Testing the Applicability of EU Law Abroad: The Italy–Albania Protocol in the Comeri, Sidilli, and Sedrata Hearings

 

Eleonora Celoria, Post-doctoral researcher, FIERI

Andreina De Leo*, Post-doctoral researcher, Maastricht University

Marcella Ferri, Research Fellow, Florence University

* Funded by the European Union (ERC, SoftEn project, 101165167, PI: Lilian Tsourdi). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them.

Photo credit: Jorge Franganillo, via Wikimedia Commons

Introduction

On 23 and 24 March 2026, the Court of Justice of the European Union (CJEU) held two hearings in the ongoing proceedings addressing the compatibility of the Italy-Albania Protocol with EU law and the broader implications of conducting asylum and return procedures extraterritorially (Joined cases C-706/25 Comeri and C-707/25 Sidilli, and C-414/25 Sedrata).

This may appear as a déjà-vu. Less than a year ago, the extraterritorial processing of asylum under the Italy-Albania Protocol had already come before the Court in the well-known Alace and Canpelli cases (discussed on this blog by Zamboni and Favilli & Marin). The focus of those cases was, however, narrower, as it concerned the criteria to designate countries as “safe countries of origin” under the Asylum Procedures Directive (APD) and their judicial review, a classification that directly enabled the use of accelerated offshore asylum procedures in Albania. Following those rulings, the Italian Government adjusted its strategy, implementing a policy based on the relocation to Albania of returnees already detained in Italian pre-removal detention centers under the Return Directive (RD). While awaiting removal, these individuals may lodge an asylum application which, according to the Government, can then be examined in Albania. Against this backdrop, the new preliminary references raise a distinct and more direct question: whether asylum and return procedures can, as such, be carried out in a third country without breaching EU law. In doing so, they place the legality of the Protocol as a whole squarely before the Court.

More specifically, the first preliminary ruling, submitted by the Court of Cassation on 20 June 2025, asks whether Articles 3, 6, 8, 15, and 16 of the Return Directive preclude transferring a migrant subject to a return order to an extraterritorial detention centre, even without a concrete prospect of removal. It also asks whether Article 9(1) of the Asylum Procedures Directive, which requires applicants to remain on a Member State’s territory, precludes keeping asylum seekers who lodge an asylum application after being transferred to a third-country detention centre within that country (see De Leo).

The second preliminary ruling, submitted by the Rome Court of Appeal on 5 November 2025, directly questions Italy’s competence to conclude an international agreement  interfering with an area of exclusive EU competence, as laid down in Articles 4(3) TEU, 3(2) TFEU and 216(1) TFEU (see Montaldo and delli Carri). Alternatively, should the Court find that Italy has such competence, it asks whether the safeguards under the APD and the Reception Conditions Directive (RCD), as well as the Charter, particularly the rights to liberty, an effective remedy, and health, preclude transferring asylum seekers to, and detaining them in, centres outside EU territory.

It is worth recalling that the stated aim of the Protocol is to “reinforce bilateral cooperation in managing migration flows,” allowing Italy to transfer migrants to Albania solely to carry out asylum and return procedures under Italian and EU law (Art. 4(3) Protocol). These procedures remain under Italian jurisdiction and are conducted in accordance with national and EU law “to the extent compatible” (Art. 4(1), Ratification Law 14/2024).

Although raising different questions, both Italian courts focused on the same core issue: whether EU law safeguards can be effectively ensured for asylum seekers and returnees detained in a third country and whether those standards are compromised merely because detention occurs outside the territory of a Member State.

Outline of the post

While awaiting further developments, and notably the Advocate General’s opinions on 23 April (Sedrata) and 11 June (Comeri and Sidilli), this post examines the key issues raised during the hearings and outlines the main arguments on whether asylum and return procedures conducted in a third country, yet under a Member State’s jurisdiction, comply with EU law.

It first considers whether EU law applies to procedures carried out in Albania and whether such procedures may interfere with its application, summarising the positions of the Italian Government and the European Commission. Two aspects are examined: the Dublin system and its founding principle of mutual trust, and whether EU safeguards can be effectively guaranteed in detention centres abroad. Finally, the post clarifies the differences between the Albania model, the so-called “return hubs,” and the notion of a safe third country, as frequently referenced during hearings and in public debate.

EU Law Beyond Territory: Parties’ Arguments on Applicability, Compliance, and Competence in the Italy-Albania Scheme

The Italian government

To justify the compatibility of the Italy-Albania Protocol and its Ratification Law with EU law, the Italian Government advanced two alternative lines of argument. Both were aimed at demonstrating that: first, the Protocol does not risk adversely affecting the internal rules of the Union and therefore does not encroach upon an area of exclusive EU competence; secondly, that the extraterritorial application of the relevant directives in a third country does not undermine their effet utile.

