Gender-Based Violence Without a Legal Gender: Imagining Single-Sex Services in Conditions of Decertification

This article considers what the implications of decertification would be for single-sex services such as domestic and sexual violence support. Some reform options attached to decertification could (re)allocate authority away from the state to organisations or individuals to determine gender criteria. What would the consequences of such re-allocation be in determining eligibility to receive or access services or excluding people on the basis of a characteristic protected under equality law? Engaging with this in the context of domestic and sexual violence support service provision raises a number of questions. Firstly, does the existence of gender-based violence and/or of the effects it produces require a stable category in order to address them? What benefits may emerge from providing single-sex spaces that could not be replicated in other settings? And finally, what criteria of exclusion and inclusion are currently used to determine access to spaces beyond legal gender status?


Introduction
The question of how single-sex services should be organised and who should be included within them, is an issue receiving significant attention in current public, policy and legal debates. Drawing on research carried out as part of the 'Future of Legal Gender' project (FLaG) between 2018 and 2022, this article considers what the implications for single-sex 1 services are if sex, and therefore legal gender status, were to be removed from birth certificates but equality law, like the Equality Act 2010, were left in place but revised to address systemic inequality. In particular, what the consequences would be for reform options which (re)allocate authority to organisations or individuals to determine gender criteria and individual status in terms of eligibility to receive or access services. The data drawn on in this article have emerged from documentary material as well as semi-structured interviews conducted between 2018 and 2020 with policy experts and employees or managers from different domestic and sexual violence non-governmental organisations (NGOs) across England and Scotland. Although this is a relatively small sample size, it offers valuable qualitative data around these issues and resonates with wider public debates happening in this area. 2 Part of the challenge of a prefigurative law reform project (see also Cooper 2023a; Grabham 2023) arises from the fact that it requires speculation about future law reform proposals and their consequences. Therefore, part of this project involves extrapolating from existing legal challenges and social issues. Although FLaG considered what the decertification of legal gender would mean for everyone-not just for trans and non-binary people-looking at how organisations, spaces and institutions currently deal with the question of how and whether trans and non-binary people should be included within single-sex spaces can help to illuminate some of the challenges that may emerge (see also Newman and Peel 2022). Unlike legal gender, several other protected characteristics under the Equality Act 2010 are not currently underpinned by a formal legal status. Therefore, legal disputes over exclusion or inclusion or one's belonging within a certain group seem to be more frequent and consequently existing jurisprudence in these areas provides some guidance as to the basis on which groups can exclude others. 3 In contrast, to date there seem to have been no reported cases about the definition of sex employed by any specific group or organisation, and as such it is less clear what a court might deem an unacceptable definition of sex in an equality context. 4 However, given the scrutiny of trans and non-binary inclusion within single-sex services at the present moment, drawing on these existing debates can provide a basis from which to extrapolate the types of concerns, challenges, and needs that may come to the fore if sex/gender were no longer underpinned by a formal legal status. This article offers an in-depth analysis of service providers' existing thinking and practices in the context of single-sex spaces, including the currently limited relevance of legal gender status as recorded on birth certificates for access to these services. This article highlights the significance competing approaches to managing spaces, particularly a new emphasis on privacy and risk management, may take on in the context of the decertification of Gender-Based Violence Without a Legal Gender: Imagining… legal gender status (see also Emerton 2023). Simultaneously, it seems important not to underestimate the impact of sustained funding pressures and resource scarcity in shaping how organisations choose to organise and their spaces in the future if current funding trends that favour mixed-sex provision continue.
Currently, public and media attention is frequently focused on whether singlesex spaces, particularly women's refuges, should include trans women. 5 Critics of proposals to simplify existing legal procedures for changing one's legal sex/gender status fear that eliminating the formalised and binary-differentiated status of women and men will hinder women's struggles for justice, equality, and emancipation; and undermine single-sex services, spaces, and activities for women (see Emmons 2019; Murray and Hunter Blackburn 2019). The increased attention to this issue seems to have been triggered, at least in part, by concerns arising around the consultation around the Gender Recognition Act 2004 (GRA). The consultation asked lay people whether changes to the GRA, the legal framework that allows a person to change their gender, would also impact the workings of the Equality Act 2010 regarding single-sex services (see also Murray and Hunter Blackburn 2019;Cowan et al. 2020;Renz 2020;Sharpe 2020). These conflicts involve disputes about what sex/ gender categories mean, who gets to make decisions and policies about them, and who gets to determine individual placement inside or outside a given sex/gender category (see also Sharpe 2020). 6 Detailed information about the issues was not included in the guidance, 7 which may also raise the question whether there are more systemic issues here, especially around public engagement with and understanding of law.
