1 Introduction

The SOGICA project ran from September 2016 until August 2020.Footnote 1 In these four years, the project consisted of different phases: (1) delineating the project’s methodology and theoretical and analytical frameworks, in particular how human rights, feminist and queer studies and the concept of intersectionality can be used as particular lenses for the analysis of SOGI asylum claims; (2) preparing and conducting fieldwork in Germany, Italy, the UK and at EU and Council of Europe levels; (3) analysing the data, writing up the results and producing detailed policy recommendations.

As explained in Chap. 1, we adopted an interdisciplinary, comparative and intersectional approach to explore the social and legal experiences of SOGI claimants. The interdisciplinary approach encapsulates legal and sociological theoretical and analytical frameworks and methods and therefore contributes to the developing field of refugee studies that take a socio-legal approach (Anderson et al. 2014; Güler et al. 2019; Khan 2016; Lukac and Eriksson 2017; Venturi 2017). As McConville and Chui (2007, p. 5) explain: ‘The non-doctrinal approaches represent a new approach of studying law in the broader social and political context with the use of a range of other methods taken from disciplines in the social sciences and humanities’. Employing a wide range of quantitative and qualitative methods,Footnote 2socio-legal approaches look at the social factors involved and the social impact of law and practice (McConville and Chui 2007, p. 20). In the context of these volumes, these approaches are especially useful for exploring the relationship between sexuality, gender (identity) and the law, and their intersections with other social relations of power. As McConville and Chui (2007, p. 22) argue, ‘the law cannot be objectively isolated’. For analysing the complexities of the social as well as the legal experiences of SOGI claimants, doctrinal positivistic approaches focusing on case law would not have been sufficient.

In order to address the issue of disparate (and occasionally low) standards across the EU’s and CoE’s member states in asylum legal adjudication, a comparative approach is necessary (El-Enany 2008; Ferreira and Kostakopoulou 2016; FRA2010a, b; Lomba 2004; Whittaker 2006). As we will explore further in Chap. 4, disparities still exist, even if they have to some extent been addressed by the establishment of the EU CEAS. By focusing on Germany, Italy and the UK, we aim to explore better and worse practices, as well as some distinctive trends that may guide asylum decision and policy-making.

The selection of the three countries was done on the basis of three factors:

  • Volume of asylum claims – these countries are among the top six EU host countries in terms of numbers of applications (EUROSTAT 2019).

  • Different adjudication procedures. While Germany adopts an inquisitorial system (where the decision-maker should take the lead in gathering evidence), and the UK adopts an adversarial system (where the evidence gathering burden is theoretically shared between decision-maker and asylum claimant, but in practice most of the burden lies on the claimant), Italy adopts a mixed system. The inquisitorial or adversarial character of the asylum adjudication system has been identified as a crucial feature that may have a bearing on the outcome of asylum claims (Committee on Migration, Refugees and Population 2009; Independent Asylum Commission 2008; Künnecke 2007; Sonnino and Denozza 2005; UNHCR2005).

  • Socio-cultural-legal context, particularly in relation to SOGI. There still exist significant differences in relation to social perceptions of sexual behaviour and gender roles across Europe as well as in protection provided by law (Gerhards 2005; Giordano 2001; Philips 2001; Waaldijk 2006). By comparing three countries that reflect different approaches to sexuality and gender identity, we were able to explore the influence of these differences on asylum adjudication.

While intersectionality is key as a theoretical concept underpinning this research (Chap. 3), it also guides our methodology. By applying intersectionality as methodology, we follow Matsuda’s approach in that we ‘ask the other question’ (Matsuda 1991). Using an intersectional approach was vital for exploring the socio-legal experiences of SOGI claimants and guided how we developed the different methods we used for the data collection across the three case study countries and at EU and CoE levels.

2 Methods

In order to achieve an analysis that offers both breadth and in-depth understanding, we used a mixed-methods approach (Blanck 1993; Epstein and King 2002; Travers 1999; Travers and Manzo 1997). Data was collected using the following methods:Footnote 3

  • 143 interviews with SOGI asylum claimants and refugees, NGOs, policy-makers, decision-makers, members of the judiciary, legal representatives, and other professionals;

  • 16 focus groups with SOGI asylum claimants and refugees;

  • 24 non-participant contextual observations of court hearings;

  • Two online surveys for SOGI asylum claimants and refugees and professionals working with SOGI asylum claimants and refugees;

  • Documentary analysis of international, European and domestic case law, policy documents, NGO reports, case files, etc.;

  • Freedom of Information (FOI) requests.

All of the above covered Germany, Italy and the UK, as well as the EU and CoE. A range of qualitative research methods were used (Bertaux 1981; Morgan 1998; Seal 2004; Seale et al. 2004). While the online surveys produced some numerical data, qualitative interviews and focus groups offered more in-depth accounts of the legal treatment of the participants’ claims and the impact on their social experience and well-being. Our ontological and epistemological standpoint is that asylum claimants and refugees are experts on the refugee experience. Hearing their knowledge and experience is essential to gain unique insights into those experiences and to verify whether international, European and domestic legal frameworks in place address appropriately their claims (Hynes 2003, p. 13). In total, 157 asylum claimants and refugees participated in the semi-structured interviews and focus groups (64 in semi-structured interviews and 93 in focus groups).

