Legal Protection as Competition for Jurisdiction: The Case of Refugee Protection through Law in the Past and at Present

Abstract

This article explores the structure of the legal protection of refugees in Europe today. To this end, it will contrast historical arrangements providing protection to refugees, namely church asylum in the late Middle Ages and refuge for religious minorities, with the current European refugee regime, that is the Common European Asylum System (CEAS), in particular the Dublin system. The central claim of this article is that a basic condition for the legal protection of refugees is the existence of multiple jurisdictions, which in turn caters for competition for jurisdiction. The official logic of the CEAS, however, endorses harmonization, unity and the hierarchy of jurisdictions rather than a competition between jurisdictions. This partially explains the difficulties under the CEAS in organizing the protection of refugees through law. In policy terms, this article supports calls for reconsidering the Dublin Regulation, since through the ‘single jurisdiction’ approach Dublin hampers the legal protection of refugees.

1. Introduction

In this article I explore the structure of legal protection, in particular the legal protection of refugees in Europe today. To this end, the article looks at arrangements providing protection to refugees in European history, namely church asylum in the late Middle Ages and refuge for religious minorities following the Counter-Reformation (also referred to as the Catholic Reformation). I will contrast the structure of the historical arrangements with the current European refugee regime, that is the CEAS, in particular the Dublin system. This historical and analytical exercise highlights the fundamental tensions within today's refugee regime and contributes to explaining partially the circumstances that enable or hamper the protection of refugees through law.

The central claim of this article is that a basic condition for the legal protection of refugees is the existence of multiple jurisdictions. This is the necessary basis for one of the driving forces for legal protection, namely competition over jurisdiction. The official logic of the CEAS, however, endorses harmonization, unity and the hierarchy of jurisdictions rather than a plurality of and competition for jurisdictions. This insight partially explains the difficulties under the CEAS in organizing the protection of refugees through law. In concrete policy terms, the findings in this article suggest that the Dublin Regulation creates an institutional structure that hampers the legal protection of refugees.

For the purposes of this article a refugee means a person who seeks protection from prosecution, punishment and persecution. This notion is both broader and narrower than the meaning of a refugee under international refugee law. Many persons who received protection under historical asylum arrangements might not qualify as refugees under current law. Conversely, the ideal type of refugee under the current European refugee regime would not fit within historical asylum arrangements.

The article focuses on jurisdiction exercised by the courts or quasi-judicial authorities and executive authorities. I thus use jurisdiction in two ways: the jurisdiction of judicial authorities to decide a legal case, and the jurisdiction of executive authorities to take action against a person. I do not look at the competence to legislate. Competition for jurisdiction means a struggle between two or more authorities with regard to who has competence to deal with the matter. ¹

I use legal protection in a specific way – to mean protection through the law. For the purposes of this article, legal protection does not so much refer to the protection of (individual) rights, but to protection by invoking law and making legal statements. In other words, the ‘legal’ aspect of legal protection is not an object of protection but a means that makes protection possible. So, under this particular understanding, the fact that legal rights or interests are protected does not necessarily amount to legal protection as such. What matters is that someone or something is protected because someone has invoked the law. My particular use of legal protection presupposes a practice whereby actors invoke the law and make legal statements.

This article takes as its cue the recent interest in the legal history of refugees and asylum. ² But unlike existing studies, by making a comparison with the historical arrangements this article also seeks to reveal how the logic and structure of the current CEAS promote and hamper the legal protection of refugees. The article focuses exclusively on historical legal arrangements for which there is evidence ³ of an actual practice. I rely on secondary literature describing these practices. But the secondary literature on which I rely is based on an examination of primary legal sources, especially case law. ⁴ For instance, my focus on church asylum was triggered by the impressive study by P. Timbal Duclaux de Martin. His innumerous references to actual case law provided a signal that we are dealing with an actual practised legal mechanism offering protection to refugees through law. ⁵

The article's focus on competition for jurisdiction stands in contrast with legitimate critiques from experts of refugee law on the fragmentation in international and EU refugee law and policies. ⁶ The lack of a fully integrated and harmonized approach in the EU caters for legal inequality, denial of responsibility to offer protection, etc. In other words, most experts of refugee law argue against competition for jurisdiction and call for integrating jurisdictions. Here lies the novelty and relevance of the article. Comparing legal protection of refugees in the past and present reveals that even under today's CEAS some crucial instances of legal protection are the result of a competition for jurisdiction. In this respect the article fits within the scholarship pointing to the benefits of legal pluralism. ⁷

Historical comparisons run the risk of showing the obvious: Past and present are simply radically different, making comparisons meaningless. Consequently, the article focuses on the structure of the legal mechanisms offering protection to refugees in the past and today. Precisely the differences and similarities in structure between the historical mechanisms and the CEAS may help us better understand the conditions that promote or hamper protection of refugees through law today.

The central part of the article is a schematic account of two distinct historic mechanisms providing protection to refugees, namely church asylum and refuge for religious refugees (Section 2). From these schematic accounts, the article identifies the structural elements that catered for the protection of refugees through law. The basic structural condition for legal protection was the existence of multiple jurisdictions. Furthermore, a central motive for authorities to provide protection through law seemed to be primarily self-interest: competition for jurisdiction (‘don't touch my refugee’). Accordingly, legal protection was often not framed as a benefit for the refugee, let alone an individual right. Finally, the main trigger for protection was the actual or imminent legal – often criminal – prosecution of the refugee, and not factual persecution. It was often the case that the refugee was not an innocent victim: He had done something wrong. The article then looks at the structure of the CEAS, in particular the Dublin system, and compares it with the historical arrangements (Section 3). The article finds that the dynamics and official logic of the CEAS are largely the exact opposite of the historical arrangements. The CEAS is mainly about promoting a single jurisdiction. Its main focus is on persecution, not prosecution. When EU legislators organize legal protection the main vehicle is individual rights, often human rights. By the same token the picture of the CEAS is not black and white. The ambition of a single jurisdiction is an official policy that does not fully correspond with actual practice. In fact, the article shows how landmark cases by the European Court of Human Rights (ECtHR) and Court of Justice of the European Union (CJEU) that improved the legal protection of refugees were triggered by competition for jurisdiction between legal authorities in the EU. The article contains a schematic table comparing the historical arrangements with the CEAS.

