Thanks to Emma Várnagy whose insightful post on the P.H. case published on the Strasbourg Observers blog inspired me to write my own take on the case, and whose comments on my text helped improve it.
Would you jump from a 7.7 metres high window? Of course not! Nobody would, although, according to the European Court of Human Rights (Court), a Roma teenager did. Reading the unanimous Chamber judgment in P.H. v Slovakia is like being transported in the reality which inspired Nobel-Prize winner Dario Fo to write his most celebrated play Accidental Death of an Anarchist. Not only are absurd explanations proffered by the defendant state in its submissions to Strasbourg, but these are then accepted by the Court who finds nothing to object to the idea that the applicant jumped from a second-floor window in order to escape from the police station from which she was about to be … (hold your breath) released.
This post is divided into four sections. The first section highlights the absurdity of the escape attempt scenario: only cartoon characters believe they may be able to continue running away after a jump of that height. It also reviews the finding of a violation of the procedural limb of Article 2 ECHR (right to life), which the absence of domestic investigations into the incident had made more or less inevitable. The second section observes that by shifting the burden of proof, the Court could have required Slovakia to provide a convincing explanation for the fall, in the absence of which a violation of Article 2’s substantive limb would have arisen. The Court, however, refrained from doing this. The third section explains that the Court did find a violation of Article 2’s substantive limb, but through a vacuous and illogical route which fully endorses the state’s absurd script of the events. The fourth section notes the by-then-to-be-anticipated rejection of the applicant’s complaints regarding Article 14 ECHR (prohibition of discrimination).
In sum, the judgment continues to add to the abundant Strasbourg jurisprudence negating a discriminatory motive in the abuse of Roma people at the hands of the police (see here, here and here). Whilst the applicant had escaped life-threatening injuries thanks to a tree branch which broke her fall and a layer of snow on the ground, the state got its own escape, so to speak, thanks to the Court and the jurisprudential routes it adopted.
1. 7.7 metres high: did the applicant jump or was she pushed, either physically or metaphorically?
The applicant in this case was a young Roma woman, aged 16 at the time of her fall and with minor mental impairment. (I shall not comment on this latter aspect but see this excellent post on the Validity blog). The Slovakian government’s story of what happened goes like this: P.H. had been caught red-handedly stealing in a shopping mall. The police were called, P.H. did not resist arrest and was taken by car to the police station. Her theft having been petty, it was decided she would be released shortly pending criminal proceedings. She asked to go to the toilet and was escorted by police officer A to a bathroom on the second floor of the police station. There Officer A turned his back on her. Next thing he knew, she had fallen from the window and laid on the ground 7.7 metres below.
No investigation into the incident was conducted, except for disciplinary proceedings against Officer A for having failed to discharge properly his duty to prevent escape, for which his salary was reduced by five per cent for one month. In these proceedings, P.H. was treated as a witness to Officer A’s offence, not as a victim. As a result, her brother – being the brother of a witness – was not heard, even though he too had been brought from the mall to the police station (in a different car) and he said he had heard his sister arriving there in an extremely upset state. When she woke up from a one-month long coma, P.H. did not remember anything about the day of the fall except, she said, for having been verbally abused and slapped by the police during the transport to the police station. No expert was heard about the question of whether P.H. might have been able to remember accurately what had happened during the car journey. Not even the police officers who had been in the car with her, let alone others, were questioned. There had been no reconstruction of the incident despite Officer A having given contradictory statements about what he had seen of the fall.
The Court was not impressed by the inaction of the defendant state which had done nothing to clarify the way the incident had unfolded, thereby giving the impression of only seeking to give the incident ‘administrative closure’ (paragraph 106). The Court unsurprisingly found a violation of Article 2 in its procedural limb.
Despite this finding, the judgment repeatedly endorses the description of the fall as an escape attempt. A more reasonable approach would have been to accept that the fall had happened in suspicious circumstances, which might have entailed the taunting or worse by Slovakian police officers. Otherwise, why would P.H. have jumped or what would have caused her to fall? Astonishingly, the judgment is written as if the Court never considers that this question is at the core of the case.