As we aimed to include a range of perspectives in our project, when recruiting participants, we looked for a diverse sample in terms of sexual orientation, sex, gender identity, country of origin and other factors such as ethnicity, religion, age and social class. We also tried to reach participants in different regions and nations (with regard to the UK) within our three countries, recognising the importance of place in terms of both where participants came from and where they settled or resided in the host country. To make our research as accessible as possible, we provided translations of the project flyer, information sheets and consent forms not only in German and Italian, but also in Arabic, Farsi, French, Turkish and Urdu.Footnote 4 Many of the asylum claimant and refugee participants were found through contacts with local, national and international NGOs offering support to asylum claimants, legal practitioners, or through contacts with other claimants. A European-wide network of SOGICA Project Friends that we had launched at the beginning of our research and that consisted of people with experience and expertise on SOGI asylum issues, supported us in recruiting participants for our research by cascading information to their contacts and networks.

In order to be able to base findings on a heterogeneous sample, it was important not to rely solely on gatekeepers,Footnote 5 but use other recruitment strategies. Gatekeepers often fear that their clients are ‘too vulnerable’ and traumatised to participate in research projects and, as we have also experienced, may, with the best of intentions, close the door to researchers. In addition, NGOs might only refer participants who they see as the ‘ideal’ sample (for example, for the reputation of the organisation; Hynes 2003, p. 14).Footnote 6 In light of these factors, to recruit participants, we used a wide range of means such as publishing the call for participants in newsletters, mailing lists, relevant publications, and on social media – our own and those of other individuals and organisations. We also distributed flyers about the research through NGOs, community groups and LGBTIQ+ cafés and bars (although this was the least successful method). Many of the participants were found through snowballing (that is, a participant suggested another participant) and direct contacts.

We are aware that, even though we adopted different recruitment strategies, we may have not reached the most isolated SOGI claimants, for instance LGBTIQ+ people who are dispersed to remote areas where they have no possibility at all to access support groups, or who have to live in concealment because of their living arrangements (living with their family, for instance).

We also noticed some differences with regard to recruiting participants in the three countries due to different structures and political and policy cultures. For instance, in contrast to the UK, where support groups for SOGI claimants have existed for at least a decade, in Italy and Germany the existence of these groups has been a fairly recent phenomenon. In Germany and Italy, most of the support organisations and groups that now exist were set up in or since 2015. Because of the more established nature of organisations in the UK, there has also been more (positive) media interest in SOGI claims and there is also a policy culture and established principles of consultation and transparency between policy-makers and ‘stakeholders’ – at least on paper. And while decision-makers in Germany and the UK were recruited through official means (judicial authorities, government departmental channels, etc.), in Italy this was to a large extent only possible through personal contacts. In general, our participants were self-selecting. Consequently, it is likely that the lawyers and decision-makers (but also NGO and other professionals) we interviewed, and on whose accounts we draw in Chaps. 4, 5, 6, 7, 8, 9 and 10, are committed to addressing failings in the SOGI asylum system, perhaps to a greater degree than a random sample of stakeholders would be.

Furthermore, there were some striking differences in the demographics of the SOGI claimants who participated in individual interviews and focus groups, especially with regard to their country of origin. Many of the Italian participants came from French-speaking countries and all of them came from Africa; in Germany, participants came from Africa, the Middle East, Asia, Eastern Europe and the Caribbean; and in the UK, participants came from Africa, Asia, the Middle East, the Caribbean and Central America. In addition, in Italy it was more difficult to find lesbian participants and cases based on gender identity. This may be also due to the different paths of arrival followed by many transgender people (for example, people from South America who have been living in Italy for many years) (Chap. 5). The different arrival paths might also be the reason why participants in Italy were younger (in the UK, for instance, more people claim ‘sur place’Footnote 7).

It needs to be said that not all of the different identifiers included in LGBTIQ+ are represented in our study. Our sample does not include any intersex claimant and only three claimants who identified as bisexual. Therefore, it needs to be kept in mind that ‘questions of access and recruitment can be central to understanding the “outcomes” of the research’ (Rapley 2007, p. 17). Nevertheless, as the demographics of our asylum claimant and refugee participants summarised in Tables 2 and 3 below demonstrate, we managed to recruit a diverse sample for our research. In Germany and the UK, we were able to recruit almost equal numbers of women and men for individual interviews and more women for the focus groups. Some other differences between the participants of the three countries emerged: the educational level of individual interview participants in Germany was higher than in Italy and the UK (which also seemed the case for focus group participants, but data is missing here; see table below). Also to note is that in the UK a wider range of legal avenues are available to claimants (Chap. 4), with some of our participants waiting for the outcome of a judicial review or fresh claim. Not reflected in the tables below is the claimants’ family status and number of children, as we did not collect information from all the participants about that matter. From the information we have, the majority of participants were single and in each country five of the individual interview participants had children (in Germany about 40% of the focus group participants had children).

All of the interview and focus group audio files were transcribed, mainly by professional transcribers who had signed our confidentiality agreement, although a small number were transcribed by members of the project team. We then uploaded the transcripts, as well as the court hearing observation notes, onto the software programme NVivo. This data analysis software is mainly used for qualitative data analysis and allows for comparative analysis. We coded the interview, focus group and observation files according to a coding framework that we developed after a coding pilot exercise involving a small number of interviews. The software then enabled us to pull out the relevant codes (or nodes, as termed in NVivo) for the analysis and writing up of our results.

