Shaping the Joint Liability Landscape? The Broader Consequences of WS v Frontex for EU Law

Abstract: The Insight delves into the CJEU judgment of WS et al. v Frontex, the first action for damages against the European Border and Coast Guard Agency, Frontex, for human rights violations at the EU's external borders. Despite the prevalence of systemic violations and heightened attention to the agency's accountability, the Court, applying a stringent causality threshold, dismissed the claim, sidestepping crucial questions of positive obligations and responsibility attribution. The analysis critiques the judgment's shortcomings in causality assessment, emphasising its broader repercussions for EU law, particularly concerning liability frameworks and accountability dynamics within the new multi-actor reality of EU integrated administration. The Insight underscores the pressing need to reevaluate the existing competence model of determining liability in EU law to address its limitations and introduces the classification of these limitations as the binary of causality and the binary of jurisdiction. The CJEU’s reluctance to establish an effective framework for joint liability not only perpetuates contested accountability gaps but also risks establishing precarious areas devoid of accountability, thereby compromising the foundational principles of the Rule of Law in the European Union. The Insight concludes with a call to address these shortcomings, emphasising that rectification is not merely a matter of procedural refinement but a crucial step towards ensuring robust accountability mechanisms in EU law.

Keywords: Frontex – integrated administration – joint liability – accountability – rule of law – shared responsibility.

I.   Introduction

On September 6, 2023, the eyes of all those working on human rights violations at the EU's external borders were fixed upon the Court of Justice of the European Union (CJEU). The Luxembourg Court (General Court) was about to issue its much-anticipated judgment on the case of WS et al. v Frontex.[1] In an environment of systematic human rights violations unravelling at the EU borders (pushbacks, illegal returns, border violence, failure to rescue)[2] and the memories of the Adriana shipwreck in June of that year that claimed 650 lives off the coast of Greece, still hauntingly fresh,[3] WS v Frontex held the promise of real change in the atmosphere of impunity at the EU external borders.

In the last three years, several investigations have also revealed the involvement of Frontex, the European Border and Coast Guard (EBCG) Agency, in such violations and substantial gaps in the agency's accountability. This was an opportunity for the Court to cover some of these gaps. If successful, it would have been the first case to find Frontex's (and thus the EU's) responsibility for human rights violations. This could have had a trickle-down positive effect on human rights across all Frontex joint operations. This promise remained unfulfilled as the Court rejected the action for damages of the Syrian family who had been illegally deported to Türkiye by the Greek authorities in a Frontex flight.

The significance of this judgment extends beyond the Syrian family or Frontex itself, carrying implications for the broader EU liability framework. In particular, the Insight commences with an analysis of Frontex’s role in European border management and the scrutiny it has faced for human rights violation. It further proceeds to scrutinise the judgment from a liability angle, highlighting shortcomings in the causality assessment and the limitations of the available legal remedies before the CJEU. Following that, the analysis is expanded to encompass repercussions that extend beyond a specific EU agency to our understanding of joint liability in the wide-reaching environment of integrated administration in the EU. The analysis closes by examining how these repercussions influence accountability dynamics and the overarching principles of the rule of law in the EU as a whole.

II.  Frontex under scrutiny: allegations, investigations, and the rule of law dilemma

To appreciate the significance of this case, it is essential to understand the broader context of Frontex's role in European migration management. Since the abolition of the internal borders within the Schengen area, we have been witnessing a growing emphasis on the control of the external borders, which is argued to be the counterweight to free movement. While the establishment of safe and legal channels of entry remains limited and predominantly discretionary and ineffective, with resettlement and humanitarian visas as the primary examples, the development of “policies of non-entrée”,[4] or “ugly-duckling policies”[5] more broadly, has been rapid.