As such, the single-sex strand of the FLaG project raises a number of crucial questions. Namely, how can we account for the workings of structural gender-based violence? Does the existence of violence and/or of the effects it produces require a stable category or stable categories to address them? What benefits may emerge from providing single-sex spaces that could not be replicated in other settings? And finally, what criteria of exclusion and inclusion are currently used to determine access to spaces beyond legal gender status? In particular, this final question may point towards the emergence of other mechanisms of exclusion, such as risk management or a greater utilisation of private spaces, that might become embedded as an alternative to a biological definition of 'single-sex' if sex/gender was to lose its formal legal underpinnings (see also Cooper and Emerton 2020). In the following 5 I am using 'trans' rather than 'transgender' or 'transsexual' to accommodate a variety of identity categories that may not necessarily align with a binary understanding of sex and gender, encompassing identities such as genderqueer and agender more readily than the medically defined 'transsexual'. 6 In particular, Question 13 of the consultation asked: "Do you think that the operation of the single-sex and separate-sex service exceptions in relation to gender reassignment in the Equality Act 2010 will be affected by changing the Gender Recognition Act?" (Minister for Women and Equalities 2018, 46). The government's own report published the following year on the workings of the Equality Act 2010 noted that any changes to the GRA would not affect the Equality Act (Women and Equalities Committee 2019, 172). 7 Although of course a government consultation should ask for public opinions, it is somewhat unclear why the government felt the need to ask a lay people about the interaction between two laws that in principle have very clear boundaries and also use different definitions of what it means for someone to change their legal status. sections, this article will provide an overview of the methodological approach underpinning the data that will be used in the later sections. The second part of this article sets out the existing legal structures that shape the legal basis on which single-sex spaces can be set up and maintained. The main analysis of the interview data that follows has been split into two parts: firstly, an outline of how service providers understand and approach gender-based violence; and secondly, an overview of alternative strategies for managing the safety of refuge spaces if legal sex status cannot be relied upon.

Method
This strand of the FLaG project focused on conducting semi-structured one-to-one interviews with experts involved in the provision of or setting policy for singlesex services between 2018 and 2022 (see further Cooper et al. 2022). 8 This article will focus on those conducted with domestic violence and sexual violence services. Participants were recruited through (a) requests for expressions of interest in interview participation circulated via relevant mailing lists; (b) snowball sampling based on recommendations from previous interviews (Heckathorn 2011); (c) direct approaches where email addresses or other contact information was publicly available. Further, all publicly listed single-sex domestic and sexual violence services in England and Wales were approached via email to gauge interest in interview participation.
This resulted in interviews with two policy experts and 14 members of staff from different domestic violence service providers across England and Wales. Additionally, a range of academic experts were consulted on a more informal basis to test the validity of some of the core interview questions. Interviews were conducted either in person or via video conferencing software or by telephone where this was the interviewees preference. The interviews were audio-recorded with the participants' permission.
As interview participants took part in their relevant professional capacity, no biographical information on participants was collected as the aim was not to have a representative population sample but rather to gather participants' expert views. Interviewees were asked questions about their work in single-sex settings and the role of legal gender in this context, as well as their expert opinion on different legal reform options in the UK, for example: "does the legal differentiation of women/girls and men/boys come up often in your work?", "should organisations have autonomy to determine what sex/gender means to them?", and "how do you currently deal with instances where people challenge gender differentiation?". The interviews were transcribed verbatim either in full or in part, depending on the content of each individual interview. Transcripts were then analysed using thematic analysis (Guest et al. 2011) to identify and analyse core themes emerging from interviews both individually and collectively in terms of the information they provided in response to the strand's research questions.

The Current Legal Framework
Single-sex religious, occupational, and cultural communities have a long history both in the UK and in other jurisdictions. While some articulate conservative gender norms, others are more radical or non-conforming: from the 1980s flourishing of lesbian feminist communities to peace camps, and efforts to establish women's lands (see Roseneil 2000;Browne 2009). In line with a longstanding history of women-only spaces and women's activism, the Equality Act does not prevent the creation or maintaining of single-sex spaces or services, which de facto discriminate as they exclude the 'opposite' sex and therefore would in principle amount to sex discrimination under s11 of the Equality Act. S27 of the Equality Act sets out that the creation of single-sex spaces is permissible where such "limited provision is a proportionate means of achieving a legitimate aim." The official explanatory notes to the Equality Act also outline that domestic violence services are indeed one instance where exclusion serves a legitimate aim, namely the provision of appropriate domestic violence services, even in instances where a local authority would choose to provide such services exclusively for women. 9 Existing empirical research on domestic violence service provision indicates that victims and survivors strongly benefit from the availability of refuge spaces and more so than from other types of services such as counselling (see e.g. Grossmann et al. 2010;Bowstead 2019;Hughes 2020;Wood et al. 2022). By their nature these services in turn are more likely to be run on a single-sex basis both for female and male victims and survivors. As such, single-sex services have a clear legal basis on which to exclude members of the opposite sex, as well as potentially trans victims and survivors. Nevertheless, existing research demonstrates that trans inclusion within single-sex services is not an uncommon practice (see Pain 2021; Women's Aid 2022). At the same time, other concerns, particularly around the availability of a diverse range of services, can lead funders (mainly local authorities) to give preference to mixed-sex services. For instance, in early 2021, Brighton and Hove Council decided to withdraw a £5 million contract from the domestic violence charity RISE after 26 years (Booker-Lewis 2021). This decision was made, at least in part, on the basis that RISE offered women-only services, while other comparable providers had agreed to make services available to men as well. In contrast to some of the concerns raised by critics of trans inclusion, RISE were operating on a trans-inclusive model and in fact set up the first UK LGBTQ+ refuge in 2016 (RISE 2021), but the decision to award funding was based primarily on the capacity to provide services to men.
A similar decision to change service providers from specialist women-only services to more general service provision from a large non-specialist provider also took place in Scotland around the same time (McVey 2021). These moves toward favouring mixed or more generic service provision seem to be part of a wider, national trend within the sector (see Nazeer 2021). Instances of funding withdrawal also acutely highlight the financial constraints that third-sector organisations operated under during the period researched. A report published by Women's Aid in 2014 highlighted that 64% of referrals to refuges were already being declined by providers due to insufficient resources (Neate 2014). This is an issue that seems to be continuing on an ongoing basis with a shortfall of 23% in terms of required refuge spaces as of 2021/2022 (Women's Aid 2021).
