Pathways to refugee protection for women: victims of violence or genuine lesbians?

Failures in UK decision-making for women seeking protection from gender-based violence vary depending on the claimant’s (purported) sexuality. These failures are at-tributable, at least in part, to the application of the particular social group Refugee Convention ground which channels claims along two distinct pathways: one path, for women assumed to be straight, focuses on the violence that threatens them; in contrast, for lesbian and bisexual women, the focus is on their sexuality. In either case, the claimant’s autonomy and individuality is eclipsed, but different stereotypes come into play depending on her (imputed) sexuality. This article argues that greater use of the political opinion Convention ground, and a holistic, rights-based approach would improve refugee status determination for all women, regardless of their sexuality.


INTRODUCTION
Law depends on distinct categories: in refugee status determination that means putting asylum claimants in different boxes. 1 They are either women fleeing gender-based violence (GBV), or lesbians fleeing homophobia. This is visible in the two pathways to establishing a claim that have evolved in the UK, depending on whether the woman's sexuality is seen as relevant to her case. For women assumed to be straight, the focus is on the form of persecution (proving a credible risk of gender-based violence); for lesbian and bi-sexual women, the focus is on the Refugee Convention ground (the credibility of their sexual identity as the basis for establishing membership of a particular social group). This ignores the reality of women's experiences and identities: many lesbian and bisexual women seeking protection in the UK are also victims of GBV but this is not fully recognised by decision-makers. Instead, official guidance and case law positions women fleeing GBV, and lesbian and bisexual asylum-claimants as two distinct categories of claimant with different criteria to meet. This matters for women claiming asylum: it has implications for the outcome of their claim as well as for their sense of the integrity of the asylum process.
The article asks why there are these parallel paths depending on the claimant's (assumed or claimed) sexuality. It shows that this phenomenon leads to decisionmaking based on partial accounts of the claimant's experiences. This results in flawed decisions and is also damaging to the claimant's sense of personhood and fair treatment. The article proposes an alternative premise to assessing the claims of women who have experienced or run the risk of experiencing GBV based on a single trajectory: refugee adjudication should start with the recognition that GBV is the reason many women -regardless of their sexuality -seek asylum and this is now explicitly recognised as the basis for providing protection in international law. 2 Having established that the threat of violence is the reason a woman needs protection, decision-makers should then consider the factors underpinning that threat in order to establish the appropriate Refugee Convention ground. This is the point at which sexuality should be a consideration for women from countries where same-sex relations are prohibited or stigmatised. At the same time, sexuality should be considered in relation to the many other factors -age, class, nationality, religion, gender identity, and others -that interact to place women at risk of gender-based violence in different ways. This is an argument for joining the dots to apply the insights of feminist scholarship and advocacy to all women asylum-claimants. 3 This approach would rely on a broader use of Refugee Convention grounds, in particular, the political opinion ground. It would recognise the agency of claimants and their right to protection rather than forcing women to shape their experiences to fit one of two stereotypesvictim of violence or victim of homophobia. The article recognises the problems attached to introducing a new narrative for women claiming asylum, and the difficulty of seeking improvements at a time when preventing regression in asylum rights overall is a challenge. However, it concludes that while GBV continues to be the cause of serious harm to many women around the world, asylum must continue to be a possible (albeit partial) remedy and States have a responsibility to institute a decision-making process that allows all women to give an account of the reasons for their claim that fully and accurately reflects their experiences.

METHODOLOGY AND THEORETICAL SOURCES
The catalyst for this article was the author's observation during the SOGICA (Sexual Orientation and Gender Identity Claims of Asylum) research project that the claims of lesbian and bisexual women participants were considered on the basis of sexual orientation, while their persecution as women who had experienced sexual and domestic violence was often invisible. 4 The arguments that are developed here are based on qualitative empirical research, specifically the individual interviews, focus groups and tribunal observations for SOGICA, and on analysis of decision papers, Home Office materials and published tribunal appeal judgments. 5 The focus is on UK asylum determination in the first two decades of the 21st century. 6 This article uses the terms lesbian and bisexual to describe women who are sexually and/or emotionally attracted to women or to people of more than one gender except where using alternative language used by research participants to describe themselves or others. 7 GBV experienced by women is defined as "violence that is directed against a woman because she is a woman or that affects women disproportionately". 8 It is recognised that many refugee women experience GBV including sexual violence. 9 It also needs to be stressed that GBV characterises the claims of many men, including men who are claiming asylum on the basis of sexual orientation, and also that people claiming on the basis of gender identity are very likely to have experienced GBV. Those experiences are not the focus of this article but many of the questions raised here are equally relevant in those instances -in particular, the question of whether decision-makers' focus on credibility in terms of sexuality and gender identity means that individuals' experiences of GBV are neglected. It is to be hoped that that some of the arguments developed here, including those relating to the problematic application of the particular social group ground, will be picked up and addressed in future research and in this way have wider value.
The theoretical underpinnings of the article are: Western feminist second wave literature; Black and post-colonial feminist critique of that literature; and feminist refugee, human rights and intersectionality scholarship, and debates from the 1980s to the present day. The article draws on this research to create new understanding of the experiences of women claiming asylum in the UK who have experienced GBV. It extends feminist research scholarship to identify discrepancies in refugee status determination depending on the (assumed) sexuality of the claimant. The significance of the article lies in the potential it creates for a more holistic understanding of women's protection needs. While the article's focus is lesbian and bisexual refugee women, decision-makers' focus on credibility in terms of sexuality has wider relevance, particularly for gay, bisexual and transgender individuals claiming asylum.
The argument set out below proceeds as follows: Section 3 considers the criteria for refugee protection that all claimants, including women, need to meet, summarising two elements that are particularly contentious for women and sexual minorities claiming asylum -demonstrating credibility and establishing a nexus with one of the Refugee Convention grounds. Section 4 provides a brief overview of the way feminist scholars and advocates have endeavoured to extend protection to women, who were not originally seen as the intended beneficiaries of the Refugee Convention, before considering what has been achieved and the remaining gaps in women's protection. In Section 5, the article considers how the claims of lesbian and bisexual women are assessed, contrasting them with claims by women whose sexuality is not specified or considered a factor. The pseudonyms used are those chosen by women participants in the SOGICA project. Section 6 considers why there are two adjudication pathways and why this matters. The penultimate section -7 -advocates an alternative approach to the claims of all women before the concluding section 8, which provides some caveats in response to the political realities women and all claimants face before summarising the main argument.