The article concludes that if multiple jurisdictions and competition for jurisdiction are indeed important conditions for legal protection, we should not so much call for more unity and the harmonization of EU refugee law in the name of legal certainty, legal equality and eliminating ‘forum shopping’. Rather, we should explore how to maintain and implement competition for jurisdiction among the various actors within the current EU framework. This means that we should reconsider certain aspects of the Dublin Regulation.

2. Asylum and refuge

2.1. Asylum in churches

For the purposes of this article I focus on church asylum in the period between the eleventh and fifteenth centuries, ⁸ because it was during this era that church asylum reached its most legalized ⁹ character. Within secular jurisdictions (cities, counties, duchies, etc.) churches and other places of Eucharistic celebration constituted a distinct pocket of jurisdiction. In these places fugitives could benefit from church asylum. Thus, when someone was being pursued or prosecuted by secular authorities for committing a crime and, as a result, entered into or took hold of religious premises (e.g., by grasping the ring on the church portal) he fell under the protection of church asylum. It meant that the secular authorities could not arrest him and take him to court. Therefore, the starting point for church asylum was primarily criminal proceedings or the attempt to arrest a suspect for the purposes of a criminal trial. As long as the fugitive remained in the church he was protected from the secular authorities. However, as soon as he left the premises he could be arrested.

There were different ways in which the asylum might end. A crucial factor in this respect was the procedure borrowed from Roman law, the intercessio, whereby the church authorities (e.g., bishops and abbots) would intervene before the prosecuting secular authorities in favour of the refugee. ¹⁰ The intercessio could result in the secular authorities granting the fugitive a laissez-passer up to the borders of its jurisdiction, which meant that the fugitive would become exiled. Asylum could also end when the fugitive – during his time in asylum – reached a settlement with the relatives of his victim. Another way of ending the asylum was when the secular authorities guaranteed – vis-à-vis the church authorities – that the fugitive would obtain a fair trial and if found guilty would not be executed. Sometimes the asylum ended as it could be shown during the asylum that the fugitive was innocent. ¹¹

In the event that the right of asylum was violated, which effectively meant that the refugee was removed from the church by force or was deceived into leaving (e.g., by false promises of protection), ¹² three responses were available. Firstly, the perpetrators could be excommunicated. ¹³ Hence, Muslims and persons who had already been excommunicated were used to apprehend fugitives. ¹⁴ Equally interesting were other ways to circumvent the right of asylum: besieging the church and cutting off supplies to the fugitive (often ending in failure because authorities lacked the resources to pay the besiegers for a long period of time), ¹⁵ or instead of dragging the fugitive out of the church, which was not allowed, the secular authorities sometimes put the fugitive in chains inside the church. ¹⁶ Secondly, the church authorities could file a kind of restitutio in integrum proceeding before the secular courts up to the highest level. Many cases were recorded where secular authorities were ordered by the highest secular courts to return the fugitive to the church authorities. ¹⁷ Thirdly, violators of church asylum could be fined by secular authorities. ¹⁸ Interestingly, such violations and their sanctions were not understood in terms of the fugitive's right to asylum, but rather the right of asylum of the church.

The trigger for asylum was actual or imminent legal proceedings against the fugitive. Secular authorities or citizens sought to apprehend the fugitive in order to have him face trial. So, the starting point for asylum was prosecution in the legal sense rather than de facto persecution. Also, the fugitive was typically not merely an innocent victim of arbitrariness, envy or bias by the secular authorities or local population. The whole point of asylum was that the fugitive would receive a fair trial, humane punishment or settlement, not complete immunity. In other words, it was often the case that the fugitive had done something wrong, but he could nevertheless benefit from asylum. ¹⁹

Church asylum was an intrinsically legal phenomenon; it was one of the first rights that churches acquired when the secular authorities recognized their distinct jurisdiction at the end of the fourth and beginning of the fifth century. In effect, it suggested that jurisdiction and asylum were almost co-constitutive: no meaningful jurisdiction without a right of asylum, and no right of asylum without jurisdiction. This also explains why a violation of asylum was not a violation of the refugee's right to asylum, but a violation of the jurisdiction of the church. Prior to the right of asylum churches offered physical protection, namely hikese, but this is to be distinguished from asylum precisely because of its lack of legal status. ²⁰

Initially the prominent motives for offering protection to refugees were purely theological. Early church doctrine rejected both capital and corporal punishment. Asylum was a way to ensure that no such punishment was to be inflicted on the refugee upon condemnation. Furthermore, originally the church also understood asylum as an opportunity to help the refugee repent for his sins and cleanse his soul. This still fits in with the early but quickly abandoned forms of church asylum where the church would be responsible for punishment instead of the secular authorities. The theological motives soon became less prominent as the church gave up its rejection of capital and corporal punishment, and penitence no longer replaced secular punishment. Still, normative motives continued to play a role. An important motive was the corrective function of asylum. Church asylum contributed to restoring the balance of justice and peace. In terms of procedural justice, through the intercessio the church authorities would obtain a guarantee from the secular authorities that the refugee would obtain a fair trial. In terms of substantive justice church asylum was a way to ensure that the refugee, when brought before the secular court and condemned, would not be given cruel and disproportionate punishment.

