Shielding Frontex
On the EU General Court’s “WS and others v Frontex”
Frontex has been under scrutiny for its alleged contributions to human rights harms for quite some time. In my (very) early days as a young scholar, it seemed that every conference, workshop or political event I attended on the topic of the EU’s border management was graced with Frontex representatives toeing the proverbial company line, echoing time and time again that Frontex takes on a mere “coordinating and supporting role”. This narrative became so repetitive that it now serves as a catalyst for cynical snickering across the room during such gatherings. This wouldn’t be problematic were it not the standard response to the increasingly louder, and more prevalent human rights critiques directed at the Agency, whereby it shields itself from any meaningful responsibility.
Dismayingly, in WS and others v Frontex, we (again) never truly get to the point where Frontex conduct is effectively and substantively tested against the human rights parameters set forth in both the Charter of Fundamental Rights (notably, Art. 4, 18 and 19 CFR), as well as in relevant EU secondary legislation (e.g., the 2016 Frontex Regulation). Instead, ‘The General Court (Sixth Chamber) hereby: 1. Dismisses the action’.
While a number of complaints have been lodged against Frontex before the EU’s courts in different procedures (e.g., annulment, transparency, failure to act), the action for damages in casu, was the first of its kind concerning human rights responsibility of Frontex and had all the ingredients to prompt the General Court to finally clarify a number of pervasive and urgent questions concerning Frontex responsibility for complicity in unlawful human rights conduct. Instead, by conflating the wrongful conduct under scrutiny, the Court prevents a critical examination of Frontex’s conduct altogether. The significance of the case thus lies in the adopted approach by the Court, which, in effect, contributes to the systematic shielding of Frontex from any responsibility for contributions to human rights harms.
Facts and Ruling
The case concerns Syrian nationals, who arrived on Milos, Greece, in 2016, and after being transferred to the Greek island of Leros declared their wish to apply for international protection. However, 6 days following said declaration, the Applicants were returned to Türkiye, following a joint return operation carried out by Frontex and Greece. With this case, the Applicants – all of whom are Syrian nationals in pursuit of international protection in the EU – sought compensation for the damage flowing from Frontex’s failure to comply with its human rights obligations under the 2016 Frontex Regulation, its Standard Operating Procedure, and its Code of Conduct.
The Applicants claim that through its supporting role in the execution of the joint return operation, Frontex violated the non-refoulement principle (Art. 19 CFR), the right to asylum (Art. 18 CFR), the prohibition of collective expulsion (Art. 19 CFR), the rights of the child (Art. 24 CFR), the prohibition of degrading treatment (Art. 4 CFR), the right to good administration (Art. 41 CFR) and the right to an effective remedy (Art. 47 CFR).
Dealing first with a number of admissibility objections, the General Court starts its analysis by recalling that for non-contractual liability to arise under EU law, a number of cumulative conditions must be met. There must be actual damage, stemming from unlawful conduct, and there must be a causal link between the alleged conduct and the damage (para. 52 – 53). Focusing first on causality, the Court then proceeds to recall that the damage ‘must be a sufficiently direct consequence of the conduct…which must be the determining cause of the damage’ (para. 56).
In what follows, the Court reasons that the Applicants wrongly base their claims on the assumption that if Frontex had complied with its fundamental rights obligations, they would have been accorded international protection in the European Union rather than having been expelled from its territory. Immediately thereafter, the Court contends that while Frontex is bound by fundamental rights obligations in the execution of its tasks, its tasks within the context of return operations are limited to providing technical and operational support and do not permit interference with the decision to grant international protection, or a return decision – both which lie in the purview of Member State competence. Mindful thereof, the Court concludes that as Frontex has no competence concerning the merits of such underlying decisions, the causal link between the damage and the conduct (namely, the decision not to grant international protection and the return decision), cannot be established and the action is dismissed, relieving Frontex from any responsibility. To fortify its findings, the Court emphasizes that the causality requirement entails that the damage must flow directly from the alleged illegality and may not be the product of the Applicant’s reaction to the illegality.
This case is bound to spark much debate, some of which has already surfaced on media platforms, ranging from the composition of the bench, the scope of the dispute and its importance in the context of proceedings against Frontex more generally. However, in what follows, I want to step away from the obstinate fundamental rights concerns flowing from Frontex’s conduct. Instead, I want to draw attention to the manner in which the Court chose to approach the question of fundamental rights responsibility and how its reasoning has accentuated the urgency for clarity on joint responsibility, the distinction between attribution and causation, and the reconcilability of the causality requirement with the right to an effective remedy.
