By Tobias Thienel

This morning, the Grand Chamber of the European Court of Human Rights decided the case if Saadi v. Italy, concerning the expulsion of a terrorist suspect to Tunisia. In argument before the Court, the Italian government and the UK government as intervener (Article 36(2) ECHR) had submitted to the Court that the protection from expulsion to a risk of torture or inhuman treatment in the receiving State (as stated, most prominently, in Soering v. United Kingdom, paras. 85-91) should not be absolute. Instead, so the argument particularly of the United Kingdom went, the risk of grave ill-treatment in the receiving State should be balanced against the danger the person concerned poses to national security. This submission meant that the UK (and Italy, which joined the British submissions in this respect) had to mount a full-blown attack on the Grand Chamber judgment in Chahal v. United Kingdom. There, the Court had held that the protection from expulsion under Article 3 was absolute, and that the behaviour of the person concerned was therefore nothing to the point, even if the person was a terrorist. The UK now argued that the Court should depart from Chahal (as clearly it could have done: Stafford v. United Kingdom, para. 68). Chahal, the UK said, had created serious difficulty for States in their fight against terrorism, in that it was no longer possible to protect society from foreign terrorists where they a) could not be deported (Chahal) and b) could not be tried due to a lack of evidence proving the commission of crimes beyond reasonable doubt. The UK further sought to deflect the force of the argument drawn from the absoluteness of Article 3: it submitted that in expulsion cases, the actual violation of the terms of Article 3 would happen in the receiving State, and not at the hands of the expelling State. The prohibition of expulsion in such a case was a positive obligation, added to the basic, negative and completely absolute prohibition of torture or inhuman treatment by the State itself. Positive obligations, the UK said, were not absolute. As might have been expected, the Court was having none of it. In a completely unanimous judgment, it reaffirmed that Article 3 ‘enshrine[d] one of the most fundamental values of democratic societies’ (its early case-law only spoke of ‘the democratic societies making up the Council of Europe’ – Soering, para. 88; of course, the extension is unremarkable), that it was without exception and could not be derogated from even in a public emergency (Article 15(2) ECHR). As in Chahal, the threat of terrorism could not call into question the absoluteness of Article 3. That cannot be anything other than completely correct. The Court did not address the United Kingdom’s argument to the effect that the prohibition of expulsion was a positive obligation, and hence not an absolute one. It seems to me that it was right to disregard the argument. The submission is plainly mistaken. It is certainly true that the positive obligations of protection deriving from Article 3 (see e.g. A. v. United Kingdom, para. 22) are not, indeed cannot, be absolute. They do not oblige the State to take all possible measures against private acts of inhuman treatment; the State must take ‘reasonable steps to prevent ill-treatment of which the authorities had or ought to have knowledge’ (Z v. United Kingdom, para. 73). But the only relevant distinction here is between negative and positive obligations: the former are absolute, the latter are not (Limbuela v. Secretary of State for the Home Department [2005] UKHL 66, [2006] 1 AC 396, paras. 53 and 77-78). The Soering obligation not to extradite or expel in the face of a real risk of torture or inhuman treatment in the receiving State is a negative obligation: it is, more precisely, a negative obligation to protect (Lord Justice Laws in Limbuela v. Secretary of State for the Home Department [2004] EWCA Civ 540, [2004] QB 1440, para. 64; yours truly, EJIL 17 (2006), 349, at 360-361), or in other words (there being no difference in substance), a prohibition on creating a risk of ill-treatment contrary to Article 3 (see S. Weber, ‘Environmental Information and the European Convention on Human Rights’, HRLJ 12 (1991), 177, at 181; yours truly, in the JICJ 4 (2006), 401, 408). It must therefore be absolute; the comparison with positive obligations is simply not apt. Another basis for the submissions of the UK was a passage from Soering, quoted by Mr Swift at the hearing in Saadi (at time code 01:13:03 and following):

[I]nherent in the whole of the Convention is a search for a fair balance between the demands of the general interest of the community and the requirements of the protection of the individual’s fundamental rights. As movement about the world becomes easier and crime takes on a larger international dimension, it is increasingly in the interest of all nations that suspected offenders who flee abroad should be brought to justice. Conversely, the establishment of safe havens for fugitives would not only result in danger for the State obliged to harbour the protected person but also tend to undermine the foundations of extradition. These considerations must also be included among the factors to be taken into account (para. 89 of Soering)

