Hostname: page-component-586b7cd67f-dsjbd Total loading time: 0 Render date: 2024-11-24T23:35:02.718Z Has data issue: false hasContentIssue false
  • English
  • Français

The Dublin Convention and the Introduction of the ‘First Entry Rule’ in the Allocation of Asylum Seekers in Europe

Published online by Cambridge University Press:  25 April 2022

Gaia Lott Open the ORCID record for Gaia Lott [Opens in a new window]
Gaia Lott*
Affiliation:
History and Civilisation Department, European University Institute, Fiesole, Italy
*
[email protected]; [email protected]
Rights & Permissions [Opens in a new window]

Abstract

The Dublin Convention (1990) was the first binding agreement on asylum between the member states of the European Community. It defined the criteria that determined responsibility for the examination of asylum applications lodged in their territories. Given the contemporary discussions about the system that derived from it, the paper reflects on one of its main criteria: the first entry principle. Drawing on archival research in Belgium, France and the United Kingdom, the essay shows how the inclusion of the first entry principle was far more than a matter of course. It was influenced by the Schengen process and the establishment of the single market, previous North-South tensions over migratory issues, and governments’ (in)capacity to predict future developments. The inclusion of the first entry principle contributed to assimilating asylum policy with migration control, creating the premises for the subsequent burden-sharing problems and readmission agreement practices.

Type
Article
Information
Contemporary European History , Volume 32 , Issue 3 , August 2023 , pp. 459 - 474
Check for updates
Copyright
Copyright © The Author(s), 2022. Published by Cambridge University Press

Introduction

On 15 June 1990, eleven of the twelve member states of the European Community (West Germany, France, Italy, Belgium, the Netherlands, Luxemburg, Ireland, the United Kingdom, Greece, Spain and Portugal) signed the Dublin Convention, the first binding agreement on asylum between European governments (the twelfth, Denmark, signed it a year later).Footnote 1 Dublin defined the criteria that determined responsibility for the examination of asylum applications. For the first time, it put down in black and white the ideas and principles that are still at the core of the allocation of asylum seekers in Europe. On the thirtieth anniversary of the convention, the debate about the system that derived from it is still open.Footnote 2 This paper aims to reflect on the historical roots of this system, retracing the interests and ideas that prevailed during the negotiations between the twelve and reflecting on the inclusion of one of the most relevant and debated criteria established by the convention: the first entry rule.

Reflecting on ‘the way to Dublin I’ and on the establishment of the first entry rule is a significant study for several reasons. First, the Dublin Convention is a cornerstone of the present ‘Dublin regime’; understanding the points of view that were included and excluded in its drafting, and why, enables us to better understand the problems that currently surround it. Second, the lack of historical literature on the issue makes addressing it urgent. Most of the existing wisdom looks at the Dublin Convention and the inclusion of the first entry rule from the perspective of refugee law or from the interpretations of the political theory of inter-state dynamics.Footnote 3 These studies do not explore the historical path that led to the signing of the convention, by which I mean a reconstruction of the roles and stakes of the different actors and their mutual influences, starting from primary sources and archival documents. Many scholars, including historians, have reflected on the birth of the European migratory system,Footnote 4 but despite this none of their analysis looks specifically at the birth of the asylum system in Europe, which is a distinct part of the broader migratory regime, characterised by specific norms and values.Footnote 5 This study aims to fill this gap. Specifically, it backs the argument that the Dublin negotiations and the inclusion of the first entry principle were heavily influenced by the imminent establishment of the single market and the Schengen process, supporting this thesis with unpublished primary sources not previously accessed by scholars.Footnote 6 Thanks to these original primary sources, I show how the inclusion of the first entry principle as a rule for allocating asylum seekers cannot be taken for granted, given previous and ongoing discussions in other fora and the very nature of the asylum issue, which is strictly connected to human rights and humanitarian concerns. I also demonstrate how other dynamics, not yet considered, played a role in this choice, such as North-South (more than East-West) tensions over migratory issues, which dated back to previous decades, and the (in)capacity of governments to predict future arrivals and the means of control that might be available. I show too how the inclusion of the first entry principle was crucial to the development of European asylum policies, from a conceptual and a practical point of view: it contributed to the assimilation of asylum policy with migration control and created the premises for the subsequent problems over burden sharing and readmission practices.

Finally, this project makes an interesting contribution to the historiography of European integration, by showing how and why intergovernmental paths prevailed over community tracks in the asylum field. This is, once again, of very specific interest, because asylum is strictly connected to human rights and humanitarian considerations as well as to European identity and values. I demonstrate how the twelve found themselves indulging in inter-state dynamics and tensions in the allocation of responsibility for asylum applications, rather than following a human rights approach, which would affect the path followed by European asylum policies up until today. This reconstruction of the historical dynamics that determined the adoption of the first entry rule among the criteria to allocate responsibility for asylum applications, including the analysis of earlier and ongoing discussions on such principles in other fora, is also unprecedented and of great relevance, given it is the subject of contemporary debates.Footnote 7

This contribution is based on both primary and secondary sources. Archival documents of the European Parliament, the European Commission, the Council of the European Union, the Council of Europe and selected member states (France and the United Kingdom) are among the primary sources. The choice of these member states derives from their primary role, together with West Germany, in the elaboration of the Dublin Convention. The study suffers from the limited reference to German primary sources; what is included comes from the archives of the EC institutions and of the Council of Europe. However, the multi-archival research (in different states and institutions) and the multiple perspective that derives from it stands as a guarantee of the reliability and thoroughness of my analysis. Furthermore, some of the secondary sources I analyse have investigated the German perspective.

The Start of the Dublin Negotiations

By the mid-1980s, the need for international cooperation on asylum emerged from the implementation of restrictive measures at the national level and the European integration process.

The combination of increasing arrivals (the twelve hosted around 65,000 asylum seekers in 1983, rising to 340,000 in 1990)Footnote 8 and the growing perception among experts and specialists of an abusive use of asylumFootnote 9 brought the twelve to introduce restrictive policies at the national level: from access to territory (visa requirements and sanctions against carriers bringing in undocumented persons) to the procedures for determining refugee status (policy of deterrence); from a stricter interpretation of the Geneva definition of refugee to the reduction of rights and benefits granted to asylum seekers and refugees (policy of dissuasion).Footnote 10 The immediate effect of these measures was to divert the flow of asylum seekers from one country to another. Despite the research that followed, showing how asylum seekers’ choice of their destination country depended on several factors (ranging from economic perspectives to migratory paths, cultural and linguistic links, diaspora ties, etc.),Footnote 11 by the mid-1980s member states were obsessed by the idea that what was enacted in neighbouring countries would affect the arrival of asylum seekers in their territories and thus competed with each other for the lowest numbers.Footnote 12

The new, lively phase of the European integration process reinforced the concern of member states over future arrivals: the achievement of the single market, meaning an area without internal frontiers in which the free movement of goods, persons, services and capital was ensured (article 8a of the EEC Treaty, as amended by the Single European Act), risked leading to a growth in the secondary movements of migrants (including asylum seekers) between member states. As stated explicitly in the Political Declaration annexed to the Single European Act, the establishment of free movement of persons inside the EC urged member states to reflect on measures to compensate for this, for example by establishing rules of entry for third-country nationals. Despite the differing interpretations of article 8a, especially concerning free movement of persons, the exchanges of the twelve on asylum avoided paralysing discussions and followed a solution-oriented approach. The twelve established the one-state responsibility principle, meaning that only one member state was to be responsible for asylum applications lodged in their territory, and focused on the criteria for determining such responsibility, without clarifying whether there should be an ‘abolition’ or an ‘easing’ of controls at EC internal frontiers or the extension of such measures to third-country nationals.Footnote 13

Despite the European Commission's proposal of a draft directive on asylum by 1988, member states chose the intergovernmental track to deal with it and in October 1986 established the Ad Hoc Group on Immigration, whose asylum subgroup would draft the Dublin Convention. The Ad Hoc Group was composed of officials from the Ministry of the Interior and of Justice responsible for asylum and derived its experience from the cooperation on security and migratory issues within the Trevi Group in the 1970s. Periodically, the Ad Hoc and Trevi groups held joint sessions at a ministerial or technical level, but they remained distinct from an organisational point of view.Footnote 14

