By Grażyna Baranowska1, Marie-Bénédicte Dembour2 and Isabel Kienzle3
This post, an output of the DISSECT’s project, is concomitantly being published on the Strasbourg Observers blog.
Regularly operated at many borders of CoE member states, pushbacks are problematic practices from a human rights perspective. They generally violate the principle of non-refoulement under Article 3 ECHR as well as, most often, the prohibition of collective expulsions under Article 4 of Protocol 4, not to mention the prohibition of inhuman and degrading treatment (Article 3 ECHR) and in some cases the right to life (Article 2 ECHR) and the prohibition of unlawful detention (Article 5 ECHR).
The Court increasingly receives complaints about pushbacks. Amongst the many challenges this case law raises is the evidencing of the incidents at the basis of the complaints. This has already been discussed on both the Strasbourg Observers and DISSECT blogs, where it was for example noted that evidence will be a key question in the pending cases regarding the EU-Belarus border, that it played a pivotal role in previous case law (with regard to Cyprus and Greece), and that domestic criminal procedures related to pushbacks can feed into proceedings undertaken at the regional level, and vice versa.
In March 2025, we submitted a third party intervention to the ECtHR in the case of A.A.N. and Others v. Greece. In it we detailed the evidentiary principles that the Court should apply in pushback cases. This blog post conveys our six main points before summarising them again.
- As the Court’s case law itself indicates, pushbacks are a European reality
The Court’s own case law has already confirmed that various European states have a practice of operating pushbacks. In its recent rulings regarding Greece, the Court has acknowledged the systemic practice of pushbacks at Greece’s land and sea borders.
When pushbacks are known to take place in a particular state, this should give rise to the presumption that the pushback complained of in a particular case against this very state is likely to have happened. How strongly this inference should be held depends on the other evidentiary elements (possibly also of a presumptive nature) with which it is combined.
- Pushback practices are surrounded by secrecy that is deliberately instigated by the state/li>
States seek to hide their pushback operations, including by intentionally failing to record their actions and by systematically confiscating the phones (and thus cameras) of the migrants whom they are pushing back. Indeed, the Court has acknowledged ‘the by definition secret and unofficial nature of [pushback] actions’. The result is that victims-survivors have little to no direct evidence to offer the Court when they bring a case before it.
- State-created evidentiary holes require the Court to adapt its evidentiary approach
In the majority of cases that make the Strasbourg case law, the evidence before the Court originates from submissions by either the applicant or the respondent state. In pushback cases, however, the applicant has little evidence to submit. As for the state, it has no interest in filling the evidentiary holes it has created precisely to escape accountability regarding its deliberate circumvention of the legal frameworks put in place in order to protect migrants’ rights, including human rights.
Such a serious human rights situation should not be left unaddressed. Instead the Court must adapt its evidentiary framework to the evidentiary challenges which pushback complaints raise. This will often require shifting the burden of proof onto the respondent state, as described below. It is nonetheless important to note, that, in the few exceptional pushback cases where strong evidence is submitted against the odds, the Court must recognise this strength and proceed with finding the factual complaint substantively established. We first make this point before turning to the shifting of the burden of proof.
- Occasionally a pushback complaint will succeed in meeting the Strasbourg ‘beyond reasonable doubt’ standard of proof
Whilst the secrecy which surrounds pushbacks means that these operations can rarely be evidenced directly, occasionally a case arises where a wide array of presumptive evidence is submitted to the Court. This could, for example, consist of a detailed account by the victim, largely corresponding to what is known through intergovernmental or civil society reporting to be practiced in the respondent state, additionally corroborated by private Whatsapp messages sent around the material time to loved ones and/or lawyers, possibly rendering precise geolocation identifiable. When the evidence submitted is such that it is no longer possible to believe that the claimed pushback did not happen, then the right course for the Court is to find the contested pushback established – and this, beyond reasonable doubt.
‘Beyond reasonable doubt’ is the standard of proof the Court generally applies when facts are disputed. Importantly, right from Ireland v the United Kingdom, the Court has made it clear that this standard can be attained at Strasbourg through strong, clear and concordant inferences or presumptions of fact. When appropriate, the Court should apply this pronouncement. In other words, when the indirect evidence before the Court is compelling, a substantive finding on its part that the facts are established is called for.
- Shifting the burden of proof amounts to a procedural establishment of the facts, which should be done on the submission by the applicant of (no more than) prima facie evidence that they were personally pushed back
Not all meritorious applicants are in a position to bring sufficient evidence for their complaint to be considered proven beyond reasonable doubt, however, as the Court is well aware. This is why a jurisprudential route for shifting the burden of proof has long been established, with the forced disappearance case law having first opened the way, which has then been followed in other case law, including regarding pushbacks.
When taken, this jurisprudential route leads to the burden of proof being reversed so that it falls to the State to rebut the presumed veracity of the applicant’s account. If the state does not convincingly do so, the Court is to proceed by holding that the facts underlying the complaint have been established. The operation of the shift is subject to two conditions being met, which could be called the context condition and the linkage condition.
As to the first condition, there must exist a context which indicates that it appears plausible that the respondent State could have acted in the way claimed by the applicant. This is the case in a variety of situations, two of which are particularly relevant to pushback complaints: (1) when the respondent state is known to already have been acting in the way described by the complainant; (2) when the state refrains from sharing evidence it has or should have in its possession about the contested events, due to the state either not having recorded the events as it should have done as a matter of course, or not having investigated them properly after they happened.
In pushback complaints, both of these situations are likely to arise. Thus it will be rare that the first condition (context condition) for shifting the burden of proof would not be met in a pushback case.
