Improving SOGI asylum adjudication: putting persecution ahead of identity

This article addresses the concern that decision-making in sexual orientation or gender identity (SOGI) asylum claims in Europe is often unfair, and that one way to remedy this is by improving the guidance provided to decision-makers when interpreting the Refugee Convention in respect of these claims. We begin by interrogating a number of different decision-making guidelines and models to assess whether they contribute to a fairer asylum system. We show that, for claims based on SOGI, success invariably depends on the decision-maker’s assessment as to whether a claimant is, or would be perceived in their home country to be, a member of a SOGI minority, and so belongs to a Particular Social Group. Such focus risks neglecting analysis of the actual risk of persecution. We set out our own recommendation for a fairer approach based on prioritising SOGI-speciﬁc Country of Origin Information (COI) and the risk of persecution, rather than focusing on whether applicants are ‘genuinely’ members of a SOGI minority. We argue that this will lead to fairer outcomes that are less likely to be over-turned on appeal and more consistent with the Refugee Convention’s spirit and letter.


HOW DO YOU SOLVE A PROBLEM LIKE SOGI ASYLUM?
This article arises from a growing concern, reflected in scholarship, that decisionmaking in sexual orientation and gender identity (SOGI) asylum claims in Europe is often unfair (a term we will define in the following section) and fails to comply with the spirit if not the letter of the Refugee Convention. 1 One reason for this may be the interpretation and application of international refugee law in different jurisdictions, suggesting that decision-making might be improved by changing the guidance provided to decision-makers when interpreting the Refugee Conventionin respect of these claims. 2 There is a range of conditions that needs to be met to decide a claim for asylum positively. These include that a well-founded fear of persecution has been established, that such fear is on account of one of the five Refugee Convention grounds (namely race, religion, nationality, membership of a particular social group (PSG), or political opinion), and that the home country authorities fail to provide protection. While all of these -and other -factors need to be in place and are non-negotiable, the balance between them may and does vary according to the decision-making body or author of the guidance in question.
This article focuses on asylum claims based on sexual orientation and gender identity (SOGI). 3 In such claims, the focus for proving credibility is generally the question of whether there is a credible fear of persecution on grounds of membership of a PSG -the group being a SOGI minority. 4 Following this, adjudicators' reasoning tends to centre on the claimant's identity -whether they are or are not 'genuinely' a member of the PSG in question. The exact details of the PSG may change, but broadly speaking it is LGBTIQþ individuals in X country where identities or behaviours related to LGBTIQþ individuals are criminalised or subjected to serious forms of legal or social discrimination.
To address the complexities and perplexities in making this assessment, there have been attempts to advance models and guidelines that can support decision-makers. Many of these attempts are born from the keen desire to find a single method that will allow for fairer SOGI asylum adjudication, but, as we will explore below, they have often been applied in diluted or inadequate ways, in the attempt to reach quicker and more straightforward decisions. In addition, some of the attempts we consider also fail to recognise the strong cultural and individual variations that exist in terms of SOGI identity, behaviour and personal development. 5 We thus explore not only the limitations inherent to some of the models discussed in this article, but also the shortcomings in their application.
Following many reports of inappropriate questioning and evidentiary practices in SOGI asylum claims across Europe for many years, 6 particularly in relation to the veracity of claimants' alleged SOGI, there have been attempts from within as well as beyond Europe to advance models and guidelines that can support decision-makers by clarifying what their role should entail. Those discussed here are the UNHCR SOGI Guidelines (Section 3.1), LaViolette's work in the Canadian context (Section 3.2), Middelkoop's work in the Dutch context (Sections 3.3), Chelvan's DSSH model -standing for Difference, Stigma, Shame and Harm -in the UK context (Section 3.4), and Jansen's work in the Netherlands (Sections 3.5). Other academic works, policy documents and Third Sector initiatives have attempted to put forward recommendations and guidance tools for SOGI asylum, including, for example, the UK Home Office, 7 the European Asylum Support Office (EASO), 8 the International Commission of Jurists, 9 Berlit, Doerig, and Storey, 10 and the Centre for Transnational Development and Collaboration. 11 However, while we refer to these in passing where relevant, these approaches focus on specific jurisdictions, incorporate or rely on the approaches we have chosen to discuss, or are less influential. In selecting models and guidelines for interrogation, we have attempted to choose ones that are as diverse and distinct from one another as possible, to avoid duplication in our analysis. Moreover, we recognise the significant differences in the aims and scope of the selected frameworks: while some provide structured guidance to decisionmakers to apply on a case-by-case basis (DSSH), others instead resemble a set of principles or checklist of factors for consideration. Therefore, we use 'models' and 'guidance' in a broad sense to cover a range of attempts by stakeholder individuals and organisations to improve SOGI asylum decision-making by focussing on the application and interpretation of international refugee law. In this article, we assess these attempts at clarification to ascertain their fairness, identify scope for improvement, and suggest avenues for developing more encompassing and effective guidance for decision-makers. Our concern in this article is not with the law -the Refugee Convention, European Union Directives and the expression of these and other instruments in national law; rather, it is with how these are translated into policy guidance and implemented by decision-makers in Europe. Our experience shows that problems with SOGI asylum extend beyond the denial of protection to people who should be granted refugee status within the current framework. Our research demonstrated that in many instances claimants ultimately receive protection, but this comes after a lengthy, costly and traumatising asylum process that involves interrogation about the individual's SOGI that is demeaning and should not be a part of any fair or effective asylum process. It is unfair because there is no comparable legal process in which heterosexual or cis-gender people are asked to prove their SOGI, and it is ineffective in the face of society's growing recognition that SOGI are fluid and cannot be proved in an immigration interview or court of appeal.
In Section 2, we explain our methodology and criteria for assessing the fairness of various models of SOGI asylum decision-making. In Section 3, we outline the key elements of the more prominent models and approaches, starting with the UNHCR guidelines, before considering some of the approaches set out by academics and practitioners working in this field. We juxtapose these with the SOGICA fieldwork and, based on the standards identified in Section 2, assess their fairness. We conclude Section 3 by summarising the advantages and disadvantages of existing approaches. In Section 4, we put forward an alternative approach, which we also critique.

METHODOLOGY AND DEFINITIONAL CONSIDERATIONS
The models we have chosen to analyse relate to decision-making in different jurisdictions, including ones outside Europe. However, our exploration is based not only on documentary analysis of publicly available materials, but also on extensive fieldwork we carried out in Germany, Italy, the UK and at EU and Council of Europe levels, between 2017 and 2019 as part of a four-year socio-legal research project -SOGICA. 12 The analysis therefore has a European focus, although we hope that our argument and recommendations for improving decision-making will be of wider interest and application.