Analysis
In paragraph 62 of the judgment, the Court points out that the Applicants incorrectly conclude that if Frontex had complied with its fundamental rights obligations (the ‘but-for test’), they would have been accorded international protection and not have been subject to return proceedings. The Court is – of course – right in this observation, as Frontex does not have the competence to rule on the merits of a request for international protection, or on the merits of a return decision.
Leaving aside, however, whether this was effectively what the Applicants argued, the manner in which the Court refutes the argument is interesting because it suggests a conflation of what the purported wrongful conduct is in the present case. The Court appears to suggest that the Applicants take issue with the underlying decision to refuse international protection and the subsequent return decision and somehow connect these decisions to Frontex. Yet – without being privy to the Applicant’s arguments in detail – it seems more likely that the Applicants object to the complicity of Frontex in the execution of these decisions – which is an altogether different question. In other words, there are three questionable but separable elements in the present case: the refusal of international protection (1), the return decision (2) and the execution of the return decision (3). The current phrasing suggests that the Applicants are contesting Frontex’s role in the former two, whereas it is far more likely that the Applicants are, in fact, contesting Frontex’s role in the third, namely, its role in the execution of those decisions (para. 6 and 15).
The (willful?) conflation of the alleged wrongful conduct, highlights a number of unresolved issues.
Joint Responsibility
Joint responsibility points to a legal construct whereby two or more actors share varying degrees of legal responsibility for their respective (identical or different) contributions to a single (human rights) harm. Whereas joint responsibility has been theorized to a certain extent under international law, this is less so under EU law. Yet, the question of joint responsibility is precisely what is at stake when Frontex assists – even in a mere supporting role – the Member States in their implementation of the EU’s policy on borders, asylum and migration.
Firstly, by conflating the alleged unlawful conduct and suggesting that Frontex cannot – by lack of competence – take decisions on merits concerning international protection and return (and not discussing the actual implementation of the decisions), the Court sidesteps the question of joint responsibility entirely. It does not clarify whether Frontex may incur responsibility for its contribution to the purported illegality through its role in the execution of the return order, nor what contributory or shared responsibility under the EU could and should look like. In other words, the purported illegality objected to by the Applicants would then not be the underlying decisions, but instead the contribution by Frontex to the execution of the return operation.
Secondly, joint responsibility may also arise when two different lines of conduct give rise to a single harm. The Court in casu appears to consider only the potentially objectionable decisions. Conversely, it does not consider that the execution of those decisions may prompt (the same) human rights violations, albeit on different grounds. Whether Greece complied with its negative and positive obligations stemming from Art. 18 – 19 CFR in implementing the Qualification Directive and Return Directive, has little to no bearing on whether Frontex complied with its own fundamental rights obligations (to not contribute to human rights violations) under the 2016 Frontex Regulation in the execution of the return operation. The latter question does not require an assessment of the merits of a protection status decision or return decision by Frontex to assure itself (positive obligation) that non-refoulement, collective expulsion, the rights of the child, good administration and the right to an effective remedy are respected.
In this case, the Court had an opportunity to shed light on the contours of EU joint responsibility. Instead, it somewhat awkwardly focuses exclusively on the underlying decisions giving rise to the expulsion, in which Frontex has no competence. This notwithstanding the fact that the Applicants contested the role of Frontex in the return operation (para. 6, 15), which is different from contesting the underlying decision giving rise to the return operation.
Attribution and Causation
In some cases, attribution has surfaced as an additional (cumulative) condition for an action for damages to give rise to compensation. In other cases, it is not mentioned, and in yet another subset of cases, attribution is conflated with causation.
Attribution and causation operate at different ends of the responsibility question. While attribution connects an actor to a particular line of (unlawful) conduct, causation links this conduct to the damage. Visualized this way, these concepts operate sequentially as two different segments of the responsibility question. Yet in the present case, the Court concludes that no direct causal link can be established between the actor (Frontex) and the damage [as Frontex has no competence in deciding on the merits of a protection status and return decision]. In doing this, the Court appears to sidestep the question of attribution, by not first separating the problematic conduct and connecting this conduct to the proper author according to their respective competencies. Instead, it jumps directly into the question of causation. Multiple tests of attribution could have been relied upon by the Court and could have generated different outcomes in the ultimate decision on responsibility. For example, the ‘effective control’ test on attribution could result in the attribution of the conduct during the return operation to Frontex, and attribution of the underlying decisions to the Member States, whereas the ‘ultimate normative control’ test on attribution could entail that instead all conduct in the sequence of events would be attributable to the implicated Member State. Different still, the ‘competence test’ on attribution connects conduct to the actors based on their Union-based competence.
