Merabishvili v. Georgia: Has the Mountain Given Birth to a Mouse?
The wait for those of us looking for much needed answers to understand what direction and coherence the Grand Chamber of the European Court of Human Rights would give to its nascent Article 18 case law (also known as ‘bad faith’ case law) has ended. A verdict has been reached in Merabashvili v. Georgia Grand Chamber judgment of the European Court of Human Rights. In a climate of retreat from human rights law and standards under the guise of domestic legalism, answers to the questions of what it means to violate the Convention in bad faith, how we prove it and what responses we owe to bad faith human rights violations have become pressing and urgent. The Grand Chamber gave us answers to the first two questions and passed on the third.
Article 18: The rediscovery of a dormant provision
Article 18 states:
“The restrictions permitted under this Convention to the said rights and freedoms shall not be applied for any purpose other than those for which they are prescribed.”
This article made its way to the Convention in recognition of the fact that states may abuse their power and use restrictions of rights to pursue illegitimate purposes and hidden agendas. The wording of the Article clearly prohibits such bad faith use of power by states. Yet Article 18 has been a dormant provision for much of the history of Strasbourg case law. The case law of the Convention operating under a structural good faith presumption downplayed the possibility of bad faith violations.
The structural good faith presumption was also reflected in how Article 18 was interpreted. The former Commission and the Court offered a very narrow reading of Article 18 by stating that it was not an autonomous article. It could only be raised in conjunction with a Convention provision that allows for restrictions and that a very high threshold of proof, either in the form of direct proof or strong circumstantial evidence, fell on the applicants to establish a bad faith violation (Merabishvili, paras 265-68). Furthermore, the Court often found it unnecessary to examine Article 18 when brought up in applications. When it did examine it, the Court never found a violation.
This was pretty much it until 2004 and Gusinsky v. Russia. The Court, moved by direct evidence proving that Mr. Gusinksy’s detention pursued the aim of coercing him to sell his media company to a state-owned one, wanted to say more than an arbitrary violation of Article 5. It did so by finding its first violation of Article 18 in conjunction with Article 5. Despite its novelty, Gusinsky was a surprisingly brief judgment. The Court simply found that the very high threshold of proof established for Article 18 had been met. There was a direct proof of bad faith, and this had to be spelt out.
In the handful of cases that followed, the Court continued its engagement with Article 18. These were all in connection with the pre-trial detention or incarceration of anti-government private actors, politicians, and civil society actors in Moldova, Russia, Ukraine, Azerbaijan and Georgia (Merabishvili paras 270-281). In these cases the Chamber judgments were very much focussed on the evidentiary questions for satisfying the high standards of proof for establishing bad faith. In two recent cases against Azerbaijan (Mammadov in 2014 and Jafarov in 2016), the Court has become more open to applicants proving bad faith not by presenting direct evidence, but also by drawing on contextual evidence (Merabishvili para. 281).
The number of new applications arguing for Article 18 violations is on the rise. They not only cover allegations of bad faith with regard detention, but also restrictions on freedom of assembly, freedom of expression, the right to fair trial and the freedom of movement.
Why was Merabishvili THE judgment to wait for?
This recent focus on Article 18 violations needed clarity and coherence around two important questions.
The first is the grand legal policy question of the very object and purpose of Article 18. We can add to this the reasons for reaching out to Article 18 in a Convention system that is structurally based on a strong good faith presumption.
In its case law prior to 2004, in many cases where the presumption of good faith might have been seriously suspect (the lack of effective investigations into killings and disappearances in the Turkish cases, for example) (Merabishvili para. 268), the Court showed no interest in entering explicitly into the territory of investigating bad faith as a matter of general policy. Instead, the Court’s good faith presumption case law found arbitrary restrictions of Convention rights post 2004. As a matter of general legal policy, we need to understand better what is distinct about finding bad faith violations as opposed to finding arbitrary violations.
Also connected to this is whether an Article 18 violation being found points to a particularly grave violation of the Convention — one that calls for specific individual and general remedies being ordered on the part of the Court and one that should automatically trigger mechanisms which had to the heightened scrutiny of the state in question by the political organs of the Council of Europe. Given that arbitrary violations and bad faith violations are not the same in nature and, therefore, should also not be in consequence, there has been a need for the Court to clarify what it means when it enters into Article 18 (or bad faith) territory.
Just weeks before Merabishvili, for example, the Committee of Ministers of the Council of Europe, in its Mammadov resolution, made the finding of an Article 18 violation by the Court its central reasoning to ask for the immediate release of opposition political Mr. Mammadov.
The second question concerned the standard of proof that needs to be met for showing that a state acted in bad faith. Given the strong presumption of good faith, the former Commission and the Court have established exceptionally high standards of proof for bad faith. The burden of proof for bad faith has been with the applicant and required that the applicant “convincingly show that the real aim of the authorities [have] not [been] the same as that proclaimed (or as [could] be reasonably inferred from the context)”.
In practice this meant that the applicant had to show both direct proof (for instance, a written agreement, or oral evidence) showing that the restriction not only lacked a legitimate purpose, but also that it had an illegitimate purpose. This is, of course, exceptionally difficult evidence for an applicant to furnish. While the Court allowed for reasonable inference from the immediate context of the restriction, it was not clear what type of contextual evidence counted. Thus what the Court should, in the absence of direct proof, take into account as evidence and whether it was reasonable in all cases to place the burden of proof on the applications, were central questions waiting for clarification.