In the following, we explain the different methods used.

2.1 Semi-structured Interviews

Most of the 143 semi-structured interviews carried out were one-to-one interviews, but 12 interviews were held with two people: either couples, friends or professional colleagues.Footnote 8 We conducted interviews in a wide range of locations in Germany, Italy, the UK and Brussels to gain understanding of regional and national differences.Footnote 9 The overall number of participants interviewed in each category of participants is summarised in Table 1:

Table 1 Number of participants
Full size table

These in-depth interviews were semi-structured, that is, we followed an interview guide but remained flexible in terms of the wording and the order of the questions, and left space for discussion of matters not raised by our questions. The purpose of the guide was to provide direction, ensuring that the interviews focused on the crucial topics we aimed to explore. In contrast to structured interviews, which consist mainly of closed-ended questions, semi-structured interviews allow the interviewer to find out about feelings and perceptions of participants that do not fit into pre-chosen options. Furthermore, too much standardisation can inhibit building trust and rapport (Burns 1994, p. 278). Semi-structured interviews focus on the participant’s perspective rather than the researcher’s and allow participants to use their own language to describe their experiences and social reality.

These are some of the advantages to a semi-structured and more flexible interviewing approach. However, a corollary is that coding becomes more difficult and the interview data is less comparable than it would be with structured interviews (Burns 1994, pp. 278–279). Entirely open-ended interviews, on the other hand, would have been too loosely structured for our purposes (Burns 1994, pp. 279–280). We followed some of Burns’ advice for questioning techniques, for instance, in reflecting back to the participants what they had said, and in using descriptive questions (for example, describing people and events) and probing questions such as ‘Can you tell me more?’ We started interviews in a friendly and supportive way, allowing participants to be in control of the flow of the information (Burns 1994, p. 281). It was important to follow what our participants said, rather than impose a predetermined agenda in order to get sufficient detail and depth of data (Rapley 2007, p. 18).

Our approach recognises the interactive nature of data collection, and rests on the assumption that an interview is always a joint production of accounts (Rapley 2007). We followed the approach Rapley (2007, p. 26, original emphasis) calls ‘“engaged, active or collaborative” interviewing’. It was vital to us to respect participants’ privacy by not asking overly personal questions. Furthermore, being honest and encouraging dialogue about what our study could achieve and what the limitations were, were essential, in our view, to building a trusting relationship (Krause 2017). We engaged with our participants’ discourse by also bringing in our own perspectives (we thus question the ‘neutrality’ of the researcher – see more on our positionality in Sect. 2.3).

As Krause (2017) suggests, especially when conducting research with refugees who have experienced human rights violations, it is vital to enable participants to speak about the issues that are important to them. Krause thus argues that ‘[c]rucially, when participants can speak out about issues that are relevant for them, they are not treated as “data sources” but as persons’ (Krause 2017, p. 20). This issue is also important from an intersectional perspective. For this reason, even though our research focused on sexuality, gender identity and ‘refugeeness’, we also asked participants questions addressing other identifiers such as ‘race’ and religion. We ‘tested’ the interview guides with several pilot interviews, asking the participants how they felt about particular questions (Burns 1994, p. 281), and amended the guide according to the feedback that we received.

SOGI asylum claimants and refugees were asked open questions about their social and legal experiences with regard to their asylum claims. In order to minimise the levels of stress, we tried to keep the interview length to an hour. We made participants aware that they could stop the interview at any time or choose not to answer particular questions. The semi-structured interviews with policy-makers, decision-makers, members of the judiciary, lawyers and NGO workers explored these actors’ positive and negative practical experiences with SOGI asylum claims. As Chaps. 4, 5, 6, 7, 8, 9 and 10 will demonstrate, these interviews provided in-depth data. Most participants gave permission for their interviews to be audio recorded and transcribed. In a few cases where participants did not give permission, notes were taken. The interviews (and the focus groups) were conducted in semi-public places (for example, spaces in universities or quiet cafés) or in places familiar to the participant (local LGBTIQ+, refugee and migrant organisation venues or the offices of law firms).

Table 2 Demographics of interview participants – asylum claimants and refugees
Full size table

2.2 Focus Groups

We conducted 16 focus groups with 93 SOGI asylum claimants and refugees in Germany (six focus groups), Italy (five focus groups) and the UK (five focus groups). Focus groups offer an opportunity for sharing and comparing views in a way that individual interviews do not. Questions in the focus groups concentrated on opinions about the asylum process for SOGI claimants, life in the respective countries and support services available.

In Germany five focus groups were held in English and one in German; in the UK, all five focus groups were conducted in English; in Italy three were conducted in English, one in French and one in a mixture of English and French. In our pilot focus groups we found that smaller groups worked better and were more interactive. Matters of sensitivity and confidentiality are also more manageable in a small group. Consequently, we decided to keep the groups fairly small (each focus group had six participants on average). Most focus groups were recruited through NGOs, thus participants often knew each other. This had the advantage of participants feeling more relaxed with each other; a disadvantage, however, was that they may not have articulated specific points which they assumed everybody already knew, and conversely, differences in opinion may not have been expressed for fear of alienating peers (Macnaghten and Myers 2007, 70).

