Today the Grand Chamber of the European Court of Human Rights held joint hearings in Al-Skeini and others v. UK (no. 55721/07) and Al-Jedda v. UK (no. 27021/08) – webcast available here, statements of facts available here. It would be no exaggeration to say that these are some of the most important cases to come before the Court in recent years, with possibly wide-ranging implications, on matters ranging from the extraterritorial application of the ECHR and the use of force generally, to occupation and targeted killings, up to the responsibility of international organizations, the relationship between the ECHR regime and the UN Security Council under Article 103 of the Charter. The Court will probably deliver its judgments by the end of the year.
Let me now try to provide a preview of some of the most important issues – particularly threshold issues – that that the two cases raise, and of the possible ways in which the Court might rule.
(Again, apologies for a long post!)
Al-Skeini in a nutshell
The case has six applicants. Five of them where killed, or were allegedly killed, by British troops on patrol in UK-occupied Basra (the facts are somewhat more complicated, and one of the applicants before the European Court is not the same as before domestic courts, but this description is accurate enough). The sixth applicant, Mr. Baha Mousa, was arrested by British troops and taken to a UK detention facility, where he was mistreated and ultimately killed. The applicants’ families asked for a full, independent and effective investigation, compliant with Art. 2 ECHR.
The threshold question is whether the applicants were even protected by the ECHR, since the killings took place outside the UK, i.e. whether they were within the UK’s jurisdiction under Article 1 ECHR. (An additional question, to which I will not be devoting any attention, is whether Baha Mousa remains a victim under the Convention, due to remedial action already taken).
There are two main strands of jurisprudence dealing with the extraterritorial application of human rights treaties generally, and the ECHR specifically. First, there is what I will call the spatial model of jurisdiction – that a state possess jurisdiction whenever it has effective overall control of an area. This is the model applied by the European Court in Loizidou, as well as by the ICJ in Wall and Congo v. Uganda. Second, there is the personal model model of jurisdiction – that a state has jurisdiction whenever it exercises authority or control over an individual. This model has been applied in numerous cases before the European Court and the now defunct Commission, e.g. in Issa, as well as by the Human Rights Committee.
The relationship between these two strands of the case law has never been clear, as the whole question of extraterritoriality was never approached in a methodical way, but even what little clarity existed was blown away by the Court’s decision in Bankovic, holding that persons killed outside a state’s territory by an aircraft where not within the state’s jurisdiction, and more generally that extraterritorial application can only be exceptional, and supposedly has to be justified by reference to general international law, for example with regard to embassies. (In my view at least, the Bankovic approach to jurisdiction is at the very least methodologically and conceptually incoherent – see more here).
In Al-Skeini, the House of Lords held that the sixth applicant was within the UK’s jurisdiction, while the other five were not. Applying Bankovic, the House held that:
(1) The spatial model of jurisdiction does not apply outside the espace juridique of the ECHR – a concept introduced but not explained by the European Court in Bankovic, designating the combined territories of ECHR member states. In other words, though Turkey had ECHR obligations to the people of Norther Cyprus because it exercised effective overall control over that area, this was only so because Cyprus was an ECHR state party. According to their Lordships, that reasoning did not extend to the UK in Iraq, because the ECHR is a regional instrument, the imposition of which in Iraq would amount to ‘human rights imperialism.’
(2) Even if the spatial model did apply in principle, the UK had no effective overall control over Basra, despite the fact that it was the occupying power in Southern Iraq, since the strength of the insurgency and the low level of its forces rendered it factually unable to guarantee ECHR rights.
(3) Whatever the validity of the personal model of jurisdiction, Bankovic was clear on the point that a mere killing would not suffice for it to engage. Therefore, the first five applicants were not within the UK’s jurisdiction.
(4) However, the sixth applicant, who was detained in a UK facility and killed there, was in fact within the UK’s jurisdiction, because a military prison has a special status in international law, akin to an embassy. The government conceded that jurisdiction attached on this basis.