The flaws in European decision-making processes that we identify reflect what we heard from participants in the SOGICA project. The research team conducted 143 interviews with policy-makers, decision-makers, members of the judiciary, legal representatives, NGO activists, and SOGI asylum claimants. Interviews were semi-structured and questions looked beyond the facts of the decision-making process to ask about claimants' perceptions of various aspects of decision-making. A standard question asked of claimants we interviewed was "In your view, what needs to change so that SOGI claims are treated more fairly?" Similarly, we held 16 focus groups in which participants were asked, amongst other questions, "Do you think there are changes needed to make it fairer for people to claim asylum because of their sexuality or gender identity?" Fieldwork also included 24 observations of asylum appeals, providing an opportunity to assess factors with an obvious bearing on fairness such as whether the proceedings were explained to the appellant, the adequacy of any interpreting provision, and whether the appellant was given a reasonable opportunity to provide an account of their experiences. In addition, we conducted two online surveys of SOGI asylum claimants and refugees, and professionals working with them. These were (if wished) completed on an anonymous basis and provided a wider pool of quantitative and qualitative data to complement the researchers' qualitative interviews, focus groups and observations of asylum appeals. Finally, further information was sought through Freedom of Information requests in the three country case studies, with responses to these requests received in the case of Germany and the UK. 13 While acknowledging the legal procedural and substantive differences between the jurisdictions on which we focussed in our research, asylum decision-making often entails three key steps in European countries: an initial or screening interview (during which the claimant is required to provide basic information about their identity, journey to host country and risk of persecution, often by filling in a form); a main interview (where an interviewer collects a broader body of evidence through a generally lengthy interview to determine whether the criteria for granting international protection has been met or not); and, where the initial decision is a refusal, the possibility of an appeal hearing (where a judicial authority considers possible grounds to reconsider the initial decision). 14 Throughout these three key steps, the evidence gathered is used to establish whether -or not -the elements of the Refugee Convention definition of refugee are present in the case in question.
The concept of fairness is at the heart of this process and, of course, may be defined in many ways. For the purposes of SOGI asylum decision-making, we prioritise three elements. First, decision-making processes and outcomes must be consistent with the criteria and standards set out in the Refugee Convention as it sits in dialogue with international human rights law. 15 Secondly, decision-making should be based on equality and non-discrimination. 16 This means it should be consistent and recognisably so, without differences in outcomes that cannot be justified based on the facts of the case. Finally, decisions should be made in a way that upholds the dignity of the individual. At the heart of the international human rights framework is the explicit recognition of the 'inherent dignity and equal and inalienable rights of all  19 Dignity is connected to 'the idea that each human being is a maker of a life plan, and that each should be treated as an end and none as the mere instrument of the ends of others'. 20 These three criteria for assessing fairness -consistency with international refugee law, with principles of equality, and with respect for human dignity -underpin our assessment of the models we consider below. We take an interdisciplinary approach to addressing the concerns of this article, recognising that while legal scholarship is the basis for exploring asylum decisionmaking, a more comprehensive analysis of SOGI-based claims must look to other fields, in particular queer scholarship and intersectionality theory, fields where scholars have emphasised the heterogeneity of individual identities and experiences. Queer theory has been a rich source of inspiration for this article, first and foremost in challenging static and binary notions of sexuality and gender identity, and interrogating heteronormative narratives. 21 In doing so, it has paved the way for the developing field of queer migration scholarship, mapping the complex ways in which the movement of people is implicated in sexual and gender identity formation. 22 Debates around homonationalism, homonormativity and 'pink-washing' 23 are of particular relevance in understanding how SOGI persecution and protection is situated in relation to neoliberal agendas -a thread we return to in our conclusion. 24 From intersectionality we take the now widely-accepted argument that to focus on only one identifier or characteristic in considering an individual's situation -here the situation of a person claiming asylum -results in a narrow and partial understanding of the claim. 25 We integrate these insights at appropriate points throughout the article, but here simply use them to highlight the need for any fair model of SOGI asylum adjudication to recognise that this is not a homogenous group of claimants.
Before proceeding, we briefly explain some choices of terminology. First, we opted for 'asylum claimant' as opposed to 'asylum seeker' owing to the increasingly negative charge the latter expression has gained in popular discourses. Secondly, we favour 'SOGI asylum claim' to 'LGBTIQþ asylum claim', in order to focus on characteristics rather than Western-biased identities and identifiers. 26 Thirdly, although sexual orientation and gender identity are separate characteristics, they raise connected issues, 27 namely in relation to individuals 'failing to conform to genderprescribed social norms and mores or for claiming their rights'. 28 In the next section, we consider some of the existing approaches to SOGI asylum decision-making that have been proposed, before reviewing them against the standards outlined above.

KEY MODELS AND APPROACHES TO SOGI ASYLUM
In this section, we discuss in some detail a number of models or approaches that are sufficiently different from each other to provide useful points of comparison. In this way, after first considering the UNHCR SOGI Guidelines that are the necessary benchmark and reference point for any model, we discuss the works of: LaViolette (an early advocate of lines of questioning focusing on feelings and emotions); Middelkoop (who favours a focus on persecution); Chelvan (who developed the focus on feelings into the four-part DSSH model); and Jansen (suggesting a shift away from the focus on feelings to prioritise other ways to establish credibility). 3.1 UNHCR as the star in the show The UNHCR 2012 SOGI Guidelines arguably constitute the most encompassing and balanced attempt undertaken to offer asylum adjudicators culturally sensitive and overall appropriate guidance on how to deal with SOGI asylum claims, both from a procedural and substantive perspective. 29 Building on a 2008 Guidance Note, 30 and born out of a desire to offer decision-makers a comprehensive reference point on practically all aspect of SOGI claims, they offer valuable and detailed advice on virtually all aspects that asylum adjudicators should bear in mind in this type of claims, including terminology, the notion of 'well-founded fear of persecution', the role of laws criminalising same-sex relations, the (ir)relevance of 'concealment' of one's SOGI, the Convention grounds that can be used by SOGI refugees, credibility assessment, the possibility of internal relocation, and several procedural and evidentiary issues. The Guidelines have thus been crucial in informing and empowering public authorities and the Third Sector, and in equipping them to deal with the particular nature of SOGI asylum claims. They are usually referred to in positive terms, for example, as usefully cautioning adjudicators against stereotyped decisionmaking. 31 That does not make the Guidelines exempt from criticism. For example, evidentiary matters (covered in paragraphs 64-66) might be more thoroughly addressed to offer queer-sensitive guidance as to the range of means of evidence claimants should be expected to provide and that authorities may request, as well as how such means of evidence should be used. This is particularly important in the light of the lack of positive guidance from other sources on this matter, and the fact that SOGICA fieldwork unearthed the inappropriate use of means of evidence, including questions of a sexual nature in Germany, a request to display signs of torture during interviews in Italy, and selective use of evidence submitted to the UK Home Office. 32 In countries where the UNHCR SOGI Guidelines are followed relatively closely, like Italy, one may see sensitive lines of questioning and sensible approaches to assessment of PSG membership. Indeed, Cristina, a UNHCR officer, told us that: [The main issue is] The ability to do an open-ended, culturally sensitive interview [. . .] it is down to the skill and the ability of the interviewer to think that every case must be seen as separately, that every case is new and give the person the possibility to speak, to tell, not to stop at what in a swift manner can be seen like the usual story. 29  Yet, in Italy we also noted the tendency by decision-makers to search for elements in claimants' testimonies corresponding with a plausible testimony (in Western terms), with hints of painful journeys of sexual awakening and inner struggles to reconcile sexuality with religious belief. 33 For example, we heard that: the Territorial Commission thinks from the point of view of who is European.
[. . .] Homosexuality must be the White one, which we've always known, the one we've always seen before our eyes or that we've read about, maybe we don't know it up close, but we can't think of it the same way as a Black guy who maybe doesn't think of himself as a homosexual, cannot even say the word homosexual because maybe he has years of internalised homophobia and it is difficult for him to understand and express it. 34 This is not, however, attributable to the UNHCR SOGI Guidelines, but rather to the way they are used (and possibly misused or ignored). 35 Although the Guidelines themselves establish that it is 'essential that decision-makers understand both the context of each refugee claim, as well as individual narratives that do not easily map onto common experiences or labels' (para. 10), decision-makers often use them selectively to require narratives to adhere closely to the 'statistically most probable' features in the experiences of SOGI asylum claimants. Consequently, relying on the UNHCR SOGI Guidelines as the main tool in such claims does not seem to be a guarantee of quality decision-making, despite the quality of the Guidelines themselves. It is therefore important to explore alternative models put forward to assist in assessing SOGI based claims.

