Feminists are divided on proposed reforms to the GRA and sex and/or gender self-determination. What are the implications for people born female?
Co-authored by Holly Lawford-Smith & Emily Vicendese
In this piece, Lorna Finlayson, Katharine Jenkins, and Rosie Worsdale give what they call ‘a feminist case against the feminist case against trans inclusivity’. Their title effectively acknowledges a deep schism within feminist theory and activism. On one side, Type A feminists argue that there is no significant distinction between trans women and non-trans women regarding their sex and/or gender (aside from their claim that trans women are oppressed by non-trans women); whereas on the other side, Type B feminists argue that the acknowledgment of important differences between trans women and non-trans women regarding their sex and/or gender is necessary for resisting sex and/or gender-based oppression.
Finlayson et al.’s reply is framed as a response to some of the recent pushback against proposed reforms to the UK’s Gender Recognition Act (GRA). Essentially, the GRA seeks to make changing one’s legally recognised sex much less onerous by replacing medical gatekeeping with self-identification alone. However, this feminist schism has a long and complex history, hence this latest dispute over the GRA can be seen as a contemporary skirmish in the ongoing Gender Wars. Type A feminists are on the side of the proposed reforms to the GRA because they endorse the claim that one’s sex and/or gender should be entirely self-determined. Type B feminists are against the proposed reforms to the GRA, arguing that leaving legal sex and/or gender categories open to pure self-identification will have a variety of unwelcome outcomes, especially for people born female.
Type B feminists—ourselves among them—have complained about the extent to which Type A feminists refuse to engage on this issue, instead resorting to insults, ostracism and accusations of bigotry, transphobia, right-wing collusion, and the like. Although Finlayson et al. do not manage to entirely avoid making such accusations, the bulk of their article is a sincere and welcome attempt to engage. The substance of their claim in the article is that our arguments (specifically, ours in this piece, and Kathleen Stock’s in this piece) ‘go[…] astray’ at various points. We think they characterise our arguments more or less correctly, but that their case against those arguments is weak, and in one important place, question-begging. In what follows, we offer a reply that is focussed on four main points: the interaction between gender self-identification and sex-based protections in law; the relation between gender identity and male violence; the assumption that trans women are women; and the claim about the infeasibility of third spaces.
Before addressing those four points, a clarification about the emphasis on male violence in our second point is required (we include sexual misconduct as a form of violence). Our previous article was about male violence, so it makes sense that Finlayson et al. focussed their critique on that. However, our analysis of male violence was not intended to exhaust all the issues relevant to the dispute between Type A and Type B feminists. Female-only spaces and events are about more than just safety from male violence. Even if it could be established that trans women pose no more threat of violence to non-trans women and girls than other non-trans women and girls, it is not clear that non-trans women should include trans women in all female-only spaces.
We believe that female-born people have the right to define themselves as such and have the right to associate with each other on that basis in certain contexts or for particular purposes. We oppose efforts to proscribe such collective self-determination and freedom of association. Some may be tempted to argue that, within the context of widespread transphobia, such self-determination and free association is equivalent to supporting White-only spaces and facilities such as those under Jim Crow in the US. We think such comparisons are patently spurious. Given the historical and ongoing subjugation of female-born people by male-born people, female-only spaces are more like African American-only spaces than White-only spaces, or, for that matter, more like trans- or queer-only spaces. A nightclub for gay men that excludes women and straight men is very different to a nightclub that excludes gay men, for obvious reasons. Despite being over fifty percent of the population, female-born people are a vulnerable group. For vulnerable groups, certain spaces and events are invaluable for solidarity, for shared and often unspoken understanding, and for time out from the burdens and risks associated with interacting with potentially hostile and prejudicial members of the dominant group. It is thus our strong belief that governments must continue to recognise the legal categories of biological sex and continue to provide resources and exemptions for female-born people on this basis, and that institutions, organisations and businesses should feel likewise empowered to do the same, where relevant.
” For vulnerable groups, certain spaces and events are invaluable for solidarity, for shared and often unspoken understanding, and for time out from the burdens and risks associated with interacting with potentially hostile and prejudicial members of the dominant group.”
With that said, we can now turn to the four points outlined above. We will argue that there is a lack of clarity in UK law about the interaction between gender self-identification and sex-based protections; that gender identity is insufficient for distinguishing trans women from males for the purpose of protecting female-born people from male violence; that the assumption that trans women are women does not defeat the precautionary principle; and that Finlayson et al. are mistaken about the infeasibility of third spaces.
