Introduction
Consensual same‑sex relations among adults remain criminalised in nearly 67 countries worldwide, with nine countries extending such laws to target transgender and gender‑nonconforming individuals (Human Rights Watch 2025). This legal landscape underscores the pervasive threat of persecution, violence and human rights violations faced by individuals based on their sexual orientation and/or gender identity and expression (SOGIE). Often referred to as SOGIE refugees, these individuals seek protection in countries that increasingly recognise SOGIE as grounds for asylum under the Refugee Convention. Yet, the reasons for individuals to migrate are complex and not solely determined by mere legal frameworks. Migration research has established that social networks are important institutions in the decision‑making of migration destination choice as well as in creating a new home in destination countries (Böcker 1994; Boyd 1989; Havinga & Böcker 1999). Research into SOGIE refugees has shown that social networks – be they based on nuclear families or Lesbian, Gay, Bisexual, Trans, Queer and other (LGBTQ+) organisations or communities – convey information of opportunities in countries of origin, facilitate the migration process and help ease the initial period in destination countries (Giametta 2020; Pullen & Franklin 2024; Wimark 2016b, 2021a).
This article seeks to expand this literature of the role of organisations in destination countries by exploring how LGBTQ+ organisations are related to spatial sorting behaviour – the process wherein individuals cluster together in locations and communities at different scales (Clark 2008) – of sexual orientation refugees. A growing body of research addresses the experiences of SOGIE refugees in the migration process (e.g. see Chossière 2021; Danisi et al. 2021; Held 2023; Hopkinson et al. 2017; Schröder 2023). However, a lack of demographic data and official statistics from migration authorities continues to pose challenges to understanding SOGIE asylum‑seekers beyond experiences (Danisi et al. 2021; García Rodríguez 2023; Shaw et al. 2022; Shaw & Verghese 2022).1 This article aims to bridge this gap using a representative sample of SOGIE asylum‑seekers in Sweden. Drawing on this sample, it investigates differences in sorting behaviour in migration both to Sweden and into place‑based LGBTQ+ organisations among asylum claimants.
LGBTQ+ organisations are central for asylum claimants, researchers and migration authorities alike, yet their involvement also presents certain challenges. While collaborating as researchers with LGBTQ+ organisations facilitates access to SOGIE refugees and ensures ethical research practices (e.g. see Chossière 2021; Dhoest 2019; Lunau & Andreassen 2023; Murray 2014a; Tschalaer 2020, 2021), it also introduces potential biases, such as geographical disparities in access and self‑selection biases among participants (Wimark 2021a, 2021b). Beyond research, these organisations often serve as vital sources of community and support for SOGIE asylum claimants, particularly in the absence of specific services from state authorities such as the Swedish Migration Agency (SMA). Despite their importance, LGBTQ+ organisations are typically concentrated in urban centres, while asylum‑seekers are frequently placed in smaller towns, limiting their ability to engage with these networks (Uteng 2006; Wimark 2021b). Moreover, some SOGIE claimants may avoid them due to fears that association with them might jeopardize their connection to other social networks (Allouche 2017; Held 2023; Wimark 2021a, 2021b).
LGBTQ+ organisations also play an essential role in policy formation and public debate (e.g. see (United Nations High Commissioner for Refugees [UNHCR] roundtable 2021). In recent years, SOGIE asylum‑seekers have become increasingly prominent in the public discourse, used simultaneously to justify humanitarian commitments and raise suspicions. As more countries in the Global North recognise sexual orientation as a basis for asylum (García Rodríguez 2023; Millbank 2021), narratives of disbelief have gained traction. In some contexts, political and media debates portray SOGIE asylum claims as fraudulent (Euronews 2023; Ferreira 2022); in others, SOGIE refugees are depicted either as suffering subjects or particularly deserving individuals (Koçak 2020; Saleh 2020). For instance, in 2023, Sweden’s right‑wing government revised its resettlement priorities to favour ‘women, girls and vulnerable refugees, among them LGBTQI individuals’, while simultaneously reducing the total annual resettlement quota from approximately 5,000 to 900 (Migrationsverket 2023).
In the next section, I account for the legal background and present theories of how SOGIE refugees sort into countries and place‑based group settings. I then explain the specifics of the data of this study: a representative sample of asylum decisions. The results are then presented, illustrating the characteristics of sexual orientation asylum‑seekers sorting into Sweden and subsequently those sorting into LGBTQ+ organisations, which are formed around place‑based activities. In the final section, I discuss some potential implications this work has for the support of SOGIE refugees in general and the knowledge production in particular.
Asylum and admission in sweden
Before delving into sorting processes of SOGIE asylum‑seekers, it is essential to understand the legal framework in Sweden that enables them to attain refugee status. Protection against refoulement – a fundamental human right – is enshrined in the Swedish Aliens Act, aligning with both the Geneva Refugee Convention (UN Nations 1951, 1967)2 and the EU Qualification Directive (Directive 2011/95/EU)3. In Chapter 4, Section 1, of the Aliens Act (SFS 2005), a refugee is defined per the Refugee Convention. Notably, Swedish law extends beyond this by explicitly recognising persecution based on gender and sexual orientation as grounds for refugee status. Thus, asylum may be awarded if individuals have a ‘well‑founded fear of persecution’ due to factors such as ‘race, nationality, religious or political opinion or because of gender, sexual orientation or other belonging to particular social group’ (SFS 2005).