Overall, since 2010 all domestic violence services have suffered dramatic funding cuts, which has led to both a cutback of services and a loss of specialist, trained staff across the sector (Bennhold 2012;Coy et al. 2007;Sanders-McDonagh et al. 2016). This is despite successive governments pledging to eradicate domestic and sexual abuse (see Home Office 2014a, 2014b. It is crucial to be attentive to the power that funding providers have in exerting pressure on organisations to change the basis on which they provide services. This is particularly so given the wider and enduring climate of government-enacted austerity and the general scarcity of funding for third-sector organisations due to a lack of ring-fencing of funding sources for these non-statutory services (Ishkanian 2014;Sanders-McDonagh et al. 2016). The need to secure funding, therefore, can in some instances lead to a loss of autonomy for third-sector organisations regarding the basis on which they govern their organisation and provide services (Weiner 1991). 10 The approach to decertification taken by the FLaG project presupposes that although decertification would remove formal legal gender status, gender would nevertheless remain a site for state intervention to address inequality (see also Cooper et al. 2022). As such, it seems important to consider how equality law currently tries to define the basis on which single-sex spaces can be offered in order to identify what the effects of decertification might be in this context. While the option to set up single-sex services is covered by equality law, at the same time s.7 of the Equality Act also protects people from discrimination if they have the protected characteristic of 'gender reassignment', and therefore it effectively prohibits discrimination against trans people and, more recently, also non-binary people. 11 This means that, in principle, service users or volunteers for single-sex services who want to use a new name, wear differently gendered clothes, or ask for a new pronoun to be used are protected from discrimination under the Equality Act, regardless of whether they 10 In practice, this may mean that smaller organisations providing specialist services are less likely to obtain vital funding, or to be able to challenge funding decisions, than larger organisations operating on a national scale despite a supposed shift towards a more localised agenda of government (Vacchelli et al. 2015). 11 In Taylor v Jaguar Land Rover Ltd [2020] ET 1304471/2018 the Employment Tribunal recognised non-binary and gender fluidity within the protected characteristic of gender reassignment. This confirmed that s7 is framed significantly wider than the GRA, which allows for a formal change of a person's legal gender status. This is because s7 explicitly covers all trans people regardless of their legal status (whereas the GRA at present requires a medical diagnosis), particularly when read in combination with the official explanatory notes to the Equality Act.
Gender-Based Violence Without a Legal Gender: Imagining… have, or want to have, any medical treatment or legal change to their gender. However, this does not mean that single-sex spaces have to automatically include trans or non-binary people and currently different single-sex spaces have differing practices regarding this, all of which are permissible under the Equality Act. 12 This means that single-sex spaces for women can exclude trans women, with or without a Gender Recognition Certificate, from their services as long as they can show that this is necessary to the way they carry out their services. Further, services run by local authorities must, in principle, also take account of individuals with the characteristic of 'gender reassignment' when considering their responsibilities under the Public Sector Equality Duty (PSED) contained in s149 of the Equality Act. This duty asks public-sector bodies to 'have due regard to' equality by thinking about their practice and possibly changing it to further promote the inclusion of those with characteristics protected by the Equality Act (Fredman 2011;Manfredi et al. 2018). Indeed, the need to account for a broad range of protected characteristics was one of the reasons Brighton and Hove council cited in deciding to withdraw funding from RISE (Booker-Lewis 2021).
As such, women's single-sex spaces cannot, by default, exclude trans women, but rather this is a possibility as long as this is justifiable due to it supporting a legitimate aim. 13 In fact, the explanatory notes to the Equality Act use the example of a service that provides counselling to victims of sexual violence as one example where the exclusion of trans people may be permissible. 14 So while the requirement to show that this is a necessary way of achieving a legitimate aim may seem daunting to service providers, it is not an insurmountable hurdle, and it is explicitly set out in the explanatory notes that domestic violence is envisaged as an instance in which these exemptions are applicable. It is therefore not obvious that removing formal legal gender status would make it impossible for providers to set rules regarding access to services, unless this was also explicitly prohibited as part of decertification. As one manager of a domestic violence service explained about their organisations' approach to defining themselves as a single-sex service: For instance, we have within our articles of association a sentence that states that we provide single-sex services. But no one would ever challenge us if we said, well, we are providing services to trans women or trans men or to men. It just, I think there is an acceptance already that there is flexibility if you have got the services set up that you can flexibly provide that support.
The provider did not perceive law, or its translation into the organisation's articles of association, as a barrier to setting up their service provision as explicitly singlesex, or amending it as necessary if the organisation perceived there to be a need for different types of provision (for a more detailed discussion of how different actors perceived the role of law in an equality context, see Emerton 2023). It is possible that this may be a more challenging issue for newer organisations compared to more established ones, however it still seems likely that funding pressures provide a greater challenge to third-sector organisations' autonomy and capacity for selfgovernance than the current legal framework around single-sex provision. Currently, providers who offer services that are limited to those with other protected characteristics, such as race or religion which are not underpinned by a formal legal status, are able to legally exclude those without the protected characteristic as defined by the relevant group. Therefore, it seems likely that in conditions of decertification it would still be possible to similarly provide single-sex services. 15 However, it may be that, over time, case law would emerge around the type of criteria that organisations can and cannot use to determine the group(s) they provide services to. A further possibility would be that even without explicit legal rules around the acceptability of specific criteria, political pressure (particularly through funding requirements) would indirectly shape the criteria used, particularly in light of the current move towards funding for mixed-sex rather than single-sex service providers (Women and Equalities Committee 2019, 163).