Self-identification and sex-based protections in law
Finlayson et al. present the proposed changes to the GRA in a way typical of the Type A feminist side, namely by downplaying its significance. “The implications of having a GRC are less substantial than is often supposed”, they say; it’s mainly about eliminating the distress caused by things like having to get married with the wrong sex (they say ‘gender’) recorded on your marriage certificate, or eliminating the contradictions between different identity documents that might ‘out’ a person as trans. The authors acknowledge that having a gender recognition certificate (GRC) makes trans prisoners more likely to be placed in the prison estate that aligns with their gender identity rather than their sex. But this is presented as part of the list of ways in which reforms to the GRC make life better for trans people.
There is no acknowledgement of the concerns that Type B feminists (and other people) have about the implications of a shift to sex self-identification. For example, the authors fail to note that only 4,910 people in the UK have GRCs under the current law, and that a shift to sex self-identification would be expected to increase that number to between 200,000 – 500,000 (which is the number of trans people in the UK according to a recent estimate by the UK Government Equalities Office. That’s a big increase in population, and it might make a big difference to the numbers of trans women in the women’s prison estate. If trans women exhibit male pattern violence, this change would have serious impacts on female people. Additionally, housing trans men in the men’s prison estate, when most trans men have a vulva and smaller stature relative to males, may be another way in which more female born people are put at increased risk. Male-born people simply do not face a comparable risk. While we do not necessarily endorse the prison system in general, especially its more punitive and dehumanising aspects, these impacts should be faced head on, and discussed.
In a similar vein, Finlayson et al. do not mention the uncertainties that exist around the interaction of the GRA and the Equalities Act. Even the Government, in their consultation document, acknowledge these uncertainties and say they ‘want to find out more’ about them (Item 10, p. 12). The Equalities Act protects both ‘sex’ and ‘gender reassignment’, but contains an exemption allowing for the exclusion of people with gender reassignment from single-sex services so long as that is a ‘proportionate means of meeting a legitimate aim’. For example, those running a women’s rape shelter might refuse to accommodate a trans woman. However, a GRC creates a legal fiction that its holder is in fact the sex it states. So it’s not clear whether this exemption applies to trans women who are legally female, or only to those who count as having ‘gender reassignment’ but do not have a GRC. Operators of single-sex services are wary of invoking the exemption, not knowing whether they have a clear legal right to do so (not to mention that some are disincentivized from doing so by the aggressive social sanctioning they might face from trans activists were they to try). If the exemption cannot be used to exclude people with a GRC, and the number of people with a GRC stands to increase massively with the proposed reforms to the GRA, then this is a serious threat to single-sex services and the interests of the women who use them, and who wouldn’t use them were they mixed sex (which includes survivors, religious women, and women from particular ethnic groups). There haven’t been legal tests of this interaction between the two protected characteristics, and the legal situation is unclear even to barristers and legal theorists. So it is far from unreasonable for Type B feminists (and others) to worry about the effects that a change to the GRA will have on the sex-based interests currently protected in the Equalities Act.
“[P]roposed reforms to the GRA [could pose]…a serious threat to single-sex services and the interests of the women who use them, and who wouldn’t use them were they mixed sex (which includes survivors, religious women, and women from particular ethnic groups).”
Relationship between gender identity and male violence
Finlayson et al. agree with Type B feminists that “men as a group systematically oppress and inflict violence on women”. The authors also agree that trans women have something in common with men, both in terms of male biology (karyotype, phenotype, hormones, etc.), and in terms of a history of male socialiation. It would be very surprising if they didn’t agree on this, given that being male is necessary for being a trans woman. Most male people are accurately perceived by others to be male, and on this basis subject to male socialisation. Trans women, especially late transitioning trans women, have often spent more of their lives being treated as boys or men than they have being treated as women. And male violence has to come from somewhere: it’s got to be biology, socialisation, or (much more likely) a highly complex and dynamic causal interaction between the two.