Notwithstanding the relevance of the corrective motive, it seems that ultimately or fundamentally the main driving force for church asylum was of a politico-legal nature. From its inception and actual workings, church asylum was most of all a matter of jurisdiction, not morality. To be more precise, the main motive behind church asylum was ongoing competition for jurisdiction between church and secular authorities. What made church asylum an effective arrangement was the fact that the church could invoke its own secular jurisdiction. It was thanks to its own distinctive secular jurisdiction that the church could penetrate the secular realm and compete with non-church authorities. The centrality of competition was not only reflected in the emergence of church asylum, but also in its demise. In effect, the practice of church asylum ended precisely when, in matters of law and government, non-church secular power became dominant from the seventeenth century onwards. When the church lost its distinct jurisdiction in virtually all public matters, notwithstanding any moral motives to correct injustice, church asylum disappeared. In the competition for jurisdiction the church had lost. ²¹

2.2. Refuge for religious minorities following the (Counter-)Reformation

If my treatment of asylum has been schematic in the sense of presenting the mechanisms as being too homogenous, I certainly have a problem when it comes to the immensely diverse ways in which refuge was offered to religious minorities. ²² Yet I must recall that my main purpose is to identify the elements and conditions under which protection was offered through the law. In this respect, I limit myself to two different mechanisms for protection offered to Protestant refugees. For analytical purposes I distinguish between the individual or traditional mechanism illustrated by the early prosecution of Protestants in the Low Countries during the early sixteenth century, and the collective or modern mechanism illustrated by the persecution of Protestants in France by the end of the seventeenth century.

In order to quash the Protestant movement in the Low Countries, Charles V issued multiple orders (plakaten) against believers of the Protestant faith. ²³ These orders required the local and regional authorities to prosecute and punish subjects for acts associated with their Protestant faith ranging from possessing Protestant literature and participating in services, to actually preaching. The prescribed punishment was in many cases execution. The idea was that local ‘police’ authorities would arrest those suspected of heresy and have local tribunals condemn them. If the condemned appealed to the higher regional courts, e.g., the Court of Holland and the Great Council of Mechelen, the latter were supposed to uphold the convictions. It meant that the prosecution was to take place through individual legal trials. But this strategy actually resulted in a protective mechanism of leniency and disobedience. The oppressive policy therefore turned out to be highly ineffective. There was overall leniency and official obstruction in the execution of Charles's policy. ²⁴ For example, suspects were often not arrested as they had been tipped off by the sheriff. The authorities also acquitted suspects or did not even prosecute them. When the suspects were found guilty they often received significantly lower penalties (e.g., exile in their own town). This practice was tolerated by the higher courts. The motives behind this obstruction were threefold. Although the authorities and the majority of the population were often far from sympathetic to the Protestant cause, they simply found the policy to be too harsh and disproportionate. More importantly, local authorities were concerned about the practicability of harsh prosecution. They believed that they could not execute the policy without seriously affecting public order. They would certainly encounter important resistance. In other words, for reasons of public order – one of their primordial responsibilities – local authorities offered de facto protection. ²⁵ Finally, the higher regional courts in particular found that Charles V, by ordering how the courts were to deal with heretics, had directly infringed upon their jurisdiction and legal practice. So, both the lower and higher courts successfully fended off the attempts by Charles V to have the Inquisition assume exclusive jurisdiction. ²⁶ Similarly, resistance against the punishment of confiscation was equally strong, not only because it was felt to be incredibly harsh on the family members of a convicted heretic, but also because it constituted a direct infringement of the legal privileges, rights and liberties of towns and provinces. ²⁷

Since the effectiveness of prosecution under Charles V was completely dependent on the co-operation of local authorities in the Low Countries, a Protestant refugee could often find refuge in another city without the need to leave the actual county or duchy, let alone the Habsburg territories. It also meant that the numbers of refugees were not comparable to the later prosecutions during the second half of the sixteenth century, which in turn may explain why refugees were not treated as a collective religious community within the city of refuge. Their legal status was simply that of an alien and protection was provided by the local authorities simply by not executing the orders to prosecute Protestant subjects.

This individual mechanism of protection for religious minorities has some commonalities with church asylum. Protection was offered partially because local authorities were jealous of their own jurisdiction. Furthermore, the protection was triggered by imminent or actual prosecution.

Different from this more traditional individual mechanism of protection is the collective mechanism providing refuge for religious refugees. ²⁸ For the purposes of my schematic account, I concentrate on the persecutions of the late seventeenth century because they resulted in the ‘purest’ form of the collective mechanism. ²⁹ Here the religious minorities were treated as a collective group by both the persecuting authorities and the authorities offering refuge. Furthermore, both persecution and protection lost their quasi-judicial character. Most of these refugees were not, as such, the subject of legal proceedings. ³⁰ A case in point is the revocation of the Edict of Nantes, which guaranteed some religious freedoms to Protestants in France, by the Edict of Fontainebleau. The Edict in itself did not constitute legal proceedings against individual Protestant subjects. ³¹ Only if Protestant subjects were to violate the Edict did they run the risk of being penalized. ³² The majority of these religious refugees were not fleeing because they were being legally prosecuted. There was certainly an environment, which had been partially created by the law, that was downright hostile to them, but the mechanism is quite different from asylum as explained above, where the context was always imminent or actual legal proceedings against individuals. In fact, it was a separate measure – not a judicial procedure – that really gave the French policy its oppressive character, namely the billeting of soldiers in Protestant households in order to persuade them to convert. Like the persecution, the protection offered to Protestant refugees in, for example, cities in the Dutch Republic lost its judicial character. The protection was not organized in the context of legal proceedings but became a matter of regulation. The authorities of the city of refuge often issued a special regulation establishing the duties, rights and privileges of refugees as members of a particular religious refugee community. Although the status of refugees may have been inferior when it came to political participation and holding public office, in economic respects it was often more favourable than the legal position of ordinary citizens (e.g., tax exemptions and the freedom to engage in particular crafts outside guilds).

