Deciding on asylum dilemmas: a conflict between role and person identities for asylum judges

ABSTRACT

Large discrepancies in asylum recognition rates across individual judges and asylum agencies constitute a cause for concern. To better understand the asylum determination process, existing research has addressed various aspects, including the identities of asylum judges and their role as street-level bureaucrats who make discretionary decisions. Building on these streams of literature, this study examines how asylum judges in Germany and Greece make decisions under conditions of high uncertainty. Drawing on original interview data with lay and administrative judges from the respective capitals, this study focuses on ‘grey area’ asylum cases and sheds light on the decision-making mechanisms used by judges. This research finds that, in the face of moral dilemmas, judges experience a conflict between their role and person identities. Depending on which identity they prioritise, their decisions may be more ‘evidence’-based or more preference-based. This article highlights the importance of the identities of asylum judges, not only as bureaucrats but also as people with a unique set of beliefs, norms and values. In doing so, it contributes to the debate on asylum determination and the field of street-level bureaucracy.

KEYWORDS:

• Refugee asylum; decision-making; identities; street-level bureaucracy

Introduction

The process of making decisions for refugee asylum applications has been described by many as a ‘black box’. While, in principle, it is a simple task of labelling and categorisation, in practice it is not always so straightforward. The staggering differences in refugee recognition rates across countries (Schuster, 2018), regions (Ramji-Nogales, Schoenholtz, & Schrag, 2007) organisations (Dallara & Lacchei, 2021) and individual judges (Keith, Holmes, & Miller, 2013) are partly reflective of the invisible intricacies of the process. These differences can also be worrisome, as they could signify a refugee asylum system that is potentially unfair or unreliable.

Scholars from various disciplines have investigated the process of asylum determination in recent years, in an attempt to understand and explain the practice of asylum policy and policy outcome inconsistencies. One of the theoretical frameworks employed for this purpose is that of street-level bureaucracy (Lipsky, 1980), a bottom-up approach to policy implementation, focusing on the individuals at the frontlines of service delivery. Street-level bureaucrats have direct interaction with clients and some discretion in making decisions while carrying out policy tasks. Depending on how they use this discretion, they shape policy outcomes.

The discretionary decisions of asylum judges, referring henceforth to both first- and second-instance decision-makers, have considerable consequences. Not only do they shape official recognition rates, but they can also have life-altering, potentially lifesaving consequences for asylum seekers. In this regard, asylum decisions also encompass a significant moral dimension (Shiff, 2021; Vianelli, Gill, & Hoellerer, 2022). This is especially apparent in more complex asylum cases that may not neatly fit within existing policy guidelines, as well as in asylum decisions that take place during pressing times, such as that of the so-called European refugee crisis of 2015–2017, on which this study focuses.

Various factors have been found to explain the discretionary decisions of asylum judges, including their individual demographic characteristics and political affiliations (Ramji-Nogales et al., 2007; Rehaag, 2011; Spirig, 2023). Nonetheless, although the importance of the ascribed identities of judges has been underscored, existing studies pay little attention to the identities the judges themselves prioritise during the process of asylum decision-making. Moreover, while the literature of street-level bureaucracy has addressed the importance of identity conflicts and moral dilemmas in policy implementation (Glyniadaki, 2021; Jensen & Pedersen, 2023; Maynard-Moody & Musheno, 2003; 2012; Tummers, Vermeeren, Steijn, & Bekkers, 2012; Vink, Tummers, Bekkers, & Musheno, 2015), the moral dimension has not been sufficiently examined in the context of asylum determination. This study aims to address both gaps.

Using original data gathered from qualitative interviews with asylum (lay) judges in Athens and Berlin, as well as with relevant legal professionals and experts in the field, and combining theoretical perspectives from public administration and social psychology, this study examines the role of judges’ identities on asylum determination. Focusing on what participants identify as challenging or ‘grey-area’ asylum claims, it sheds light on the decision-making mechanisms used by judges. This research finds that the moral dilemmas asylum judges encounter lead to identity conflicts between their professional role identity as judges and their person identity (see Burke & Stets, 2009) as compassionate – or less so – individuals. Depending on which identity judges prioritise, their decisions would be more ‘evidence’-based, or more aligned to personal normative preferences.

In short, this article advances the argument that, to better understand how asylum decisions take place in practice, especially in challenging asylum cases and under pressing conditions, judges’ self-views should also be considered, not only as bureaucrats but also as persons with a given set of values, beliefs and normative understandings. This approach is in line with Lipsky’s attempt to highlight the ‘human face’ of bureaucrats, while it also adds to the debate on discretionary decision-making in policy implementation.

More specifically, this study makes a three-fold contribution to the existing literature. First, building on existing studies that analyse judges’ subjective sense-making and interpretation of asylum claims (Jensen, 2023; Vetters, 2022), this research also addresses their self-understandings as decision-makers. Second, acknowledging that asylum decision-making entails moral dilemmas (Shiff, 2021), this study shows how these dilemmas translate into identity conflicts between the role and person identities of judges. Third, although previous accounts emphasise the negative impact of judges’ discretionary decisions on asylum claimants (Giametta, 2020; Johannesson, 2018; Magalhães, 2016, 2018), this research shows there are also instances of ‘positive discretion’, where judges make an explicit effort to help claimants.

The remainder of the article is divided into four parts. The section immediately below provides an overview of the existing literature on asylum determination and on moral dilemmas and identity conflicts at the street level. It is followed by a description of the research methods and data used in this study. The next section presents the proposed decision-making mechanism of asylum judges, and the one after that illustrates the different parts of this mechanism, using interview data extracts. The article ends with some concluding remarks, highlighting how the role and person identities of asylum judges may play a critical role in asylum decision-making in the face of moral dilemmas.