Had the Court clarified its stance on attribution (by adopting the competence test on attribution, for example), it is likely that the underlying decisions would be attributed to the Member States as the rightful author, whereas the execution of the return order (at least in part) would be attributed to Frontex. This would subsequently generate two questions of responsibility: fundamental rights responsibility for the decisions of the expelling Member State and fundamental rights responsibility for the role in the return operation of Frontex. This seems to be in line with what the Applicants intended and explains why a separate procedure was lodged against Greece before the ECHR (para. 6).
Furthermore, not engaging with attribution and instead conflating the alleged wrongful conduct, implies that the next step of the responsibility question – the test of causality – cannot be met in this case. How can the causal link between the damage and conduct be direct and determinative to establish Frontex’s responsibility if the conduct that the Court scrutinizes is the conduct of which Frontex is not actually the author?
By leaving the question of attribution unresolved, the Applicants are left guessing which test of attribution could and should apply, making it hard to definitely discern who did what in violation of what rule of EU law, let alone convincingly argue on what grounds Frontex may be held responsible for its complicity in the human rights harm.
This very same question of attribution and the subsequent question of causality drives at the very heart of the joint and several liability cases concerning Kočner and Europol currently pending appeal before the ECJ. It underscores the urgency for a legally certain and systematic method to and distinction between attribution and causation, and clarification as to the applicable test of attribution in multi-actor operational settings.
Final Remarks
Focusing exclusively on the alleged illegality of the decisions rather than Frontex’s role in the return operation, the Court seems to cherry-pick the arguments brought forward by the Applicants and leaves open burning questions on the contours of EU joint responsibility, the burden, method, and standard of proof of joint responsibility, the rule of attribution, and the applicable test of causation.
Consequently, the Court rids itself (and Frontex by extension) of the hot-responsibility-potato, by recalling that Frontex only has a supportive and coordinating role, thereby designating the Member State as the sole author of the potentially unlawful conduct. This has two important implications:
- Member States may still incur responsibility for unlawful human rights conduct in these scenarios. However, it is plausible (if not likely) that domestic courts, as well as the ECtHR, may very well apply the ‘effective control’ test of attribution, alleviating the Member State of responsibility, and de facto throwing the hot potato back in the direction of Frontex, as the body that executed the return operation and thus exerted physical control over the operation. This blame-shifting works to the detriment of the individual Applicants and their right to an effective remedy, who are caught in a limbo of complicit actors shunning their responsibility through misaligned responsibility rules adopted by different courts.
- By emphasizing that Frontex acts in a merely supporting role, the Court implicitly suggests that Frontex conduct will almost always be at the behest of or intimately tied to Member State conduct. Indeed, in situations of joint operational action, it is unlikely that there will ever be a scenario where Frontex operates independently from the Member States. What is then the purpose of the action for damages if the causality requirement is so strict that it cannot be met, and operational and supportive conduct by Frontex cannot be subject to judicial scrutiny if there are underlying Member State decisions giving rise to the operational and supportive Frontex conduct?
With the Kočner v Europol appeal, the pending Hamoudi v Frontex case, and hopefully an appeal in this case, the CJEU has ample opportunity to take a stance on these topics, where the General Court neglected to do so.
Sincere thanks to Merijn Chamon and Thomas Streinz for comments on an earlier draft. All errors or omissions remain mine.
The distinction between Frontex’s authority and corresponding liability is not dissimilar from that of the police in same situations. I doubt the police is ever deemed liable for executing orders made by courts or executive bodies – they may be liable for how they executed these decision through not following the proper procedures in executions and possibly infringing on any human rights in the execution of the decisions but they are arguably never liable for having executed a decision in and of itself especially if they are decisions over which they have no right to question or veto or refuse to participate in. Extending liability to Frontex for having partaken in the decision through its execution would also arguably entail that Frontex has discretion in executing or not.
Why guess what the applicants pleaded when it is summarized in points 57-59? In sum they claimed:
“…that Frontex’s unlawful conduct before, during and after the return operation caused them actual and certain damage of a material and non-material nature. More specifically, they argue that, if Frontex had not infringed its obligations relating to the protection of fundamental rights in the context of joint operations, in particular the principle of non-refoulement, the right to asylum, the prohibition of collective expulsion, the rights of the child, the prohibition of degrading treatment, the right to good administration and to an effective remedy, they would not have been unlawfully returned to Türkiye and they would have obtained the international protection to which they were entitled, given their Syrian nationality and the situation in Syria at the material time.”