MEETING THE DEFINITION OF A REFUGEE: THE PARTICULAR
SOCIAL GROUP AND CREDIBILITY Turning to the Refugee Convention, as is now widely acknowledged, it was drafted to address harms that have not historically been seen to map onto those experienced by women. 10 Claimants for international protection under the Refugee Convention of 1951 and its 1967 protocol must meet several criteria in order to be recognised as refugees, of which two elements have proved particularly difficult for women. They must demonstrate a "well-founded fear of being persecuted" from which their home state cannot or will not protect them, and the persecution must be "for reasons of race, religion, nationality, membership of a particular social group or political opinion. . ." 11 Neither sex nor gender are included in the five grounds. For this reason, much early debate about how to bring women under the Convention's protection centred on the value and viability of adding gender as a further ground. 12 The general consensus is now that this is not an option that should be pursued. 13 Reasons given vary: many observers recognise the risks of revisiting the refugee definition at a time when there is at best "compassion-fatigue" and at worst outright hostility to asylum claimants and migrants in many "refugee-receiving" countries. At the same time, establishing sex or gender as a ground would be highly problematic in application: either it would be reserved for the claims of women, sexual and gender minorities, implying that heterosexual and cisgender men have no sex or gender; or the universality of gender as an identifier would be recognised, but at the risk of the ground losing all specificity in its application. More importantly, a new ground is not necessary: while the focus of this article is GBV, women's diverse experiences of persecution can be met using the existing Convention grounds if those are interpreted through a gendered lens. 14 In addition to the nexus, the concept of credibility is central to asylum claims in establishing a "well-founded fear" and is, moreover, often the basis for negative decisions. 15 Credibility is often attributed a special meaning in the context of asylum: The term "credibility" is used a good deal in the context both of asylum appeals and of decisions whether a person is a victim of trafficking, and we have detected a tendency to treat it as having some special technical meaning. But in truth it connotes no more than whether the applicant's account is to be believed. In making that assessment the decision-maker will have to take account all factors that may bear on that question. 16 In the UK, a key element of refugee status determination is identifying the material facts of the case and then going on to "assess their credibility against the correct standard of proof" 17 ; however, it has been argued that decision-makers confuse credibility with proof. 18 Moreover, as this article goes on to consider, what "facts" are considered material for establishing credibility varies depending on whether there is a sexual orientation element to the claim. For women claiming GBV without this element, credibility equates to whether the decision-maker believes the woman's account of being at risk of violence such as forced marriage, trafficking, or rape. However, for claims with a sexual orientation basis, which generally also include GBV, credibility means proving that the applicant is genuinely lesbian or bisexual, which involves providing evidence of her activities and relationships in the UK, rather than only whether she is at risk of violence in her country of origin. 19

REFUGEE WOMEN AND GENDER-BASED VIOLENCE
In light of the factors mentioned above, writers have not focused on formal revisions to the Convention but rather on interpreting it in gender-sensitive ways. These have been the focus of work by refugee scholars who, since the 1980s, have sought to bridge the protection gap for women in refugee protection, regardless of sexuality. 20 Explicitly or implicitly, they build on Western feminism of the 1970s, including the recognition that "the personal is political" which analysed how legal, political and cultural distinctions between public and private domains have historically situated the exploitation and abuse of women in the domestic or non-political realm, and so outside the responsibility of the state and society. 21 While the personal is political also exposed inequalities in labour between men's paid work outside the home and women's unpaid work within it, 22 the phrase is strongly associated with feminisms that defined sexual and domestic violence as a primary cause, consequence and site of women's oppression and pillar of patriarchy. 23 These Western feminist works have been strongly critiqued by other feminist writers for homogenising and prioritising the experiences of white Western women. 24 This early body of work has also been criticised for privileging women's experiences of violence as the paramount concern for feminist scholars. 25 However, it is entirely possible to recognise the validity of those criticisms while still giving credit to the significance of this branch of feminism in identifying the place of violence in the perpetuation of women's inequality. In the context of international refugee law, it is certainly part of the chain of analysis that led to the point at which violence against women by non-state actors where there is no state protection is now recognised, at least in principle, as the basis for claiming asylum. Just as feminists have challenged the public-private distinction to identify violence against women as fundamental to women's subordination and a public and political concern, so refugee scholars have identified GBV, whoever the perpetrator, as a serious harm that constitutes persecution for asylum purposes. 26 Persecution is famously undefined in the Refugee Convention, giving the basis for ambiguity as to the threshold that needs to be met to qualify for protection. The definition used by the Refugee Women's Legal Group in their 1998 guidelines provides a clear starting point that can be applied beyond gender-based claims: "'Persecution' ¼ Serious Harm þ The Failure of State Protection." 27 This negates the assumption that persecution must be at the hands of a state official or agent. In this way, the work of feminist refugee scholars continues the project of exposing and interrogating assumptions related to public-private distinctions begun by earlier feminist writers and activists. 28 However, Western feminist refugee scholars have also been informed by Black and post-colonialist feminist critiques of feminisms that homogenise the experiences and values of white Western women. 29 Macklin, Crawley and others, for example, have disaggregated "women" as a group by highlighting the different ways women experience persecution: A woman may be persecuted as a woman (e.g. raped) for reasons unrelated to gender (e.g. activity in a political party), not persecuted as a woman but still because of gender (e.g. flogged for refusing to wear a veil), and persecuted as and because she is a woman (e.g. female genital mutilation). 30 Queer asylum scholarship, while not the focus of this article, also builds on critical race theory to dismantle essentialism: Tschaeler, for example, shows how the racialisation of Black lesbian women claiming asylum in Germany determines decisions about their claims. 31 If decision-makers were to recognise this diversity as the basis for claiming asylum, it would be clear that all five Convention grounds are as applicable to women as they are to men, although they may take different forms: taking religion as a ground, for example, girls may be denied access to education on religious grounds in a different way to boys, as is the case in Afghanistan at the time of writing. 32 Broader application of all the Convention grounds to women claiming asylum is the approach favoured by many refugee women scholars and advocates and is also the approach taken in this article; however, it is not what has transpired in practice. 33 Instead, the particular social group ground has become the default for women's claims, including in the UK, in a way that has superficial advantages but raises concerns that have been widely discussed and are revisited here. 34 Case law on this matter dates back to the 1980s, but in the UK it is the case of Shah and Islam that for the first time established that women could constitute a particular social group. 35 The case involved two women from Pakistan at risk of violence from their husbands although their experiences were very different in other ways. 36 The political opinion ground was explicitly ruled out (Lord Steyn said "I was not attracted by this argument"), 37 while less than a decade later, another case also firmly rejected the application of political opinion to women. 38 In light of the rejection of women's claims on political opinion grounds, the affirmation that they might constitute a particular social group is welcome; however, it should be seen as a mixed 30  blessing for women and LGBTQIþ minorities -who also achieved particular social group recognition through this case. 39 While it created the standard for recognising women fearing GBV as refugees, there is confusion about the necessary specificity of any particular social group. Moreover, the question of whether the criteria for group membership are cumulative or alternative creates further difficulties for women as discussed below.
