Struggling with humanitarianism : changes of migration policies concerning unaccompanied minor asylum seekers in Switzerland from 1951 to 2015

(1)

Universiteit van Amsterdam

Graduate School of Social Sciences

Master in Sociology - Migration and Ethnic Studies

Master’s Thesis

Carola Schabert

carola.schabert@gmx.ch

Student number: 11128178

Supervised by:

Dr. Simona Vezzoli (First Supervisor)

Prof. Dr. Hein de Haas (Second Supervisor)

Amsterdam, 30

th

_{June 2016}

Struggling with humanitarianism

Changes of migration policies concerning unaccompanied minor asylum

seekers in Switzerland from 1951 to 2015

(2)

Abstract

In 2015, the number of unaccompanied minor asylum seekers heading towards Switzerland rose to an unprecedented level of 2’749 UMAs. This is almost seven times higher than the average of the previous decade. Although international regulations like the Convention on the Rights of the Child provide child-specific rights, national policies are often not in the best interests of the child, which is a primary prin-ciple of the convention. This paper looks at the evolution of national migration policies concerning unaccompanied minor asylum seekers in Switzerland between 1951 and 2015. The changes in migra-tion policy restrictiveness and the drivers behind these changes are examined. Although Switzerland earns a reputation of having a humanitarian tradition, the results show that the restrictiveness at the national level has been increasing and the Convention on the Rights of the Child is implemented on a minimum level with a focus on the eventual removal of the child. Regarding the drivers of the policy changes, several factors like the migration stock, the recognition rate, the Swiss people’s party and the annual GDP per capita show high correlations with policy restrictiveness. The results contradict Swit-zerland’s humanitarian reputation and rather seem to point to a liberal paradox effect.

(3)

III

List of abbreviations

ANAG Bundesgesetz über Aufenthalt und Niederlassung der Ausländer (Federal Law of Stay and Residence of Foreigners)

Art Article

AS Asylum Seekers

AsylG Asylgesetz (Federal Law on Asylum) of 1998

AsylV 1 Asylverordnung 1 über Verfahrensfragen (Asylum Ordinance relating to the Asylum Procedure)

AuG Ausländergesetz (Federal Law on Foreigners)

BFF Bundesamt für Flüchtlinge (Federal Office for Refugees) BFM Bundesamt für Migration (Federal Office for Migration) BFS Bundesamt für Statistik (Swiss Federal Statistical Office) CRC Convention on the Rights of the Child

CRSR Convention Relating to the Status of Refugees

DAWES Dismiss an Application without Entering into the Substance of the Case DEMIG Determinants of International Migration POLICY database

DRC Declaration on the Rights of the Child 1959

Dublin II Council Regulation No 343/2003 of 18. February 2003

Dublin III Regulation (EU) No 604/2013 of the European Parliament and of the Council of 26 of June 2013

EVZ Empfangs- und Verfahrenszentren (Reception Centres)

GC General Comment

IMES Bundesamt für Zuwanderung, Integration und Auswanderung (Federal Office for Im-migration, Integration and Emigration)

NEE Nichteintretensentscheid (DAWES = Dismiss an Application without Entering into the Substance of the Case)

SEM Staatssekretariat für Migration (State Secretariat for Migration) SFH Schweizerische Flüchtlingshilfe (Swiss Refugee Council)

(6)

TestV Testphasen zu den Beschleunigungsmassnahmen im Asylbereich (Pilot Phase for an Acceleration of the Asylum Application Procedure)

UMAs Unaccompanied Minor Asylum Seekers

UN United Nations

UNHCR United Nations High Commissioner for Refugees

ZEMIS Verordnung über das Zentrale Migrationsinformationssystem (Central System of Mig-ration Information)

(7)

1

1 Introduction

Although most of the minor asylum seekers are accompanied by adults, a growing number of unac-companied children is arriving in Europe since the 1990s (Arnold et al. 2014: 468). In 2015, the number of unaccompanied minor asylum seekers heading towards Switzerland rose to an unprecedented level of 2’749 UMAs. This is almost seven times higher than the average of the previous decade. These chil-dren are extremely vulnerable because they often had traumatizing experiences either in the present or near past. Their vulnerability can be identified on three different levels: they are children, asylum seekers and unaccompanied. Therefore, they should have the right to special protection. According to Article 3 of the Convention on the Rights of the Child of 1989 (CRC), the protection has to be in ‘the best interest of the child’. However, who defines what is in the best interest of the child? This is a complex question as there are several aspects to consider. While for children under the age of 18, parents are still in charge as primary caregivers, when it comes to the protection of unaccompanied minors, the role of the state becomes more prominent (Arnold et al. 2014: 470). States have to protect the child and make sure that the child is granted the same rights as every child living in the same country, as stated in Article 22 of the CRC. Therefore, every child who is seeking refugee status is enti-tled to the same protection as any other child. This, however, is exactly the ambiguity unaccompanied minor asylum seekers find themselves in. While they are protected as children under the CRC, they are at the same time asylum seekers who have less rights and protection, as these rights tend to be more vulnerable to the political situation. International conventions leave the implementation of their reg-ulations up to national governments. To which extent states are willing to meet rights as the principle of the best interest of the child should become clear after looking at national legislations.

Research has been done on migration policy-making (de Haas and Natter 2015; de Haas et al. 2014; Bonjour 2011) as well as on the topic of the UMAs (Arnold et al. 2014; Barbulescu and Grugel 2016; Bhabha and Young 1999; Derluyn and Broekaert 2008). Research is also more generally conducted on the evolution of asylum policies in Europe (Hatton 2009) and on the role of the state concerning inter-national migration and its policies (Massey 1999; Joppke 1997, 1998). Holzer et al. (2000: 1189) focus on the Swiss context and the evolution of its migration policies. They describe that Switzerland has a reputation of having a humanitarian tradition in asylum policy making. Also the Swiss government itself emphasises the humanitarian tradition, referring to the accommodation of Hungarian and Czechoslo-vakian refugees during the cold war and refugees from the Balkans during the 1990s (EDA 2016: Hu-manitäre Tradition). Due to Switzerland’s humanitarian tradition, one would expect the UMAs to be granted a broad sets of rights. As I am, however, not aware of any research being done, focusing on migration policies for the specific migration category of UMAs in Switzerland, this thesis examines the role of the Swiss government in dealing with UMAs, their migration policies and how these policies have changed. The focus is on which rights UMAs have and how they are treated on a national level,

(8)

as children or as asylum seekers, as well as how Switzerland deals with the ambiguous position of the UMAs. Comparing national policies with international conventions should help to understand how Switzerland deals with UMAs. The CRC of 1989 can be taken as the most important convention for the rights of UMAs, defining them principally as children. Even though UMAs represent a sub-group of asylum seekers, this analysis is seeking to understand this group as children and less as asylum seekers. An additional aspect in determining how states are dealing with UMAs is the restrictiveness of the relevant migration policies. How migration policies have changed over time and whether they have become more or less restrictive represents an important aspect of how states are seeking to influence the UMAs rights. Additionally, the role of the state in influencing the UMAs position is determined by several factors which are discussed in the theoretical section of the paper.