Furthermore, they also listed the specific damage caused to them allegedly by Frontex:
“Thus, they would not have had to suffer material damage consisting of (i) the amount spent to travel to Greece; (ii) the cost of renting a house in Saruj and the cost of purchasing furniture; (iii) the expenses incurred in fleeing to Iraq; (iv) the rent paid in Iraq; (v) the electricity costs for their home in Iraq; (vi) the children’s school fees in Iraq; (vii) the subsistence costs in Iraq; and (viii) the cost of legal aid for assistance with their complaints against Frontex. Similarly, they would not have had to suffer non-material damage consisting of (i) feelings of anguish, particularly on the part of the children, caused by the return flight to Türkiye, on account of their separation during that flight, their being prohibited from speaking and the presence of uniformed escort officers and police officers and (ii) a feeling of fear and suffering linked to an extremely difficult and dangerous journey to Iraq in the snow-covered mountains because of the fear of being returned to Syria by the Turkish authorities.”
Even with the most generous reading, it is hard to discern what human rights violation is Frontex responsible for that directly led to e.g. spending money to rent a house in Greece. GC couldn’t have come to a different conclusion unless they themselves rewrote the submission to make it make sense. Unfortunately, quality of legal representation matters and I personally cannot see what alchemy could the CJEU perform to make the appeal successful.
I also do not understand the criticism of the formation as a chamber of 3 is a default GC chamber and this case is not complicated. I’m even more perplexed by comparisons to Kocner as Kocner concerns very specific Europol actions, namely Europol’s involvement in criminal investigation and active collaboration with MS authorities… all in all, very confusing.
Thank you for the comments and the opportunity to clarify.
The careful reference to the Applicant pleadings, is in response to the reasoning adopted by the Court, which suggests that the Applicants argued that the violations of fundamental rights lies with the rejection of the international protection claim, and in the decision to return the individuals. I fully agree that it is quite clear from the the damage references (that you cite) and the application made the Applicants in the case, that they are contesting the execution of the return decision, as opposed to the return decision and rejection of international protection itself. All the more why I was perplexed that the Court chose to refute causality by reference to the latter, when the Applicants are claiming that it is in fact the former that violated their fundamental rights.
As to the alchemy upon appeal remark – it would be interesting to hear the Court on the actual question, namely whether the execution of the return decision by Frontex violated any fundamental rights. This brings me to your point on uncertainty as to the actual fundamental rights violations. Under the non-refoulement principle, there are a number of positive (procedural and substantive) obligations that must be met (by States) to be considered compliant with this fundamental rights commitment. Yet, to date, there is no clarity whether these concrete positive obligations as they apply to States apply to Frontex in the same manner as they do to the Member States. The EBCG Regulation in its various amended forms, hints that it does, but the CJEU has yet to take a stance on this. Concretely, the entire operation including the relevant decisions and the execution of the return decision happened within a mere matter of days, which suggests that the procedural and substantive positive obligations under the non-refoulement obligation (to investigate the country of return, appeals to the rejection decision etc) may not have been respected. Regardless of whether they were or were not respected however, the Court did not delve into this, by ignoring this question altogether – which you point out was quite clearly, the actual question in the application.
I think we have to agree to disagree on the point of the Chamber formation – having followed the developments of judicial review quite closely, I would have assumed that the first-ever action for damages concerning Frontex was sufficiently complex to warrant a different formation, not in the least because it concerns complicated questions of joint responsibility under EU law that have not yet been definitively resolved. This was recently reaffirmed in a large-scale conference bringing together scholars in different fields working on shared responsibility relating to Frontex (https://idpbarcelona.net/wp-content/uploads/2023/09/SHARED_Programa.pdf)
Finally, the reference to Kočner v Europol: this case strikes me as a relevant case, precisely because the Europol Regulation takes joint responsibility one step further, and explicitly introduces joint and several liability to deal with circumstances involving EU and MS involvement in operational action. Of course, the content/agency is entirely different and the provision on joint and several liability is limited in material scope to data processing. But when looking closer at the case, it is clear that the provisions on joint and several liability seek to address problematic cases of attribution in the interest of an effective remedy for individuals, which is exactly what we are frequently confronted with in Frontex cases. While an automatic transplant from Europol to Frontex seems unlikely, whatever the ECJ decides in this case, may serve as a source of inspiration for the development of more advanced regimes of joint responsibility under EU law as they apply to Frontex as well.
Again, happy to continue the conversation and feel free to reach out in case of disagreement and concerns.