13 October 2023

A Hidden Success

Why the EU General Court’s Frontex Judgment is Better Than it Seems

Following the EU General Court’s dismissal of the complaint of WS and other asylum seekers against Frontex in its ruling on September 6, 2023, scholarly commentary has largely expressed disappointment. However, a more optimistic way of reading the judgement is also possible. By declaring the lawsuit admissible, the court confirmed that factual misconduct by Frontex can be addressed with action for damages claims – and this in itself is a major step forward in the system of fundamental rights protection in the European Union.

Facts of the Case and the General Court’s Judgement

To briefly recall: In 2016, WS and several other Syrian nationals arrived in Greece and expressed their intention to apply for international protection in the EU. Only eleven days after their arrival, their deportation to Türkiye was carried out in a joint return operation by Greek authorities together with Frontex.

With their complaint, the deported applicants sought compensation for the damage resulting from Frontex’s failure to comply with its obligations under the then-applicable Frontex Regulation, the agency’s Standard Operating Procedures and its Code of Conduct. In essence, the applicants invoked a violation of their right to asylum and the prohibition of refoulement under Art. 18, 19, 4 ChFR, the rights of the child under Art. 24 ChFR, as well as the right to good administration and to an effective remedy under Art. 41, 47 ChFR.

The General Court found the applicants’ complaint under Art. 268, 340 para 2 TFEU to be admissible. As I argue below, this in itself is a great success. When it comes to the merits, however, the judgment remains doctrinally weak – which justifies the harsh scholarly criticism. The General Court’s main consideration here was that Frontex’s contribution could not be considered causal for the relevant damage. It justified this with an argument that as imprecise as it is widespread, namely that Frontex merely assisted in the factual process of deportation.

A Major Step for the Protection of Fundamental Rights

The complaint is of particular importance because it was the first of its kind. After decades of discussion among scholars and practitioners on how Frontex could possibly be held responsible for its violations of fundamental rights, two main approaches have emerged. While some advocate for the route via the annulment procedure under Art. 263 TFEU, or the closely related action for failure to act under Art. 265 TFEU, others suggest employing the action for damages under Art. 340 para 2 TFEU. Both approaches have recently been presented before the CJEU.

While each approach has its advantages depending on the specific circumstances of the case, the action for damages seems structurally more promising. This is because Frontex mainly acts in non-formally binding forms, and only the action for damages allows to challenge factual conduct.

Against this background, it comes as no surprise that the scholarly response to the General Court’s judgement of 6 September was to a large extent marked by disappointment. Joyce de Coninck, for instance, saw the General Court’s judgement as contributing ‘to the systematic shielding of Frontex from any responsibility for contributions to human rights harms’, while Christopher Paskowski observed the perpetuation of an ‘administration without responsibility’.

Such criticism is important and necessary.  But it should not be overlooked that the judgment nevertheless represents an important step towards the protection of fundamental rights against the EU border administration. Even though the applicants lost in the first instance and even though the General Court has fallen for a standard sham argument, the case of WS et al can already be considered a success. For since September 6, there can be no doubt that the action for damages is a suitable way to address factual misconduct by Frontex before the Court of Justice of the European Union. As Mariana Gliakti has put it, the main success of this case is that it has paved the way for further proceedings of its kind.

Adapting the EU Legal Protection System to Administrative Realities

Viewed from this angle, the General Court’s judgement appears as an important step in the gradual process of adapting the EU legal protection system to administrative realities at the EU’s external borders. No doubt, it would have been even better if the General Court had not argued beside the point on the question of causation. But a look into the CJEU’s case law in other areas, such as EU agricultural subsidies or EU financial support, suggests that the judicial development of the EU legal protection system tends to proceed neither quickly nor steadily.

