Refugee roulette in the Common European Asylum System : subjectivity in Greek refugee status determination

(1)

Refugee Roulette in the Common

European Asylum System: Subjectivity in

Greek Refugee Status Determination

Author: Linda Mullan

Supervisor: Polly Pallister-Wilkins

Second Reader: Darshan Vigneswaran

Borders, Geopolitics and Conflict

International Relations Master’s Thesis

26

th

June 2015

U

NIVERSITY OF

A

MSTERDAM

(2)

This thesis explores the subjectivity and agency of refugee status decision-makers in the Greek asylum system, and considers the implications this has for a Common European Asylum System. It uses Lipsky’s theory of street-level bureaucrats and Bourdieu’s theories of field, habitus, and dispositions as the main theoretical

framework for this examination. This framework is then applied to data gathered from semi-structured interviews and discourse analysis.

The analysis determines that refugee status decision-makers are street-level

bureaucrats who exercise agency in a variety of ways that are shaped by the structure of their field. Furthermore, it finds that there is considerable subjectivity in the refugee status decision-making process, which is incorporated in the language of the very international legal instruments that make up the existing Common European Asylum System. This means that there is a substantial element of chance in the success or failure of an asylum application based on the dispositions of the

interviewers involved. These findings have significant implications for a Common European Asylum System, as they mean that homogenous, universal standards for refugee status determination are unattainable.

(4)

1. Introduction

Questions of identity, of who we are and where we belong, are an intrinsic part of the human condition. The answers to these questions are not just matters for the

individual, but for the state as well. Nowhere is this more evident than in the field of official migration regimes. Through these regimes, founded on principles of inclusion and exclusion, states reconstitute the identities of individuals and groups and ascribe degrees of belonging to them. The form these regimes should take, and the criteria they adopt to filter the included from the excluded, is a particularly salient matter in the European Union, where issues of European identity arise alongside those of national identity (Smith, 1992).

Consequently, there has been a drive for migration policy harmonization in the European Union, which has gained momentum since the implementation of the open borders system within the Schengen Area (Givens & Luedtke, 2003). This has meant that states whose geographical borders constitute both national and EU borders are subject to scrutiny from other Member States. Under the Dublin system, these states have specific responsibilities regarding the management of migrants. The Dublin requirement that all asylum seekers register in their European country of entry particularly affects border states, and they are also expected to maintain certain reception standards under the EU Reception Conditions Directive (Hatton, 2005). The Dublin system has also led to increased investment in border controls and policing, which has led to the use of the label ‘Fortress Europe’ by both proponents and detractors of restricted access to entry (Geddes, 2000). This increasingly restrictive stance has limited access to residence and citizenship throughout Europe, and for many would-be immigrants who succeed in reaching European territory, the asylum process is one of few routes to legal residence open to them (Council on Foreign Relations, 2015).

This situation is magnified in Greece. Its role as an EU border state means that it carries all of the responsibilities described above. However, in a country where migration law is dominated by a strong jus sanguinis principle, where second generation immigrants born and raised in the country were, until very recently,

ineligible for citizenship unless they could meet an almost insurmountable standard of proof, the legal admission of non-nationals is a contentious issue in public debate

(5)

(Triandafyllidou, 2009). Furthermore, the political and economic fallout of the 2009 Eurozone crisis has led many in Greece, Europe and elsewhere, to re-examine Greece’s identity as a European state (Karamouzi, 2014), which has the potential to impact its level of cooperation with EU directives.

These issues illustrate the challenges facing those who advocate for the EU goal of establishing a Common European Asylum System. Prior to 2011, Greek law did not officially recognise asylum status, but rather practiced a “politics of invisibility”, in that policy and practice towards asylum seekers was informal, and had been

incorporated into the general asylum system with little consideration for the Geneva Convention Relating to the Status of Refugees (Rozakou, 2012, p.563;

Triandafyllidou, 2009). The 2011 European Court of Human Rights (ECtHR) judgement, M.S.S. v Belgium and Greece, thrust the treatment of asylum seekers in Greece into the international spotlight and determined that Belgium and other EU Member States could not return asylum seekers to Greece due to the reception conditions and history of mistreatment there. This judgement constituted a massive political and social challenge to Greece. The 2012 revised Greek Action Plan for Migration and Asylum Management attempted to address the ECtHR’s concerns, and provided for a new asylum system, which came into effect in 2013.

This thesis is concerned with the Greek asylum system in the wake of the 2011 ECtHR judgement, specifically the period from June 2013 to the present day during which the new asylum system has been in place. It focuses on the government employees who conduct the asylum interviews and decide on applicants’ status. Taking a bottom-up approach to policy analysis and practice, I examine the standards that workers in the decision-making process are supposed to adhere to, how they navigate a role that can encompass contradictory aspects, and how they make decisions on asylum applications. This paper will draw upon Lipsky’s theory of street-level bureaucrats, which holds that public servants who directly interact with the public play a fundamental role in the formation of public policy (Lipsky, 2010). It will use this theory to answer the primary research question: To what extent are interviewers in the Greek asylum system ‘street-level bureaucrats’? It will also draw upon Bourdieu and Giddens’ theories of structure and agency to answer the secondary research question: If this is the case, how do they exercise agency in refugee status

(6)

This thesis argues that the subjective nature of the application assessment procedure and the exercise of agency within it enables the dispositions of individual decision-makers to shape and define the application of the law in a way that profoundly influences the lives of asylum seekers. It further argues that this subjectivity, and the variation in decision-making that it causes, presents a significant barrier to the implementation of a Common European Asylum System.

Before beginning the analysis, it is necessary to clarify a few key terms. Firstly, the terms refugee and asylum seeker must be defined. The UNHCR declares that “a person…does not become a refugee because of recognition, but is recognised because he is a refugee” (UNHCR, 2011, p.9). However, as this research focuses on refugee status decision-makers, for the purposes of this thesis a refugee will be defined as someone who has been recognised as such and granted international protection. An asylum seeker will be defined as someone who declares themselves to be a refugee but has not yet received official recognition from a state authority (unhcr.org). Secondly, the term ‘subjectivity’, a central concept in this thesis, also needs

clarification. As a philosophical concept it can be traced back to Descartes, and has been a source of considerable debate (Hall, 2004). For the purposes of this research it will be defined simply as “the limited, error-prone perspective of the individual” (ibid., p.3).

This thesis is structured as follows: Chapter 2 elaborates on the legal evolution of the Common European Asylum System (CEAS), and its incorporation into the Greek asylum system. Chapter 3 lays out my literature review, which discusses key debates on CEAS and practices of inclusion and exclusion. Chapter 4 sets out the theoretical framework, in which the relevant theory is discussed, with particular focus on the theories of Lipsky and Bourdieu. Chapter 5 describes the methodology adopted, justifies why an institutional ethnographic framework was selected and discusses its strengths and weaknesses. Chapter 6 presents and discusses my results and their significance, and Chapter 7 details my conclusions.