2. Why the right questions should have made the Court shift the burden of proof onto the state
The applicant had no recollection of the fall and could not explain, let alone evidence, the exact details of what had happened to her. This, however, was no legally valid reason for the Court to reject her complaint regarding the substantive limb of Article 2, whose violation should have been found through a correct understanding and apposite application of the Latin adage ‘Actori incumbit probatio’.
The adage, which remains an important foundation of the law of evidence, can be translated as follows: the person who raises an allegation must prove it. Contrary to what a superficial reading of it could suggest, it does not mean that in international human rights law the burden of proof necessarily falls on the applicant whose complaint is to be rejected if she is unable to prove it. The point of the adage is that any party to a case (not just the complainant) is expected to prove their allegation.
Thus, in P.H., the Court could have ruled as follows: the applicant has proven her allegation that she was at the police station and that she experienced a life-threatening event. In response, the defendant state has alleged that her fall was an escape attempt. However, the state has failed to prove this allegation. The Court finds the applicant’s allegation proven to a sufficient standard and declares the complaint regarding Article 2’s substantive limb established.
Admittedly the Strasbourg Court is rarely inclined to apply the Latin adage in a way which involves, right from the start, a distribution of the burden of proof between the parties. A longer reasoning about the adage, however, would also have ultimately driven the Court to shift the burden of proof onto the defendant state. This is because it is widely recognised in international human rights law that there exist a range of circumstances where a human rights court should free the applicant from the burden of proof and place it onto the defendant state.
Christopher Roberts has identified three situations where this typically happens:
- The applicant’s complaint fits a clearly established pattern of human rights violations;
- Knowledge about the incident at the basis of the applicant’s complaint rests exclusively with the defendant state;
- The complaint relates to an injury sustained by the applicant whilst under the control of the state, without the cause of this injury clearly not being attributable to the state.
The common thread running through these situations is that the state has some important explaining to do in each of them. If the state fails to come up with a convincing explanation, the presumption arises that the applicant’s complaint is valid, thereby giving the court a legal ground for declaring the violation complained of but not fully proven by the applicant established.
Any single one of the scenarios highlighted by Roberts is theoretically sufficient to trigger the shift in the burden of proof. Strikingly, in P.H. v Slovakia, all three were at play:
- P.H.’s fall could be connected to rampant antigypsyism in Slovakia;
- Knowledge about the precise circumstances of the fall was or should have been in the hands of the defendant state which, instead, did nothing to try to unearth it;
- The fall had occurred when the applicant was under the control of the state -or one could even say, given both her young age and mental impairment, in its care- and the state had only offered an absurd explanation for it.
Especially in the current state of the Court’s jurisprudence regarding racial discrimination (more on this below), the first point had virtually no chance of being adopted. The rejection of the other two points, however, is surprising given the total absence of effective investigation on the part of Slovakia and the lack of any ‘innocent’ cause explaining the fall, with no suggestion, for example, that P.H. might have been depressed, suicidal, otherwise of unsound mind or not in control of her actions.
To summarise, even leaving aside the ‘short’ route of directly distributing the burden of proof between the parties and the ‘longer’ route of justifying a shift in the burden of proof by reference to an established pattern of violations, the Court could and should have found a violation of the substantive limb of Article 2 on the grounds either that the applicant had sustained a life-threatening fall whilst in the hands of the state with the state having offered only an implausible explanation for the fall or/and that the state had been the party best placed to produce knowledge about the incident and had not done this
3. The Court’s finding of a violation of Article 2’s substantive limb: illogical and vacuous
Although the Court did find a violation of the substantive limb of Article 2 in P.H., this was achieved through an unsatisfactory reasoning which comes across as the opposite of a vindication of the applicant’s complaint. In short, the Court completely endorses the version of the fall presented by the state – no questions asked .