Due to Switzerland’s humanitarian tradition, which goes with the hosting of the International Commit-tee of the Red Cross and the Geneva Convention of 1949, back to the 19th_{century, a historical} perspec-tive seems to be interesting to take. Additionally, states have been dealing with unaccompanied mi-nors for a long time. As Arnold et al. (2014: 468) describe, this has been a global concern since the First and Second World War. Keeping this in mind, looking at the migration policies concerning UMAs from a historical perspective since the end of the Second World War might be a fruitful way to go. As a take-off point serves the 1951 Convention Relating to the Status of Refugees (CRSR) which for the first time defined a refugee and his or her rights officially. The research question guiding this thesis will therefore be as follows:

How have Swiss migration policies concerning unaccompanied minor asylum seekers evolved between 1951 and 2015? What have been the drivers of these changes?

To support this research question, two sub-questions were developed.

- How do migration policies concerning unaccompanied minor asylum seekers differ from the Convention on the Rights of the Child 1989?

- Have Swiss migration policies concerning unaccompanied minor asylum seekers become more restrictive since 1951?

The analysis is based on a mixed-methods approach, containing a qualitative change-tracking policy analysis and descriptive statistics. In the first part of the analysis, the migration policy changes and their changes in restrictiveness are investigated. In the second part, the focus lies on the influence of previous migration flows on the restrictiveness of migration policies concerning UMAs.

The analysis builds upon the theoretical concepts of the role of the state, policy making and the liberal paradox. By doing this analysis, the goal is to add to a better understanding of how states are dealing

(9)

3 with migration policies concerning the ambiguous position of UMAs. This topic is not just of relevance for Switzerland because bringing together knowledge of different cases can improve the understand-ing of the bigger picture. Addunderstand-ing an additional case might be helpful to get a better grasp of migration policy processes and how states are shaping them.

The analysis is structured as follows. The theoretical framework contains, besides previous research, the role of the state and the concept of the liberal paradox. In addition to the role of the state in shaping migration policies and policy-making processes, the liberal paradox helps to understand states’ interests which is expected to shed some light on the evolution of migration policies concerning UMAs. After the theoretical framework, the Swiss asylum procedure is discussed in chapter three. In the fourth chapter, the methodological framework is laid out which addresses the method and data. The results follow in chapter five and are structured according to the research questions. Thereafter, a discussion and finally a brief conclusion round off this analysis.

(10)

2 Theoretical framework

The AsylG’s definition of a refugee relies upon the definition of the CRSR (see Appendix I). Based on the criteria mentioned in the CRSR, the AsylG defines a refugee as a person who receives serious dis-advantages as a consequence of these criteria. Serious disdis-advantages are danger of life, body or free-dom as well as unbearable psychological pressure. In addition to this, the AsylG also defines persons who are not considered as refugees. In reservation of the CRSR, not a refugee is a person coping with serious disadvantages due to refusal to do military service or desertion (§ III, Art 3, AsylG). Also not a refugee is a person, in reservation of the CRSR, who has serious disadvantages due to his or her be-haviour after having left the country of origin, which is not a continuation of the convictions or orien-tation already existing in the country of origin (§ IV, Art 3, AsylG).

In the national law, no definition is given on the term of asylum seeker. However, the SEM defines on their website asylum seekers as persons applying for asylum in Switzerland and who are in the exami-nation procedure. In principle, they have a temporary residence permit in Switzerland for the duration of the procedure (SEM 2015f: Ausweis N).

An UMA has, in addition to being an asylum seeker, two other characteristics. It is a minor and unac-companied. The CRC and the ordinance of Swiss asylum law (AsylV1) define a child as a human being under the age of eighteen years (Art1, CRC; Art1a, AsylV1). Being unaccompanied is defined by the UN as children who have been “separated from both parents and other relatives and are not being cared for by an adult who, by law or custom, is responsible for doing so” (Art 7, CRC GC No. 6). On a national level, the federal law on asylum (AsylG) and the asylum ordinance (AsylV1) do not imply a concrete definition of the term of an unaccompanied minor. The state secretariat for migration (SEM), however, provides an asylum and return compendium which contains a more precise definition of the CRC Gen-eral Comment No. 6 of 2005. The compendium defines a minor as accompanied if he or she is being accompanied by one parent or an adult legally responsible for the minor1_{(Handbuch Asyl und} Rück-kehr 2015: 5). Regardless of the situation, the definition of the status of the minor must be handled in a restrictive manner as defining a child as accompanied or unaccompanied leads to different kinds of rights and restrictions (Handbuch Asyl und Rückkehr 2015: 5). Restrictive in this way means that the official authorities have to be careful with determining a child as accompanied, as the minor will then be registered together with the adult and loses its UMA rights.

1_{An adult sibling already living in Switzerland does not count as parental care and the minor is therefore defined} as unaccompanied. If a minor is accompanied by another adult besides the parents, the definition is not very clear. If the minor has been living together in the same household with an adult relative in the country of origin and the relative has been responsible for the child, the minor can be defined as accompanied (Handbuch Asyl und Rückkehr 2015: 6). An exception like this can be made if the relatives are willing to take responsibility for the child during their time in Switzerland and if it is in the best interests of the child. Additionally, official authorities have to determine the relatives as being the legal representatives of the child. Is a minor arriving with adults who are not relatives, the minor is generally defined as unaccompanied.

(11)

5 To summarize, an UMA is an asylum seeker in a country except for the country of origin, under the age of eighteen, not being accompanied by at least one of his or her parents or another adult holding responsibility for the child. This definition only focuses on unaccompanied minors who actually seek asylum. This can be problematic as it leaves aside all children who are for any reason not seeking asy-lum. However, getting reliable data on the situation of unaccompanied minors not seeking asylum is difficult as they often disappear. Therefore and due to the fact that the Swiss government in its legis-lation uses the UMA term when referring to the rights and restrictions of unaccompanied minors, it seems to be most straight-forward to take over this definition.

2.1 Previous literature

Several scholars looked at the situation of the UMAs from different perspectives and in different con-texts (Arnold et al. 2014; Barbulescu and Grugel 2016; Bhabha and Young 1999; Derluyn and Broekaert 2008). Bhabha and Young (1999) as well as Villarreal (2004) examined the situation in the USA. Both articles argue that UMAs should not be treated in the same manner as adult asylum seekers. While Babha and Young (1999: 124-125) argue for a child-appropriate asylum procedure, implementing the CRC, Villarreal argues that children are fleeing due to child-specific reasons which are not acknowl-edged as official reasons for getting granted asylum. She therefore argues for child-specific refugee reasons. Arnold et al. (2014) look at the role of the guardian in determining the best interest of the child in Belgium, Ireland and the Netherlands. They study the question to which extent these three countries meet the minimum standards of the guardianship and the best interests of the child, written down in European law. The main finding of the authors is that all the three cases show divergences in guardianship law and practice (Arnold et al. 2014: 502). The best interests of the child principle is in all three states not clearly determined and the states do not meet the regulations of the CRC. Due to the fact that there are still large variations in national practice, European institutions should encourage states to go beyond their standards in order to meet the principles of the CRC (Arnold et al. 2014: 504). Barbulescu and Grugel (2016) focus on civil society-state struggles around the repatriation of unac-companied minors in Spain. They argue that although pro-rights civil society organizations were ini-tially able to force the state to act in accordance with international human rights obligations in relation to repatriation, the Spanish state was able to react energetically and regained control. The government made bilateral agreements with countries of origin of the minors and in this way was able to delegate its responsibility partly to other states that have poor human rights records and high levels of violence (Barbulescu and Grugel 2016: 14-15). The authors argue that states are, by relocating people they are legally responsible for, able to use ‘weapons’ which are unique to states.

