Refugee Camps at EU External Borders, the Question of the Union’s Responsibility, and the Potential of EU Public Liability Law

‘The EU hotspot approach as implemented in Greece is the single most worrying fundamental rights issue that we are confronting anywhere in the European Union’. This quote by the head of the EU Agency for Fundamental Rights (FRA) might sound drastic. Yet, it is not far-fetched. EU bodies, national institutions, international organisations including the Council of Europe, and NGOs, have, during the past four years, continuously documented that the asylum processing centres at the EU external borders lead to fundamental rights violations on a daily basis. The EU hotspot administration indeed jeopardises the respect for fundamental rights and the rule of law as enshrined in Article 2 TEU.

Usually, when something is going wrong, a first step towards improvement is to ask: who is responsible? And yet, with regard to EU hotspots, this question is still subject to debate. Responsibilities are effectively blurred by the sheer number of actors operating in those centres combined with a lack of legal clarity. On the political level, this leads to responsibility-shifting between the European Commission, Greece and local municipalities. On the legal level, so far, only Greece as the host Member State is considered responsible, namely under the ECHR. The considerable involvement of the Commission and EU agencies—in particular Frontex and the European Asylum Support Office (EASO)— however suggests to look to EU law and to examine whether and to what extent the European Union is legally responsible.

It is argued here that EU public liability law—more specifically: an action for damages against the Union or its agencies Frontex and EASO—has a particular potential in this context. First, it would help secure the right to an effective remedy to concerned individuals. Second, it would thereby serve to address systemic deficiencies in the EU hotspot administration. Third, it could ultimately provide an answer to the crucial question of whether the Union is responsible for fundamental rights violations in EU hotspots.

1 – The violation of fundamental rights in EU hotspots—systemic deficiencies

In 2015, the Commission put forward the EU hotspot approach as part of the European Agenda on Migration. While the approach is implemented both in Italy and Greece, this contribution focuses on the latter. Each of the five EU hotspots in Greece, located on Aegean islands, consist of a refugee camp, an administrative complex, and, in some cases, a pre-removal detention facility. In March 2016, with the implementation of the EU-Turkey Statement, the EU hotspots were transformed into return centres meaning that the asylum procedure and the reception conditions were adapted to the aim of return. Currently, about 41,000 persons are staying in those camps.

The approach of ‘processing asylum claims at borders, particularly when these centres are located in relatively remote locations, creates fundamental rights challenges that appear almost unsurmountable’. This assessment by FRA seems plausible given the empirical evidence provided by the already four-years long ‘hotspot experiment’. More specifically, FRA finds fundamental rights risks with regard to, inter alia, Articles 1, 4, 5(3), 6, 7, 18 and 19, 20 and 21, 24, 25 and 26, 41 and 47 of the EU Charter on Fundamental Rights (ChFR). Two aspects deserve particular attention.

First, the reception conditions are far from complying with any standard of EU secondary law and wholly inadequate for human beings: Shelter is insufficient (if there is any), there exists exposure to extreme weather conditions, a high risk of sexual, gender-based and other forms of violence, a lack of medical services despite widespread physical and severe psychological health issues, insufficient and inadequate sanitary facilities, and a lack of access to education or social services. Taken as a whole, the reception conditions arguably amount to a violation of Article 4 ChFR prohibiting inhuman or degrading treatment, at least insofar as vulnerable persons are concerned. This follows from the standards established by the CJEU from N.S. to Jawo, taking into account the case law of the ECtHR from M.S.S. to Tarakhel. Concerning EU hotspots specifically, the ECtHR seems to slowly change its jurisprudence: In contrast to earlier decisions concerning the situation in March 2016, a violation of Article 3 ECHR was found in more recent interim measures concerning vulnerable persons. Even if one assumes that a violation of Article 4 ChFR can be found only for vulnerable persons, this still affects a considerable number of people.