This shift towards external border control, driven by political developments of intolerance and nationalistic protectionism and labelled as a response to the “refugee crisis” since 2015, has led to the securitisation of border control. In this era of securitisation, Frontex has emerged as a securitising agent and a pivotal actor in integrated border management.[6]

In recent years, Frontex has faced scrutiny and allegations of human rights violations, particularly concerning pushbacks in Greece and Hungary.[7] Media reports in November 2020 implicated the agency in pushbacks at the Greek-Turkish border, triggering investigations by various bodies, including the Frontex Scrutiny Working Group of the European Parliament and the European Anti-fraud Office (OLAF).[8] These investigations revealed instances of concealed evidence and fundamental rights violations, culminating in the resignation of Frontex's Executive Director, Fabrice Leggeri. Still, despite the steps towards increased internal accountability taken by the agency’s Regulation (e.g. individual complaints mechanism, fundamental rights monitors) much is to be desired for the overall accountability of the agency.[9] The lack of answerability of the agency undermines respect for the rule of law to an extent that bears the question of whether Frontex, EU External Borders and the Rule of Law are three elements that are impossible to coexist.[10]

III. From Aleppo to the CJEU: WS et al. challenge Frontex in Court

The case of WS et al. v Frontex marks a significant milestone in the evolving accountability landscape at the EU external borders. The first action for damages against Frontex for human rights violations concerns a family of Syrian refugees who sought asylum in Greece but were instead refouled to Turkey in a Frontex-coordinated return flight. The claimants alleged material and immaterial damages resulting from Frontex's actions before, during, and after the return operation.

The applicants, who escaped Aleppo at the height of the war in Syria, arrived on the Greek island of Milos in October 2016 and formally expressed their interest in applying for international protection. However, their request was never registered, and six days later, the family was transferred to Türkiye in a joint return operation carried out by Frontex and Greece. From there, they subsequently moved to Iraq, where they currently reside.

The return of refugees without allowing them to apply for asylum or challenge the deportation order results in international and European law violations, including the right to asylum, the prohibition of non-refoulement, the Return Directive, and the EBCG Regulation. The applicants made use of the agency's individual complaints mechanism to lodge a complaint regarding their return to Türkiye. They also brought a case against Greece before the European Court of Human Rights, ultimately reaching a friendly settlement.[11] Concerning the internal complaint, the agency found that it did not concern the agency's actions but only the Greek authorities and, thus, forwarded it to the national authorities. Failing to address Frontex's role in the return operation, the applicants turned to the CJEU.

The applicants claimed that their damage followed the failure of the agency to comply with its legal obligations regarding: a) return operations, b) the operational plan, c) the role of the Frontex coordinating officer, d) the evaluation of the operation, e) the protection of fundamental rights, and f) the individual complaints mechanism (arts 16, 22, 26, 28, 34 and 72 EBCG Regulation 2016/1624, steps 1-5 Frontex Standard Operating Procedure, art. 4 Code of Conduct for Frontex return operations).[12] In particular, they argued that Frontex had acted in violation of its obligations under the principle of non-refoulement (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). Thereby, the applicants argued that Frontex assisted in carrying out the unlawful deportation and claimed that Frontex's unlawful conduct before, during, and after the return operation caused them material and non-material damage. This includes expenses incurred in their journey, such as travel costs, rent, and subsistence in Türkiye and subsequently in Iraq, as well as legal aid. They also cited non-material damage, such as anxiety, fear, and suffering during the return and subsequent journey.

IV. The findings of the Court: no direct causal link

The conditions for EU liability, as outlined in art. 340 TFEU, include a) unlawful conduct, b) actual damage, and c) a causal link between the alleged conduct and the damage. The Court in WS v Frontex emphasises the cumulative nature of these conditions. The Court has developed a three-part test for non-contractual liability: a) a rule must exist intended to confer rights upon individuals, b) which has suffered a sufficiently serious breach, and c) a causal link must exist between the damage and the unlawful conduct.[13]

The applicants argued that Frontex's failure to uphold fundamental rights led to their unlawful return, resulting in material and non-material harm. The Court, however, scrutinised the causal link between Frontex's conduct and the alleged damages. It asserted that the damage must be a “sufficiently direct consequence of the conduct”, with the conduct being the “determining cause of the damage”. The direct causal link must be proven by the applicants.