First, the Government argued that the Albanian centres should be considered “representations of Member States” abroad under the APD and RCD, and that EU law applies indirectly through its incorporation via the Ratification Law. On this basis, the Government maintained that the Italy-Albania Protocol is purely technical and logistical: it does not alter the substance of the legal framework on pre-removal detention, asylum procedures or reception conditions, which remain fully applicable because the legislator has chosen to extend EU standards to the situations covered by the Protocol. By virtue of this legal extension, the Government maintained that third-country nationals detained in Albania receive treatment equivalent to that in Italy. It further argued that instruments such as the Dublin Regulation, though not explicitly mentioned in the Protocol, continue to apply because the individuals remain continuously under Italian jurisdiction. Accordingly, the Protocol would neither interfere with the Common European Asylum System (CEAS) nor encroach upon the Union’s external competences.

With respect to the RD, the Government argued that the transfer to Albania does not qualify as a “return” within the meaning of Article 3, but it rather amounts to a temporary relocation to a different pre-removal detention facility, legally assimilated to ones located on Italian territory. Accordingly, it maintained that compliance with the Directive’s objectives and provisions, including Articles 15 and 16 on detention and conditions of detention, are not undermined, as the material conditions would be fully respected. Regarding the APD, the Government contended that Article 9 is not violated when returnees apply for asylum after being transferred to Albania, since they remain under Italian jurisdiction, and thus do not risk refoulement, and continue to benefit from the full application of both the APD and RCD.

Alternatively, the Government argued that EU law could apply directly through a functional interpretation of “territory.” In the absence of a clear EU definition of “border” or “transit zones,” the centres in Albania could be regarded as a fictio iuris, functionally equivalent to transit zones located on Italian soil. On this view, the legal equivalence of procedures in Albania and Italy would justify treating the Albanian centres as falling within the scope of EU law, in the same way as their territorial counterparts.

While the Dublin Regulation was not explicitly mentioned by the referring judges, it emerged as a contentious issue. In particular, the risk of circumventing Dublin criteria for unaccompanied minors (Article 8) and family members (Articles 9–10) was debated during the Sedrata hearing. The Italian Government maintained that the Regulation remains applicable because the Albanian centres qualify as “transit zones” under Dublin III. This, however, raised a fundamental question from the bench: on what legal basis could another Member State be required to accept a Dublin transfer from a non-EU centre, and how would this not interfere with common EU rules? Questions which were left unresolved.

The European Commission

The Commission, by contrast, adopted a strict literal reading of the directives, distinguishing clearly between jurisdiction and territory. While acknowledging that the individuals remain under Italian jurisdiction, it stressed that asylum and reception rules are territorially anchored and do not apply automatically outside Member State territory. Nonetheless, the Commission did not view the Protocol as violating EU law or encroaching on exclusive EU competences, so long as the legislator’s choice to apply EU standards effectively preserves the effet utile of the acquis. Its reasoning followed two separate lines, depending on the instrument in question.

As for the RD, the Commission noted that it does not strictly define its scope in territorial terms but rather sets standards to ensure the effectiveness of returns from EU territory. Since Italy has committed to applying these standards in Albania, the Commission did not see any inherent incompatibility with the temporary transfer of returnees outside the territory pending repatriation. Yet, this conclusion was subject to strict conditions: the Commission emphasised that detention should have been previously judicially authorised, that all guarantees under Articles 15–18 of the Directive are fully respected, and that the transfer to Albania must not constitute the execution of the return decision. In essence, for the Commission the Directive does not require detention to occur on the territory of the Member State: what is rather decisive is that the Member State fully complies with its concrete obligations.

As for the APD and the RCD, the Commission rejected the qualification of the Albanian centres as “transit zones.” It stressed that the notion of territory is autonomous under EU law and cannot be extended to third countries, and that thus a “transit zone” must be located within the Union’s territory. When pressed by the bench on whether an international definition of “transit zone” exists, it indicated that this was ultimately irrelevant, as the notion must be understood as a matter of EU law. Accordingly, the Protocol cannot extend EU territory, and the direct applicability of the asylum acquis is thus excluded.

That said, the Commission did not view the mere extension, via national law, of EU standards to procedures in Albania as automatically undermining the directives’ objectives or interfering with internal EU competences under the ERTA doctrine. Such a violation would only arise if the arrangement risked affecting common EU rules: a risk the Commission considered unproven here. Article 9 APD illustrates this approach: if the asylum application is made in Italy, transfer to Albania is precluded. Conversely, if lodged only after transfer, while the Directive does not apply directly, its purpose, i.e. to prevent removal before assessing refoulement risks, is fulfilled because Italy retains jurisdiction and applies EU standards.

