellence.

5.4 However, there is widespread misunderstanding and/or disregard for this provision in the Acts, possibly on the basis of poor legal advice. For example, the ‘LTA Policy and guidance on trans people playing tennis’16 states:

5.4.1 ‘You should accept people in the gender they present… should someone have… genuine and substantive concerns about the woman’s physical strength, stamina or physique that may put other female competitors at a disadvantage, they should refer their concerns to the LTA Safe and Inclusive Tennis Team.’ Consequently, despite tennis being played in single sex categories for reasons of fairness in relation to strength stamina or physique, the LTA permits individuals to play in opposite sex categories on the basis of self-identified gender identity unless anyone complains.

5.4.2 This is problematic. The EHRC explains ‘a trans woman who does not hold a GRC and is therefore legally male would be treated as male for the purposes of the sex discrimination provisions, and a trans woman with a GRC would be treated as female’17. Further, a transwoman with a GRC who has therefore changed legal sex may, or even must, still be excluded from female sport categories, for fairness and/or safety reasons. If tennis is played in single sex categories, which it is, this is only legal if it is a ‘gender (meaning sex) affected activity’. If it is a ‘gender affected activity’, transwomen with or without a GRC would be ineligible for female categories presumably for reasons of fairness in order to avoid indirect discrimination against females. Further, permitting some but not all biological males to compete in female categories may constitute direct discrimination against other biological males. This is particularly so in sports participation settings where transgender people are likely to have made no medical transition at all. Yet, the LTA appears to disregard Section 195 of the EA by advising inclusion of some males (who self-identify their gender) within female categories according to the gender (presumably ‘gender identity’ rather than sex) in which they present. This appears to contravene the requirements of the EA and may indirectly discriminate against females.

5.5 The legal test for single sex categories in order to avoid indirect sex discrimination against females, is that activities are ‘gender affected’. If this is disregarded and ‘gender identity’ is used instead of biological sex to determine eligibility into single sex sport this will have a significant disproportionate effect on female categories as follows:

5.5.1 The sex gap in sports where competition is integral to participation is significant and presented using Sport England Data May 2018/1918.

5.5.2 Consequently, eligibility based on ‘gender identity’ rather than biological sex where competition is integral to participation disproportionately affects female categories.

Sport	% Transwomen in Female Categories1	Numbers of Transwomen in Female Categories1	% Transmen in Male Categories1	Numbers of Transmen in Male Categories1

Team Sports	1.0-4.4	7 373 - 32 932	0.1 – 0.2	2 174 – 4 784

Football	2.5 – 10.5	5 528 – 24 690	0.03 – 0.1	638 – 1 404
Cricket	2.4 – 9.8	779 - 3 481	0.04 – 0.1	96 - 212
Rugby Union	1.5 – 6.5	563 - 2 515	0.1 – 0.1	110 - 242
Golf	1.5 – 6.5	2 457 - 10 976	0.1 – 0.1	479 – 1 055
Basketball	1.1 – 4.9	665 - 2 968	0.1 – 0.2	174 - 383
Rugby League	0.6 – 2.7	128 - 571	0.1 – 0.3	61 - 135
Track and Field Athletics	0.6 – 2.5	412 - 1 841	0.2 – 0.4	221 - 485
Tennis	0.5 – 2.1	1 378 – 6 156	0.2 – 0.4	886 – 1 949

Note: Lower estimate uses the lower GEO2 0.3% transgender estimate and assumes equal numbers of transwomen and transmen. Higher estimate uses Stonewall’s 1% transgender estimate3 and the Collin et al1 distribution of a 67/33 transwomen/transmen split, equating to a population prevalence of 1.34% for transwomen and 0.66% for transmen. All estimates use Sport England sex participation data for May 2018/1918.

5.5.3 It appears permitting ‘gender identity’ eligibility for single sex sport categories meets neither the legal requirements of both Acts in relation to ‘gender affected activities’ or represents a proportionate means of achieving a legitimate aim, the legal test for the other single sex provisions in the Acts.

Recommendations:

Government issues guidance for the Sports Councils and governing bodies of sport clarifying:

reference to ‘gender affected activities’ refers to activities affected by biological sex not legal sex or ‘gender identity’;
these are legally permitted for reasons of fairness and safety to avoid indirect discrimination against females;
transgender people who have changed legal sex can or even must be excluded from categories of the opposite sex for reasons of fairness and/or safety.
this provision relates to sex categories rather than case by case assessment.

Academic publishing

Finally, both Acts are widely misunderstood or misrepresented within the academic literature. Government guidance and clarification regarding terminology and legal provision of single sex sport specifically to ensure no indirect discrimination against females on the basis of biological sex, would be welcome. I have had comments from two academic reviewers of my work19, looking at the impact of self-identification of gender on the sports participation of girls and women in relation to the Acts, stating the following:

‘The use of ‘the biological differences’ and ‘the sexes’ is particularly reductionist.’

‘Putting biologically male and biologically female next to transgender is transphobic’.

‘The Authors should remember that not just gender, but sex too, is a social construction.’

‘Using the term opposite biological sex is not only sexist, homophobic and transphobic, it is biologically incorrect’

‘Steps need to be taken to rewrite much of this article to avoid contributing to the moral panic over trans women competing in women’s sport’

If academic work exploring the fair, safe and legal inclusion of girls and women in sport in relation to the law is rejected for such reasons this amounts to censorship and a restraint on academic freedom.

Recommendation:

Government issues guidelines for academic publishers and universities clarifying the terminology and content of both Acts with respect to the legal requirement for single sex sport for reasons of fairness and safety of female participants in ‘gender affected activities’.

November 2020

References