The Court found that the applicants' claims lack a “direct causal link”. It noted that certain expenses, like smugglers' fees to Greece, predate the return operation and cannot be directly attributed to Frontex. More importantly, the Court asserted that the applicant’s claims are based on the incorrect assumption that their damage would have been avoided had Frontex acted lawfully and fulfilled its fundamental rights-related obligations.

While the Court acknowledged the agency’s fundamental rights obligation under the EBCG Regulation[14] and the EU Charter of Fundamental Rights, it also highlighted Frontex's limited role in return operations. It emphasised that the agency is tasked with providing technical and operational support to the Member States and does not have access to the merits of return decisions or decisions regarding the granting of international protection, which belong to the sole competence of the Member State.

Given Frontex’s lack of competence regarding return and asylum decisions, the Court held that there is no direct causal link between Frontex's conduct and the alleged damage. Any incurred damage is, in principle, Greece's sole responsibility. Consequently, the Court dismissed the action for damages in its entirety. Given that the conditions for EU liability apply cumulatively and that the condition of causality was not fulfilled, the Court deemed it unnecessary to examine the other conditions per its settled case law.[15]

V.  Legal remedies and litigation avenues before the CJEU

The CJEU can ensure the enforcement of EU law using an action for damages, an action for annulment, or an action for failure to act. All these instruments are, in principle, available to applicants who wish to challenge the conduct (actions and omissions) of Frontex also in the context of human rights violations. However, strict admissibility requirements and other procedural obstacles stand in the applicant’s way.

Firstly, an act of the agency can be declared void, or the failure to act contrary to the Treaties, as a result of the legality review of the Court performed under arts 263 and 265 TFEU. In an attempt to seek a legality review of acts and omissions of Frontex, the individual faces strict accessibility requirements. The reviewability of the acts of the agency can also be challenged, as most of the acts of the agency do not have legal effects vis-à-vis individuals. The stringent accessibility requirements for individuals before the CJEU and the limited reviewability of the agency's actions pose substantive hurdles that make the legality review options (action for annulment and failure to act) hardly attainable.

Secondly, the action for damages under art. 268 in conjunction with art. 340 TFEU emerges as the most appropriate litigation route for individuals seeking Frontex's accountability. This remedy can address the liability of Frontex directly and has the potential to make good any damages caused by the agency in the course of its activities. However, it presents unique procedural challenges, including the burden of proof and establishing a direct causal link in a multi-actor environment.[16]

VI. In light of precedent: evolving perspectives on causation and joint liability

In the last years, several complaints have been lodged before the CJEU regarding the lack of transparency in the work of the agency[17] and an action for annulment concerning a call for a tender.[18] The first action concerning the legality of the acts of the agency in relation to human rights violations was submitted pursuant to art. 265 TFEU (action for failure to act), asking the Court to assess the legality of the decision of the Frontex Executive Director to not suspend or terminate Frontex operations at the Aegean in light of serious and persistent human rights violations in accordance with art. 46(4) EBCG Regulation. Running against admissibility hurdles, the Court considered this action inadmissible.[19]

WS v Frontex was, however, the first case concerning Frontex's liability for human rights violations. There the aforementioned challenges regarding an action for damages proved to be crippling obstacles in holding the agency accountable. In particular, consistent with its earlier case law,[20] the Court in WS v Frontex ruled that the damage invoked must be the necessary and sufficiently direct result of the alleged illegality.[21] The mere fact that the unlawful conduct constituted a necessary condition (a condition sine qua non) for the damage to arise, in the sense that the damage would not have arisen in the absence of such conduct, is not sufficient to establish a causal link. The actions of the national authorities would be the only direct causal link to the damage. The fact that the agency had no access to the relevant asylum or return decision and, therefore, no influence upon them was equal in the eyes of the Court to a complete lack of a direct causal link to Frontex.