(7)

2. Background of the European and Greek Asylum Systems

Informal practices of asylum can be found throughout history (Fassin, 2013). But, as Arendt famously claimed, “The trouble arose when it appeared that the new

categories of persecuted were far too numerous to be handled by an unofficial practice destined for exceptional cases” (Arendt, 1951, p.294). In order to understand the context that refugee status decision-makers operate within, this chapter will provide a brief overview of the development of the official asylum system in Europe.

2.1 Overview of the legal evolution of the European Asylum System

The modern concept of international political asylum was established in Article 14 of the Universal Declaration of Human Rights, and further defined and expanded upon in the UN 1951 Convention Relating to the Status of Refugees (hereinafter the Refugee Convention) and the 1969 Protocol Relating to the Status of Refugees (hereinafter the Refugee Protocol). In the wake of WWII it was designed to provide a universal form of protection against persecution and displacement (1951 Convention Relating to the Status of Refugees, Article 1). Furthermore, the ideological conflict of the Cold War, combined with post-WWII labour shortages, meant that the relatively few asylum seekers from Communist countries in the wake of the Convention were positively received in the West (Czaika, 2009; Jordan, Stråth & Triandafyllidou, 2003).

However, the definition of a refugee in Article 1 of the Convention is far from exhaustive, and leaves considerable scope for interpretation. It defines a refugee as a person who "owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality, and is unable to, or owing to such fear, is unwilling to avail himself of the protection of that country." (UN 1951 Convention Relating to the Status of Refugees, Article 1.A.2). However, nowhere in the

Convention does it define the terms “well-founded fear” or “persecution”, nor does it clarify how these criteria are to be assessed. Consequently, there has been

considerable scope for variation in national law. In response to this variation, and the perceived threat it posed to the Schengen Agreement, EU Member States developed the Dublin Convention in 1990, to establish a more uniform evaluation of asylum

(8)

claims throughout the EU, as well as to determine state responsibility to individuals. Yet, as argued by Sandra Lavenex:

“Refugee policies are a paradigmatic example of such a policy field where supranational and transnational actors disagree not only as to the scope of Europeanization, but also about the factual representation of the refugee problem and the normative orientation of a common policy. With the current ‘crisis’ of the post-war regime, this disagreement arises over precisely who constitutes a refugee deserving protection, what kind of protection should be offered, and by whom.” (Lavenex, 2001, p.855).

In spite of these challenges, steps to reconcile these differences were taken. The need for greater harmonization was acknowledged in the 1997 Treaty of Amsterdam and reaffirmed during the 1999 EU Council Summit in Tampere, in which negotiations started on the creation of a Common European Asylum System (CEAS) (European Commission, 2014). The Tampere Conclusions included the short-term goals of establishing: (1) a mechanism for determining the state responsible for an asylum application; (2) common asylum procedures; (3) common, minimum reception conditions; and (4) an approximation of rules for the recognition and content of refugee status (Tampere Presidency Conclusions, Section A, Subsection II). The long-term goals were the establishment of a common European asylum procedure and a uniform status for those receiving international protection (ibid.).

In 2003, as an attempt to meet the first of these short term goals, the Dublin II Regulation was adopted, which was designed to address the perceived gaps in the original Dublin Convention. It included revised criteria to establish the state

responsible for reviewing an asylum application, and aimed to end ‘asylum shopping’ and ‘refugees in orbit’ (Hatton, 2005). It cemented the enforcement of the ‘first host country’ principle (i.e. the first signatory country which an asylum seeker enters has exclusive responsibility for processing their claim) through the implementation of the fingerprint database, European Dactyloscopy (EURODAC). It also extended this principle of first contact to include ‘safe third countries’ outside of the European Union (Lavenex, 2001; Hatton, 2005). In the wake of Tampere, the Reception Conditions Directive (2003), the Refugee Qualifications Directive (2004), and the Asylum Procedures Directive (2005) were also adopted as part of Phase I of CEAS.

(9)

The Reception Conditions Directive laid down minimum standards for the reception of asylum seekers in Member States; the Refugee Qualifications Directive sets out standards as to who qualifies as a beneficiary of international protection; and the Asylum Procedures Directive aimed to establish common standards of safeguards and guarantees to access a fair and efficient asylum procedure (European Commission Department of Migration and Home Affairs, 2015).

In 2004 the Hague Programme was adopted by the European Council to evaluate these Phase I legal instruments, and help to implement Phase II. However, the evaluation of Phase I and recommendations for amendments to the instruments in place took longer than anticipated to complete (Allard, 2010). Consequently, the European Council adopted the Stockholm Programme in 2009, which aimed to fully implement CEAS by 2012 (Presidency of the European Union, 2009). Revisions of the legal instruments listed in the prior paragraph were implemented, and Dublin III was approved in 2013, as an attempted remedy for the perceived ‘burden shifting’ rather than ‘burden sharing’ effects of Dublin II (Smythies & Ramazzotti, 2013). Nonetheless, the timeframe for the Programme again proved to be unrealistic. The Stockholm Programme ended in 2014, with the ultimate goal of a Common European Asylum System still unachieved (UNHCR, 2014).

The collective action issues that have played a role in these delays (Thielemann, 2005) manifest in a number of ways. Differing interpretations of the existing regulations by states could form a significant barrier to the achievement of CEAS. The issues identified by Lavenex (2001), of ascertaining who offers and receives protection, and the form that protection takes, are still far from resolved. The

assumption that all Member States offer an equal level of protection to asylum seekers was categorically rejected by the European Court of Human Rights (ECtHR) in the landmark case M.S.S. v Belgium and Greece in 2011. The court found that by transferring M.S.S. to Greece, the Belgian authorities had violated his rights under Article Three of the European Convention on Human Rights (ECHR), which prohibits subjecting people to torture, or to inhuman or degrading treatment. Belgium was condemned not only for exposing M.S.S. to the risk of indirect refoulement, but also direct refoulement (ECtHR (GC), M.S.S. v. Belgium and Greece, Appl. No. 30696/09 para. 366-367)

(10)

The M.S.S. judgement’s clear assertion of the variation in Member States’ protection standards highlights a central problem of one of the key pieces of CEAS regulation: the Refugee Qualifications Directive. The Qualifications Directive essentially restates the definition of a refugee in the 1951 Convention, and provides some of the

interpretive guidance that was lacking in the original document (Storey, 2008). It also created the category of subsidiary protection for individuals who do not qualify for refugee status, but are deemed to be at risk of serious harm if returned to their country of origin (Allard, 2010). However, the Directive, with its emphasis on minimum standards, avoids laying down true common standards. Furthermore, EU Member States have failed to uniformly implement the Directive’s standards into their national law (Allard, 2010).