However, in their section titled ‘The Basis of Risk’, Finlayson et al. “dispute the claim that these features—either singly or in conjunction—are the correct basis for determining the risk that an individual poses in terms of violence against women”. Instead, the authors argue that gender identity, the key distinguishing feature between trans women and non-trans males, might be sufficient for eliminating the risk posed by males to females. The strongest claim the authors make is that “it is, to say the very least, not obvious that gender identity makes no difference to the way in which either biological and social factors manifest themselves” [our emphasis]. They don’t go so far as to claim that that biology and socialisation have no role to play in males’ propensity for violence, but argue that it is highly plausible that gender identity plays a significant role too. If that’s right, then it supports an empirical hypothesis: trans women won’t exhibit the same patterns of violence towards women and girls that other males do.
So what is gender identity, and how could it eliminate a person’s propensity to violence? At the most basic level, to have a gender identity is to take oneself to be a member of a gender category. Trans women are males who take themselves to be members of the category “woman”. Just why people take themselves to be members of a particular gender category is poorly understood and hotly contested: is there a neurological basis? If not, why do some people take themselves to be members of a gender category for which they do not have the corresponding physical features? For our own part, we take ourselves to be members of the category “woman” because we are female, i.e., we each have the karyotype and phenotype necessary for producing ova and becoming pregnant. We don’t take ourselves to have ‘gender identities’ beyond simply applying the concept “woman” to ourselves in accordance with our physical features. But not all females take themselves to be members of the category “woman” (e.g. trans men and some non-binary people), thus what it means to have a gender identity—and even what it means to be a member of a gender category—becomes very unclear. Moreover, it seems that there are many different reasons why people assert that they have gender identities. For some people, it is explained by dysphoria, but not all trans people have dysphoria.
One of the authors, Katharine Jenkins, has written here and here on the subject, arguing that what it is to have a gender identity is to take the culturally particular gender norms in operation in one’s social environment to apply to you. For example, people who have female gender identities are people who feel themselves accountable to norms applied to females, whether or not others take them to be accountable to those norms. For Jenkins, then, gender identity operates as a kind of filter for socialisation, making some males responsive to female gender norms and not responsive to male gender norms. If male gender norms play a role in producing male propensity to violence, but gender identity is highly influential over whether those norms influence behaviour, then it’s plausible that males who have female gender identities (i.e. trans women) may be resistant to male gender norms and thus the influence towards violent behaviours, and susceptible to female gender norms that lessen likelihood of violent behaviours.
There are several problems with this thesis. Firstly, it is an empirical claim, yet it is presented in a highly speculative way. If we had the data that showed trans women as a sub-class of males to not have the same propensity to violence as males as a whole, then Jenkins’ gender identity thesis might serve well as an explanatory hypothesis. But because we do not have this data (and in fact, the limited data we do have suggests the opposite), to use this thesis as a justification for the legal institutionalisation of self-identification is to put the cart squarely before the horse. Moreover, while it’s obviously true that individual trans women adopt a selection of explicitly gendered norms, such as those related to gender presentation (e.g. women’s clothes, pronouns, etc.), it’s not clear that this is always because they have the kind of gender identity that Jenkins postulates. Furthermore, on Jenkins’ proposal the trans woman doesn’t have to take all feminine norms to apply to her, only more feminine norms than masculine norms. So it could be that a person counts on this definition as having the relevant gender identity while not seeing the specific norms against violence as applying to them. The trans umbrella encompasses many different categories of persons, from those who occasionally cross-dress through to those who suffer full-blown dysphoria. Even if some such people have this kind of gender identity, it’s not likely that all do. And it’s even less clear that a gender identity’s proposed norm-filtering effect would apply to norms that are less explicitly gendered, such as expectations about the amount of smiling one does, or how comfortable one is with taking up physical space, or whether one sees violence as an acceptable means to getting what one wants.
Secondly, gender socialisation is not merely the internalisation of norms, but the result of being treated in particular ways due to being perceived to be a member of a particular sex. It is deeply embedded in human history and in all aspects of the way we are treated by other people, from the time we are small babies all the way through our lives. Gender socialisation can be extremely subtle, from differences in the amount of eye contact parents give to boy and girl babies, through to the extremely obvious, like encouraging care-based work in females and leadership in males. Male privilege is granted externally—a male person cannot refuse to be believed more easily than his female counterpart would be; or refuse to be assumed to be cleverer by others than his female counterpart would be; or easily refuse the social network opportunities that are accorded to him because of his sex (and this is partly because he cannot easily know that this is why they are accorded to him). It’s implausible that being systematically treated—or mistreated—in certain ways would not have a significant effect on one’s psychology, in particular one’s self-confidence, and one’s sense of entitlement. It is extremely unlikely that a boy growing up with a female gender identity could simply block all the effects of socialisation that would give other males without that gender identity a higher propensity to violence.