Regarding the Dublin system, the Commission held that the Regulation does not apply outside EU territory, revealing a potential paradox noted by Advocate General Emiliou during Sedrata: if EU law does not formally apply, can it still be circumvented?

More broadly, the question arises whether exclusion from the scope of the Regulation and Directives is merely a consequence of its incidental inapplicability in a third country, or the result of a deliberate choice by Italy, and namely, transferring individuals to Albania. In the latter scenario, how can it be argued that the Protocol and Ratification Law do not effectively allow Italy to evade its obligations under EU law?

As President Lenaerts highlighted, the key concern of the ERTA doctrine is preventing Member States from using external bilateral agreements to regulate matters that fall within EU law, in ways that could undermine its objectives. In other words, it is not enough to claim that EU law does not directly apply in a third country, but what matters is whether the external action could affect the effective internal functioning and uniformity of the EU legal framework. In this respect, the Commission’s position appeared unsatisfactory, as it failed to address the broader systemic implications for the CEAS and the structural risks of circumvention inherent in the scheme.

This brings us to our legal assessment, where we examine the unresolved issues in the approaches of the Italian Government and the Commission regarding the challenges of extraterritorialising asylum procedures and applying EU law in a third country.

Assessing the Potential Interference of Procedures Conducted in a third country with the EU Migration and Asylum Acquis

All in all, both the Italian Government and the European Commission ultimately arrive at a broadly similar conclusion: the Italy-Albania Protocol does not entail a violation of the Union’s exclusive competences, nor does it compromise the effet utile of the relevant directives, insofar as EU standards can be ensured, even in a third country. However, this conclusion leaves several key issues unresolved. First, it does not fully address the implications for the Dublin system. Indeed, the extension of Dublin obligations outside the Union could highly interfere with the CEAS and the principle of mutual trust. Second, the assumption that equivalent standards can be guaranteed extraterritorially remains highly problematic, as the territorial scope of EU asylum law reflects structural constraints necessary to ensure full compliance with procedural and substantive guarantees. These two issues will be addressed separately in the following sections.

The (application of the) Dublin system: a paradigmatic interference with the CEAS

The application of the Dublin system emerged as a highly sensitive issue during the Sedrata hearing. While EURODAC was not discussed, the reasoning for the Dublin Regulation can be extended to it, in light of recital 54 of EURODAC Regulation. The Italian Government and the Commission took sharply divergent positions: the Government qualifies the centres as “transit zones” under Dublin, a view opposed by the Commission, which raises several critical questions.

First, how can the Albanian centres be considered “representations of Member States” under the APD and the RCD, or “transit zones” under the Dublin Regulation? This argument is unpersuasive, as it would imply that the notion of “transit zone” differs across two CEAS instruments, risking to undermine the system’s internal coherence.

Secondly, if the Dublin system were applied to asylum requests lodged in the Albanian centers, other Member States would be obliged to accept transfers, as the Italian Government argued. From a theoretical perspective, this would directly interfere with the CEAS and undermine the principle of mutual trust on which it rests (Lenaerts, 2017). In fact, the presumption of compliance with EU law standards appears questionable in itself when it is applied to an asylum request lodged outside the territory of a Member State — albeit under its jurisdiction. Yet, save for specific exceptions, that principle prevents other States from verifying whether, in a given case, Italy has actually complied with EU law in processing an asylum application, notably that it has allocated responsibility in accordance with the procedural guarantees laid down in the Dublin Regulation and with the right to an effective remedy against the transfer decision. In short, extending the Regulation to applications lodged in the Italian centres in Albania would run counter to the principle of mutual trust underpinning the Dublin system and, ultimately, the CEAS.

Third, from a practical perspective, applying the Dublin Regulation outside Italy would adversely affect the rights of asylum seekers lodging an application in another Member State and seeking reunification with a family member detained in Albania (Article 10 of the Dublin Regulation). Unless the applicants were also transferred to Albania — which would amount to their de facto detention — they would remain in Italy, thereby rendering effective reunification  impossible. This impossibility, which stems directly from the detention of the family member in Albania rather than in a detention facility on Italian territory, would breach their rights to family life and to the child’s best interests, as laid down in the Dublin Regulation and in Articles 7 and 24 of the Charter.

By contrast, if, as the Commission argues, the Dublin Regulation does not apply, Italy could evade its obligations under Articles 8— 10 of the Regulation and Articles 7 and 24 of the Charter. Once again, such circumvention would be achievable simply through the deliberate extraterritorialisation of asylum procedures. In a nutshell, whether applied or not, the Dublin Regulation illustrates the systemic and structural interference with the CEAS created by the Protocol and Ratification Law.

Ensuring Effective Application of  EU Standards and Rights in a Third Country: Mission Impossibile?