This failure is compounded by the fact that for those states who have fully or partially transposed the Directive into their national law, mistranslations and the use of

approximate or alternative language can muddy the interpretation and effect of the law, and ensures that “the common text does not remain common” (ibid., p.317). Indeed, the UNHCR has found that “the chances of an individual asylum-seeker to find protection in the EU can vary nearly seventy-fold, depending on where he or she applies” (Feller, 2008, p.217). For example, one category for subsidiary protection is individuals under threat of serious harm due to indiscriminate violence in a situation of international or internal armed conflict (Qualifications Directive, 2004, Art 15(c)). However, what constitutes either international or internal armed conflict is not precisely defined. Consequently, UNHCR research revealed that in 2007, France considered the conflict in Iraq to be an internal armed conflict, Sweden did not, and Germany’s jurisdictions were divided (UNHCR, 2007).

2.2 CEAS and the Greek Asylum System

In spite of these challenges and contradictions, the influence of the European efforts to build a more homogenized asylum process can be clearly seen in Greek national law. After the Dublin Convention came into force in 1997, and perhaps as a response to the sudden spike in immigration that resulted from the instability in the Balkans in

(11)

the 1990s, Greece began its first regularisation programme in 1998. This was the first time access to citizenship had been granted to non-Greeks since the inception of the modern Greek state in 1829 (Triandafyllidou, 2009; Karyotis, 2012). However, the subsequent immigration system that developed under the Ministry of Public Order, administered by the police, was riddled with human rights abuses, and was virtually impossible to access (Human Rights Watch, 2012; UNHCR, 2010; Cabot, 2012). There was a distinct lack of political will to substantively reform the immigration and asylum system between 1998 and 2011, as migrants were viewed by the political elites and many in the general public as more of a convenient flexible labour force rather than permanent additions to Greece who were entitled to equal rights and citizenship (Triandafyllidou, 2009). Indeed, during this time the political rhetoric against migrants grew, and Golden Dawn, the far-right anti-immigration party, rose to become the third largest political party in Greece in 2012 (Ellinas, 2013; Rodgers, 2012). It is against this backdrop that the new Greek asylum system came into being in 2013.

(12)

3. Debate on the Common European Asylum System:Literature Review

These challenges at a national level are indicative of the complex nature of the development and interpretation of national and transnational law and policy. This chapter will examine some of the official texts, literature and key debates surrounding the development of CEAS. This includes a brief overview of theories on the

feasibility of CEAS, the compatibility of the EU principles of freedom and security, and the need for ambiguity in policy formation. It then positions this thesis in the debate.

The central premise of CEAS, and indeed that of the European Union itself, is one of inclusion and exclusion. Article 2 of the Treaty of the European Union states that the maintenance and development of the EU would be as “an area of freedom, security and justice”. However two of those criteria, freedom and security, have proven somewhat difficult to reconcile. The Tampere Conclusions describe the challenge of the Treaty of Amsterdam as “to ensure that freedom, which includes the right to move freely throughout the Union, can be enjoyed in conditions of security and justice accessible to all” (Tampere Presidency Conclusions, Preamble). This implies a perception of security that differentiates between the ‘safe inside’ and the ‘unsafe outside’ (Monar, 2000, p.5). As a result, “people from the ‘outside’ which actually or potentially threaten the ‘safe inside’, must be kept outside or brought under

appropriate control” (ibid.). This is the underlying premise of the ‘Fortress Europe’ mentality (Geddes, 2000).

To counter this however, the Tampere Conclusions also state that this freedom of movement should “not be regarded as the exclusive preserve of the Union’s own citizens” and that “it would be in contradiction with Europe’s traditions to deny such freedom to those whose circumstances lead them to justifiably seek access to our territory” (Tampere Conclusions, Preamble). It further affirms that the aim is “an open and secure European Union” (ibid.) (emphasis added). This aim is far from straightforward. The Tampere Conclusions do not specify under which

‘circumstances’ someone may seek access, nor what makes such an action

‘justifiable’. Such ambiguity is sometimes “a necessary recourse to accommodate conflicting interests in order to devise draftable legislation” (Yanow, 1990, p.220).

(13)

Nonetheless, it is inevitable that the outcome of this ambiguity is differing

interpretations and national standards, something which fundamentally conflicts with the purported goal of harmonization. This conflict is further sustained by the

undeniable truth that the EU is made up of states with conflicting interests. This is evident in the fact that there is disagreement among Member States as to the location of the very borders of the EU and principle of freedom of movement, as evidenced by the UK and Ireland’s exemption from many provisions of the Schengen Agreement (Monar, 2000).

Consequently, the feasibility of CEAS has been called into question. The substance of the asylum policies, as laid down by the Treaty of Amsterdam and followed by the Hague and Stockholm Programmes, has been subject to considerable scrutiny.

Scholars have termed the process, with its emphasis on minimum standards, as a ‘race to the bottom’ (Hatton, 2005) and a ‘common market of deflection’ (Noll, 2000). Many of the measures taken, including the limitations on the access to domestic asylum procedures and the Dublin Regulation’s ‘safe third country’ rule are seen as violations of the spirit of the 1951 Refugee Convention (Lavenex, 2001). The emphasis on security, and the integration of migration policy into a security

framework has been termed ‘securitization’ (Huysmans, 2000). This securitization of migration in Europe has been attributed by many to the spike in immigration to Western Europe after the fall of the Iron Curtain and the crises in the Balkans, as well as those in the Middle East and Africa (Czaika, 2009; Hatton, 2005). Other authors however, have deemed the so-called ‘immigration crisis’ a political and social construct, designed to create a ‘border spectacle’ which reinforces perceptions of the racialised illegal migrant, and helps to delineate the boundaries between the ‘safe’ EU and ‘unsafe’ outside (De Genova, 2002; Huysmans, 2000).

Additionally, the securitization of migration has highlighted the conflicting nature of discussions on internal state security and human rights issues. While the original 1951 Refugee Convention emphasised asylum as a universal human right, EU asylum policies since the 1990s, both at a national and transnational level, have been framed in terms of formulating ‘compensatory measures’ to safeguard internal security (Geddes, 2000; Bigo, 1996). Indeed, a number of studies have shown that the

(14)

measures implemented from the 1990 Dublin Convention emphasise restrictive measures rather than an accessible system (Hatton, 2005; Lavenex, 2001).

The rate of variation that has resulted from this emphasis on minimum standards, when combined with the Hierarchy of Criteria of the Dublin Regulation (Dublin III, Articles 7-15) which removes agency from asylum seekers as to their country of application, calls into question the fairness that CEAS claims to aim for. Applications that may be accepted in one state can be rejected in another. As a result, the current open-border system that aims to prevent illegitimate claims through ‘asylum

shopping’ fails to adequately distinguish between illegitimate and legitimate claims on an EU-wide basis (Allard, 2010; Smythies & Ramazzotti, 2013). This poses a challenge to the view of the EU as the paradigm of the network state, one defined as a space of flows rather than territorial nation-states, as conceptualised by Castells (2000). Asylum, and immigration as a whole, is linked to citizenship, one of the cornerstones of the nation-state. To cede power on this issue is to redefine political identities (Jordan et al, 2010). It is perhaps to be expected then that national

interpretations of these international standards differ so widely, and that states are so reluctant to cede full control of the matter. Both national and EU identities position themselves in contrast to an outside group, so where there are differing perceptions of who constitutes this outgroup, harmonization is difficult, and potentially impossible (ibid.).