Like asylum, collective refuge for religious minorities was possible because there were multiple jurisdictions: The religious minorities needed a separate place to which they could flee. ³³ But for the rest collective refuge was less a matter of law and jurisdiction. Firstly, the trigger for refuge was not legal proceedings against individual members of a religious minority but de facto hostility by the authorities and the Catholic population vis-à-vis the religious minority as a group. It also follows from this non-legal trigger that the religious refugee was not a suspect or actual criminal as in the case of asylum, but an innocent victim. Secondly, unlike asylum where the prosecuting authorities often had the de facto authority to get their hands on the refugee, from a practical military perspective it was not possible for the persecuting authorities to retrieve their subjects from the country of refuge (e.g., the King of France did not have the military means to retrieve Protestant subjects taking refuge in England). In other words, the protective force of refuge for religious minorities seemed to have largely been derived from the military strength and physical and geographical positioning of the receiving state. Thirdly, the protection offered to religious refugees by the receiving authorities took the form of regulations. Although these measures were legal they took the form of privileges and not legal proceedings or court cases. Furthermore, the privileges were granted to the religious community and not to individuals. Interestingly, those privileges often were aimed at protecting not the refugees against persecution (or prosecution) by the ‘exiling’ authorities, but newcomers against the local population of the receiving state. ³⁴ In any event, collective refuge for religious minorities constituted an actual benefit for refugees. This stands in contrast to asylum, which was not structured as a right of the refugee but the church. Thus, in an almost paradoxical way, while collective refuge for religious minorities constituted a collective benefit for refugees, it was not structured as offering legal protection against persecuting and prosecuting authorities. Finally, the non-judicial character of the later collective refuge for religious refugees is also reflected in the fact that the refugees were considered to be innocent fellow believers, victims of unjust persecution. By contrast, the church offered asylum to refugees not because they were innocent (often they were not), but because secular authorities claimed jurisdiction over them. Also under the early forms of the individual protection of religious refugees, local authorities disobeyed the orders to prosecute Protestant subjects not because they believed that the Protestant suspects were innocent victims. Rather, they believed that the orders to prosecute were impractical and violated their own jurisdiction.

Apart from the pursuit of religious homogeneity (for the persecuting authorities) and religious solidarity (for the protecting authorities), it seems that there was some win-win motive behind collective refuge. ³⁵ The exiling authorities could rid themselves of their religious enemies, while the authorities offering refuge could benefit from the economic and financial potential offered by these refugees. ³⁶ This is another contrast with church asylum, which was generally not in the economic interest of the church. It cost money and was cumbersome. As explained above, the motive behind church asylum was not economic but politico-legal: competition for jurisdiction. In this respect I should note that the motives of the King of France were particularly complex: While he may have wanted to get rid of his religious enemies, he also wanted to have a realm with as many subjects as possible. So, the Edict of Fontainebleau ordered Protestant ministers to leave within a fortnight, yet non-clerical Protestant subjects were ordered to stay. ³⁷ The ministers were officially exiled. However, the bulk of Protestant refugees from France were not officially exiled, but had fled from France illegally. ³⁸

In short, the motives for asylum were highly politico-legal: competition for jurisdiction. In the case of collective refuge for religious refugees the motive was political, religious and especially economic. Furthermore, asylum was an individual and legal mechanism. By contrast, collective refuge for religious minorities was much more a de facto (or executive) and collective arrangement. Interestingly, in the case of asylum both competing authorities wanted control over the refugee. By contrast, refuge for religious minorities seemed to be more like an outright abandonment of or disconnection from them. Unlike asylum the persecuting authorities did not claim that the religious refugees had to be returned. Finally, with collective refuge for religious refugees, innocence and victimhood became a key characteristic of these refugees. In fact, it was not asylum, but collective refuge for religious refugees that was the true non-juridical precursor to our modern European refugee regime. ³⁹

2.3. Regulation, management and administrative law

Although this article focuses on particular historical legal mechanisms offering protection to refugees, a history of modern refugee law is not within its scope. Yet I should say something about the emergence of modern refugee law since it introduces a new logic and approach that is crucial in terms of legal protection, namely management and administrative law. The collective mechanism for protecting religious refugees already hinted at this new modus operandi. It echoes a collective and de facto logic that is deeply rooted in the current refugee regime and reflects an underlying structure within modern migration policy in general. Also, with the advent of religious refugees came the moral category of innocent victims of persecution.

Apart from these novelties introduced by the protection of religious refugees, the protection of refugees in general was, until the middle of the nineteenth century, still structured as a matter of criminal justice and the ‘classic’ law of nations. ⁴⁰ Protection was organized around the legal question of whether there was a duty to extradite the refugee or whether the receiving state had the right not to extradite. The trigger was typically a refugee escaping criminal prosecution in his home country. Prior to the French Revolution, refugees who were prosecuted for ordinary crimes would receive protection, since they would fall under the normal criminal jurisdiction of the receiving state. By contrast, if the refugee was prosecuted for political crimes, the right not to extradite would be less clear. It became a matter of foreign relations and whether the refusal to extradite could give rise to diplomatic actions or constitute a cause of war. From the French Revolution onwards this logic became reversed. It was precisely political refugees who deserved asylum, whereas refugees prosecuted for ordinary crimes had to be extradited. As an analogy with the religious refugee, the political refugee prosecuted for his beliefs became the quintessential category of a refugee who truly deserved protection. This reversal became crystallized from the middle of the nineteenth century onwards, when leading European states entered into extradition treaties, reserving asylum for political refugees only. The upshot of this development was that refugee protection became almost completely detached from the logic of criminal law and criminal jurisdiction.