How Do Third-Sector Organisations Understand Gender-Based Violence?
In the year March 2016 -March 2017, the Crime Survey for England and Wales (CSEW) estimated that 20% of women and 4% of men have experienced some type of sexual assault since the age of 16, equivalent to 3.4 million female and 631,000 male victims. Further, 3.1% of women (510,000) and 0.8% of men (138,000) aged 16-59 had experienced a sexual assault in the last year (Office for National Statistics 2018). Women's Aid reported in 2013 that, based on their estimates, more than 1 million women in the UK experienced domestic violence over the span of a single year; and that on average a quarter of women will experience domestic violence during their lifetime (Women's Aid 2013). It is likely that, if anything, these figures underestimate the prevalence of domestic abuse and sexual violence, given the pervasive issue of underreporting in this context (see Poppleton and Molina 2020 16 ). It is vital that resources and services are available for people who have experienced sexual violence; and while available statistics for both sexual violence and domestic abuse suggest that men also suffer these experiences and are likely to under-report them, women still make up the majority of victims and survivors (Home Office 2010). However, just looking at the prevalence of sexual violence does not necessarily tell us which such services should be provided on a single-sex basis, in the sense of biological sex determined at birth and recorded on birth certificates, and which ones should be provided on a single-gender basis, in the sense of allowing people to self-identify as being a particular gender, or which ones should be provided to all genders. 17 Generally single-sex service providers justify what is de-facto discrimination, as they of course generally exclude men, by referring to what they perceive as the importance of providing single-sex services to people, and most frequently to women, who have experienced interpersonal violence. This is often framed around an explicit or implicit narrative about the oppression and discrimination that women and girls face in the wider world and the gender-based harms they have already experienced. Refuges and service providers often present themselves explicitly as a 'safe space' against the outside world and, as such, operating on a single-sex basis becomes a fundamental organising principle: We work with self-identifying women; so women who are born as women have automatic access to our services, and those women who are self-identifying as women also have the right to access our services. But in terms of the ethos of the organisation and the rationale for why we work in a women-only way, the gender status of our clients is significant, because we do not work with men.
The emphasis on being a single-sex space, in the context of sexual and domestic violence services, seems to link back to the development of knowledge about these specific types of violence, and particularly to radical feminist critiques of how society approaches sexual and domestic violence. Feminist research from the 1960s onwards emphasised the link between different forms of violence and thus contributed to establishing that women have a shared experience of gender-based violence (Dobash and Dobash 1979;Russell 1982;Schur 1984;Cameron and Frazer 1987;MacKinnon 1989). Feminist researchers posited that sexual violence was not just an individual crime but a practice which reflected and reproduced structural inequality through the infliction of racialised and/or gendered terror (see Radford and Stanko 1991). In the late 1980s, feminist scholar Liz Kelly (1987) proposed her 'continuum of sexual violence.' This suggested that women's experience of sexual harm could not be contained within legal parameters that defined sexual offences. Kelly's theory was based on in-depth interviews with women about their experiences throughout their lives. Her research showed that women experienced many unwanted sexual and violent acts within what could be considered 'consensual' relationships e.g., marriage or long-established intimate relationships. Kelly's continuum of gendered/ sexual violence defined a collection of behaviours, from sexual harassment to sexualised murder, with the same social and political function: preserving male power by making women feel unsafe. These structural analyses focused on intersectional power relations. This is in contrast to many commonly held views of sexual violence, which emphasise the idea of dangerous strangers as the main threat to women's safety and which often also contain highly prejudicial racialised assumptions within that. In contrast, feminist research emphasises the everyday nature of sexual violence and its pervasiveness within contemporary society, as well as its links to wider gender norms (Peters et al. 2002). Women's domestic violence services and refuges in some instances evolved directly out of feminist activism although most do not necessarily align with a radical feminist political perspective (see Mann 2000); and a number of current service providers still refer explicitly to their ethos as a feminist space. Several interviewees explicitly described their organisations in reference to feminism: I think we've got an understanding of ourselves, you know, as feminist, you know, we are a feminist organisation that came from radical feminism. We are a radical feminist organisation in our understanding of patriarchy and the construction of gender and all of that. (Employee of a domestic/sexual violence organisation) The organisation was set up [in the 1970s] and it came about as a result of a group of feminists coming together and establishing a grassroots … well, it wasn't even an organisation in those days, it was just a place for women to come together that was women-only and felt to be a safe environment for women to be able to discuss quite sensitive or difficult issues, and to share their experiences. It evolved from that sort of that sort of place. (Manager of a domestic/sexual violence organisation) As such, feminism offered a way of accounting for gender-based violence, but also served as a guiding principle in the running of the organisation. A number of interviewees pointed out their emphasis on collective decision-making, as well as a commitment to feminist working practices, which meant for instance being particularly attentive to the additional challenges faced by female employees with children or other caring responsibilities. Drawing on this feminist ethos that underpinned most of the organisations involved in interviews, most interviewees described a structural account of gender and of gender-based violence: For me, it's about recognising the gender dimension of violence and abuse. I know we are a domestic violence charity, but I think also kind of making those connections across the wider 'violence against women and girls' agenda. And I think if we lose that gender perspective then I think we lose some of the specificity. We lose sight of the structural inequalities that women and girls face in the world today and have done for millennia. It really has been a battle to get that gender dimension recognised. In terms of being able to collate data to be able to demonstrate. Ninety percent of