Table 3 Demographics of focus groups participants
Full size table

In line with ethical standards – guided by the University of Sussex, academic and professional bodies and also our own principles – we tried to make our information sheets and consent forms as detailed as possible (Sect. 2.3). Nonetheless, this meant that going through these documents with a group of participants and making sure that every participant understood what participation involved, was challenging. Some participants found the information sheet and the consent form too bureaucratic, and some participants were illiterate or nearly illiterate, so we allowed time to explain everything carefully and in detail. All participants signed the consent form, but we also made sure that we received consent verbally. In a few focus groups (in Italy) assistants helped with translation, filling out forms and taking notes. We moderated the focus groups and endeavoured to encourage all participants to speak using appropriate strategies. For example, where one participant tended to dominate the conversation, the interviewer would look away from that person and turn their body towards somebody who had contributed less (Macnaghten and Myers 2007). Sometimes participants asked us to use simpler and more basic language, and we also discovered that descriptors such as ‘men loving other men’ and ‘women loving other women’ were sometimes preferable to ‘being gay or lesbian’, as some participants did not identify that way (even if in relation to demographic questions, participants mostly adapted to those categories). In the pilot carried out in Italy, we also discovered that the original opening question we had set seemed to require a ‘capacity of abstraction’ that made some people uncomfortable from the outset.Footnote 10 Therefore, we changed it to a question asking more specifically about the arrival in the host country.Footnote 11

We had some ethical concerns about conducting focus groups with asylum claimants and refugees (Sect. 2.3), based on the potential re-traumatisation participants might experience when listening to other participants’ stories. Yet, according to the feedback we received from participants, these focus group discussions were largely positive experiences. The focus groups offered participants a way to reflect on particular issues within a group setting, sometimes for the first time, and to become aware that other people faced similar issues. Often there were passionate discussions in the groups, especially when participants discussed the decision-making process and how to prove their sexual orientation.

2.3 Observations in Courts

We conducted 24 non-participant contextual observations of court hearings of asylum appeals in Germany (ten), Italy (three) and the UK (11) between February 2018 and April 2019.Footnote 12 Using a guide, these observations focused on how the different actors involved dealt with asylum claimants’ SOGI and related aspects of their claims.Footnote 13 As Burns (1994, p. 265) points out, ‘the functional distinction between participant and non-participant observation is ambiguous as it is impossible to avoid interactions in social situations. The aim of non-participant observation is to observe unobtrusively by minimising interactions with participants. The hearings we observed in Germany and the UK were open to the public and no authorisation was required, however, we needed assistance from lawyers and NGO workers/volunteers to identify forthcoming SOGI asylum appeals, as the detail and nature of appeals is not published. In Italy, we could only observe two tribunal hearings directly, and one indirectly via an interpreter and a lawyer, whom we interviewed when the hearing ended. Hearings are not public and in seven cases the relevant judges did not authorise the attendance of third parties. In these cases, we were denied access despite the claimants’ consent and sometimes despite our presence at the venue of the hearing.

While we tried to observe unobtrusively, to respect and avoid distracting the claimants, at times we made our presence as researchers known to the judges. In all three countries, lawyers and NGO workers fed back to us that they thought our presence made a difference as to how the hearings were conducted and their outcome.Footnote 14 When we refer to the court hearings we observed, we specify the court, the broad geographical location and the year the hearing took place, but omit further details to protect the anonymity of the claimants.

2.4 Online Surveys

Between August 2018 and March 2019, SOGI asylum claimants and refugees in Europe, and those supporting them, were invited to complete an online questionnaire about their experiences with SOGI asylum procedures and wider social experiences. Although these volumes focus on a comparison between the three country case studies – Germany, Italy and UK – the online survey included participants from across Europe, contributing to a broader understanding of the situation of SOGI claimants in Europe. There were two separate surveys: one for claimants and one for people who work with or support them.Footnote 15 These surveys had the following aims: to provide some quantitative data and further qualitative material across Europe to complement the detailed fieldwork described above and inform the project’s policy recommendations; to provide complete anonymity to people who did not feel comfortable participating in interviews or focus groups, but wanted the chance to have their voices included in the research; to broaden the opportunities for contributing to the research to the many individuals who expressed an interest in the project and could not be accommodated throughout the fieldwork, both in the case study countries and in other countries.

We developed the questionnaires according to what is described in the literature as ‘model questionnaire’ (Burns 1994, p. 349), including an introduction, body of survey and demographic questions. The (numbered) questions were grouped into logical sections with a smooth transition between them, and some ‘lighter’ questions at the end (Burns 1994, pp. 349–358). Most of these questions were ‘closed questions’, in other words, participants were able to choose from a range of options, but with the option to provide additional responses in free text form. Some questions had a sliding scale from one to ten (for instance, ‘how easy/difficult is it…’). The surveys were made available in different languagesFootnote 16, and participants were offered a document with a range of answers to potential queries they could have on the surveys.Footnote 17

Information about our research and links to the online survey was distributed through our website, social media, SOGICA’s quarterly newsletters, Project Friends, LGBTIQ+ and refugee mailing lists, and our professional networks. In total, 157 supporters and 82 claimants filled in the online surveys, but not everyone answered all the questions. As the survey was based on a non-representative sample and received a relatively low response rate, we treat the quantitative aspects of the data with care in the subsequent chapters. We use the European-wide quantitative data to provide some background to particular issues, while we use the surveys’ qualitative data to complement our fieldwork in Germany, Italy and the UK. A full analysis of the results of the survey can be found elsewhere.Footnote 18 Here, we provide a summary of the demographics of survey respondents.