States had always considered asylum at the core of their sovereignty rights, given its link with the power of governments to establish rules of entry for third-country nationals and asylum's connection with national values and culture.Footnote 15 The General Declaration on articles 13 and 19, annexed to the Single European Act, expressed well the protective jealousy member states had for their sovereign rights in the field. The French and the British were the strongest opponents of Community competence. The British stressed in 1988 that one of the aims of the Ad Hoc Group was to ‘divert the European Commission from its intention of intervention’, because asylum had to remain an intergovernmental issue.Footnote 16 The interest in a quick solution, combined with the principle of subsidiarity promoted by the president of the European Commission at the time, Jacques Delors, led the commission to indulge the twelve and withdraw the draft directive;Footnote 17 the European Commission was permitted to participate as an observer in the meetings of the twelve. This, together with the references to the Community System by the Dublin Convention, encouraged scholars to define the work of the Ad Hoc Group as a ‘pioneer of the third pillar approach’.Footnote 18

The inclusion of asylum in the agenda of the Ad Hoc Group on Immigration should not be assumed. If migration was part of the general concern of member states by the mid-1980s, asylum was not, drawing the attention only of specialists. We have to wait until the early 1990s, with the end of the Cold War and the exodus of over a million people from Eastern Europe, Albania and Yugoslavia, to see the ‘asylum crisis’ perceived as such by the twelve. By the mid-1980s, only West Germany (and to a lesser extent Denmark) felt overburdened by the growing numbers of asylum seekers in its territory and saw asylum as a domestic problem.Footnote 19 West Germany received almost 60 per cent of all the applications lodged in the territory of the twelve, registering around 650,000 applications between 1985 and 1990, hosting more than 100,000 asylum seekers per year from 1988 (compare this with France, the second member state for asylum numbers, which registered fewer than 222,000 applications between 1985 and 1990 and hosted between 30,000 and 60,000 asylum seekers per year from 1988).Footnote 20

Germany's geographical proximity to Eastern Europe played a specific role in attracting applicants; the open-border with East Berlin until the end of 1986, the economic prospects it offered and its liberal constitutional provision on asylum, which was considered a sacred post-Nazi heritage, had all played their part.Footnote 21 The growing numbers, together with the decentralised management of the phenomenon, which allowed other actors (Länder [constituent or federal states], especially Bavaria) in addition to the central government to raise their voice, opened a broad public discussion on refugees by the mid-1980s.Footnote 22 Besides asylum seekers, West Germany hosted a large guest worker population, which remained in the country with their families after the migration closure of the 1970s, and many ethnic Germans returned to the country at the end of the Second World War. This composite immigrant population reinforced the perception of being overburdened, fuelling public discussions.Footnote 23 However, the debate was stuck in a stalemate because of the symbolic value of the constitutional right of asylum, which seemed to prevent any change.Footnote 24 Given these internal difficulties, Bonn tried to exploit the EC context to legitimise the enaction of measures in the asylum field, which would have been hard to approve at a national level. According to the British and the French Ministry of Foreign Affairs, by the mid-1980s the Germans insisted that asylum was given a high profile and encouraged discussion of it in ‘all possible meetings’, including in the Ad Hoc Group, in the European Political Cooperation meetings and in the Council of Europe.Footnote 25 In 1991, Chancellor Kohl claimed there was a link between EC asylum policy and constitutional revision.Footnote 26 Against this background, Bonn can be considered the driving force behind the start of the Dublin negotiations.

The Cold War and the Dublin Negotiations

It is worth reflecting on the impact of Cold War dynamics and the events of autumn 1989 on the Dublin negotiations. Cold War dynamics were intertwined with asylum policies in Europe from the start: authoritarian regimes in Eastern Europe had determined the flow of asylum seekers from or transiting through their territories to those of the twelve since the 1950s. Despite this, European asylum policies and debates had always been less influenced by Cold War dynamics compared with other contexts, such as that of the United States.Footnote 27

By the second half of the 1980s the increasing flow of people fleeing the collapsing Soviet Union gave rise to new concerns, especially in the countries that bordered on the region (West Germany).Footnote 28 However, until 1991–2 no discussions took place inside the Ad Hoc Group and events in Eastern Europe at the end of the 1980s did not determine changes to the agenda of the negotiations of the Dublin Convention.Footnote 29 This is understandable in light of the narrow and technical approach of the asylum subgroup, which rarely linked its reasoning to broader themes like foreign policy. Other more political settings were sensitive and debated the point – the Visa subgroup or European Political Cooperation, for example.Footnote 30 But as I will demonstrate, some passages of the Dublin Convention (article 3.5 in particular, which enables the twelve to send asylum seekers back to third-country nations) took the ‘eastern concern’ into consideration and created the premise to deal with it in the following years.

The Responsibility Principle

Together with its inclusion in the title of the Dublin Convention, the principle of ‘responsibility’ can be considered the core element of the document. The basic idea conveyed by the convention is that responsibility for the examination of an asylum application belongs to one member state alone and, specifically, to the state ‘that takes the biggest part in the asylum seeker's entrance’.Footnote 31 This ‘biggest part’ can be played voluntarily, according to the so-called authorisation principle (article 5), or involuntarily, connected to the lack of control of national (and community) borders (articles 6 and 7).

The responsibility principle was the first idea around which member states developed their negotiations. Family unity (article 4) and cultural links (article 9) were added in June 1987, following requests from Denmark and Spain, the former promoting the Scandinavian approach to family and cultural rights, the latter subjected to strong pressure from the Latin American community.Footnote 32 Even though family unity came first according to the treaty, it was considered an exception to the responsibility principle: articles 5 and 6 were ‘really the nub of the whole Convention’.Footnote 33

The dominant role of the responsibility principle was related strictly to the process of European integration and the establishment of the single market: the need to establish rules of access for third-country nationals in an area without internal frontiers entailed responsibility to check external (community) borders. What emerges clearly from the discussions among the twelve is the fear of some ‘destination countries’ (France and Denmark above all) becoming overburdened because of their generous asylum policies, alongside the ‘irresponsibility’ of some entry countries, meaning the states lying on the borders of the community. This fear was related to previous tensions that went beyond asylum and focused on broader migration policies. Specifically, Mediterranean governments were in the dock: a 1989 ISOPLAN survey, commissioned by the European Commission, had shown that the possibility of irregular entry and stay (because of a lack of monitoring and control by national authorities), combined with the low level of rights and benefits recognised for migrants in southern Europe, incentivised secondary movements and the region risked becoming a ‘waiting room’ for northern Europe.Footnote 34

Against this backdrop, the negotiations between the twelve and European integration were exploited to force ‘irresponsible states’ to start implementing a ‘proper migration and asylum policy’. The Dublin Convention was perceived as a means to impose responsibility on those states that, until then, had not practised responsibility, thus shifting the problem from one country to another. The combination of the principle of one-state responsibility and the introduction of common rules for allocating asylum seekers was meant to bridge this gap between northern and southern states.

