With Great Power Comes No Responsibility: On Frontex’s (Lack of) Accountability Following WS

In 2019, amid allegations concerning its border management operations, the European Border and Coast Guard Agency (Frontex) ‘categorically denie[d] any involvement’ in fundamental rights violations. Turns out, those allegations were true. Or rather, as found by the European Anti-Fraud Office (OLAF), there had been cases of serious misconduct by the agency’s officials, who actively implemented measures to prevent any observation or investigation of possible violations of the fundamental rights of migrants.

To jog our memory, Frontex is the EU Agency that supports the Member States in the management of the EU’s external borders. With a staff that will grow to 10000 standing corps by 2027, Frontex provides crucial assistance in managing migration flows. This, by its very nature, involves a series of fundamental rights considerations: not least, the risk of forcing migrants to return to an unsafe country. In such a context, where, as seen, internal accountability mechanisms seem disregarded, the focus turns to courts, whose task is precisely to ensure the protection of fundamental rights and remedies for the victims of violations.

In this respect, last September, the General Court (GC) of the Court of Justice of the European Union (CJEU) ruled on the very first action for damages brought against Frontex in relation with a return operation carried out in conjunction with the Greek border authority: WS and Others v Frontex. Yet it dismissed the claim since it could not find a causal link between the agency’s conduct and the damage suffered by the applicants. In light of Frontex’s ever-growing role in border control, the ruling arguably has negative implications on the EU’s accountability for fundamental rights violations: indeed, as we will see, overlooking Frontex’s real responsibility, WS contributes to reinforcing a liability vacuum that is ill-suited to the alleged ‘comprehensive system of legal remedies’ often invoked by the CJEU.

The WS case

The case originated from a group of Syrian nationals who, a few days after arriving in Greece and expressing their will to apply for international protection, were returned to Turkey in a joint operation by Frontex and Greece. Once there, fearing a forced return to Syria, the applicants fled to Iraq, where they filed a complaint before Frontex’s Fundamental Rights Officer (FRO), i.e., the official in charge of monitoring and promoting the agency’s respect for fundamental rights and responsible for the internal fundamental rights complaint mechanism. This was followed by a second complaint for not obtaining any answer to the first one.

Strikingly, the FRO’s response contained no information on the results of the investigation carried out by Greece since the relevant police report was classified as confidential; moreover, albeit without addressing the agency’s role in the return operation nor the subject of the second complaint, it found that ‘Frontex had complied with its obligations’.

The applicants then filed an action for damages before the CJEU (Articles 268 and 340 TFEU), claiming that Frontex had violated several of their rights enshrined by the Charter of Fundamental Rights of the EU and mentioned in Frontex’s Regulation and Code of Conduct. As a result, they would have incurred various damages, including travel expenses to Greece, living expenses in Iraq, as well as non-material harm due to their treatment during the deportation process (in which, for instance, children were separated from their families) and the conditions after arriving in Turkey.

What the Court has done

In its judgement, the CJ first declared the action admissible, rejecting Frontex’s argument whereby the applicants were actually seeking to obtain an identical result to that of an action of annulment (Article 263 TFEU) against the FRO’s letter closing the complaint mechanism. Instead, the Court considered that, while the action under Article 263 TFEU, if successful, would only have required a fresh examination of the applicants’ complaint, the action for damages sought something more: namely, relief from damages occurred not only as a consequence of the FRO letter but also of Frontex’s unlawful conduct ‘before, during and after the return operation’.

In this regard, the GC has been praised for clarifying that, in fact, the action for damages could represent an effective remedy against Frontex, which typically operates through material action rather than (binding) acts, thus making it extremely difficult for an action for annulment to remedy violations.

In the merits, however, the Court’s reasoning turned against the applicants. Indeed, after recalling the three (cumulative) conditions for non-contractual liability to incur – unlawful conduct by the EU, the occurrence of actual damage, and the causal link between the two – the GC went directly to the analysis of the causal link between Frontex’s conduct and the alleged damages. Then, the Court quickly found that none of such damages was the ‘direct consequence’ of Frontex’s conduct: neither (and rightly so) the smugglers’ fees from Syria to Greece, nor, most crucially, all those incurred during and after the return operation. This was mainly because, whereas the applicants argued that had it not been for Frontex, they would have received international protection by Greece and hence would not have been harmed, the agency’s task was only, under the letter of its own Regulation, ‘technical and operational’: formally, it has no role in taking the return decision, which instead falls within the competence of the member state.