Asylum determination and the role of identities

In 2015, European countries recorded 1,349,638 asylum applications (EASO, 2016), an unprecedented number compared to previous years. This placed asylum processing at the top of political agendas, while it also ignited a revitalisation of academic interest in the topic (Schuster, 2018). Asylum determination has been studied through the lens of various scholarly disciplines and traditions over the years. I take here the perspective of street-level bureaucracy (Affolter, 2021a, 2021b; Dahlvik, 2017; Dallara & Lacchei, 2021; Miaz, 2017), which focuses on the practice of policy and the use of discretion by implementing actors. I explore the link between street-level bureaucrats’ moral dilemmas and identities (Maynard-Moody & Musheno, 2003; 2012; Tummers et al., 2012; Vink et al., 2015; Zacka, 2017), and I also draw from social psychology and Identity Theory (Burke & Stets, 2009).

Asylum judges as street-level bureaucrats

In many respects, asylum judges are very close to the archetype of the street-level bureaucrats that Lipsky (1980) introduced over 40 years ago: they are public servants, have direct interactions with clients, and exercise a fair amount of discretion when making daily decisions (Lens, 2012). Although these bureaucrats’ use of discretion has been problematised in studies of several disciplines, including law (e.g. Keith et al., 2013) sociology (e.g. Schittenhelm & Schneider, 2017), political science (Spirig, 2023) and even economics (e.g. Chen et al., 2016), the angle of public administration, and street-level bureaucracy, remains less explored (Billand & Steinmetz, 2017).

From this perspective, asylum determination presents an interesting case. As per the Geneva Convention, asylum judges must apply the law with consistency for all claimants, while at the same time, they must examine each individual application on a case-by-case basis (2013/32/EU directive). This means that discretionary decisions are an inherent part of the process and some variation in asylum outcomes is therefore inevitable. Nonetheless, too much variation is undesirable. Pointing out the stark differences in recognition rates across countries (e.g. Table A1 in Appendix), regions, organisations and individual decision-makers, several studies have suggested that asylum applicants are often subjected to ‘lottery-like’ conditions (e.g. Ramji-Nogales et al., 2007; Rehaag, 2011). They essentially argue that asylum applications are not always judged on the basis of clearly defined and consistent criteria, but rather in accordance with individual, organisational, or systemic biases.

While contextual idiosyncrasies are also of high importance in asylum determination (Gill et al., 2022), this study focuses on the internal tensions and sense-making processes of individual decision-makers. More specifically, it examines the discretionary responses of asylum judges to everyday moral dilemmas, in line with recent studies that address ‘creative and morally conscious bureaucratic action’ (Shiff, 2021). In doing so, it also adds to existing efforts to inform street-level bureaucracy through social psychological theorising (e.g. Jensen & Pedersen, 2023; Zacka, 2017).

As previous research shows, individual asylum judges face various difficulties during the asylum determination process. Communication with applicants, for example, is likely to be challenging. This may be due to judges’ lack of understanding of claimants’ local cultural contexts, claimants’ traumatic stories which necessitate sensitive and uncomfortable exchanges (Dallara & Lacchei, 2021), or language barriers and interpretation shortcomings (Hedlund & Johannesson, 2023). Moreover, in pressing times when asylum application backlogs rapidly increase, these challenges may be further compounded by intensified workload and time pressure to meet quota targets (Schittenhelm & Schneider, 2017).

In addition, uncertainty for asylum judges may stem from gaps in or contradictions of the policy framework itself, allowing for differing interpretations by policy practitioners (Miaz, 2017). A notable example, historically, has been the question of what constitutes ‘a particular social group’, as per the Geneva Convention definition. Since the 1980s, feminist critics have argued that gender should be included in this definition (Arbel, Dauvergne, & Millbank, 2015), as women in certain areas, especially conflict zones, are systematically subjected to persecution. Although the UNHCR has since issued guidelines to tackle this issue, it has been widely acknowledged as a source of asylum outcome inconsistencies over the years (Freedman, 2015).

Further to the list of challenges, there are also asylum claims for which it is almost impossible to provide tangible supporting evidence. For cases pertaining to persecution based on one’s sexual orientation, for instance, applicants may not always be able to find evidence to prove this, or the evidence they can provide is not necessarily regarded as reliable and persuasive (Dallara & Lacchei, 2021; Giametta, 2020). This could potentially also have a flipside, where some applicants use such claims to increase their chances of receiving asylum, precisely because they are difficult to prove, making asylum judges (even more) suspicious of such applicants.

For all these types of challenging cases, it often comes down to judges to decide on the claimant’s ‘credibility’, meaning whether and which of the applicants’ statements and other evidence submitted can be accepted and considered (UNHCR, 2013, p. 27). The issue of credibility has indeed been at the heart of many studies discussing the process of asylum determination (Affolter, 2021b; Dahlvik, 2017; Magalhães, 2016; Miaz, 2017; Sorgoni, 2019; Tomkinson, 2018; Verhaeghe, Jacobs, & Maryns, 2023). This also has parallels with the notion of ‘deservingness’, which is central to street-level bureaucracy research and theory (Jilke & Tummers, 2018). In both lines of research, the focal point is that bureaucrats’ judgements of their clients shape their use of bureaucratic discretion.

Dealing with moral dilemmas

Determining a person’s credibility or deservingness inevitably entails a moral dimension. To some degree, making decisions that encompass moral complexities is an integral part of delivering human services at street level. Existing studies in street-level bureaucracy have extensively discussed moral dilemmas faced by police officers (Maynard-Moody & Musheno, 2003, 2012) or medical professionals (Tummers et al., 2012; Vink et al., 2015), while lesser attention has been dedicated to those faced by asylum judges. Yet, much like a doctor deciding whether to discontinue an expensive experimental treatment for an unresponsive patient, or a community police officer deciding whether to arrest an underprivileged delinquent juvenile, an asylum judge must decide whether returning home incubates a life-threatening danger for an asylum applicant.