We concur with the Commission that the territorial nature of EU asylum law imposes inherent limits on extending its directives beyond the Union. However, we diverge from the view that voluntarily applying EU standards abroad via national law automatically resolves compatibility concerns. This divergence is anchored in the structural reality that the territorial scope of the directives is not merely formal: it reflects the impossibility of fully guaranteeing procedural and material standards outside the Union. Applying EU law where these guarantees cannot be secured risks undermining the directives’ objectives, the uniformity of EU law, and may raise concerns under the ERTA doctrine (see: De Leo & Celoria, and Montaldo).

As highlighted by the referring judges and the lawyers representing the migrants, a central issue is the gap between the guarantees afforded to migrants detained in Italy and those in the Albanian centres. The Government asserts that procedures in Albania are identical to those in Italy, yet legal equivalence on paper does not ensure effective protection in practice. In our view, the fiction that Albanian centres are legally treated as part of a Member State is insufficient to secure genuine compliance with EU standards on the ground.

The main concerns identified by the referring courts and discussed during the hearings include: the right to liberty under Article 6 of the Charter, notably the requirement for immediate release if detention is unlawful; guarantees related to detention conditions, including access for lawyers, family members, national, international, and non-governmental organisations, and access to healthcare; and the right to a fair trial and effective remedy under Article 47, particularly the right to be “advised, defended, and represented.” As emphasized by the rapporteur judge of Comeri and Sidilli case, these discrepancies risk creating a distinction between “two types of asylum seekers,” or even “two types of detainees,” solely based on where and when an application is lodged.

Both the RD and the RCD require that a person “shall be released immediately” if detention is found unlawful (Article 15(1) RD; Article 9(3) RCD). This requirement is inherently impossible to reconcile with the Italy–Albania scheme. Under Article 6(2) of the Protocol, Italian authorities must “take the necessary measures to ensure the permanence of migrants within the areas, preventing their unauthorised exit into the territory of the Republic of Albania, both during and after the completion of administrative procedures, irrespective of the final outcome.” In practice, release on Albanian territory is impossible: individuals can only be freed once transferred back to Italy. Therefore, even when detention is considered unlawful, the person remains under continuous coercive control throughout the transfer, i.e. from the detention center to the port, during maritime transport, and until release in Italy (typically in Bari or Brindisi): a process that may last from 24 hours to several days, depending on logistical conditions. In this respect, the Italian National Guarantor has recently noted that, if transfer cannot occur the same day, the person remains in a “separate area” of the same detention centre, a practice arguably failing short of the “immediate release” requirement. During this period, individuals are subject to a “coercive measure that deprives [them] of [their] freedom of movement and isolates [them] from the rest of the population” (CJEU, FMS, para. 223). Both the Albanian centres and the transport arrangements display the defining features of detention: a closed, restricted space where movements are limited and monitored and exit is not voluntary (CJEU, FMS, para. 231). The European Court of Human Rights similarly confirms that confinement on ships may constitute de facto detention, even if nominally for the person’s interest (Khlaifia and Others v Italy, paras. 70–71). In sum, the extraterritorial setting alters the legal consequences of detention, making immediate release practically unfeasible and raising serious questions about compliance with EU law.

Extraterritoriality also affects effective access to detention centres. Under national law implementing EU standards (Articles 16(2) and (4) RD; Article 10(4) RCD), a wide range of actors, e.g. parliamentarians, national monitoring bodies, UNHCR representatives, lawyers, family members, religious ministers, and civil society organisations, should be granted access (Articles 21 DPR 394/99; 7 D.Lgs. 142/2015; Ministerial Directive 19 May 2022). By contrast, Article 9(2) of the Protocol limits access to “lawyers, their assistants” and “international organisations and EU agencies,” raising questions about the interplay of the two regimes. Even if Italy seeks to apply its broader framework, as argued during the hearing, effective implementation depends on Albanian authorities’ consent. This is because access requires entry into Albanian territory, which remains under Albanian sovereignty. Article 9(2) subjects access to “applicable EU, national and Albanian law,” while Article 6(1) confirms Albanian responsibility for public order and security outside the centres and during transfers. Consequently, Albania may lawfully restrict access based on domestic entry, residence, or public order rules, limiting in practice the actors able to monitor detention conditions. While the Italian Government argued that this poses no practical concern, citing visa-free entry for holders of Italian passports or residence permits, the effectiveness of EU law guarantees cannot depend on conditions outside the Member State’s control. Making access contingent on the rules or discretion of a third country introduces uncertainty incompatible with the requirement to ensure EU rights in an effective and uniform manner. Moreover, denials imposed by Albanian authorities cannot be effectively challenged before an Italian court, which undermines the right to an effective remedy for those having the right to access the detention centres.