LaViolette and the Canadian template 36
While the UNHCR Guidelines may be the primary and official source of guidance for decision-making on SOGI asylum, the work of Nicole LaViolette is seen by many as laying important foundations for adjudication in this field. There is no explicit 'LaViolette' model, but an approach can be extrapolated from a reading of her work, in particular her paper written for the Immigration and Refugee Board in Canada in 2004 and updated in 2015. 37 Although it was written specifically for the Canadian asylum context, and thus aiming primarily at addressing the range of shortcomings of Canadian SOGI asylum decision-making at the time (in assessing, e.g., claimants' credibility and the persecution threshold), both the paper and LaViolette's approach 33 Ibid. in general have had a broader reach and have certainly influenced SOGI asylum literature internationally, 38 as will be seen in the discussion of other approaches below. LaViolette hones in on the difficulty that SOGI claimants have in demonstrating not only that they have experienced discrimination, but also that they meet the threshold for persecution because of the lack of independent documentation of the human rights abuses they face: In sum, there continues to be significant obstacles facing sexual minority claimants in meeting the threshold of persecution rather than discrimination. One of these hurdles is the fact that independent human rights documentation continues to be difficult to obtain for many parts of the world. 39 This underlines the need for rigorous and reliable COI in these claims. This was confirmed by one of the SOGICA interviewees who identified the 'lack of knowledge of countries of origin and COI' as a factor relevant to the outcomes of claims and went on to prioritise the need for COI on SOGI -preferably on all categories of SOGI (lesbians, gays, trans, etc.), but if not possible, at least on some of those identities (Helen, EASO).
In the examples taken from Canadian case law, LaViolette shows her understanding of the value of an intersectional analysis, pointing out that '[m]any identity markers may intersect to create particular vulnerabilities to persecution', 40 before moving on to consider how PSG membership may be established. She confirms that it is the responsibility of the claimant to provide evidence to support their claim and that the burden of proof lies with them, while also highlighting the need to avoid cultural and other biases in denying the claimant's plausibility. She further argues that: decision-makers should be very cautious about assessing a claimant's membership of a PSG based on the perception of the claimant's physical appearance, mannerisms or manner of dress, or on any expectations of how they are to behave. Such conclusions may constitute stereotyping unless based on a sound evidentiary foundation, such as evidence of how the claimant would be perceived in his or her home country. 41 Here she criticises stereotyping on the part of decision-makers, while at the same time highlighting that what matters is the perception of the claimant in their country of origin. Despite this cautioning, stereotyping on the part of decision-makers remains prevalent and affects outcomes: Diana, claiming asylum in Germany, was told by her interviewer 'You don't look like a trans', while Mariya, an NGO worker in the same country, confirmed that 'even if they [decision-makers] deal with refugees on a daily basis, they still have internalised, unconscious stereotypes and ideas and that cannot be eliminated all at once. And that always plays a role in the assessment'.
LaViolette thus concludes that: The one aspect of the lives of sexual minorities that is universal is the pervasive societal rejection of their sexual orientation or gender identity. There is no country where LGBTI can grow up free of prejudice, discrimination, persecution or repression. The result is that most LGBTI will struggle with their sexual orientation or gender identity at some point in their lives, and that struggle will often move them away, or place them in opposition of their families, friends, communities and society in general. It is suggested that inquiries into areas that touch upon the claimant's personal experience as LGBTI will provide the strongest basis for assessing the credibility of the claimant on the issue of whether he or she is a member of the particular social group. 42 This focus on personal experience corresponds to many of the testimonies we heard: [Credibility] is based solely on the applicant's story and, when the applicant's personal story is considered coherent, plausible and sufficiently detailed with reference to what he declares of his orientation and also of the country from which he comes, then credibility is normally accepted. 43 LaViolette nowhere sets out a model for decision-making or template for decisionmakers, however, in an earlier variation of the paper dated 2004, she proposes a model to question claimants about their sexual orientation in order to establish PSG membership, again stressing that these are areas of inquiry and that there can be 'no 41  true answers to these questions'. 44 In the 2004 paper, there are three suggested areas of inquiry: personal and family; lesbian and gay contacts in the country of origin and in Canada; and experience/knowledge of discrimination and persecution. Questions for the claimant are suggested for each of these areas, such as (for personal and family) 'At what point in their lives did they come to realise their homosexual or bisexual sexual orientation?' 45 and (for experience/knowledge of discrimination and persecution) 'What does the claimant fear will happen to them if they return to their country of origin?' and 'What leads them to believe they will be persecuted?' 46 These questions correspond to the current focus of interviewers on claimants' internal emotional processes, as discussed further below. LaViolette's work is important for two reasons. First, she is an exemplar for much SOGI asylum scholarship in grouping her discussion, albeit not explicitly, into two areas: factors relating to the potential persecution and the persecutors, such as COI and criminalisation, and factors relating to the identity of the persecuted individual, such as credibility as a member of a PSG. Secondly, her work has been very influential and her approach reverberates in most of the work we discuss below, specifically the focus on identity rather than behaviour as the appropriate basis for questioning. 47

Middelkoop and the Dutch perspective
Writing in the context of the Netherlands, and on the basis of an analysis of 13 cases, Louis Middelkoop asks how the Dutch Immigration and Naturalisation Service (IND) defines the PSG of homosexuals and applies that definition in practice, before outlining his alternative approach to refugee status determination. 48 Aiming primarily to address shortcomings in establishing membership of a PSG, he explains the way that sexual orientation may manifest itself, distinguishing three ways: behavioural, emotional and through self-identification. 49 Pointing out that the PSG may cover people who are 'actually' and also those who are perceived to be gay, 50 he contrasts 'someone who experiences sexual or romantic feelings towards someone of the same sex' and who is gay according to the Refugee Convention, with '[a] person who only engages in same sex activities without experiencing feelings of attraction [and] is not persecuted for who he or she is'. 51 This act/identity distinction has come to dominate many debates about how SOGI asylum claims should be proven, with this debate centring on whether sexual 44  orientation is verifiable by what one does or who one is. 52 Middelkoop does not endorse this approach and goes on to deconstruct the questions and answers in interview transcripts from the 13 cases analysed to show how Dutch officials' questioning appeared to build on the model of homosexual identity formation developed by the psychologist Vivienne Cass, 53 and as discussed in the context of SOGI asylum by Berg and Millbank. 54 Middelkoop says: According to this model, identity is developed in a number of stages: initial confusion about sexual identity, comparison of sexual identity with others, tolerance of own identity, acceptance of it, and pride and integration of the sexual identity with the overall personal identity. It appears from the present research that IND officials understand sexual identity development similarly to this model. 55 The expectation that a claimant be able to describe their sexuality in emotional terms (who they are) and not simply behavioural ones (what they do) can be observed in other jurisdictions as seen in the following interview excerpt where Ruben, claiming in the UK, describes what happened in his Home Office interview: SOGICA interviewer: And this [sexual encounter] is 15 years ago, when you were a teenager?