Derluyn and Broekaert (2008) look from two different perspectives, a psychological and a legal per-spective, at the situation of unaccompanied refugee children in Belgium. From a legal perper-spective, the

(12)

minors are principally seen as asylum seekers. From a psychological perspective, unaccompanied ref-ugee minors are seen as children which often have traumatizing experiences in the present and in the past. Having these two perspectives, the authors argue that depending on which perspective the gov-ernment takes as a starting point, the care and reception system will be constructed differently (Der-luyn and Broekaert 2008: 320-323). The authors find that states concretise the CRC on a minimum level although unaccompanied refugee children and adolescents are a vulnerable group in need of special care and protection and the CRC guarantees the right to special protection for unaccompanied refugee children until the age of majority, (Derluyn and Broekaert 2008: 327). Providing unaccompanied minors only with temporary protection and residence documents until the age of majority leads to minimal standards of care and reception. The government, by taking in the legal perspective, provides lower quality of care standards to unaccompanied minor asylum seekers compared to native-born minors. Therefore, Derluyn and Broekaert (2008: 328) call for a psychological perspective which should be taken as a starting point for the entire reception and care system, treating them in the first place as children.

These studies showed the relevance of examining UMAs and the national context. As Holzer et al. (2000) studied in detail the Swiss context of the asylum system, their study is used to get a better understanding of the Swiss context. In addition to this, Ruedin et al. (2015) also looked at the evolution of immigration and integration policies in Switzerland, covering the time period for 1848 to 2014. The authors divide the history of migration policies in three periods: the expansive period, the restrictive period and the expanding period (Ruedin et al. 2015: 8). The expansive period, dominated by stability and few policy changes, lasted from 1848 until 1921 (Ruedin et al. 2015: 10). During the restrictive period which lasted from 1921 to 1974 almost all policy changes were towards a more restrictive end. The expanding period, lasting since 1974, is dominated by increasingly expansive migration policies and an increase in politicized and contested immigration (Ruedin et al. 2015: 11). Since 1968, popular initiatives are increasingly used as instruments to shape policies and an increase in the number of regulations took place. The study of Holzer et al. (2000) only covers the expanding period, looking at the asylum applications and its determinants in Switzerland from 1986 to 1995. They argue that the number of applications is influenced by the Swiss government’s deterrence, the economy in the send-ing and receivsend-ing country, key events in the international system, as for example a war, and the recog-nition rate (Holzer et al. 2000: 1188). Holzer et al. (2000: 1183, 1185) argue that since the early 1980s, Switzerland, like many Western European countries, experienced a steep increase in the number of applications which lead to the situation that asylum policies became one of the most fiercely debated issues in Swiss domestic politics and xenophobia started rising. This resulted in the adoption of seven measures, which are the modification of the asylum law in 1983, a legislation with an executive order which provisionally created a new federal office to administer refugee and migrant programs in 1985,

(13)

7 an asylum law modification in 1986, two simultaneously adopted ordinances, an urgent federal reso-lution in 1990, and two ordinances in 1991 (Holzer et al. 2000: 1183). Holzer et al. (2000: ibid.) look at these new rules and whether their goals have led to a substantial reduction in the number of asylum applicants. The authors come to the conclusion that states do shape the number of asylum applications, however, not under all conditions (Holzer et al. 2000: 1205). The deterrence measures are not success-ful if the push factors in a nearby region reach a critical level (ibid.). Nonetheless, the recognition rate has an impact on the number of asylum applications (Holzer et al. 2000: 1201). This means that official authorities can use the recognition rate as a highly questionable steering tool to influence future ap-plications (ibid.). In addition, the government increasingly started to use a policy of temporary protec-tion to be able to send asylum seekers back after the situaprotec-tion has calmed down in their country of origin (Holzer et al. 2000: 1205). Holzer et al. (2000: 1184) furthermore show that the number of asy-lum applicants and the admission rate are interrelated, which means that the Swiss authorities change the interpretation of the reasons for recognizing refugees without altering the legal refugee status. In this way, a uniform application of the global norms is, according to the authors, no longer guaranteed as the governments make the asylum decisions dependent on other issues than the motivation of the asylum seeker (ibid.). This strategy stands in contrast with the, at least rhetorically, communicated humanitarian tradition of Switzerland which the Swiss government still highlights (ibid.). With the new AsylG of 1998, a new provision containing the protection for refugees fleeing from violence and war has been adopted (ibid.). This enabled the government to decide in a general fashion how many per-sons fleeing from a war will be given the right of temporary protection (ibid.). Although this might sound generous at first, the authors argue that this was designed to prevent persons from going through the costly asylum procedure where everyone has to be treated individually (Holzer et al. 2000: 1205-1206). This again points in the direction of the Swiss authorities verbally promoting new human-itarian mechanisms, however, actually not being willing to break decisively with the old regime of re-strictive asylum politics. Finally, Holzer et al. (2000: 1206) state that this demonstrates that the trade-off between international commitments and unilateral measures is one of the crucial conflicts in the asylum policy of Western European states.

To summarize, the articles on UMAs show the importance of providing child-specific rights and treat-ments during the asylum procedure. They, however, also show the shortcoming of governtreat-ments in carrying out these tasks. This underlines the importance of investigating the national context like Holzer et al. did. To examine the national context of migration policies concerning UMAs, the liberal paradox will be used. This theoretical concept provides the opportunity of investigating international migration policies by taking into account the role of the state and the national context.