survivors of domestic violence are women. Not to detract from the perpetrating bit [referring to women as perpe-trators of sexual violence], but just to kind of recognise the structural inequalities are really important. (Manager of a domestic/sexual violence organisation) A gender-based analysis was vital, for these interviewees, in understanding how violence operates in society and to provide services that could effectively address it and support survivors. The interviewee above emphasised that, in their group session, this specific organisation focused particularly on exploring broader experiences of gender inequality and the way gender norms shape people's experience of society and their daily lives in particular ways. Without this structural approach, there is then a risk that gender-based violence becomes reduced to individual events and the actions of individual perpetrators, rather than being understood as a systemic problem. One of the challenges that may remain in conditions of decertification, is that of resources to address structural inequality and oppression. Currently, structural accounts of gender-based violence have become intertwined with the need to continuously apply for funding to ensure the viability of the services provided. Due to the competitive nature of such funding applications, organisations must evidence a need for their services, and as such the need for statistical data about gender-based violence was a recurring theme across interviews: One of my issues then is about recording, because every time we go for a bid for example, you have to give evidence what the numbers are women experiencing this. If you lose the category of woman and it becomes diluted, then how do you evidence that structural inequality? I have been working on a bid all weekend and I've been able to draw disabled women out of that; BAME [Black, Asian and minority ethnic] women-because it's quite a big national bid that we were working on. If we lost that category [women], what would we have in its place? (Manager of a domestic/sexual violence organisation) Currently, the prevailing funding arrangements for domestic and sexual violence services and the increasing scarcity of funding shape organisations' capacity to provide services and who such services can be provided to. Although services currently don't generally require formal legal documentation to evidence someone's gender there is nevertheless a question to what extent current practices are based on an implicit assumption that woman/man are largely stable and coherent categories. As noted above, decertification might see the criteria that organisations use to determine who can access services, and therefore who for their purposes is a man or a woman, shaped by both judicial and political interventions. While it is possible to imagine that decertification could lead to the development of a huge range of different ways to define various kinds of genders, it seems more likely that at least in the context of equality law over time there would be a formalisation of some commonly accepted ways of defining genders, perhaps along similar lines of division as religious groupings with some taking a more 'liberal' approach (e.g., self-definition) while others default to a kind of gender orthodoxy, by relying for example on biological characteristics over social ones (for further discussion, see Renz and Cooper 2022).
However, in divergence from the structural account of gender and violence outlined above, some of the current public discourse about trans women's access to and inclusion within women's spaces seems to focus primarily on bodily configurations as sources of violence in contrast to some strands of radical feminism. Adrienne Rich (1980) argued that the existence of sexual violence in our society was generally defended with what she termed the "penis with a life of its own" argument-essentially that most societies take as a given the patriarchal rights of men over women's bodies and therefore mobilise an adolescent model of the male sex drive which "once triggered cannot take responsibility for itself or take no for an answer" (645). Concerns about the inclusion of trans women, or anyone who is perceived as having a 'male' body, within women's services often seem to centre specifically on the concerns identified by Rich about the presence of threatening bodies and body parts, and discussions about which bodies are perceived as threatening and why. This also seems to intersect with the erroneous construction of trans people, and trans women in particularly, as inherently threatening to cis women (Sharpe 2020;Vincent et al. 2020). Media reporting and public discussions about this topic often focus on the idea that people who were not assigned female at birth will enter women's spaces without having undergone genital surgery. This links to other areas where concerns over trans people's presence in intimate spaces have been expressed, for instance in the context of public bathroom usage. Here, Westbrook and Schilt (2014) suggest, assumptions about gender are reduced to purely biological binary understanding of sex, which may not be the case in less 'sensitive' contexts.
It's [trans inclusion] never come up, as far as I am aware, in terms of the staff volunteers, trustees and women who are already engaging with our services. It's mainly come up as an issue via our social media platforms from members of the local … well we don't know whether they are actually local … but members of communities outside the organisation who, as I said before, feel that through doing that [including trans women], we are making our service accessible to men (Employee of a domestic/sexual violence organisation).
Notably, and as highlighted in this quote, these concerns are expressed by people not currently working in this sector. As such, this debate seems to be partially taking place in the realm of the imaginary, where victims and perpetrators of violence as well as the presumed trans women interlocutor are all hypothetical cyphers rather than necessarily real people. These debates construct women as inherently more vulnerable and men or rather cis-masculine bodies as inherently more likely to be perpetrators of violence (see also Peel and Newman 2023). The framing of some bodies as inherently linked to gender-based violence can, of course, also serve to reproduce existing power dynamics around definitions of sexual violence, how these definitions include and exclude certain actors and actions, and how these definitions are shaped by the meanings attached to gender and other social categories. It seems vital to be attentive to how these power dynamics are mobilised in contemporary public discourses around this issue.
However, this is not to suggest that concerns about policing spaces based on gender are only expressed by people not involved in service provision. In fact some service providers were explicitly concerned about the prospect of decertification, which they saw as enabling predatory men to enter spaces otherwise perceived as safe: They [perpetrators] do try to access services by claiming to be victims all of the time. And I think it would just make it more difficult to identify which of those are trans women that need a service, and which are actually male perpetrators.