Taking the first of these two obstacles, while it is universally agreed that a particular social group must be defined by more than a shared experience of persecution, there is little agreement beyond that as to how broad or narrow a group may be for women claiming on the basis of GBV. Groups have been defined as widely as "Women in Afghanistan" or -by the appellant in the same case -as narrowly as "lone Afghan women who have suffered sexual assault, have female children and are without social or family protection". 40 Turning to policy guidance rather than case law, in the UK, the Home Office Country Policy Information Notes (CPINs) brief decision-makers on particular issues in particular countries. There are Notes relating to women and GBV, and notes relating to sexual orientation and gender identity and expression (SOGIE). 41 For women and claims based on GBV, the specificity of the group that is required varies: sometimes the particular social group is as broad as "women", sometimes it is narrowed to, for example, "Women and girls in fear of FGM", sometimes it is both broad and narrow, as in "Women and girls in Nigeria, including those in fear of FGM, form a particular social group". 42 In contrast, LGBTQIþ claimants are covered in single Notes generally titled following the formula for Iran: "Country policy and information note: sexual orientation and gender identity, Iran, June 2019". Moreover, there is no narrowing of the group for LGBTQIþ claims. All Notes are modelled on the wording used here: LGBTI persons form a particular social group (PSG) in Kenya within the meaning of the Refugee Convention because they share an innate Refugee Survey Quarterly 9 characteristic or a common background that cannot be changed, or share a characteristic or belief that is so fundamental to identity or conscience that a person should not be forced to renounce it and have a distinct identity because it is perceived as being different by the surrounding society. [emphasis in original] 43 Fundamental to this article is the recognition that many lesbian and bisexual women are not only persecuted on the basis of their sexuality as covered by the SOGIE notes, but also subject to GBV and should be covered by the notes relating to these forms of persecution in particular countries. This makes the particular social group criteria that a woman needs to meet unclear. If a woman from Kenya has been threatened or undergone FGM because she is (perceived as) a lesbian, does she need to demonstrate membership of the group that is "Women and girls in fear of FGM" and the group that is " LGBTQIþ" persons? 44 Of course, for a woman in this situation, all these elements -fear of FGM, being a woman and being a lesbian -are relevant to her claim and also related. However, this is not what is happening. It is clear that the law cannot recognise intersectionalities: 45 A Home Office letter refusing protection to a lesbian woman who had experienced extreme GBV stated "You claim to have a well-founded fear of persecution in Malawi, on the basis of your membership of a particular social group (a lesbian)." 46 In an appeal by an Albanian woman who had been physically abused by her alcoholic father long before she told her family she was a lesbian, the following questions were asked: 1. Was the Appellant a member of a social group as a result of sexual orientation, alternatively; 2. Was the Appellant a member of a social group by virtue of being a woman; 3. Was the Appellant a member of a social group as a result of her having refused a forced marriage? 47 These questions are posed as alternatives. It is not recognised here that group membership may be based on sexual orientation and being a woman and refusing a forced marriage. Moreover, in such cases, it appears that sexuality trumps gender. In the case above, consideration of questions two and three was subsequently dropped. Instead it is stated: " [o]f particular concern was the issue as to whether lesbians in Albania formed a particular social group and whether or not a sufficiency of protection was available to them. Confusion as to what exactly the particular social group should be is compounded by the question of how high the evidentiary bar is in requiring claimants to meet one or two criteria. The United Nations High Commissioner for Refugees (UNHCR) definition of a particular social group has two limbs: the "immutability" and the "social perception" elements: the former requires that a group is united by a shared fundamental, innate or unalterable characteristic; the latter requires that the group be "cognizable" and set apart from larger society. 49 According to UNHCR guidance, these should be seen as alternatives, with only one element required. 50 However, the European Union Qualification Directive takes a cumulative approach, requiring both elements to be present. 51 This was rejected by the House of Lords in 2006 as propounding "a test more stringent than is warranted by international authority" and the alternative approach was confirmed by the Upper Tribunal in 2020. 52 Nevertheless, the cumulative approach is replicated in UK guidance as well as in the Borders and Nationality Act 2022. 53 The implications of this are that women claiming asylum need to show both that they are part of a group that shares an immutable characteristic, and also that they are perceived to be part of a group that is differentiated from larger society in some way. 54 This raises a high evidentiary bar that may disproportionately affect some women, for example, those who have been trafficked are likely to find it difficult to prove that they are perceived as having a distinct identity. 55 Similarly, a woman fleeing domestic violence may find this limb of the criteria difficult if GBV is endemic in the country she comes from, as it is in most countries. How cognisable can a group be if it constitutes as much as half of society? While the size of the group should be irrelevant, it is generally recognised that public and political opinion is resistant to any measure that appears to open the "floodgates" to large numbers of asylum claimants. 56  it may be relatively easy to meet the "social perception" criteria, that sexual minorities are a stigmatised group in the country of origin where "homosexuality" is criminalised in that country. However, lesbian and -in particular -bisexual women often find it difficult to demonstrate the "immutability" element and may be expected to "pass" for heterosexual. 57 There is evidence that claims are routinely refused because the claimant is not deemed "credible" as a lesbian or bisexual woman. 58 Protecting women from GBV is not specified in the Refugee Convention, forcing advocates to engage with the legal niceties of the particular social group criteria. However, International Human Rights Law does explicitly give States responsibility for protecting women from GBV, effectively addressing the gaps in the Refugee Convention. 59 CEDAW was the original instrument here, specifying that discrimination against women includes GBV and is therefore a violation of women's human rights. 60 However, more recently, the Istanbul Convention has explicitly set out States' responsibilities to asylum-seeking women as a human rights duty encouraging them to take measures to ensure that GBV is treated as persecution within the meaning of the Refugee Convention. 61 Moreover, while refugee law is an individual solution that functions on a case-by-case basis and does not attempt (rightly or wrongly) to change the practices of the "refugee-producing" State -as Anker points out, its role is "palliative" 62 -human rights law does seek to set standards and promote change beyond an individual level. 63 Together, the elements of international law described above provide solid justification for providing refugee status to women at risk of GBV who are unable to secure the protection of the country of origin. There have been notable cases that have confirmed this protection. 64 A number of countries have had gender guidelines in place for many years. 65 Moreover, the application of a gender lens to asylum extends beyond the formal decision-making process to include consideration of what else needs to exist to ensure women can make a full and fair claim for protection, for example, the provision of childcare to ensure women do not have to be interviewed in the presence of their children. 66 However, this is a description of progress in theory, and not necessarily what is happening in practice. Furthermore, in the UK (as in many countries), any gender-sensitive asylum measures need to be seen in the context of a broader climate that is hostile to migrants and asylum-claimants.