One common observation in the existing literature is that moral dilemmas manifest into identity conflicts for street-level bureaucrats. These are conceptualised as tensions within one’s role prescriptions, stemming from contradicting role expectations. Tummers and colleagues (2012) identify three such conflicts: the ‘policy–professional’ conflict between what a policy requires and what the role prescribes; the ‘policy–client’ conflict between what a policy states and what a client needs; and the ‘organisational–professional’ conflict between organisational demands and professional values. Adding to these, Vink and colleagues (2015) discuss the ‘professional–client’ conflict between one’s professional values and their clients’ demands. In these, as well as in more recent studies (Jensen & Pedersen, 2023), the street-level bureaucrats’ role identity is seen as the key ‘explanatory variable’ behind their decisions.

However, it can be argued that judges, as well as other professionals, do not process information and make decisions solely as bureaucrats. Particularly in politically controversial and socially divisive fields, such as migration management, policy implementers are likely to have their own stance in the public debate, which may impact their discretionary decisions (e.g. Bell, Ter-Mkrtchyan, Wehde, & Smith, 2021; Glyniadaki, 2022). Especially when working under high pressure and in conditions of uncertainty, they often need to come up with new solutions to unprecedented problems. When confronted with cases that are not adequately covered by existing policies and protocols, for example, street-level bureaucrats, including judges, may respond with ‘moral problem-solving’ (Shiff, 2021). Individual judges’ ‘intimate conviction’ (Kobelinsky, 2019) or ‘subjective impressions’ of claimants (Dahlvik, 2017) are also found to affect their decision-making. That being so, for a more comprehensive understanding of identity-informed decision-making at the street level, it is important to look beyond bureaucrats’ role identities.

According to Identity Theory (Burke & Stets, 2009), there are three bases of identities: role, social and person (Table A3 in Appendix). Role identity refers to the position an individual holds within a social structure or an organisation. A professional identity is therefore a role identity. There is also social identity, which refers to a person’s identification with a social group, such as a racial/ethnic group or even a sports club. Lastly, and with greater relevance to this research, there is person identity, which describes an individual’s unique idiosyncratic characteristics or their ‘core self’, and which transfers across situations, roles and group memberships (Ibid.). Being ‘empathetic’ or ‘compassionate’ would be examples of person identities.

It follows then that, in the face of moral dilemmas, judges make decisions not only as bureaucrats, but also as persons with a given set of values and beliefs, and normative understandings, or ‘moral predispositions’ (Zacka, 2017). Existing studies have already touched upon this, discussing the importance of judges’ attitudes towards policy (Keith et al., 2013), or the weight they place on their ‘inner belief’ (Kobelinsky, 2019) or ‘feeling’ (Miaz, 2017) during decision-making. Shiff (2021) more explicitly describes the mismatch or ‘discordance’ asylum judges may experience between the perceived deservingness of a claim, based on shared moral schemas, versus the lack of legal fit under codified law. As Vetters (2022) also finds, an asylum judge ultimately wants to be able to ‘face oneself in the mirror’ at the end of the day, meaning to make decisions that live up to their personal sense of justice.

Method & data

This study employs a qualitative methodological approach, based on semi-structured interviews with 24 asylum (lay) judges in Athens in Berlin (Table 1), which took place between June 2017 and December 2018. Of these participants, 17 were making first-instance asylum decisions (9 in Athens and 8 in Berlin) and 7 second-instance decisions (5 in Athens and 2 in Berlin). The majority (20) were lay judges, meaning they were public servants with an educational background in social sciences or law. In addition, 3 of the participants were administrative judges (members of the judiciary), making decisions at the appeals stage (2 in Berlin and 1 in Athens).

Table 1. List of main participants: Asylum (Lay) Judges.

		Official Role	Referred to as	Decision Instance	Education Background	Gender	Age
Athens	1	Caseworker	Lay Judge	1st	Social Sciences	M	40–45
	2	Caseworker	Lay Judge	1st	Social Sciences	F	35–40
	3	Caseworker	Lay Judge	1st	Social Sciences	F	50–55
	4	Caseworker	Lay Judge	1st	Social Sciences	F	40–45
	5	Caseworker	Lay Judge	1st	Social Sciences	F	40–45
	6	Caseworker	Lay Judge	1st	Social Sciences	F	30–35
	7	Caseworker	Lay Judge	1st	Social Sciences	F	40–45
	8	Caseworker	Lay Judge	1st	Social Sciences	F	40–45
	9	Caseworker	Lay Judge	1st	Social Sciences	F	35–40
	10	Appeals Committee Member	Lay Judge	2nd	Law	F	30–35
	11	Appeals Committee Member	Lay Judge	2nd	Law	F	35–40
	12	Appeals Committee Member	Lay Judge	2nd	Law	F	30–35
	13	Administrative Judge	Judge	2nd	Law	F	50–55
Berlin	14	Caseworker	Lay Judge	1st	Law	M	40–45
	15	Caseworker	Lay Judge	1st	Social Sciences	M	25–30
	16	Caseworker	Lay Judge	1st	Social Sciences	F	20–25
	17	Caseworker	Lay Judge	1st	Social Sciences	F	30–35
	18	Caseworker	Lay Judge	1st	Social Sciences	F	25–30
	19	Caseworker	Lay Judge	1st	Social Sciences	F	30–35
	20	Caseworker	Lay Judge	1st	Law	F	45–50
	21	Caseworker	Lay Judge	1st	Social Sciences	F	20–25
	22	Caseworker	Lay Judge	1st	Social Sciences	M	30–35
	23	Administrative Judge	Judge	2nd	Law	M	40–45
	24	Administrative Judge	Judge	2nd	Law	F	35–40

First-instance decisions were taken in both cities by lay judges, known as ‘caseworkers’. At the second instance, however, there was variation. At the time of the data collection, the Greek asylum system was undergoing a period of transition. According to the ‘old’ system, appeal decisions were taken by a committee of 3 lay judges, comprised of 2 public servants and 1 lawyer appointed by the UNHCR. Responsibility for these decisions was later transferred to single administrative judges, which made the overall process similar to that of Berlin. Nonetheless, the ‘old’ and ‘new’ systems were operating simultaneously during this transition phase, the period when this research was conducted. Of second-instance decision-makers from Athens, four participants were lay judges (lawyers appointed by UNHCR) and 1 was an official administrative judge.