Similarly, under Article 10(4) of RCD, lawyers, family members, and civil society organisations have the right to communicate with and visit detainees, and any limitations must not make access impossible or excessively difficult. In the Albanian centres, however, restrictions appear structural and generalised rather than exceptional, making them difficult to reconcile with Articles 7 and 47 of the Charter and the proportionality requirement under Article 52(1). The Government’s argument that similar limitations exist in Italy is unpersuasive: in Albania, restrictions are the norm, access is more complex, time-consuming, and costly, and a recent monitoring report by the Italian National Guarantor for the Rights of Persons Deprived of Liberty notes that some detainees were not informed of their transfer, further hindering visits. Crucially, as noted above, the ultimate decision rests with Albanian authorities, meaning that limitations stemming from a third State’s sovereignty cannot be effectively challenged or remedied in Italy, thereby weakening the effective enjoyment of EU-law-derived rights.

Additionally, discrepancies in safeguards for lawyers are particularly pronounced. In Italy, appointed lawyers can access detention centres without prior authorisation and often be appointed in person after initial telephone contact. In Albania, this is not possible, nor foreseen by the Protocol or the Ratification Law, which allows travel at public expense only “when remote connection is not feasible” during the detention validation hearing (Article 4(5), Law 14/2024). This effectively precludes in-person meetings before or after the hearing to prepare a defence or appeal a negative decision. Given the current five-day deadline to challenge a validation decision, detainees without immediate lawyer access face heightened risk of ineffective representation. In-person meetings are often essential to identify vulnerabilities or health conditions that may render detention unlawful, tasks difficult to perform remotely, particularly when access depends on the private centre manager’s discretion and mobile phone use is restricted. Physical distance, limited contacts, and communication restrictions materially weaken lawyer-client interaction, undermining the right to effective assistance and representation under Article 47 of the Charter. Remote hearings further exacerbate these issues. By analogy with the Court’s reasoning in FP and Others (C‑760/22), videoconference participation is not inherently incompatible with a fair and public hearing, but its appropriateness depends on the individual being able to follow proceedings, be heard without technical obstacles, and communicate effectively and confidentially with their lawyer. In Italy, lawyers attending remote validation hearings can choose to be in the courtroom with the judge or at the centre with their client: a flexibility that is simply unavailable for detainees in Albania, further undermining effective representation.

Finally, significant disparities exist regarding the right to health, guaranteed under Article 16(3) RD and Article 17(2) RCD. Detainees in Albania face substantial limitations in medical services, particularly psychiatric and addiction care. Unlike in Italy, where detainees are integrated into the National Health System, no equivalent framework operates in Albania, and on-site medical teams cannot provide comparable treatment. These deficiencies have already led to serious consequences, including self-harm and suicide attempts (see, report, here). A recent report by the National Guarantor for the Rights of Persons Deprived of their Liberty found that, of 26 detainees for whom updated medical assessments were requested, 25 were deemed unfit for detention and released, highlighting the centres’ inadequacy to ensure proper care. While Article 4(8) of the Protocol obliges Albanian authorities to provide treatment in emergencies beyond Italian capacity on site, healthcare in a third country is by definition different from that provided by a Member State’s National Health System.

Taken together, the disparities analysed above make detention in centres located in a third country significantly more burdensome than detention in Italy, casting doubts as for its compliance with the principle of proportionality. There is no justification for imposing a harsher coercive measure than would be applied in Italy, particularly as the transfer to Albania does not demonstrably enhance the effectiveness of return procedures under the RD. On the contrary, a report by the Italian National Guarantor for the Rights of Persons Deprived of their Liberty shows that only 56 of 192 individuals transferred to Albania were ultimately returned to their country of origin, representing a return rate of approximately 29%, which is lower than the general return rate.

Conclusion

In conclusion, while the Italian Government frames the issue in terms of formal equivalence between procedures carried out in Albania and those in Italy, and the European Commission emphasizes that the voluntary application of EU standards suffices to ensure compliance, our assessment shows that this approach does not adequately address the structural limitations inherent in the de facto extraterritorial application of EU migration and asylum law.

While awaiting the Court’s decision, we maintain that the strict territorial scope of EU law in this area is not merely formal but reflects a substantive requirement: it is meant to operate within the territory of Member States, where its guarantees can be fully implemented and monitored. Remaining under the jurisdiction of a Member State alone is insufficient to ensure compliance, as EU law cannot be effectively applied outside the Union without undermining its purpose, coherence, and the procedural and material protections that are inherently tied to the territorial context in which they are guaranteed.

Two key issues arise. First, if the Dublin Regulation applies to applications lodged in Albania, the Protocol and Ratification Law would undermine mutual trust and the functioning of the Dublin system, since other Member States cannot verify compliance with EU law outside the Union. Even if it does not apply, the deliberate delocalization of asylum procedures to a third country allows circumvention of EU obligations. Second, procedural and material safeguards ensured in Italy cannot be fully replicated in a third country.