LESBIAN AND BISEXUAL WOMEN CLAIMING ASYLUM IN THE UK
Whether positive or negative, the developments above do not extend sufficiently to claims by lesbian and bisexual women where the connection with GBV has not been fully made. Retracing the steps above to the case of Shah and Islam, this was also a catalyst for recognising claims based on sexual orientation, including those of lesbian and bisexual women. However, at that point, the paths of women claimants take different directions depending not on whether women are or are not at risk of GBV, but rather on whether they are or are not claiming persecution based on their sexuality. The result is that the refugee status determination process, in particular the credibility aspect of it, diverges for women depending on their (imputed) 67 sexuality. Women whose claims relate to the threat of GBV but whose sexuality is not seen as a material fact will need to demonstrate that they are at risk of forced marriage, FGM, "honour-"based violence, rape, or sexual violence. In contrast, women whose have experienced these same forms of violence because they are lesbian or bisexual and are claiming protection as a result will have to prove their sexuality, while GBV is either absent from the claim or plays a secondary role in the determination process. While they also need to demonstrate a risk of being persecuted, if they are from a country where "homosexuality" is criminalised or stigmatised, this is likely to be apparent from country information and thus accepted by decision-makers, hence the case falls or not on the credibility of their claim to be a lesbian or bisexual woman. Although neither of these evidentiary challenges are easy, they are very different.
If a grant of asylum depends on converting one's experiences into a recognisable narrative, the narrative that is recognised is very different for lesbian and bisexual women on the one hand and victims of GBV on the other, despite the obvious Refugee Survey Quarterly 13 overlap in these claimants' identities and experiences. 68 Taking a notable case about GBV violence as a benchmark for claims that do not involve sexual orientation, that of NS, a woman from Afghanistan, while there was discussion in her successful appeal as to the precise definition of the particular social group to which she belonged, it was ultimately found to be "women in Afghanistan". 69 The appellant had been raped and threatened with forced marriage. Factors that were considered included that it would be "unduly harsh" to return a lone woman with two daughters to Afghanistan. Evidence submitted related to human rights in Afghanistan, in particular "the situation of women in Afghanistan" which is considered over six pages of the judgment, and also to the appellant's post-traumatic stress disorder. Her fear of persecution was found to be credible and relocation was ruled out based on testimony that: the state is unable to protect women, and that the task of protection of women falls to the male members of her family, especially their husbands. Women who find themselves without the effective protection of their families are in a very vulnerable and dangerous position. 70 Similarly, in a more recent case concerning a woman from Nigeria, it was accepted "that as a former victim of trafficking her claim for asylum falls for consideration under the Refugee Convention as a member of a particular social group". 71 The issue before the tribunal was whether she would be at risk of being trafficked again if returned. In both cases, it is the threat of harm on return that is at issue, not the behaviour of the appellant once she was in the UK. In contrast, the focus of decisionmakers in the claims made by lesbian and bisexual women at risk of GBV, including those interviewed by the author, is very different. In these cases, the sexuality of the appellant, including how she expressed it after arriving in the UK, is a primary consideration as the following cases demonstrate. NN, a Pakistani woman, claimed asylum in the UK in 2015. The basis of her claim was that she was a lesbian. She had been forced into marriage, beaten by her husband and threatened by her father. Her application was refused and her appeal to the First Tier Tribunal was turned down. The Upper Tribunal appeal "submitted that in its focus on that historical account the [First Tier] Tribunal lost sight of the central question to be determined: is the Appellant in fact gay?" It found in favour of the appellant: "I accept and find as fact that the Appellant is gay. I do so for the following reasons: The Appellant has given detailed and compelling evidence about her realisation that she was gay, her struggle with that fact in the context of her cultural background, and her reconciliation with her faith. . ..". Forced marriage as the basis for claiming asylum is recognised but only in the final paragraph of the judgment. 72 It is the appellant's sexuality per se, and not the way it intersects with and is the basis for the violence she experienced, that is the key factor in a positive decision.
Angel 73 was a 35-year-old Zimbabwean woman at the time of her interview for this research. She fled Zimbabwe when she was caught having sex with her same-sex partner. She had been raped and forced into marriage and had a daughter through the marriage, who she had to leave behind. 74 Despite being an avowed political activist, supporting women's groups in Zimbabwe as well as a self-identified feminist activist for refugee women's rights in the UK, her claim was not based on political ground, nor was GBV the focus. She was initially refused, with the refusal letter repeating the basis of her claim, including the fact that after a marriage had been arranged for her, her husband forced her to have sex and beat her up: "When you went to the police, your husband told them that you loved girls. . . and so the police also beat you up . . ..You were raped by 2 men on [date]". 75 However, the refusal letter stated "[a]s you have been inconsistent throughout your account and not provided what are considered to be reasonable explanations, it is not accepted that you are a lesbian".