As this research is part of a larger project, in addition to interviews with asylum judges, it also draws from 21 interviews with experts in the field (Table 2), including 6 top-level bureaucrats in relevant public agencies and ministries (4 in Athens and 3 in Berlin), as well as 12 refugee asylum lawyers and legal support officers (6 in Athens and 8 in Berlin). While the latter set of interviews is not the focus of the analysis below, it did assist towards a greater contextual understanding of asylum processes across the two capitals. Most notably, interviews with asylum lawyers helped underscore the ‘grey areas’ of asylum decision-making, meaning the types of asylum claims for which it is considered harder to provide evidence, and on which it is therefore harder to come to a decision.

Table 2. List of other participants: asylum experts and laywers.

		Role	Organisation	Gender	Age
Athens	1	Top-level Bureaucrat	Ministry of Migration Policy*	M	50–55
	2	Top-level Bureaucrat	Ministry of Migration Policy	M	55–60
	3	Spokesperson	Asylum Service, Ministry of Migration Policy	F	50–55
	4	EU Administrative Employee	EASO* Athens	M	40–45
	5	Asylum Lawyer	UNHCR	F	25–30
	6	Asylum Lawyer	UNHCR	F	25–30
	7	Asylum Lawyer	(pro bono 1:1)	F	25–30
	8	Asylum Lawyer	Local NGO	F	30–35
	9	Asylum Lawyer	Save the Children	F	25–30
	10	Asylum Lawyer	Save the Children	F	25–30
Berlin	11	Top-level Bureaucrat	Federal Office for Migration and Refugees	F	40–45
	12	Top-level Bureaucrat	Federal Office for Migration and Refugees	F	40–45
	13	Manager of Caseworkers	Federal Office for Migration and Refugees	M	30–35
	14	Asylum Lawyer	(pro bono 1:1)	F	65–70
	15	Asylum Lawyer	Private Office	F	30–35
	16	Asylum Lawyer	Private Office	F	45–50
	17	Asylum Lawyer	Private Office	F	35–40
	18	Asylum Lawyer	Private Office	M	35–40
	19	Asylum Lawyer	Private Office	M	30–35
	20	Legal Support to Asylum Seekers	Local NGO	M	50–55
	21	Legal Support to Asylum Seekers	Local NGO	F	25–30

*Notes: The table indicates the names of these organisations at the time of fieldwork. The Greek Ministry of Migration Policy has subsequently been renamed Ministry of Migration and Asylum. The European Asylum Support Office (EASO) has subsequently become European Union Agency for Asylum (EUAA).

I recruited most research participants through formal request phone calls or emails submitted to their respective organisations. I also identified some of the participants through informal networks of professionals in migration management as well as through the snowball technique. My requests to attend asylum interviews, which are generally closed to the public, were declined. I conducted most interviews at each participant’s place of work, or a nearby public place, while two were conducted online. Each lasted from 40 to 90 min. Having received the consent of participants, I audio-recorded all but one, where the participant had declined to give consent. I then transcribed them verbatim and analysed them thematically (Braun & Clark, 2006) through the qualitative analysis software NVivo. The initial themes I identified were: ‘moral dilemmas’, ‘identities’, and ‘decision-making’. At a later stage of analysis, some of the secondary themes that emerged were: ‘role identity’, ‘person identity’, and ‘“evidence”-based’ decisions and ‘preference-based’ decisions.

Two points are important to clarify here. First, although there were organisational differences across the respective asylum agencies in Athens and Berlin – for example, the availability of material resources – analysing the impact of organisational-level dynamics is beyond the scope of this article. The focus of analysis is instead on individual judges and their subjective interpretations of the asylum decision-making process. As such, I analyse the accounts of German and Greek asylum judges not in comparison to each other but in parallel, as the following sections show. Second, I present below the theoretical contribution of this article, namely the identity-informed decision-making mechanism of asylum judges, followed by a discussion of this study’s empirical findings. Although theoretical, this contribution is largely informed by empirical evidence, as interview data analysis and theoretical development proceeded iteratively.

The decision-making mechanism

A key observation of this study is that, when encountering challenging asylum cases, judges experience a ‘role–person’ identity conflict, where their professional role prescriptions are incongruent with their personal and moral inclinations. The latter includes their individual values, beliefs and normative understandings of the issue at hand, which may play a critical role in the face of moral dilemmas. The outcome of this identity conflict seems to shape the asylum decisions judges make, thereby shaping policy outcomes.

To illustrate this ‘role–person’ identity conflict, I present here a map depicting the decision-making mechanism of asylum judges, based on the analysis of the interview data in this study. As Figure 1 shows, there are various potential routes in the proposed mechanism. In chronological order, the first step is ‘case assessment’. This is when asylum judges examine the information available to them (interview transcripts, supportive documents, etc.) and decide whether the asylum claim in front of them clearly fits the legal definition of a ‘refugee’. If yes, they proceed to issuing a positive asylum decision. If not, the decision will be one of rejection.

Figure 1. Decision-making mechanism for asylum applications.

However, as already explained, there is also a category of ‘grey area’ asylum cases, where the claims cannot be easily accepted or rejected. This could be because the claimant could not provide a clear and coherent story, or because the evidence in support of the persecution claim (e.g. ‘I was chased by the Taliban’) was not strong enough. It could also be that, due to the nature of the claim (e.g. homosexuality), it is very difficult to prove. In any case, this is when asylum judges encounter, what Zacka (2017) calls, an ‘impossible situation’.

In such ambiguous cases, the issue of claimants' credibility and deservingness takes centre stage, and the dilemma that comes with it is a moral one. On the one hand, asylum judges must allow a fair chance to each applicant whose claim is under examination. As such, they should not be quick to reject complicated cases. On the other hand, giving false positives can also prove to be highly problematic, both for policy implementation and politically. External pressures may also come into play at this stage, including the dominant political discourse, public preferences, principal–agent, peer-to-peer and bureaucrat-client dynamics (Table A4 in Appendix; see also Miaz, 2017).