Until the beginning of the twentieth century, the dominant refugee context was still protecting the refugee against prosecution and extradition claims from the country of origin. This context changed from the First World War and Second World War onwards, with increasingly large numbers of refugees now seeking protection from non-protection. The country of origin (if it still exists) does not necessarily prosecute the refugee; it simply fails to offer minimal protection. The relevance for the purposes of this article is that countries of origin cease to claim jurisdiction over the refugees. As a result, the decision to offer or refuse refugee protection stops being a matter of international relations and the classic law of nations. To be sure, modern refugee protection is still very much governed by international law instruments. Yet, contrary to the classic law of nations, these international law instruments are not so much horizontal instruments regulating the conditions under which offering or refusing refugee protection constitutes a cause of war or diplomatic reprisals. In fact, though imperfect, they aim at establishing a vertical relationship between the refugee and the receiving state. In short, by the mid-twentieth century, the nineteenth century elements that had catered for a kind of competition for jurisdiction largely disappeared (extradition, criminal jurisdiction and just cause of war). ⁴¹

Furthermore, not only did refugee protection cease to be a matter of criminal/international jurisdiction. In the twentieth century refugee policy became a matter that was closely connected to migration policy. The pairing of the two policy fields resulted in refugee policy adopting the structure of migration policy, namely regulation, administrative law and, later on, management. ⁴² The dominance of the administrative law and the management paradigm is now deeply rooted in the EU context. In the member states refugee law is typically a subset of the law on aliens, which is largely governed by administrative law. Pursuant to Article 79 of the Treaty on the Functioning of the European Union (TFEU) the official objective of the EU's common migration policy is the following: ‘efficient management of migration flows’. Any reference to juridical practice has disappeared and the legal subjects have vanished into a flow. Article 79 deals with migration policy while Article 78 deals with refugee policy. However, from the early beginnings of the CEAS, the EU understood the two policy areas as being closely related. ⁴³ Furthermore, technically speaking the efficient management of migration flows under Article 79 TFEU presupposes that asylum seekers and refugees are filtered out and transferred to the CEAS under Article 78 TFEU. When under the Article 78 regime asylum seekers see their application being rejected and they cease to have a legal title to stay, they again end up under Article 79 TFEU. Under Article 79(2)(c) TFEU the EU adopts measures regarding ‘illegal immigration and unauthorised residence, including removal and repatriation of persons residing without authorisation’. In short, according to the EU's own rationale and legal techniques, the efficient management of migration flows must include the efficient management of refugee flows. The point here certainly is not to show that the efficient management of migration and refugee flows is a malign policy objective, nor do I deny the ethical pull of portraying the ideal type of modern refugee as an innocent victim of persecution for justly held political beliefs who deserves our protection, which is very similar to the religious refugee. Rather, the point here is to show how in the past the protection of refugees was also possible thanks to less benign and morally laudable dynamics, in particular, competition for jurisdiction.

3. Comparing asylum and refuge with the CEAS: Multiple versus single jurisdiction

The historical arrangements offering legal protection to refugees are characterized by a plurality of jurisdictions, often competing for priority. The protection offered to refugees was a matter of claiming jurisdiction over the refugee. The protection of the refugee was not so much based on the refugee's individual – legal – claim to protection. Authorities offered protection to refugees to claim and defend their own jurisdiction. By contrast, the official logic and ambitions of the CEAS ⁴⁴ are about jurisdictional integration and unification, whereby the protection of refugees is based on the individual rights of the refugee, fully in line with the approach of international refugee law pursuant to the Geneva Refugee Convention. The claim of this article is that precisely this focus on unity, integration and individual rights may unintentionally hamper the protection of refugees through law. Conversely, contrary to the official logic of the CEAS, this section will show how some of today's crucial improvements in the protection of refugees have been facilitated by the seemingly outdated mechanism of competition for jurisdiction.

The CEAS logic of jurisdictional integration and unity is most clearly expressed in the Dublin system. The crux of the Dublin Regulation is that the member state of first entry must process the refugee application. It rules out the possibility for refugees to arbitrate among the various jurisdictions of the Dublin member states. In other words, in the EU a refugee only gets to see one jurisdiction when it comes to asylum: vis-à-vis the refugee seeking access to the European asylum system, the Dublin countries present themselves as a single jurisdiction. The presumption underlying the Dublin mechanism is unity and commonality. Although, technically speaking, member states maintain their distinct and separate jurisdictions, they all act as one and the same when it comes to asylum. This is because the various regulations and directives of the CEAS are supposed to guarantee a homogenous and equivalent asylum approach across the CEAS member states. Hence, in theory it should not matter in what member state a refugee ends up asking for asylum. Furthermore, any potential heterogeneity in the interpretation of the common asylum system can supposedly be addressed by the CJEU through preliminary questions. ⁴⁵

The logic of jurisdictional integration and unity becomes even more prominent in the proposals to improve the CEAS. Typical failures of the current CEAS are the inadequate practices of processing asylum applications and the reception conditions in the member states that are dealing with large numbers of asylum applicants. It means that the presumed unity and commonality in asylum practices among member states do not hold true. Furthermore, the burdens are not being fairly shared among the member states. The seemingly logical response to these shortcomings of the CEAS is to enhance the integration and the unity of the system. Accordingly, EU legislation that still takes the shape of a ‘directive’ should be transformed into an instrument of direct integration ‘regulation’, e.g., a proposal to recast the Asylum Procedures Directive into the Asylum Procedures Regulation. ⁴⁶ Probably the most far-reaching proposal in terms of jurisdictional integration and unity is the recommendation by the EU Commission ⁴⁷ and refugee law experts ⁴⁸ to establish a single EU body responsible for dealing with all asylum applications in the EU.

Under the official logic of the CEAS, the protection of refugees is not based on the jurisdictional claims of member states, but on the individual rights of refugees. The CEAS originated out of genuine concerns for the rights and liberties of refugees. In effect, it seeks to prevent an ‘asylum lottery’, ‘refugees in orbit’ (caused by member states refusing to deal with certain asylum applications), legal uncertainty, unfair procedures, etc. So, the motives are well intended as far as refugees are concerned. Furthermore, where under international law it is still debatable whether an individual right to asylum actually exists, under the CEAS a refugee has such a right. ⁴⁹

Although the official motives of the CEAS are well intended vis-à-vis the refugee, the political dynamics in the member states are less welcoming towards refugees. In effect, many states simply want to avoid rather than receive refugees. This political dynamic stands in contrast to the context of the historical arrangements protecting refugees. Especially in the context of asylum, both prosecuting and protecting authorities claimed jurisdiction over the refugee. Until the nineteenth century a large, growing population (inter alia through immigration) was considered to be an economic and military benefit. ⁵⁰ This changed somewhere around the end of the nineteenth century and the beginning of the twentieth century.