(14)

2.2 The role of the state, its policies and the liberal paradox

Massey (1999) looks at the role of the state in international migration at the dawn of the 21. Century. By looking at how states intervene in migration processes, Massey (1999: 307) defines immigration policies as the “outcome of a political process through which competing interests interact within a bureaucratic, legislative, judicial, and public arenas […]” to regulate the flows of migration. Although national politics aim at restricting immigration, factors like the global economy or the internal consti-tutional order of liberal democracies turned out to be obstacles making the enforcement of restrictive policies challenging (Massey 1999: 313-314). Increasing civil rights for immigrants undermine the ca-pacity of states to control immigration. However, Joppke (1998) argues that the rise of a liberal doc-trine of human rights is not sufficient to protect the rights of immigrants. There is an additional need for the means to guarantee those rights within a specific national polity which requires a written con-stitution and a strong, independent judiciary. Massey (1999: 315) takes up this argument and defines five factors which determine the efficacy of migration policy restriction. For Switzerland, Massey grades the five factors as follows: a high strength of bureaucracy, strength of constitutional protections and independence of judiciary, a moderate demand for entry and a low tradition of immigration. Joppke (1998: 271) tries to answer the question why states accept unwanted migration and argues that rather than externally, liberal states are internally impaired in controlling unwanted immigration. Not international human rights but legal and moral constraints keep European states from pursuing rigorous zero-immigration policies (Joppke 1998: 272). The post-Second World War period lead with the implementation of international human rights to the situation that not only states but also individ-uals are recognized as subjects of international law (Joppke 1997: 292). However, international human rights have to be understood as ‘soft laws’. States are not losing ability to determine entry and exit of migrants due to the rise of an international human rights regime (Joppke 1997: 260). States’ sover-eignty has never been absolute and will never be. It is the self-limited soversover-eignty of liberal states rather than globally limited sovereignty which explains states politics vis-à-vis immigrants (Joppke 1998: 292). The two increasingly conflicting principles ‘popular sovereignty’ and the ‘protection of human rights’ of liberal nation-states are the ones shaping asylum policies (Joppke 1997: 259). And within this contradiction, liberal states’ most powerful weapon against undesirable migration remains exclusion (Joppke 1997: 295).

While Joppke and Massey generally aim at understanding the role of the state in regulating migration, policy-making approaches focus on the processes of how policies evolve and by which determinants they are shaped. Migration policies can broadly be defined as “rules (i.e. laws, regulations, and measures) that national states define and implement with the (often only implicitly stated) objective of affecting the volume, origin, direction, and internal composition of immigration flows” (Czaika and de Haas 2013: 489).

(15)

9 Bonjour (2011: 94), in line with Joppke, argues that while scholars often state that national policy mak-ers have lost the power of policy-making to the court or supra-national actors, in fact they are still the ones holding the power over migration policy-making. Bonjour, like Joppke as well, also wants to an-swer the question why states accept unwanted migration. She looks at policy-making of family migra-tion in the Netherlands, as this is a very undesirable kind of migramigra-tion for a country that refuses to become an immigration country. The case shows that ideas, ideologies and ethical considerations play a crucial role in migration policy-making processes (Bonjour 2011: 111). National policy makers, free of external constraints, determine the dynamics of domestic policy-making, which is to a large extent a morality play (Bonjour 2011: 115).

De Haas and Natter (2015) looked at the determinants of migration policies and whether the political orientation of governments matter. Based on their analysis, the authors do not find a clear association between the political orientation of governments and the restrictiveness of migration policies. Left-wing parties cannot simply be classified as pro and right-Left-wing parties as anti-migration (de Haas and Natter 2015: 4-5). Left-wing actors such as unions for instance be for more restrictive migration policies to protect the market, whereas right-wing actors such as employers can be for more liberal policies to get more liberal markets. The authors, however, find that other factors such as economic growth and unemployment, recent immigration levels and political system factors such as electoral systems or the level of federalism play a role. Considering the previous migration flows, high levels of immigration can cause a public backlash against immigration which then leads to a push for more restrictive policies. The authors also argue that migration policy making is not entirely determined by internal domestic affairs. Supra-national policy making and international diffusion of policy practices have an increasing influence, which is in line with Sassen’s (1996: 9) argumentation. Furthermore, migration policies have to be understood as a ‘mixed bag’ of policy measures targeting different migrant groups and policy categories which means that political parties’ orientation is likely to vary among types of policies and target groups (de Haas and Natter 2015: 6).

Referring to the understanding of migration policies as mixed bags, de Haas et al. (2014: 22) also look at migration policy changes in 45 countries since 1945 and show that the overall changes of migration policies have been towards less restrictiveness. Focusing, however, on the different target groups and different polity areas, trends differ. While entry and integration policies have become less restrictive, border control and exit policies have become more restrictive. For irregular migrants, family members and asylum seekers, trends are towards more restrictiveness, while for low-skilled workers or students, trends are the opposite (de Haas et al. 2014: 23-24). This underlines the argument that migration pol-icies have to be understood as a tool of selection rather than as an instrument affecting migration numbers. In this way, states are able to control who is allowed to immigrate and who is not (de Haas et al. 2014: 23).

(16)

Besides policy-making processes, another theoretical concept helps to understand how states shape migration policies, the embedded liberalism. The theoretical concept of embedded liberalism is, inter

alia, ascribed to Hollifield. According to Hollifield (1992: 569), international migration is influenced by

national and international policies and has an economic and a political dimension. The globalization of markets (economic dimension) as well as an increase in rights-based liberalism, based on civil and so-cial rights (political dimension), lead to an increase of international migration (Hollifield 1992: 590). Due to these dimensions, migration represents, according to Hollifield (1992: 591), a critical dilemma for governments of liberal states, challenging the sovereignty and autonomy of the nation-state. The theoretical concept of embedded liberalism, or the liberal paradox how Hollifield defines it, deter-mines this challenge. Embedded refers to the liberal notions of rights (Hollifield 1992: 574). Hollifield (1992: 575) argues that human and civil rights have spilled over into international relations. This is what makes it so difficult for nation-states to regulate immigration. Especially for refugees, who are as non-citizens among the most vulnerable individuals in liberal societies, the embedded liberalism serves as an explanation. In other words, the rights-based politics of international relations challenge the na-tion-state’s sovereignty. States have to find a middle-path between liberalism and nationalism (Holli-field 1992: 582). While economic forces push for open borders, security concerns and powerful politi-cal forces simultaneously push towards closure (Hollifield 2004: 885). In this way, “international mi-gration reveals a contradiction between the main economic purpose […] to promote exchange and the national perquisites of sovereignty and citizenship” (Hollifield 1992: 568). On the one hand, states must keep economies and societies open to trade, investment and migration, on the other hand, the move-ment of people involves greater political risks (Hollifield 2004: 885).This is the ‘liberal paradox’ states are dealing with.

Concerning migrants’ rights, most of the rights come from the legal and constitutional protection which are with the extension to non-nationals guaranteed to all members of society (Hollifield 2004: 901). If states, by regulating international migration, do not respect the human or civil rights of the individual, the liberal state risks undermining its own legitimacy (Hollifield 2004: 901-901).