Angel appealed and she explained to the author that while her first refusal was because the Home Office did not believe she was a lesbian, she lost her appeal because the judge did not believe that lesbians in Zimbabwe were at risk: Because, when we went there [to the Home Office], all along the argument even from when I went for my interview, was my sexuality, that is the thing that they were arguing about. That no, they don't believe from the interview. And when we were going to court, we were all going focused on my sexuality, because this is what they had said, only to get there, to be told that "ok, even if you are gay in Zimbabwe, it is legal for women to be gay according to law, but it is illegal for male". 76 The fact that she was a victim of rape and forced marriage was incidental to her case. Angel explained: "I remember with my refusal the judge said I didn't say a lot about my rape, so he does not consider that it is a correctional rape". Her experience of being raped was included as a part of her narrative but, because Angel did not describe it in the way that the judge expected for correctional rape in sexual orientation cases, this became an element that undermined her case. Angel described how hard she had found it talking about the sexual violence she had experienced in court when she was trying to forget it, and felt that the fact that she did not say much about it was held against her. 77 The difficulty that victims of rape and sexual violence have in talking about their experiences is widely recognised in guidance and research on women's asylum and GBV, 78 but does not appear to have been recognised by the judge in Angel's case. Angel was finally granted asylum after making a fresh claim but there was little sense of euphoria because she continued to be separated from her daughter and faced the huge task of securing family reunion. Jayne 79 was another interviewee who had experienced sexual abuse and violence, repeatedly and by more than one close family member. She also had a child but had been able to bring him with her to the UK. She shared her decision papers with the author and her case was presented largely in terms of sexual orientation. For example, her original refusal letter stated "You claim to have a well-founded fear of persecution in Malawi, on the basis of your membership of a particular social group (a lesbian)." It went on to state that "your claimed sexuality has been rejected". 80 The judge in the First Tier Tribunal refused her appeal, finding that "[t]he appellant is not a lesbian, nor would she be perceived to be a lesbian on return to Malawi". This was largely because of the failure of her ex-partner and teenage son to appear as witnesses to support her claim to be a lesbian. There was some consideration of the GBV she had experienced, with the judge stating: "If her account, that she had been already raped and persecuted by her family for her sexuality, was true, her delay in claiming in a safe country which is known for its tolerance is even less credible". 81 When she was finally granted asylum it was after providing fresh evidence of her sexual relationships and involvement in the LGBTQIþ community in the UK, including 15 letters from individuals stating that they are "certain [Jayne] is a lesbian" or "sure [Jayne] is a lesbian" or "aware that [Jayne] is a lesbian". Evidence included a statement from her son, now 19, that "I met mum's girlfriend [X] on several occasions. . . . They slept in same room and sometimes bathed together". The appellant herself provided a testimonial that said "More recently, I have had sexual relations with [X] who lives here in [UK city]. She has prepared a witness statement". Jayne submitted medical records, including a note from a consultant clinical psychologist that stated she was "suffering from PTSD and depression as a result of her past experiences of rape, abuse and persecution on account of her sexuality. The GP notes additionally discuss our client's discovery of being HIV positive as a result of the sexual assault she endured in Malawi". 82 Her eventual grant of asylum was based on recognition that she was, indeed, a lesbian.
As a final example, in a First Tier Tribunal appeal in 2018, a woman who explained that she had been raped by her cousin to "cure her" when he found out she was a lesbian, and subsequently trafficked to the UK and recognised as a victim of modern slavery, was refused asylum initially because of inconsistencies in her account and because the Secretary of State for the Home Department "was adversely impressed by the fact that Miss M had not been involved in any lesbian relationship since coming to the United Kingdom". 83 The Tribunal judge was not satisfied that she was or would be perceived as a lesbian despite her "mannish" appearance. Her experience of rape was not a focus of the appeal. She was eventually granted protection by the Upper Tribunal. 84 In all these cases, both at initial decision-making and appeal stage, the violence the woman experienced took second place to the question of whether she "really" was a lesbian. The previous sections have shown the way that women claiming asylum on the basis of GBV are assessed differently depending on whether sexuality is considered relevant to their claim. The next section identifies the harmful implications of this phenomenon.