However, despite the critique that many of the above scholars provide on the feasibility of harmonisation in migration policy, few of them question one of the underlying motivations behind it, that of ‘burden-sharing’. Lavenex, although she references the concept, does not analyse it in any way. Noll, although a critic of many of the CEAS instruments, praises them as “a dam-breaker construction with regard to the many dilemmas of fiscal burden-sharing” (Noll, 2003, p.245). Hatton supports this evaluation. He believes that the harmonisation policies have been “too tough” in their requirements on Member States, but that “there is an important role for enhanced burden-sharing arrangements” (Hatton, 2005, p.106). The same can be said of Allard, who refers to the need to “ease the burden on states of harbour” (Allard, 2010, p.297). Allard and Noll also support the claim that national asylum policies, not EU

(15)

States. However Thielemann (2004), another advocate of burden-sharing, argues that European coordination efforts have served to deflect some asylum seekers to less developed countries, but have done little to address the existing distribution of asylum seekers in the EU. This emphasis on the implementation and impact of the policies is a common thread throughout the literature. Few of the authors examined here analyse the language used in law and policy-making, but rather echo it uncritically. This thesis seeks to fill that gap.

This thesis provides support for arguments challenging the feasibility of the true harmonisation of national asylum policies throughout the EU. It also supports theories that hold that the securitization of migration policy can act to limit the principle of freedom of movement, and its extension beyond EU citizens. Additionally, the findings of this research support Yanow’s (1990) argument that while the ambiguity of official language may be necessary to ensure agreement to it, it leaves considerable scope for interpretation. However, unlike literature discussed in this chapter, this thesis does not explore the impact that harmonisation has on burden-sharing, but rather seeks to problematize the language used in this discourse and look at the impact that this language may have on harmonization.

(16)

4. Street-level Bureaucrats and the Interaction between Structure and Agency: Theoretical Framework

CEAS, and the category of asylum as a whole, reconstitutes applicants as either ‘worthy’ or ‘unworthy’ guests; refugees or unwanted economic migrants (Butler, 2004; Rozakou, 2012). This chapter will set out the rationale for focusing on role of the individual decision-makers in this system at the micro level of implementation, and elaborate on the relevant theory that will be used for the examination of this role. It will focus particularly on Lipsky’s theory of street-level bureaucracy and how it is identified, and how structure and agency interact in this context.

Bigo describes three social universes of border control (Bigo, 2014). The first two he identifies are predominately related to border guards. He describes the first universe as militarized, preoccupied with containment and patrols. The second is more concerned with internal security, ‘filtering’ and the management of flows of people. The third is less of a physical and more of a virtual universe, inhabited by data analysts who filter migrants as ‘big data’ rather than physical bodies. It is in the second universe that the central premise of this thesis lies. While Bigo focused on the role of border guards, his second universe, engaged with the classification and

filtering of individuals, can also be extended to the asylum system and the bureaucrats who operate within it. This is because through the filtering process of the asylum system, the border between those included and excluded by the state becomes manifest.

This classification and filtering through the use of the Refugee Convention and Protocol, far from being the implementation of an invariable set of rules, is a work of continuous interpretation (Fassin, 2013). This interpretation is not just carried out by national law and policy-makers, but also by the individual bureaucrats who assess asylum claims and decide on the applicants’ status, and can be considerably

influenced by individual perceptions and beliefs. A US study of asylum adjudication found that “when an asylum seeker stands before an official or court who will decide whether she will be deported or may remain in the United States, the result may be determined as much or more by who that official is, or where the court is located, as it is by the facts and law of the case” (Ramji-Nogales, Schoenholtz & Schrag, 2007, p.302). They found a large degree of statistical variation in decisions made by

(17)

adjudicators. In one instance, a judge was 1820% more likely to grant an applicant for relief than another judge in the same courthouse. They also found that an asylum applicant assigned to a female judge had a 44% better chance of a favourable

judgement than an applicant assigned to a male judge (ibid.). This led the authors to coin the term ‘refugee roulette’ to describe the element of chance involved. These findings clearly demonstrate that the personal ideologies and dispositions of decision-makers have an undeniable impact on asylum status decisions.

Engel and Munger (2003) theorise that the law becomes active in peoples’ lives through both formal and informal discursive events. The formal processes, such as judicial procedures and asylum interviews, offer themselves up readily for

examination. The informal discussions, such as those between successful candidates and prospective applicants, between lawyers and their clients, or between interviewers and their colleagues, are, by the very fact of their informality, much less transparent, and their impact is difficult to decipher. These discussions may influence anything from the information deemed important to the language used in the official interview (Cabot, 2013). These informal discussions of law and protection occur at every level of the asylum process, and can affect the dispositions of all participants. The

assessment of asylum claims does not occur in a vacuum.

Cabot, in her study of NGO workers and asylum seekers, examined how the narratives given by asylum seekers were often influenced by the perceived

expectations of the listeners. She also explored how the lawyers that screened cases to determine applicants’ eligibility for representation used “feelings, impressions,

narrative cues and visual tropes” to make decisions (Cabot, 2013: 458). She found that employees “actively elicited and recrafted [asylum seekers] narratives before issuing a decision” (Cabot, 2013: 456). The legal criteria, as well as the design of legal and bureaucratic forms, significantly contributes to what is considered a ‘good’ or ‘plausible’ narrative (Riles, 2001; 2007). Both Cabot (2013) and Fassin (2013) identify truth, trust, and credibility as core issues for those assessing asylum claims. The performative aspect of the asylum seeker’s narrative can be as important to the success of their claim as the content, and the interpretation of this performance is highly dependent on the subjective perceptions of the interviewer (Fassin, 2013;

(18)

In light of the impact that the dispositions of individual decision-makers can have on the outcomes of asylum claims, and the fact that the success or failure of these claims can potentially be a matter of life or death, it is important to closely examine the agency of these governmental bureaucrats, and the structures in which they work. The role of asylum interviewers appears to fit rather neatly within Lipsky’s 1969 theory of street-level bureaucrats, i.e. “public service workers who interact with citizens in the course of their jobs, and who have substantial discretion in the execution of their work” (Lipsky, 1969, p.3). He theorises that “decisions of street-level bureaucrats, the routines they establish and the devices they invent to cope with uncertainties and work pressures, effectively become the public policies they carry out” (Lipsky, 1969, p.xii, emphasis in original). As front-line service providers, the individuals

conducting asylum interviews have considerable capacity to influence the

implementation of policies, and potentially subvert or direct it to reflect their own interests (Sabitier, 1986; Lipsky, 1969).

I will ascertain the extent to which their role is one of ‘street-level bureaucrats’ as conceptualised by Lipsky. This means ascertaining the extent to which the following conditions are present: Firstly, they are employees who are constantly called upon by the government to interact with citizens (or in this case, non-citizen members of the public) in the regular course of their jobs. Secondly, they have a significant level of independence in job decision-making. Thirdly, they potentially have an extensive impact on the lives of their clients. (Lipsky, 1969, p.3).