Thus, in our view, relocating individuals to Albania undermines the effet utile of the CEAS, creating tangible discrepancies in treatment that formal equivalence cannot remedy. This, in turn, risks breaching the principle of loyal cooperation and jeopardizes both the uniform application of EU law and the trust-based cooperation that underpins the EU legal order.

A Final Note: Why Albania Is Different from Other Externalisation Models

Throughout the hearings, “return hubs” were frequently cited as a possible analogy to justify the legality of the Albania model. However, as repeatedly emphasized by all parties, the two schemes are fundamentally different from a legal perspective. The Commission’s proposed reform of the notion of a ‘return decision,’ which provides the legal basis for return hubs, presupposes a formal removal decision to a country other than the state of origin with which an agreement exists. In that context, the transfer constitutes an actual return under EU law (discussed, in this blog, here). By contrast, the Albania model does not involve a removal but a temporary relocation of the detention stage of the return procedure under the full jurisdiction of the Member State, prior to any formal removal to the individual’s country of origin or habitual residence. Consequently, it cannot be assimilated to the return hub framework.

Similarly, the Albania model cannot be framed within the reformed notion of a safe third country without a connection requirement (discussed, in this blog, here), as no inadmissibility decision is taken in Italy based on Albania’s acceptance of the asylum application. Instead, the procedure merely relocates the processing of the asylum claim to Albania while keeping the substantive application under Italian jurisdiction.

This distinction highlights that equating the Albania model with return hubs or safe third country schemes overlook its unique legal and procedural features. It underscores the complexity of assessing its compliance with EU law and why the pending preliminary rulings are essential to clarify the boundaries and practical limits of this peculiar migration management scheme.

Migrant deaths in the Mediterranean: What can the EU do?

 

Steve Peers

The latest tragedies in the Mediterranean add to the enormous and growing death toll of migrants crossing that sea. Today EU ministers will consider the issue as a matter of urgency, and the Commission is due to propose an EU strategy next month. Here are some thoughts as to what parts of the EU response should be.

First of all, it is necessary to reinstate a major search and rescue operation. The EU’s ‘Triton’ operation which replaced the Italian ‘Mare Nostrum’ operation from the end of last year is obviously inadequate as regards saving lives – which is not its main purpose in the first place. Equally the assumption that the termination of Mare Nostrum would reduce the number of migrants willing to undertake the Mediterranean crossing – and die in the attempt – was obviously mistaken.

The new operation need not be an EU operation as such, due to legal questions about how much the EU as such can set up a search and rescue operation. Possibly the EU, or at least a large group of its individual Member States, can contribute toward the coordination and funding of a joint operation, rather than leave this entirely up to the Italian government as before.

Secondly, the EU has an established system for border surveillance, known as Eurosur. Its main focus in border control but it has a subsidiary role as regards detecting migrants in need of rescue. It could be focussed more on the humanitarian aspect, if necessary by means of a rapid amendment to the legislation establishing it – or alternatively, if there are legal doubts about this, by a parallel agreement among the Member States.  

Thirdly, the EU should address the important role of the private rescue of migrants. At present, the EU Directive prohibiting the smuggling of migrants specifies that any assistance to irregular migrants is prohibited – even if the person or organisation concerned is not acting for profit. There is only an option for Member States to exempt humanitarian assistance from the prohibition. This legislation should be amended as soon as possible to make this exception mandatory, clarifying that this also exempts private sea vessels which assist those in need (in accordance with the law of the sea) also benefit from the exception. Private vessels which lose money as a result of offering assistance (for instance, because they were diverted from catching fish) should be compensated by national or EU funds if necessary. This amendment could be fast-tracked, pending a more comprehensive review of the legislation to consider the best criminal law response to smugglers and traffickers.

Fourthly, there is a need to address the situation in countries of origin and transit, so that fewer people feel the need to make this risky journey in the first place. This also takes account of realpolitik: it’s not socially, economically or politically feasible for the EU to offer protection to everyone in the world that needs it. Of course, solving the conflicts and poverty that cause desperate people to flee in the first place is far easier said than done. But it would certainly be possible to fully review EU foreign policies and development aid funding with a view to addressing the root causes of migration.

In terms of a legal response, there are several other external aspects of EU migration policy. For those who have a need for international protection, many already receive support or protection in neighbouring and transit countries, and the EU could step up its contribution to such assistance.  A proportion of the people concerned are already resettled by Member States, and the EU could increase its support for this resettlement process. It might be useful to amend EU asylum legislation to confirm that the persons concerned have refugee status in the Member States concerned, in accordance with the EU’s ‘Qualification Directive’, upon arrival.