SOGIE Refugees and Sorting to Certain Countries
There are many theoretical ways to frame the reasons for SOGIE refugees’ sorting patterns – i.e. the process in which individuals cluster together in certain countries, places and place‑based group settings. Many scholars have used historical–structuralist approaches to explain why SOGIE migrants primarily seek safety in the Global North. These approaches emphasise historical and structural factors such as colonial legacies, warfare and international inequalities (cf. Massey et al. 1999). Previous studies on SOGIE refugees often highlight colonialism as a structural source, attributing the criminalisation of same‑sex activities in former colonies to this historical legacy (e.g. see Lee 2018; Mulé 2020). Additionally, scholars draw on Puar’s (2007) concept of ‘homonationalism’, which describes how certain Western states portray themselves as progressive by accepting LGBT individuals while depicting others as backwards. Such narratives often demand adherence to Western, white‑centric conceptions of LGBT identities (Akin 2017; Giametta 2017; Murray 2014a).
Yet, historical–structuralist frameworks often overlook individual agency. Despite facing persecution, SOGIE individuals make deliberate decisions regarding their destination countries and places of settlement. Moreover, while economic considerations may not be the primary motivator for fleeing persecution, they can influence migration decisions. This is not to say that SOGIE individuals are purely rational actors attempting to maximise their incomes, commonly explored in economic sorting models (c.f. Borjas 1989). Rather, it highlights that individuals choose to move to certain countries and, upon arrival, enter place‑based group settings shaped by their life courses, in which their socioeconomic situation is one component (Elder et al. 2003). The decision to move and cluster together is based on the interplay between life course trajectories – developed over time in a complex process of interrelations in different environments, neighbourhoods, locales and spaces – in the origin countries and the possibilities available in destination countries (Wingens et al. 2011). This is not to say that the life course approach can replace other approaches, as it often downplays geo‑political relations of power in which uneven differences are reproduced (Garelli & Tazzioli 2021; Murray 2014b).
Moreover, the life course approach to the decision‑making process has received criticism for being modelled on a heterosexual life course (Elder et al. 2003), and the main reason to flee for SOGIE individuals may still be grounded in persecution based on sexual orientation and/or gender identity/expression. In queer migration studies, the traditional heteronormative life course approach has been expanded to centre sexual exploration and development alongside economic factors (e.g. see Di Feliciantonio & Gadelha 2016; Lewis 2014; Wimark 2016a). These studies underscore the role of queer communities as magnets in destination cities, which provide spaces for sexual freedom and community‑building while simultaneously acknowledging power relations underpinning the heteronormative production of spaces (Weston 1995). Similarly, certain countries are perceived as havens for sexual freedom, attracting SOGIE individuals seeking refuge (Dhoest 2019; Di Feliciantonio & Gadelha 2016; Murray 2014a).
Queer migration studies also highlight the role of migrant networks, where kinship and friendship ties sustain migration and facilitate the exchange of information and experiences between origin and destination areas (c.f. Weston 1995). Migrant networks have similarly been recognised as playing an essential part between origin and destination countries (Böcker 1994; Massey et al. 1999). While SOGIE refugees may have been ostracised from their families, studies suggest that family connections remain influential beyond migration (Wimark 2016b, 2021a). Additionally, the proliferation of social media enables the dissemination of information and experiences, potentially reducing migration costs and facilitating migration decisions (Levitt 1998).
Social network initiatives also highlight that entities other than solely family and friends form part of a pool of possibilities for potential migrants in countries of origin. The ‘migration industry’ encompasses various actors, including non‑governmental organisations (NGOs), employers and smugglers, who either facilitate or exploit migration processes (Boswell & Geddes 2010). While the impact of the migration industry on SOGIE refugees remains understudied, LGBTQ+ NGOs in destination countries engage in activism and outreach efforts, shaping perceptions of migration possibilities and providing support networks for potential migrants in their origins.
By broadening the theoretical understanding beyond historical–structuralist approaches to include individual‑centred explanations, insights can be gained into the diverse motivations of the sorting of SOGIE refugees. This approach recognises that sexual orientation and gender identity/expression may not always be the primary reasons for migration nor for group formation; individuals may migrate for study or work opportunities, only to seek international protection later. It also highlights the importance of social networks and organisations, both for the facilitation of the migration process and for continuous group sorting in the new country of residence. Conversely, not all asylum‑seekers with SOGIE claims necessarily identify as LGBTQ+ individuals. The methodological approach in this study is described within this theoretical framework.