“[G]ender socialisation is not merely the internalisation of norms, but the result of being treated in particular ways due to being perceived to be a member of a particular sex. It is deeply embedded in human history and in all aspects of the way we are treated by other people, from the time we are small babies all the way through our lives.”
Thirdly, gender identity, whether it is sufficient to alter the likelihood of a male ending up violent or not, is not a useful defining feature of gender categories for legal purposes. Because gender identity is a purely mental phenomenon, it is only verified by first-personal introspection. There are no scientific tests that can tell us what someone’s gender identity is. This means that the most reliable way we have of knowing someone’s gender identity is for them to tell us. An analogous case is pain: when deciding how much pain medication to give a patient, hospital staff have to ask patients to rate their pain on a scale of 1-10, because there is no external measure. While observable behaviours may offer some guide, when it comes to internal states, verbal reports will in most cases trump observable behaviours. This means that the idea of gender identity is actually doing very little work for Finalyson et al. in the context of this debate, because all we actually have access to is a person’s self-report, not any fact of gender identity itself. We can’t define a group for legal purposes based on a proposed feature that is unobservable and unverifiable. Even if Finlayson et al. are right about the norm-filtering effect of gender identity (which we think they probably are not), we just don’t know how closely self-report tracks the kind of gender identity that they postulate. It is therefore not a good feature on which to rest their case for the distinction between trans women and other males.
The precautionary approach, and ‘trans women are women’
Even if Finlayson et al. were right that gender identity might be a feature that controls for a lower rater of violence in trans women compared to non-trans males, feminists don’t need to take this mere possibility as decisive against trans women’s exclusion. Pending empirical support for that speculative empirical claim, they would be justified in adopting a precautionary approach. Given everything we know about males, Type B feminists argue, we ought to preserve female-only space at least until we know for sure that there’s a difference-that-makes-a-difference (and maybe even after that; see our further considerations above). The authors, to their credit, anticipate this prospective reply. But despite their prescience, they do not take the opportunity to explain (in the section ‘Playing it Safe’) why in fact a precautionary approach is not justified. They simply sidestep the issue and end up begging the question, arguing that if trans women are women then there’s no case for keeping them out of any women-only spaces or events in the first place. And they say that Type B feminists beg this question too, by assuming that trans women are not women. However, neither of us have arrived at a settled view, just yet, about whether at least some trans women (e.g. fully-passing transsexuals who are treated by the wider society as women) are women. That is because neither of us have arrived at a settled view, just yet, about whether gender can be fully decoupled from sex, so as to allow occupation of the social position ‘woman’ by an alternative route than being female, or not. One of us (Holly) is tempted to think not, whereas the other of us (Emily) leans more toward endorsing the sex/gender distinction and thinking the two conceivably come fully apart. But we don’t need to take a stand on that here. We both agree that trans women are not female (that’s what the “trans” in “trans woman” means), and we think that an adequately intersectional feminism must continue to acknowledge the existence and interests of females qua females, regardless of whether it also understands the meaning of the term “woman” in such a way that includes some or all trans women.
In their opening section, Finlayson et al. write that because some feminists believe that trans women are women, they ‘are part of the ‘constituency’ that is feminism’s primary concern’. And in the section ‘Playing it Safe’, they write that if trans women are women, then their safety matters just as much as non-trans women’s, and that the two cannot appropriately be traded off. So trans women should be included in all female-only spaces. But given that their speculation about gender identity yields an empirical prediction for which there is not yet sufficient confirming evidence, and given that a precautionary approach would ordinarily be justified in light of that, inclusion is effectively an experiment with women’s safety for ideological reasons. In the worst case, it amounts to saying that trans women’s safety matters more than non-trans women’s safety: if including trans women in all female-only spaces and events brings more costs for non-trans women than benefits for trans women, then inclusion prioritises the interests of trans women over non-trans women.
“[I]f including trans women in all female-only spaces and events brings more costs for non-trans women than benefits for trans women, then inclusion prioritises the interests of trans women over non-trans women.