While Joppke and Bonjour, referring to the power structure, argue that international migration is in-ternally determined within the nation-state, Massey, Sassen, de Haas et al. and Hollifield argue for an increasing influence of external, international policies determining states’ strategies in dealing with international migration. Boswell (2007: 86) adds another dimension, arguing that if nation-states are determined by international regulations, the question is on which ideals do these norms ground? Ac-cording to Boswell, the embedded liberalism states that the international norms appear from hege-monic liberal states’ interests in spreading their worldview but it does not state whether these inter-ests are based on the motivation of an ethical commitment to values of liberty and justice or simply on states’ interest in national security and free trade. Boswell (2007: 86) argues that this is an im-portant difference because if the latter option holds true, the capacity to provide a robust support for

(17)

11 refugee and immigrant rights will be limited. Referring to states’ motivation, Boswell (2007: 89) argues that states’ legitimacy is based on four preconditions, namely fairness, accumulation (of wealth), se-curity and institutional legitimacy, which do, however, not all have to be given in all migration policy types.

According to these theoretical concepts, we can expect for the current analysis that the Swiss govern-ment finds itself in a dilemma between the implegovern-mentation of liberalism in the form of international human rights and nationalism, the maintenance of the sovereignty. Due to the increasing importance of international conventions which Switzerland has to implement in order to keep its own legitimacy, one would expect more open borders since the end of the Second World War. However, as Joppke and Bonjour emphasize, the mere ratification of international human rights is not sufficient. The Swiss gov-ernment remains, according to them, still in the power position of how to deal with immigration. In other words, the nation-state plays a crucial role in shaping migration and holds responsibility of how international human rights are implemented. Concerning the humanitarian tradition of Switzerland, one would expect a sincere respect for the international human rights. Especially in the case of child protection, the state can be expected to be holding a liberal position, regarding international human rights, because this might be a means to earn a favourable reputation. Like Bonjour states, policy-making is to a large extent a morality play. To which extent the Swiss government acts in favour of liberalism and therefore against nationalism is examined in this paper. Due to the state’s tensions of interests of on the one hand international regulations and on the other hand sovereignty, I expect the Swiss government and its policy-making to hold a balance between these two opposites.

Regarding the drivers of these changes, several factors might be relevant according to the theory. De Haas and Natter showed that previous migration flows, the economic growth and unemployment can shape migration policies. Concerning the previous migration flows, the underlying assumption is that increasing migration flows will raise internal voices against immigration and will push the Swiss gov-ernment to act in favour of nationalism and therefore against liberalism (de Haas and Natter 2015: 7; Massey 1999). This means that in times of higher migration flows, migration policies concerning UMAs may become more restrictive. They also looked at the effect of the political orientation of governments on the restrictiveness of migration policies (de Haas and Natter 2015). Although they could not confirm this effect, the variable is tested again, due to the explained backlash effect and the focus on one specific target group. The assumption is that previous migration flows, or alternatively the migration stock, might lead to a public backlash which gives the political parties acting in favour of anti-immigra-tion policies more power. According to Holzer et al. (2000), the recognianti-immigra-tion rate has an impact on mi-gration flows. The recognition rate is also likely to influence or to be influenced by mimi-gration policies.

(18)

3 Swiss asylum procedure

Generally, the Swiss law concerning UMAs follows the international guidelines of the 1989 Convention on the Rights of the Child, especially taking into account the best interests of the child principle (SEM 2015a: 6). Several policies determine the current legal situation of UMAs in Switzerland. The most fundamental policy of the Swiss Confederation is the federal constitution of 1999. Article 11 of the constitution addresses the protection of children and adolescents. It states that minors have the right to special protection of their integrity and their own development (§ I, Art 11). Furthermore, minors can rely on their rights within the context of their ability to judge (§ II, Art 11) (see Appendix II for the original German version).

The federal law on asylum (AsylG) is the primary policy determining the UMAs’ rights. Article 17 of the AsylG refers to the situation of the UMAs. It states that special regulations of the asylum procedure are provided for minors (§ II). Asylum applications of UMAs have to be dealt with priority (§ IIbis). Additionally, during the asylum procedure of UMAs, the responsible canton has to determine a guard-ian, representing the interests of the child (§ III). The federal council regulates the access to legal rep-resentation in the reception centres and if a minor is suspected of having reached the age of 18, the SEM can order an age determination test which is explained later on (§ IIIbis, IV; see Appendix II). Besides the priority of UMAs applications, they are following the same application procedure as adult asylum seekers. A brief understanding of the procedure’s functioning is given in the following section, based on the asylum and return compendium (Handbuch Asyl und Rückkehr 2015: chapter B6, C1 and C10, Appendix III provides a visualization of the process). The compendium is a reference book pro-vided by the SEM for staff members of the secretariat dealing with asylum procedures. The reference book contains one chapter about the guidelines of the UMAs in French. A German translation is already available, however without official approval2_.

An application can be handed in at an EVZ (reception centre), at a border control point of an interna-tional airport in Switzerland or at a general border control point on the Swiss border. The first two questions which have to be answered when an asylum seeker suspected of being a minor has arrived are whether the asylum seeker is under the age of 18 and whether he or she is accompanied or not. If an application is handed in at one of the airports, Zürich or Geneva, the applicant has to stay at the airport’s transit zone for the whole procedure. Both airports have their own EVZ and the SEM decides there, during the stay in the transit zone, whether an applicant is allowed to enter Switzerland. In case of a ‘Dismiss an Application without Entering into the Substance of the Case’ (DAWES) or a negative

2_{In this analysis, the French version of chapter C10 of the asylum and return compendium is used with support}

(19)

13 decision, the applicant is deported from the transit zone. A DAWES decision is the case when Switzer-land does not hold responsibility for the applicant, most of the time due to the fact that another Dublin state holds responsibility or the applicant previously stayed in another safe country. If the application is not submitted at an airport, the applicant will be brought to an EVZ with a maximum stay of 90 days. This is also the case for UMAs older than 14 years. UMAs younger than 14 years old, should as far as possible be accommodated privately. From the period of time in the EVZ or at the airport on, the UMA has the right to a guardian. If no guardian is available for the UMA immediately, the cantonal authority has to determine a reliable person who takes care of the UMA for the interim.

As soon as the application is submitted, the examination procedure starts with a preparatory phase which should last no longer than three weeks. During a short questioning, the applicant is asked about his or her identity, the reasons for the application, fingerprints are taken and information on the ap-plicant is conducted. After the brief questioning, the SEM decides on whether Switzerland is responsi-ble for the case or whether the applicant has be returned to another Dublin state or a safe third country. If Switzerland holds responsibility, a first interview will be conducted. During this first hearing, the applicant is interviewed on the reasons for asylum. For the hearing, the UMA, discerning or not, has the right to be supported by a legal representative, taking care of the UMAs interests and rights. How-ever, the legal representative does not have to be present at the hearing (Handbuch Asyl und Rückkehr 2015: 16). After this hearing, the SEM takes a second decision. The decision can either be a DAWES, a negative or a positive decision or a temporary admission. In case of a positive decision (B status), a temporary admission (F status) or if a decision cannot be taken based on the first interview (N status), the applicant is handed over to the canton who takes care of all the living conditions during the out-standing examination procedure. On the cantonal level, the accommodation situation for UMAs differs from canton to canton. Some cantons provide special locations for UMAs as for example Basel City or Bern does, other cantons host them in private accommodations or in a children’s home. However, there are also still cantons who accommodate UMAs in normal reception centres for asylum seekers which should be avoided according to the UN Committee on the rights of the child (Tages-Anzeiger 7.10.2015: Vertrieben, allein, minderjährig; Art 69, CRC/C/CHE/2–4 2015).