Similarly, one interviewee mentioned that they shared a building with another group that offered a trans women's group and that subsequently several of her employees had expressed concern about this as it made them feel uncomfortable. The question about the authenticity of some genders or gender identities over others is not a new one. More often, it is framed as competition for resources, for instance in the context of who should be allowed to participate in women only shortlists (Murray and Hunter Blackburn 2019), while in the context of refuges it becomes framed as a question of safety. Most strikingly, while most groups had policies about who they provided services for and who was or was not included within their services, which for a majority of women-only -services interviewed included trans women, in their day-to-day work, legal gender status seemed rarely relevant, especially as many people fleeing domestic violence lack access to the necessary documentation. Instead, how bodies were perceived seemed much more salient, although this was not included explicitly within any policies. As mentioned previously, most interviewees articulated a clear understanding of gender as relationally and structurally constituted, which for some provided a rationale as to why trans women should be excluded from women's services because their experience of violence, oppression, patriarchy was fundamentally different from cis women. However, in practice, exclusion or considerations about exclusion happened on the basis of perception or aesthetics, and of course these may overlap with experiential factors as how one is perceived likely also shapes one's experience of society, but nevertheless would not generally be considered the main factors that constitute gender: But also … I think people who might be perceived as men … it's a difficult issue… it's a really complex issue. We did have a volunteer who identified as non-binary and we were sort of in discussions with her, because what we didn't want to say is: "you are not welcome"; we wanted to discuss it. She hadn't identified that way when she started volunteering for us. And, actually, what happened was she, well she said she was wanting to step back and take a break and then disengaged. (Employee of a domestic/sexual violence organisation) I think in terms of our own services, just pragmatically, trying to provide services for women-that are provided for good reason really-because many will have very negative experiences at the hands of men and people with male bodies then, it would make that more difficult, I think, genuinely. (Employee of a domestic/sexual violence organisation) It seems challenging, if not impossible, to separate issues of how one's gender is perceived from the way in which 'acceptable' gender performance intersects with problematic notions of class, culture, race, and disability (see Deliovsky 2008;Inckle 2014). This poses particular challenges for those with a gender that does not match their sex at birth and an inability to pass is generally linked to less social acceptance and a greater incidence of discrimination (see Hines 2007). Therefore, inclusion/exclusion practices that include aesthetics or perception of gender as a proxy for safety would seem to have a high likelihood of replicating wider social practices of exclusion and discrimination that affect already marginalised groups most acutely.
As a consequence, in part, of the tensions between different accounts of gender and its relationship to bodily difference, some interviewees saw the question of trans inclusion as an issue that was causing divisions between different feminist organisations working to provide domestic abuse services: We have been doing self-identification for years. So, like, my anger with lots of places, like lots of other feminist women who use places like [organisation name] as an excuse to be transphobic is, it's unforgiveable, really. They are not protecting our women only spaces and our women only spaces are fine. Loads of us who have been supporting and open to trans women accessing our services for years are fine, actually, and we've been doing self-identification for years. (Manager of a domestic/sexual violence organisation) While not all interviewees provided trans inclusive services, all were acutely aware of the tension this issue was causing between different groups and often would explicitly frame their position in comparison to, or contrast with, those taken by other organisations.
While views around trans inclusion varied between service providers, there was a common narrative around gender being primarily constructed through structural and relational experiences of oppression, discrimination and violence. 18 A number of service providers linked this structural account of gender back to their ethos as women-focused and as feminist organisations. This, more so than the concerns about male perpetrators of violence identifying as women for nefarious purposes, seemed to be the main driver behind reluctance to include trans women or further open up the category of who constituted a 'woman' for the purposes of their services, as an interviewee often felt that trans women had not experienced the same lifelong disadvantages on account of their gender in the same way that cis women had.
In contrast to either a structural account of gender or a focus on specific bodies, some interviewees, particularly from LGBT-focused organisations working on domestic violence saw gender more as a matter of personal identity and would frequently refer to it as a 'personal' matter that should be left up to individuals to decide. Organisations that approached gender in this way generally favoured trans inclusion and a self-identification model, which they saw as the only feasible approach to determining someone else's gender. In these accounts, the inclusion of trans women was framed as a logical extension of domestic violence services, particularly due to the higher risk of violence that trans women face due to their gender compared to the population average. However, this is not to say that these organisations did not also view gender-based violence as a structural issue; and several of these interviews described gender both as something structural, and as a matter of personal identity, mirroring the tensions in the accounts of others involved in contemporary equality work (see Cooper 2020).

Risk-Management and Privacy as Solutions?