Italy was the first defendant, accused of behaving as a ‘cavalier seul’ because of its ‘open door’ migration policy, exerting limited control over its borders and over people inside its territory ‘to the detriment of northern neighbours’ (France, above all).Footnote 35 Concerning the specifics of asylum, by the mid-1980s Italy was the only country among the twelve that had not signed the New York Protocol (1967), which overcame the geographical and temporal limits of the Geneva Convention. This implied that non-European asylum seekers applying for asylum in Italy did not remain in the country but were re-distributed in Western Europe and all over the world. Italian participation in the Dublin process depended on its adherence to the protocol.Footnote 36 Given the geographical proximity to its southern neighbour, France was the member state most concerned about the Italian migration and asylum system. By the mid-1980s, Franco-Italian tension emerged not only in the Ad Hoc Group but also in bilateral meetings; it conditioned Italian inclusion in the contemporary Schengen process, which involved France, West Germany and the Benelux countries.Footnote 37

The First Entry Principle in the Asylum Debate and in Practice

By stating the first entry principle (articles 6.1 and 7.1), that is, the first country entered by non-authorised asylum seekers will be considered responsible for their applications, Dublin represents the first international agreement to put down in black and white the idea that transit in itself (without the asylum seeker's claim for protection and without any other link to a country) entails responsibility for asylum applications.Footnote 38 The Schengen Convention, which was signed in June 1990 to implement the 1985 agreement, approved the same some days later.Footnote 39

The idea that the transit of an asylum seeker in the territory of a member state can determine the state's responsibility for examining his or her application derives from article 31.1 of the Geneva Convention and from the lack of an individual right to asylum in international law.Footnote 40 The underlying assumption of the first entry principle is that applicants have no right to choose their country of asylum and should claim status in the first state in which they arrive.

The debates between the twelve show that the idea of leaving asylum seekers completely free to choose where to apply for protection was quashed by the mid-1980s.Footnote 41 However, the position of other fora that dealt with the same issue at the time was slightly different. The UNHCR, the first reference for any reasoning on refugees, suggested taking the asylum seeker's intention into account ‘as far as possible’; the Council of Europe, active on asylum since the 1960s, referred to a ‘non decisive role . . . not an absolute criterion’ of such intention and to the need to strike a balance between the interests of the first and second asylum countries and of the asylum seeker.Footnote 42 At EC level the European Parliament, whose debates in the 1980s increasingly concerned refugee matters,Footnote 43 defended the right of applicants to choose freely their country of asylum and the draft directive of the European Commission included the asylum seeker's intention in the criteria for allocating responsibility for asylum claims in some cases.Footnote 44

If such a point was then still controversial, the assumption that an asylum seeker had to seek protection in the first country, where possible, and the related link between transit and responsibility on asylum applications were far more than a matter of course in any of these fora.Footnote 45 These two elements called national interests into question as well as the idea of considering the asylum seeker's preferences and relationship with the host country. The discussions inside the fora that dealt with this at the time show the varying nuances of the issue.

Starting with the UNHCR, the organisation made a point of stating that transit in itself could not be considered enough to entail responsibility for asylum applications. The preparatory work of the Geneva Convention had already excluded the requirement for asylum seekers to show that they had not been able to find refuge in one country in order to enter another.Footnote 46 In 1979, the UNHCR underlined how ‘the duration and nature of any sojourn of the asylum seeker in a country had to be taken into consideration’ and that asylum should not be refused ‘solely on the ground that it could be sought from another state’. The organisation did not exclude the possibility of sending an asylum seeker back to another country, but only in case of ‘previous protection . . . connection or close links’ with such state.Footnote 47 The UNHCR's subsequent study of the irregular movement of asylum seekers and refugees came to the conclusion that asylum seekers could be sent back to another country without infringing the non-refoulement principle only if they had left a country where they had received adequate protection, which is different from a transit country. The study was never adopted by the Executive Committee, but this point was recalled in other statements made by the organisation, such as the Conclusion n.58 of 1989, and has been maintained until today.Footnote 48 The relationship between transit and responsibility was vigorously debated inside the UNHCR, producing a somewhat ambiguous position that has survived over time.Footnote 49 The inclusion of the first entry principle in the Dublin Convention would be defined as ‘wholly inappropriate’ by the UNHCR, but it was not over-critical, preferring to wait and see how the convention worked in practice.Footnote 50 Europe and the UNHCR were at a low point in their relationship at the time, which contributed to the UNHCR's distanced position.Footnote 51

At the outset, the Ad Hoc Committee of Experts on the Legal Aspect of Territorial Asylum, Refugees and Stateless Persons (CAHAR), a committee established by the Council of Europe after the failure of the UN Conference on Territorial Asylum in 1977, recalled the UNHCR's reasoning when it included the ‘clause en route’. This clause established that the asylum seeker had to live 100 days in a (transit) country for that country to be responsible for his or her application; something similar to the ‘close links’, which were presented as such.Footnote 52 However, this approach was not unanimously accepted. While the traditional transit countries (Turkey, Greece, Austria and Italy) supported the 100 days provision, other states (Benelux and the Scandinavian countries) agreed with the ‘clause en route’ principle (they had promoted it in the first place) but asked for a shorter period of stay in the transit country combined with a longer period in the destination country. The ‘100 days’ clause was completed with the formula, ‘can be reduced to a period of not less than 30 days’, but this compromise did not convince all parties. France and Switzerland did not agree with the very principle of ‘clause en route’.Footnote 53 The negotiations became stuck in a stalemate and were suspended for almost two years, re-starting only in May 1986. The second round of discussions moved from ‘a revision of concepts . . . a new approach’ promoted by the Swiss, who replaced the ‘clause en route’ with the first entry principle. At the same time, West Germany, partially backed by Denmark, pushed for the introduction of a burden-sharing mechanism into the agreement to encourage acceptance of the first entry principle.Footnote 54 The substitution of the ‘clause en route’ with such provisions (which looked solely at government interests) demonstrates how respect for the asylum seeker's link with the host country was not crucial for the CAHAR delegations. The UNHCR expressed concern about the shift in the group's approach, but it remained unheard.Footnote 55 However, no compromise was found in this case either: from January 1988, several representatives suggested waiting for the outcome of the EC and Schengen deliberations. The two groups had no formal connections with the CAHAR, but the majority of the CAHAR was composed of delegations of the twelve (often the very same officials of the Schengen and Ad Hoc Group). By June 1989 this idea prevailed – discussions were suspended and would never be resumed.Footnote 56

Like the Dublin Convention, the European Commission's draft directive on asylum included the first entry principle (article 9). It also added the right for asylum seekers to legally apply to enter another member state of their choice within three months of their arrival, in the case of applications lodged at borders or in transit, which meant when the asylum seeker had not entered EC territory through a visa or a permit to stay.Footnote 57 Even if this clause was probably difficult to implement, it can be interpreted as an attempt to compensate for the first entry principle by considering the asylum seeker's intention. But such an intention was not taken into account in the case of applications lodged after an irregular stay in a member state, when the first entry principle ruled (article 10). However, the draft directive was shelved, and this principle was never implemented.Footnote 58

In sum, none of the fora discussing and reflecting on the issue at the time took the first entry principle for granted. Quite the opposite – all of them tried to include the asylum seeker's perspective in some form, to take into account his or her preferences/link with the host country, valorising his or her rights.

Before moving on to an analysis of the inclusion of the first entry principle into the Dublin Convention, national practices are worthy of notice. By the mid-1980s, member states did not consider that transit in itself entailed responsibility for asylum applications, requiring a minimum stay of asylum seekers in a country (even a member state) to determine its responsibility for their claims.Footnote 59 Denmark was an exception; it introduced a regulation in 1986 that considered transit enough to entail responsibility for asylum applications, in its efforts to limit the growing number of arrivals.Footnote 60 At the same time, France and the United Kingdom recognised the possibility of rejecting asylum seekers at the border, if they had transited through ‘safe third countries’, where they could have asked for protection. However, rejection was not automatic – the right to appeal was granted (even if not always with suspensory effects) and a specific procedure worked as a guarantee for the asylum seeker.Footnote 61

The First Entry Principle in the Dublin Convention

During the negotiations between the twelve, Greece, Italy and Spain, which were about to become first entry countries (in other historical circumstances, different migration paths or allocation rules made other countries, like Denmark or France, first entry countries), tried to counterbalance the first entry principle by proposing something similar to the ‘clause en route’. Most of the discussions that prolonged the negotiations between the twelve were related to this ‘cut-off point’, as the British defined it, and assorted proposals were put on the table.Footnote 62 Denmark, France and the Netherlands were the most firm in defending the first entry principle, because ‘most asylum seekers arrive by land or sea and therefore from neighbouring European countries’.Footnote 63 A compromise was found in requiring a ‘proof’ of irregular entry and in the addition of paragraph 2 to article 6, which stated: ‘That state shall cease to be responsible, however, if it is proved that the applicant has been living in the Member State where the application for asylum was made at least six months before making his application for asylum. In that case it is the latter Member State which is responsible for examining the application for asylum’. This was a step forward compared with the Schengen Convention, which had not even mentioned or discussed such an option, but it was a solution that left the first entry countries dissatisfied.