Amidst a lack of protocol-based solutions, this study suggests that, for asylum judges, this moral dilemma becomes an identity conflict between their role identity as bureaucrats and their person identities as compassionate (or less so) individuals. Their asylum decisions are thereby shaped by how they resolve this conflict. If role identity proves more salient in the given context, the judges’ asylum decision will be based on the direction of the information at hand, even if this information is very limited. If the ‘evidence’ in support of the applicant’s claim suggests a ‘well-founded fear’, the asylum decision will be positive, regardless of a judge’s personal stance on the case. If the ‘evidence’ that contradicts the claim seems stronger than that which supports it, then the decision will be negative.

Alternatively, if the asylum judge’s person identity dominates, their decision will be in accordance with the values, beliefs and normative understandings this identity encompasses. There may be religious, political, or humanitarian underpinnings, as previously noted. Irrespective of these, a judge may take a more compassionate or less compassionate stance towards asylum seekers. In the more compassionate scenario, the limited evidence will be examined under a positive light, giving rise to a positive asylum decision. However, if the judge’s person identity is less compassionate towards asylum-seekers, the ‘evidence’ will be interpreted through a negative filter, resulting in a negative decision outcome (Figure 1).

Asylum determination in practice

To illustrate the different stages of the decision-making mechanism introduced above, I present here empirical evidence from this study’s interview data. Again, these stages involve: (a) the recognition of ‘grey area’ cases by the asylum judges, (b) the moral conflicts that manifest as ‘role–person’ conflicts, and (c) the ‘role’-based resolution, or (d) the ‘person’-based resolution.

The grey areas

While the more challenging and ambiguous asylum cases constitute a minority of the cases according to the accounts of the participants in this study, they are nevertheless particularly meaningful, considering they are more likely to lead to inconsistent asylum outcomes. As already noted, a variety of asylum cases may fall into this ‘grey area’ range. Some are known categories of cases that frequently present themselves in this grouping and are largely expected. These include cases when it is nearly impossible to establish the ‘the facts’, or when ‘the facts’ are established but still may be subject to interpretation.

It’s not so much the law, but the facts. This is very, very difficult because, as you know, asylum seekers normally don’t have any documentation with them, or they don’t show it to the authorities […] You never know whether what they state is true, because they have an interest in telling a story which will get them asylum. […] On the one hand, you are not supposed to set unattainable evidence standards, and on the other hand you are constantly being lied to, and it is very difficult to say, “Yes, I believe this person. I am sure, I have no doubt that this person is telling the truth”. (Judge 24, Berlin)

On the one hand, the law is the law, but the interpretation of the facts is subjective. I mean, for one person, it is perhaps enough [to know] there were 5 security incidents in that specific area. They will say, “Pfff, the other area had 100”. But, someone else might say, “Yes, 5, but out of these 5 [people who died] the 4 concerned”, let us say, “persons with military profiles. And my guy here has a military profile, too. Therefore, I judge that this guy will be in danger”. [The law] does not tell you not to do this. It cannot tell you what decision to make. It is up to the [decision-maker]. (Lay Judge 4, Athens)

In the first quote, the judge from Berlin describes the difficulty of deciding on whether an applicant’s claim is true, given the recurrent issue of lack-of-evidence. The second quote addresses the various layers of analysis in which an asylum judge may possibly engage when interpreting the facts. These observations echo studies that highlight the challenging process of fact-finding and fact-interpretation, and their connection with credibility assessment (Dahlvik, 2017; Jensen, 2023; Tomkinson, 2018; Verhaeghe et al., 2023). It is also worth acknowledging here that the asylum judges’ in-depth understanding of the related legal provisions could mean that almost any decision is a lawful decision, as long as it is well argued in black and white.

In addition to the above, there are other, more specific cases that can be challenging for asylum judges. The first example below refers to mentally ill claimants and the second to traumatised minors.

On 4 out of the 5 days, I encounter problems where there is no protocol to solve that problem, you know? … So, for example, […] I get the electronic file and I see the picture and where the person is from, yada yada … But then I go downstairs and I meet that person and she is … to say it in a politically correct way … she is mentally unfit to perform an interview […] I was talking to her, you know, but it was impossible to find a connection. What do you do with that type of person? She couldn’t even answer “What’s your name?” or “How are you feeling? Are you ready for the interview?” She was like … “Out of the window. There. They’re coming for me”. Like, OK, wow … So, what do you do? There’s no protocol! (Lay Judge 15, Berlin)

It is difficult to handle cases of unaccompanied minors who, often arriving from North African countries, live in hostels for minors, and state they are victims of rape. There, you have to deal with a 16–17-year-old kid who is all closed in on themselves. […] During the interview, the applicant offers too little information about their case, while we have their background from the NGO responsible for them and we try to figure out an objective decision on that person. And all that while I don’t have training, or expertise, in unaccompanied minors and victims of torture. But, I do have all this weight on me to make an objective decision. (Lay Judge 5, Athens)

The first lay judge here appears overwhelmed and underprepared for such challenging cases, while the second explicitly states that the weight of the responsibility involved in making such a decision feels disproportionate, given their lack of relevant training. These examples further illustrate that asylum judges often come across cases which they find confusing and challenging, and for which they cannot provide easy yes or no answers (see also Dallara & Lacchei, 2021). Whether it is about establishing the facts, interpreting them, or handling highly vulnerable claimants, making fair and objective asylum decisions can sometimes seem an elusive goal. Inevitably, such discretionary decisions require individual moral judgements.

The identity conflicts

As mentioned earlier, in such ‘grey area’ cases, moral dilemmas manifest as identity conflicts between the role and person identities of asylum judges. The following quotes show how the ‘professional’ is very closely linked with the ‘personal’. This, in turn, means there is a variation in individual approaches to dealing with difficult cases, which can relate to a discrepancy in decision outcomes.