When it comes to judicial protection against the integrated border administration, countless studies have found serious shortcomings in the EU legal protection system. This is hardly surprising, given that the system was designed at a time when EU bodies did not yet interact with individuals in areas that are highly sensitive to fundamental rights, let alone in direct physical contact. Today, however, the outdated design of the EU protection system makes it almost impossible for asylum seekers to exercise their right to an effective remedy under Art. 47 ChFR against EU agencies, despite these agencies directly interacting with them and systematically impacting their fundamental rights. As a result, the rule of law is undermined in an area where its benefits are most urgently needed. There can hence be no doubt that the CJEU must urgently reform its interpretation of the legal protection system.

The Action for Damages as a Makeshift Fundamental Rights Remedy

In this context, the judgement of September 6 confirms that judicial protection against EU bodies can be achieved through the action for damages. In other words, the judgement confirms that the action for damages functions as a makeshift fundamental rights remedy.

Concerning the admissibility of the claims of WS et al, Frontex had argued that the applicants should have lodged an action for annulment against the latest response letter of the agency’s Fundamental Rights Office within the relevant time limit, and that therefore, their later action for damages was inadmissible. The General Court took the occasion to recall and specify the relation between Art. 263 TFEU and Art. 340 para 2 TFEU. Although the action for damages is a self-standing form of action in the EU system of remedies, it must be declared inadmissible ‘where it is actually aimed at securing withdrawal of an individual decision’ (para 24). Nonetheless, the General Court then continued, ‘it would be contrary to the autonomy of an action for damages, and to the effectiveness of the system of remedies’ to consider an action for damages inadmissible ‘on the sole ground that it might lead to a result comparable to the results of an action for annulment’ (para 25). Therefore, the General Court held, ‘an action for damages may also be able to nullify the legal effects of a decision which has become final where the applicant seeks greater benefit, but including that which it could obtain from an annulling judgement’ (para 26).

Remarkably, the General Court did not even object to the applicants not only seeking compensation for their material and immaterial damage but also requesting the Court to ‘find that Frontex engaged in improper conduct with regard to them’ (para 17). Dedicating only a few paragraphs to this issue, it considered it ‘clear’ that the present action for damages ‘does not have the same purpose or the same effect as an action for annulment’ as the maximum effect of an action for annulment would have been a fresh examination by Frontex’s Fundamental Rights Office. On this basis, the General Court deemed the action for damages admissible (para 27 to 29).

The judgment thus clearly confirms two claims that have been made repeatedly by scholarship. The first, as made for instance by Timo Rademacher, is that the action for damages slowly evolves into a form of declaratory relief. Albeit far from being designed for such purposes, Art. 340 para 2 TFEU has proven to be the only procedure in the EU legal protection system that is flexible enough to allow for judicial review of factual conduct. The second and closely related claim, made for instance by Angela Ward and by Melanie Fink, is that the action for damages functions as a makeshift fundamental rights remedy. Put differently, Art. 340 para 2 TFEU secures access to the CJEU in cases where EU bodies act in a non-formally binding manner, which is the case in the majority of border administration scenarios.

From a practical standpoint, the core benefit of the judgement of September 6 is that it enables future complainants against Frontex to rely on Art. 340 para 2 TFEU without lengthy discussions of admissibility. From now on, a simple reference to the General Court’s argument in WS et al should suffice.

Contingency of Facts and Difficulties of Proof

Speaking about future cases, it seems that slightly different constellations might have more chances to succeed on the merits. This is not merely a hint from the ivory tower; it is obvious that the well-documented lack of documentation and proof remain significant hurdles to litigation against Frontex. Based on what can be seen from the judgement, there can be no doubt that the representatives of WS et al have done an excellent job in this regard.

However, it should be kept in mind that the case of WS et al was about a deportation, not about a pushback. As observed by Omer Shatz, this difference is important. While a pushback is unlawful per se, a deportation can be either lawful or unlawful. Among other factors, the lawfulness depends on whether the country of destination is safe for the individual deportee. In the case at hand, the decisive question hence was whether Turkey can be considered as a safe third country or not. Given that the Court of Justice had made clear years ago already that it was not willing to decide this highly contentious issue, it seemed unlikely that the General Court would take a stance on this matter. In cases of pushbacks, however, the General Court might have less difficulties to identify these as unlawful.