The six-month stay required by article 6.2 could be considered an implicit expression of ‘strong links’ with a country. However, the debate between the twelve shows how it was added to take into consideration the concerns of Mediterranean states, rather than to recognise the importance of a link between the asylum seeker and the host country.Footnote 64 The compromise was found thanks to Greece (which proposed a first half-way solution) and France (which managed to find an agreement on it), two states belonging to opposite fronts but which did not defend the strictest positions in their respective groups and were interested in exiting the impasse during their presidency.Footnote 65 Governments more detached from the debate, like the British (because of their isolated geographical position) and the German (because of their entry and destination nature: most asylum seekers entering EC territory via West Germany remained in West Germany), helped the mediation between ‘northern destination’ and ‘southern entry’ states. The British, for instance, tried to convince the Danes by arguing that ‘worse texts . . . e.g. pressure for on arrival appeal’ could be approved: it would have been worse if the compromise for irregular entry had been that responsibility was given in its entirety to destination countries.Footnote 66

Mediterranean states eventually accepted the compromise of article 6.2, not only because of their asymmetrical bargaining power, deriving from their inadequate entry and migration policies, broadly speaking, and the limited numbers of asylum seekers in their territories at the time.Footnote 67 They also accepted the compromise because, by the mid-1980s, most applicants in the EC were ‘over-stayers’, rather than border applicants, and their hope was that entry would be difficult to prove for those who entered irregularly.Footnote 68 By looking backwards to previous migratory paths and means of control, Mediterranean states underestimated the future implications of the Dublin Convention. The perception that the Dublin Convention was part of the European integration process also played a role in the acceptance of the system by southern European states. In addition to the convention's references to the European Community and its treaties, the contemporary Schengen process had a strong influence from this perspective. Given the similar content of the Dublin Convention and Chapter VII of the 1990 Schengen Convention, as well as the prominent role of the Five of Schengen in the Ad Hoc Group, Dublin was perceived to be strictly connected with the Schengen discussions. Schengen was interpreted unanimously as a pioneer of the European integration process, especially in relation to the establishment of the single market, and it created strong pressure on some of the non-participating member states (the Mediterranean countries, above all).Footnote 69 The agreement on the Dublin Convention was seen by these states as an instrument to facilitate their inclusion in the Schengen process, according to the dynamic of a type of package-deal. Since asylum was less strategic than other provisions of the Schengen Convention, the compromise was considered worthwhile.

The agreement on article 6.2 was arrived at just before large inflows of migrants appeared in Western Europe between 1991 and 1992, following the end of the Cold War. The numbers of asylum seekers interested first and foremost West Germany, which had been less involved in the debate over paragraph 6.2. However, what is interesting here is that Bonn did not try to introduce a burden-sharing mechanism in the negotiations over the first entry rule (as it had been trying to do in the CAHAR), which might have helped West Germany manage the situation in the 1990s. This was probably attributable to the CAHAR experience itself, which proved how difficult it was to find a compromise. Given the importance of a speedy agreement between the twelve, Bonn preferred to avoid adding contentious points to the discussions at this stage, relying on a different provision of the Dublin Convention, article 3.5, to deal with arrivals from Eastern Europe.Footnote 70

A final point: article 6.2 requires an amount of time to confirm the responsibility of the destination country. This shows once more that the clause was not introduced to indulge the asylum seeker's intention or his/her link with the host country but to find a compromise between the interests of entry and destination states. At the same time, this compromise reversed the logic followed until that moment by the UNHCR and the Council of Europe, demonstrating the different bargaining power of the governments interested in the measure. The developments that followed would further confirm this last point: irregular entry would turn out to be hard to prove, but supplying evidence of the six-month stay would be even harder.Footnote 71 Faced with these difficulties, the effective operation of the convention came through the introduction of the EURODAC system, a means to trace entry by registering fingerprints in a common European system, which made tracking stays almost irrelevant. The reasoning on its establishment dates back to the Dublin negotiations.Footnote 72

Asylum Seekers and Irregular Immigrants

By the mid-1980s, in all the twelve member states, the first entry principle played a guiding role in determining responsibility for readmission agreements for third-country nationals. Readmission agreements establish procedures for the identification and return of persons (respectively citizens or third-country nationals) who do not, or no longer, fulfil the conditions of entry to, presence or residence in, EU territories. The ‘first generation of readmission agreements’ between EC states, dating back to the 1950s, not only covered people in transit but required a minimum stay (generally a fortnight) in order to ask a state to readmit the migrant. In the second generation of agreements, in the 1980s, the mere ‘transit’ of an irregular migrant in a state became enough to oblige that state to readmit him or her into its territory. Scholars connect this restrictive shift to the broader trend of closure in the migration policy of the time.Footnote 73

What matters for this analysis is that the Dublin Convention extended this practice to asylum seekers. Internal documents, and the conclusions to the ministerial meeting in Copenhagen in December 1987, stated explicitly that in the case of irregular entry, the asylum seeker had to be sent back to their country of entry, ‘notably in the framework of a readmission agreement’.Footnote 74 The same Dublin Convention would be defined as a ‘multilateral readmission agreement’.

Putting asylum seekers together with irregular immigrants is a crucial conceptual shift in European asylum policies. Looking more deeply into the decisions and debates of the 1980s, however, it emerges that the distinction between the two phenomena was more on paper than in actual practice. Not only the rhetoric, including that of internal documents, on ‘bogus asylum’, but also several provisions taken in the migration field reveal this; those on visas and sanctions for carriers (and the growing practice of border detention) did not consider specific clauses or exceptions for asylum seekers.Footnote 75 The concern raised about abusive use of asylum was a constant in the twelve's meetings, especially at ministerial level, and affected their approach. The UNHCR, the Council of Europe, the European Parliament and the same European Commission warned the twelve of the risk of assimilating the two phenomena, but their warnings went unheard.Footnote 76 This assimilation had relevant concrete implications, which risked depriving asylum seekers of the main guarantee granted to refugees by the Geneva Convention (the protection against direct and indirect refoulement). This was also problematic, given the declaratory nature of refugee status, meaning that every asylum seeker is a potential refugee because he or she ‘does not become a refugee because of recognition but is recognized because he is a refugee’.Footnote 77 At the same time, it determined the association of asylum policies with migration control, rather than with human rights. Even if the right of asylum was (and still is) not unanimously considered a human right (unlike the right to seek asylum), it is strictly related to the human rights field.Footnote 78 The choice to assimilate asylum policies with migration control was a result of the historical context of the mid-1980s, but it has shaped European asylum policies up until today. Dublin overrode once and for all any consideration for asylum seekers’ needs and rights (no relevance given to their intention, nor links with the host country) and ignored alternative, broader approaches to asylum, promoted by the UNHCR at the time, such as fighting against the root causes of flight or international liability for countries of origin.Footnote 79

The implicit equivalence given to asylum seekers and irregular immigrants was consistent with the historical record of international negotiations on refugees, which have always been characterised by a mix of humanitarian considerations and the fear states have of the abusive use of asylum. Refraining from leaving a ‘blank cheque’ for asylum seekers was predominant during the negotiations of the Geneva Convention, as well as the 1967 New York Protocol and the General Assembly Declaration on Territorial Asylum.Footnote 80 By the mid-1980s, there were historical contingencies exacerbating this tendency: the spread of mixed flows, the new country of origin of many asylum seekers, the security approach of the Ministries of the Interior and Justice represented in the Ad Hoc Group,Footnote 81 as well as the dominance of the responsibility principle and the intergovernmental tensions over migration. States were focused on making themselves responsible for the entry of third-country nationals and shifting the responsibility (and related burden) to their neighbours as best as they could, no matter what the conceptual implications. The fear that such an approach might convey a repressive image of the convention was diffuse among the twelve and prompted the Group of the Coordinators (established by the European Council of Rhodes in December 1988) to synchronise efforts on free circulation, to state publicly in 1989 that ‘Free movement is a dimension of a people's Europe . . . often obscured by the importance of the measures to be taken to ensure that the opening of borders does not result in increased insecurity . . . but should not lead to a misinterpretation of these efforts’.Footnote 82 This failed to reshape the image of the convention.