In general, the issue is very political, and I try to stay completely neutral. And because I have learned to do so as a lawyer, it is easier for me. […] There were other colleagues, who I would meet with during the break, and they would tell me, “Great, yet another negative decision”. Or, in contrast, others who were entirely like, “Okay, whatever the applicants say is true. Poor them”. So, there were people from both sides. (Lay Judge 14, Berlin)

There are some of my colleagues who are less scientists and more ‘humanitarians’. […] Humanitarianism is hypocritical, because you can’t say you give someone asylum because you are a humanitarian. No! You give them asylum because your country has signed the Geneva Convention, and you have the obligation, if they fulfil the requirements of the Geneva Convention, to give them asylum. Even if you are not a humanitarian, and even if you dislike [the applicants] … The ‘humanitarians’ can only bring negative consequences because they distance the discussion from the real problem. If there is a real problem, it needs to be solved in a scientific way. (Lay Judge 10, Athens)

For these participants, a ‘neutral’ approach stripped of emotional attachment and personal inclinations is the preferred route. This is also more widely deemed to be the proper professional strategy (Schneider, 2019). The same does not appear to be true, however, for their colleagues in this study: by describing the preferred strategy of others as false or insufficient, these participants position themselves on the opposite side of this role-person conflict (see also Johannesson, 2018). For them, the professional role identity is more salient.

Nonetheless, just as some judges view their colleagues as being too biased and too ‘humanitarian’, others view theirs as too cold-hearted and lacking in empathy. The first two excerpts below come from a lay judge on the Board of Appeals in Athens, who was also the president of her committee, making second-instance decisions together with two other members. The second set comes from an administrative judge in Berlin who also makes second-instance decisions. As these participants note, some asylum judges may be too economical in their use of discretion, especially when it is in the (best) interests of claimants.

My percentage of cases was around 33% recognition, which was very good, very positive. So, in 33% of the cases, I got a positive decision, overturning the first-degree decisions. There was [another committee] president who had 1 positive case out of 110!

I remember when I was fighting for the Afghanistan case, a lady who had previously worked as a secretary at a ministry told me: “Okay, in Greece, too, we have terrorist attacks. We are not in danger”. I said “Are you seriously comparing Kabul to Athens now!? … In Kabul there are bombs exploding, for Christ’s sake!”. I think there are just different perspectives. Entirely. […]. Those [like me] who had worked for NGOs before becoming judges had a different kind of sensitivity or understanding. Those who had come from other positions, ministries, administrative positions, or academic positions, would reach decisions that could be considered harsher. They are stricter and more conservative. (Lay Judge 12, Athens)

In my old chamber I think I was, together with one colleague, the person more inclined to say “Yeah, I believe what they say” and “No, I don’t think they can be treated like this”, or that they should be given some leverage. And there were others who had a greater tendency to say that they don’t believe them, and “No, they are responsible for their own actions”.

So, [this claimant] was at this market with this cow, and then all this military came, and then it was a mass panic. That was the way he told it […] A woman from the Bundesamt¹, sitting in her nice warm office […], wrote “It’s totally illogical that there was a mass panic because it’s contrary to life experience that there is mass panic if the militia comes” […] But it’s not [Germany], it was in Eritrea, in a village! And then the next sentence was, “It’s also contrary to life experience” – [as in] my personal experience sitting in the office – “that you would leave the hospital, because the hospital is a place where you should get help and not get arrested”. Yeah, it should be … But it was obvious you failed to put yourself into the other individual’s [position] (Judge 23, Berlin)

By stating what they are critical of, these judges indicate their preferred approach. The lay judge from Athens offers an example of a colleague who lacks awareness of, and compassion towards, asylum seekers and the danger to which some are exposed. She attributes this divergence in perspectives to different previous employment statuses among judges, characterising those who had not previously worked for NGOs as ‘stricter’ and ‘more conservative’, lacking ‘sensitivity and understanding’. The same theme is evident in the words of the judge from Berlin. By sharing that he was among those judges more inclined to believe applicant claims, and by criticising other judges’ lack of empathy and cultural sensitivity, he signals that, for him, a more compassionate approach is more appropriate. As it appears here, different degrees of empathy and compassion towards asylum seekers during the asylum determination process are likely to lead to different decision outcomes.

The paragraphs above underscore the importance of identities in asylum determination and illustrate the recurring tension between role and person identities. While the ‘professional’ must be unquestionably present, there are differing views as to the degree to which the ‘person’ identity should be allowed to take part in the decision-making process. Similar tensions have been identified in previous studies, where the process of asylum decision-making is framed as a boundary work ‘between emotions and materialities’ (Schneider, 2019), influencing whether a judge, having made their decision, can subsequently ‘sleep at night’ (Affolter, 2021a).

The role resolution

As illustrated in Figure 1, when the professional role identity ‘wins over’ and ‘silences’ the person identity, the asylum decision will be based only on how the supporting ‘evidence’ meets the policy requirements. Even if they have a strong ‘inner belief’ (Kobelinsky, 2019) about how a case should be decided, or regard a specific policy as nonsensical, a judge’s decision will be based on the tangible pieces of information at hand that can justify it on paper.

I am trying to back up my decisions as well as possible, based on international sources, regardless of whether I agree or not with the decision I make. I mean, often there might be an internal conflict … Because you have to base a lot on the credibility of the applicant, your personal view may often be different from what is written on paper, or what was said [during the interview]. But you cannot exactly justify this discrepancy. It’s just that your instinct is telling you, “What they’re telling me has been well-learned, but it is not their life”. […] Often, you might be in that grey zone, where they tell a story that you cannot refute. So, there, I try to justify my decision based on the information and not put my own opinion first. (Lay Judge 2, Athens)

You have a child that came over and then there is one parent who comes over as well … And, you can decide, either you give the parent family protection, or you give them their own refugee status […] But, the outcome is super different. If you do give them family protection they can’t … if they have another child, they can’t really ask for this child to come over to Germany because they only got it for their first child […]. So, I always used to give them their own protection [status] […] Now, they said, in our [organisational] policy, we do have to give family protection. So, I still don’t agree. I still think it’s a stupid rule but, in the future, I will probably give family protection because they tell me this is how we do it [here] … Even though I personally think this is super stupid, that’s what policies are there for. (Lay Judge 20, Berlin)

As the first judge’s ‘instinct’ pushes her in a different direction from that which the evidence points towards, the resolution of this tension comes through the path of an ‘evidence’-based decision. This is so even though she does ‘not agree’ with the very decision she makes. Similarly, the second quote discusses a case where policies appear ‘super stupid’ in the eyes of the participant. Again, the asylum decision is based on what an asylum judge must do, as opposed to what they personally think is fair and appropriate.