Second, a deportation to Turkey, at least in the vast majority of cases, would be in breach of the Asylum Procedures Directive, since Turkey cannot be considered as safe third country or first country of asylum. This is, despite the differing decision of the Greek Council of State, in line with the view of the Greek administration (and the Administrative Court of Munich). Considering the situation in Turkey, it seems that, at least for the vast majority of persons, the deportation would amount to a violation of the non-refoulement principle as enshrined in Articles 4, 18, 19(2) ChFR. This follows from the minimum standards established by the ECtHR in Ilias and Ahmad with regard to Article 3 ECHR. (The CJEU has not yet established the constitutional standards following from Articles 4, 18, 19(2) ChFR: The decision in Alheto concerns a specific case, and the decision in LH  remains to be awaited). With regard to the situation in Turkey specifically, an individual complaint before the ECtHR is pending.

Those two aspects speak in favour of describing the implementation of the EU hotspot as systemically deficient. Both a breach of Article 4 ChFR as well as breach of the non-refoulement principle as enshrined in Article 4, 18, 19(2) ChFR meet the threshold of being relevant for Article 2 TEU. Further, both breaches are systemic in the sense of widespread or inherent to the situation: An arguable limitation to the sub-group of vulnerable persons does not hinder the qualification as systemic. Due to the design of EU hotspots as return centres, the question whether deportations to Turkey violate the non-refoulement principle is, despite the relatively low numbers of returns, of structural relevance.

2 – The considerable involvement of the Union in the EU hotspot administration

Against this background, it is worthwhile to have a closer look at the involvement of the Union in the EU hotspot administration. From the perspective of EU administrative law, the distinctive characteristic of EU hotspots, in comparison to other asylum processing centres at EU external borders, is the close administrative cooperation between Union bodies and national authorities. This becomes clear already from Article 2(23) Frontex Regulation defining a ‘hotspot area’ as an area ‘in which the host Member State, the Commission, relevant Union agencies and participating Member States cooperate, with the aim of managing an existing or potential disproportionate migratory challenge characterised by a significant increase in the number of migrants arriving at the external borders’.

The EU hotspot administration can hence be described as the paradigm example for advanced vertical administrative cooperation within the integrated European asylum administration. This means that several EU agencies—such as Frontex, EASO, Europol, and Eurojust—cooperate with several national authorities—such as asylum service, reception service, police, and army. In practice, international organisations such as UNHCR and IOM, several NGOs, and a private security company operate in those centres in addition.

The operational level—the role of Frontex and EASO

On the operational level, migration management support teams (MMST) deployed by the EU agencies support the Greek authorities. The distinctive feature of the MMST lies, inter alia, in the close inter-agency cooperation. While Frontex supports in particular by registering applicants and escorting deportations to Turkey, EASO supports notably by conducting asylum interviews and drafting legal opinions recommending the acceptance or rejection of the concerned individual’s claim for international protection.

With a view to EU public liability law, it should be kept in mind that the responsibility to issue administrative decisions lies with the host Member State. The role of Frontex and EASO is to provide non-formally binding administrative support. However, the line between formally-binding and non-formally binding is not that easy to draw: Non-formally binding administrative conduct can have de facto binding effects on national authorities, as illustrated by EASO’s involvement in the assessment of asylum claims. And non-formally binding administrative conduct can have quite significant effects on individuals, in particular since the reformed Frontex Regulation does not exclude the use of force by Frontex MMST staff.

The coordination and monitoring level—the role of the Commission and the EURTF

On the coordination and monitoring level, responsibility lies with the European Commission, who is supported by Frontex, EASO, and the other relevant EU agencies in this respect. Article 40(3) Frontex Regulation provides that the ‘Commission, in cooperation with the host Member State and the relevant Union bodies, offices and agencies (…) shall be responsible for the coordination of the activities of the migration management support teams.’ The Commission performs this task within the framework of the EU Regional Task Force (EURTF). The EURTF is a coordination structure which has been established without a clear legal basis and operates under non-public ‘terms of cooperation’ and ‘rules of procedure’.

With a view to EU public liability law, it should be noted that the Commission’s mandate includes the supervisory obligation to ensure that the EU hotspot approach is implemented in line with EU law. This becomes clear already from Article 40(3) Frontex Regulation, read in light of its Article 1 and recitals. Further, and more importantly, this follows from Article 17(1) TEU, as interpreted by the CJEU in Ledra, as well as from Article 51 ChFR.