CONSIDERING THE IMPLICATIONS
The focus on sexuality cannot be attributed solely to adjudicators. It tends to be solicitors who advise claimants on how best to present their case, either to the Home Office or, after a refusal, to the tribunal, in which case they will be instructing a barrister. It is likely therefore that if lesbian and bisexual women who have experienced GBV present their claim as based on sexuality rather than violence, this is based on advice from their advocate. 85 However, it also correlates with Home Office guidance to decision-makers. In addition to the Country Policy Information Notes discussed above, the Home Office has guidance for decision-makers that includes "Gender issues in the asylum claim" (last updated 2018) and "Sexual orientation in asylum applications" (2016). There is cross-referencing within these documents, however, decision-makers, operating under pressure and with limits to the amount of time they can spend on any one claim, must prioritise their use of resources. Therefore, while an official given the case of a bisexual woman from Bangladesh who is threatened with "honour-based" violence will have available to them the guidance on gender and on sexual orientation, as well as the Country Policy Information Notes for "Bangladesh on Women fearing gender-based violence" (2020) and "Sexual orientation and gender identity Bangladesh" (2020), it is likely that one dimension -GBV or sexual orientation -will be privileged at the expense of recognition of the interconnections between the two. This is demonstrated in the case studies detailed above. Moreover, if the particular social group ground is privileged over the other four possible Convention grounds, this is in line with the direction taken in all of the country guidance. While more than one Convention ground may apply to a case, all the Home Office CPINs direct decision-makers to the particular social group ground and only to that ground. 86 84 Ibid. 85 For example, in a First Tier Tribunal hearing observed by the author, the appellant's barrister summed up the case as follows: "We would argue this is a case based on credibility. The Secretary of State says Cameroon is not safe for lesbians. So this case stands on whether my client is a lesbian. We understand there is a previous judgment. We argue that the evidence before you today is so significantly different that it shows the appellant is a lesbian and she should be before you. Since her previous hearing the appellant is in a new relationship with Miss X and she is here before you. The evidence provided by the appellant and this witness is broadly consistent and we would argue this shows they are in a genuine relationship." (First Tier Tribunal hearing, Sept. 2018). The evidence of the case studies and analysis of Home Office guidance above shows that GBV and sexual orientation are still too often seen as separate identifiers. Despite the acknowledgement of intersectionality that is now standard in scholarship, echoed in UNHCR guidance, and that is critical to understanding the nature of GBV experienced by lesbian and bisexual women, 87 international refugee law struggles to conceive individuals as multi-dimensional. This ignores the material realities of women's experiences as lesbian, racialised, of a certain age and class, and the many other factors that contribute to the specificity of their treatment and experiencesand which may be relevant to an asylum claim. And while this may be true of law in general, it is particularly so in the UK, where asylum decision-making replicates the adversarial nature of criminal law based on polarities such as recognition or nonrecognition. 88 Moreover, not only are women's sexuality and their experiences of violence treated in isolation, it is also the case, based on the examples provided above, that for lesbian and bisexual women, their sexuality trumps their experience of violence as the prime element of their claim. This may, in part, be a strategic decision on the part of advocates. Some forms of GBV, ones that are also experienced by many women in the UK, such as rape and domestic violence, may be seen as too mundane and recognisable and therefore not sufficiently serious to be the basis of a successful claim. There are also likely to be evidentiary difficulties in GBV claims. Experiences of violence will often have happened in the home or without witnesses and there is unlikely to be documentary evidence to support a claim. While the Home Office guidance on credibility emphasises that claimants should be given the benefit of the doubt where they provide a coherent narrative, critics claim that this is not the case in practice. 89 The same evidentiary difficulties apply to claims based on sexual orientation. The difference is that what needs to be proved is not that someone has been raped or forced into marriage, but that they have been or risk being abused because of their sexuality, so the "fact" to be proved becomes whether they are "homosexual" or not. Where claimants apply on the basis of sexual orientation-based persecution and are refused because of lack of evidence and go on to appeal, as many claimants do, 90 they can then spend the often lengthy period before their appeal hearing accumulating "evidence" that they are lesbian or bisexual, evidence such as letters from support groups, photos of participation in Pride events, and witness statements from new partners. Although decision-makers may discredit such evidence as selfserving, they are generally recognised as evidence of future risk due to sexual orientation, while options for securing new evidence of past experiences of GBV are unlikely to exist. 91 Advocates may therefore find it expedient to encourage their clients to focus on the sexual orientation element of their case over the elements of violence.
The future-focused nature of the Refugee Convention definition may also explain why lesbian and bisexual women are well advised to focus on their fear of future persecution as members of a sexual minority rather than as women threatened with violence, perhaps on the basis of having already experienced it. UN Women express the concern that "the Refugee Convention is forward looking and requires that there be a well-founded fear of persecution in the future, which raises complication for past victims of rape". 92 As in the US, the UK formally recognises past persecution as an indicator of future risk. 93 However, lawyers may see a difficulty in proving that a woman who has undergone FGM is at risk of further harm. 94 They may need to develop subtle hard-to-support arguments about the level of trauma that would be experienced on return being equivalent to persecution. In comparison, the case of a lesbian woman in flight from a country where same-sex relationships are criminalised may be easier to make. Basing a claim on sexual orientation rather than the risk of violence may also help to avoid refusal on the basis of an internal flight alternative: if "homosexuality" is penalised in a country, this is most commonly across the whole territory, whereas if a woman is threatened with FGM in her home town, it might be argued that she can return to another part of the country. There may be good arguments why this is not a reasonable expectation, but these will need to be developed and proven, while looking to what the law of a country states in relation to sexual relations is much simpler -and quicker for lawyers paid through legal aid which increasingly fails to cover the costs of their time. 95 Finally, there may be a general reluctance to recognise GBV, whether it relates to sexual minorities or not, for the reasons suggested by Baillot et al. They found a failure on the part of all parties to fully recognise experiences of rape in women's asylum claims, which they attributed to factors including distancing on the part of decision-makers to protect themselves from trauma. 96 The discussion above suggests that there are a number of cultural, psychological, legal, and strategic reasons why the claims of lesbian and bisexual women who have experienced GBV are viewed primarily through the lens of sexual orientation, overlooking the extreme violence that many of these women have experienced. One response might be that this is not a problem. If women's protection claims are addressed largely by deploying the particular social group ground but with different pathways for claims based on GBV and sexual orientation, and if lesbian and bisexual women are routinely assessed in relation to the latter and not the former, the only concern should be that the end result is the same, namely that refugee status is granted where merited. Does it matter whether claims are accepted or refused primarily because the claimant is a lesbian or bisexual, or is a woman who has experienced GBV? However, the end result may not be the same, as the cases above illustrate: if a claim is presented as based on sexual orientation, that triggers a determination process almost entirely focused on the claimant's credibility in terms of whether she is "genuinely" a lesbian or bisexual, often evidenced by her behaviour after claiming asylum. The claim is then often rejected on the basis of minor inconsistencies in the claimant's narrative. If the same woman's experiences of GBV are made central, the basis for proving her case prioritises the body of Country of Origin Information available, to ask whether she would be at risk of the same kind of violence that caused her to flee if she is returned to her country of origin. These are two significantly different lines of argument and evidentiary processes, and may therefore result in different final decisions. Neither type of claim is easy and both are intrusive in different ways, but in channelling the claim into the sexual orientation box at the expense of other aspects, the evidentiary process for lesbian and bisexual women fails to reflect the reasons why they need protection: that they experience GBV because of their sexuality. The focus becomes the individual's identity, established largely through the coherence of her narrative and jeopardised by the kinds of memory lapses that are typical of people who have experienced trauma and violence. 97 Moreover, questioning in this pathway often makes unrealistic demands of the claimant, expecting women who have fled extreme violence and have recently arrived in a country where they may not know the language and have few resources to prove their sexuality by going to gay clubs and dating. 98 Rather than looking for evidence to confirm the claimant's credibility in the country of origin's political and social conditions for women, including lesbian and bisexual women, the focus is on the claimant's behaviour after she left her country of origin and made her claim -what has she done in the UK to demonstrate she is a lesbian or bisexual. This is the wrong question.