Not only are the political dynamics different, but so is the nature of the refugee movements. Today's refugee movement consists mainly of large numbers of people who can often be categorized as belonging to groups. This corresponds with the phenomenon of the second wave of religious refugees fleeing persecution. By contrast the first wave of religious refugees and church asylum pertained to smaller numbers of refugees and were always provoked by actual or imminent individual prosecution.

For the purposes of clarity we may schematically contrast the official logic and dynamics of the CEAS with the historical arrangements offering protection to refugees as follows in Figure 1.

Figure 1. Comparison of refugee regimes

Over the last few years, the CEAS has undergone legislative amendments through the so-called recast, making some important improvements in the legal protection of refugees, e.g., the recast Asylum Procedures Directive. ⁵¹ These improvements have been greatly influenced by the case law of the CJEU and the ECtHR. ⁵² The contributions of the European courts fit nicely within the official logic and ambitions of the CEAS. At first glance, the case law is an expression of jurisdictional integration and the promotion of the fundamental rights of refugees. However, in what follows, I propose an alternative reading exploring how unexpectedly the mechanism of competition for jurisdiction still played a crucial role in some leading cases.

The Hirsi case ⁵³ is arguably the example par excellence of where the legal protection of refugees is made possible thanks to the logic of human rights. It is the universal scope of human rights that enables the protection of refugees to be extended beyond territorial borders. Italian coastguards and customs officials were conducting push-back operations on the Mediterranean high seas. The Italian officials intercepted possible refugees and returned them to Libya, without establishing their identity and offering them the opportunity to apply for asylum. The ECtHR ruled that the Italian operations resulted in a violation of Article 3 ECHR as the returnees had faced a serious risk of inhumane and degrading treatment in Libya. The Italians had violated the prohibition on refoulement. At first glance, the Hirsi case is an expression of the extraterritorial scope of human rights, in particular the ECHR. In other words, the legal protection to be offered to refugees following the Hirsi ruling is a reflection of the universal logic of human rights. I believe that this is a legitimate characterization of the Hirsi case. Yet there is an alternative reading that may be even more productive in explaining how the mechanics of legal protection really work, for we might also understand the Hirsi case as competition between jurisdictional claims over the refugees.

A brief passage in the Hirsi judgment (paragraphs 70–82), where the ECtHR establishes that the ECHR applies to this case, is crucial. Pursuant to Article 1 ECHR the contracting parties shall secure the rights and freedoms of the Convention for everyone within their jurisdiction. Accordingly, the Italians argued they did not have jurisdiction on the high seas and hence the ECHR did not apply. However, the Court found on the basis of previous case law and standard international law notions of jurisdiction that the Italians did exercise jurisdiction within the meaning of the ECHR. So, it is the struggle over jurisdiction between the Italian authorities and the ECtHR that actually paves the way for human rights to enter the fray.

Certainly the competition for jurisdiction is more complex than the mechanisms in our historical examples, for at the surface it is not about competing claims of jurisdiction. Quite the contrary, the Italian authorities were not claiming jurisdiction. Rather, they were denying jurisdiction. However, upon closer scrutiny there is a jurisdictional struggle taking place, albeit more involute. Firstly, the Italian authorities denied jurisdiction in order to exclude the application of the ECHR and thus to deny the jurisdictional claims of the ECtHR. Conversely, by refusing Italy's denial of jurisdiction, the Court could establish its own jurisdiction over the case. Secondly, although Italy denied jurisdiction over the refugees in one sense, it also implicitly claimed jurisdiction in another sense. Although the facts of the case do not mention it explicitly, the Italian officials most likely issued orders and directives towards the refugees during the actual operations (e.g., orders to remain seated or not move, form a queue, embark, disembark). By doing so the Italian officials claimed obedience or at least compliance by the refugees. This compliance was necessary for the orderly execution of the operations. Probably if the refugees had not complied, the officials would have considered themselves to have been legally authorized to apply coercive measures.

Of course it is technically possible for the Italian officials not to have issued any orders and commands to the refugees: The officials could simply have guided the behaviour of the refugees through physical means only. Also, it is possible that the officials took coercive measures without believing they were legally authorized to do so. But this would have meant that the Italian authorities had openly admitted that they had operated totally outside the law: They took measures vis-à-vis people over whom they lacked any legal authority. This position is difficult to substantiate, at least publicly. In other words, if the Italian authorities claimed to have acted legally, they necessarily must have claimed some kind of jurisdiction over the refugees. In effect, the Italian authorities denied jurisdiction in one way in order to claim jurisdiction over the refugees in another way. The Italian officials needed a form of jurisdiction over the refugees allowing them to issue legal directives and even coercive measures without the obligation to process asylum claims. In short, underneath the obvious universalistic logic of human rights governing the Hirsi case, we find competition for jurisdiction which is very similar to the historical arrangements offering protection to refugees, albeit more hidden and complex.

Like the Hirsi case, the obvious reading of the landmark cases of MSS and NS also amounts to the primacy of the human rights logic within the CEAS. ⁵⁴ The question in both cases was essentially the same. Does a Dublin country violate the prohibition on degrading and inhumane treatment when it returns an asylum applicant to the Dublin country responsible for examining the application, if the asylum procedures and reception conditions in the latter country would result in the degrading and inhumane treatment of the applicant? The European courts found that such a violation was possible (depending on the facts). It followed that under such circumstances the obligation under the Dublin Regulation to transfer the applicant must be set aside. Furthermore, if the Dublin country hosting the applicant cannot find, within a reasonable length of time, another Dublin country responsible for examining the application, it must process the application itself. In short, at first glance this is a clear case of human rights (Article 3 ECHR and Article 4 Charter of Fundamental Rights of the European Union, or CFREU) trumping the Dublin system.