If the application gets rejected, the SEM additionally has to check whether a deportation to the country of origin is ‘permitted, reasonable and possible’. The removal must not be in contradiction with the CRC principles (Handbuch Asyl und Rückkehr 2015: 18). In order to determine the factor ‘reasonable’, the SEM has to take into account age, maturity, degree of dependence, relation to the guardian, re-sources of the guardian, education, degree of integration in relation to the duration of the stay in Switzerland and chances as well as risks of reintegrating in the country of origin (Handbuch Asyl und Rückkehr 2015: 19). Is one of the three factors not given, the applicant cannot be deported and gets a temporary admission. Is the removal permitted, reasonable and possible, the SEM enforces a removal order. Against the asylum decision and the removal order, an appeal within a certain period of time,

(20)

varying between 5 and 30 days depending on the type of decision, can be submitted (SFH 2016: Asylum procedure).

To sum up, after having identified an UMA, the SEM basically has to take three decisions. First, the SEM has to decide whether Switzerland holds responsibility of the case or whether an application has to be dismissed (DAWES). Second, if DAWES is not the case, the SEM has to make a material decision based on the applicant’s refugee qualities. And third, if a case gets rejected, the SEM has to decide whether a removal back to the country of origin is permitted, reasonable and possible.

In the following section, on the one hand the test phase will be introduced. The test phase is a current project which is testing new regulations of the asylum procedure in a special reception centre. On the other hand, two aspects exclusively relevant for UMAs are introduced. These are the age determina-tion test and the role of the guardian.

3.1 Test phase

With the urgent changes of the AslyG from the 28. September 2012, a test phase has been imple-mented in order to make the asylum procedure faster (TestV, see Appendix II). The test phase initially should last no longer than two years but has been extended until the 28. September 2019. A separate reception centre in Zürich is testing the conditions of the changes of the AsylG. All asylum seekers, also UMAs, are randomly distributed between the initial reception centres and the special reception centre of the test phase. Applicants being allocated to the test phase must not have any advantages or disad-vantages in the outcome of the procedure compared to the other applicants (Art 6, TestV). The maxi-mum length of stay in the test centre is 140 days (§ VI, Art 9, TestV). The centre is mainly used for the preparatory phase of the asylum procedure.

The goal of the test phase is to speed up the application procedure. Therefore, the single phases during the procedure are shortened in the test phase. The preparatory phase lasts for Dublin proceedings up to 10 days, for the other proceedings up to 21 days. After the preparatory phase ends, the accelerated proceedings start which last between 8 to 10 working days. If a case cannot be dealt with within the accelerated procedure, the case will switch to the initial procedure. The time for lodging an appeal in the test centre is 10 days (Art 38, TestV).

UMAs are accommodated in a separate part of the centre, however, still in the same centre as the other asylum seekers. During their stay in the test centre, the UMAs also have the right to a guardian. However, the role of the guardian will be carried out by a legal representative and not by a separate person (§ I, Art 5, TestV). As soon as the minor is handed over to a canton, an original guardian takes over. This guardian has the same functions as in the normal asylum procedure. (§ II, Art 5, TestV).

(21)

15 Considering the special conditions of the test phase, a few aspects need to be discussed further. First of all, it can be problematic to accommodate UMAs in the same reception centres as other asylum seekers. UMAs are extremely vulnerable and in need of special protection. Therefore, it is not ade-quate to mix them in the same centre with all the other asylum seekers. The situation of being in the asylum procedure, not knowing what is going to happen or still waiting for a decision can be psycho-logically and physically challenging and can lead to tensions within the centres. Minors should not get into such a situation and should be protected against any kind of violation or threatening.

Second, the role of the guardian seems to be rather problematic. Legal representatives have to, be-sides informing and consulting asylum seekers, attending hearings, giving statements on rejected ap-plications and writing petitions of review, represent the interests of the UMA in the test centre (Art 28, TestV). Besides the fact that legal representatives do not have the necessary knowledge to carry out this job, the specific criteria of what exactly the role of the guardian implies are not defined. The TestV does not state clearly what is meant by looking after the minor’s interests. Additionally, although the legal representative might be able to take care of the UMA’s legal rights, an UMA’s rights go beyond legal rights. Article 22 of the CRC clearly states that a child seeking asylum has the right to enjoy the rights set forth in the convention. This, however, might not be the case if the role of the guardian is carried out by a legal representative. Furthermore, as soon as the canton holds the responsibility for the minor, the person of trust switches from the legal representative of the test centre to a ‘normal’ guardian provided by the canton. This switch could also be problematic as an UMA with potentially traumatic experiences and without a family environment is especially in need of stability and support. In 2016, a national vote on the amendments of the test phase took place. The vote contained that additional test centres should be build, taking over the amendments which are introduced in the al-ready existing test centre. The national vote was adopted with a majority of 66.8, although at the date of the vote, a Swiss newspaper published an article on UMAs in test centres being forced to do a genital test. This controversial method is used in addition to the in the next section explained age determina-tion test. A physician visually checks the development of the genitals to determine the age of the minor and support the controversial age determination test (Tages-Anzeiger 5.6.2016: Junge Flüchtlinge müssen zum Genitaltest). With the adoption of the amendments, the genital test will be extended to the additional test centres which will be build.

3.2 Age determination test

An UMA has to prove his or her age in order to get the specific protection for minors. Does the SEM doubt the claim of the minor’s age, an age determination test can be carried out (§ III bis, Art 17, AsylG). The asylum and return compendium distinguishes between strong and weak pieces of evidence. While

(22)

identification papers, acknowledgement of reasons for lack of identification papers or acknowledge-ment of stateacknowledge-ments on age are seen as strong pieces of evidence, physical appearance and x-ray tests on bones are seen as weak pieces of evidence. Considering the x-ray hand bone test, if the result pre-sents an age older than 19, the applicant is not categorized as a minor. This is due to the fact that a difference up to three years is usual, according to the compendium (Handbuch Asyl und Rückkehr 2015d: 14). In other words, if an applicant’s age is guessed at 19, he or she can be between 18 and 20 years old. The compendium states that the x-ray test is not a reliable proof and therefore not sufficient to certainly determine an applicant’s age. Although the Swiss government admits this and also from a scientific point of view (Derluyn and Broekaert 2007: 321; Bhabha and Young 1999: 91-92), using this method is doubtful, it is still used. In relation to this, a current media debate broke out after data was published on the age distribution of asylum seekers. This data showed a peak at the age of eighteen. Several newspapers criticized the SEM therefore, as this could reflect a systematic over-estimation of applicants’ age. This could be in the SEM’s interest, as these applicants then do not have the right to child-special protection which safes official authorities’ resources (Tages-Anzeiger 12.5.2016: Auf-fallend viele 18-jährige unter Asylbewerbern).