Based on the role-resolution of the ‘role-person’ identity conflict, an asylum decision will be positive or negative, depending on the direction towards which the ‘evidence’ leans. This role-informed decision-making strategy echoes previously observed implementation approaches where street-level bureaucrats make an effort to distance and detach themselves from their clients and from the discretionary decisions they make about them (e.g. Affolter, 2021a; Schneider, 2019). However, it runs contrary to those who suggest that policy implementers tend to prioritise their own policy preferences (e.g. Bell et al., 2021; Spirig, 2023).

The person resolution

The person resolution to the role-person identity conflict is where the personal stance of asylum judges is the one that ultimately guides their decision. This is not to say that the policies and regulations in place are being ignored or bypassed. After all, with every asylum decision, the asylum judges must demonstrate on paper that they do implement policy. It is important to remember, however, that it is within their discretion to back their decisions based on the specific articles or directives they deem most fitting. In this regard, the door is open for them, should they so wish, to cherry pick the specific legal provisions they go by, according to those which tally most with their personal values, norms and beliefs.

As already noted, some judges take a more compassionate stance towards asylum seekers than others, whatever the reason. In practice, therefore, if their person identity is more salient than their role identity, their asylum decisions are more likely to benefit claimants. The ‘person resolution’ to the person-role identity conflict is illustrated below in the words of a lay judge from Berlin.

Every time, when I am unsure, I really try to tell, if I am not sure that the risk of them being found and [persecuted] is higher than, you know, not. So, I just go with granting refugee protection […] So far, I’ve never been against the guidelines. What I have done is, if the guidelines keep it kind of open, then I try to interpret the guidelines differently … If the guidelines say, “homosexuality in Ghana is not a problem”, I never write a decision saying that homosexuality is a problem in Ghana. That’s clear and I can’t write anything against it. I know it would never go through, there is no point. But if they are saying “some situations can be so bad for women that they cannot return by themselves”, then I can make a case about how it is really, really bad for that woman. (Lay Judge 16, Berlin)

This segment indicates this judge’s positive predisposition towards asylum seekers. When uncertain, she leans towards granting them refugee protection. Accordingly, when presenting her decision in writing, she uses all the room for discretion the existing guideless allow to support a positive decision. As mentioned earlier, while the guidelines themselves are fixed, the extent to which a specific grey area case may fit within the given guidelines can be manipulated.

A rather similar example, from an Athenian lay judge, is quoted below.

I had a family of Kurds from Iraqi Kurdistan, where there is an ambiguity as to whether you should give subsidiary protection or not … They had a little kid, aged two and a half, with leukaemia […] the kid’s hair had fallen out, a very beautiful little child. And these people said it clearly: “We came here to save our child. We could not find medicine; we could not find hospitals.” This is not a reason for refugee protection. But, on the other hand, we have the bad luck of living in a country that has no [proper] migration policy […]. There is a weight on us that is unfair, and has nothing to do with our training, either. So, I took this case, I really stretched it, and I gave them subsidiary protection. […] Because [otherwise], the second degree would have to reject them [again] and refer them to the humanitarian procedure, which means four years of waiting until you get a decision. Who out of those in need of humanitarian protection can wait for four years? Half of them are going to die! So, you are forced into a situation that causes you, you know, internal … anguish. To find a solution. (Lay Judge 7, Athens)

This quote vividly illustrates another person-based resolution example, where the person identity is the one that primarily influences the judge’s decision. The lay judge here ‘stretches’ the case legally, so as to provide protection for a child in need. Even though she acknowledges that this claim does not qualify for refugee or subsidiary protection, she interprets the available evidence in accordance with her own sense of justice and labels the case worthy of subsidiary protection. She therefore goes the ‘extra mile’ (Belabas & Gerrits, 2017) and ‘bends the rules’ (James & Julian, 2021), using her professional discretion in line with her own understanding of what a fair and just decision entails (see also Dahlvik, 2017). In these last two examples, we see the ‘person’ identity guiding the judges’ decisions, echoing previous studies that have highlighted the crucial role of the personal convictions, feelings, ethics and ethos of judges in the asylum determination process (Affolter, 2021a; Kobelinsky, 2019; Miaz, 2017).

Concluding discussion

This article has sought to shed light on the decision-making mechanisms of asylum judges, focusing on lay and administrative judges in Athens and Berlin during and soon after the so-called refugee crisis era of 2015-2017. Existing research on asylum determination has addressed the moral dilemmas judges encounter (Shiff, 2021), and the importance of their identities in decision outcomes (Ramji-Nogales et al., 2007). This study has built on this research and the street-level bureaucracy literature on asylum adjudication (Affolter, 2021a, 2021b; Dahlvik, 2017; Dallara & Lacchei, 2021; Miaz, 2017), as well as on discretionary decisions in the face of moral dilemmas (Tummers et al., 2012; Vink et al., 2015). Also drawing from original interview data with judges who make first- and second-instance asylum decisions, this study has examined how judges decide on ‘grey area’ asylum cases, and how their decision-making is informed by their individual identities.