3 – The Potential of EU Public Liability Law—enforcing EU law from below

The Commission, Frontex, and EASO are hence closely involved in the EU hotspot administration which is systemically deficient, and leads to fundamental rights violations in individual cases. This gives rise to the crucial question: Can the Union be held responsible? A legal regime which could provide an answer to this question would ideally grant the right to an effective remedy to the concerned individual and enforce the rule of EU law more generally, while at the same time allowing for the attribution of responsibility among the involved actors.

It is argued here that EU public liability law has a particular potential in this context due to its subjective and objective legal protection function combined with its attribution function. More specifically, the particular potential lies in the action for damages against the Union or its agencies—as codified in Article 340(2) TFEU respectively Article 97(4), 98 Frontex Regulation, and Article 45(3) EASO Regulation. In the latter case, the agency would be liable under its founding Regulation in a first degree, and the Union, since it cannot exclude its liability under Article 340(2) TFEU by adopting secondary law, in a second degree.

To begin with, it seems that, among the approaches addressing systemic deficiencies by enforcing EU law, one can distinguish between top-down procedures, initiated by the Commission as guardian of the treaties, and bottom-up procedures, initiated by individuals. Both the preliminary reference procedure, as the standard mechanism in the internal market, as well as procedures in which individuals claim their rights directly before the CJEU, as standard mechanism in competition or state aid law, form part of the latter.

In the case of EU hotspots, any procedure depending on the Commission’s initiative seems unsuitable to enforce EU law due to the Commission’s involvement in the EU hotspot administration. The preliminary reference procedure is moreover of little use already because an action for damages against the Union cannot be brought before national courts. What remains are the procedures granting the individual direct access to the CJEU.

The action for damages is the most suitable procedure in this context. Notably, it could grant the right to an effective remedy, enshrined in Article 47 ChFR, in a particularly challenging context. The increasingly integrated European administration more generally raises challenges as to how to guarantee the right to an effective remedy. In the case of EU hotspots, the challenge arises, inter alia, because the relevant administrative conduct is of non-formally binding nature and consists in omissions to comply with supervisory obligations. While the action for annulment does not provide a remedy in those cases, the action for damages does. This is indeed the reason why the action of damages has become the main action ensuring the right to an effective remedy—as examined in particular by Timo Rademacher, and as analysed with regard to Frontex in particular by Melanie Fink. Finally, EU public liability law has an attribution function: an action for damages against the Union would not exclude liability of the host Member State or the other Member States under the case law following Francovich. Quite to the contrary, EU public liability law allows to assess each contribution, and the Union and the Member States can be held jointly liable.

Against this background, one might wonder: If the situation in the EU hotspots is really so bad, and if EU public liability law really has such potential, why did nobody file an action for damages against the Union yet? To be sure, the CJEU’s dismissal of the action for annulment against the EU-Turkey Statement, which was in essence directed against the implementation of the return policy in the EU Hotspots, does not preclude an action for damages against the Union based on the systemically deficient EU hotspot administration: The CJEU’s finding, namely that the Union did not conclude the EU Turkey Statement, is not relevant to the question of whether the Union is liable due to its administrative involvement in the EU hotspot administration. Rather, practical obstacles such as insufficient capacity of legal aid may provide the reasons: The few lawyers working under extreme pressure in the EU hotspots might come to the conclusion that it is simply not feasible to invest a considerable amount of time and resources in a procedure with uncertain outcome.

4 – The critical question of who is responsible—holding the Union liable?

Now, assumed that a person succeeded in filing an action against the Union before the CJEU, and that he or she claimed damages invoking the dire living conditions in the EU hotspot or his or her deportation to Turkey: Would the Union indeed be held liable?

Finding an answer to this question requires a close analysis of the extensive case law on EU public liability law. According to the CJEU’s jurisprudence, non-contractual liability under Article 340(2) TFEU arises if unlawful conduct of a Union body, qualifying as a sufficiently serious breach of a rule conferring rights on individuals, has caused a damage. Liability under Articles 97(4), 98 Frontex Regulation and respectively Article 45(3) EASO Regulation arises under the same conditions. Given the scope of this post, the argument here is limited to considering on the basis of which administrative conduct liability might arise, and shortly outlining two crucial legal issues.

