30 January 2024

Shielding Frontex 2.0

The One with the Impossible Proof

When WS and others v Frontex came out in September of 2023, many of us working around questions of legal responsibility of Frontex for complicity in unlawful human rights conduct, were surprised by the brevity with which the General Court (GC) dismissed the action (see here, here, here, here, here and here). The GC left open the question of if, when, and how independent or joint human rights responsibility would arise when Frontex is engaged in shared operational conduct with the Member States. The prevailing sentiment at the time appeared to be that the GC would have another opportunity to clarify these pressing questions in Hamoudi v Frontex. But again, the action is dismissed. This time not on the basis of an obscure re-interpretation of the Applicant’s claim, but instead, on the basis of an unattainably high and unrealistic burden, standard and method of proof. In doing so, the General Court again eschews from clarifying the nature, conditions and consequences of both independent and joint human rights responsibility of Frontex. Taken together, these cases raise the question whether there are any viable forms of judicial recourse for fundamental rights violations committed or contributed to by the EU’s Border and Coastguard Agency.

Facts and Ruling

In April 2020 Alaa Hamoudi arrived irregularly by boat from Türkiye on Greek territory in pursuit of asylum. Shortly after disembarkation, he claims to have been intercepted by the Greek authorities, who confiscated his cell phone and sent him back out to sea with 21 others, where he claims to have seen a Frontex-operated airplane on two different occasions tracking the movements of the boat he was on, described as “a life raft without any means of propulsion”. On the night of 28 and 29 April, Hamoudi and others were subsequently intercepted by the Turkish coast guard and taken into detention, where his passport was confiscated.

With this action, Hamoudi sought compensation for the damage incurred during and after the push-back operation in the Aegean sea. He attributes this damage to Frontex’s alleged failure to comply with its fundamental rights obligations under the 2019 Frontex Regulation and the Charter of Fundamental Rights, relating inter alia to the non-refoulement principle and the prohibition of collective expulsion. Applying the ‘the but-for’ test of causality (see here), and mindful of the two Frontex operations underway in the region at the time, the Applicant identifies Frontex as the “true author” of the push-back and claims that Frontex should additionally be held responsible for its “aiding and assisting in the commission of the unlawful collective expulsion”. The case thus presented the GC with an explicit opportunity to clarify the nature, conditions, and consequences of both independent and joint human rights responsibility of Frontex.

The GC clarified that in an action for damages there is no specific chronological order to follow when examining the cumulative conditions of unlawful conduct (1), actual damage suffered (2), and the causal link between the damage and the unlawful conduct (3), and subsequently embarks on its investigation of the ‘actual damage’.

Unfettered Production of Evidence

The GC first recalls its settled case law whereby damages will only be awarded insofar as the Applicant can adduce conclusive evidence of actual and certain damage, which cannot be assessed in the abstract, but instead “in relation to the specific facts characterizing each case in point”. This immediately raises an under-scrutinized but critical question in cases of alleged push-backs at sea: how to obtain evidence of what would be an illegal and covert operation? I can’t help but wonder whether my own first instinct would be to meticulously document an ongoing push-back in the middle of the night, without the tools to do so, on a raft lacking propulsion capacity and without any certainty of survival.

Having provided a written declaration, a Bellingcat article concerning the events, and four photographs taken as screenshots from two YouTube videos recorded by third parties as evidence, the General Court rejected it as being manifestly insufficient, noting that the witness statement lacked credibility and reliability, and that the Applicant is not identifiable in the photographic material. Specifically, the Court held that the witness statement is insufficiently specific as the Applicant could not recall the exact date during which the push-back transpired, that the Bellingcat article “does not feature the applicant’s name”, that in the videos of which screenshots were taken, the Applicant “is not looking directly at the camera”, and finally that “it is not possible to establish the date or the location of the events concerned by the screenshots”.

But how can this evidentiary burden ever be met? It is hardly plausible that the Greek Coast Guard or Frontex would voluntarily corroborate their complicity in a seemingly clear-cut violation of (customary) international law, that the 22 individuals onboard the vessel (including the Applicant) would all be sufficiently acquainted with each other beyond the confines of their shared experience, so as to corroborate each other’s experience years after the facts, and that the Applicant – not knowing that he was being filmed given the origins of the video material – would adequately pose for a video so as to be identifiable for future litigation should he survive the pushback.

The ongoing Frontex operations in the same geographical zone suggest a significant imbalance in the evidentiary burden borne by the Applicant when compared to that of Frontex. Essentially, the GC appears to be asking of the Applicant an ‘impossible proof’ (‘probatio diabolica’), whereby what is demanded of the Applicant both in terms of individual experience and Frontex’s role therein, is simply impossible to prove given the material circumstances of the case. Conversely, Frontex – with two active operations in the region – was not called upon to prove or disprove the Applicant’s claims.

When facing specific evidentiary difficulties, or in situations where the defending State has access to information capable of corroborating or refuting the Applicant’s allegations, the ECtHR’s approach typically involves either sharing or reversing the burden of proof (see here, here, and here). In casu, the GC not only maintains the exclusive burden but articulates an unsustainably high standard and method of proof, not on par with the Applicant’s (vulnerable) status as an asylum seeker. Given the Charter-based allegations at stake – which correspond to provisions in the ECHR – it is striking that there is no engagement with the ECtHR’s case law and how the respective evidentiary rules have developed in analogous cases, in line with Article 52(3) CFR.