There’s also an existing legal framework for Member States to consider applications for asylum lodged at their consulates in third (non-EU) countries. At the end of 2013, the EU’s main court (the CJEU) already ruled in the Koushkaki case that the EU’s current visa code requires Member States to issue visas to applicants if the criteria for issue are met. The visa code mainly applies to short-term visas, not visas for people who need long-term international protection. However, the code also provides for rules on the issue of a short-term visa with ‘limited territorial validity’ (LTV). These visas are not valid in all Schengen states, but rather valid usually in only the single Schengen State which issues them.  

According to the current visa code, an LTV visa ‘shall be issued…when the Member State concerned considers it necessary on humanitarian grounds, for reasons of national interest or because of international obligations’. The important point is that an LTV visa can be issued where the usual conditions for issuing a visa are not met, for instance where there is insufficient evidence of an intention to return to the country of origin. Obviously, where a person has a genuine protection need, a reluctance to return to her country of origin is perfectly understandable; indeed, it is built into the very definition of refugee or subsidiary protection status (ie a well-founded fear of suffering persecution or serious harm in that country).

It’s arguable the 2013 CJEU ruling in the Koushkaki judgment also applies to LTV visas, in light of the word ‘shall’. Admittedly, that word is then qualified by the words ‘considers it necessary’. But arguably, at least in cases involving a potential international protection need, the EU Charter of Fundamental Rights requires that where a person applies for a visa from a (Schengen) Member State, the existence of such a need must be considered if it is alleged, and an LTV visa must be issued if such a protection need exists. It can hardly be denied that EU law (and therefore the Charter) applies whenever a third-country national applies for a visa from a Schengen Member State.

So the EU and its Member States could agree that this is the correct interpretation of the existing law, and take action to put it into effect, for instance adopting guidelines or standard rules for consulates on how to apply such obligations. EU funds could assist with any additional costs for Member States.

There’s also an immediate opportunity to address this issue by means of legislation, since proposed amendments to the visa code that would fully overhaul it are currently under discussion in the Council and the European Parliament (EP). These proposals raise many different issues, and the Council and EP are still at an early stage of discussing them. But it would be possible to split up the proposal and adopt an amendment to the LTV rules on a fast-track basis.

In a previous blog post, I suggested a simple amendment to the visa code to this end. This would simply amend the rules to state that an LTV visa ‘shall be issued…when it is necessary in order to ensure the international protection of the person concerned in accordance with Directive 2011/95 [the Qualification Directive], or when the Member State concerned considers it necessary…’.

In the longer term, other difficult and controversial aspects of an external protection policy could be developed. It would be possible to develop ‘joint external processing’ of asylum claims, as long as such processing in no way prejudices applications made at the border or on the territory of Member States, and entails the entry and stay of persons with international protection needs either on the territory of Member States, or on the territory of third countries which offer an equivalent level of protection. Nor should joint processing prejudice the resettlement of those groups of persons (such as Syrian refugees) whose need for international protection is obvious.

The problems associated with such joint processing should not deter the EU from taking the immediate steps outlined here that could reduce the death toll on the Mediterranean by stepping up rescue efforts and by giving more people with an established protection need a form of safe passage to the EU – either by means of resettlement of established refugees or assessment of asylum-seekers’ protection needs as an adjunct to the visa application process. There is a moral imperative for the EU to act swiftly and effectively to address the issue.

 

Photo credit: Daily Mail

Barnard & Peers: chapter 26

External processing of applications for international protection in the EU

Steve Peers

Last autumn's huge loss of lives near Lampedusa, when hundreds of migrants drowned in the Mediterranean, was one of the latest and most dramatic death tolls in the recent history of irregular crossing of that sea. It ought to have led to a complete rethink of EU policy toward border controls and visas, but did not - doubtless because of the belief that far-right parties would capitalise on the increased public concern about migration that would result.

Of course, irregular migration flows are 'mixed': they consist of some people with a legitimate need for international protection, as well as some people who don't have such a need. The former group has a right to enter and stay on Member States' territory, while the latter group in principle does not - although this obviously doesn't mean that such migrants should be left to drown.

While no potential solution to this dilemma is simple, or would solve all the related problems, a move towards greater external processing of asylum seekers would clearly be a step in the right direction. It would mean that at least some of those migrants with a legitimate need for international protection would be able to obtain safe passage to EU Member States' territory without having to risk crossing the Mediterranean in unsafe vessels, having had to pay unscrupulous smugglers a fortune to arrange their journey.

For the last decade, the idea of external processing of asylum applications was tarnished by its association with suggestions made by Tony Blair for a new approach to asylum seekers coming to Europe.  His ideas, which seemed to have been worked out on the back of an envelope, and were clearly inspired by Australia's 'Pacific solution' to refugees, envisaged external processing as the exclusive route for asylum seekers to obtain refuge in the EU.