Ruben: Yes, so, first I don't remember all of it, and second it wasn't like this deep relation, at some point I explained to her, this kind of relationships is different from here in the UK or Europe. It is not like we were going very intimate. And she kept asking me about very detailed information which I don't have. It is not only that I don't remember, basically I don't have, I wasn't very deep in this relationship to this boy. At some point I had to explain to her I know him for sex. I have to bring it frankly. I was trying to explain it in other way, but at some point she provoked me that much that I told her I just knowing him for having sex with him, and that is it.
Ruben went on to describe the Home Office interviewer's reluctance to drop this question: And then she returned back to my relationship with this guy [. . .] and I didn't want to be close to him and classmates so I didn't look very suspicious. Somehow she didn't accept that. She asked me like, I stayed with him like two years, and she was like how come two years you don't know what he is interested in. And at this point actually when I get very, I have to say upset, I just told her I need to explain something, I just was meeting him for sex, and that is it. Outside this I was just his classmate.
Here we see what can be the result of expecting asylum adjudicators -with large workloads and in all likelihood covering claims made on grounds other than SOGIto build verification of human sexuality into the asylum interview. Unsurprisingly, the interview becomes a tick-box exercise where the interviewee must provide a standard narrative to meet the interviewer's definition of what a gay relationship looks like.
Middelkoop's study leads him to argue for 'a shift in focus from assessing whether it is credible that the asylum seeker is gay to whether elements in the narrative indicate that the actors of persecution perceive him to be gay'. 56 The question of whether or not the claimant is gay then becomes irrelevant, because what matters is whether or not the persecutor believes the claimant to be gay. The burden of proof remains with the claimant, but now they have to prove that they are perceived to belong to a sexual minority by the persecutor. The relevant questions for decisionmakers 'then become whether the asylum seeker is likely to want to live an openly gay life, whether the authorities will therefore perceive him as gay, and whether he will be persecuted'. 57 Focussing on the persecutor's perception, however, has not been the approach prioritised in other models, including the DSSH model to which we now turn.

Chelvan's DSSH model
Chelvan, a practising barrister in the UK, is the creator of the DSSH model, first presented in 2011, and comprising Difference, Stigma, Shame, and Harm. 58 Chelvan sees these as 'basic characteristics or elements that are likely to be common' in many journeys of SOGI claimants. 59 The DSSH model has been presented as a means to address inappropriate lines of questioning and, simultaneously, inadequacies in credibility assessment.
'Difference' is connected to claimants' non-conformity with 'gender sex-roles, gradual recognition of same-sex attraction or 'gender difference', and 'association with other LGBTIs'. 'Stigma' relates to the non-approval of such difference by family, 56  neighbours, and religious and political leaders on the basis of social, cultural, and religious norms. 'Shame' consists of the 'internalised feelings' and 'avoidance strategies' that claimants develop as a consequence of stigma, including isolation and broader impact of being the 'other'. Finally, 'harm' pertains to the 'specific forms of persecution that may be perpetrated upon LGBTI persons' and, more broadly, negative consequences of physical, socio-economic, legal, spiritual, and medical-psychological nature that SOGI claimants derive from their difference (or are unable to address effectively owing to their difference), both at the hands of State and non-State actors. 60 The model relies on 'trigger questions' that should take the form of open questions related to the four key elements identified, with the aim of eliciting a more detailed narrative. 61 It is true that many SOGI claimants may not feel 'shame' or may not have suffered any harm in the past and simply fear future harm. Furthermore, it has been suggested that the DSSH model has not departed 'radically enough from the assumptions of universal "gay" identities'. 62 Addressing these and other possible points of contention, Chelvan recognises that the model 'will not apply to every applicant', 63 and 'will not provide a one-size-fits-all recipe equally applicable in all relevant cases'. 64 Rather than providing a list of questions, the model includes examples of themes that asylum adjudicators may wish to explore. 65 Chelvan also warns against 'stereotypical recipes' of what is 'difference' and points out that such realisation of difference may come early or late in one's life. 66 We devote considerable space to the DSSH model because of the discussion it has provoked in conjunction with its widespread use, including by UNHCR and EASO (see below). At a national level, Cyprus, Finland, Poland and the UK asylum authorities acknowledge using the DSSH model, 67 and asylum authorities in New Zealand and Sweden have also been reported as using it. 68 In the UK, the Home Office 2016 Asylum Policy instruction on sexual orientation makes reference to stigmatisation, shame, feelings of difference, and harm, but amongst many other issues and without explicit reference to the DSSH model. 69 Although this does not necessarily read as a fully-fledged endorsement of the model, it is undisputable that Home Office decision-makers have been trained on the DSSH model to help them explore credibility in sexual orientation asylum claim interviews in an effective and sensible way, 70 and that they use it in their asylum decisions. 71 One of the participants in the SOGICA project confirmed this as follows: So, I would sort of say 'right, I am now going to ask you about your sexuality' [. . .] then we would move on to talking about the realisation of their sexuality, so sort of beginning from their childhood. . . the method that we use, it is called the DSSH method [. . .] So, with the initial difference, what we are trying to establish is how they felt growing up, how was their childhood different from those of other people around them. [. . .] with LGBT applicants we learn that often they come from quite conservative countries, so growing up the realisation of their sexuality is quite a painful process for them. [. . .] and it is about talking to them about how they worked through that process. How they came to that realisation. It is the most important part of a LGBT claim. 72 The DSSH model's elements can be seen here in the focus on the claimant's internal emotional processes. However, a different Home Office decision-maker, Emily, spoke of the DSSH model in slightly more cautious terms: 'I would use it, but I wouldn't base an entire decision on it because every case is different'.
Another SOGICA participant saw the adoption of the DSSH model as a positive move: 'to be fair to the Home Office [. . .] the Chelvan model tries to move them beyond stereotypes' (Joseph, NGO volunteer, UK). However, the way the model has been deployed has been criticised by the UK Independent Chief Inspector of Borders and Immigration: 'the Home Office's current approach to DSSH seems to us to be a "half-way house" and rather confusing'. The Inspector recommended that the Home Office '[p]rovides more detail about the DSSH model in its training for caseworkers so that it can contribute to the quality of interviewing'. 73 Responding to a lack of consensus about the model on the part of stakeholders, the Government's reply to the recommendation of the Independent Chief Inspector was that '[w]e will consider the scope for providing greater detail and clarity in our training on the DSSH model but will want to do this in consultation with other partners as there is not necessarily consensus amongst them on the efficacy of the model'. 74 The DSSH model has been applied, if not formally adopted, by authoritative sources: it is an element of the EASO Training Curriculum that cascades learning through a 'train-the-trainers' approach across EU Member States. 75 It is also thought to have been endorsed by UNHCR. 76 Paragraph 62 of the UNHCR SOGI Guidelines states that '[e]xploring elements around the applicant's personal perceptions, feelings and experiences of difference, stigma and shame are usually more likely to help the decision maker ascertain the applicant's sexual orientation or gender identity, rather than a focus on sexual practices'. Yet, there is no clear reference to the model, only to three out of the four elements of this model as 'usually more likely' to be relevant in SOGI claims. It is thus doubtful whether this constitutes an endorsement as such. 77 Moreover, paragraph 63 of the Guidelines advocates 'openended questions that are crafted in a non-judgmental way' and states that 'there is no magic formula of questions to ask and no set of "right" answers in response'. In relation to what questions interviewers should ask, and on the matter of 'non-conformity' in particular, para. 63(v) of the Guidelines states that: LGBTI applicants may have grown up in cultures where their sexuality and/or gender identity is shameful or taboo. As a result, they may struggle with their sexual orientation or gender identity at some point in their lives. This may move them away from, or place them in opposition to their families, friends, communities and society in general. Experiences of disapproval and of 'being different' or the 'other' may result in feelings of shame, stigmatization or isolation.