Methods
The data utilised in this study are derived from a representative sample of all adults who received a first‑instance asylum decision in 2020 from the SMA. First‑instance asylum decisions relate to applicants applying for asylum in Sweden or at the border, excluding quota refugees. In 2020, the SMA made decisions on 15,478 adult asylum applicants, of which 1,263 were Dublin regulation cases and 1,591 were retracted or deregistered.4 Therefore, the sample was drawn from a total individual population of 12,624 adult asylum‑seekers who received a first‑instance asylum decision. The grant rate for this population was 24.8% after excluding Dublin and retracted/written‑off cases. Importantly, the majority of applicants arrived before the coronavirus disease 2019 outbreak. Also, the average processing time for first‑instance cases by SMA was over 300 days in 2020 (Migrationsverket 2021), meaning that the overwhelming majority arrived before the outbreak. Pandemic‑related issues should not be a factor in this study. Ethical approval from the Swedish Ethical Review Authority was obtained for the project (Dnr 2021‑06258‑01 and 2022‑04063‑02).
The sample includes 3,588 randomised individuals’ decisions, representing 28.4% of the total adult population of asylum‑seekers who received a first‑instance decision in 2020. All asylum decisions were provided in PDF format by the SMA. The decisions include distinct information about the applicant (e.g. gender, birth date, country of origin [CoO]), decision outcomes, application specifics, and asylum claims stated by the applicants. Asylum claims often require interpretation, as they may not always be explicitly stated concerning legal grounds. In this study, asylum decisions were coded using categories defined in the law, including ‘race, nationality, religious or political opinion or because of gender, sexual orientation or other belonging to a particular social group’. Due to the difficulty in differentiating between race and nationality, these categories were merged. Meanwhile, the category of ‘gender’ was further subdivided into gender and gender identity/expression. A detailed description of the coding process can be found in Wimark and Almberg (2024), which involved oscillating between data and legal definitions and delimitations as outlined in preparatory works to the Aliens Act and the UNHCR Handbook on Procedures and Criteria for Determining Refugee Status (UNHCR 2019).
A research assistant conducted the coding. To ensure high inter‑rater reliability, the author coded a random sample containing approximately 5% of all decisions (180 decisions) and an additional random sample containing 20% of the identified sexual orientation asylum decisions (51 decisions). The coding results were compared using standard inter‑rater reliability measures, demonstrating high agreement between coders (lowest score for Cohen’s kappa = 0.91). Values over 0.81 for Cohen’s kappa are generally considered to indicate almost‑perfect agreement (see, Altman 1991; Landis & Koch 1977).
In the Results section, two distinct subpopulations are delineated to attempt to capture the differences between those who are a part of social networks created by LGBTQ+ organisations in Sweden and those who are not: first, those who have claimed asylum based on their sexual orientation (SO claims), and second, those who have been affiliated with or engaged in LGBTQ+ organisations in Sweden and have claimed asylum based on their sexual orientation (SO LGBTQ). Standard nonparametric tests were used for nominal data, and t tests were applied to interval data to assess differences between groups. The analysis first compared all non‑SO asylum claimants with those making SO claims (non‑SO claims, n = 3.331; SO claims, n = 257), followed by a comparison between SO and SO LGBTQ+ claimants (non‑SO LGBTQ, n = 198; SO LGBTQ, n = 59).
There are several limitations in this study. The coding process involved analysing asylum decisions authored by case officers; this reflects their interpretations rather than the applicants’ views or objective reality. Notably, ‘gender identity/expression’ is among the categories subject to interpretation. While the entire asylum decision and the sections where case officers summarised claims to identify any potentially missed claims were screened, it is still possible that relevant texts were overlooked or omitted from the assessment section by case officers. Case officers may also choose not to include certain information in the asylum decisions that might affect the measurements of the two subpopulations. For example, some case officers might have been reluctant to report engagement in LGBTQ+ organisations for different reasons.
Manual coding of data inherently carries drawbacks, and the interpretation of certain variables poses challenges. For instance, the conceptual validity of variables such as ‘Days in Sweden’ may not accurately capture the actual duration of time spent in Sweden. While time is a crucial factor in asylum applications, as it has been argued to affect the decision outcomes (Millbank 2009), variations among applicants necessitated a simplified understanding that focused on the last entry in relation to the current application. Consequently, the time spent in Sweden may not be fully represented.
Results
This section explores the spatial sorting of sexual orientation asylum claimants compared to other asylum‑seekers. Moreover, the discrepancies between the two subsets within the SOGIE population, those making SO claims and individuals affiliated with LGBTQ+ organisations in Sweden who also make SO claims, are analysed. I start by illustrating the prevalence of SOGIE claimants relative to the entire sample of asylum‑seekers to set the context. Subsequently, a comprehensive overview of the individual demographics and characteristics is provided for SO claimants and SO individuals engaged in LGBTQ+ organisations in Sweden. Finally, the specific case nature of the asylum cases within each subgroup is examined.
