The UK and non-refoulement

Some time ago, I posted a suggestion that, from a certain perspective, the international law norm prohibiting the refoulement of people to countries in which they risk being tortured could be viewed as part of global administrative law. This norm, enshrined in Article 3 of the Convention against Torture, is now also undoubtedly part of customary international law, and quite possibly part of the body of jus cogens. I am even less sure than I was previously of the GAL-relevance of non-refoulement, and i haven't had achance to read the case carefully, so I'll keep this brief; but I wanted to mention it, because it's important: the recent judgment of the House of Lords that upholds an earlier decision that the radical Muslim cleric Abu Qatada can be deported to Jordan, on the basis of diplomatic assurances as to his treatment should he be handed over. The judgment was joined with the cases of two other individuals that the UK was seeking to deport to Algeria, raising many of the same issues.

The first thing that I want to note is that, in strictly legal terms, I cannot see that the judgment represents a departure from or an exception to the universal and non-derogable requirement of non-refoulement. Contrary to the suggestion in some newspapers, (The Guardian, for example), the deportation of Abu-Qatada has not been approved "in spite of fears he may be tortured" - this would clearly be in violation of the UK's obligations under both international and European human rights law (the latter being the explicit basis upon which this judgment was rendered). Rather, the Lords upheld a decision by the Special Immigration Appeals Commission that any fears of torture were in the circumstances of the present case unfounded.

Nonetheless, this judgment has been strongly criticised, both in the press and by international human rights NGOs, and it is easy to see why: it does seem to mark a significant shift in emphasis from both the SIAC's treatment of "diplomatic assurances", and the Law Lords' willingness to protect the rights of detainees more generally. In particular, the House of Lords refused to undertake a full review of the decision of the SIAC, holding that the restriction of appeals to questions of law only prevented them from doing so. Therefore, although all parties that "accepted that neither Algeria, in the case of RB and U, nor Jordan, in the case of Mr Othman, was a country to which the appellants could safely have been returned had the United Kingdom not received assurances from the respective Governments as to the way in which they would be treated" (para. 107), the Lords refused to consider whether the assurances actually did reduce the risk of torture below the "susbstantial" level necessary to activate the non-refoulement prohibition, viewing this as a question of fact on which they had no jurisdiction to rule. Rather, they insisted that "the only ground upon which those conclusions can be attacked on an appeal restricted to questions of law is irrationality" (para. 117).

This is a hugely important finding. The House of Lords has effectively denied itself the capacity to oversee the factual conclusions of the SIAC as to whether assurances effectively remove the substantial risk of torture, and has limited itself to the much narrower - and much harder to establish - issue of whether the SIAC had acted "irrationally" in finding so. To give an example of just how much leeway this gives the lower tribunal, consider the following discussion from the judgment relating to the proposed deportation to Algeria. The SIAC had affirmed that 4 conditions had to be met for assurances to be acceptable (para. 23):

i) the terms of the assurances had to be such that, if they were fulfilled, the person returned would not be subjected to treatment contrary to Article 3;

ii) the assurances had to be given in good faith;

iii) there had to be a sound objective basis for believing that the assurances would be fulfilled;

iv) fulfilment of the assurances had to be capable of being verified.

The Lords then recalled that (para. 29)

So far as the fourth condition was concerned, the United Kingdom government had sought to persuade the Algeria Government to agree to monitoring, but had not succeeded. For reasons given in the decision in relation to Y, SIAC concluded that there was nothing sinister in this. There were other ways in which the performance of the Algerian assurances could be verified. British Embassy officials would be permitted to maintain contact with RB, if not in detention, and prolonged detention would itself be indicative of a breach of the assurances. Amnesty International and other non-governmental agencies could be relied upon to find out if the assurances were breached and to publicise the fact. Accordingly SIAC found that the fourth condition was satisfied.

This seems... unsatisfactory; apart from anything else, surely one of the purposes in ongoing monitoring is to enable potential breaches to be identified and prevented early, not simply actual breaches to be acknowledged ex post. Yet, despite these and other potential weaknesses, the self-limiting approach of the House of Lords meant that their review of this aspect of the SIAC's decision was limited to the following brief passages:

I have described earlier in this opinion the consideration given by SIAC to the reliance that could be placed on the Algerian assurances. This had particular regard to the general conditions in Algeria at the time that the assurances were given, the attitude of the Algerian authorities to the observance of human rights, the degree of control exercised by the Algerian authorities over the DRS, the internal security service, and the manner in which the performance of the assurances could be verified. SIAC paid careful regard to all relevant matters and applied to them the proper test of whether they amounted to substantial grounds for believing that RB and U would be at real risk of inhuman treatment if returned to Algeria.

SIAC gave consideration to the reasons why Algeria was not prepared to agree to monitoring and concluded that this was not indicative of bad faith and that there were alternative ways of ascertaining whether there was compliance with the assurances. These conclusions were not irrational. The contention that the assurances did not, on their true construction, protect against inhuman treatment was not well founded.

For these reasons the irrationality challenge to SIAC’s conclusions does not succeed. I would reject the appeals brought by RB and U.

This gives a pretty clear indication of the limited nature of the approach adopted by the House of Lords in this case: there is, it seems to me, a potentially huge difference between "there is no substantial risk of torture" and "it was not irrational to conclude that there was no substantial risk of torture". It's a long judgment, and there's a lot of important stuff in it that I haven't the time to read at present (including relating to the use of closed sessions in the Abu Qatada decision not only in relation to the threat that he posed to UK national security, but also in evaluating the risk of torture and the effectiveness of assurances - see paras. 76-98). In any event, even if the law remains effectively unchanged, this case does seem to indicate an important shift in emphasis in UK jurisprudence on the issue of assurances and non-refoulement.

An appeal has already been lodged before the European Court of Human Rights, and my own view is that they will block the deportation. Always risky, making predictions - on the upside, however, this one is likely to take at least two years, so even those few who have read this far are unlikely to remember it if I got it wrong...