Perhaps unsurprisingly, the majority of the 82 SOGI claimants who answered the survey were claiming asylum in the three SOGICA case countries, as it was here where we had the most contacts (Fig. 1). It is likely that some of the SOGICA claimants we interviewed also participated in the survey. However, we also reached respondents in many other European countries. Seventeen percent of the respondents were claiming asylum in Austria, Belgium, Denmark, Greece, Hungary, Ireland, Portugal, Spain, Sweden, Switzerland and the Netherlands. Unfortunately, we do not know where 29% of the respondents claimed asylum, as these respondents did not disclose that information.

Fig. 1
A pie chart, with the values in percentage. Not disclosed. 29. Germany, 28. UK and Others, 17. Italy, 9.

In what country are you claiming asylum?

Full size image

In terms of countries of origin, the survey reached a sample slightly different from our interview and focus group participants (Tables 2 and 3). As in our qualitative research sample, a high percentage of survey respondents were from Uganda (16%), Nigeria (7%) and Jamaica (4%), however our survey reached more SOGI claimants from Syria (9%). Twenty-eight percent of survey respondents came from other countries, including some that were not represented in our other methods’ sample, such as Algeria, Armenia, El Salvador, North Macedonia, Sierra Leone and Turkey.Footnote 19

Regarding gender, gender identity and sexual orientation, the survey respondents’ self-identification was fairly similar to our interviewees’ and group participants’ self-identification: 34% described their gender or gender identity as male, 23% as female, 7% as trans, 5% as queer, and 1% as ‘other’ (for sexuality, see Fig. 2).Footnote 20

Fig. 2
A pie chart with the values in percentage. Gay, 38. Lesbian, 23. Not disclosed, 22. Heterosexual, 7. Bisexual, 5. Others, 3. Queer, 2.

How would you describe your sexuality?

Full size image

In terms of religious identity, age and educational background, the survey respondents’ demographics matched those of our interviewees (Tables 2 and 3). The majority were Christian (Fig. 3), between 25 and 34 years old (45%, Fig. 4), and their highest level of education completed was further or higher education (38%, Fig. 5).

Fig. 3
A pie chart has five segments. The maximum share is 32% for Christians, followed by 29 for not disclosed, 17 for Muslims, and 11 for atheists and others.

How would you describe your religious or non-religious identity?

Full size image
Fig. 4
A bar graph plots the percentage of different age groups. The age group 25 to 34 has the highest bar value of 44%, and the lowest is 1% for the 55 to 64 category.

How old are you?

Full size image
Fig. 5
A bar graph plots the percentage of education level. The highest value is 38% for further or higher education, and the lowest is 1% for primary school.

What is the highest level of education that you have completed?

Full size image

From the 157 people who answered the ‘survey for people who work with or support LGBTIQ+ people claiming asylum’, the majority (41%) were working or volunteering with an LGBTIQ+ organisation or with an organisation providing legal advice and/or representation (19%) (Fig. 6).

Fig. 6
A bar graph plots the percentage for nine organizations. The highest value is 41% for L G B T Q I + organizations, and the lowest is 2% for asylum, refugee, and other migrant organizations.

What kind of organisation do you work or volunteer with?

Full size image

Also here, perhaps unsurprisingly, the majority of respondents were working in the UK (39%), Italy (17%), and Germany (9%), and it is likely that some of our interviewees participated in the survey. Other countries where respondents were working (34%) were Austria, Belgium, Cyprus, Denmark, France, Greece, Ireland, Malta, Norway, Portugal, Slovenia, Spain, Sweden, Switzerland, and the Netherlands.

The data of the surveys, which were created using Qualtrics Survey Software, was analysed using the statistical software package SPSS. However, we did not carry out any bivariate and multivariate analysis or statistical tests to measure the correlation between variables (McConville and Chui 2007, p. 62). For our purposes, univariate analysis (looking at only one variable at a time) was sufficient. When we refer to survey respondents in this publication, we use codes: C corresponding to claimants and S to supporters.

2.5 Documentary Analysis

During the course of our research, we analysed a variety of documents: international treaties and international courts’ and committees’ decisions, European and domestic legislation, case law, policy documents, NGO reports, case files, etc. These documents were available through publicly accessible sources (for example, Refworld) and provided by the research participants. This material was used to support the analysis of the fieldwork primary data in the chapters that follow. We also produced four case law tables for European, German, Italian and UK case law, which contributed to our analysis.Footnote 21

2.6 Freedom of Information Requests

In order to gain more official information about SOGI asylum statistics, training offered to interviewers and decision-makers, etc., we submitted freedom of information (FOI) requests to the relevant authorities in Germany, Italy and the UK.Footnote 22 Freedom of information laws are the means by which members of the public or NGOs are entitled to ask for and receive information held by national governments. Legislation varies from one European country to another.Footnote 23

FOI is ‘a relatively new research innovation in academia’ (Walby and Luscombe 2018, p. 10). In recent years, it has been recognised that FOI ‘is a powerful tool available to researchers’ (Savage and Hyde 2014, p. 315), with more and more researchers using it (Walby and Luscombe 2018, p. 2), although, as some authors argue, researchers are still not making sufficient use of it (Bunt 2018; Savage and Hyde 2014; Turnbull 2015; Walby and Luscombe 2017).