Therefore, the Court dismissed the action in its entirety. In so doing, however, it blatantly underestimated the role of Frontex both in general and in the case at stake.

What the Court has not done

The Court’s approach could be criticised at least for three reasons. The first is the choice, probably decisive for the formulation of its reasoning, to start the analysis with the assessment of causation rather than the illegality of Frontex’s behaviour. While consistent with its jurisprudence, this approach was not required by law (and indeed, in various situations, even when a particular requirement was not met, the Court still proceeded with the assessment of the others). Still, it allowed the GC to refrain from distinguishing the precise conduct(s) which led to the damages and hence the agency’s contribution to each of them. Instead, perhaps relying on an (unfortunate) argument by the applicants (para. 57–58), it assumed that the unlawful conduct underlying all the damages was the return decision per se.

In other sections of the ruling, however, it is clear that the applicants were seeking relief from the agency’s behaviour ‘before, during and after’ the return operation (para. 6, 15, and 28). Even accepting that Frontex had no responsibility in the return decision, that leaves out potential damages occurring as a result of the operation itself: namely, the non-material harm caused by the separation of children from their families. Such damages could not have been caused by the return decision, but instead by the management of the return flights, over which Frontex does have an extensive responsibility (see Art. 28 of Frontex Regulation).

Second, in general, the GC largely disregarded the possibility of shared responsibility for infractions in joint operations between member states and EU agencies. Indeed, while this kind of joint liability is underdeveloped in the EU, it is a crucial concept in international law. Had the Court applied, for instance, an ‘effective control’ like the one commonly used to assess joint responsibility of international organizations, it would probably have concluded that Frontex had a greater role in the execution of the return order than mere technical support.

And indeed, joint liability could prove crucial in granting relief to victims of fundamental rights violations in joint operations where attribution cannot be definitively established, as the CJEU has recently held in Kočner v Europol. Furthermore, in that case, the Advocate General also recalled that joint and several liability are not unknown to European legal systems (para. 50), which is relevant since the EU’s non-contractual liability takes inspiration from the principles common to the member states (Art. 340(2) TFEU). Yet in WS, rather than considering whether an ‘outdated system of remedies’ had to be updated, the GC avoided the issue altogether treating Frontex’s causal influence to the violation as a remote one.

Last, the Court’s reasoning did not consider a significant factor that could have impacted the analysis of the causal link: Frontex has specific obligations to respect fundamental rights itself and to oversee the member state it is assisting. This flows, firstly, from the provision in its Regulation which requires both member states and the agency to comply with fundamental rights during the entire return operation (Art. 50(3)). Moreover, the agency has the duty to actively supervise the respect of fundamental rights in each return operation (Art. 50(6)) and, crucially, to ensure that individuals are not returned to a country where there is a risk they might be expelled against the principle of non-refoulement (Art. 80). Briefly put, Frontex has to make sure that the operation as a whole, with its implementation, complies with fundamental rights.

In this respect, it is worth mentioning that a return decision is not carved in stone: if concerns arise that a destination country may not be safe, for instance following allegations of non-compliance with the non-refoulement principle (like here, here, and here), Frontex would have the duty to either suspend the operation or, at the very least, notify its Fundamental Rights Officer (FRO) and the cooperating member state (Art. 25(4) and 28(6)–(7)).

Conversely, the failure to do so should have been at least mentioned by the Court, since the agency’s responsibility can also be engaged through an omission against a legal obligation to act, like in the case just shown. And indeed, this kind of inaction would reach the threshold required for a conduct to be unlawful, as it is a sufficiently serious breach of specific supervisory obligations conferring rights on individuals. Finally, if willing enough, the Court could have established a causal link between the violation of supervisory obligations and the damages. This is because the supervisory duties of Union bodies are specifically designed to prevent illegal actions by Member States: such responsibilities would be rendered meaningless if they were assigned no role in the causation assessment.

The GC’s reluctance to acknowledge these considerations is worrisome not only in itself, but also as it reinforces the liability vacuum surrounding Frontex.

Frontex: Great Power, No Responsibility?