Prevalence of sexual orientation and gender identity/expression claimants
A common assumption is that SOGIE claimants constitute a marginal population within asylum‑seekers. However, this analysis challenges this notion. Table 1 illustrates the prevalence of SOGIE claimants compared to other asylum claimant groups. The Refugee Status Claim categories in the first column in Table 1 indicate if an applicant has made at least one claim. An applicant might have several claims, so the sum of each row will not show 100%. The data in the second column marked ‘All’ represent all the applicants in the sample. Since it is a sample, a 95% confidence interval (CI) is also reported, indicating within which interval the ‘true’ share for all applicants in the year 2020 was located. The fifth column marked ‘Convention Refugees’ further presents data only for applicants with refugee status claims referring to the Refugee Convention, i.e. occluding those seeking subsidiary protection. Two conclusions can be drawn from these data.
Table 1
Refugee status claims per claim category, first‑instance decisions, 2020. Each row presents the share of claimants who have stated at least the claim category in question.
| REFUGEE STATUS CLAIM CATEGORY | ALL | CONFIDENCE INTERVAL (95%) | CONVENTION REFUGEES | CONFIDENCE INTERVAL (95%) | ||
|---|---|---|---|---|---|---|
| Lower | Upper | Lower | Upper | |||
| Political opinion | 45.4% | 43.8% | 47.1% | 58.7% | 56.9% | 60.5% |
| Religious belief | 21.2% | 19.9% | 22.6% | 27.4% | 25.8% | 29.1% |
| Gender | 19.4% | 18.1% | 20.7% | 25.0% | 23.5% | 26.7% |
| Race/ethnicity/nationality | 10.5% | 9.5% | 11.6% | 13.6% | 12.4% | 14.9% |
| Sexual orientation | 7.2% | 6.3% | 8.1% | 9.3% | 8.2% | 10.4% |
| Particular social group | 4.9% | 4.2% | 5.6% | 6.3% | 5.5% | 7.3% |
| Gender identity/expression | 0.4% | 0.2% | 0.7% | 0.5% | 0.3% | 0.9% |
| Several refugee status claims | 25.8% | 24.3% | 27.2% | 33.3% | 31.6% | 35.1% |
| No refugee status claims | 22.7% | 21.3% | 24.1% | N.A. | N.A. | N.A. |
| N | 3588 | 2771 | ||||
[i] Source: Swedish Migration Agency; calculations were performed by Angelica Dovenryd Almberg and Thomas Wimark.
First, the findings show that 7.2% (95% CI 6.3–8.1%) of asylum applicants in the study claimed sexual orientation as at least one of their grounds for seeking asylum. Considering only Convention Refugees, this figure is 9.3% (95% CI 8.2–10.4%). Contrary to common belief, this proportion is reasonably close to the number of asylum applicants citing race/ethnicity/nationality as a basis for their claims (10.5% and 13.6%). Similar trends have been observed internationally (e.g. see Bird et al. 2022; Rehaag 2023). Second, the analysis reveals that the number of asylum applicants citing gender identity and/or gender expression as grounds for seeking asylum is exceedingly small; less than 1% of applicants stated gender identity and/or gender expression (0.4%, 95% CI 0.2–0.8%) and, among Convention Refugees, only 0.5% did (95% CI 0,3–0,9%). Reliable international comparisons for this group are scarce, although a Canadian study indicates a similarly low proportion (Rehaag 2023). This finding aligns with the consensus among scholars that the number of gender identity/expression minorities among asylum claimants is indeed small (Danisi et al. 2021).
Given the minimal representation of gender identity and/or gender expression claimants in the sample, presenting further detailed data on individual and case characteristics for this group will not yield meaningful insights. So, the subsequent analysis will focus on the significant presence of sexual orientation claimants.
Sexual orientation claimants’ sorting characteristics and origins
As discussed in the Introduction, empirical data regarding the individual sorting characteristics of SOGIE refugees are lacking in international research. Table 2 provides an overview of the individual characteristics of sexual orientation asylum claimants, comparing them to other asylum‑seekers. The table has three columns reporting data, each representing shares of the populations (except for the age variable representing average age). The first column, marked with ‘All’, represents all the claimants in the sample. The second column, marked with ‘SO claim’, represents applicants who have reported sexual orientation as at least one of their asylum claims. The third column, marked with ‘SO LGBTQ’, denotes all SO applicants who have been engaged in LGBTQ+ organisations. For example, reading the row ‘Male’ should be interpreted in the following manner: ‘among all asylum‑seekers, 60.7% were registered as male in their identification document’ (column ‘All’); ‘among those that have reported sexual orientation as at least one of their asylum claims, 71.2% were male (column ‘SO claim’); and, ‘among sexual orientation applicants who have been engaged in LGBTQ+ organisations, 71.7% were male’ (column ‘SO LGBTQ’). Since it is a sample, a significance test was also conducted. If an asterisk accompanies the result, it also indicates how sure one can be that the finding is also true for all applicants in 2020 (alpha level values; * = p < 0.05, ** = p < 0.01, *** = p < 0.001). From the table, three conclusions can be drawn.