This interpretation reflects the general approach the Court adopts in its case law of the competence model, which is examined further in sections VII.3. and VIII.2. Despite this broader approach, the issue of causation is to be determined on a case-by-case basis. Thus, room can be found in the Court’s case law for less restrictive interpretations. In Krohn, the Court held that the causal link is not severed by an implementing act of the state if the latter was not acting independently but under the binding instructions of the Union.16 While the CJEU has often affirmed the competence model in its case law, which allocates liability on the basis of normative control and de iure powers, it has held in KYDEP that such instructions, which are treated as de facto binding by the member state, can leave the causal link intact.[22]

More importantly, in Missir Mamachi di Lusignano v Commission, the EU Civil Service Tribunal acknowledged that certain unlawful conduct did not necessarily need to be the sole cause of the damage for the causal link to be deemed sufficiently direct.[23] Nevertheless, it would be necessary to establish a direct causal link with each cause.
 The strict interpretation of the causation criterion under the competence model creates, in most cases, a binary distinction in the attribution of responsibility, where either the member state or the agency can be found to have caused the damage. Missir Mamachi di Lusignano, however, along with Brasserie[24] and Bergaderm,[25] which have brought the actions for damages for state (Francovich liability[26]) and EU liability (art. 340) closer, applying a common test,[27] have left room for the allocation of joint liability to multiple responsible actors.

The strongest jurisprudential evidence of the joint liability of the EU and its member states is to be found in Kampffmeyer, where the Court ruled on damages the applicants had suffered due to the failure of the Commission to exercise its supervisory powers under EU law appropriately.[28] In this case, the Commission had unlawfully approved a protectionist (of the national economy) measure of the Member State, which resulted in damages for the company with respect to its maise import activities. Thus, in Kampffmeyer, the liability of both the EU and the Member State was acknowledged.

In the recent Kočner v Europol,[29] where the applicants asked for the joint liability of Europol, the EU's law enforcement agency, and the Slovak authorities for damages related to the right to family and private life (art. 7 EU Charter) from unlawful data processing, the General Court repeated its restrictive view towards causality. In particular, the Court held that the applicants failed to establish a sufficiently direct causal link to Europol's conduct.[30] The case is currently under appeal before the Court of Justice, and it shall be intriguing to witness the Court's response to Advocate General Rantos's viewpoint on the nature of the responsibility of Europol. In particular, in his accompanying opinion, AG Rantos promotes a more permissive stance on the causation requirement based on the textual, contextual, teleological and historical interpretation of art. 340 TFEU. He concludes the interpretation explicitly advocating in favour of joint and several liability, according to which the unlawful data processing can be attributed to both parties.[31]

VII.     Consequences of WS for Frontex Accountability

vii.1. Rapid reactions and resounding criticism

The reactions to WS v Frontex by academic commentators were rapid and fuming. Numerous blog posts emerged within days and weeks from the publication of the judgment, delving into specific aspects of substantive human rights,[32] responsibility and attribution,[33] procedural matters and liability[34] and other remedies.[35]

A prevailing sentiment among critics is the perceived imposition of an "unreasonably and unnecessarily high threshold" for the causal link requirement,[36] which virtually precludes any meaningful accountability for Frontex, leaving the agency seemingly immune to breaches of its obligations. This critique is captured in the notion of "argumentative short-circuit", underscoring the observation that Frontex's lack of competence in assessing asylum applications or return decisions should not absolve the agency from respecting migrants' human rights.[37] Similarly, Gareth Davies rejects the notion of a “just following orders” defence for Frontex, emphasising that, as an expert organisation tasked with monitoring and ensuring compliance with fundamental rights during return flights, the agency is supposed to be “the grown-up in the room”.[38] Davies implies political motivations behind the judgment, asserting that courts should not prioritise making politicians happy. De Coninck also criticises the judgment for seemingly "cherry-picking the arguments brought forward by the Applicants," suggesting a selective approach that may have contributed to the perceived shortcomings of the decision.[39]