This study suggests that, under conditions of considerable uncertainty, the inherent moral dilemma of asylum decisions is pronounced, often leading to an identity conflict between the judges’ role identity as policy implementers and their person identity as compassionate (or less so) individuals. This observation compliments existing research which links moral dilemmas with identity conflicts in street-level discretionary decisions (e,g. Jensen & Pedersen, 2023; Maynard-Moody & Musheno, 2012). It does so, first, by conceptualising the identities of judges identities as understood by policy implementers themselves and, second, by looking at identity tensions that go beyond their professional role identity.

To illustrate how this role-person identity conflict shapes asylum determination, this article has delineated the different steps judges follow before reaching a positive or negative asylum decision (Figure 1). In ‘clear-cut’ asylum cases, the process is relatively simple: if the case neatly fits the legal refugee definition it will be a positive decision and, if not, a negative one. In grey area cases, however, when there are no obvious right and wrong decisions, there are two possible scenarios. If their role identity dominates, their decision will be based on the available ‘evidence’ provided, however limited this may be. Alternatively, if their person identity dominates, their decision will be more heavily influenced by their degree of compassion towards claimants. The preceding empirical section has demonstrated these assertions through specific examples from the participants’ accounts. It has discussed asylum cases that judges perceive as challenging, and has demonstrated what role and person identity-informed asylum decisions may mean in practice.

Another, less anticipated, finding of this study is the presence of positive discretion by asylum judges. Much of the relevant literature takes the view that asylum judges are biased against asylum claimants, which makes their use of professional discretion mostly negative, meaning contrary to the interests of claimants (e.g. Giametta, 2020; Johannesson, 2018; Magalhães, 2016, 2018). However, this study has shown that some judges are also positively predisposed towards asylum claimants and keen to make discretionary decisions to help them. Even those who prioritise their professional identity and tend to make ‘evidence’-based decisions often have some degree of compassion towards asylum seekers, which they try to suppress (see also Kobelinsky, 2019). Although this observation is not at the centre of this research’s focus and findings, it does point towards the need for further acknowledgement of positive instances of street-level discretion (see also Belabas & Gerrits, 2017; Glyniadaki, 2022; James & Julian, 2021).

It is important here to address a limitation of this study. As the data analysed consists of individual qualitative interviews, the judges’ accounts cannot be neutral and unbiased. The participants of this study, like humans in general, are likely to want to be seen in a positive light, and therefore share only information that portrays them as good professionals and compassionate people. Nevertheless, their experiences and perceptions of the intricacies of the asylum determination still offer a meaningful and valuable contribution to our understanding of this process.

The overall approach of this study, and especially the emphasis on the daily challenges judges face, partly serves to ‘humanise’ asylum judges, which is among the main premises of street-level bureaucracy literature (Lipsky, 1980). What this article has perhaps more explicitly brought to light is the extent to which the bureaucrat is intertwined with the human. The recurring tension between professional and personal sides of asylum judges indicates that the person cannot be entirely disentangled from the professional. Indeed, such a complete separation would not be desirable, even if it meant more consistent asylum statistics. Future research could further explore the identity tensions judges encounter during asylum determination, tensions which may go beyond the conflicting expectations of their role prescriptions, and which may also influence asylum decision outcomes.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Public agency where first instance decisions take place.

Appendix

Table A1. 1st Instance recognition rates in Germany and Greece for applicants from Syria, Afghanistan and Iraq.

	Syria			Afghanistan			Iraq
	2015	2016	2017	2015	2016	2017	2015	2016	2017
Germany	97.7%	99.3%	95.2%	72.8%	60.1%	46.6%	98.3%	76.7%	63%
Greece	99.6%	55.3%	83.5%	55.2%	46.6%	75.6%	64.7%	63.9%	63%

Source: Eurostat.

Table A2. List of role conflicts among professionals at the street level.

List of Conflicts	Examples
The ‘policy–professional’ conflict	Medical doctor having to perform an abortion to save a mother’s life, but abortion being illegal.
The ‘policy–client’ conflict	Police officers having to implement ‘zero tolerance’ policies without accounting for the clients’ circumstances.
The ‘organisational–professional’ conflict	Doctors prioritising quality of service but managers insisting they have to meet a fixed number of examinations per week.
The ‘professional–client’ conflict	Doctor looking after a patient who requests euthanasia.

Note: Table compiled by the author, based on Tummers et al. (2012) and Vink et al. (2015).

Table A3. Defining features of person, role and social identities

Features	Person Identity	Role Identity
Bases	Individual self-concept	Expectations tied to social positions
Definitions	Meanings that define a person as a unique individual	Meanings tied to a role
Behaviour	Independent of others	Complementary to others
Self-Reference	Me	Me as role

Source: Burke and Stets (2009, p. 129), shortened by author.

Table A4. Sources/types of pressure as described by participants.

Source /Type of Pressure		(Lay) Judges
		Germany		Greece
		1st inst.	2nd inst.	1st inst.	2nd inst.
1	Rapid changes in policy framework	√		√
2	Unclear direction of overall legal framework	√		√
3	Low implementability (time frame/availability of relevant professionals etc.)	√		√
Societal
4	Media pressure	√		√	√
5	Public/social environment pressure	√		√	√
Organisational
6	Limited training	√		√	√
7	Heavy workload/time pressure to perform task	√	√	√	√
8	Weak administrative capacity			√	√
9	Principal–agent pressure	√			√
10	Precarious working conditions			√	√
Work-Related
11	Conflicting expectations (i.e., support for law and humanitarian values)
		√	√	√	√
12	Emotionally heavy work	√	√	√	√
13	Lack of evidence to support claim (i.e., age, vulnerability etc.)
		√	√	√	√
14	Communication compromised by interpreter	√	√	√	√
Claimant-Related
15	Cultural/language barriers	√	√	√	√
16	Communication difficulties due to trauma	√	√	√	√
17	Traumatised applicants/not mentally well	√	√	√	√
18	Illiterate applicants/poor story narrators	√	√	√	√
19	Changes in claimants’ country of origin situation	√	√	√	√

Note: Table constructed by the author, based on the interviews with participants.