The Court starts by rehearsing at paragraph 111 evidentiary principles which are now well-established in its Article 2 jurisprudence. It cites a long passage from the Fanziyeva judgment, thereby recalling:
- the vulnerable state of persons in custody, triggering a suspicion when they unexplainably die;
- the positive obligation on the state to take protective action when a detainee’s life is at risk;
- the application by the Court of the standard of ‘proof beyond reasonable doubt’, which can nonetheless be achieved through inferences and presumptions of fact;
- the fact that the burden of proof may be regarded as resting on the authorities if exclusive knowledge about the events in issue rests with the authorities.
These principles could have been expected to favour the applicant in P.H. They did not, due to the inapposite way the Court applied them.
Turning to the facts of the case, paragraph 112 starts by reiterating that P.H.’s injuries were caused by her fall. It adds that ‘there has not been any allegation of any direct involvement of any third party in it’. This statement is crucial, for it suggests that there was nothing untowardly about the fall, no external element to it, and therefore no reason to shift the burden of proof. Importantly, however, the statement fails to make justice to the submissions of the applicant, as summarised by the Court itself at paragraph 86. To quote: ‘The applicant stated it was unclear whether she had fallen by herself. Assuming that that was the case, there must have been a very good reason for her to try to escape in such a way and the State had failed to prevent it. As it was possible that the treatment she had been exposed to in the police car had continued at the police station, it was likewise possible that such treatment was the reason for her attempt to escape’ (emphases added). Read this as you may, clearly, the applicant had not excluded that the fall could have been caused by a third party. Astoundingly, the Court ignores this.
At this stage, it is worth going into what the applicant and the Court said about the car journey. The applicant had complained of ill-treatment contrary to Article 3 having happened in the car -and which might have continued at the police station and driven her to jump. Early in the judgment, the Court had dismissed this complaint. This was in my view entirely reasonable, given her recollection of ill-treatment in the car was only supported by her brother saying that she had arrived at the station highly disturbed. What is not acceptable, however, is that the Court uses the rejection of the Article 3 complaint to reject the Article 2’s substantive complaint. The beginning of paragraph 112 has already been quoted. It continues: ‘an allegation of ill-treatment [in the car] that should have preceded the fall and in fact have given a cause to it has not been found established to the requisite standard’. Such a formulation keeps the burden of proof squarely on the applicant. This is wrong given the 7.7 metre fall remained unexplained, and it was for the state to explain its cause. That ill-treatment in the car had not been considered proven by the applicant beyond reasonable doubt should have been neither here nor there as far as the Article 2 complaint was concerned. It certainly should have provided no excuse for the state to be relieved from the obligation to explain the cause of the fall. I am personally agnostic as to whether any ill-treatment happened in the car, but whether it did or not, the fall from the window is highly suspicious.
Admittedly, the state did provide an explanation for the fall, namely, the fact that P.H. had attempted to escape. The problem, thus, becomes that the Court accepts this explanation. Accordingly, paragraph 113 concedes that precautions should have been taken by the state to protect life (but without acknowledging that danger to life in this case may have arisen from outside the scenario of an escape attempt on the part of the applicant). Continuing with the idea that the fall was an escape attempt, paragraph 114 observes that there had been a duty to watch detainees in order to prevent their escape; domestic proceedings had established that Officer A had failed to discharge this duty. It is this failure by Officer A upon which the Court relies for finding, in paragraph 115, that Article 2 had been violated in its substantive limb.
The finding is problematic at multiple levels. First, it is illogical within the confines of the judgment, in that it contradicts the Court’s earlier finding that there had been a violation of the procedural limb of Article 2. From having remonstrated the defendant state for having failed to attempt to clarify in any way what had happened, now the sequence of events is presented by the Court as if it had been clearly established: P.H. attempted to escape; Officer A had failed to prevent her escape.
Second, the finding is illogical at a more systematic level, in that it goes against the subsidiary character of the Strasbourg system of human rights protection. The system’s whole point is to incentivise states to do the right thing at domestic level, hopefully making an intervention by the Court unnecessary. In the present case, one cannot help thinking that had the Court really believed the fall to have been the result of Officer A failing to prevent an escape by the applicant, it would have had no reason to find a violation of the Convention, since the state had itself corrected its failing by pursuing disciplinary proceedings (which the Court nowhere describes as unsatisfactory) against Officer A. On the facts of the case as presented by the Court, the matter should have been found to have been handled properly at domestic level rather than being the object of a violation of the Convention.