Conclusions

I have shown how the negotiations of the Dublin Convention started from a West German proposal, which matched with other member states’ interest in discussing asylum issues in light of the new characteristics of the phenomenon in the EC, the growing mutual impact of national policies in the field and the process of European integration, which established a common need for rules of entry for third-country nationals. While retracing the historical roots of the Dublin system, I have reflected on the inclusion of the first entry principle in the criteria established by the convention. This was a turning point for EC asylum policies: none of the international fora that dealt with asylum at the time considered such a principle to entail responsibility and most member states had applied it on a systematic basis only to irregular immigrants.

The adoption of the first entry principle created the premise for the problem of burden distribution, which would become a point of fracture inside the community in the years that followed. The numbers and trajectories of asylum seekers made it tricky to predict such developments by the end of the 1980s, and the difficulties in the ‘workability’ of the convention postponed the emergence of the issue for some years. However, as I have shown, all the grounds for the problem were there: in addition to the attitude of ‘shifting responsibility’ of most member states and the north-south fracture, the adoption of the authorisation principle, together with restrictive entry measures, made irregular entry (and the implementation of the first entry rule) or family reunification the only way for an asylum seeker to reach Europe.Footnote 83

The Council Resolution on Host Third Countries of 1992 and the signing of numerous readmission agreements valid for asylum seekers in the early 1990s shed a new light on the historical relevance of the Dublin Convention. Most of the literature emphasises the role of article 3.5 of the treaty for the readmission agreements that followed, leaving member states, West Germany above all, a free hand to deal with the increasing number of arrivals from Eastern Europe.Footnote 84 However, the first entry principle was also strongly legitimising. Despite the Dutch proposal in the Ad Hoc Group (to require a certain amount of time in a third host country to send an asylum seeker back to that country, the transit criterion), most of the 1990s readmission agreements considered mere transit enough and made no distinction between asylum seekers and irregular immigrants.Footnote 85 The UNHCR criticised the implementation of the ‘transit’ principle for asylum seekers in these agreements. However, other issues were more urgent (namely the lack of protection against refoulement and of specific guarantees for asylum seekers) and the organisation left aside any campaigns against it.Footnote 86

Looking at the Dublin Convention through the lens of the historiography of European integration, I have shown that despite the initial efforts of the European Commission, states’ jealousy of sovereignty rights in the asylum field led member states to follow an intergovernmental path, leaving European institutions out of the discussion and enabling governments to fully control and determine the outcome of negotiations.

The European integration process played a primary role in shaping the content of these discussions, especially concerning the adoption of the first entry rule. The establishment of an area without internal frontiers, in which the free movement of persons was ensured, implied a higher risk of secondary movements of migrants (including asylum seekers) from the member states that were less vigilant but also less attractive (coinciding with EC border countries) to the more attractive states. The historical context at the end of the 1980s might lead us to think of Cold War influences on the asylum discussions between the twelve, but it was north-south rather than East-West tensions that conditioned the content of their negotiations. The exchanges inside the Ad Hoc Group came to reflect tensions over migratory issues that dated back to the previous decades and did not have much to do with asylum in itself. By establishing common rules to allocate asylum seekers and, among these rules, the first entry principle, northern states tried to impose more responsibility (and the related burden) on their southern neighbours, seeing them as a loophole in their national migration (and asylum) systems. The British Home Office asserted that the main aim of the Dublin Convention was to ‘penalise the Member State responsible for allowing the entry to the EC of an asylum seeker . . . the logic here is that this is the best way of promoting common standards of frontier vigilance’.Footnote 87

Franco-Italian tensions were particularly relevant. Mediterranean states accepted the first entry rule promoted by the Dublin Convention because they underestimated the future implications of such a principle and, above all, did not want to be excluded from what was seen as a step in the European integration process. Dublin's provisions (including the first entry rule) were very similar to the clauses of Chapter VII of the 1990 Schengen Convention, which was unanimously interpreted as a pioneer of the European integration process. By accepting the Dublin compromise, Mediterranean countries tried to facilitate their inclusion in the Schengen process. Since asylum was less strategic than other clauses of the Schengen Convention, the compromise was considered worthwhile.

Against this background, the track followed by the Dublin negotiations is consistent with the history of the European migratory system, a consistency we should not take for granted. The asylum realm, differently to the broader migration field, is strictly connected to human rights and humanitarian considerations, as well as European identity and values. By following the path and approach I have retraced in this article, the twelve indulged in inter-state dynamics and tensions related to the broader migratory issue, bunching together asylum seekers and irregular immigrants, rather than following a human rights approach, which would have given value to the asylum seeker's perspective, to his or her preferences and link with the host country, or would have implied a broader approach to the issue, including a discussion of the means to fight against the root causes of flight and assess international liability. This choice has affected the path followed by European asylum policies and continues to do so in the contemporary world.

References

1 Convention determining the state responsible for examining applications for asylum lodged in one of the member states of the European Communities – Dublin Convention, OJ C 254, 19.08.1997. By ‘asylum’ the paper means the protection granted under the 1951 Geneva Convention and the 1967 New York Protocol, despite the word referring to a broader and more complex concept. This is due to a need for synthesis and because the negotiations of the time and the Dublin Convention (article 1) defined ‘asylum’ in these terms.

2 The Dublin Convention was turned into a European Community Regulation in 2003 (Reg. (EC) n.343/2003), the so-called Dublin II, reformed in 2013 in the so-called Dublin III (Reg. (EU) n.604/2013), in effect at the time of writing. On the criticism towards Dublin III see, among others, European Commission, ‘Country Responsible for Asylum Application (Dublin Regulation)’, available at https://ec.europa.eu/home-affairs/what-we-do/policies/asylum/examination-of-applicants_en (last visited 2 May 2020).

3 See, among others, the works of (in alphabetical order) Christina Boswell, Matthew J. Gibney, Guy S. Goodwin-Gill, Atle Grahl-Madsen, Kay Hailbronner, Christian Joppke, Sandra Lavenex, Gil Loescher, Goran Melander.

4 See, among others, Comte, Emmanuel, The History of the European Migration Regime: Germany's Strategic Hegemony (New York: Routledge, 2018)Google Scholar and Paoli, Simone, Frontiera Sud: l'Italia e la nascita dell'Europa di Schengen (Milano: Mondadori, 2018)Google Scholar.

5 Gibney, Matthew J., The Ethics and Politics of Asylum: Liberal Democracy and the Response to Refugees (New York: Cambridge University Press, 2004), 15–20CrossRefGoogle Scholar.

6 By ‘Schengen process’ the paper means the negotiations between France, West Germany, Belgium, Luxemburg and the Netherlands that determined the signature of the 1985 Schengen Agreement on the gradual abolition of checks at their common borders and of the 1990 convention on the implementation of the 1985 Agreement.

7 See, among others, European Parliament News, ‘EU Asylum Rules: Reform of the Dublin System’, 24 July 2019, europarl.europa, available at: https://www.europarl.europa.eu/news/en/headlines/priorities/refugees/20180615STO05927/eu-asylum-rules-reform-of-the-dublin-system (accessed 2 May 2020).

8 Statistical data on migration and asylum in these years are approximate, given the scarce reliable sources and the different national statistical systems. Three major factors determined the increase of arrivals of asylum seekers by the mid-1980s: the growing instability of some developing countries torn by civil wars, economic crisis, natural disasters or a combination of the three; the technological advances in transport and communication; and the closure of foreign labour recruitment by most European countries, as a result of the economic crisis of the previous decade. For an historical review of this period see Loescher, Gil, ‘The European Community and Refugees’, International Affairs, 65 (1989), 617–36CrossRefGoogle Scholar.