On point (1), their Lordships’ reasoning on the spatial model and the espace juridique point has been heavily and rightly criticized – see especially Ralph Wilde’s case note on Al-Skeini in the AJIL and his article on the concept of the ECHR’s legal space in the European Human Rights Law Review, as well as Tobias Thienel’s article in the JICJ. Technically, the House of Lords put the espace juridique concept to a much more radical use that the European Court did in Bankovic. Likewise, as Judge Bratza pointed out in a question at the hearings, Issa at least is directly contrary to the proposition that the spatial model can only apply within ‘Europe’, and so is probably Ilascu. More fundamentally, if universality truly is the foundation of human rights, why should it matter that the ECHR is a regional treaty for the purpose of its extraterritorial application, particularly when Article 1 does not say so? Jurisdiction either means control of a territory, or it does not.
On point (2), the issue of whether the belligerent occupation threshold of effective control and the human rights, jurisdiction threshold of effective overall control is a complex one, on which reasonable people can disagree. The threshold should in either case be met only when the obligations imposed could be realistically complied with. In that regard, the English courts have in my view underestimated the flexibility inherent in the positive obligation to secure human rights under Article 1 ECHR, which requires states to do only what they reasonably can, and have thus exaggerated the adverse implications that considering the two thresholds to be the same could potentially have.
On point (3), I think it is fair to say that Bankovic does preclude the application of the personal model to the first five applicants (although they argue that it does not) – but that does not make it any less wrongly decided.
But how are we then to explain the fact that the sixth applicant was within the UK’s jurisdiction? The justification given by the House of Lords – that a military prison is analogous to an embassy – to my mind simply defies common sense. A prison is in no way like an embassy, as I imagine any prisoner would be able to attest, if we say invited him to a cocktail party at an appropriate locale in Belgravia. It most certainly doesn’t have any ‘special status’ in international law.
The only thing common to a prison and to an embassy are that they both operate on the basis of the territorial state’s consent. But if this were the reason why the extraterritorial state’s jurisdiction was to exist, not only is there no justification given for why this would be so – one would imagine that non-consensual interventions would generally be more likely to affect the human rights of the population – but consent can be given to many things, like to the presence of foreign forces in general.
What in my view explains the result of Al-Skeini are not the intricacies of the concept of jurisdiction in Art. 1 ECHR, real or imagined, but the tensions in the policy considerations underpinning the law. On the one hand, like the European Court in Bankovic, the English courts in Al-Skeini did not want to open the floodgates of litigation by considering every individual against whom force was used as falling under the protection of the Convention. They did not want to micromanage the use of force in the field, especially when some of the killings in question may even have been justified. On the other hand, however, nothing could have justified the mistreatment and killing of a defenseless prisoner. Baha Mousa simply had to be protected – and this is where the prison somehow became analogous to an embassy.
Al-Skeini: options before the Court
Compared to the House of Lords’ judgment, the Court could do better or worse, and much depends on just how many issues the Court will reach.
Thus, from the standpoint of the spatial model, the Court could do the following:
(1) Affirm that the effective overall control of an area conception of jurisdiction applies outside the ECHR’s espace juridique, as the Chamber had done in Issa, or, conversely, agree with the House of Lords that the spatial model extends only to the territories of the ECHR states parties – certainly the worst possible result;
(2) Assuming that the spatial model does apply, the question then would be whether Basra was under the UK’s effective overall control. The Court could either say that the UK’s status as an occupying power necessarily meant that it was in control of Basra, despite all the difficulties that it had encountered, and that hence all six applicants were within the UK’s jurisdiction, or it could agree with the English courts that Basra was not under the UK’s effective control;
(3) A finding that Basra, was, in fact, an area under the UK’s jurisdiction would dispose entirely of the preliminary question of extraterritorial application. If, however, the Court were to find that Basra was not under the UK’s control, it then may decide to apply the attenuated version of the spatial model as control over places in order to bring Baha Mousa under the UK’s jurisdiction upon his detention, as it did in Al-Saadoon. Doing so, however, would leave outside the ECHR’s scope any events that took place between Baha Mousa’s arrest and his ultimate transfer to the UK detention facility, as well pointed out both by the applicants in their argument, and by Judge Ann Power in a very probing question, in which she asked the government to explain the basis for their concession that Baha Mousa was protected by the Convention.