The general sentiment of the comments reflecting a broader scepticism regarding the coherence of the judgment is encapsulated in the authors’ criticisms of the precedent as “worrying and discouraging”,[40] “curious and misguided”[41] and one more instance where Frontex “got away (…) with human rights violations”.[42] In a cautiously optimistic outlook, Cornelisse wonders if the Court of Justice in appeal can “sort out this mess”.[43]

vii.2. Unaddressed aspects: what the Court didn't say

In a more in-depth examination, it becomes evident that the answers provided by the Court might not align precisely with the questions posed, while aspects essential to address questions regarding the liability of Frontex in return operations are overlooked.

The most evident oversight in the judgment concerns the misinterpretation by the Court of the applicants’ argument. The Court points out that the applicants incorrectly argue that if Frontex had complied with its fundamental rights obligations, “they would have been accorded international protection and not have been subject to return proceedings”.[44] In doing so, the Court substantially misses the mark, as the applicants did not challenge the adoption of the (negative) asylum and return decisions (which in this case were never issued), but the execution of the (non-existent) return decision by Frontex. In this sense, the question answered was not necessarily the question asked.

With this ill-advised and regrettable manoeuvre, the Court managed to sidestep the issue of the agency's positive obligations. Frontex has positive obligations under human rights law (EU Charter), while central in this respect are the positive duties that have been set out in the EBCG Regulation, such as the obligation of the agency to ensure respect for fundamental rights during all its activities (art. 34 EBCG Regulation 2016/1624),[45] including its return operations (art. 28(3)). In particular, in performing its tasks, Frontex is mandated to ensure that no person is returned in violation of the principle of non-refoulement, including chain refoulement, as was the case in WS (art. 34(2)). The agency is also obliged to take into account the special needs of children during its operations (art. 34(3)), while all operations shall be monitored by forced-return monitors to ensure compliance with fundamental rights (art. 28(6)). On the basis of its positive obligations, the agency must take active steps to protect against interference with a given right. Based on the agency’s supervisory, reporting and communication duties, the agency could have, in the least, been here reasonably expected not to enter into the merits of the return decision but to ensure that such a decision had been issued and that the return fulfils the procedural requirements of the Returns Directive.[46] The Frontex return escorts could also have been expected to ensure that the return is conducted in conditions that are compatible with the rights of the children on board and that the coercive measures were necessary and proportionate.

Failure to abide by these positive obligations can make the agency complicit in a violation that could have been prevented with its appropriate intervention. Additionally, its monitoring duties can lead to the conclusion that the agency had “presumed knowledge” of the violation, which would trigger its responsibility in case of inaction.[47] The key argument here is that the agency was under a positive obligation to verify the issuance of a return decision. Moreover, considering the well-documented issues of systematic violations in Greece, the agency should have exercised extra caution. This is highlighted by the agency’s own Fundamental Rights Officer (FRO), who recommended the suspension of operations in Greece due to serious and persistent human rights violations on the basis of art. 46 of the 2019 EBCG Regulation.[48] This indicates a recognised need for heightened scrutiny and precautionary measures. If the agency chose to turn a blind eye to a predictable and reliable threat to human rights, its responsibility for failing to abide by its positive obligations indicated above needs to be activated.

By sidestepping these issues, the Court failed to provide answers to the essential questions: How should the positive obligations of the agency in the context of coordinating joint return operations be realised in practice? Did the agency fail to abide by its positive obligations to protect the family from the alleged violations? Did this failure constitute an unlawful act? Is Frontex liable for this failure?