Regardless of which theoretical understanding of gender and gender-based violence a specific organisation subscribed to, they all had to find strategies to manage access to their services so that these would only be accessible to employees, volunteers and service users. This also meant that it needed to be possible to exclude anyone who did not fall within these categories and particularly perpetrators of domestic violence. Beyond just policies that specify who can enter or be part of a given space, organisations often tried to manage perceived concerns around safety through detailed risk management or safeguarding processes. This is one strategy that is commonly used to manage the challenges of "gender's informalisation" in a variety of contexts (Cooper and Emerton 2020, 15). Risk-management becomes particularly noticeable in the context of gender-based violence services, as it is frequently used to try and mitigate a variety of 'risks' from drug use to physical violence. One interviewee outlined their safety policy in some detail during an interview: If somebody tried to access a group in a hostile and threatening way, we have an intercom that we are listening to. We wouldn't be buzzing somebody in unless we were expecting them, or even if someone did gain access to the building, we have a protocol. If people feel threatened, they phone the police. We have an on-call manager on call all the time. We have instant recording forms and they go, where necessary, to the board, [for example] when we need to put procedures in place. We are a learning organisation and so we are constantly updating our policies, but also looking to identify if there are any risks. Our staff and volunteers hopefully know the policies and are trained on the bits that they need to know. (Employee of a domestic/sexual violence organisation) In this account, 'hostility' by people trying to enter was used as a criterion to screen out visitors to the building in advance and prevent entry if necessary. This formed part of a wider risk management process involving form filling, different levels of authority and the capacity to amend procedures to mitigate new or emerging risks in the future. The process outlined here in many ways seems to be a very procedural approach to risk and safety-it is almost an iterative approach to ensuring safety (see Rose 2000;Mythen et al. 2009;Werth 2019). It also functions as a temporal process with multiple steps that taken together over time are intended to produce safety, by trying to pre-empt future-and at this stage only imagined-harm. For other organisations risk management was comparatively more straightforward, and focused on the exclusion of men from spaces more generally: And also, you have got women coming into a refuge who have been subjected to domestic violence by men. I think if you had men close around, you know, that would have a disastrous effect, really. The other thing is also from the safety point of view. At the moment, and quite frankly, if a man turns up then we know that there is a risk. Having men in the organisation. It's more difficult I think to then manage that risk. (Employee of a domestic/ sexual violence organisation) There may be possible interpretations of what risk management does in this instance. Is it meant to avoid wrongful exclusion of legitimate service users? Does it allow policy authors and participants to narrate a process of ensuring safety? This also raises a further question about the role 'space' plays in this context, as space here is both pre-existing, but in the sense of being a 'safe' space also emerging through the implementation of procedure. Space in this sense, rather than bodies, becomes a proxy or carrier for safety. Different organisations may approach this differently, with trans inclusionary groups emphasising procedures that are intended to ensure safety in a space and trans exclusionary groups emphasising inclusion/exclusion criteria that focus on specific bodies. Risk management also includes a strong subjective element. What is perceived as hostile or threatening can vary hugely from one person to the next and across different contexts. What happens to risk management and the idea of safety if someone is perceived as hostile for irrational or biased reasons for instance? To what extent would services increasingly rely on these types of risk management procedures if gender became perceived as a less relevant or permissible basis on which to exclude people? As such, it might be a possibility that rather than trying to find criteria through which to define gender that are both acceptable in a legal context and in the eyes of service funders, risk management procedures would instead implicitly incorporate some proxy identifiers for gender either by more explicitly focusing on bodies or by considering patterns of behaviour more associated with one gender such as physical violence.
A further option for navigation the question of inclusion or exclusion of differently gendered subjects is the move towards providing individualised services that cater to specific groups. As one interviewee noted: My feeling is that it's quite hard for [trans] women to be integrated into a generic women's organisation. I think that a very specific service is needed to meet their needs. What I would like to see as part of the [XXX] is a dedicated service for all trans women as part of what we deliver. (Manager of a domestic/sexual violence organisation) There seems to be tension within the availability of third-sector funding in the sense that while very recently some funders have chosen to only fund general services; there is also an assumption on part of service providers that funding is frequently tied to a degree of specialisation in service provision and can often come with explicit conditions around running a sex/gender segregated service. Part of the perceived advantage of single-sex spaces is the ability to provide a safe environment for survivors of sexual or domestic violence to live together and explore their shared experiences. As one interviewee noted in her description of the shared space in their refuge: They come together, so they have house meetings every week and that's facilitated by the staff team there. The women tend to form friendships. We will do things together which is really lovely. The children, of course, always kind of-you know-develop those friendships because they've moved from, they've given up their friends and they've given up their home with mum. I think it's useful to have it [communal space] so that they can bond. (Manager of a domestic/sexual violence organisation) However, it should be noted that actual women's refuges, as they used to exist-and often still exist in popular imagination-are increasingly rare. While many refuges previously provided shared accommodation, in which women and girls used common areas like living rooms and kitchens communally, this now seems to be phased out in favour of providing individual accommodation, with few or no communal spaces. Even the quote above was in reference to a refuge that only had a garden and one room as a shared space, while the actual living areas consisted of self-contained flats. Similarly, one interviewee who was the CEO of a larger sexual violence charity noted that their refuges now all had individual bedrooms and bathrooms but shared kitchen facilities. While she thought shared facilities were important to allow people to support each other and share their experiences, clients generally preferred private accommodation with no communal areas. This changing design and set-up of refuge spaces may also impact on decertification's effects on specialist service providers that currently differentiate on sex/gender grounds. Gender-segregation may be increasingly less relevant if there is an overall trend towards only providing individual accommodation where service users are unlikely to encounter other service users.
Service providers often perceived a move towards both greater privacy within a refuge and a greater degree of specialisation for organisations as an opportunity to cater to the specific needs of different groups, as the needs of men who have experienced domestic violence within a heterosexual relationship, for instance, may be quite different from those of women who have experienced violence within a lesbian relationship. However, there may also be something that gets lost if services become increasingly individualised or specialised. As one interviewee highlighted, even though they ran a refuge for BAME women only, they also made efforts to bring these women together with women from their other refuges: I think it was a long, hard battle to be able to get that recognition of how religion and culture and tradition intersect with gender and what the impact of that is and where there is lots of shared and common space. We do do things when we bring the women together and so we have survivor events and the women from the refuge come and they join in and their children join in. There is also that need for that specific space as well. One of the reasons-having been in the movement for so long-is the racism that BAME women suffered when they were in generic spaces and about protecting them from that. (Manager of a domestic/sexual violence organisation) While it is of course important to identify different kinds of experiences of inequality, violence and discrimination, it is also important to identify commonalities across groups and experiences. The same concerns that were expressed by a service provider about the move away from communal accommodation, namely that it makes it harder for service users to share experiences with each other and support each other, also seem relevant here. If sex/gender were decertified, this might simply accelerate/ exacerbate the move to providing different services to different groups if gendersegregation based on legal status is not an option, with organisation and funders determining which groups need specific services and how the groups themselves are defined. At the same time, the moves towards individualisation/specialisation and a reliance on risk management can obscure or invisibilise the underpinning accounts of gender which shape how such procedures become operationalised in practice.