FOI requests can be seen as ‘an innovative research technique for qualitative researchers’ (Turnbull 2015), and are especially useful when combined with other methods (Savage and Hyde 2014), but like all methods of data collection, they have advantages and disadvantages. The main advantage of FOI requests is that ‘[b]y providing a way to access information produced and/or assembled by public authorities, freedom of information requests allow data to be accessed that otherwise could not be, at least without lengthy negotiations with public authorities’ (Savage and Hyde 2014, p. 308). FOI allow researchers to gain access to information and documents that are not publicly available, especially valuable in situations where it is difficult to gain access to gatekeepers, as was the case in Germany where we could not gain access to BAMF staff.

However, there are also disadvantages, as we discovered. Making FOI requests can be time-consuming when public authorities do not provide the information needed and the researcher is ‘fishing around for information’ (Turnbull 2015). The path that researchers need to take to gain the required information is unpredictable and dependent on the person who coordinates the request (Walby and Luscombe 2018, p. 10).

Our FOI requests were drafted in consultation with relevant stakeholders (in Germany, the LSVD; in Italy the UNHCRItaly; in the UK, the UKLGIG and Asylum Research Consultancy). We had different experiences with how public authorities responded to our FOI requests and, more generally, our research.

In Germany, we tried several times to gain access to interviewers and decision-makers, but the BAMF rejected our request on each occasion.Footnote 24 In order to gain the information we needed, we prepared comprehensive FOI requests containing 30 questions. As these questions were addressed to different public authorities, we liaised with the party ‘Die Linke’, who submitted these questions as a parliamentary request (‘Kleine Anfrage’). The government responded to the request within 4 weeks.Footnote 25 Nonetheless, the government’s response was not comprehensive and some of the information requested could not be provided due the country’s federal structure. For instance, in its response, the government stated that individual federal states are responsible for reception and that the government did not have information about how federal states respond to vulnerabilities. The government also stated that they did not have information about specific accommodation for SOGI claimants, or what happens if claimants ask to be moved from their accommodation centre. Furthermore, no statistical information about SOGI claims exists (Chap. 4).

In Italy, the process of gaining access to information was more difficult. In February 2019, the FOI request was sent to the National Commission for the Right of Asylum (‘Commissione Nazionale Asilo’) and the Minister of Internal Affairs (Department for civil liberties and immigrants). We received no confirmation that they had received the FOI request and they did not respond to the request, nor to reminders sent in May 2019. In fact, all the participants in our fieldwork in Italy, including decision-makers, were found through personal contacts after attempts to make contact at an institutional level failed. It is rare for Italian public administrative bodies to respond to such information requests from citizens, although they are formally obliged to do so ‘in a reasonable time’. We then sent a complaint to an opposition party’s member of Parliament, including a request to submit, through him, a parliamentary question (‘interrogazione parlamentare’). However, we did not receive a reply. As context, the political climate in Italy throughout 2017 and 2018 became tense, which our analysis of the asylum reforms highlights (Chap. 6).

In the UK, the process was also far from quick or transparent. Three FOI requests were submitted in January 2019: one to the Ministry of Justice and two to the Home Office.Footnote 26 The FOI request to the Ministry of Justice as well as the one to the Home Office regarding detention were answered but elicited no new information, largely on the basis that neither department held the information we were requesting. With regard to the third FOI request, which was the most detailed, covering decision-making, we only received a full answer to our request in June 2019 (FOI reference 52467), thus much later than the specified 20 working days, despite further emails and complaints on our part to both the Information Commissioner’s Office and to the Home Office.Footnote 27 Nonetheless, much of the information provided was either not relevant or did not directly answer the questions posed in our FOI request. In addition, the Home Office stated that it was withholding information regarding the decision-making process on ‘public interest’ grounds:

Regarding your request for information and the questions used to identify the basis for an asylum claim, we do hold the information, but have decided to exempt this information under section 31 of the FOIA 2000. Section 31(1)(e) allows us to exempt information if its disclosure would or would be likely to prejudice Law Enforcement – the operation of immigration controls. This exemption requires us to consider whether, in every respect the public interest in maintaining the exemption stated above, outweighs the public interest in disclosing the information. Arguments for and against disclosure in terms of the public interest are set out in the attached Annex 1. (p. 3)

In short, the way the FOI requests in Italy and the UK were handled raises important concerns regarding democratic accountability of officials and transparency in public policy, not only in relation to the substance of the responses we received, but also in terms of procedural failures that make freedom of information a right that in some cases exists only on paper.