The shifting of the burden of proof depends on a second condition being met as well: the linkage condition. This condition requires the applicant to show the plausibility that their particular complaint relates to the context which contributes to give rise to the presumption. The standard of proof to be applied to the linkage condition is the one known as prima facie (a Latin expression that can be translated as ‘at first appearance’ or ‘on the face of it’).
Accordingly, in pushback cases, next to the applicant’s statement, just one of the following pieces of evidence should be enough for the Court to consider the linkage condition fulfilled:
- a more detailed account from the applicant of what happened to them;
- a picture of the pushback operation taken against the odds, where the applicant appears;
- a witness statement emanating from a fellow migrant who testifies having seen the applicant being pushed back;
- a social media message recounting the event and the presence of the applicant in it, sent shortly after the pushback in question occurred;
- confirmation by whatever means of the applicant’s location during, immediately before or immediately after the pushback.
That no more than just one piece of prima facie evidence is needed for the linkage condition to be considered met, was confirmed by the Court in A.R.E v. Greece, where the Court stated (in our translation from the French, and with emphasis added):
The applicant provided a number of elements which, even taken separately, could constitute prima facie evidence in favour of her version of the events.
In other words, it is not the combination of the evidentiary elements that would have allowed the Court to conclude that the applicant had proven her case prima facie. Just one element would have sufficed for this.
It is crucial that the Court keeps to the proper meaning of prima facie when it considers whether to shift the burden of proof in pushback cases. This is for two distinct but mutually reinforcing reasons.
First, if the evidence submitted by the applicant of their involvement in a particular pushback is more than prima facie and the case lends itself to be found proven beyond reasonable doubt through presumptions and inferences, then this should be done, with the result that the evidentiary route of shifting the burden of proof should deliberately and expressly not been followed by the Court. This is because a substantive establishment of the facts is preferable to a mere procedural one, achieved ‘only’ because the state did not rebut the presumption which arose against it.
Second, if the Court misqualifies as prima facie a piece or bundle of evidence which is actually stronger than just plausible – thus being probable, compelling or even beyond reasonable doubt – this may wrongly be taken in future cases to indicate that prima facie evidence requires more probative value than it normatively does. In other words, the risk is that prima facie comes to be interpreted as denoting a higher standard of proof than it does. In addition to possibly affecting the resolution of the case in question, this would undermine the whole evidentiary system put in place by the Court.
- The role of the Court in alleviating the evidentiary inequality of arms between the applicant and the state
Clearly, the state-instigated secrecy that surrounds pushbacks creates a stark inequality of evidentiary arms between the applicant and the state.
In short, the state does all it can to suppress evidence, putting the applicant in a very challenging if not impossible position to prove their case. In addition, most pushback survivors do not have access to any legal representation, let alone access to counsels who have developed a specialist expertise in finding ways to seeking to prove pushback complaints. This makes it important for the Court to be particularly careful in the way it handles evidentiary issues in this case law.
The concealing practices adopted by states should persuade the Court to do all in its power to alleviatie the evidentiary inequality of arms between the parties. In particular, it should not only envisage shifting the burden of proof onto the state when the two conditions for this are met. It should also pro-actively request information from the respondent state, as it is empowered to do under Article 38 ECHR and Rule 44A of the Rules of Court. In pushback cases, it would, for example, often be very useful for the Court to request from the state the list of registrations of migrants at the relevant time and place, as well as administrative documents listing the coast guards and police officers in duty at the relevant time and place, their location and tasks. Information requests have binding effect on respondent states. Both the replies the state offers and the applicant’s comments on them should put the Court in a better position to establish the facts, possibly enabling it to do this substantively (thanks to the strong, concordant presumptions to which the information request process would have given rise) rather than just ‘procedurally’.
Summary
It is evident from the Court’s case law that European states conduct pushbacks and that pushbacks are secret and unofficial in nature. Due to states’ deliberate actions, pushback survivors have little to no direct evidence to offer the Court.
Occasionally a pushback complaint will nonetheless succeed in meeting the Strasbourg ‘beyond reasonable doubt’ standard of proof. In the total or near-total absence of direct evidence, this standard will generally be reached through strong, clear and concordant inferences and presumptions of fact.
It is crucial that in such a case the Court acknowledges that its findings of fact are reached substantially, on the application of a high standard of proof – thus without the mechanism of shifting the burden of proof having been activated or a prima facie standard of proof having been applied.
In most cases, however, pushback survivors are not in a position to bring sufficient evidence for their complaint to be considered proven beyond reasonable doubt. In such cases, after the applicant provides prima facie evidence of their involvement in the pushback in question, the burden of proof is to be shifted to the respondent state.
When examining whether the linkage condition for shifting the burden of proof is met, it is crucial that the Court keeps to the proper meaning of prima facie.
Given the stark inequality of evidentiary arms between the applicant and the state, the Court should more systematically request information from states and draw adverse inferences when the state refrains from submitting crucial evidence.
1Grażyna Baranowska is Professor of Migration Law and Human Rights at the Friedrich-Alexander-Universität Erlangen-Nürnberg and the Vice-Chair of the UN Working Group on Enforced and Involuntary Disappearances
2Marie-Bénédicte Dembour is Professor of Law and Anthropology at the University of Ghent and PI of the Horizon 2020 ERC project ‘DISSECT: Evidence in International Human Rights Adjudication’ (AdG-2018-834044)
3Isabel Kienzle is a doctoral researcher and research associate at the Chair for Migration Law and Human Rights at the Friedrich-Alexander-Universität Erlangen-Nürnberg. Her PhD project focuses on evidence in international human rights adjudication on pushbacks.