3.3 Role of the guardian

According to international standards and national law, every UMA has the right to a guardian, taking care of the UMAs’ interests. The Dublin III regulation states that member states have to ensure a rep-resentative for the UMA who assists and represents the minor with respect to all procedures (§ II, Art 6, Dublin III 2013). The regulation defines a representative as “a person or an organisation appointed by the competent bodies in order to assist and represent an unaccompanied minor in procedures pro-vided for in this Regulation with a view to ensuring the best interests of the child and exercising legal capacity for the minor where necessary” (Dublin III 2013: L180/36).

On the national level, the AsylG states that the federal council determines the role, competence and tasks of the guardian and regulates the access to legal representation in the reception centres (VI, Art 17). The guardian is determined by cantonal authorities and will participate at the airport or the EVZ if for the further procedure crucial steps are taken, during the procedure at the cantonal level and during the Dublin procedure. A more detailed description of the role of the guardian is in the AsylV 1. The AsylV 1 states that the guardian within the Dublin procedure needs to have knowledge on the federal law on asylum and the rights relating to the Dublin procedure. The guardian accommodates and sup-ports the minor during the asylum or Dublin procedure and fulfils the tasks of giving advice before and during the questioning, support by the obtaining of evidence and assistance while dealing with local authorities and of dealing with the health system (§ III, Art 7, AsylV 1). In addition to these legal regu-lations, the asylum and return compendium defines the role of the guardian as follows: the role of the

(23)

17 guardian is very diverse, covering the representation of the minor’s interests, administrative tasks and organizational tasks like following the legal procedure or organizing medical care. In addition to this, the guardian has to as far as possible be able to be reached by the minor (Handbuch Asyl und Rückkehr 2015: 9).

After this overview over the Swiss asylum procedure and the special treatments for UMAS, the next chapter contains the methodological part of the analysis. After a brief instruction of the methodology, the method, the data collection and the operationalization are introduced.

(24)

4 Methodological framework

The underlying research question of this thesis required a combination of a qualitative and a quantita-tive analysis. Therefore, the study relies on a mixed-methods approach: A change-tracking policy

anal-ysis and a descriptive statistical analanal-ysis of the drivers of policy changes.

The purpose of the change-tracking policy analysis is to determine the migration policy changes and their relative restrictiveness based on international and national policies. Within the change-tracking policy analysis, a qualitative and a quantitative approach are applied. The qualitative approach serves to describe in detail and understand the evolution of migration policies and their changes in restric-tiveness and to add meaning to the graphs based on the quantitative approach. The quantitative ap-proach serves to determine the changes in restrictiveness. It supports the qualitative analysis with a visualization of the policy changes and adds to determining the drivers behind the changes in migration policy restrictiveness. Object of inquiry are specific migration policies on the national and international level concerning UMAs in Switzerland from 1951 to 2015. By using a systematic approach, the written policies are quantified into codes. The policy documents are examined in detail and compared with each other in chronological order to determine moments of change, identify the type of change and whether it increases or decreases restrictiveness. The analysis mainly relies on two sources of data, the already existing DEMIG POLICY database (Determinants of International Migration) and documents collected directly from official authorities’ websites.

The descriptive statistics are aimed at explaining the drivers of the changes observed in the first set of analysis. To determine the drivers of the policy changes, inferential statistical analysis of the data such as a regression was considered. However, the focus on one country and the availability of data only since 1995 mean that the requirements for the statistical analysis cannot be met. Therefore, descrip-tive statistics, in the form of graphs and correlations, are used alternadescrip-tively. Although the correlations clearly cannot provide the same amount of information as a regression analysis would, they seem to be the most accurate way to go under the given circumstances. The dependent variable is the relative

restrictiveness of migration policies concerning UMAs. The main independent variables derived from

the theory and previous research are previous migration flows, the share of migration stock, GDP, unemployment rate, strength of SVP and the recognition rate. The figures for the independent varia-bles are based on the data from the Bundesamt für Statistik (BFS, Swiss federal statistical office) and the SEM.

4.1 Justification of method

Using the method of a change-tracking policy analysis turned out to be the most accurate method to determine how in Switzerland migration policies concerning UMAs have evolved over time. As the focus of this analysis is not on the comparison of different countries but on one specific target group

(25)

19 within one country, a comparative policy analysis approach did not seem to make much sense. The change-tracking policy analysis allows to determine how migration policies have changed over time and allows to include country specific details. As the restrictiveness measured according to this ap-proach is always relative to the previous situation and does not provide an absolute measure, this allows for a more detailed analysis of the development of the policies and in which direction the re-strictiveness is evolving. Furthermore, using a historical approach helps to understand how a situation became to be as it currently is and can provide important background information which is necessary to properly understand the dynamics of migration and the role of the state.

Within the change-tracking analysis, a more qualitative and a more quantitative approach are applied. By using a qualitative change-tracking policy analysis, the opportunity is given to examine in detail the single policy changes and how they differ from the previous situation. This approach is helpful as it adds to the in-depth understanding of the policies’ evolvement by gathering far richer information compared to only using a quantitative coding scheme. However, as one aim of this analysis is to deter-mine the drivers of the policies’ evolvement, and in order to support the qualitative analysis, a more quantitative coding scheme is applied additionally. The coding scheme helps to visualize the evolve-ment of the policies and breaks down the qualitatively gathered information to a comparable level. The underlying idea of combining both approaches is that this leads to a better understanding of the object of inquiry. While the qualitative analysis provides rich and detailed information, this type of measurement does not provide solidly comparable data. Whereas, the quantitative coding does pro-vide comparable data but probably leaves aside important details. Therefore, a combination of the two approaches seems most adequate.

4.2 Data collection

Most of the data was collected directly from the official authorities’ websites. Besides collecting rele-vant documents independently, the DEMIG POLICY database was used to identify relerele-vant policies. The DEMIG POLICY database has been constructed between 2010 and 2014 and is part of the DEMIG project (Determinants of International Migration: A Theoretical and Empirical Assessment of Policy, Origin, and Destination Effects) (de Haas et al. 2015: 2). The database provides migration policy changes from 45 countries around the world, most of the time covering a timeframe from 1945 to 2013. The CRSR of 1951 was determined as the take-off year as this policy can be seen as an important change for refugees and asylum seekers.

The DEMIG POLICY database focuses on ‘policies on paper’ which are the laws, regulations and measures enacted by the state to regulate migration. The database does not include policy discourses and the implementation of policies (de Haas et al. 2015: 4). The same approach was adopted for the current analysis. Regarding policies on paper, national and international policies are taken into account.

(26)

As the DEMIG POLICY database mainly focuses on national policies, relevant international policies are included additionally. These policy documents are crucial as they often define minimum standards the member states have to apply and therefore can be suspected of having an impact on the situation of the UMAs. Additionally, international policies were examined to answer the question to which extent national policies meet international standards. The international policies are usually accessible on the official websites of the responsible organization like the UN or the EU.