9 This perception was strictly related to the appearance of mixed flows – migration flows where asylum seekers who fulfilled the Geneva requirements were mixed with economic migrants and people fleeing circumstances not included in the Geneva Convention (such as wars, famine or natural disasters) – and the growth of the ‘asylum shopping’ phenomenon – asylum seekers asking for protection in more than one state. On the two phenomena see Brochmann, Grete and Hammar, Thomas, eds., Mechanisms of Immigration Control: A Comparative Analysis of European Regulation Policies (London: Bloomsbury Academic, 1999), 12–15Google Scholar. Some scholars stress how the non-European origin of the new asylum seekers, who started to come from all over the world, contributed to a less favourable attitude towards them. Joly, Danièle and Cohen, Robin, eds., Reluctant Hosts: Europe and its Refugees (Aldershot: Avebury, 1989), 147Google Scholar.

10 For an overview of these policies see Martin, David A., ed., The New Asylum Seekers: Refugee Law in the 1980's: The Ninth Sokol Colloquium on International Law (Dordrecht: Boston Press, 1988)CrossRefGoogle Scholar. These restrictions in the asylum field were consistent with the closures introduced by the twelve concerning immigration in the previous decade. On the immigration restrictions of the 1970s, see among others Guiraudon, Virginie, Les politiques d'immigration en Europe: Allemagne, France, Pays-Bas (Paris: L'Harmattan, 2000)Google Scholar.

11 For an interesting study and overview of the literature on the point see European Commission, Asylum Migration to the European Union: Patterns of Origin and Destination (Luxembourg: Official Publications of the European Communities, 1998) and Kritzman-Amir, Tally, ‘Not in My Backyard: On the Morality of Responsibility Sharing in Refugee Law’, Brooklyn Journal of International Law, 2 (2009), 355–94Google Scholar.

12 On this concern and competition see, among others, Fullerton, Maryellen, ‘Restricting the Flow of Asylum-Seekers in Belgium, Denmark, the Federal Republic of Germany and the Netherlands: New Challenges to the Geneva Convention Relating to the Status of Refugees and the European Convention on Human Rights’, Virginia Journal of International Law, 33 (1989), 33–114Google Scholar.

13 The United Kingdom, Greece, Denmark and Ireland were the most restrictive on third-country nationals’ free movement rights. Archives du Ministère des Affaires Étrangères (thereafter AMAE), La Courneuve, 1914INVA5 Sauve (Ministère des Affaires Intérieurs (thereafter MAI), Direction Liberté Publiques et Affaires Juridiques) – Ministère des Affaires Etrangères, Direction des Français à l'extérieurs et des Etrangères en France (thereafter MAE-DFEF), LIB/CAB II/BJ/N.12, 06.01.1989.

14 AMAE, 101SUP17 Trevi Group and Ad Hoc Group on Immigration, 14–15 Dec. 1989. On the relations between the two groups and the previous works of the Trevi Group see Lavenex, Sandra, The Europeanisation of Refugee Policies: Between Human Rights and Internal Security (Aldershot: Ashgate, 2001)Google Scholar and Bigo, Didier, L'Europe de la sécurité intérieure (Paris: Institut des hautes études de la sécurité Intérieure, 1992)Google Scholar.

15 Callovi, Giuseppe, ‘Regulation of Immigration in 1993: Pieces of the European Community Jig-Saw Puzzle’, International Migration Review, 2 (1992), 353–72, 365Google Scholar. Among member states, West Germany, France, Italy and Spain had constitutional provisions which recognised the right of asylum.

16 National Archives (thereafter NA), Kew Garden, FCO 30/7564 Cabinet Office ‘Note’ 15.04.1988.

17 Historical Archives of the European Commission (thereafter HAEC), Brussels, BDT-315.2014 1044 70.253.032 Williamson (EC Commission, General Secretariat), SG (88) 3312, 25.03.1988. The principle of subsidiarity promoted by the commission was an initiative to come into play only if intergovernmental work did not make any satisfactory progress.

18 Hailbronner, Kay and Thiery, Claus, ‘Schengen II and Dublin: Responsibility for Asylum Applications in Europe’, Common Market Law Review, 34 (1997), 957–89, 962CrossRefGoogle Scholar. Asylum will be included in the Third Pillar by the Treaty on European Union, under article K1. OJ C 326, 26.10.2012.

19 The introduction of a more restrictive asylum law in 1986 determined a drastic reduction of asylum applications in Denmark. NA, FCO 30/9053 Hatful (Home Office, thereafter HO) – Spencer (Foreign and Commonwealth Office, thereafter FCO), 17 Dec. 1990. For a synthetic overview of this period see Loescher, Gil, Beyond Charity: International Cooperation and the Global Refugee Crisis (Oxford: Oxford University Press, 1993)Google Scholar.

20 John Salt, Current Trends in International Migration in Europe, Council of Europe, CDMG (2001), 23 Nov. 2001.

21 Article 16 of the Basic Law stated: ‘persons persecuted on political grounds shall enjoy the right of asylum’, meaning asylum seekers had always to be granted access to the German territory. Friedrich Heckmann and Wolfgang Bosswick, eds., Migration Policies: A Comparative Perspective (Stuttgart: Enke, 1995), 324–7. The proximity of Eastern Europe implied arrivals from those territories but also from developing countries through the Schonfeld airport (East Berlin). NA, FCO 58/4488 FCO, Western European Department, ‘Home Secretary's Visit to Bonn’, 23 Sept. 1986.

22 Debates on the opportunity to reform the liberal asylum policy dated back to the 1950s but involved only experts at the time. Poutrus, Patrice G., ‘Asylum in Postwar Germany: Refugee Admission Policies and Their Practical Implementation in the Federal Republic and the GDR Between the Late 1940s and the Mid-1970s’, Journal of Contemporary History, 49 (2014), 115–33CrossRefGoogle Scholar.

23 Stokes, Lauren, ‘The Permanent Refugee Crisis in the Federal Republic of Germany, 1949–’, Central European History, 52 (2019), 19–44CrossRefGoogle Scholar.

24 Constitutional amendments required a qualified majority, but only the Christian Democratic Union and Christian Social Union were ready to do it; the Free Democratic Party and the Social Democratic Party were against it. For an overview of the West Germany asylum policy see Munch, U., Asylpolitik in der Bundesrepublic Deutschland. Entwicklung und Alternativen (Opladen: Leske & Budrich, 1993)CrossRefGoogle Scholar and Bosswick, Wolfgang, ‘Development of Asylum Policy in Germany’, Journal of Refugee Studies, 43 (2000), 43–60CrossRefGoogle Scholar.

25 AMAE, 1851INVA34 Renouard (MAE-DFEF) Note 05 Nov. 1986. This primary role of West Germany is consistent with Emmanuel Comte's thesis that Bonn was the most relevant actor in the establishment of an European migration system; see Comte, Emmanuel, The History of the European Migration Regime: Germany's Strategic Hegemony (New York: Routledge, 2018)Google Scholar.

26 NA, FCO 30/9904 Levi (British Embassy Bonn) – Arthur (FCO) 09 Aug. 1991. Germany eventually amended its constitution in 1993; the Dublin Convention, but also the massive arrivals of the early 1990s, played a role in this decision. Similar dynamics influenced the approval of stricter asylum laws by most member states in the early 1990s.

27 Ferris, Elizabeth G., Refugees and World Politics (New York: Praeger, 1985), 111Google Scholar.

28 AMAE, 3302TOPO2918 German Federal Ministry of the Interior ‘Concept regarding the refugee problem’, 25 Sept. 1990.

29 As was the case for the Schengen Convention, whose broader approach exposed it more to the developments in Eastern Europe. AMAE, 101SUP104 Tel n.26492-94 Calimajou (MAE) – Administrative Distribution, 18 Dec. 1989.