The mountain and the mouse : The Merabishvili Grand Chamber Judgment
Despite the clamour and high expectation Article 18 being better clarified, this judgment of the Grand Chamber is, as we say in Turkish (and as I recently realised in many other Council of Europe languages, and in Latin, too), ‘the mountain that gave birth to a mouse’. The finding by nine to eight of an Article 18 violation in conjunction with Article 5 (with four concurring separate opinions alongside eight dissenting opinions) with respect to the detention of Mr Merabishvili begs more questions than it answers.
The Grand Chamber judgment does show awareness that the nascent Article 18 case law needs to be clarified (para 282). It undertakes a (brief) discussion of the travaux of the Convention (para 154), reaches out to European Union law (paras 155-156), Inter-American Court of Human Rights case law (paras 157-169), extradition law (paras 161-165) and French domestic administrative law (paras 166-168). The way it then synthesises what the future for Article 18 holds, however, makes for a puzzling read.
First, concerning the object and the purpose of Article 18 and how it fits with the Court’s structural good faith presumption, the judgment does not develop any general principles about when the Court turns to Article 18 and what is distinct about entering into Article 18 territory. In a rather understated fashion it states that Article 18 ‘is intended to complement the restriction clauses in the Convention.’ (para 283).
The judgment then proceeds to develop a novel approach to how Article 18 operates under a structural presumption of good faith. This approach starts by positing a ‘plurality of purposes presumption’ (para 292). According to the Court, the plural legitimate and illegitimate purposes can co-exist simultaneously in the actions of states when they restrict rights. What should trigger an Article 18 consideration is whether a bad faith purpose is a ‘fundamental aspect of the case’ at any given time (para 291). That is to say that a state’s restriction of a convention right may start with good faith (that is, a restriction to rights that can be acceptable from a Convention perspective), but this may then become a predominantly bad faith practice if evidence can prove so.
This new predominance test for bad faith is an unexpected turn for the future of Article 18. It complicates rather than clarifies the object and purpose of Article 18 by relativising bad faith, and, along with it, good faith. More importantly, the plurality of purposes presumption turns bad faith into a banal state of affairs. It normalises its occurrence so long as it is not a predominant reason for restricting rights.
According to the plurality of purposes presumption and the predominance test, whilst a state may have a little of bad faith in Time1, it may have a lot of it in Time 2. This also means that a little bit of bad faith (putting the difficulty of proving this to one side) would not be a violation of Article 18, but a lot of it – if it became predominant – would.
Perhaps then unsurprisingly, there is no discussion in the judgment about what distinct consequences should follow from the finding of an Article 18 violation as opposed to a violation of the Convention in good faith. Specifically there is no discussion in the judgment about remedial individual measures beyond compensation and no discussion of general measures.
On the second issue, that of the very high threshold of standard of proof, the Court offers a clear and, at first glance, progressive answer. It finds that there is no reason for the burden of proof test for Article 18 to deviate from the rest of the Convention (para 310 and 316). It, therefore, holds that the burden of proof does not need to fall exclusively on applicants and that the standard of proof should be ‚beyond reasonable proof’. The Court also clearly states that when investigating Article 18 violations it is empowered to examine any material irrespective of its origin and that it can obtain material on its own motion (para 311). This is a progressive development of the case law of Article 18, but one that also begs for some caution for the following reasons.
First, the Court holds that all evidence about Article 18 must be about “primary facts, or contextual facts or sequences of events which can form the basis of inferences about the primary facts” (para 317). This means that evidence about systemic decay in human rights protections, say for example, by sustained attacks on the independence of the judiciary, cannot constitute evidence of bad faith, if they cannot provide immediate inferences about restrictions of rights at a certain time and place.
Second, it is not clear whether equalising the burden of proof standards with the rest of the Convention is a net improvement when this is read together with the newly introduced predominant purpose test. The predominant purpose test means that if states can show more legitimate purpose than illegitimate purpose the Court cannot make a bad faith finding based on the mere existence of bad faith. Indeed, the eight dissenting judges in this case, whilst agreeing with this new novel theory, disagree that bad faith found at a certain time in the detention of Mr Merabishvili can be seen as predominant. Conversely, judges with four separate concurring opinions hold that the predominant purpose test is flawed, because any amount of bad faith should trigger a violation of Article 18 and holding a contrary view strips Article 18 of any distinct value. This disagreement shows that much of the future case law may host similar discussions amongst judges about whether bad faith is predominant or not.
What future for our mouse?
The Grand Chamber had two big questions to settle in this case. This first was of the very purpose of delving into and declaring an Article 18 violation of the Convention and the second was the standard of proof and the kinds of evidence required for showing illegitimate purposes for restrictions of rights.
In a much divided Grand Chamber judgment (only five out of seventeen judges agree with both the reasoning and the outcome), the Court relaxed the high burden of proof standards for establishing bad faith, but introduced a novel and overarching requirement asking for bad faith to be the predominant motive in rights restrictions. It also rejected the relevance of contextual evidence that points to a systemic undermining of domestic rights protections by governments. In future cases this may effectively mean taking what the Court gives with one hand with the other. If one were pessimistic, one may even say that the Court has helped actors with bad faith to conceal it better.
Many thanks indeed Basak for your insightful piece, as ever. I agree with you about your concerns as to how the ‚predominant purpose‘ test will be applied in future cases.
I have written a piece on the case (having acted as the applicant’s legal representative) for Open Democracy: https://www.opendemocracy.net/od-russia/philip-leach/georgia-strasbourgs-scrutiny-of-the-misuse-of-power
Also pending in Strasbourg is the Grand Chamber case of Denisov v Ukraine, in which Article 18 has been raised in conjunction with Article 6 (an issue on which the European Court judges have been divided to date).