As Jansen points out, '[f]rom the consistent use of the word "may" in this passage it is clear that UNHCR is aware of the possibility that LGBTI asylum seekers do not struggle with their sexual orientation or gender identity and that they are not LGBTI people, and that the vast majority of individuals we interview will be able to discuss them in some capacity. Those characteristics are difference and stigma. Many people will additionally have experienced shame and harm.
The reason this model is so useful is that it allows participants to conduct interviews in a way that is non-judgmental and employs non-intrusive and non-sexual questions. For RSD interviews, it is also an alternative means of assessing credibility for individuals who are not 'out' or have not engaged in same-sex partnerships. The DSSH model is based upon open-ended questions and employs a non-adversarial approach. 79 (bold as in original text) The DSSH model is thus presented as a nuanced approach, adaptable to the diversity of SOGI claimants, and a positive alternative to the more intrusive, sexualised, and adversarial techniques used by asylum authorities in some countries. It has many supporters, with one suggesting it as an optional protocol to the Refugee Convention; 80 however, the risk of asylum authorities applying it in a simplistic and damaging way has also materialised. While sometimes facilitating positive outcomes, the DSSH model may have the unfortunate effect of narrowing the mind-set of decisionmakers. They may be less accepting of different narratives, less common experiences of sexuality, and accounts that combine socio-economic, cultural or religious factors. Certain expectations may develop, as is patent from what we heard from Qasim, a UK decision-maker: When you have someone who is a genuine LGBT applicant, they are really enjoyable interviews. My last LGBT applicant was really, really enjoyable to talk to, because you are learning so much about the kind of things they feel inside, when they are growing up and that kind of conflict and how they deal with it and how they lived with that sort of sense of shame whilst in secrecy and then how it felt when they were allowed to open up. 81 Qasim's words show a welcome openness to listen to the claimant's own account, but at the same time they show an expectation that certain experiences and emotions 78  will be present -conflict and shame -and that these are the trigger for recognising a 'genuine' applicant. The absence of one or more of the common themes identified by the DSSH model indeed constitutes an ideal pretext to undermine a SOGI claimant's credibility. The feeling of 'difference', for example, may be peripheral to the experiences of many SOGI asylum claimants, and will not be a universal element in all claimants' narratives. 82 However, it is a basis for assessing credibility, as this UK Home Office decision reveals: When asked to explain how you came to realise that you were different from other people, you responded 'I used to read poems like Shakespeare and Ilham Al Mohammed Iqbal'.
When asked to explain how this made you realise that you were different, you failed to explain. 83 Some SOGICA participants saw such an expectation of 'difference' as dehumanising. Alphaeus described how he felt on being asked about feeling 'different' by officials: 'So, in that instance it makes you inhuman somehow. [. . .] Because it's like you were human and now you're becoming inhuman. "How did you start feeling that you're different?" Really? Really? Me, I'm like "no", but you just have to answer'.
Similarly, the absence of feelings of shame may -in the eyes of a decision-makerundermine the claimant's credibility, even if insights from cultural anthropology and psychology deny the absolute need for a 'struggle' to accept one's sexual orientation: 84 in many ways is quite a useful model but then case workers apply it inflexibly at the Home Office [. . .] [two claimants] in their statements, they experienced stigma and they experienced horrible oppression from their families, but they never felt ashamed. So, the Home Office felt they can't be gay because they didn't feel shame and I am thinking 'well, actually, they are in their 20s, they have been here for 4 or 5 years, Pakistani culture is changing, with satellite TV you can watch telly from all over the world, why should they feel shame?' But this is this model of difference and shame and stigma, [it] is applied inflexibly. 85 An expectation of shame may be re-framed in terms of 'inner conflict', as in this decision from Germany: 'The interviewer also wants to know whether he was unsure about this new homosexual side that he had just discovered; and whether this led to any inner conflict about his homosexuality '. 86 Harm is often absent as well, as claimants may have only recently started exploring their sexuality, as Marlen, a lawyer in Germany, told us: 'I have a client who comes from Morocco [. . .] And it was very hard for him to talk about it [his sexuality] [. . .] and then he also told me that he somehow also only started to meet men here'. When decision-makers focus on past harm and not risk of future harm, the 'harm' element becomes problematic.
As noted by Grønningsaeter, there are signs that the DSSH model is used in Norway, and she points out that '[s]uch notions fail to take into account the diversity that exists in human sexuality and therefore creates a risk that the claims for refugee status from lesbian, gay and bisexual asylum applicants are rejected on the wrong ground'. 87 The model has thus been criticised for being used by decision-makers in a way that normalises and homogenises claimants' sexuality and identity and how these should be expressed, thus replacing old stereotypes with new (Western) ones about the -assumed to be stable and linear -ways in which individuals realise their sexual orientation and what emotions accompany that process. 88 This undermines antiessentialist approaches to sexuality, gender and a range of other individual characteristics, which are 'socially constructed and therefore fluid and contestable'. 89 As 'sexual orientation identities are the result of social constructivist forces', 90 SOGI asylum decision-makers need to engage with those forces and all their complexity and nuances. This would produce fairer outcomes in terms of dignity, self-determination and diversity in asylum decision-making. Finally, the model risks being deployed in a way that disadvantages particular groups of claimants. Dawson and Gerber, writing about female claimants in particular, have highlighted the fact that the DSSH model 'continues to perpetuate ontological constructions of non-heterosexuality as linear and ultimately fixed', which affects not only non-heterosexual women but potentially also 'transgender and gender diverse, intersex, asexual and other non-normative sexualities and practices'. 91 Similarly, Amanda, an NGO worker in Brussels interviewed for SOGICA, reiterated this concern in relation to asexual people, questioning how applicable the DSSH model would be in such cases.
Some authors wish to retain the DSSH model but strip it of one or another of its elements. Gomez, for example, argues that 'shame' should be removed from the model because it 'is worryingly susceptible to reliance on stereotypes by adjudicators'. 92 Yet, there is a real risk that any of the elements in the DSSH model can be misused if privileged over other elements, held as essential despite the endless variety of experiences within SOGI minorities, or applied in inflexible ways that rely on fixed, linear and universalized conceptions of SOGI. 93 In this debate, one can see the tensions that exist between the strategic need to set out a strong claim while doing justice to the truth of each individual claimant, in all their complexity. Some thus make an overall cautious assessment, as, for example, Amanda: We have heard mixed opinions so far [. . .] we have heard people are very supportive because it offers some positive guidance [. . .] But then we have also heard some legal practitioners and NGOs who are very worried because it is becoming another dogma [. . .] which then boxes asylum seekers again into certain categories and if they don't prove those elements, then they are again out of the system. [. . .] I think it filled a gap where a gap was totally present, and something needed to be done, but I understand in creating a formulaic kind of approach, yes, and a tick box exercise, then that obviously excludes to the detriment of that applicant. . . yes, it is very tricky.
If the DSSH model has not achieved all that might be hoped, in part because of the way it has been implemented, other approaches should be considered and we conclude this section with a final example.

