Research design

In this research, I employed an ethnographic research design. Ethnography is based on an understanding of knowledge as situated. Through ethnographic research, I aimed to come to a

‘thick’ description of my topic within its context: not only demonstrating people’s practices but the different meanings around them (Gibson-Graham, 2014). To grasp these meanings I combined various qualitative research methods. Qualitative methods allow insights into people’s behaviors, experiences, and opinions, as well as the meanings they attach to these (Silverman, 2020).

4.3.1 Data collection methods

To come to a thick description of the asylum preparation process, I used data from four sources: interviews with NGO workers, participant observation, IND documents, and court reports.

Due to ethical considerations, that I outline in section 4.4.1, I chose not to interview undocumented migrants but researched NGO practices from the perspective of NGO workers. Most information was derived from fourteen semi-structured interviews with NGO workers, conducted between October 2021 and January 2022. In these interviews, I explored the preparation practices that NGO workers engage in, and how they see the IND’s demands and their own role. In interviews with NGO coordinators, I asked about the organizational structure and how work processes are created. I met another student doing her thesis research on a very similar topic. We chose to conduct interviews together to avoid over-asking participants. In practice, we had each prepared our own interview questions and generally switched the interviewer role halfway. The NGO workers’ answers to her questions did not always directly address my research but provided valuable context. The interviews generally lasted an hour and were conducted in Dutch. With the participants’

permission, we audio-recorded the interviews. We used two interview guides: one for employees directly doing the preparation process, and one for coordinators within NGOs. The interview guides we used can be found in appendices B and C.

Next to these interviews, I conducted five months of participant observation at the

organization Solidarity with Undocumented Migrants (SUM, a pseudonym). Participant observation entails gathering data and observing while participating in a research population’s daily activities, interactions, and events (DeWalt & DeWalt, 2011). As an intern within the social-legal team, I participated in meetings, joined events, and discussed cases and opinions with colleagues. For this

period, I worked in the NGO’s office two days a week, during which I wrote fieldnotes regularly. I wrote my fieldnotes chronologically, but later distinguished entries that demonstrated the general organizational context and those related to LGBTI(Q+) cases. Participant observation provided valuable background information to decide what to research, what questions to ask during interviews, and how to understand the answers given. During my internship, I ‘prepared’ an LGBTI(Q+) client for an asylum claim myself. I did not use information from this case, but the instructions I received from other NGO workers provided interesting insights into how they approach the preparation process.

I supplemented these primary data with secondary data: publicly available working instructions of the IND and publicized court cases of LGBTI(Q+) asylum seekers. The working instructions instruct IND hearing- and decision-officials on how to conduct the asylum interview (IND, 2021a), how to judge the credibility of an asylum seeker’s story (IND, 2015a), and how to judge cases in which the reason to request asylum is someone’s sexual orientation (IND, 2019).

These instructions provided me with valuable data about the asylum process for LGBTI(Q+) asylum seekers and the IND’s demands for a credible story. Additionally, I selected recent court cases in which LGBTI(Q+) asylum seekers had been rejected and appealed the decision. These court cases are published anonymously by Dutch courts on Rechtspraak.nl. They demonstrated the application of the IND’s demands in practice. I had a selection of sixteen court cases. I drew on three cases explicitly in my empirical chapters to demonstrate how certain IND demands are applied in practice. I used the court and IND documents to triangulate my findings from the participant observation and interviews. They helped to understand the legal-institutional context in which the preparation process takes place, as well as the narrative demands LGBTI(Q+) asylum seekers must adhere to.

4.3.2 Sampling strategies

I employed various sampling strategies. The participant observation data were collected through my internship at SUM. This was a convenience sample (see Taherdoost, 2016), as I had already secured an internship here before starting my research. Most interviewees were also found through convenience sampling, as I mostly interviewed SUM employees. A number of participants from other organizations were found through snowball sampling (see ibid.): SUM employees recommended people from other organizations that I could speak to, or put me in contact with them. The high workload that is common in these organizations makes that NGO workers often have little time to speak to researchers. Snowball sampling allowed SUM employees to ‘vouch for’

my research, which helped me access more participants. I further employed purposive sampling

(see ibid.) by reaching out to other organizations by email. While I secured some interviews this way, I noticed that without ‘vouching’, people were more hesitant or simply did not respond. The difficulty accessing the research population systematically, combined with the ethnographic nature of the study justifies my non-random sampling approach (see Higginbottom, 2004).