Table 2
Individual characteristics of asylum applicants in first‑instance decisions, 2020.
| VARIABLE | ALL | SO CLAIM | SO LGBTQ |
|---|---|---|---|
| Male sex (passport) | 60.7% | 71.2%*** | 71.7% |
| Age (mean in years) | 35.6 | 32.3*** | 30.9 |
| Family applicant | 15.1% | 1.2%*** | 0.0% |
| Child in application | 16.6% | 2.3%*** | 3.3% |
| Country of Origin | |||
| Rest of Africa | 13.2% | 44.7%*** | 58.3%** |
| Former USSR | 12.2% | 10.9% | 8.3% |
| Syria | 9.6% | 1.9%*** | 0.0% |
| Uzbekistan | 8.9% | 1.6%*** | 1.7% |
| Rest of MENA | 7.4% | 10.5% | 11.7% |
| Rest of Asia | 7.3% | 4.3% | 1.7% |
| Iraq | 6.7% | 7.0% | 8.3% |
| Latin America | 6.5% | 10.1%* | 0.0%** |
| Iran | 6.1% | 3.1%* | 5.5% |
| Afghanistan | 4.5% | 2.3% | 0.0% |
| Turkey | 4.1% | 0.8%** | 5.0%** |
| Palestine | 3.4% | 0.8%* | 0.0% |
| Eritrea | 3.2% | 0.0%** | 0.0% |
| Western Balkans | 2.7% | 1.2% | 0.0% |
| Other/Stateless | 4.0% | 0.8%** | 0.0% |
| CoO Same‑sex activity criminalized, HRW | 54.5% | 64.2%*** | 71.7% |
| N | 3588 | 257 | 59 |
[i] Note: The table only shows countries with more than 110 claimants. Others have been grouped together into regions. The variable ‘Family applicant’ refers to whether the applicant sought asylum together with another adult family individual. The variable ‘CoO Same‑sex activity criminalized, HRW’, refers to whether the applicant originated from a country where same‑sex activities were criminalized, as reported by Human Rights Watch in 2023. Source: Swedish Migration Agency; calculations were performed by Angelica Dovenryd Almberg and Thomas Wimark.
First, comparing the first two columns of data reveals that individuals with sexual orientation claims differ from applicants in general. Sexual orientation claimants are predominantly male and younger. They also apply for asylum alone and originate from different countries than general asylum applicants. This aligns with findings from previous studies (Bird et al. 2022; Danisi et al. 2021; Hopkinson et al. 2017; Shaw et al. 2021).
Second, there is a notable difference between SO claimants and those SO claimants who are engaged in LGBTQ+ organisations in Sweden by comparing the second and third columns of data in the CoO. This discrepancy is particularly pronounced for the share of claimants originating from the Rest of Africa and Latin America. While nearly 45% of SO claimants originate from the Rest of Africa, close to 60% of those engaged in LGBTQ+ organisations originate from the Rest of Africa. Conversely, none of the SO claimants engaged in LGBTQ+ organisations originate from Latin America, despite around 10% of SO claimants being from this region. This uneven sorting could underscore a potential selective engagement of LGBTQ+ organisations or indicate that SO claimants from some regions lack information about or avoid LGBTQ+ organisations. This suggests that LGBTQ+ organisations in Sweden predominantly interact with specific subsets of SO claimants rather than representing the entire population.
Third, there are discernible selective sorting effects. Certain countries are more frequently selected as destinations among SO claimants compared to all applicants, comparing columns one and two of the data. Comparing the Swedish data with data from other countries in Table 3 further highlights this, although making interpretations only using the numbers in the table should be made with caution. Table 3 attempts to compare data by presenting the five most common countries of origin for SO claimants using data from Sweden, Canada, the United States and the United Kingdom. Uganda and Nigeria emerge as the top two countries of origin in the Swedish dataset, both known for outlawing same‑sex relations between consenting adults (Human Rights Watch 2025). Similarly, these countries feature prominently in Canada and the United Kingdom but not in the United States.
Table 3
Country of origin among sexual orientation claimants in Canada, the United States, the United Kingdom and Sweden.
| Rank | CoO | SWE | CoO | CANADA | CoO | UK | CoO | USA |
|---|---|---|---|---|---|---|---|---|
| 1 | Uganda | 13.2% | Nigeria | 26.8% | Pakistan | 25.1% | El Salvador | 28.0% |
| 2 | Nigeria | 8.2% | Uganda | 5.3% | Bangladesh | 10.5% | Honduras | 14.8% |
| 3 | Iraq | 7.0% | Jamaica | 4.5% | Nigeria | 9.4% | Mexico | 12.1% |
| 4 | Ethiopia | 6.6% | Cameroon | 3.9% | Uganda | 7.0% | Guatemala | 8.4% |
| 5 | Ukraine | 4.3% | Ghana | 3.3% | Iran | 5.9% | Ghana | 7.8% |
[i] Source: Canada: refugee claims made 2013–2021 (Rehaag 2023), USA: Fear claims that led to interviews by asylum officers 2007–2017 (Shaw et al. 2021), UK: Initial decisions where sexual orientation was part of the claim 2015–2021 (Home Office 2022), Sweden: Initial decisions where sexual orientation was part of the claim 2020. Calculations were performed by Thomas Wimark.