Jansen and a further perspective from the Netherlands
The emphasis on claimants' feelings and emotions visible in some of the approaches discussed so far represents a clear rejection of the model for decision-making in SOGI cases that was prevalent before the 2014 CJEU judgment in A, B and C, 94 which ruled out questions about sexual activity and sexually explicit material as the basis for determining credibility. However, concerns began to emerge in some European countries in the years following that judgment that the pendulum had swung too far in a particular direction. Such concerns are most clearly articulated by Sabine Jansen in her report for COC in 2019, but have been voiced in relation to official practice in the UK, Norway and Sweden, 95 and reflect a lack of 'room for a different experience'. 96 Using Dutch case law and IND case files, Jansen sets out to understand current SOGI asylum decision-making in the Netherlands as the basis for making recommendations for improving decision-making, in particular addressing existing shortcomings with interviewing, credibility assessment and evidence issues. She draws attention to the Netherlands' Working Guidelines 2015/9 for 'Interviewing and decision-making in cases in which LGBT orientation has been put forward as an asylum motive' and the themes they lay out as guidelines for assessing LGBT credibility. These are summarised by Jansen as (our paraphrasing): private life, including religion; present and past relationships, including country of origin LGBT contacts and knowledge; Dutch LGBT contacts and information; country of origin discrimination and persecution; and 'future after (forced) return to the country of origin'. 97 Jansen traces the sources for these points, including the UNHCR Guidelines, pointing out that they are not prescriptive in requiring that claimants struggle with self-acceptance. 98 She goes on to problematise the use of the Cass model of staged identity development in SOGI asylum, building on the work of Berg and Millbank: This staged model of homosexual identity development poses the problem that it is based on the experience of white, middle-class men in the United States. Although it is presented as generally valid, a model like this cannot be applied to other groups, such as women or people from non-western countries. Even in western countries alone, there are big differences between the experiences of lesbians and those of gay men. 99 Jansen then demonstrates the flawed outcomes of a model of decision-making reliant on a staged process of SOGI minority identity development that differentiates between behaviour and identity and privileges the latter. Specifically, she identifies the following features of the IND decision-making process: a focus on processes of awareness and self-acceptance on the part of the claimant; the expectation of an internal struggle (particularly where religion is concerned); the prioritisation of feelings and identities; and the de-prioritisation of statements by partners and others. To Jansen's mind, a new dominant stereotype has been created: On the basis of the current policy, it is expected that lesbian, gay and bisexual asylum seekers first go through a process of guilt and shame before they come to terms with their sexual orientation. This is unacceptable. Strictly speaking, the question 'What did it do to you, when you discovered you were gay/lesbian/bisexual?' is an open-ended question, but the files clearly show that many asylum seekers do not understand this question. Of course, there are LGBTIs who are ashamed or feel unhappy because they are different from the rest, but it is stereotypical to expect negative emotions in all cases. . .. Jansen further points out that the expectation that claimants have a negative selfopinion is inconsistent with claimants' experiences: 'In many of the examined files, the person concerned says they have not struggled with the sexual orientation and did not have any problems with self-acceptance. Sometimes, there is a sense of relieve [sic] when it becomes clear to them what is going on. The problems lie rather with the others, with their environment'. 101 In addition, this stereotype is not only based on a Western male model of sexuality, but it may also fail to take account of different educational levels and verbal abilities on the part of claimants. 102 This expectation and the fact that it may not correspond to claimants' lived experiences tallies with the testimonies given by some of the SOGICA participants. On the one hand, we heard from Maria Grazia, a decision-maker in Italy, that: The first experience of attraction towards a person of the same sex in my opinion is important because in a context in which homosexuality is defined as a crime or socially reprehensible, it must necessarily have been significant for the applicant, both because it was experienced with shame, and because it is problematic and in any case emotionally strong.
In contrast, when asked if he was scared before his interview with the Italian Territorial Commission, Kamel told us: 'Absolutely not, because I don't feel there is a shame that I have to hide. Even when I was raped in prison. I said it. Even when my father beat me, even when they left me in a 1 meter x 1 meter room and didn't give me food'.
In setting out her proposals for how decision-making should be carried out, Jansen starts by stating that the core requirement for SOGI claimants to articulate processes of awareness and self-acceptance must be abandoned. She continues to propose greater recognition than at present for forms of evidence such as partner statements, photographs and witness evidence. She prioritises the 'four-eye' principle of assessment where credibility is in question, whereby a second officer becomes involved with the adjudication process. Jansen also recognises the significance of COI that is disaggregated to a greater extent than is presently the case, calling for COI 'that makes a distinction between different sub-groups'. 103 However, the concluding recommendations are less a proposal for a new and coherent approach to SOGI decision-making and rather a checklist of what to do and what not to do, based on a strong conviction that the current approach goes too far in one particular direction to the detriment of SOGI claimants who do not fit a particular mould.

Taking stock of the options
While there are distinct elements to each of the approaches discussed -and these were the basis for our selection -they are not conflicting models or approaches but rather emphasise different elements of the Refugee Convention definition and UNHCR SOGI Guidelines as the focus for decision-making. Moreover, they largely and properly include many of the same elements that will always need to be included in the adjudication process, many of which were included in the SOGICA project policy recommendations, such as avoiding stereotypes and the importance of COI for these applications. 104 However, two of the models particularly indicate a course that is based on interrogating the claimant's feelings as the basis for establishing credibility. Both LaViolette and Chelvan argue that the impact of societal and/or State discrimination and/or persecution will impact on individuals in a way that is likely to have common elements that are the basis for credibility questioning. LaViolette suggests as a possible area to explore: 'What did they [the claimant] personally believe about homosexuality at the time they came to realize that they were lesbian or gay?' 105 Chelvan suggests that '[i]n most cases, it may be helpful in the credibility assessment process to focus on: The applicant's feelings about her/his own sexual orientation or gender identity, including changes in this (gradual self-acceptance vs. gradual internalisation of stigma, or both)'. 106 As discussed above, these approaches correspond with the assumption of a staged process of SOGI development that is implicitly based on the work of Vivienne Cass and involving an internal conflict which may or may not lead to final acceptance of SOGI on the part of the claimant.
As we have shown, not all SOGICA participants were able to articulate a SOGI narrative in these terms. The other two writers we have discussed -Middelkoop and Jansen -take a different approach. While Middelkoop appears to begin with the assumption that PSG should be the focus, launching directly into a consideration of manifestations of sexual orientation, 107 he ultimately questions whether it is 'appropriate for a government agency to determine who is and is not actually gay'. 108 The alternative Middelkoop suggests is to focus on whether a claimant would be perceived to be gay: 'a refocus on the outward and the visible and whether it attracts persecution'. 109 Jansen too problematises approaches that centre on the claimant's narrative of awareness of their SOGI. She suggests greater reliance on selfidentification as well as other forms of 'evidence'. 110 Our research findings suggest that the approaches outlined in this article, while helpful in promoting fairer SOGI asylum adjudication in many cases, would not result in an effective process in all cases. Specifically, they fail to conform to the elements that underpin our conceptualisation of fairness for the purposes of this analysis as set out in Section 2, nor do they adequately recognise the heterogeneity of SOGI asylum claimants and their experiences, thus failing to adequately reflect insights from queer migration and intersectionality research fields. We thus need to consider an alternative approach.

Persecution, persecutor, or persecuted?