Frontex could incur liability based on its registration of applicants in the EU hotspots and based on its escorting of deportations to Turkey. The former contributes, at least insofar as vulnerable persons are concerned, to keeping applicants in conditions incompatible with Article 4 ChFR, and the latter, at least in most cases, to a violation of the non-refoulement principle as enshrined in Articles 4, 18, 19(2) ChFR. Both could be in breach of Frontex’s obligation to respect fundamental rights under Articles 1, 36(2), 44(3), 48 Frontex Regulation, Article 51 ChFR. Further, the conclusion of the relevant Operating Plan, or the omission to withdraw from the administrative cooperation despite knowledge about systemic fundamental rights violations could be in breach of Articles 1, 36(2), 46(4) Frontex Regulation, Article 51 ChFR. (On supervisory obligations conferring rights upon individuals see the CJEU’s case law, notably Ledra.) In the same vein, EASO could incur liability based on its conducting of asylum interviews, drafting legal opinions, and adopting the relevant Operating Plan and the Standard Operating Procedures, which could be in breach of EASO’s obligations to respect fundamental rights. Finally, the Commission could incur liability based on its failure to adequately exercise its supervisory obligations. The failure to ensure the implementation of the EU hotspot approach in compliance with EU law could amount to a breach of Article 40(3) Frontex Regulation, Article 17(1) TEU, Article 51 ChFR. (On administrative omission see the CJEU’s case law, Kampffmeyer, and more recently Ledra, which confirms that the Commission’s omission to effectively ensure that Member States act in compliance with EU law may incur liability.)

To be sure, several legal issues would need to be resolved. Notably, the question arises to which entity administrative conduct of staff seconded to the EU agencies must be attributed. To give an example, the question is whether the conduct of a German officer seconded to Frontex and deployed to Greece as part of an MMST is to be considered as an act of Germany, of Greece, or of Frontex. Existing doctrinal analysis mainly suggests attribution to the host Member State due to the internal decision-making structure. However, one could also argue that the external appearance of the conduct towards a reasonable addressee must be taken into account in addition—which means that the appearance of the seconded staff’s conduct as conduct of the agency speaks in favour of attribution to the latter. The CJEU’s decision in A.G.M.-COS.MET as well as the right to a remedy, which cannot be effectively exercised if the individual is required to analyse the agency’s internal decision-making structure in order to know against whom to file an action, suggest such a reading.

Another legal issue arises in the context of causation, namely: whether non-formally binding administrative conduct may incur liability. The question is whether the ‘sufficiently direct link’ required for causation is ‘broken’ by the administrative decision of the host Member State. In contrast to its earlier jurisprudence, the CJEU in KYDEP and similar cases acknowledged that even a telefax by the Commission may, in principle, incur liability of the Union. It remains to be discussed whether later case law again overturned the KYDEP doctrine. Another approach, proposed by Melanie Fink, is to transfer the differentiation between primary and attributed responsibility, based on the Draft Articles on State Responsibility, into EU public liability law. A further discussion of those issues would go far beyond the scope of this contribution.

Whether the Union actually is liable for fundamental rights violations in EU hotspots hence remains to be answered. In other words, the potential of EU public liability law in the context of EU hotspots remains to be unfolded. And this, to begin with, requires a closer doctrinal analysis of the CJEU’s case law.

5 – EU public liability law as a limit to externalisation policies

Current EU migration and asylum policy relies, not fully, but to an important extent, on externalising the challenge of dealing with enhanced forced migration towards Europe. The challenge is often either put on third countries, or, where this is not possible, on Member States located at the EU external border. This approach leads to large scale fundamental rights violations—despite the difficulties of ECHR and EU law to address situations characterised by extraterritorialisation and outsourcing.

EU hotspots can be described as a paradigm example in this regard. As externalisation has an ‘out of sight, out of mind’ effect, it seems possible to forget about daily fundamental rights violations at the EU’s external borders. EU constitutional law however calls into question whether mere externalisation to Member States located at the EU external border is really sufficient to wash the Union’s and the other Member States’ hands of responsibility. This would indeed be quite strange, not only in light of the noble values enshrined in Article 2 TEU, but also given that the European Asylum System is conceived as a Common one.