30 NA, FCO 58/4488 Stagg (FCO) – Barnett (FCO), 10 Nov. 1986. At the same time, the Council of Europe had already discussed the issue in 1988, under an Austrian request (once again, an Eastern European bordering country).

31 AMAE, 101SUP4 Renouard, Schengen II-7, 05 Oct. 1988.

32 Historical Archives of the Council of the European Union (thereafter HACEU), CM2 WGI 44 General Secretariat of the Council, SN 1830/87 WGI 89, 22.06.1987. The Dublin Convention adopted a restrictive concept of family unity. At the same time, cultural links were among the facultative clauses of the convention.

33 NA, FCO 30/8222 Morris (HO) – Langdon (HO), 23.10.1989.

34 HAEC, BDT/224/94 1180 Isoplan Etude Nov. 1989. For an historical overview of the Mediterranean asylum and migration policy see Russell King and Richard Black, Southern Europe and the New Immigrations (Brighton: Sussex Academic Press, 1997).

35 NA, FCO 30/9904 Waterworth (British Embassy Bonn) – Wilkinson (Private Secretary, Home Office), 20 Aug. 1991. For an historical overview of the Italian migration and asylum policies see Einaudi, Luca, Le politiche dell'immigrazione in Italia dall'unità a oggi (Rome: Laterza, 2007)Google Scholar and Michele Colucci, Storia dell'immigrazione straniera in Italia : dal 1945 ai nostri giorni (Rome: Carocci, 2018).

36 The Martelli Law (39/1990) determined Italy's adhesion to the 1967 protocol. On the internal and external dynamics that led to the adoption of the Martelli Law see Finotelli, Claudia and Sciortino, Giuseppe, ‘The Importance of Being Southern: The Making of Policies of Immigration Control in Italy’, European Journal of Migration and Law, 11 (2009), 119–38CrossRefGoogle Scholar.

37 AMAE, 101SUP13 Tel.646 Perol (French Embassy Rome) – MAE 12.06.1990.

38 The reasoning developed in this and the following section works for (non-authorised) land transit. Air transit was disciplined by separate clauses of the convention (articles 7.2 and 7.3). On the ‘transit’ concept in asylum policy see Fitzgerald, David Scott, Refuge Beyond Reach: How Rich Democracies Repel Asylum Seekers (Oxford: Oxford University Press, 2019)CrossRefGoogle Scholar.

39 Article 30e of the Convention Implementing the Schengen Agreement. The analysis of the relation between the Schengen and the Dublin Convention is beyond the scope of this article. It is enough to state that the five of the Schengen Group came to an agreement before the twelve and sent the same negotiators of the two groups to the table, which promoted the same core ideas. France and West Germany were particularly aligned in both fora. AMAE 101SUP8 Renouard, Note n.8329, 18 Apr., 1989.

40 Article 31.1 of the Geneva Convention states: ‘Contracting states shall not impose penalties, on account of their illegal entry or presence, on refugees who, coming directly from a territory where their life or freedom was threatened in the sense of Article 1, enter or are present in their territory without authorization’. There was a debate on the interpretation of the word ‘directly’. On the point see Hurwitz, Agnes, The Collective Responsibility of States to Protect Refugees (Oxford: Oxford University Press, 2009), 129–35CrossRefGoogle Scholar.

41 The twelve indirectly recognised asylum seekers’ right to choose their country of asylum in case of no visa requirements and of airport or port transit (article 7), taking the 1960s Benelux integration agreements as an example. It is interesting to stress that the Schengen Group did not provide free choice in these cases, where the first entry rule worked.

42 UNHCR Executive Committee, Conclusion n.15 (1979), available at https://www.unhcr.org/excom/exconc/3ae68c960/refugees-asylum-country.html and AMAE, 1930INVA5068 Council of Europe, Parliamentary Assembly, Commission on migrations, refugees and demography AS/PR (34) 15, 13 Aug. 1982.

43 See, e.g., Historical Archives of the European Union (thereafter HAEU), Fiesole, PE1-18213, B1-112/84 (Lizin, Socialist Group) adopted on 15 Apr. 1984 on gender persecution or PE2-13311, B2-512/87 (Socialist Group) adopted on 18 June 1987 on national asylum policies.

44 European Parliament, Res A2-227/86, OJ C 099, 13 Apr. 1987. For the European Commission, see below. Other contexts dealt with the matter at the time, such as the European consultations on the arrivals of asylum seekers and refugees in Europe, but they remained on a general level and are therefore excluded from the analysis.

45 Antonio Fortin, UNHCR Senior Legal Officer in London, quoted in US Committee for Refugees, At Fortress Europe's Moat: The Safe Third Country Concept, July 1997, 9.

46 This position reflected the concerns of the High Commissioner Van Heuven Goedhart. Goodwin-Gill, Guy S., The Refugee in International Law (Oxford: Oxford University Press, 1996), 88Google Scholar.

47 UNHCR Executive Committee, Conclusion n.15 (1979).

48 UNHCR Executive Committee, Conclusion n.58 (1989), available at https://www.unhcr.org/excom/exconc/3ae68c4380/problem-refugees-asylum-seekers-move-irregular-manner-country-already-found.html (last visited 2 May 2020). The non-refoulement principle prohibits states from sending a person to a place where s/he is at risk of irreparable harm upon return, including persecution, torture, ill treatment or other serious human rights violations. The idea of ‘protection’ was related strictly to the debate on the meaning of ‘safe’ third country. The analysis of these discussions is beyond the scope of this article; for a review of the literature on this point see Coleman, Nils, European Readmission Policy: Third Country Interests and Refugee Rights (Leiden: Martinus Nijhoff, 2009), 223–317CrossRefGoogle Scholar and Lavenex, Sandra, Safe Third Countries (Budapest: Central European University Press, 1999)Google Scholar.

49 Zoom interview with Professor Christopher Hein (Professor in Refugee Law at LUISS University, former UNHCR official and former president of Consiglio Italiano per i Rifugiati), 25 May 2020.

50 Revisiting the Dublin Convention: some reflections by UNHCR in response to the commission staff working paper SEC (2000), 522.

51 By Europe, the paper refers here to the twelve, whose governments had frequent tensions with the high commissioner because of critical stands of the organisation (see West Germany around 1983) and lack of funding. On the relation between UNHCR and European governments see Loescher, Gil, The UNHCR and World Politics: A Perilous Path (Oxford: Oxford University Press, 2001)CrossRefGoogle Scholar.

52 The length and the nature of the stay were a matter of discussion. The issue of the nature of the stay was resolved thanks to the secretariat of the council, who suggested the formula ‘with the consent of the authorities’, where the consent did not need to be formally expressed, recalling the provisions of the agreement on transfer of responsibility for refugees. AMAE, 1851INVA2 Council of Europe, CAHAR (83) 20 def., 12 July 1983.

53 Council of Europe, CAHAR (84) 4 final, 22 May 1984, available at https://rm.coe.int/0900001680911b5e (last visited 2 May 2020).

54 AMAE, 3302TOPO2918 Council of Europe, CAHAR (87) 1, 03 Feb. 1987.

55 The UNHCR participated as an observer to the CAHAR. The Italian, Turkish and Spanish delegations tried to again put the issue of the ‘link’ between asylum seekers and host countries on the table but at a late stage of the discussions, and they did not fight for it. Council of Europe, CAHAR (88) 3, 18 Apr. 1988, available at https://rm.coe.int/09000016809125e4 (last visited 2 May 2020).

56 Council of Europe, Conclusions of the 427th Meeting of the Ministers’ Deputies, 12–15 June and 19 June 1989, available at https://rm.coe.int/090000168091d694. Although the convention was never signed, part of the unadopted agreement was included in Parliamentary Assembly, Recommendation 1236 (1994), available at http://assembly.coe.int/nw/xml/XRef/Xref-XML2HTML-en.asp?fileid=15270&lang=en (last visited 2 May 2020).

57 Visa and permits of stay were considered an indirect expression of the asylum seeker's intention. HAEC, BDT-315.2014 1044 70.253.032 Braun and Perissich (EC Commission, DG III Internal Market), Note III/D/1, 28 Jan. 1988.