As explored in the theoretical framework, studies on SOGIE often revolve around the notion that historical–structural processes drive migration or certain countries are favoured due to their LGBTQ+ policies. However, the data suggest that the reasons behind choosing a particular destination are more nuanced. Previous research on self‑selection among refugees offers valuable insights into potential explanations. Various factors contribute to the attractiveness of certain countries over others (Havinga & Böcker 1999). Research on refugee self‑selection highlights how elements such as colonial, linguistic and cultural ties (e.g. United Kingdom with Uganda, Pakistan and Bangladesh), and geographical proximity (e.g. Latin Americans favouring the United States) play key roles (Castles & Miller 2003; Díaz‑SáNchez et al. 2021; Havinga & Böcker 1999; van Tubergen et al. 2023). The United States may also appeal more to qualified and skilled migrants than Sweden (Birgier et al. 2018), potentially appealing to individuals from less‑developed countries in Africa to choose Sweden as their destination.
Moreover, ties through migrant communities and networks could also significantly influence migration decisions (Havinga & Böcker 1999). For instance, Sweden boasts a sizable migrant community from Iraq (Hammar et al. 2023), which could explain why Iraq ranks third among asylum‑seekers. The high ranking of Uganda and Nigeria may also be plausible, as these two countries have established ‘sexual orientation’ ties through smaller migrant networks. For instance, as early as 2013, the primary Swedish LGBTQ+ organisation published country reports for Uganda and Nigeria (Petersson 2013b, 2013a). These indicate that SO refugees from these countries have been present in Sweden for some time, despite the overall low number of residents from these countries in Sweden.
Sexual orientation claimants sorting into LGBTQ+ organisations dissimilarities
This final section explores the specifics of asylum cases involving SO claimants, shedding light on their uniqueness and the differences between those involved in LGBTQ+ organisations. Table 4 should be interpreted in the same way as Table 2. From the table, several conclusions can be drawn.
Table 4
Case specifics for asylum applicants in first‑instance decisions, 2020.
| VARIABLE | ALL | SO CLAIM | SO LGBTQ |
|---|---|---|---|
| Derivative right | 1.6% | 0.0%* | 0.0% |
| Harm | 62.0% | 70.0%** | 68.3% |
| Days in Sweden (mean) | 484 | 525 | 463 |
| Processing days (mean) | 384 | 447*** | 520* |
| Late applications | 2.1% | 12.8%*** | 10.0% |
| Several refugee status claims | 25.7% | 37.0%*** | 36.7% |
| Identity proven | 67.9% | 72.0% | 75.0% |
| Written evidence applicant | 46.7% | 53.3%* | 91.7%*** |
| Written evidence SMA | 16.5% | 16.0% | 21.7% |
| CoO report | |||
| Lifos | 50.5% | 25.7%*** | 18.3%* |
| EASO | 15.7% | 10.1%* | 11.8% |
| Landinfo | 13.8% | 5.5%*** | 6.8% |
| HRW | 5.5% | 10.1%*** | 11.9% |
[i] Note: The variable ‘Derivative’ rights indicates if a decision outcome was granted on other family members claims and assessments. ‘Harm’ refers to applicants that claimed to have been the subject of harm in their country of origin. Source: Swedish Migration Agency; calculations were performed by Angelica Dovenryd Almberg and Thomas Wimark.
Although comprehensive data are limited, there is a common perception that SO claimants face greater harm, supported by studies from the United States (Bird et al. 2022; Hopkinson et al. 2017). Table 4 indeed shows higher reported harm among SO claimants than other applicants. Late applications – referring to asylum applications submitted late or sexual orientation claims added during the asylum process – are also more frequent in this group, often due to unawareness of eligibility based on sexuality (Berg & Millbank 2009), which can lead to credibility issues (Asanovic et al. 2017). However, this study found no significant delay in application timing for SO claimants in Sweden. Notably, they are more likely to add or expand refugee claims during the process, indicating a distinct pattern in how their cases evolve.
The results also indicate that those with SO claims who are engaged in LGBTQ+ organisations in Sweden differ from those who are not. In the literature, the significance of evidence in credibility assessments is extensively discussed (Berg & Millbank 2009; Giametta 2017; Murray 2014a). The applicant’s narrative typically serves as the primary evidence, with supplementary written information potentially validating the claims. A distinct contrast emerged concerning the submission of written evidence by the claimants. SO claimants were six percentage units more likely to provide written proof, while those with engagement in LGBTQ+ organisations were 45 percentage units more inclined to do so. This suggests that those sorting into LGBTQ+ organisations differ from regular SO claimants.