The possibilities of the spatial model would thus be exhausted, and the Court would then have to examine the case from the standpoint of the personal model:
(1) It could reject the validity of the personal model altogether, or narrow its application to some arbitrarily selected exceptional circumstances, and thus exclude all of the applicants from the UK’s jurisdiction, unless it opted to fiddle with the spatial model to protect Baha Mousa, as above; in that regard, as a matter of policy, the problem with the personal model is precisely that it cannot be limited by reference to any non-arbitrary criterion – it simply boils down to the proposition that the state has the duty to respect human rights whenever it has the capacity to violate them;
(2) In a similar vein, it could adopt the approach of the English Court of Appeal, say that the personal model requires physical custody, and that accordingly only Baha Mousa was within the UK’s jurisdiction; inter alia, it could use its recent Medvedyev gloss of Bankovic as applying only to ‘instantaneous extraterritorial acts,’ which a killing is but detention is not, to justify this result;
(3) Finally, it could radically depart from Bankovic, and say that the five applicants killed by UK troops on patrol were also under the UK’s control, and accordingly within its jurisdiction. This would in my view be a welcome, but not very likely development, for the reasons given above.
The Court may of course very well surprise us. Whatever it does will inevitably be a consequence of how it perceives the tension between universality and effectiveness on the particular facts of the case, and on their broader policy implications, especially with regard to the personal model.
Al-Jedda in a nutshell
Turning now to the other of the two ‘Al-‘ cases heard today, Al-Jedda adds a whole new layer of complexity – or actually two or three. At first glance, the jurisdiction issue is the same as with Baha Mousa in Al-Skeini, and would accordingly be covered by the UK government’s concession, any broader principle notwithstanding. Though Al-Jedda was a shady character detained by British forces in Iraq, he was NOT detained under the law of occupation, nor on a criminal charge in pre-trial detention, as in Al-Saadoon. He was rather detained under the authority granted to the US and UK to detain preventively by the UN Security Council, in Resolution 1546. On the other hand, absent a derogation Article 5 ECHR does not allow for such preventive security detention.
This gives rise to two lines of argument:
(1) First, relying on Behrami, the UK argues that the acts of its soldiers in Iraq, which were there under UNSC authorization, were subsequent to that authorization NOT to be attributed to the UK, but to the United Nations. Accordingly, if the acts of UK soldiers in Iraq were not attributable to the UK, then the UK could not have exercised Art. 1 jurisdiction over Al-Jedda, no more than Turkey could have exercised jurisdiction in Northern Cyprus if it had no soldiers of own controlling the territory.
(2) Second, even if Al-Jedda’s detention was attributable to it, the UK argues that its detention authority prevailed over the Article 5 ECHR prohibition of preventive detention, on the basis of Article 103 of the UN Charter.
In my own view and that of my co-author, Tatjana Papic, in the view of practically every other author (see, e.g., Kjetil Larsen’s EJIL piece), and most importantly, that of the International Law Commission (see here, at 10), the first of these arguments must be wrong, because Behrami itself was wrongly decided. When Al-Jedda came before the House of Lords, however, their Lordships were somewhat more equivocal. They too thought that it defies reality and common sense to say that the acts of US and UK troops in Iraq are actually attributable to the UN, but they wouldn’t be so undiplomatic as to say that Behrami was wrongly decided. They therefore (in my view, quite unpersuasively) distinguished it on the facts. They then went on to find that the UNSC resolution did in fact prevail over Art. 5 ECHR by virtue of Art. 103 of the Charter, and that its prohibition of preventive detention was accordingly displaced or qualified.
Now, before the Grand Chamber, the UK government has again raised its attribution argument, claiming that on this matter Al-Jedda is indistinguishable from Behrami. And on this they are quite correct – the only difference between Iraq and Kosovo is the existence of UNMIK, a UN civilian presence. The military missions, however, are exactly the same. KFOR is subjected to no more UN control than the UK and US troops in Iraq.
On the other hand, rather than arguing that Behrami was wrongly decided, for obvious reasons of strategic litigation the applicants now maintain that the House of Lords’ distinguishing of Behrami was entirely persuasive, all the while pretending that there nothing wrong with Behrami in the first place, for example by treating it as being perfectly consistent with the ILC’s DARIO, when the ILC itself and its special rapporteur quite openly reject it. In truth, this adds an air of unreality, or a whiff of the Emperor’s new clothes, to the whole proceedings.
On attribution, the applicants do quite rightly point out that saying that the acts of UK forces in Iraq were attributable to the UN would be inconsistent with the European Court’s decision and judgment in Al-Saadoon (this was also pointed out by Francesco Messineo in the comments to this post).