Other questions that remained unaddressed due to the Court’s strategic focus concern, for instance, were whether the actions of the FRO in relation to handling the related individual complaint were challengeable before the CJEU under art. 263 TFEU.[49] This unresolved issue leaves still open the question regarding the judicial review of decisions made by Frontex in the context of its internal individual complaints mechanism. Furthermore, by immediately focusing on causality, the Court crucially sidesteps the questions of attribution and legal qualification of the conduct. As rightly observed, “in order to assess causality between unlawful conduct and damages, it is necessary to establish, first which actor has done what (attribution) and secondly, how that particular conduct can be qualified under the law (i.e. as unlawful or lawful). Only after this has been clarified is it possible to assess the causality between the unlawful nature of the act and damages”.[50]

vii.3. Frontex Liability under the Competence Model

In WS v Frontex, the Court followed the competence model, which is the basic model of responsibility in EU law.[51] The competence model attributes responsibility on the basis of formal decision-making competence within the EU system. In areas where the EU has exclusive competence to adopt binding acts, it also carries the exclusive responsibility for unlawful, harmful outcomes. In areas of shared competence (where EBCG operations can be placed), responsibility can be allocated based on priorly agreed-upon regimes of formal competencies. The CJEU has often reaffirmed the competence model in its case law, allocating liability on the basis of normative control or, in other words, on the basis of legal decision-making powers. The determinative questions are: who exercised legal control, whether the actor had the authority to issue legally binding orders, or whether it had formal discretion to determine its conduct. For instance, in Krohn I, the Court rejected the liability of the Member State as it came to the conclusion that it had no formal discretion to derogate from the instructions of the Commission.[52]

Consequently, the competence model does not leave much space for factual control, double attribution or complementary (indirect) responsibility, which are necessary elements in determining joint responsibility within EBCG operations and the accountability of Frontex, resulting in a gap in accountability for acts that impact on fundamental rights.

This gap was found incompatible with the safeguarding of fundamental rights by the European Parliament as early as 2015 before the extensive expansion of the mandate and capabilities of the agency in 2016 and 2019:

“[...] Frontex coordination activity cannot, in practice be dissociated from the Member State activity carried out under its coordination, so that Frontex (and thereby the EU through it) could also have a direct or indirect impact on individuals’ rights and trigger, at the very least, the EU’s extra-contractual responsibility [...], whereas such responsibility cannot be avoided simply because of the existence of administrative arrangements with the Member States involved in a Frontex-coordinated operation when such arrangements have an impact on fundamental rights”.[53]

VIII. Broader Consequences of WS for EU Law

viii.1. The ecosystem of integrated administration

By adopting an exceptionally stringent interpretation of causality and avoiding pronouncing itself on the above-highlighted issues, the Court risks setting in motion snowball consequences not only in the broader area of border management but also throughout EU law.

Interpreted so narrowly by the Court, the direct causal link proves too strict of a requirement for the liability of Frontex since its actions, by definition, occur in an environment of “many hands”,[54] where a nexus of different responsibilities exists,[55] and a severality of acts and omissions by different actors may cause (also in the sense of contributing to) the harmful result. In fact, the mere involvement of the Member State can be sufficient for the CJEU to break the chain of causation and prevent the liability of the agency.

The applicability of this observation extends beyond EBCG operations to numerous complex structures and multi-actor operational settings under the umbrella of shared or integrated administration. Enhanced cooperation and coordination have been imperative for the function of a common area without internal borders in a mixed intergovernmental/supranational environment. This has been essential to ensure uniform standards and the consistent implementation and enforcement of EU law at the member state level. The centralisation of administrative bodies and practices at the EU level, forming a “common European model,[56] coupled with the Europeanization of administrative structures in various Member States, [57] has given rise to a European administrative space.[58] Within it, multiple network structures constitute the modern EU integrated administration.[59] These structures of close administrative cooperation take various forms, including cooperation among administrations (e.g. joint controllership over personal data between the European Central Bank and National Competent Authorities under the Single Supervisory Mechanism),[60] composite administrative procedures with input from diverse administrative actors from both member states and the EU (e.g. in the area of access to information),[61] or agencification and the joint structures the agencies have created (e.g. EBCG operations).[62]

In this ecosystem of integrated administration, the network structures jointly exercise powers in a system of shared sovereignty.[63] The European administrative space evolves, thus, into a “three-dimensional concept with complex vertical, horizontal and diagonal relations among the actors”.[64] This reality of integrated administration is counter to the more traditional model of EU administration, i.e. the two-level system of “executive federalism”. Under this model, the decision-making powers belong to the EU, while the Member State authorities are responsible for implementing the EU rules on the ground, which creates a dual distinction of administrative procedures.[65]

viii.2. Is there space in EU law for joint liability?