Concluding Thoughts
In line with the wider public contestation of gender's meaning (Cooper 2020;Peel and Newman 2020), a tension emerged within a number of interviews, where interviewees embraced both an account of gender defined through structural oppression, and a willingness to recognise that gender could also be an identity, which people may choose to embrace and may experience as innate. This conflict is clearly not resolved through the existence of a legal gender status, and the need to confirm a person's legal gender was never invoked by interviewees as a way to resolve disputes about someone's belonging within a group. As such, removing legal gender status may make little difference to existing conflicts. One possibility is that decertification would intensify the clash between these different accounts, if there is a lack of clarity around what criteria of inclusion/exclusion are permissible without reference to a formal legal status.
For decertification to advance equality and reduce the tension between competing accounts of gender, it would be important to ensure that simultaneously other protections are in place-for example, more funding for refuges and NGOs. It may be that people even continue to hold contradictory understandings of gender at the same time in the same way that everyone is capable of holding contradictory views about a number of issues. As such, law's ability and relevance in intervening in conflicts around gender's meaning may be inherently limited. However, it also seems that any move towards decertification would need to be highly attentive towards the structural underpinnings of gender, and in particular the ways in which they continue to have a disproportionate negative, and at times explicitly violent, effect on some groups over others, particularly women, disabled people, racialised minorities and those whose gender has historically not been supported by the state and civil society. In line with the approach to decertification taken by the FLaG project more broadly (see Renz and Cooper 2022), decertification in the context of work around domestic violence should not amount to denying the current relevance of gender to specific social phenomena, but rather should aim at reducing its social significance over time by tackling gender inequality.
British law's focus on specified, defined 'protected characteristics' places definition in the spotlight. What 'sex', 'man' and 'woman' mean in the Equality Act is already subject to dispute (Malleson 2018;Sharpe 2020) despite these terms' formal contribution to legal personhood. If decertification caused legal gender status to be abandoned, identifying group membership might more closely resemble the approach taken to those other protected characteristics, that are not accorded formal legal status but are nevertheless recognised in equality law. This could allow multiple sex/gender groupings to be recognised (including non-binary, agender, genderqueer) in relation to direct and indirect discrimination alongside the PSED. In principle, service providers could, then, take different approaches to defining what it means to be a single-sex/gender space, but of course law may well seek to also prohibit certain definitional possibilities if these are deemed to be particularly undesirable on equality or other grounds (Cooper and Renz 2016).
It is possible that a move towards decertification would then simply maintain the status quo for domestic violence services. Given that single-sex provision is currently not linked to an individual's legal sex/gender status, organisations could continue to define their services as they see fit, be that women-only in the sense of excluding trans women, or women-only but with the inclusion of trans women and non-binary people. However, in conditions of decertification the GRA would necessarily cease to be relevant, meaning that there may no longer be a legal definition of what it means to be transitioning genders (see also Cooper et al. 2022). This would necessitate other definitional approaches such as relying on biological factors, which would of course create evidential challenges. One regulatory option might be that groups whose objectives are explicitly 'equality-seeking' are given greater discretion in setting their membership requirements or the terms on which they provide services (see also Cooper 2023b). Of course this becomes less straightforward in instances where there are competing equality-based interests as is frequently the case when women-only services exclude trans women (Boyle 2011, 510). In such instances a provider may well want to argue that they are working towards equality but that this requires the exclusion of another group.
Decertification could, in such instance, also lead to legal challenges about rules of inclusion/exclusion similarly to those seen in the UK around other protected characteristics 19 or as those seen around gender other jurisdictions. For instance, in Vancouver Rape Relief Society v. Nixon, 20 Kimberley Nixon, a trans woman, challenged a rape crisis service's policy of excluding trans women from their definition of being a women-only service (Findlay 2003). The Canadian Supreme Court in this instance upheld the service's policy of only allowing volunteers who had experienced gender-based oppression from birth (Boyle 2011). However, even though the Supreme Court found against Nixon, Vancouver Rape Relief Society subsequently experienced a decline in funding (Clément 2019). They then lost access to local government funding entirely (Agahi 2019) due to their trans exclusion policy. This illustrates the capacity of local governments and larger private funders to shape thirdsector organisations. This may be perceived as a positive move when such funding decisions prioritise inclusion and the accessibility of services to a greater range of people. However, it also makes organisations subject to the political priorities of the day (see also Emerton 2023) and may, in the long run, make it more difficult to develop expertise and provide continuity, particularly if there is an overall trend towards giving precedence to the cheapest provider rather than the one with the most relevant expertise.
As such, decertification, particularly in the context of providing vital services, would need to be attentive to the kinds of pressures imposed by funding regimes that may undermine wider efforts towards achieving gender-equality. Consequently, the FLaG "Legislative principles for the decertification of sex and gender" emphasise the value of providing gender-specific provision for the purpose of addressing social subordination, unfairness and violence (Cooper et al. 2022, 38). However, this would also need to be supported by better funding for service provision and facilities, rather than taking place in contemporary conditions of austerity, which may lead to the inadvertent exacerbation of austerity.