As well as fulfilling these requirements, Lipsky’s street-level bureaucrats must also work under the following conditions: Their available resources are inadequate (organizational and/or personal); their work proceeds in circumstances where there exists clear physical and/or psychological threat, and/or the bureaucrat’s authority is regularly challenged; and expectations about job performance are ambiguous and/or contradictory, and include unattainable idealized dimensions. (Lipsky, 1969, p.4). After establishing the extent to which members of the asylum interview board meet the criteria of street-level bureaucrat, it is necessary to establish how they operate within their role. This includes determining how they are socialized in the system after training, and if they develop mechanisms to make their task easier, in order to

(19)

alleviate the bureaucratic difficulties inherent in the working conditions above. If so, what are they? What are the trade-offs and consequences? Furthermore, do certain categories of applicants receive special treatment? Do they demonstrate agency, and if so, how? Through these questions and others like them, I hope to ascertain how street-level bureaucrats operate, and what dispositions they develop. In exploring the role of street-level bureaucrats, Satyamurti (1979) has done considerable research on the ‘cognitive dissonance’ that can affect employees when their conditions of work, and the responses of their clienteles, do not approximate to the ideals and values of their professional ethics, or to their training. Cuff (1980) found that immigration officers found it difficult to ‘pull off’ “a convincing version of his or her personal identity that wove together policy implementation, humanitarian ethics, technical proficiency and personal empathy” (Jordan et al, 2010, p.216). After all, as Palumbo and Calista argue: “Administrators, no less than legislators, seek power, esteem and monetary rewards that inevitably shape how (or which) goals are achieved” (Palumbo & Calista, 1990, p.7).

As discussed earlier, the asylum system closely correlates with Bigo’s second

universe of border control (Bigo, 2014). This universe consists of a social structure, or ‘field’ to use Bourdieu’s term, with its own governing rules, hierarchy and

participants. It is also constituted as a habitus, which is the way in which the objective structures of the field influence the subjective actions and thoughts of the agents within it, and are formed into a set of dispositions (Bourdieu, 1988). It is through this lens that I wish to examine the everyday practices and routines of bureaucrats

working within the asylum system, and particularly any gaps between the objective structures of the social world and the practices of individuals. If, as Fassin argues, “truth [in asylum applications] is what institutions decide to be true” (Fassin, 2013: 59), it is crucial to examine how the individuals employed by institutions operate and form their decisions.

The question of agency in street-level bureaucrats touches on a larger debate, one of structure versus agency. Bourdieu has been referred to by some as a structuration theorist (Jones & Karsten, 2008). However, this label comes not from Bourdieu’s own work, but from the influence his work had on Antony Giddens, who coined the term.

(20)

Giddens analyses the relationship between individuals and society, and proposes that structure and agency are a mutually constitutive duality (Giddens, 1984). Like Bourdieu, he holds that social structures (fields) are continuously being created through the flow of everyday social practice. He defines structure as “rules and resources, organised as properties of social systems” (Giddens, 1984: 33), which can involve both virtual internal structures and objective external structures (Stones, 2005). He shares Foucault’s view of power as relational, and asserts that individuals have the power to transform structures (Jones & Karsten, 2008). However, his view of agents as highly autonomous and voluntaristic differs from Bourdieu, who views agency as more heavily shaped by structure. For Giddens however, this weaker concept of agency amounts to a form of determinism. The agency versus structure debate is fiercely fought, and Giddens has numerous critics on this subject (Bhaskar, 1979; Callinicos, 1985; Layder, 1985). While an in-depth exploration of this debate is beyond the scope of this thesis, it will utilise the aforementioned aspects from the work of both theorists to add an additional, deeper dimension to Lipsky’s theory of street-level bureaucrats.

(21)

5. Methodology 5.1 Research Design

In order to answer the research questions, I used a qualitative research design. In this chapter I will discuss how I developed a design that draws upon institutional

ethnography and utilises semi-structured interviews and discourse analysis. Some aspects of grounded theory were also incorporated. I will justify my choices and detail the strengths and weaknesses of the methods used.

As stated in the introduction, this thesis seeks to take a bottom up approach to examine the internal environment and processes of the asylum interview procedure, and investigate the dispositions and agency of the employees. Consequently, institutional ethnography, originally developed by Dorothy E. Smith as a feminist, Marxist methodology to study social organisation and processes (Smith, 1987; Walby, 2005), provides a useful framework for the explication of these processes. While the theoretical framework applied in this thesis is neither predominantly feminist nor Marxist, I will argue that the method lends itself to broader application (Devault, 2006).

Institutional ethnographies treat work activities as the fundamental grounding of social life, in line with the Marxist focus on social relations (i.e. involuntary

participation in social relationships which are necessary to survive and constitute an economic structure) (ibid.). The methodology focuses on texts, and particularly their organisational and discursive power (Smith, 1999). Smith uses this methodology to demonstrate how knowledge is socially created and is done so at work (ibid.). This is because “the activity or operation of a text is dependent on the reader’s interpretative practices. These too are constituents of social relations rather than merely the

idiosyncrasies of individuals” (Smith, 1990, p.121). This focus on interpretation and the subjectivity it entails is particularly relevant for the secondary research question of this thesis, which is concerned with how street-level bureaucrats interpret legal

instruments and other sources of information to determine the status of asylum applications. It is clear that while this methodology was developed as a feminist tool, its focus on individual experience need not be applied exclusively to women’s experience, but that of all individuals working at a local level. Consequently, this

(22)

regardless of gender, both as subjects and agents of the dominant discourses and practices of asylum procedures.

In order to analyse the dominant discourses of the refugee status determination procedure, I conducted a discourse analysis of some of the key interpretative tools of asylum status used in the Greek asylum system and throughout Europe, namely the Qualifications Directive (2004) and the UNHCR Handbook for Determining Refugee Status and Guidelines on International Protection. These texts were selected as combined they represent both the binding and non-binding frameworks for

interpretation that decision-makers in the asylum system draw on. As discussed in Section 2.2 however, the Qualifications Directive is interpreted differently under different national legal systems. My research in this area was inhibited by my lack of knowledge of the Greek language, a point that will be explored in more detail in Section 5.4. As a result, I support my analysis of the Qualifications Directive and its implementation in Greek law through scholarly articles and the testimony of my interviewees.

Discourse analysis is a means of examining how “social interaction is influenced by cognitive scripts, categories and rationalities” (Torfing, 1999, p.81-82). In adopting this method, I do not claim to produce an objective, universal truth, but rather to provide a logical interpretation of the ideological discourse these texts produce. I seek to explore how discursive practices shape possibility for action, or, to place it in Bourdieu’s terms, how discursive practices shape the dispositions of those who operate in the field of the Greek asylum system.

To study the interaction between structure and agency, and the dispositions within this field, I also conducted a series of semi-structured interviews with people who work in the Greek asylum interview proceedings, including decision-makers, lawyers and interpreters. The semi-structured form of the interview was adopted to allow for representative sampling, complete answers, and flexibility in the question form. While this research method may restrict the generalisation of my findings, it should provide a deeper, fuller picture of employees’ perspectives than the more rigid structure of a questionnaire which large-N data collection would require, and has fewer access issues. Due to potential access restrictions and ethical concerns, which will be

(23)

expounded in Section 5.3, all the interviews conducted took place outside of the workplace.