It is this passage that Chahal purported to explain (para. 81 of Chahal), but from which Chahal may be thought to have departed. But however that may be, Chahal was certainly right to take the view that it did, and the passage from Soering relied on by the UK, on the face of it and as read by the UK, does go too far. In particular, the suggestion that a ‘search for a fair balance’ is ‘inherent in the whole of the Convention’ may well be true, but is beside the point, and quite dangerous, in the context of Article 3. As in the extradition context in Soering, the idea of a balancing exercise as ‘inherent in the whole of the Convention’ might potentially be applied in any other case, to deny the absolute protection that Article 3 is there to provide. Who’s to say, for instance, that the argument will not be deployed in the ‘ticking time bomb’ scenario? It is true that human rights are all about balancing the interests of the individual and society, but in the absolute rights such as Article 3, the Convention has decided where the balance lies: torture and inhuman and degrading treatment and punishment are so utterly beyond the pale that they cannot be required or justified by any considerations of proportionality. Whatever falls within Article 3 is necessarily disproportionate, by the clear command of the article itself. The Court would be quite wrong, therefore, to set up any form of proportionality test as an additional requirement before the prohibition cuts in. This is related, if somewhat remotely, to another argument deployed at the hearing (at around 01:08:00 and following): the State is obliged to protect the lives of all under its jurisdiction, not least through measures taken to prevent crime (see e.g. Osman v. United Kingdom, para. 115). With a greater or lesser degree of precision, this argument is very frequently made, in legal as in political discourse. But to make this argument is to forget that the obligation on the State to act is necessarily conditioned by the limitations of the Convention on the way it can act; as the Osman case puts it (para. 116):

Another relevant consideration is the need to ensure that the police exercise their powers to control and prevent crime in a manner which fully respects the due process and other guarantees which legitimately place restraints on the scope of their action to investigate crime and bring offenders to justice, including the guarantees contained in Articles 5 and 8 of the Convention.

The Concurring Opinion of Judge Myjer, joined by Judge Zagrebelsky, in Saadi makes a very similar point, although the Judges may perhaps be criticised for placing the limit too highly, in saying ‘only’ that States ‘must not resort to methods which undermine the very values they seek to protect’ (a valuable discussion of the indirect threat of terrorism to our liberties, through counter-terrorism legislation, follows). The UK had also argued that the danger a person poses to the national security of the State ought to be balanced against the risk of that person being subjected to Article 3 ill-treatment following removal to another State. The Court responds to the argument thus (para. 139):

The Court considers that the argument based on the balancing of the risk of harm if the person is sent back against the dangerousness he or she represents to the community if not sent back is misconceived. The concepts of "risk" and "dangerousness" in this context do not lend themselves to a balancing test because they are notions that can only be assessed independently of each other. Either the evidence adduced before the Court reveals that there is a substantial risk if the person is sent back or it does not. The prospect that he may pose a serious threat to the community if not returned does not reduce in any way the degree of risk of ill treatment that the person may be subject to on return.

Taken on its own, that must be correct. The fact that someone is dangerous does not mean that he or she is less likely in fact to be tortured after expulsion from a Contracting State of the ECHR. The two questions are both questions of fact, and are, as the Court rightly says, independent of one another. [Judge Zupan?i? had made the same point at the hearing (at time code 01:19:00 and following), and has made it again in his Concurring Opinion, at para. 2] However, I think the Court has misunderstood the argument of the United Kingdom. That argument was not that a person proved to be very dangerous would, by the same token, incur a lesser risk of being subjected to ill-treatment abroad. The argument, as made by Mr Swift at the oral hearing, was that the more dangerous the person, the higher the risk of ill-treatment that Article 3 would tolerate: Article 3 would only prohibit the expulsion of a very dangerous person if the risk of ill-treatment was particularly high. The balancing test, then, was to apply not at the stage of finding the facts (of dangerousness and of the risk of torture), but at the stage of applying the law to the established facts. This misunderstanding, as I believe there has been, does not of course mean that the UK was right. Indeed, the Court returned to this argument in the very next paragraph from the one just quoted. It felt, based on its apparent misunderstanding, that this was a second argument in addition to what we might term the ‘balancing on the facts’ submission, but it did consider the point. It said this (para. 140):

With regard to the second branch of the United Kingdom Government’s arguments, to the effect that where an applicant presents a threat to national security, stronger evidence must be adduced to prove that there is a risk of ill-treatment (see paragraph 122 above), the Court observes that such an approach is not compatible with the absolute nature of the protection afforded by Article 3 either. It amounts to asserting that, in the absence of evidence meeting a higher standard, protection of national security justifies accepting more readily a risk of ill-treatment for the individual.

So it does. Of course, the Court had already rejected that argument, following Chahal in holding that Article 3, and the Soering dimension with it, was absolute. That conclusion, I think, is beyond reproach.