Furthermore, the analysis revealed two significant disparities regarding the assessment by the SMA. The first concerns the use of CoO reports. While case officers predominantly relied on CoO reports authored by the SMA Lifos agency for regular cases, SO claimants in LGBTQ+ organisations more often involved alternate CoO reports. This discrepancy could stem from a lack of detailed information pertaining to the circumstances of LGBTQ+ individuals in the respective countries in the Lifos reports. As a result, case officers may have turned to alternative sources to better understand the specific challenges faced by SO claimants. Another explanation would be that claimants in LGBTQ+ organisations are better informed, driving case officers to use more varied CoO reports in their decisions. Second, the processing time by the SMA is the longest for claimants with experience from LGBTQ+ organisations; even when those with late claims are removed, the difference remains. This result could reflect a larger complexity of cases where the SO claimants have LGBTQ+ organisational experience, making cases harder to decide. This would be supported because this group is also more likely to present written evidence. A less likely alternative explanation would be that the SMA would deliberately delay cases in which the applicant has experience from LGBTQ+ organisations.
Discussion
The Results section has presented the spatial sorting tendencies among sexual orientation asylum‑seekers to Sweden. Additionally, it has highlighted the unique characteristics and differences between those involved in LGBTQ+ organisations and those not, i.e. those sorting into LGBTQ+ organisations and those not. The findings reveal that certain CoOs are prominently represented among those engaged in LGBTQ+ organisations, while others are underrepresented. Several factors may explain these sorting differentiations. One possibility is selective engagement by LGBTQ+ organisations with specific CoO groups, supported by the observation that the two most common countries of origin are those with which the main LGBTQ+ organisation has longstanding ties, suggesting stronger social network connections. Another potential explanation is the preferences of claimants, where individuals from certain CoO may avoid LGBTQ+ organisations. Alternatively, it may be due to a lack of information among claimants from certain CoOs. Overall, the results indicate an uneven outreach by LGBTQ+ organisations.
Moreover, the results indicate that those who engage with LGBTQ+ organisations are more likely to present written evidence and are assessed differently by the SMA in relation to CoO reports and processing times. This may suggest that individuals involved with LGBTQ+ organisations in Sweden are better prepared upon arrival. Another likely explanation, however, is that LGBTQ+ organisations equip their members more effectively and inform them about the importance of written evidence. Additionally, these organisations likely provide better knowledge about rights and access to lawyers with expertise in SOGIE asylum. Consequently, the SMA may need to justify its asylum decisions more thoroughly, relying less on its own CoO reports and taking longer to process these cases.
Previous research has demonstrated that LGBTQ+ organisations play a crucial role in the asylum process for SOGIE asylum‑seekers by enhancing social, economic and cultural capital. They help build networks, provide material assistance, aid in accessing healthcare and assist in preparing claims (Giametta 2020; Pullen & Franklin 2024; Wimark 2021a). Furthermore, welfare states depend on these organisations and allocate funding through them, such as through the Asylum, Migration and Integration Fund by the European Union (Migrationsverket 2025). However, research also shows that their geographical outreach is uneven; some SO claimants may avoid them due to feelings of not belonging or fears that association with them might lead to marginalisation (Allouche 2017; Held 2023; Wimark 2021a, 2021b). Additionally, some claimants may be ostracised if members of the organisation engage in policing activities to assist only those who ‘truly are gay’ (Giametta 2020). The research presented here seems to confirm some of these concerns. It further contributes to this knowledge by highlighting that not only is the outreach uneven, but clear advantages also appear to be measurable for those who are part of an organisation regarding both the preparation and assessment of their cases. This implies a system in which some SOGIE asylum‑seekers are afforded advantages over others.
Conclusions
Over a decade ago, Berg and Millbank (2009) hinted at a major influx of asylum‑seekers with sexual orientation claims. However, scant data exist to quantify the scope of SOGIE asylum‑seekers, let alone their sorting characteristics in relation to choosing their destination country and to engage with LGBTQ+ organisations. This knowledge gap has been a significant challenge in migration scholarship concerning sexual minorities (Danisi et al. 2021; García Rodríguez 2023; Shaw et al. 2022; Shaw & Verghese 2022). To address this lack, this study utilised a representative sample of first‑instance asylum decisions in Sweden to shed light on the sorting processes of SOGIE asylum‑seekers. Moreover, this study sought to explore the potential role of LGBTQ+ organisations in the sorting of refugees.
The findings confirm that refugees with sexual orientation claims are now close to comparable in size to other refugee claim categories, such as race/ethnicity/nationality. This demographic shift suggests that SOGIE asylum‑seekers are no longer a marginal group (also, see Home Office 2022). This warrants attention from state authorities in planning refugee reception and processing refugee populations, particularly those disproportionately affected by harm, as has been shown in this study. However, revealing numbers could also raise concerns. No longer ‘few enough in number not to threaten dominant systems’ (Luibhéid 2008: 179–180), SOGIE refugees now find themselves subject to increasing scrutiny and politicisation by state actors. In some countries, such as Sweden, LGBTQ+ refugees are still exploited as a pretext to justify restrictive migration policies (e.g. see Migrationsverket 2023). In others, like the United Kingdom, their very existence is becoming questioned (e.g. see Euronews 2023). Interestingly, while sexual orientation claimants now represent a larger number of asylum claimants and appear to be under political scrutiny, gender identity/expression claimants, who, according to this study, remain a relatively small proportion of asylum claimants, do not (yet) pose a perceived threat to the system.