There may therefore be implications in terms of recognition rates for women claimants in that the process for establishing a claim based on sexual orientation is particularly difficult because it asks claimants to prove the unprovable. 99 However, refusal rates are only one consideration. 100 Equally important is the individual's sense that they have gone through a meaningful and fair process. Angel was forced into marriage and repeatedly raped because she is a lesbian. She was refused in the first instance because the Home Office doubted her sexual identity and refused again at first appeal because, while it was recognised that she is "genuinely" a lesbian, she comes from a country where only male homosexuality is criminalised. Her claim was finally accepted, but she is unlikely to have come out of the asylum process feeling that she has had a fair hearing and that her experiences have been understood and recognised. Questioning to establish whether an individual is or is not a lesbian may have improved since the controversy over the Home Office asking offensive and intrusive questions in 2014 and the CJEU case that year establishing that this is impermissible, 101 but it remains a degrading experience, as for Jayne whose son and ex-partners among others were required to provide written statements identifying her as a lesbian on the basis of sexual experiences and shared bath-times in order to finally secure her protection. 102 In this situation, eventually securing refugee status may be at the expense of someone's personhood, dignity and any real understanding of why they need protection.
Focusing on sexuality rather than GBV may also mean that opportunities to make connections with international human rights are lost. While the Home Office guidance on gender in asylum claims refers decision-makers to CEDAW and the Istanbul Convention, the guidance on sexual orientation in asylum claims does not. 103 If a Home Office decision-maker assessing the claim of a lesbian or bi-sexual woman, and operating under recognised pressures of time, refers only to the sexual orientation guidance, the protection that should be provided by international human rights instruments may be overlooked. 104 Lastly, there are practical reasons why it matters if not all the reasons why women are claiming protection are taken fully into account. For example, the trauma caused by experiences of sexual and GBV may have implications for a claimant's ability to present a coherent narrative. 105 If the key question is whether she is genuinely a lesbian or bisexual, this may not be recognised. The extensive work that advocates have done to improve the process for women in relation to interviewing people who have experienced trauma and issues relating to separation from children may not be applied. Lesbian and bisexual women may not benefit from, for example, recognition of the time that it takes for people to reveal experiences of sexual violence or of the impact of trauma on memory recall. 106 Similarly, if decision-making prioritises confirming the claimant's sexuality, an inappropriate burden of proof may be imposed -dismissing a claimant on the basis of coherence (as a lesbian or bisexual woman) rather than recognising the effect that her traumatic experiences will have had on her ability to provide a coherent and consistent account and giving her the benefit of the doubt. 107 The problem, of course, is not that sexual orientation is the basis of lesbian and bisexual women's claims when, in some cases, GBV is the reason they need protection and therefore that the prioritisation should be reversed. That would be to marginalise a different aspect of women's identity. Rather the decision should be based on the totality of the claimant's experience which is often that the serious harm women are threatened with because of their sexuality takes the form of GBV. This more integrated approach is not theoretically incompatible with the application of the particular social group ground. Just as there does not need to be only one Convention ground applied to a claim, equally there can be more than one particular social group. 108 Markard, for one, suggests that the particular social group ground has developed broadly at the same time as gender-based perspectives of refugee law, and that it is flexible enough to accommodate the totality of lesbian women's reasons for claiming protection based on both gender and sexual orientation. She argues that this has been the case avant la lettre since the 1990s, even if adjudicators have not explicitly or consciously applied the concept of intersectionality. 109 However, while Markard argues that the flexibility of the particular social group ground makes it suitable for recognising intersectionality in claims and includes lesbian women in her examples, that is not the approach advocated here. Whether necessarily or simply in practice, the particular social group ground has served to put women in silos -either lesbian/bisexual women or (purportedly) heterosexual women. The stereotyping that accompanies the application of the particular social group ground is a central part of the problem because women escaping violence, including lesbian and bisexual women, need to demonstrate that they meet both limbs of the particular social group criteria to be recognised in the UK. 110 Even if decision-making were to encompass both sexuality and experiences of GBV, some claimants would still need to demonstrate that they were a lesbian or bisexual women who had experienced violence on that basis (the immutable characteristic requirement) and that they would be perceived as such (the social perception element). It is the latter requirement that encourages stereotypes. In fact, taking an intersectional approach but relying on the particular social group ground as currently implemented would increase the burden of proof and perhaps create insuperable obstacles for women because of the contradictory stereotypes in play. For women who have experienced violence, conforming to stereotypes of victimhood and vulnerability may facilitate their claim 111 ; for lesbians and bisexuals fleeing homophobia, the prevailing stereotype is of "out-and-proud" LGBTQIþ people able and expected to live openly for the first time. 112 The artificiality of these stereotypes becomes clear when one considers the case of a woman who has experienced sexual violence because she is a lesbian or bisexual. Moreover, conforming to the stereotype of a happy clubber and also that of an abject victim of violence at one and the same time would clearly be very difficult. There are competing and irreconcilable stereotypes for each of these archetypes, and presenting a demeanour that combines both is unlikely.

AN ALTERNATIVE PATHWAY
There are different ways to tell any story. 113 The choices that are being made about how to tell the stories of lesbian and bisexual women do not necessarily serve these women well, either in the short or the longer term. Rather than continue to rely exclusively on the particular social group ground, a number of feminist refugee scholars have suggested making greater use of all the Refugee Convention grounds, in particular, political opinion for women fleeing GBV. 114 It is recognised that "[p]olitical opinion should be understood in the broad sense, to incorporate any opinion on any matter in which the machinery of State, government, society, or policy may be engaged". 115 As Mulligan argues: "Thus, if there is no effective prosecutorial system available to a woman in the country where the rape has occurred, or if her life is endangered by bringing the issue to the attention of the authorities, then the applicant has a well founded fear of persecution on account of political opinion." 116 This article supports that argument and suggests that it is equally relevant to the claims of lesbian and bisexual women fleeing GBV. The key to establishing any claim for protection is the nexus -that there is a threat of persecution on account of at least one of the convention grounds. For women threatened with sexual and domestic violence, Crawley and other writers have long argued that their claims should be seen as political in that women are explicitly or implicitly rejecting the gender norms imposed on them. 117 The basis for doing this is as clear in relation to lesbian and bisexual women as it is in relation to any other woman who has experienced GBV. A similar nexus exists for some lesbian and bisexual women who have been forced into marriage, raped, or undergone FGM -or been threatened with any of these abuses -again because of their failure to comply with different, but also gender-based norms. By refusing to comply with oppressive gender-based expectations, a woman is implicitly or explicitly rejecting the misogyny and/or the homophobia that she is subject to in her country of origin. This should be seen either as an expression of a political opinion or as the basis for having a political opinion imputed to her.