According to institutional ethnography methodology, the interview subject is not a topic in and of themselves, but his/her experiences are used to reveal the institution’s organisational arrangement to the researcher (Walby, 2005). This is compatible with my research question on the role of street-level bureaucrats as it provides a means to explore the role of an individual within a larger organisational structure. The use of interviews provided an opportunity to observe how those working in the asylum interview proceedings discuss their social interactions with their colleagues and their clients (the asylum seekers), and how the dominate discourses shape these

interactions and the interview subjects’ narratives.

The use of interviews as a data source is not without its issues. It is impossible to eliminate subjectivity and reactivity in the interview and its analysis (Bernard, 2006). The accuracy of reported behaviour can be another concern. People can be inaccurate when reporting their own behaviours, as human memory can be unreliable.

Furthermore, the interview is a social encounter, so those involved have a personal stake in it and often wish to portray themselves in a positive light (ibid.). I was aware of this when conducting my research, and gathered as many respondents as possible within the allotted timeframe in order determine if their reports correlated and reduce the impact of this effect. The validity of my research and findings will be discussed further in Section 5.4.

I adopted a grounded theory approach to the data I collected using the above methods. I did not have a fixed research question in mind when I began my research, but rather sought to identify key themes as I reviewed texts and interviews and allowed the research questions to emerge from that process (Strauss & Corbin, 1990). This approach was particularly useful for data processing and triangulation (see Section 5.4).

5.2 Setting and Participants 5.2.1 Case Selection

(24)

important to investigate the impact that extra-local institutions have at a local level. This was particularly true given the aforementioned extent of national variation with regards to refugee status determination. Due to restrictions on time and resources for this research, a cross-case analysis was not feasible. Instead, it was necessary to select a single case that was representative of the challenges facing EU Member States as they aim to comply with the obligations they acquired when they ratified the various conventions and treaties outlined in Section 2.1.

Accordingly, Greece was selected. As one of the first twelve signatories of the Dublin Convention in 1990, Greece has been involved in the development of the instruments of CEAS since the start. However, for much of the time since then, Greece has had one of the lowest rates of refugee recognition among EU Member States (UNHCR, 2010). The ECtHR case of M.S.S. v Belgium was a decisive moment for a state whose asylum system had been the subject of considerable international scrutiny and condemnation. It was also a turning point for the Dublin system in its revocation of the assumption of safety in EU Member States. After this judgement, the Greek state introduced the revised Greek Action Plan for Migration and Asylum Management in 2012, which provided for the implementation of a new asylum system beginning in June 2013. Greece’s failure to fully implement its obligations under the CEAS instruments, and subsequent measures to reform and comply fully with them, make it a useful case study for the examination of refugee status decision-makers. This is because it required a significant shift in both the procedure and mentality of the national asylum system, which has been highly scrutinized and well-documented. As the only asylum centre in Greece for non-detainees was in Athens, which was also the base for almost all of the international organisations and NGOs working with asylum seekers, I selected Athens as the centre for my research.

5.2.2 Participants

As the subject of this thesis required knowledgeable and experienced experts as interview subjects, I used purposive sampling to select my participants. The snowball method was used to recruit my participants, in which existing contacts recruited future participants from among their acquaintances (Bernard, 2006; Goodman, 1961). I originally formed some contacts in the Greek asylum system during a small migration research project hosted by the Netherlands Institute in Athens, where I had the

(25)

opportunity to meet a number of academics and specialists in the field. Through these contacts I was able to meet with a number of people who worked in the asylum interview procedure. These participants included decision-makers in the first and second instance in the new system; decision-makers in the second instance in the backlog committees who worked with the appeals backlog from the old police

system; lawyers; and interpreters. Ten interviews were conducted in all, amounting to over twenty hours of interview data. Interviews were conducted outside of the

interviewees’ workplace in most cases, the exception being an out-of-hours office meeting with the lawyer who participated. This was done as a number of participants had recently ceased working in the asylum centre, and the more informal nature of the interviews encouraged greater openness in their answers.

Snowball sampling was selected it enabled me to access to a number of people in a rather niche community (Bernard, 2006). By having my respondents recruit members of their professional network and make the initial contact, potential respondents had greater privacy, which was particularly important if they were reluctant to participate. (Heckathorn, 1997; 2002).

5.3 Ethical Considerations

The research methods described above were designed to comply with high ethical standards. By using snowball sampling and adopting pseudonyms for participants in my research, I have striven to protect their privacy and autonomy. The fact that my subject matter was a highly sensitive interview process, the outcomes of which could have a life or death impact on the applicants, was a significant factor in my decision to limit my methods of data gathering to discourse analysis and interviews. The use of participant observation to balance the potential reliability issues inherent in interview was briefly considered. However, as participant observation can influence

proceedings due to the observer-expectancy effect, I felt it was inappropriate for me to be present during such sensitive proceedings (Bernard, 2006).

5.4 Validity

The use of exclusively qualitative data collection in this thesis limits its external validity, as the results cannot be easily replicated. This research makes no claims to objectivity as the analysis has been undoubtedly coloured by my personal experiences

(26)

and theoretical approach. Nonetheless, a brief discussion of the internal and external validity of this thesis will acknowledge strengths and weaknesses in my methods and may highlight potential challenges for future research.

Firstly, Greece was selected as a case study as it provided numerous examples of the challenges facing EU Member States as they attempt to comply with their obligations under CEAS. However, as I have documented, there are many discrepancies at a national level in the EU. Therefore, while the case of Greece may be representative, it cannot be considered a universal model. Furthermore, my research in Greece was historically specific, as it occurred soon after the election of a political party that has radically different views on migration policies than its predecessors, before they implemented any significant changes to the asylum procedure. As a result, researchers conducting similar work in future may find the relevant policies and practices quite different.

Additionally, the flexibility of face-to-face interviews means they cannot be replicated precisely (Creswell, 2009). However, their semi-structured format means they can be replicated to a degree, and triangulation was used to increase the validity of the findings. I applied both simple triangulation, where conclusions reached during interviews are discussed with participants, and discrepancies in interpretation resolved; and broad triangulation, where data generated from existing statistics, discourse analysis, and the face-to-face interviews was compared and used to provide contextual support for alterations and adaptations to the research methods if necessary (ibid.).

The internal validity of research should be strong, as the questions asked in the face-to-face interviews accounted for a broad range of potential influencing factors in the interpretation and determination of refugee status, which was continually developed and expanded as research progress. The discourse analysis of existing legal and interpretive documents also explored a wide variety of interpretations in order to ascertain the core themes. In doing so, I have done my utmost to ensure the validity of the data gathered. However, the ultimate goal of this thesis is not to establish

objective facts about the asylum decision-making process, but rather to provide a rich account of the experiences of those involved in it. My findings should be read in this light.