The ecosystem of integrated administration creates an increasing need for shared responsibility frames in case of wrongdoing. However, the EU joint liability frameworks have failed to keep pace with the constantly expanding ecosystem of integrated administration.

The application of shared responsibility in the practice of the CJEU is rather challenging. The limited case law produced by the CJEU on the joint liability of the Union and its Member States is barely adequate to form the basis for decisive predictions on joint liability cases.

a) The Kampffmeyer Avenue

The Court has acknowledged the possibility of the EU and its Member States being jointly liable in Kampffmeyr in 1967. However, this possibility has remained theoretical since, despite the convergence of the two tests, the state (Francovich) and the EU liability have to be decided by different courts. The CJEU has exclusive jurisdiction over the non-contractual liability of EU Institutions,[66] bodies and agencies, while compensation ought to member state liability under EU law needs to be claimed before domestic courts.[67] These rules can only be interpreted as dividing jurisdiction in a way that actions for damages attributed to the Union are dealt with by the CJEU, and those attributed to member states are dealt with by domestic courts.[68]

Thus, in Kampffmeyer, the CJEU had to stay in the proceedings until the German courts reached a conclusion about the liability of the German authorities. The German court followed the example of the CJEU and ordered a stay of proceedings, waiting for a ruling at the EU level. As a result, the applicants found themselves trapped in this legal battle for two decades.[69] Eventually, after the completion of the domestic proceedings, none of the applicants returned to the CJEU to examine the liability of the EU. The case was ultimately removed from the Court’s records.[70] In practice, since Kampffmeyer, the Court tends to reject claims for damages against the EU when the possibility exists to join the case against the member state and the case against the EU in the way described above in order for compensation to be sought against the member states before the national courts.[71]

b) The Unifrex Avenue

Another possible scenario is that any case concerning the liability of the EU when that liability can be shared with a Member State, will be dismissed by the CJEU unless the applicants have not secured compensation at the national level. Given the different jurisdictions for the separate actions for damages, the CJEU has ruled in Dietz v Commission that a legal remedy needs to be sought first before the national courts.[72] The issue of the exhaustion of the local remedies could be resolved by reasonably arguing the lack of available remedies at the domestic level regarding the liability of the EU, as the CJEU has exclusive jurisdiction. The case law of the Court in Unifrex, however, points in a different direction. There, the Court held that in a case concerning the implementation of an EU measure by the national authorities, the applicant needs to contest that measure before national courts before resorting to an action for damages against the EU. According to the Court, the availability of domestic remedies depends on whether these can ensure effective legal protection and result in compensation for damages.[73] This interpretation, strictly prioritising the domestic remedies against the Member State, as these would be sufficient to ensure effective legal protection, makes effective legal protection the sole purpose of the justice system. As a result, this would allow EU actors to be left unaccountable if their unlawfulness coincides with that of a Member State.

c) The binaries of causality and jurisdiction and the competence model

Thus, the question arises: Is there scope for (non-contractual) joint liability in EU law? Some commentators have adamantly argued that EU law leaves no space for joint liability.[74] My analysis suggests that the potential for joint liability exists,[75] albeit hindered by procedural challenges in its implementation.[76] I identify two procedural challenges, namely a) the binary of causality and b) the binary of jurisdiction.[77] According to the binary of causality, the stringent interpretation of the causality criterion for liability creates, in most cases, a binary distinction in the attribution of responsibility, which allows for only one actor to be held responsible, leaving little space for the acknowledgement of joint liability. The binary of jurisdiction refers to the different judicial fora which need to deal with the EU and the Member State liability separately, which hinders the execution of joint liability. Unless the liability of both (all) actors can be brought under the same judicial forum (e.g. the ECtHR after the accession of the EU to the ECHR) or unless a way is found to bridge the different proceedings (e.g. a solution based on the principle of subsidiarity),[78] the possibility for joint liability before EU law will remain theoretical.