The approaches we have identified as the most common in SOGI adjudication focus on SOGI identity and vary as to whether the focus is on the 'fundamental characteristic' or the 'social perception' approach as the basis for determining SOGI credibility. What is common to them all, and causes difficulties, is that they tend to take the claimant's identity and the question of whether they belong to a PSG as the starting point. This is problematic, as has been recognised by queer scholars, in leading to 'essentialist juridical discourses of asylum [that] produce the refugee as one with a fixed, timeless, and universally homogenous identity'. 111 While the sources we discussed above take pains to recognise the heterogeneity of SOGI, establishing PSG membership inevitably means the focus of the decision-making process is on how SOGI is established. Replacing questions about sexual behaviour with a focus on the claimant's self-awareness and individual narrative (as happened in the period after the judgment in A, B and C) is largely positive, but, as Jansen in particular points out, this risks marginalising some claimants who are not able to meet decision-makers' expectations of a journey of self-discovery. She and Middelkoop prioritise other possible avenues for decision-makers, including those that might demonstrate a perception of PSG membership by the persecutor. However, that is not the only question that needs to be asked for Convention purposes.
If we return to the need to establish a causal nexus between PSG membership and persecution, giving sufficient weight to identifying (the threat of) persecution and its source, alongside the question of PSG and identity, leads to a difference in approach and one that is fairer and corresponds better to the principles of the Refugee Convention. It is after all the threat of persecution by State or non-State actors that is the catalyst for the claimant's need to seek protection. Yet, in many of the cases analysed by the SOGICA team, consideration of the persecutory environment was almost entirely absent. While the prurient nature of decision-making that existed prior to the 2014 CJEU ruling is no longer lawful, it has been replaced by an obsessive focus on other ways of demonstrating SOGI that are also intrusive and often demeaning to the claimant. This is particularly the case in public appeal hearings, where questions such as whether the claimant is or is not a lesbian and, if they are, whether they would live openly if returned, are debated between lawyers, immigration officials and judges, while the individual at the centre of inquiry is silent. In a SOGICA observation of an Upper Tribunal hearing in the UK in February 2018, this debate between the judge and the Home Office presenting officer took place in front of the appellant: Judge: What about the issue of whether she did indeed live discreetly. This is someone who had relationships in Uganda, in this country, who went to groups. It doesn't seem to me that is what is normally [meant] by discreet? Home Office presenting officer: As Lord Rogers made the point, 112 there are infinite gradations of living discreetly. My submission is the appellant for her own personal reasons had conducted herself and would on her return [live] discreetly in Uganda. This is not a particularly shocking example, rather it highlights the small and cumulative ways in which the privacy of SOGI minorities is intruded upon, and the personal details of their lives, including how they would or would not behave in the future, are dissected in public.
Here we wish to clarify that we are not suggesting that the models discussed in this paper ignore the 'well-founded fear of persecution' or fail to recognise the importance of COI in demonstrating that well-founded fear. 113 However, in many contributions, this is or appears to be secondary to the key question: is the claimant credibly lesbian or gay? 114 For example, the work by Middelkoop analysed here, after initial comments, begin with the question of 'manifestations of sexual orientation'.
At this point it is important to consider the context of SOGI asylum decisionmaking, given that the purpose of this article, and of the work we have discussed, is to improve SOGI asylum adjudication in light of the realities of European asylum and immigration systems. In Europe, asylum and immigration officials, regardless of their individual sympathies, will be operating in a climate of hostility to refugees heightened by inflammatory media reports. 115 Equally important, first instance caseworkers and decision-makers are generally not lawyers nor do they have extensive legal training. They operate within resource constraints that prevent them from reading lengthy and nuanced academic or legal contributions such as those discussed in this article. As a result, any approach will inevitably be simplified and condensed into something that is workable for adjudicators who need to keep up-to-date on changing legislation and policy in a number of areas -not just SOGI. This explains the tendency to use one approach, and in recent years this has been the expectation of an emotional journey as reported across a number of European countries and found in SOGICA's research. As one example of this, Jayne's appeal refusal included the statement by the tribunal judge: 'What struck me was that her account lacked any kind of emotional depth or detail of her journey towards her sexuality in a place where such relationships are criminalised and taboo'. 116 One problem is the incongruence between legal and academic discourse, on the one hand, and the policy and decision-making environment on the other. The writers discussed here do not suggest that their approaches provide a single script to be applied to all claims and, as the UNHCR Guidelines point out, '[d]eveloping a list of questions in preparation of the interview may be helpful, however, it is important to bear in mind that there is no magic formula of questions to ask and no set of "right" answers in response'. 117 However, when detailed and nuanced approaches filter down to the level of decision-making, they appear to be both simplified and taken as the approach to be used in all situations. As Jansen points out, and as mentioned above, for many claimants, the problem is not their inner sense of identity, but quite simply a hostile and discriminatory external environment. To focus almost exclusively on the individual claimant's personal struggles and emotional development is to lose sight of the structural factors -including but not only homo-and transphobia, racism and inequality -that are universally prevalent and impact on all claimants. This rightly requires that decision-makers pay adequate attention to historical, geopolitical, social, and economic aspects of homophobic and transphobic violence. 118 Nonetheless, the expectation that claimants experience and articulate these factors in a particular and similar way explains the inconsistencies that are present in decisionmaking, because different individuals are not (and should not be expected to be) able to do this to the same degree or in a similar manner. Nor does this approach have relevance in cases where a claimant is persecuted not because they belong to a SOGI minority, but because they are perceived as being so -the distinction pointed out by Middelkoop and mentioned above. It is thus crucial that decision-makers consider both structural and individual factors that may have led to a risk of persecution and the need to seek sanctuary elsewhere, 119 against the backdrop of good quality and relevant COI.
Having identified what we see as the main problems inherent in the application of current approaches, we now develop our alternative. The element of PSG that entails consideration of a claimant's 'true' or perceived identity sadly remains an inevitable one given the Refugee Convention provisions. However, in addition to encouraging more frequent use of the other four Convention groups in SOGI cases, 120 there is scope for establishing PSG membership through acceptance of self-identification as a default position and more consistent recognition of the principle of the benefit of the doubt. What is needed is good faith on the part of decision-makers in believing that people do not want to leave their homes and families unless they have a very strong reason to do so, something that the principle of the benefit of the doubt -a key principle in international refugee law -requires them to do at any rate. 121 Moreover, and this is at the centre of our proposal, this should be accompanied by a greater focus on conditions in the country of origin (with all that entails in terms of resource allocation and guidance), given the widespread recognition that this is a particular need in SOGI asylum decision-making and the risk that a lack of COI about SOGI persecution will be interpreted as evidence that persecution does not exist. 122 What we suggest here is a clearer delineation between the two elements of the adjudication process: the evidentiary stage (including interviewing) and the decision made on the basis of that evidence. It is the former that needs revision, by focusing more on persecution and the threat of persecution and the evidentiary basis for that. Prioritising COI in the first, evidentiary stage of the claims process would then provide a sounder basis for decision-makers at the second stage of assessing whether, based on the established facts, claimants meet the criteria as laid out in Article 1 of the Refugee Convention.