They were incompatible with human rights obligations, and entailed a degree of cooperation from third countries which either would not have been offered, or should not have been accepted if it was. This approach might have entailed agreements with Khadafy's Libya, or Putin's Russia, to host refugees and asylum seekers. If Putin were currently in a position to release many thousands of refugees who wished to make their way to the EU, the Union's reaction to his annexation of Crimea and sabre - rattling in eastern Ukraine would be even milder than it is already - if that's possible.

A new approach

Blair's ideas have been kicked into the long grass several times by the EU. But certainly the idea of joint external processing in principle has merit, if it has completely different premises from his suggestions. In particular, such processing should in no way prejudice applications made at the border or on the territory of Member States, and must entail the entry and stay of persons with international protection needs either on the territory of Member States, or on the territory of third countries which offer an equivalent level of protection. Nor should joint processing prejudice the resettlement of those groups of persons (such as Syrian refugees) whose need for international protection is obvious.

However, the EU will need time to work out the details of joint processing, for instance how to allocate the beneficiaries of international protection between Member States. In the meantime, there is now a perfect opportunity to adopt rules on purely national external processing of asylum applications, in the form of the proposed amendments to the Schengen visa code.

This is the second of four posts on this blog dealing with these proposed amendments. The first post dealt with the impact of the proposals on EU citizens’ third-country national family members, and the remaining posts will deal with the substance of the proposals as regards short-term Schengen visas and the newly proposed ‘touring visa’.

The proposed amendments to the visa code retain the existing possibility for Member States to issue a short-term visa with ‘limited territorial validity’ (LTV), ie the visa is not valid in all Schengen states, but rather valid usually in only the single Schengen State which issues it. But admission to only one Member State is clearly better than facing a risk of harm outside the EU. Once the visa expires, the person concerned can be given a longer-term residence document; in fact, the EU’s qualification Directive requires that a recognised refugee or beneficiary of subsidiary protection must receive a residence permit, and the EU Directive on asylum procedures specifies that in principle asylum-seekers cannot be removed from a Member State’s territory before a decision is made on their application.

The new proposal would make no substantive changes to the current rules in the visa code on LTV visas (see Article 22 of the proposal, as compared to Article 25 of the current code). However, it would be open to the European Parliament and the Council to insist that changes ought to be made.

According to the current visa code, and the proposed amendments, an LTV visa ‘shall be issued…when the Member State concerned considers it necessary on humanitarian grounds, for reasons of national interest or because of international obligations’. The important point is that an LTV visa can be issued where the usual conditions for issuing a visa are not met, for instance where there is insufficient evidence of an intention to return to the country of origin. Obviously, where a person has a genuine protection need, a reluctance to return to her country of origin is perfectly understandable; indeed, it is built into the very definition of refugee or subsidiary protection status (ie a well-founded fear of suffering persecution or serious harm in that country).

It should be noted that the CJEU has recently ruled in the Koushkaki judgment that in principle an ordinary Schengen visa must be issued when the applicant satisfies the criteria to obtain one, subject to a wide degree of discretion for Member States’ authorities to assess whether those criteria are satisfied. Does the same rule apply to LTV visas? At first sight, it does, due to the word ‘shall’, although that is qualified by the words ‘considers it necessary’.

Arguably, at least in cases involving a potential international protection need, the EU Charter of Fundamental Rights requires that where a person applies for a visa from a (Schengen) Member State, the existence of such a need must be considered if it is alleged, and an LTV visa must be issued if such a protection need exists. The Member State’s Charter obligations would also be satisfied if it issues an LTV visa to the person concerned when the application for international protection is made, and considers the merits of the application while that person is on its territory. It can hardly be denied that EU law (and therefore the Charter) applies whenever a third-country national applies for a visa from a Schengen Member State.

The substantive law applicable to the consideration of the application would be the EU’s Qualification Directive, since that Directive does not limit its geographical application. However, the EU’s legislation on asylum procedures and reception conditions only applies to applications made at the border or on the territory of Member States. But since the EU Charter applies to applications for visas made on the territories of third States, it must follow that some basic standards compliant with the Charter would apply to the procedures and reception conditions would still be applicable. There would be no need to decide which Member State is responsible for considering the application, since the EU’s Dublin rules are subject to the same geographic limitations.

If this interpretation is correct, the current and proposed visa codes already include implicit rules covering those applying for international protection. However, it would be preferable to include express rules to this effect. There would be no need for elaborate provisions on this issue, since the details of such purely national forms of external processing of asylum applications do not need to be harmonised in great detail. It would be sufficient to provide that an LTV visa ‘shall be issued…when it is necessary in order to ensure the international protection of the person concerned in accordance with Directive 2011/95 [the Qualification Directive], or when the Member State concerned considers it necessary…’.

This would be a modest but important step towards reducing the appalling death toll of those migrants who cross the Mediterranean in search of a new life in the European Union.


Barnard & Peers: chapter 26