Upon closer scrutiny and similarly to the Hirsi judgment, one finds that underneath the logic of human rights in the MSS and NS cases, competition for jurisdiction was taking place. In a particularly insightful analysis Morgades-Gil showed how the protection of human rights in both cases was actually triggered by the so-called sovereignty clause under the Dublin Regulation. ⁵⁵ The sovereignty clause ⁵⁶ allows a member state to examine an asylum application even if such an examination is not its responsibility under the Dublin Regulation. The sovereignty clause is an explicit acknowledgment of the state's prerogative to grant asylum: ⁵⁷ Doing so is an expression of the state's sovereignty.

In MSS the Belgian government argued that it was bound by the Dublin Regulation to transfer the applicant and that it could only use the sovereignty clause in exceptional situations which did not apply to the case at hand (paragraphs 326–7). Conversely, the ECtHR considered that Belgium was not fully bound by EU law to transfer the applicant, because it retained its sovereignty to examine the application itself. ⁵⁸ The sovereignty clause allowed Belgium to escape from its Dublin obligation to transfer the applicant and provided Belgium with the legal opportunity to prevent an Article 3 ECHR violation. In effect, as in the Hirsi case, a struggle over jurisdiction took place. The Belgian government denied its own jurisdiction, while the ECtHR precisely acknowledged the Belgian state's prerogative to grant asylum. In effect, the ECtHR found that jurisdiction comes at a price as a prerogative may also trigger an obligation: noblesse oblige. ⁵⁹

In NS the CJEU found that a member state may not transfer an asylum seeker to the Dublin country responsible for examining the asylum application if the asylum procedures and reception conditions may result in a real risk of inhuman and degrading treatment as prohibited by Article 4 CFREU (paragraph 94). Yet in order to check whether or not the transfer of an asylum seeker would amount to a violation of fundamental rights, the CJEU first had to establish that the Charter applied to the case at hand. Again, the sovereignty clause played a crucial role. The trigger for the case was the fact that the UK government had not honoured the request by the asylum seeker to make use of the sovereignty clause and to examine his application instead of Germany (the country responsible under the Dublin Regulation) (paragraph 40). The question before the Court was whether a member state when making a decision on the use of the sovereignty clause was implementing EU law and thus triggering the application of the Charter (Article 51 CFREU) (paragraphs 50 and 68–9). The Court found that a decision on the use of the sovereignty clause constitutes the implementation of EU law.

So, in both MSS and NS the European courts used the separate authority or jurisdiction of each member state to grant asylum to make way for human rights protection under the ECHR and CFREU respectively. Interestingly, each Court mobilized the sovereignty clause in a slightly different way. The ECtHR concluded that the sovereignty clause enables a Dublin country to retain a separate discretion in asylum matters, which means that it can escape the fully binding effect of the EU regime, i.e., the Dublin Regulation. This allowed the ECtHR to directly examine whether there was a violation of the Convention without applying the Bosphorus doctrine. In short, the ECtHR used the sovereignty clause to set aside the EU regime and to directly apply the Convention regime. By contrast, the CJEU argued that the sovereignty clause is an integral part of the EU regime. Hence, when a Dublin country makes a decision on the use of the sovereignty clause it is implementing EU law triggering the CFREU. In other words, the CJEU used the sovereignty clause not to set aside EU law but to allow the direct application of the EU Charter. What matters for present purposes is to see how at two levels a plurality of jurisdictions created a legal mechanism to protect refugees. Firstly, each individual Dublin country retains separate jurisdiction vis-à-vis the integrated Dublin system. Secondly, each European Court has its own separate jurisdiction which allows each Court to come up with its own construction of the sovereignty clause. The ultimate upshot is that the CEAS ceases to present itself as a singular jurisdiction vis-à-vis the refugee, paving the way for various opportunities for legal protection.

The European courts have ruled that when reviewing decisions on international protection and expulsion domestic courts should not only look at points of law but also the facts of the case. Furthermore, the courts should subject the decisions to a rigorous scrutiny. ⁶⁰ This case law directly resulted in the recasting of the Asylum Procedures Directive incorporating in Article 46(3) the obligation for the domestic courts of the member states to conduct a full ex nunc examination of both the facts and points of law when reviewing a decision refusing international protection. At first glance, it is a clear example of how the logic of human rights improves EU legislation for the benefit of refugees. By the same token, we may also read into this example an underlying logic of competition for jurisdiction which is very similar to the historical mechanism of asylum. In the case of church asylum, the refugee sought protection from a church authority or jurisdiction against the prosecuting authorities. The fact of being within the jurisdiction of the church authority provided the refugee with some immunity from prosecution. But not complete immunity. The church asylum operated as a safeguard that the refugee would obtain a fair trial and would not be subject to cruel punishment; it did not mean that the refugee would not be prosecuted, tried, condemned and punished. Similarly, in the EU context the refugee appeals to the jurisdiction of the ECtHR and CJEU for protection against the decision of domestic authorities. The European courts do not offer full immunity but simply intervene in order to ensure that the refugee can obtain a fair and thorough examination of his case. So, they respect the jurisdiction of a domestic court in so far as it lives up to the standards of fair procedures and rigorous scrutiny.

Still, things become different when the refugee resorts to the ECtHR to establish whether Article 3 has been violated. Here the Court will not simply defer to the fair procedures of domestic courts but conduct its own investigation. For the purposes of the present article, we can interpret this mechanism as an outright struggle over jurisdiction whereby the Court assumes full jurisdiction over the refugee and the case. It echoes earlier forms of church asylum whereby the church authority simply took over the prosecution of the refugee from the secular authorities.

To conclude my alternative reading of some influential recent cases through the lens of competition for jurisdiction, let us briefly consider one of the landmark cases of European asylum law, namely the Soering case. ⁶¹ In Soering the ECtHR held that Article 3 ECHR prohibits refoulement, i.e., expulsion, if there is a real risk that the person to be expelled will suffer ill treatment in the country of origin. The ruling is the basis for many cases filed by failed asylum seekers and has become a pillar of European asylum law. Yet the mechanism underlying Soering is very similar to the historical arrangement of asylum.