Regarding national policies, besides the DEMIG POLICY database, the Asylgesetz (AsylG) and its changes are taken into account. The AsylG is the actual legislation, to a large extent regulating the rights and restrictions of the UMAs. It is the most crucial policy document determining the UMAs’ rights. Additionally, the AsylG shows how international regulations are implemented in national law. While the DEMIG POLICY database includes the most influential changes of the AsylG, all versions of the AsylG of 1998 are examined here. As not all versions of the old AsylG of 1979 were not accessible, for the old AsylG only the most important amendments were included. Because data on the independ-ent variables often only cover the last 20 years or less, focusing on the new AsylG of 1998 seems to make sense. Besides the AsylG, some other national policy documents are added to the DEMIG sample which are considered as being relevant for the UMAs. Most of these documents can be found on the official homepage of the federal council admin.ch. Data on the migration flows and other independent variables are from the BFS and the SEM.

One other aspect which is relevant to mention in relation to the DEMIG POLICY database is the fact that the database identifies the specific migration groups who are targeted by every single policy. However, the database does not include the target group of UMAs. Therefore, to collect the relevant documents out of the database, two alternative migrant categories were chosen: refugees, asylum

seekers and other vulnerable people and all migrants. Furthermore, the DEMIG POLICY database

con-tains four policy areas: border and land control, legal entry and stay, integration and exit. Considering the two migrant categories mentioned above, most migration policy changes within the time-period of interest are within the legal entry and stay category. Within the given time-frame, the database provides 33 relevant policy changes for UMAs. What has to be mentioned, however, is that for the database, individual changes within one amendment were coded. This differs from the current analysis, where several changes within one amendment were coded as one change according to the overall effect. In other words, a policy document of the current analysis is likely to include more than one policy change which would be listed separately in the DEMIG POLICY database. Most of the relevant documents contained in the DEMIG POLICY database were retrieved directly from the official author-ity’s website admin.ch. A few older documents were not accessible anymore. The information of these documents is then based on the DEMIG POLICY database, newspaper articles and scientific articles.

(27)

21

4.3 Types of measurement

This section contains the measurement of the dependent variable restrictiveness and the different in-dependent variables. The variable of changes in policy restrictiveness was coded according to the scheme as of the DEMIG POLICY database. The coding scheme was taken over from Mayda’s (2010: 1262) research on a panel data analysis of 14 OECD countries on the determinants of bilateral migra-tion flows between 1980 and 1995. A simple +1/0/-1 coding scheme was applied. The scheme is guided by five criteria based on DEMIG which can be found in Appendix VI. They were adapted to the target group of the UMAs. According to these criteria, if a migration policy has become more restrictive, the restrictiveness is coded with +1. If a migration policy has become less restrictive, the restrictiveness is coded with -1. And if migration policy overall has not become more or less restrictive, the restrictive-ness is coded with 0 (de Haas et al. 2015). While coding the changes in restrictiverestrictive-ness, a subjective discretion always plays a role. However, by using the same coding scheme as for the DEMIG POLICY database, subjective discretion should be minimized. In this way, by maximising transparency, the re-liability and consistency of the data should be increased (de Haas et al. 2015: 3). Additionally, by taking over the measurement of the DEMIG POLICY database, the current analysis might add to the existing database which is in line with the database’s philosophy (de Haas et al. 2015: 17-18).

A few aspects have to be mentioned additionally. First of all, it is important to notice that the restric-tiveness coded this way is based on an ordinal scale. This means that the restricrestric-tiveness is just relative to the previous policy document and does not say anything about the absolute restrictiveness of the policies. Furthermore, the relative restrictiveness is measured per policy document. This means that the coding reflects an overall effect which does not automatically mean that only more or only less restrictive components are included in the coded policy. It might be that for example a policy docu-ment includes two aspects increasing the restrictiveness and one aspect decreasing the restrictiveness. This policy change will then be coded as increasing restrictiveness. This approach is based on the cri-terion that changes in restrictiveness trends are the main interest of the analysis. Rather than looking at single aspects of migration policies, the focus is on the general changes of the situation for the UMAs in Switzerland.

In addition to this, another aspects has to be mentioned. The coding of the change in restrictiveness does not say anything about the dimension of the change. A small change of restrictiveness towards more restrictive will be coded the same as a major change towards increasing restrictiveness. As this inevitably leads to a loss of information, a detailed descriptive overview over the policy changes is provided in the change-tracking analysis. However, to reach a solid sample with comparable data, the focus lies on the direction of the change of restrictiveness instead of the dimension of the change.

(28)

The most crucial independent variable is the previous migration flows. Two different variations of this variable are included in the correlations. On the one hand, the migration flow of the UMAs and on the other hand the migration flow of all asylum seekers. Both variations of migration flows are measured by the number of yearly asylum applications of the specific group in Switzerland. The data on the mi-gration flows of the UMAs is provided by the Central System of Mimi-gration Information (ZEMIS). The data is, however, only available since 2004. Data on the number of applications of all asylum seekers is available since 1986. Besides previous migration flows, the share of migration stock might play a role as well. This variable again contains two possible variations. On the one hand it includes the annual share of the migration stock, meaning all foreigners living in Switzerland. On the other hand, the share contains former asylum seekers, currently having a (temporary) residence permit, either status B3 (ref-ugees) or F (temporary admission). The data on the share of all migrants is available since 1981 and the data on the share of foreigners with a B or F status since 1995.

Additional variables with a possible impact are the economic situation, the recognition rate and the political orientation of the government. The economic situation can, according to de Haas and Natter (2015) be measured by two indicators, the GDP and the unemployment rate. The GDP in the current analysis is the annual GDP per capita in Swiss Francs which is provided by the BFS for the years 1990-2014. The unemployment rate is defined as the percentage of persons being registered at a govern-mental employment office in relation to the permanent Swiss population available since 2005. The recognition rate is the percentage of all UMAs getting a refugee status (status B). The temporary ad-mission rate is the percentage of all UMAs getting a temporary adad-mission (status F). The total recogni-tion rate is the summary of both rates. For the current analysis, the total recognirecogni-tion rate for UMAs since 2004 is used. And finally, the political orientation is measured with regard to the Swiss People’s Party (SVP). The SVP predominates the issue of migration and, besides being the strongest political party in the parliament, they achieved a swing to the right of most political parties. Therefore, it is by far the most influential political party regarding the issue of migration and asylum. The SVP is measured as the percentage of seats the SVP has in the national council which is available since 1971.

4.4 Data description

After having collected and selected the data, a dataset with 60 policy documents was the result. Each policy document was studied and the changes in policy restrictiveness coded. International policies and the changes of the federal law on asylum AsylG were coded separately and the relative restrictive-ness was determined in relation to the previous version. Additional national policies relevant for the

3_{In the data provided by the BFS, the status B group is listed together with the status Ci group. The Ci status is}

given to close relatives of high skilled migrants temporarily working in Switzerland. As this group can be consid-ered as insignificantly small, these figures are nonetheless used for the status B group.