To select applicable IND documents and to find court cases that demonstrate narrative demands, I employed purposive sampling (see Taherdoost, 2016). I included the IND working instructions that addressed LGBTI(Q+) cases, credibility, or asylum hearings. For the court cases, I searched for cases on Rechtspraak.nl that contained the words ‘LHBT’, ‘geloofwaardig’ (credible), and ‘asiel’ (asylum). I collected these documents in October 2021. In a preliminary selection, I included all cases that had been decided on after 2019 as these decisions were based on the most recent IND working instruction. I limited my search to cases decided upon in the court in The Hague, as this is a large court that regularly publishes cases. I selected the cases in which the credibility of applicants’ sexual orientation was the reason their claims were rejected. This provided me with sixteen court cases, which gave background information. I used three cases explicitly in my empirical chapters as these provided the clearest demonstration of a certain IND demand.

4.3.3 Sample

Below is an overview of my interview participants. The participants worked for five different organizations in three Dutch cities. Most people worked for SUM, where I also conducted

participant observation. A full list of the IND and court documents I used can be found in Appendix D.

Table 1: Overview and characteristics of participants Name

(pseudonym) Organization

(pseudonyms) Function Age Sex

1 Anna Solidarity with

Undocumented Migrants (SUM)

Social-legal employee +/- 30 F

2 Lucia SUM Social-legal employee +/- 25 F

3 Sophie SUM Social-legal employee +/- 30 F

4 Nicole Migrant Support

Network

Netherlands (MSN)

Social-legal employee - team supporting formerly

unaccompanied underaged asylum seekers

+/- 45 F

5 Isabel SUM Senior social-legal employee +/- 55 F

6 Eva MigrantNetwork Social-legal employee +/- 45 F

7 Christine MigrantNetwork Social-legal employee +/- 50 F

8 Annemarie MSN Legal helpdesk,

jurisprudence +/- 30 F

9 Thomas Undocumented

Migrants Basic Needs Support (UMBNS)

Social support coordinator +/- 35 M

10 Laura SUM Social-legal supervisor +/- 30 F

11 Paulien MigrantNetwork Social-legal supervisor +/- 35 F

12 Judith SUM Social-legal employee +/- 60 F

13 Pieter Migrant and

Refugee

Organization (MRO)

General coordinator +/- 40 M

14 Sam SUM Social-legal volunteer +/- 25 NB

4.3.4 Data analysis

I used various approaches to analyze my data. For my interview data, I employed thematic analysis to understand the main themes that emerged. I transcribed the interviews and

subsequently used Atlas.ti to code my data. In line with Strauss and Corbin (1990), I started with

open coding to break down my data and categorize it. I then proceeded with axial coding, where I grouped these smaller sub-categories into larger groups (ibid.). Lastly, I engaged in selective coding, in which I selected the central themes around which my other data categories were focused (ibid.). This coding process allowed me to identify the main themes in my interview data, from which I could discern a more theoretical understanding of my findings and review them in light of existing literature (Bryman, 2012). I combined an inductive and a deductive approach, as I based my conclusions on the main themes found in the data while having already some ideas about possible themes derived from the literature (ibid.).

Furthermore, I thoroughly read through my observation fieldnotes and selected entries with information that addressed my topic. I organized my fieldnotes based on the subject each entry was addressing, distinguishing between general notes about SUM as an organization and entries about LGBTI(Q+) cases. I discerned which of the main themes from the interview data came back in my fieldnotes, and I used the observation data as context for understanding the interview data. Similarly, I searched for themes emerging out of my interview data in the court cases I used. In this way, the themes I found in my interview data provided a framework to understand my

observation data and court cases. Lastly, I selected the information from the IND working instructions that demonstrated how they assess applicants’ credibility and important themes in LGBTI(Q+) cases. I compared and supplemented the information from these working instructions with my findings from the interviews and court cases to understand where and how the different sources of data might differ.