3 Ethical Implications: Doing Research with SOGI Refugees

It might be said that fulfilling ethical standards is important for any project. However, due to the particular situation in which SOGI asylum claimants find themselves, considering ethical implications when conducting research with this group of participants was particularly important (Held 2019). Before we started the research, several ethical issues were identified and ethical approval from the University of Sussex’ Ethics Committee obtained.Footnote 28

The interviews and focus groups with asylum claimants and refugees invariably involved individuals who had gone through difficult life experiences. Depression, PTSD and anxiety are common in the refugee population (Chap. 9). Therefore, the retelling of traumatic events can have a considerable impact on research participants’ mental health. If their SOGI was the basis for the harm inflicted on them, then talking about these aspects of themselves, and their experiences in this regard, may be re-traumatising. Yet, as Stevenson and Willott (2006) point out, we may not always have the right understanding of what topics will be sensitive for a participant.Footnote 29 There may be other intersecting aspects of SOGI claimants’ identities that are more difficult to talk about and foresee. Furthermore, policy-makers, members of the judiciary, legal representatives, and NGO workers interviewed delved into intimate aspects of the lives and experiences of SOGI claimants, which also risked causing them some distress and anxiety about their and their clients’ confidentiality.

It was important to us not to cause psychological harm by asking questions in an insensitive way or probing too much about experiences that might have caused distress to the participant (Krause 2017). We tried to minimise risks to asylum claimant and refugee participants by offering to refer them to agencies and services (where available) capable of dealing professionally and in a supportive way with any mental or physical health issues. We also provided a list of support services available in the relevant countries.Footnote 30 Nonetheless, we were also aware that listening to participants’ potential disclosures of (sexual) violence and trauma might be distressing for us and impact on our well-being (Krause 2017, p. 4). Therefore, before going into the field, we received training on ‘vicarious trauma’from Freedom from Torture, to learn about the signs of vicarious trauma and how to practice self-care in the course of our fieldwork.

We obtained informed consent from all research participants, who were given comprehensive information about the project. This consisted of a clear and honest discussion with the researchers, alongside an information pack (in their preferred language where possible), which explained the project aims, purpose of the data collection, methods, data storage, information on how to withdraw consent and contact information for further enquiries.Footnote 31 This information allowed individuals to consciously decide whether they felt emotionally and mentally prepared to share and discuss their experiences. Participants were given time to reflect and ask questions on the information given. Consent forms included questions about consent to audio-record interviews, with information on how all data would be anonymised for publication.Footnote 32 With regard to the self-completed surveys, submitting a response to the survey implied consent (this was stated on the information sheet preceding the survey).Footnote 33 Participants were given the opportunity to withdraw consent, and request the destruction of any data relating to them, at any time up to the publication of an outcome, without giving any reason, and without repercussion or penalty for the participant. This was explained clearly by both the researcher and on the information sheet provided to each participant.

Many of our participants would typically be seen as ‘vulnerable’. Yet, the issue of ‘vulnerability’ is complex (Krause 2017). As further discussed in Chap. 4, asylum claimants and refugees might in some circumstances be officially defined as ‘vulnerable’, – for example ‘transsexual and intersex people’ for the purposes of immigration detention in the UK (Home Office2018), and they might also in some sense be ‘vulnerable’ because they are more likely to be exposed to human rights violations (Chap. 3). Nonetheless, not all asylum claimants and refugees are vulnerable or would like to be identified as such in the everyday sense of the word, which can have infantilising connotations; most individuals seeking protection have needed to be immensely strong and resourceful, and might not consider themselves vulnerable (Stevenson and Willott 2006). In fact, we might think of asylum systems as not addressing pre-existing vulnerabilities but actually imposing vulnerability. In addition, vulnerability has also been described as a term that is ‘overused’ (James 2020, p. 33). Taking intersectionality seriously means that when assessing vulnerability, we ‘need to consider both the individual circumstances of each research participant, rather than see them as a homogenous group’ (Stevenson and Willott 2006, p. 383). By focusing on hardship and using victimising notions, people seeking asylum are often portrayed as having similar experiences including the shared experience of vulnerability (Krause 2017), but they may find this label patronising. Instead of homogenising SOGI claimants as a ‘vulnerable group’, the intersecting characteristics of each person need to be taken into account when assessing individual vulnerability. It is also important to highlight asylum claimants’ and refugees’ agency. For instance, many are themselves involved in establishing networks, support groups, and other forms of refugee activism (Bhimji 2016). In this respect, while it might be crucial to anonymise data to avoid risks, it is also important to grant participants the autonomy to decide for themselves whether they want their accounts to be anonymised or not (Clark-Kazak 2017; Krause 2017). In our research, about a quarter of the SOGI claimants and refugee participants chose to be named, while the other three-quarters chose to be anonymised (some provided a pseudonym, for others we chose one), or did not mind either way. To be consistent, we refer to all participants only by first name, including professionals. Occasionally, when we refer to very sensitive matters, we use neither a real name nor a pseudonym.