A claim structured in this way has a better chance of allowing a woman to provide an account of her reasons for claiming protection that is an honest reflection of her experiences and allowing her to feel that she has been heard and validated at the end of the process, as well as there being more likelihood of a fair decision. This could enable a more multidimensional determination process and crucially one that both recognises the experiences of GBV many lesbian and bisexual women have undergone and does so in a way that positions them as actors making choices about their lives rather than only victims of violence. This would be enabled by the kind of broad application of "political" that writers such as Crawley have long advocated.
This would also avoid the prurient focus on whether a claimant is "genuinely" lesbian or bisexual that then leads to decision-making based on the internal coherence of a claimant's narrative that is at odds with the reality of how incoherent people's memories are, particularly of traumatic events. At the same time, the particular social group ground leads to stereotyping, and worse, conflicting stereotypes that force claimants to choose between different types of persecution and the corresponding stereotypes. Those problems will certainly not be resolved by greater use of the grounds of political opinion, but there is more likelihood that the claimant will not be encouraged to conform to stereotypes that deny her agency.

CONCLUSION
As a prelude to concluding, some qualifications are required. First of all, it should be pointed out that this approach is not relevant to the claims of all women. Refugee scholars have explored the heterogeneous nature of gender-based claims. 118 The arguments here, based on the authors' research, are made specifically in relation to the subset of claims by lesbian and bisexual women who have experienced GBV and who are not currently encouraged to integrate that experience into their claim.
Secondly, attempting to give women claimants greater agency by using the political opinion ground may create or reinforce new stereotypes. At present, the prevailing narrative in international human rights and refugee law remains that of the vulnerable woman who is a victim of her state or culture and who needs "special protections". 119 Events in Afghanistan at the time of writing are a good example: on the one hand we see women's rights, often juxtaposed with the needs of children, repeatedly at the forefront of the reasons why the West must intervene and support people both in leaving and once they are in the UK. 120 Undoubtedly, the "womenand-children" framing device creates support for refugees at a moment of urgent need. 121 Yet, deploying those tropes perpetuates the "saving women" narratives that in the longer term do not serve women well on an individual basis or in terms of wider discourse and power relations. 122 However, by positioning women claimants, including lesbian and bisexual women, as agents exercising free will, a new paradigm is created that is also problematic. The wretched victim of backward regimes who is saved by the West is replaced by the free-thinking subject who is out of place in her primitive country of origin and is welcomed by the West. 123 Furthermore, some women fleeing GBV, including some lesbian and bisexual women, do secure protection as members of a particular social group, even if only on appeal and after months or years in limbo. Razack points out that "women's claims are most likely to succeed when they present themselves as victims of dysfunctional, exceptionally patriarchal cultures and states." 124 This applies to lesbian and bisexual women as much as those fleeing GBV and assumed to be heterosexual, the difference being the stereotype of the kind of victim that applies in each case. Advocates are unlikely to risk jeopardising the claims of individual women by rejecting tried-andtested cultural stereotypes of particular social groups based on either sexuality or experiences of violence, particularly given the case law and official guidance all of which direct women and sexual minority claimants towards the particular social group ground. 125 However, while there are arguments for this kind of "strategic essentialism" in the context of asylum, these hegemonic and artificial narratives only work for some individuals and even for these few, a successful grant of asylum often comes at a high cost. 126 Finally, in the current climate of disbelief, any improvement to asylum decision-making seems unlikely, particularly through strategies likely to be seen as opening the "floodgates" to large numbers of women, and it is not surprising if short-term expediency trumps more realistic narratives that speak to the totality of claimants' experiences. 127 Despite these provisos, most people would agree that the cases described above demonstrate a flawed system. Even if one sets aside the harm to the claimants, each of these women ultimately received refugee status but after months or years of waiting and at huge cost to all concerned -financial costs, but far more importantly, costs in terms of claimants' time, physical and mental health. Returning to the case of Jayne described above, if she had been assessed in the way proposed here, her claim would not have been initially refused because it would have been considered on the grounds of political opinion as a woman fleeing GBV because of her refusal to conform to gender norms in her country of origin. The determination process may still have been intrusive but less so, and would not have required her son and partners to provide testimonials stating that she "is a lesbian".
It remains the case that many women experience GBV, including for reasons that relate to their sexual orientation. They should be able to claim protection in a process that acknowledges the reasons why they need protection rather than questioning that prioritises who they are. Avoiding the default use of the particular social group ground in favour of political opinion may lead to improvements in ensuring that all women, regardless of their sexuality, are subject to the same credibility bar in terms of showing their well-founded fear, and presage a renewed focus on the individual nature of women's experiences and claims that is central to refugee status determination.
At the moment, many lesbian and bisexual women are required to prove the wrong "fact" -that they are lesbians or bisexual rather than that they are at risk of GBV because they are lesbians or bisexual. Whether or not they have a "well-founded fear of persecution" is lost in the question of whether or not they are a lesbian or bisexual woman. As a result, they come out of the process, often after years, with a decision that effectively says to them: "Yes, you are lesbian or bisexual" (if they are granted refugee status) or "No, you're not" (if refused). This undermines the right to self-determination which every individual should enjoy. It means that some women secure protection at the expense of recognition of who they are and what they have been through. What women need to hear is: "Yes, we recognise that you are at risk of GBV because you are a lesbian or because you are bisexual".
Refugee Survey Quarterly 27