Many Type A feminists argue that this is justified, because non-trans women are a dominant group who should give up their privilege in favour of trans women, the more marginalised and vulnerable group. However, this focuses on the non-trans/trans axis of advantage while squarely ignoring the male/female axis. Female people have been subordinated and exploited by male people at least since the advent of agriculture. That’s the class of female people; not the class of ‘feminine’ people, or the class of people with a female gender identity. In Margaret Atwood’s The Handmaid’s Tale, Gilead does not care about your gender identity: you are selected for subjugation in accordance with your potential reproductive capacities, that is, whether or not you are female. We find it absurd that some feminists, in their hurry to address the mistreatment of transgender people, overlook–or sometimes even actively deny–the ongoing and often brutal reality of sex-based oppression and exploitation. At the very least, Type A feminists should incorporate sex (defined in terms of reproductive role) as an axis of oppression in their intersectional analyses.
Finlayson et al. go so far as to claim that “trans women’s relationship to the structure we call ‘patriarchy’ has more in common with the relation in which cis women stand to that structure than with cis men’s relationship to it”. We reject this claim outright, at least for most cases. Many groups are disadvantaged by patriarchy, including women, men, trans people, gender non-conforming people, and gay, lesbian, and bisexual people. That doesn’t make all of these people somehow in the same class as non-trans women. Gay men, for example, are sanctioned for being gender non-conforming, and are subject to male violence. But that doesn’t mean they should be included within female-only spaces. Female-only spaces are not refuges for the vulnerable (and it’s a tired gender stereotype that women should be kind and caring enough—indeed, inclusive enough—to welcome all vulnerable people into their spaces).
The hostility and violence that trans women are subjected to qua trans women is not a result of being perceived to be women, but comes from being perceived to be men who are “pretending” to be women or are “deluded” about being women (see this paper by Talia Mae Bettcher). This means that despite the claims of trans activists, most trans women are not subject to either misogyny or ‘transmisogyny’. (This is compatible with some trans women being subject to misogyny, i.e. those who do pass as women.) In this regard, trans women have more in common with effeminate gay men than they do with non-trans women. Both effeminate gay men and trans women face extra contempt from misogynist homophobes, because they are seen not only as gender non-conforming males, but as adopting some of the traits of another despised group: women.
Consider how it is more acceptable to be a “tomboy” than a “sissy”—it is considered far worse for a male to stoop to femininity than it is for a female to adopt masculinity. But this is not to say that gender non-conforming males are subject to misogyny any more than to use an ableist slur against an able-bodied person is to subject them to ableism. To hate someone more because they are female-like is misogynistic toward females, not toward the male who is hated for behaving like a female. Thus it is simply false that trans women’s relationship to patriarchy has more in common with non-trans women’s relationship than non-trans men’s relationship. This argument for including trans women as women also fails.
We suspect that readers will have been sympathetic to Finlayson et al.’s claim that “the proposal to instigate ‘third spaces’ is less of a proposal than a way of washing one’s hands of the issue” (in the section, ‘A Political Question’). They say that people on our side often seem sympathetic to the situation trans people are in by proposing third spaces; but in fact, “nobody is seriously proposing to provide them”, which means that in practice, trans people must use the spaces that correspond to their sex. Again, we do not think this is quite right. First of all, when trans activists have demanded third spaces, they have tended to get them (e.g. Topshop making all its changing rooms gender neutral in 2017 after criticism from a single trans activist). Second of all, if trans activists would put their enormous lobbying power into achieving gains such as third spaces—which we need anyway for non-binary people—they would be likely to get them. Consider how much more they are currently attempting to get, and in some quarters, succeeding in getting: some lesbians to accept them as partners, even though they have penises; some feminists to argue for their inclusion in women-only spaces; the government to initiate a shift toward sex self-identification whose initial consultation involved exactly zero women’s groups. (Imagine what Dr Adrian Harrop could accomplish if he turned his attentions from taking down women’s groups’ campaign materials and toward securing gender neutral bathrooms!)
If such spaces were used by trans people, non-binary people, trans and non-binary allies, and non-trans people who couldn’t care less which spaces they used, then it would be simply false that the use of such spaces would force someone to ‘out’ themselves as trans. Indeed, it would be easy enough to nudge non-trans people to use such spaces too, by making them slightly more accessible, convenient, or well-resourced. The University of Melbourne, where we both work, has recently taken this approach in the new Arts West building, including an “All Genders” bathroom between the male and female bathrooms on every floor. Each of the “All Genders” bathrooms has better facilities than the sex-segregated ones, including more privacy, more space, and private basins. We see third spaces as a workable solution to the fierce debate over female-only spaces, which to our mind, should be kept exactly that: female-only.