The need for granulated and up-to-date COI is stressed by many of the authors discussed here, most recently by Chelvan in his thematic review of the Home Office's COI products that was commissioned by the Independent Chief Inspector for Borders and Immigration. 123 In this, Chelvan calls for 'separate sections in the [Home Office] reports with respect to lesbian and bisexual women, trans and gender expression applicants and intersex applicants' to enable a 'user-friendly approach and accurately determine risk to these groups'. 124 We would also suggest that Home Office COI guidance -Country Policy Information Notes (CPINs) -are reformatted to lead with COI information, rather than beginning with an 'Assessment' guidance section which has a subsection on 'Credibility'. 125 This would be helpful in prioritising key country information and avoiding the risk of decision-makers assessing an individual's credibility based on assumptions about their SOGI in advance of consideration of country conditions. This recommendation would probably be welcomed by Daphne, a lesbian woman who spent six years going through the UK asylum system before finally receiving refugee status. In Daphne's view, it was COI that caused the Home Office to cede the case. Her claim was successful on appeal without her having to appear in court, and when asked why she thought that was and what had changed since her initial application, she explained that her lawyers had 'worked very, very, very hard to find somebody to write [a] Kenya report, so when they see that report, and the other things I had put in, [the Home Office] said there is no point for you to come to the court'. Up-to-date and disaggregated data on the specific harms experienced by lesbian women in Kenya was thus critical to establishing Daphne's case. If the relevant Home Office CPIN had provided more up-to-date information about the situation for lesbian women in Kenya, and Daphne's solicitor in her initial claim had used this information, she might have received a fair decision far sooner than she did. 126 Recognising the importance of COI could alleviate some of the dehumanising elements of current procedures. Current common lines of questioning could be avoided, such as 'when/how did you first realise/discover you were LGBTI?' 127 or, as in a refusal by the UK Home Office: You state that you went to church and ignored what the bible said about homosexuality and that you felt 'normal' and 'OK' with this. Given that you are practicing Christian, your failure to raise any potential conflict in relation to your behaviour and belief raises doubts regarding your credibility. 128 Decision-makers might instead start by asking: 'What do you fear would happen to you if you were returned to your country of origin?' Claimants' answers could then be mapped onto what is known about the treatment of SOGI minorities in the country of origin and coherence between the two would be a good indicator of whether refugee status should be granted. This would avoid an excessive focus on the claimant's 'true' SOGI and would help to address the problem of stereotyping. Stereotyping is rightly recognised as a problem in SOGI asylum decision-making, yet this was not adequately addressed by the CJEU's ruling in A, B, and C, which only precluded asylum decisions from being taken solely on the basis of stereotypes. 129 Efforts to eliminate adjudicators' stereotypes are often made through recommendations for (better) training of decision-makers. LaViolette supports what she terms cultural competency training, while the DSSH model has been set out as a chapter of a larger multi-disciplinary training manual. 130 Likewise, the avoidance of stereotypes and the need for better training is included within the SOGICA policy recommendations in recognition of their importance. 131 Yet, it is unrealistic to imagine that individuals can free themselves entirely from stereotype-based assumptions; that will always remain a problem while one person's assessment of another's SOGI remains the core element of the decision. Again, more emphasis on disaggregated and good quality COI may help to address this problem.
Such shift of emphasis would also provide a fairer burden of proof. The UNHCR position is that: while the burden of proof in principle rests on the applicant, the duty to ascertain and evaluate all the relevant facts is shared between the applicant and the examiner. Indeed, in some cases, it may be for the examiner to use all the means at his disposal to produce the necessary evidence in support of the application. 132 However, many States choose to emphasise that the burden lies with the claimant. The UK Home Office, for example, states that '[a]s in all claims, the burden is on the applicant to establish their case'. 133 Greater emphasis on the persecutory environment would require States to invest in producing more comprehensive and up-todate COI, ideally in partnership with one another, with EASO and with experienced NGOs in both host countries and countries of origin. The requirement for asylum decision-makers to assess claims based on rigorous COI would inevitably shift the burden of proof to something more equitable and reduce the frequency of refusals based on very minor inconsistencies on the part of the claimants, as is so often the case at present. 134 At the same time, this shift also incurs risks as we recognise in our concluding remarks.

Concluding remarks
All the models and guidance tools explored above have shortcomings. At the same time, even models that have been criticised on several accounts, such as LaViolette's or the DSSH model, are acknowledged as useful in preparing SOGI claimants for their interviews. 135 The question remains: how to include those who do not fit those models and guidance tools? SOGI asylum claims currently invariably centre on identity. Often demeaning and meaningless debates between lawyers and officials across the person at the heart of it all, about whether they are gay or not, etc., lead to questions that spin out into debates between academics, lawyers, and policy-makers about how SOGI is determined -by behaviour or identity. These are questions that cannot be answered and, even if they could, would not show whether an individual is at risk of persecution. The change that took place in Europe following the 2014 CJEU judgment was positive in one way -largely eliminating the prurient focus on sexual activity -but it simply shifted that focus towards the expectation of a narrative based on childhood memories, moments of discovery, an inner struggle and a process of self-awareness. Whether or not one thinks this shift has gone too far, it is still the PSG component of the refugee definition that takes precedence over the need to demonstrate a fear of persecution. It is still therefore the case that decision-making revolves around the need to establish SOGI, albeit now on the basis of who one is rather than what one does.
Claimants may no longer be asked to give graphic accounts of their sexual history but they are still expected to reveal many other personal and private aspects of their lives in ways that undermine their dignity and which would be unthinkable to most people. SOGI claimants are routinely treated in ways that no other individual who has not been charged with an offence is likely to be treated. 136 Prince Emrah (Germany) told us: 'Usually I am so shy, to tell this stuff. I'm not so open. . . I don't like to talk about this kind of stuff'.
We wish to stress that in critiquing existing approaches, we are not arguing against them, but rather pointing out that if any of these is taken as the only approach, then the result will not be a fair adjudication process for all claimants. The models we have looked at are not competing models; for example, they all emphasise to some extent the need to avoid stereotypes and ask open-ended questions, as well as the need for better training of decision-makers. Equally, all the authors discussed here draw attention to the different elements of the refugee definition, including the fact that what matters is whether the claimant would be perceived to be LGBTIQþ. However, many of the approaches above fail to follow through on the recognition that perception on the part of the persecutor is critical, instead returning to an emphasis on the ways that the claimant's SOGI should be verified -implicitly accepting the legitimacy of this question.
We conclude by recognising the difficulties and possible flaws in our preferred approach to SOGI asylum decision-making. First, and particularly in the context of anti-asylum sentiments, it is likely that policy-makers will claim that if the focus is not on the claimant's 'true' identity but the risk of persecution, then claimants will be more easily able to abuse the system by conforming to the kinds of gay stereotypes that most of us would like to see abandoned. Yet, those stereotypes are already in operation and they do not work for many claimants.
There is also the risk that, by diverting focus from the individual's 'true' sexuality or gender identity towards the risk of persecution in the state they come from, we will bolster the binaries of refugee-producing and refugee-receiving, and of more and less 'civilised' States that queer migration scholarship has helped to debunk. A nowextensive body of work associated with the concepts of homonationalism and 'pinkwashing' has pinpointed the way that States exploit queer bodies and lives to pursue neoliberal agendas. 137 Asylum decision-making is an obvious sphere in which to see this played out and an enhanced focus on COI may make it more likely that 136 It is important to highlight that victims of trafficking and FGM will also have to answer intrusive and potentially re-traumatising questions during their asylum application, and that we are highlighting problems that are symptomatic of asylum determination beyond SOGI-based claims. consistency in outcomes based on granulated COI reflecting the diversity of SOGI claimants' experiences; and, by focusing on the causes of flight, it would avoid the intrusive and demeaning lines of interrogation that so often breach the dignity of claimants. We hope that the discussion this article generates will contribute to incrementally improving the SOGI asylum process.