It has been highlighted that the uniqueness of the refugee experience brings high levels of distrust with it – because of mistreatment in countries of origin, but also mistreatment during the asylum process (Hynes 2003, p. 13). SOGI claimants in particular might have had experiences of not being able to trust people and ‘be out’ with regard to their SOGI. In the host country they might have had negative experiences with government officials, housing and other social service providers, and the discrimination they experience adds to their levels of mistrust. This will be compounded by the power differentials between researchers and researched and the fact that researchers often treat research participants solely as ‘data source’ and not as active subjects with rights, dignity and agency (Krause 2017).Footnote 34 The asylum system disempowers SOGI claimants, and often they have internalised disempowering messages. Traditional approaches to research can make asylum claimants and refugees feel exploited (Cochrane 2015).

The three researchers who conducted interviews with SOGI claimants are all cis-gendered (one male, two female) – one gay, one lesbian and one heterosexual – and White academics with considerable social, economic and cultural capital. Consequently, power inequalities between us and our participants were likely to exist on grounds of ‘refugeeness’, gender, ‘race’, socio-economic status and political rights. For instance, one of the participants in a focus group in Italy complained about the fact that we were White researchers, stating that ‘we cannot understand’ in full depth their stories and feelings, and that researchers are not necessarily aware of the pain they go through (focus group no. 1, northern Italy). In the interactive process of data collection, power differentials (real or assumed) need to be addressed, while acknowledging that power is relative and exists in all relationships, in other words, it can shift and change. It was thus important to not only consider the intersecting identities of participants, but also those of the researchers. Our gender, sexuality, ‘race’ and nationality will necessarily have had an impact on how open participants felt to talk about their experiences. For instance, one NGO volunteer we interviewed (Thomas, Germany) felt that LGBTIQ+ asylum claimants and refugees might not tell him about their experiences with discrimination ‘out of politeness and hospitality’.

While we were aware that power differentials could not be eliminated, we aimed to conduct research that brought reciprocal benefits, and established ‘ethical relationships between researchers and participants that are responsive to the needs, concerns and values of participants’ (Mackenzie et al. 2007, p. 307). Our aim was to do research with refugees instead of for or on refugees (Hynes 2003, p. 14) and to work collaboratively as much as possible. Even in so doing, however, power imbalances were unavoidable: three years into the research and having developed ongoing relationships with a number of our participants, we sometimes found ourselves in situations where journalists or event organisers contacted us asking if we knew of any individuals claiming asylum on SOGI grounds who might wish to speak at a conference or appear in a programme. We had to decide which of our many participants to contact about such opportunities, which might not only be enjoyable experiences for the individual in question, but also constitute ‘evidence’ to include in an appeal bundle. While we always stressed the advantages and disadvantages of such public engagement opportunities to individuals, the underlying power differential remained in place.

In order to achieve our collaborative objectives so far as possible, we established an Advisory Board consisting of five members with expertise in socio-legal research and SOGI asylum claims, including two beneficiaries of international protection, who oversaw the methodological and ethical soundness of the research, as well as offered advice on any aspect of the activities carried out to achieve our aims and objectives. In the spirit of knowledge exchange, we were keen to collaborate with research participants and stakeholders throughout the project, and tried to make this a genuinely two-way process so that we could also give something in return. All researchers were engaged with local NGOs and gave presentations and workshops for LGBTIQ+, refugee and migrant organisations and offered various forms of support to individual and NGO participants, in light of occasional requests. We provided letters for participants confirming that they had participated in our study, which they could use as part of their claims and appeals. We kept in contact with many participants throughout the project and offered emotional and practical support. We invited participants to attend and speak at conferences and events we organised, and connected them with other people and groups, which often helped in breaking down isolation. We sent out a regular newsletter to participants (and increasingly to other people who subscribed to it – a total of more than 1,200 individuals by January 2020) to provide updates on the project.Footnote 35 We also created a database of resources for use by SOGI claimants, practitioners and researchers alike.Footnote 36

While there is a risk that interviews can re-traumatise participants, the potential therapeutic effect of telling their story in a safe environment has also been pointed out (Harrell-Bond and Voutira 2007). In particular, the focus groups, which provided a space for open discussions, seemed to have had cathartic effects on participants (Giulio, referring to focus groups no. 1 and 2, northern Italy). As Rosette, who for the first time visited LeTRa when we had the focus group (no. 3, Bavaria, Germany), described her feelings: ‘Yeah. I can talk and express myself, and at least listen to people, how other people really feel’. Inspired by the same focus group, Ayeta said ‘I feel that we should be always… we have to select some few days that we can be together and talk about issues and everybody’s issues’. People were excited about and committed to participating in the study. As Kennedy (Italy) explained: ‘and that is the reason why I am so happy, you know, in doing these interviews, for us to pass a message across to the governments, not just Nigeria, to Africa, governments, to Africa leaders’.

We tried to avoid raising any expectations on the part of participants that participating in this research would enhance outstanding asylum applications, but very often participants were keen to take part in order to make their voices heard, and above all to help improve the lives and experiences of SOGI refugees in the future (see also Venturi 2017). For instance, Winifred, who participated in focus group no. 4 in Bavaria, Germany, and who is a researcher herself, said:

We really appreciate taking part in this interview. We really appreciate it. It is a very big opportunity, even though we don’t know what will come out. Or we know it will help us in the future, but we really appreciate it. Maybe by then, those people who will be in our shoes now, they will have a better future, they will be given different opportunities. So we really appreciate being part of this interview.

In the following chapters, we hope we succeed in making the voices of our participants heard.