(27)

6. Refugee Status Determination:Data Analysis and Results

This chapter will provide an account of my research findings and place them in the context of my theoretical framework. It will address the research questions and, by means of the analysis of my data through the lens of the theory, provide support for my answers. It discusses the definition of street-level bureaucrats, and its

correspondence to the role of refugee status decision-makers. This chapter also looks at the structure of the Greek asylum system, the dominant discourse within it and its impact on the habitus, and explores how refugee status decision-makers exercise agency within this structure.

6.1 Refugee Status Decision-Makers as Street-Level Bureaucrats In this section I will address the first research question: To what extent are

interviewers in the Greek asylum system ‘street-level bureaucrats’? As discussed in Chapter 4, Lipsky’s definition of a street-level bureaucrat includes specific criteria. These criteria will be analysed step-by-step to determine the extent to which those conditions are present in the role of Greek Refugee Status Decision-Makers (RSDMs).

First, Lipsky states that street-level bureaucrats “are employees who are constantly called upon by the government to interact with citizens in the regular course of their jobs” (Lipsky, 1969, p.2). In the case of RSDMs, by the very nature of their job they do not interact with citizens, but rather non-citizens seeking international protection. However, expanding the definition from ‘citizens’ to ‘members of the public’ remains within the spirit of Lipsky’s original definition. In this interpretation, RSDMs easily fit within this category, as a main aspect of their role is to conduct face-to-face

interviews with asylum seekers. The regularity of this interaction varied depending on the category of RSDM. For those working in the first instance in the new asylum system, and in the second instance in the backlog committees, it was standard to conduct two or three interviews per day. For those working in the second instance in the new asylum system, recent changes in policy and the role of the rapporteur meant that face-to-face interviews were rare and sporadic (Andreas, 2015; Katerina, 2015). In this respect, although the role of those working in the second instance in the new committees is one in which they should encounter members of the public in the

(28)

suspended interviews in the second-instance, except when declared necessary by the rapporteur in exceptional circumstances, to prevent a backlog of cases (ibid.). This places them in unusual situation, as in theory their job fulfils this attribute of a street-level bureaucrat, but in practice they often do not.

A secondary criterion according to Lipsky is that “they have non-voluntary clienteles” and that “the clienteles, for the most part, do not serve as primary reference groups for these bureaucrats” (Lipsky, 1969, p.2). This means that the clients of street-level bureaucrats have no ability to determine who they will interact with and on what terms. It also means that the bureaucrats themselves cannot determine who their clients will be (ibid.). Furthermore, their role parameters and accountability are not determined by their clientele, but by other reference groups who assess and evaluate their performance. The first criterion, that of ‘non-voluntary clienteles’, is easily established for RSDMs. As demonstrated earlier, the Dublin system is designed to limit the asylum seeker’s choice of his or her country of application. Furthermore, when applying within a country, an asylum seeker cannot choose who they will submit their application to, nor who will conduct his/her interview (Eleni, 2015; Andreas, 2015). However, the asylum seeker can specify their preferred gender of interviewers (Alexandra, 2015). The RSDM cannot select which cases he/she will review either. When the new asylum system was established, the second instance committees were originally intended to specialise in cases from specific countries but the disparate workloads that resulted from this method of organisation meant that it was quickly abandoned (Katerina, 2015; Sofia, 2015). In the current system, RSDMs receive cases on the basis of their availability.

There are a number of reference groups for RSDMs. They must answer to their seniors in the asylum system, particularly with regard to the number of decisions they produce and the rate at which they produce them (Katerina, 2015; Andreas, 2015; Eleni, 2015). Furthermore, everyone I interviewed was highly aware of the failings of the old asylum system, and felt that their work in Greece was under particular scrutiny from international institutions such as the European Commission and the UNHCR. My participants each expressed feelings of responsibility towards the asylum seekers and an awareness of the significance of their decisions for the applicants. However, they did not feel directly accountable to the individual asylum seekers themselves. This was supported by the testimonies of Sofia and Alexandra, both of whom had

(29)

worked as lawyers representing refugees and asylum seekers before working as RSDMs. Both participants expressed a feeling that they were now “working for the other side”. When pressed and asked if that meant they felt that working for the state placed them in opposition to asylum seekers they disagreed with that interpretation, and said that they did what they could to help applicants. However, while neither felt that their role was to critique or undermine the applicants, I received the distinct impression that they felt their ability to advocate for the refugee was more curtailed in the role of RSDM than it had been in their previous positions. Alexandra said that she felt more distance from the asylum seekers and felt that she was judging them rather than representing them. I believe this is symptomatic of the fact that the refugees were no longer their primary reference group.

Another related secondary characteristic of street-level bureaucrats is that “they have limited control – although extensive influence – over clientele performance,

accompanied in part by high expectations and demands concerning that performance” (Lipsky, 1969, p.3). RSDMs cannot control or predict the behaviours and actions of the applicants, but through their own attitudes and questions they can have a

considerable impact on an asylum seeker’s performance in the interview. Many of the participants I interviewed emphasised the need to establish trust in the interview proceedings to ensure the applicant was comfortable about telling their story.

However, for Sofia, first-instance decision-maker in the new system, while trust was important, gaining the applicant’s full trust was impossible. She felt that the roles that the system cast them in, her as adjudicator and the applicant as supplicant, meant that the interview could never be a true dialogue between equals, and there would always be a degree of distance which curtailed openness.

Eleni, who had worked on two backlog committees, reported that attitudes towards establishing trust could vary. She told me: “when I was working with another committee, they were very prejudiced, and they had no interest in making someone feel at ease. Then we had some tensions between us so I changed the committee, and the next one was very nice, like offering some candies, offering some markers for the kids, and it was a totally different atmosphere for the interview” (Eleni, 2015). An example she gave of the atmosphere in the first committee she worked in was an instance where an applicant who had been charged with a crime was brought into the

(30)

favour of letting the uniformed police officer who accompanied him remain in the room. She believed this had a significant impact on the level of trust and the applicant’s openness during the interview.

As well as setting the atmosphere, the RSDM can influence the behaviour and success of the applicant in the interview by the tone and phrasing of the questions they ask. Giorgios, a pro-bono lawyer who works with asylum seekers, gave the following example: “If you ask someone why they came to Greece, of course they will say to work. But if you ask them why they left their home country, then you might hear the truth” (Giorgios, 2015). The asylum interview, in seeking to establish a “well-founded fear of persecution”, often navigates extremely sensitive subjects, and applicants can be asked to relive traumatic experiences. One example that came up during my research was that of the difficulty of establishing fear of persecution due to sexuality. Eleni felt there were very few questions she could reasonably ask to establish if someone was homosexual. Other RSDMs are perhaps less discriminating in their line of questioning. Giorgios reported that in one interview he attended with a client who was claiming asylum on the grounds of his sexuality, the interviewer asked his client which sexual positions he preferred (Giorgios, 2015). This invasive line of

questioning embarrassed the client. As the applicant’s responses in the interview are a key determinate of the success or failure of their application, this aspect of the street-level bureaucrat’s role is particularly crucial.