Unfettered Assessment of Evidence

Next, the GC elaborates on the principle of unfettered ‘assessment’ of evidence, following which the probative value of evidence is an exclusive prerogative of the Court. Accordingly, the document’s evidentiary value depends on its origin, creation circumstances, recipient, content, and overall reliability.

The document relied upon in casu, concerns the Applicant’s own statement. Here too, the GC dismisses the evidence as unreliable, on the basis of lacking specificity concerning the other individuals onboard with the Applicant, the fact that the Applicant didn’t remember the exact date of the push-back and that the Applicant’s statement was drafted only once the Applicant was being assisted by a non-governmental organization more than a year after the facts.

The Court’s rejection of the credibility of the evidence is striking on two counts, relating first to the practice of the ECtHR in analogous expulsion cases, and relating second, to studies on behavioral psychology, traumatology and the credibility of accounts provided by asylum seekers.

The ECtHR has typically held that the Applicant should be given the benefit of the doubt in the face of evidentiary uncertainty (see here). This aligns with the aforementioned approach whereby the ECtHR may reconsider the burden and subsequent standard of proof when an individual encounters difficulties in acquiring evidence and where governmental authorities have access to information not readily available to the Applicant capable of dispelling or corroborating the presented claims. In the Court’s order however, there is no explicit or implicit engagement with these considerations, including those consolidated under EU law, notwithstanding the fact that the case in essence concerns fundamental rights claims that correspond to fundamental rights under the ECHR and thus, in line with Article 52(3) CFR, should be interpreted in line therewith.

Secondly, bearing in mind that “scientific studies in psychology and traumatology suggest that standard credibility criteria, such as behavior or the coherence and plausibility of statements are also deficient” (see here), it is puzzling that such considerations (seemingly) did not factor into the evidentiary assessment by the Court. Studies in traumatology and behavioral psychology highlight that reliance on memory for the purpose of corroborating past events is challenging, including for asylum seekers, particularly as concerns recalling (for example) temporal information, proper names, and peripheral memories (see here). In addition, the impact of trauma and post-traumatic stress disorder in migration related cases, complexifies coherent and plausible memory recollection for asylum-seekers (see here, here, here, and here) Bearing this in mind, the broad-strokes approach adopted by the Court in dismissing the credibility of the Applicant statement, appears hard to reconcile with both the ECtHR’s practice on expulsion of asylum seekers, as well as (interdisciplinary) work on asylum law and its interplay with psychology and traumatology (for additional discussions of credibility in asylum cases, see here, and here).

Looking Beyond Hamoudi

Taken on its own, the case merely provides room to ruminate on evidentiary standards, which may appear rather trivial. The case becomes more problematic however, when assessed in tandem with WS and Others on the one hand, and in relation to other (unsuccessful) applications lodged against Frontex on the other hand. Frontex’s human rights record has been subjected to pervasive criticism (see for the most recent examples here, here and here). Yet, attempts to hold Frontex to account in Luxembourg are largely unsuccessful.

In both WS and others v Frontex, as well as Hamoudi v Frontex, the General Court was presented with the opportunity to provide clarification on the cumulative conditions of causation, unlawful conduct, and damage that must be met in an action for damages. Yet, in both cases, the General Court sidestepped any substantive appraisal of Frontex’s involvement in alleged unlawful human rights conduct. In WS and others v Frontex, the Court sidestepped the question of how causation (and attribution) should be determined in operational action concerning both Member States and Frontex, by conflating the conduct of both actors, and concluding that the test of causation connecting unlawful conduct to Frontex cannot be met. Clearer rules on joint liability would arguably prevent the Court from sidestepping substantive appraisals of whether Frontex’s conduct amounted to a violation of its human rights obligations. In Hamoudi v Frontex, the Court sidestepped the question of unlawful conduct and actual damage, focusing instead, on an extremely stringent – impossible – standard and method of proof.

There are a number of – some even understandable – reasons for the Court’s apparent reticence to engage with a substantive appraisal on the matter, but the urgency for interpretation and clarification on the nature, the conditions and the consequences of legal responsibility of Frontex for alleged complicity in human rights abuses, cannot be overstated. There have been transparency cases lodged against Frontex albeit without much success (see here, and pending case here), there have been two unsuccessful actions for failure to act (see here and here), and now after WS and others v Frontex, another unsuccessful action for damages in Hamoudi.

So, what explains these unsuccessful cases: is it truly that the degree of Frontex involvement simply does not meet the thresholds for fundamental rights responsibility? Or are we dealing with reasons of a more systemic nature, shielding Frontex – and EU entities more generally – from findings of human rights responsibility? All eyes are on the Court of Justice in handling the appeals of these cases, mindful of the right to an effective remedy and the EU’s ‘complete system of remedies’. But while we wait, it may be worth considering that there may be something amiss with the EU’s human rights responsibility regime tout court.”