Would the Court have been uncomfortable with its own endorsement of the defendant state’s version of events and would it have wanted to ‘give’ something to the applicant by finding a violation of the substantive limb of Article 2? If so, it needed to properly acknowledge what had happened to the applicant.
As it is, and it is the third and possibly most objectionable result from a human rights perspective, the Court’s acceptance of the state’s absurd scenario makes a mockery of the applicant’s submissions, thereby emptying the finding of a substantive violation of Article 2 (no less!) of any real substance. Nowhere in this part of the judgment does the Court’s reasoning capture the seriousness of what was at stake. Instead it simply rehashes the version of the defendant state according to which P.H. had tried to escape and an inattentive police officer had inadvertently and unfortunately failed to stop her escape attempt. A finding of violation of the substantive limb of Article 2 is normally immediately understood to be extremely serious; in this particular case it is vacuous.
4. The rejection of the Article 14 complaints: disingenuous
In my reading of the judgment, the applicant almost lost her life due to actions which are ultimately attributable to the defendant state. And the only explanation I can see for this is that she was targeted either as a Roma or as a person suffering from a mental impairment or both. In my view, it is hardly possible to explain what happened to her without bringing in a discrimination motive. At the very least, the state has not explained the fall convincingly outside a discriminatory framework. The Court, however, refrains from even declaring the Article 14 complaints admissible, let alone shifting in their regard the burden of proof onto the state.
The judgment notes that a third-party intervention by ERRC had ‘argued that police services in Slovakia were contaminated by institutional “anti-Gypsyism”, and that Romani women with mental disabilities were a particularly easy target for abuse by the police’ (paragraph 116). The Court does not consider the report, however, strong enough evidence for accepting that the Article 14 complaints deserve an examination on the merits. It rejects them as manifestly ill-founded under Article 35 of the Convention, thereby implicitly indicating that in its view the applicant has not even succeeded in making a prima facie case of racial discrimination.
This rejection is justified by the Court in two short paragraphs, whose central arguments run as follow: ‘as no ill-treatment in the police car has been established, no question of discrimination in that treatment arises’ (paragraph 118) and ‘the material in [the Court’s] possession discloses no appearance that discrimination on the basis of origin or mental disability played any role in the ensuing investigation being ineffective’ (paragraph 119). The first argument makes sense, the second only holds because the Court has refrained from shifting the burden of proof.
Twenty years ago, Judge Bonello scolded the Court for a jurisprudence which, despite recurrently acknowledging that members of vulnerable minorities had suffered violations of Articles 2 and 3 of the Convention, persistently failed to link these most serious human rights violations to the ethnicity of the applicants (dissenting opinion in Anguelova v Bulgaria, paragraph 3). P.H. is unfortunately a testimony that not enough has changed since then.
In a never-ending circle, the Court refuses to shift the burden of proof onto the state because it sees not one ounce of evidence of discrimination in the case to begin with; thereby it puts itself in no position to find racial discrimination established, given the burden of proof has not been shifted. One step leads to the other, and vice-versa. Does one need to observe that this vicious circle reasoning does not bode well for human rights protection in Europe?
 See, however, as a counter-example where the ECtHR distributes the burden of proof, e.g., J.K. and Others v Sweden, paras 91-98. The African Court of Human and People’s Rights commonly distributes the burden of proof between the parties.
 Roberts acknowledges that the above operation may occasionally lead a state to be wrongly found to have abused human rights. Though reasonable, the presumption will not always be correct in practice; it is, after all, only a presumption. This, however, has no serious negative impact from a human rights perspective. The state is not criminally sanctioned. At most, the undeserved finding of a violation will have prompted additional action for the protection of human rights. By contrast, failing to apply the presumption in a situation where the state was in fact abusing human rights, but without the victim having had the capacity to prove this, is highly damaging not only to the individual victim but also to society at large.
 Várnagy’s post documents how, throughout the judgment, the Court tends to accept the allegations of the state and to dismiss those of the applicant.