First of all, Soering was not fleeing persecution for any of the standard Refugee Convention reasons. He was a murderer facing ordinary criminal prosecution in the US and most likely would have ended up on death row. So, the figure of Soering corresponds very much with the classical refugee seeking church asylum. Secondly, a central question for the Court was its own competence in the matter: Is the Court competent to hear cases concerning potential Article 3 violations (as opposed to actual violations)? And does the Convention extend to violations that may take place outside the jurisdiction of a contracting party? The Court decided in the affirmative and claimed its jurisdiction over the case. Thirdly, Soering was also sought by other authorities. The US and German authorities were trying to exercise their criminal jurisdiction over Soering by obtaining his extradition from the UK. It is a classic case of asylum in the context of extradition which fully corresponds with the historical model of church asylum. Moreover, the Court explicitly found it relevant that the German authorities were also claiming Soering, for it offered the UK an alternative to extraditing him to the US. In other words, the multiple claims of jurisdiction over Soering made protection possible. Finally, in a classical reasoning echoing the official rationale underlying church asylum, the Court ruled against extradition because there were no sufficient guarantees that Soering would not face capital punishment and cruel or ill treatment. In short, even a pillar of EU asylum law is very much based on mechanisms almost identical to the historical arrangements offering protection to refugees.

Not only competition for jurisdiction at the supranational and transnational level creates opportunities for the legal protection of refugees at the local level, but also the constitutional (as in staatsrechtlich) structures within member states dividing competences and jurisdiction (e.g., the role of states versus the federal authorities in Germany, and city authorities versus national authorities in the Netherlands and Belgium). City authorities provide protection to rejected asylum seekers on the basis not of human rights or an individual right to asylum, but their own distinct and exclusive jurisdiction or competence in matters of public order, so offering protection for the sake of exercising power. ⁶²

4. Conclusions

Further to my comparative analyses of the historic and current refugee regimes, it may be tempting to draw some normative and prescriptive conclusions about what should be done. However, the method adopted in this article is rather ill-suited to support any material prescriptive conclusions. I have relied on legal history in order to identify conditions for the legal protection of refugees and compare those circumstances with the current European refugee regime. We must manage our expectations when it comes to the practical benefits of historical analyses: History cannot provide us with guarantees for what will work in the future.

Furthermore, the article did not address legitimate concerns about the downside of competing jurisdictions. It is beyond the scope of this article to do so, but I should mention probably the most serious normative objection. Promoting the plurality of jurisdiction would increase the ineffectiveness of EU refugee policy and compromise the legal protection of refugees. The approach would only aggravate the central problem of EU refugee policy: political and legal fragmentation.

This objection ultimately relies on an empirical assumption about the necessary connection between fragmentation and effectiveness. Examining this assumption lies beyond the scope of this article. Still, at first glance the connection seems more nuanced. There are many examples in other policy areas where the plurality of jurisdictions is compatible with and even beneficial to effective policies. This objection should consider these examples. ⁶³

But even if this normative objection is empirically sound and a plurality of jurisdictions and competition for jurisdiction hampers effective refugee policy, this does not necessarily damage the central aim of this article, for I am trying to show the connection between protection through law and competition for jurisdiction. If jurists are concerned about legal protection they should reconsider an outright promotion of integrating jurisdictions in refugee matters. However, I am not arguing in this article that legal protection is the only objective which is worthwhile pursuing. Personally, I believe that jurists should focus on and champion legal protection because experts in other disciplines are less likely and competent to take up the cause. But this leaves completely unaffected the possibility of there being good and overriding reasons for giving priority to effective policy over legal protection. In short, even if competition for jurisdiction hampers effective policy, it need not affect the argument posed in this article.

However, the plurality of jurisdictions most likely leads to diverging legal practices among jurisdictions causing legal inequality. Also, there is a risk of a race to the bottom. Legal inequality can be a serious problem. Yet in a way the plurality of jurisdictions is precisely a mitigating factor. What matters is whether there are sufficient alternative jurisdictions (local, national and supranational) available to which the refugee can appeal for protection. Similarly, the plurality of competing jurisdictions may operate as a mitigating factor against a race to the bottom. In theory the classical cure for a race to the bottom is centralized intervention. But this often presumes that the race to the bottom is caused by a collective action problem and moral hazard. In the context of refugee policy, it is much more a problem of how national policymakers can deal with (the perceptions of) national and local anti-refugee sentiments. It is not immediately clear how an integrated approach could stop such a race to the bottom.

Bearing in mind this legitimate normative objection against competition for jurisdiction, as well as the mitigating factors, by comparing the historical legal mechanisms with the CEAS we gained some insights into what worked and did not work in the past and what were possible factors for successes and failures. So, for now, my conclusions are fourfold.

Firstly, a basic condition for legal protection is the existence of multiple jurisdictions. This is the necessary basis for one of the major driving forces of legal protection, namely competition for jurisdiction. Secondly, the official logic of the current CEAS, in particular the Dublin system, is harmonization, unity and the hierarchy of jurisdictions. This logic hampers the legal protection of refugees. Furthermore, some crucial instances where the legal protection of refugees has been improved under the CEAS were actually thanks to competition between multiple jurisdictions in Europe. Thirdly and closely related to the second claim, the historic precursor to the CEAS in terms of refugee protection is the collective protection of religious refugees following the Counter-Reformation. The logic of this mechanism is more akin to regulation and management, rather than protection through legal and judicial procedures as was the case for church asylum. In short, the difficulties of protecting refugees through law under the CEAS could be partially explained as a lack of legal legacy. Finally, if anything like a policy recommendation could follow from the mainly descriptive findings in this article then it should be a call for institutionalizing competition for and among jurisdictions within the European refugee regime rather than aiming for more unity in refugee policy through establishing one single jurisdiction.