If assessed in context, WS represents but one example of the challenges in holding Frontex legally accountable for violations of fundamental rights during its operations. These concerns are amplified by the fact that the agency has been recently granted significantly more authority in response to the migration ‘crises’. And this not only de iure, through the conferral of proper executive powers, but also de facto: indeed, even though asylum decisions remain under member states’ competence, Frontex’s role, including information provision, can significantly influence such decisions. On top of that, since political discourse appears to lean more towards effective migration control rather than finding a balance with the protection of fundamental rights, it becomes crucial to scrutinize how such powers will be used.

At the same time, though, the 2019 Regulation has not significantly improved Frontex’s monitoring system, which currently rests on the Consultative Forum on Fundamental Rights and the FRO. On the one hand, the role of the first is only general, its reports are not binding, and the agency’s follow-up to them is largely unsatisfactory. On the other hand, only a few cases are deemed admissible to the complaint mechanism before the FRO, and, as testified by WS (para. 9, 12, and 16), this mechanism has significant shortcomings. Most importantly, it is not independent since the FRO is appointed by the agency’s management board, the same entity it is supposed to scrutinize.

Finally, concerning internal reporting of violations, despite the obligation to report serious incidents concerning human rights there remains no sanction for officials who fail to do so. Consequently, in practice, little prevents the same situation found by the OLAF investigation from happening.

Against this backdrop, judicial review at the EU level fails to perform its task of ensuring accountability. First, as seen, this is because the Court refused to acknowledge joint liability together with its over-formalistic reading of Frontex’s role. In this respect, while it is true that the Court has recently granted damages in relation to unlawful conduct held together by Slovakia and Europol, the CJEU found that Europol’s joint and several liability stemmed specifically from the text of the agency’s Regulation, by way of exception to the Union’s general liability regime. Conversely, the same is not true for Frontex’s legal framework, which deliberately remains vague on the division of responsibilities and hence facilitates ‘blame-shifting’.

Second, even if the Court in WS had come to a different conclusion, it would still not have been easy for the applicants to obtain relief: indeed, in a similar case (Hamoudi and Others v Frontex), the GC held that the applicant, who had been subject to a push-back, did not adequately demonstrate the damages, essentially imposing an ‘impossible burden of proof’ on individuals who usually migrate under degrading conditions. Moreover, in cases of joint responsibility, the CJEU typically requires that an applicant first pursue remedies at the national level before it addresses the Union’s liability.

Not that the other judicial avenues before the CJEU are more effective. Indeed, the action for annulment can only be brought against formal (binding) acts, thus excluding most of Frontex’s operational activities from review. Lastly, in actions for failure to act, the GC seems to be content with a generic answer from Frontex to a request to terminate its activities in the Aegean Sea potentially violating fundamental rights, thereby fulfilling its obligation to define its position.

To conclude, this leaves Frontex without a proper mechanism of accountability other than political ones, and victims without proper remedy under EU law and forces them to seek relief before national courts. Leaving aside the challenges posed by the lack of transparency of member states’ operations, this way could indeed prove effective also thanks to the possibility of ultimately ending up at the European Court of Human Rights, which is particularly open towards the difficult situation of migrants. Still, the fact remains Frontex’s involvement in fundamental rights violations would remain per se unscrutinised since the EU is (still) not part of the Council of Europe.

...What Now?

As seen, the recent outcome of WS, together with the difficulties of pursuing other legal avenues against Frontex for fundamental rights violations, the high standards of proof required by the Court in Hamoudi, and the limited responsiveness of Frontex’s own fundamental rights mechanisms, brings us to conclude that despite the Frontex’s power – sensibly increased after the 2019 Reform – with such power comes little, if any, responsibility.

In this respect, it is to be hoped that the CJEU will develop clearer attribution rules giving justice to violations arising from the interaction between the EU and its member states. This hope now lies on the Court of Justice, which has been called upon to review the judgement of the General Court. Otherwise, the long-awaited first action for damages against Frontex will be just one in a series of failed attempts to hold the agency accountable.

Put differently, the Court in WS may well have ‘paved the way for further proceedings of its kind’. Yet, without a change either in the legislation or the approach, this newly paved path might remain largely untraveled.

Francesco Mauri

(bibliographical references can be found in the links throughout the text)

---

The contents of the article represent solely the ideas and opinions of the author and in no way the opinions of Bocconi University or the IUS@B association.