The binaries of causality and of jurisdiction express the “either-or” mindset of the Court, which is inspired by the competence model, as shown in section VII.2. The competence model, as the source of these procedural hindrances, can essentially inhibit the accountability not only of Frontex but of any EU actor involved in the ecosystem of integrated administration.[79] It can thus be convincingly argued that it is inadequate to fully address the responsibilities of “many hands” in the contemporary ecosystem of EU integrated administration. The reality of the European administrative space needs to be accompanied by appropriate accountability tools that are able to address the challenges of shared administration. The two-dimensional logic of the competence model, however, corresponds to the outdated model of executive federalism.

d) Legal Accountability Gap and the Rule of Law

In WS v Frontex, the Court had the opportunity to recover lost ground and develop a joint liability framework that can deal with today's realities of the EU's integrated administration. It could have developed further its own case law on positive obligations to examine exceptions to the strict causality criterion when liability can occur as a result of the breach of positive obligations. It chose, however, to ignore the relevant questions, risking to authoritatively confirm a widely argued accountability gap.[80]

The consequences of this choice extend beyond Frontex itself to all instances of EU liability in the context of integrated administration. Taking into account the extent of the multiple administrative network structures that constitute the European administrative space, failure to address the issue of joint liability gives rise to a major accountability gap. This gap poses a critical threat to the Rule of Law within the European Union. By neglecting to establish a framework for joint responsibility, the EU runs the risk of creating areas of no accountability, effectively forming a no man's land for responsibility in EU integrated administration.

IX. Conclusion: shaping the EU joint liability landscape?

A highly anticipated judgment, WS et al. v Frontex was the first occasion in which the CJEU had the opportunity to rule on the liability of the European Border and Coast Guard Agency for human rights violations. The Court, applying an intensely high threshold for causality, based its dismissal of the applicant’s claim on the lack of competence by the agency to influence asylum or return decisions. In doing so, the Court chose to circumvent the essential questions of positive obligations, unlawfulness and attribution of responsibility. At the end of the day, we are none the wiser about violations at the external borders or the legal enforceability of the obligations of the agency.

More importantly, this unsatisfactory judgment brings the CJEU's causal link requirements into focus, which leaves a critical accountability gap in “many hands” environments. The analysis underscores the pressing need for a reevaluation of the existing competence model of assessing EU liability, particularly in its limitations concerning the liability of multiple actors. Such limitations can be classified as the binary of causation (limitations including factual control, double attribution, and complementary/indirect responsibility) and the binary of jurisdiction (limitation in implementing joint responsibility under the same judicial forum).

The identified accountability gaps, incompatible with fundamental rights protection, take on added significance given the substantial expansion of the agency's mandate and capabilities. More importantly, this Insight highlights the broader consequences that the Courts’ inability (or unwillingness?) to deal with the liability of multiple actors can have upon EU law more generally. The Court’s non-contractual liability framework is maladapted to the changing multi-actor European administrative space. The case of WS v Frontex encapsulates the complexities and challenges inherent in holding an EU Institution, body or agency accountable in today’s complex ecosystem of integrated administration.

The Court’s stance not only perpetuates a widely contested accountability gap but also extends its ramifications to all liability instances within EU integrated administration. Given the intricate web of administrative networks in the European administrative space, the absence of an effective framework for joint liability creates a perilous accountability void, posing threats to the Rule of Law in the European Union. The Court's reluctance risks establishing areas devoid of accountability, resulting in a precarious no man's land for responsibility within integrated administration. Addressing these shortcomings is not merely a matter of procedural refinement; it is a crucial step towards enhancing joint responsibility determination and ensuring robust accountability mechanisms in EU law.