58 Gaia Lott, ‘Between Restrictiveness and Humanitarianism: EC Institutions and the Asylum Policies of the 1980s’, in Sara Lorenzini, Umberto Tulli and Ilaria Zamburlini, eds., The Human Rights Breakthrough of the 1970s (London: Bloomsbury, 2021).

59 Belgium and West Germany required a three month stay; the Netherlands, Spain, Portugal and Italy considered transit (broadly interpreted) not enough. FCO 30/8227 Ad Hoc Group on Immigration (thereafter Ad Hoc Group), SN 2499/89 WGI 444, 04.09.1989. These time requirements were consistent with the UNHCR principle of ‘previous protection’.

60 The 1986 ‘Danish clause’ derived from the Scandinavian practice; ‘Société française pour le droit international’, 30th Colloque de Caen: droit d'asile et des réfugiés (Paris: Pedone, 1997), 200.

61 AMAE, 1914INVA5 Sauve – MAE-DFEF, LIB/ECT/MIS/N.108 FD-DS/MB, 27.04.1989. Moreover, as the French recognised, the procedure was rarely applied.

62 Proposals ranged from three months to two weeks to three days in the transit country. HACEU, CM2 WGI 45 Ad Hoc Group, SN 1223/88 WGI 257, 14.04.1988.

63 HACEU, SCH/II 58 Schengen Agreement, Group II, SCH/II (88) 9, 02.03.1988. The French expressed this position in the Schengen Group, but it worked for the twelve's discussions as well.

64 HACEU, CM2 WGI 45 Ad Hoc Group, SN/63/88 WGI 221, 15 Feb. 1988.

65 AMAE, 101SUP1 MAE, Coordination Mission for the Free Circulation of Persons Note, 18 Apr. 1989.

66 NA, FCO 30/8228 Arthur Hodge (British Embassy Copenhagen), 13 Oct. 1989.

67 HAEC, BAC 41/1989 N.1046 Wenceslas de Lobkowicz (EC Commission, DG III Internal Market) ‘CAHAR 21–5 Sept. 1987’, 23 Oct. 1987.

68 During the negotiations Denmark, the Netherlands and the United Kingdom suggested unsuccessfully to rely on ‘reasonable presumption’ rather than ‘proof’ of irregular entry. HACEU, CM2 WGI 77.3 Ad Hoc Group, SN 3144/89 WGI 483, 18 Oct. 1989.

69 On this point see Paoli, Simone, Frontiera Sud: l'Italia e la nascita dell'Europa di Schengen (Milano: Mondadori, 2018)Google Scholar.

70 Germany would put the ‘burden sharing’ issue on the negotiation table of the twelve some years later (1994) with no success; see Draft Council Resolution on Burden-sharing with Regard to the Admission and Residence of Refugees, July 1994, Doc. No. 7773/94 ASIM 124.

71 Commission of the European Communities, Staff Working Paper ‘Revisiting the Dublin Convention’ SEC (2000) 522.

72 NA, FCO 30/9903 Lord Waddington (Home Secretary) ‘Asylum and Immigration Appeal’, 12 July 1991.

73 Daphne Bouteillet-Paquet, ‘Passing the Buck: A Critical Analysis of the Readmission Policy Implemented by the European Union and Its Member States’, European Journal of Migration and Law, 3 (2003), 359–78, 362. By ‘irregular immigrant’ the paper refers to the definition given by the European Commission: ‘people who move to a new place of residence or transit outside the regulatory norms of the sending, transit and receiving countries’, European Commission, European Migration Network Glossary, available at https://ec.europa.eu/home-affairs/what-we-do/networks/european_migration_network/glossary_search/irregular-migration_en (accessed 2 May 2020).

74 NA, FCO 30/7564 Ad Hoc Group, SN 3712/1/87 WGI 203, 07 Jan. 1988.

75 See, among others, AMAE, 101SUP8 MAI, Minister Cabinet, International Affairs Advisor, CAB/FN/NL/N.830, 18 Apr. 1989.

76 See, among others, NA, HO 394/991 ‘Dublin and Schengen Convention: UNHCR's Position’, WGI 853, undated but filed in a 1991 folder.

77 UNHCR, ‘Handbook on Procedures and Criteria for Determining Refugee Status under the 1951 Convention and the 1967 Protocol Relating to the Status of Refugees’, HCR/IP/4/Eng/REV.1 Re-edited, Geneva, Jan. 1992, UNHCR 1979. On paper the Dublin Convention protected asylum seekers from the risk of direct and indirect refoulement, because member states were considered safe countries of origin and safe third countries. However, this idea will be problematised by national courts in the early 1990s, especially with regard to indirect refoulement. Hailbronner, Kay, ‘Asylum Law Reform in the German Constitution’, American University Journal of International Law and Policy, 4 (1994), 159–80Google Scholar.

78 To deepen the broader historical discussion on the link between the right of asylum and human rights, see Colloquy on European Law, The Law of Asylum and Refugees. Present Tendencies and Future Perspectives: Proceedings of the Sixteenth Colloquy on European Law (Strasbourg: Council of Europe, 1987).

79 AMAE, 3302TOPO2918 French Mission in Geneva – MAE-DFEF n.648/FAE 20.05.1987. For a synthetic overview of a different approach to asylum policies see Hathaway, James C., ‘Reconceiving Refugee Law as Human Rights Protection’, Journal of Refugee Studies 4, 2 (1991), 113–31CrossRefGoogle Scholar.

80 Zoom interview with Professor David Martin (International Law Professor at the University of Virginia), 15 May 2020.

81 Joppke, Christian, ed., Challenge to the Nation-State: Immigration in Western Europe and the United States (New York: Oxford University Press, 1998), 155CrossRefGoogle Scholar. On certain topics (e.g. rights of asylum seekers) a distinct approach was perceivable between the representatives of Foreign Affairs and of the Interior and Justice departments. However, regarding criteria determining responsibility for asylum applications, archival research shows that the approach of the two ministries was not so different.

82 AMAE, 101SUP2 Coordinator Group Free Movement of Persons, CIRC 3651/1/89 REV 1, 20 Nov. 1989. The works of the Ad Hoc Group had been criticised by NGOs and the European Parliament since the beginning, because of their secret character that excluded democratic accountability. See, e.g., HAEU, PE3-24386 joint resolution replacing the previous motions by the single parliamentary groups (B3-1208; B3-1209; B3-1227; B3-1232; B3-1248), 14 June 1990.

83 Noll, Gregor, ‘Prisoners’ Dilemma in Fortress Europe: On the Prospects for Equitable Burden-Sharing in the European Union’, German Yearbook of International Law, 40 (1997), 405–37, 434Google Scholar.

84 Between 1990 and 2000, 124 readmission agreements were signed (there were twelve between 1950 and 1970 and two between 1980 and 1990). Switzerland, Germany and the Netherlands were the most active in these years. Federal Migration Service of Russia and IOM, Manual on Readmission for Experts and Practitioners (2010), available at https://publications.iom.int/books/manual-readmission-experts-and-practitioners-0 (last visited 3 Feb. 2020).

85 NA, HO 394/991 Ad Hoc Group, SN 2881/91 WGI 837 and WGI 842, 24 July 1991. The 1991 Schengen–Poland agreement and the 1994 EU specimen readmission agreement did the same.

86 NA, FCO 30/9904 Stephenson (FCO) – Montgomery (FCO), 16 Sept. 1991 and UNHCR, UNHCR Position on Readmission Agreements, Protection Elsewhere and Asylum Policy, 1994, available at http://www.refworld.org/docid/3ae6b31cb8.html (last visited 3 Feb. 2020). Referring to the 1990s readmission agreements, some authors speak of ‘neo-refoulement practice’. Jennifer Hyndman and Alison Mountz, ‘Another Brick in the Wall? Neo-Refoulement and the Externalization of Asylum by Australia and Europe’, Government and Opposition, 43 (2008), 249–69.

87 NA, FCO 30/9053 Morris – Dent (HO), 21 May 1990.