There is a further absurdity here, arising from the UK’s own argument for the invasion of Iraq in 2003 – that it was implicitly authorized by the UNSC to do so – and its present claim that all UN authorized actions are attributable to the UN only and exclusively, rather than to the states which actually carry them out. This would simply mean that our eyes were deceiving us back in 2003 – it was actually the UN which invaded Iraq!
Now, assuming that attribution is somehow out of the way, what of the apparent norm conflict between UNSC Res 1546 and Art. 5 ECHR? International constitutionalists should rejoice! Citing Prof. Rudolf Berndardt in the Simma Charter Commentary, the UK has now argued that the UN Charter is the constitution of the international community, which prevails over other norms.
The applicants, on the other hand, emphasized the European Court’s own qualification of the ECHR as the ‘constitutional instrument of European public order,’ and the ECHR as a self-contained regime, which could give way to other norms only through a derogation. Not only was UNSCRes 1546 an authorization, rather than an obligation, under the Charter, which would render Art. 103 inapplicable, but even if it did apply the Security Council couldn’t just extinguish the ECHR on a whim – the applicants thus raised the specter of the UNSC requiring states to impose the death penalty on terrorists, thereby displacing the ECHR.
In their view, the Court could rely on the ECJ’s decision in Kadi, and say that UNSC resolutions could not affect human rights protections under the ECHR as far as the ECHR itself is concerned. Or, the Court could indeed rely on its own decision in Bosphorus, and say that UNSC resolutions could potentially only displace the ECHR if the UN provided equivalent protection of human rights, which it obviously does not in this particular instance.
Both of these avenues are in my view deeply problematic. The first because the basis of Kadi is the ECJ’s belief that EU law is an independent legal system, which international law can penetrate only on its own (dualist) terms. For the ECJ, the situation was no different than if a domestic constitutional court reviewed domestic legislation against its own constitution. But even if the EU legal system is now exactly that (and I don’t buy that entirely), the ECHR is most emphatically not. As for Bosphorus, the equivalent protection method of norm conflict avoidance can work for norms which are hierarchically equal – in that case, norms binding states under the ECHR and under EU law. But it logically cannot apply if one of the norms – here Art. 103 of the Charter – is hierarchically superior. An inferior norm simply cannot condition when the superior one is to apply.
(For more on all theses issues, see my norm conflict and human rights piece in the Duke JCIL; see also here).
Al-Jedda: options before the Court
So, what can the Court itself do? Logically, it first has to deal with the issue of attribution. If it finds that Al-Jedda’s detention was not attributable to the UK, the case is dead that very instant. If not – we might hear more about how the Court sees itself and the Convention within the fragmented international legal order, and may even witness a clash of competing constitutional visions. Thus:
(i) The Court can apply Behrami, and rule that the UK’s acts were attributable to the UN, to predictably massive outcry;
(ii) It can ‘distinguish’ Behrami, just like the House of Lords (probably the most likely result);
(ii) It can actually overrule Behrami, to the great joy of many, but that would not be very likely;
(2) Norm conflict – if option (1)(i) was off the table
(i) The Court can say the UK had no obligation under UNSCR 1546, and that accordingly Art. 103 of the Charter does not apply, or that the resolution and Art. 5 ECHR could somehow be interpreted harmoniously (though they truly cannot);
(ii) It can pull a Kadi, and declare independence from international law, by saying that the ECHR is an autonomous legal order which is unconcerned with Art. 103;
(iii) It can employ Bosphorus, but under the circumstances that would only be a marginally more diplomatic way of doing (ii);
(iv) It can, like the House of Lords, acknowledge the primacy of the Charter and the Security Council, and have the resolution allow security detention under Art. 5 – but that would open the doors for possible abuse in the future.
Whichever way you look at it, the choice that the Court is facing here is not easy. I would say that options (i) and (iv) are more likely, but I may well be wrong. And, unlike the attribution issue, this is certainly not a question capable of a ‘right’ or ‘wrong’ answer in objective terms.
Whatever the Court does, these will be some of the most interesting judgments that it has ever delivered. Not only do both cases pose serious legal and technical questions, they also lead us to ask deeper, more fundamental questions, such as what is actually a ‘constitution’ or an independent legal order. Likewise, both cases pose serious difficulties for the Court, both as a matter of law and as a matter of policy. To a large extent these difficulties are of the Court’s own making, particularly because of its decisions in Bankovic and Behrami. In all fairness to the Court, however, no other court would find them any easier. Now all we have to do is wait and see.