Moreover, this study adds nuance to existing research on SOGIE asylum‑seekers by suggesting a potential risk for bias in knowledge production, as those sorting into LGBTQ+ organisations differ in relation to where they originate from, how much written evidence they present and how much time it takes for the SMA to process their cases. While current approaches often centre around LGBTQ+ organisations as primary sources of information (e.g. see Chossière 2021; Dhoest 2019; Lunau & Andreassen 2023; Lunau & Schröder 2025; Murray 2014a; Schröder 2023; Tschalaer 2020, 2021), these findings caution against the overreliance on such channels. This study demonstrates that significant segments of SOGIE asylum‑seekers, particularly those originating from certain countries, are not necessarily connected to these organisations. This underscores the limitations of LGBTQ+ organisations’ outreach and raises concerns about the representativeness of knowledge derived from them.
The findings are based on asylum decisions, which have limitations, particularly the subjective and selective nature of case officer interpretations. As case officers shape the decisions, information may be emphasised or downplayed to support their stance. For example, when assessing engagement with LGBTQ+ organisations, references were often found in narrative summaries and through formal evidence like membership letters. While it is unlikely that officers deliberately omit such documentation, mentions of affiliations may still go unrecorded. Additionally, some applicants may choose not to disclose organisational involvement, depending on their country of origin.
While scholars often rely on LGBTQ+ organisations to facilitate research, asylum authorities similarly depend on these organisations to promote the well‑being of SOGIE asylum‑seekers and to foster community‑building among them, as they do not themselves offer such services. Research also indicates that greater investment in LGBTQ+ organisations can have positive impacts on their well‑being (Istiko et al. 2024). In Sweden, the SMA frequently consults with the country’s primary LGBTQ+ organisation to help shape policy and ensure well‑being. At the same time, it is not uncommon for ‘regular’ NGOs to refer SOGIE asylum‑seekers to LGBTQ+ organisations (Mahn et al. 2021; Ongwech et al. 2024). Although these practices support the well‑being of asylum‑seekers, this study highlights the limitations of such approaches. As discussed in the theoretical section, NGOs play multiple roles that extend beyond direct support. They are embedded in global networks, acting as facilitators in the migration process and shaping perceptions of migration opportunities. Additionally, they provide crucial support networks for potential migrants in their countries of origin. By acknowledging these broader roles and the limitations identified in this study, policymakers can work towards more inclusive and comprehensive frameworks for supporting SOGIE asylum‑seekers.
Acknowledgements
Angelica Dovenryd Almberg played an essential role as the main coder of the data in this study. The support of the Swedish Research Council for Health, Working Life and Welfare under Grant GD‑2021/0028, which provided essential funding for this research, is gratefully acknowledged. I also extend my sincere appreciation to all members of the expert reference group for their valuable time and contributions: Silas Aliki (lawyer, Folkets advokatbyrå), Johannes Björhn (lawyer, Asylrättscentrum), Ann‑Christin Cederborg (Professor, Stockholm University), Anna Lindblad (Deputy Head of Legal Affairs, Migrationsverket), Joakim Lundqvist (lawyer, Folkets advokatbyrå), Alexandra Politis Palm (lawyer, RFSL ungdom/Newcomers Youth) and Siri Sandin (lawyer, Asylrättscentrum). The research team consisted of Danielle Drozdzewski, Rebecca Thorburn Stern and myself, and the study would not have been possible without the discussions held within the group.
Competing Interests
The author has no competing interests to declare.
Data Availability Statement
The data that support the findings of this study are not publicly available due to privacy and ethical restrictions. The data are available from the Swedish Migration Agency. Restrictions apply to the availability of these data, which were used under ethical approval from the Swedish Ethical Review Authority for this study.
Biographical Note
Thomas Wimark is a Senior Lecturer and Associate Professor at the Department of Human Geography at Uppsala University. His research focuses on migration, urban planning and marginalized groups, with a particular emphasis on how normative views of sexuality, gender and race impact (queer) individuals migrating within and between countries. His academic work draws on Swedish register data as well as qualitative research methods. Wimark has published articles in high‑ranked journals such as the International Migration Review, the Journal of Ethnic and Migration Studies, Social and Cultural Geography and Housing Studies. Additionally, he serves on the editorial board of Gender, Place and Culture.
Funding Details
This work was supported by the Swedish Research Council for Health, Working Life and Welfare under Grant GD‑2021/0028.
Notes
[5] The United Kingdom is an exception to this since it presents ‘experimental’ statistics on sexual orientation. One of the issues with the statistics presented is that case officers are instructed to flag cases based on sexual orientation, but a report has shown a low level of adherence to the instruction (Danisi et al. 2021).
[6] The 1951 Convention Relating to the Status of Refugees and the 1967 Protocol Relating to the Status of Refugees.
[7] Directive 2011/95/EU of the European Parliament and of the Council of 13 December 2011 on standards for the qualification of third‑country nationals or stateless persons as beneficiaries of international protection, for a uniform status for refugees or for persons eligible for subsidiary protection, and for the content of the protection granted (recast), OJ L 337/9.