Furthermore, the existence of material and non-material damage might be easier to prove in cases that slightly differ from that of WS et al. Here, the applicants claimed compensation for material and non-material damage that had occurred due their onward flight from Türkiye to Iraq (para 58 and 59). While this onward journey can hardly be said to result from a free decision of the applicants, as the General Court suggested, the fact remains that the relevant damages occurred not as the temporally most direct consequence of the deportation as such. Although the applicants’ arguments for causation in this regard might convince in light of the CJEU’s own case law, the reference to the comparatively long chain of causation might still be a weak point from a strategic point of view. And indeed, the General Court was eager to make use of this weakness in order to support its own doctrinally weak arguments in the context of causation  (para 67 to 70). In cases where the invoked damage is a more direct consequence of the process of deportation, this additional hurdle could be avoided.

The Question of Causation and De Facto Bindingness as Relevant Threshold

As regards the case of WS et al, it remains to be seen if and how the Court of Justice will decide. As Gareth Davies put it, ‘things do not look good for Frontex’ and ‘the Court of Justice must now sort out this mess on appeal’. This being said, some doctrinal weaknesses of the General Court’s reasoning are so obvious – and have been highlighted so immediately and consistently – that it will be very difficult for the Court of Justice to not be more precise. And when it comes to holding Frontex accountable, precision helps.

As mentioned above, the General Court’s brief argument in essence repeats the standard answer of Frontex itself to allegations of misconduct. The point is as commonplace as it is inconsistent, suggesting that the causal link between Frontex’s conduct and the resulting damage is not established because Frontex ‘only assists’. In other words, the argument is that Frontex does not have the competence to issue return decisions, and therefore its operations cannot be considered causal for damages resulting from the execution of the deportations.

As shown by Joyce de Coninck and Christopher Paskowski, this argument is plainly wrong. The General Court made at least two very basic mistakes.

First, it failed to define Frontex’s misconduct. Instead of beginning its assessment with the identification of what exactly Frontex did wrong, the General Court jumped directly to the criterion of the causal link (para 55) and based its assessment of causation on a general reference of ‘Frontex misconduct before, during and after the deportation’ (para 57). This is problematic because a meaningful evaluation of causation necessarily presupposes a prior definition of the conduct at stake. Although the conditions for EU liability are cumulative, as the General Court rightly stresses (para 53), it seems logically impossible to assess the causal link between the conduct and the damage without prior identification of the relevant conduct.

The identification of relevant misconduct requires distinguishing between the rejection of an asylum claim, the issuance of a deportation order, and the execution of a deportation – and to recognise that Frontex’s conduct forms part of the latter phase. On this basis, the Court of Justice will then have to identify what exactly Frontex did wrong. From what can be discerned from the judgement, this will include a failure to raise objections to an unlawful deportation, a failure to insist on the respect of EU law, and ultimately a failure to withdraw its administrative support altogether.

The second basic mistake of the General Court’s analysis was made in the context of causation. Here, the General Court started with the wrong question. Instead of asking whether Frontex’s conduct properly defined was causal for the resulting damage, the General Court implicitly asked which conduct was causal for the resulting damage (para 62). It then concluded that the national decision to reject the applicants’ asylum claims and to issue their deportation orders was causal for the deportation (para 64 and 65). While this statement is correct in itself, it ignores that there can be multiple causes for one result, i.e., that both a national administrative decision and EU administrative support can be causal for a resulting damage. Based on its wrong assumption, however, the General Court then observed that Frontex was not competent to issue the relevant administrative decisions, and from this concluded that Frontex’s conduct could hence not have been causal (para 66). Again, the argument here misses the point. While it is of course true that Frontex’s competences are limited, this as such does not say anything on whether its factual conduct is causal for a resulting damage. It will hence be for the Court of Justice to ask the right question, namely whether Frontex’s participation in the process of deportation was causal for the damage resulting from the deportation, and to answer this question on the basis of an extensive case law analysis.

The issue of causation is complex and requires delving into the CJEU’s ca