Street-level bureaucrats are also required to have “a significant level of independence in job decision-making” (Lipsky, 1969, p.2). RSDMs are intended to be independent from political structures. There are certain checks and balances in place. For example, all interviews are recorded, and an asylum seeker can request a copy of the notes from his/her interview. Furthermore, the appeals system is structured so that there is the possibility of the procedure being reviewed at each level: first by a second instance committee; then by a national court; then, as seen in the case of M.S.S., perhaps by the ECtHR. However, Andreas, a former decision-maker in the second instance, reported that for him and many of his colleagues there was pressure from senior members in the asylum system to limit the number of positive decisions they made. He said “what they unofficially do is that they put pressure on the interviewers, the people who conduct the interviews to say that, ‘Okay, you have done 20% asylum positive this month.’ You don’t take another positive decision’” (Andreas, 2015).

(31)

This report was particularly concerning in light of the next criterion for street-level bureaucrats: “They potentially have an extensive impact on the lives of their clients” (Lipsky, 1969, p.2). This is undoubtedly the case for RSDMs, given that to be eligible for international protection means “to have a well-founded fear of persecution”(1951 Refugee Convention), or in the case of subsidiary protection, to be “at risk of serious harm”(Refugee Qualifications Directive). This means that any decision made by an RSDM will profoundly affect the lives of the applicants. A positive decision means legal residence in the country; a negative decision could potentially be fatal. The Dublin system ensures that asylum seekers only have one opportunity to apply for protection in the EU, so when appeal procedures have been exhausted, the applicant must leave the EU altogether.

As well as fulfilling the above requirements, Lipsky’s street-level bureaucrats are also likely to work under certain conditions. The first of these conditions is that “available resources are inadequate (organizational and/or personal)” (Lipsky, 1969, p.4). The Greek asylum system lacks fundamental resources, which may due in part to the current economic crisis in the country. The asylum centre in Athens is the only one in Greece for non-detainees, with queues of between two and three hundred people waiting outside each morning. However, only thirty applications are processed per day (UNHCR, 2014). Due to budget issues, this situation is unlikely to change. When the new system was set up in 2013, most of its budget came from the EU and the UN (ibid.). However, in September 2014 UN and EU funding was scaled back, and the majority of the asylum system’s budget was covered by the national government. This led to major cuts. For example, the Appeals Committees were suspended between July and September 2013, and the number of committees was then nearly halved from nineteen to ten (Greek Council for Refugees, 2014). Furthermore, many of my respondents reported inconsistency in their salary payments and contracts. Alexandra, who worked on a backlog committee, reported that at the time of interview the backlog committees had been suspended for two months and she was unsure when they would resume. Sofia, who conducted interviews in the first instance, said that she had been working as part of a temporary, EU-funded programme that had ended a few weeks before our interview, and she hadn’t been paid for several months. These payment difficulties were not exclusive to the government employees. Both of the asylum interpreters I spoke to said that their

(32)

NGO had been struggling to pay them. One interpreter, Nick, said that even when they did receive a salary it was not enough to live on, requiring him to work two jobs. Funding is not the only scarce resource in the Greek asylum system. Most of those I interviewed reported receiving very little formal training before they started the job. For the backlog committees, Alexandra told me there was a five day training

programme. In the new system, there was a three day UNHCR training course before beginning work, although this was not compulsory. Katerina reported that she missed her training due to prior work commitments, and consequently had to start the job with nothing more than some reading material and her colleagues to guide her, as she had no prior interviewing experience. Other respondents reported occasional, sporadic training days or information lectures throughout their employment, but for the most part many felt that the bulk of their knowledge had come from their educational background, personal research, and interaction with their colleagues. Additionally, my respondents reported a lack of psychological support, both in the backlog committees and in the new system. Many said that they had had mental and emotional difficulty dealing with the cases that they heard on a day-to-day basis. Eleni said that she, along with some of her colleagues in the backlog committees, had requested group therapy to help with this, and her request had been denied. These are just a few examples of the resource scarcity that RSDMs in Greece face as street-level bureaucrats. The final two criteria are, in this case, related. For Lipksy’s street-level bureaucrats, “work proceeds in circumstances where there exists clear physical and/or

psychological threat, and/or the bureaucrat’s authority is regularly challenged” and “expectations about job performance are ambiguous and/or contradictory, and include unattainable idealized dimensions” (Lipsky, 1969, p.4). The role of RSDM bears strong similarities to the judiciary, and, as Lipsky argued in his case of lower court judges, unrealistic expectations can be considered a threat to the idealised view of their role. This closely correlates with the accounts given in my interviews. Eleni and Alexandra told me that in the backlog committee they were expected to conduct four interviews per day, something that rarely, if ever, happened. Furthermore, they had to deliver a decision within thirty days. In the new system, decision-makers in the first instance had sixty days to write a decision, and in the second instance they had sixty days in cases where an interview was conducted and thirty days in cases where there was no second interview. Andreas reported that in the new system there was

(33)

considerable pressure to submit decisions in a much shorter time frame. He said: “they wanted the process to be quicker and to show to the EU that now we’re very quick, and they use to press the more vulnerable, politically vulnerable commissions to – not to issue on 30 days that the law says but to issue in five days, in one day, because they used to say if it is an easy application, you can read what the rapporteur proposes”. However, he was reluctant to rely on the rapporteur’s proposal, as he said he had found in some instances they had taken paragraphs from reports out of context to support their decision, and did not present a full picture of the risks the applicant would face if they were returned to their home country. The role of the rapporteur will be elaborated on in the next section, which discusses the organisational structure of the system.

The pressure to conduct as many interviews as possible and to produce decisions as quickly as possible, presents a challenge to the RSDMs. Several of my participants reported the difficulty of trying to fully research a case, the country conditions, and the relevant law within the time frame given (Katerina, 2015; Sofia, 2015, Andreas, 2015). This necessitated a compromise between the quantity and the quality of the decisions produced, and created the ‘cognitive dissonance’ described by Satyamurti (1979). Both Katerina, an RSDM, and Nick, an interpreter, used the word ‘machine’ to describe how they felt in their roles. They did not feel as if they were treated like individuals with emotions and needs, but rather were valued on the basis of how much they could produce. Andreas took pride in the fact that he, along with members of his committee, had often waited until the last possible day to issue a decision, in order to take the time to research the case in as much depth as possible, and said that his committee had often produced 36 page decisions when the average was 8-12 pages. In this way, he worked to resist the threat that pressure posed to the idealised

dimensions of his role, and the way in which his authority was challenged by it, as far as he could within the law.

Based on the evidence provided, it can therefore be concluded that in general the role of Greek refugee status decision-makers fully corresponds to that of street-level bureaucrats, according to Lipsky’s definition. However, there are some reservations for appeals committees in the new asylum system. As previously stated, during the time period examined in this thesis, there was a policy change that curtailed

Refugee roulette in the Common European Asylum System : subjectivity in Greek refugee status determination