MT (Algeria) & Ors v Secretary of State for the Home Department

This is the open judgment of the court in three appeals from determinations of the Special Immigration Appeals Tribunal (‘SIAC’) in each of which SIAC dismissed an appeal from a deportation decision of the Secretary of State for the Home Department (‘the Secretary of State’) made on the ground of national security. In this court the appellants have been called MT, RB and U. Before SIAC they were called Y, BB and U respectively and we will use these initials in this judgment.

SIAC was created by section 1 of the Special Immigration Appeals Commission Act 1997 (‘the 1997 Act’). By section 2 of SIACA as amended, a person may appeal to SIAC if he would have been entitled to appeal against the decision under sections 82(1), 83(2) or 83A(2) of the Nationality, Immigration and Asylum Act 2002 but for a certificate under section 97 of that Act (‘the 2002 Act’). By section 97(1) an appeal under those sections may not be brought if the Secretary of State certifies that the decision was made or taken wholly or in part on a ground listed in subsection 97(2), namely (a) in the interests of national security or (b) in the interests of the relationship between the United Kingdom and any other country.

These appeals raise closely related but in some respects different issues. They each centre principally on whether there is a real risk that the particular appellant will be subjected to treatment contrary to article 3 of the European Convention on Human Rights (‘the Convention’) if he is returned to Algeria. It is common ground that that is the test under article 3: see eg Chahal v United Kingdom (1996) 23 EHRR 413 (‘Chahal’) at [97]. Article 3 of course provides:

In the case of each appellant the Secretary of State decided to deport him to Algeria on the ground that his deportation would be conducive to the public good because he was a danger to national security. By section 7 of SIACA, this court has jurisdiction to entertain an appeal from such a decision but it is limited to “any question of law material to” the decision. In each case the appellant submits that SIAC erred in law in concluding that there were no substantial grounds for believing that he would be exposed to a real risk of being subjected to torture or to inhuman or degrading treatment or punishment contrary to article 3 of the Convention if returned to Algeria.

We will consider the issues in this order; the appellants’ objections to the use of closed material, which are common to all three appeals, the appeal of Y, the jurisdiction of this court, which is principally relevant to the appeals of BB and U, the appeal of BB, the appeal of U and our overall conclusions

The issue is whether it was open to SIAC to use closed, as well as open, material in reaching its conclusions. SIAC did rely on, or at least refer to, such material in reaching its conclusions, adverse to all of the appellants, on risk on return; and in reaching its conclusion, contested before us in closed proceedings, as to the danger to national security presented by BB. The appellants contended that that procedure was unlawful, an unlawfulness that was not cured by the system of special advocates. The Secretary of State should either have made his case on the basis of material that was disclosed to the appellant, or have accepted that the appellant could not lawfully be removed to Algeria.

SIAC’s procedure was that envisaged by section 5 of the 1997 Act, section 97 of the 2002 Act and rule 4 of the SIAC (Procedure) Rules 2003 (‘the SIAC rules’). In order to demonstrate that SIAC had erred in law the appellants had to displace or explain that statutory scheme. Their case was that SIAC was nonetheless prevented from relying on undisclosed evidence by both of, or by a combination of, the jurisprudence of the ECtHR and the rules of English common law.

This case was supported by the intervener, Liberty, who had the advantage, as did the court, of submissions by Miss Dinah Rose QC. Miss Rose formulated the issue as being:

We have already noted that by section 97(3) of the 2002 Act SIAC’s jurisdiction is engaged when the Secretary of State certifies that his decision was made wholly or partly in reliance on information which should not be made public (a) in the interests of national security; (b) in the interests of the relationship between the United Kingdom and another country; or (c) otherwise in the public interest. SIAC’s procedure is governed by the 1997 Act, and rules made under that Act. Section 5(3)(a) of the 1997 Act provides that rules may be made which enable proceedings before SIAC to take place without the appellant being given full particulars of the reasons for the decision. Rule 4(1) of the SIAC rules provides:

The argument changed its shape somewhat in the course of the proceedings, but the final statement of it rested very strongly on what was described as the procedural as well as the substantive component of article 3. The state’s obligation is not only not to deport persons to a state where they face a real risk of torture, but also to ensure that the proceedings in which that issue was considered were fair. Proceedings are not fair if evidence is used that is not seen by the appellant and his (open) advocates.

There is no doubt that the national court must engage in rigorous scrutiny of an article 3 case, in view of the absolute character of the article 3 obligation and the fundamental values that it enshrines: see for instance Chahal at [96]. However, it does not follow from that that presence throughout of the applicant, or the giving to him of all of the evidence, is a necessary component of the rigorous scrutiny. That presence and participation is of course a very important element in the process of getting to the correct answer, since the applicant may be able as no-one else is to correct factual errors. But absence of that element does not in itself disqualify what otherwise is a rigorous scrutiny. Rather, overall imperative for the applicant to be present is in order to respect the values of openness and legality. Those are logically different from the value of getting the correct answer.

The only authority that Miss Rose was able to show us that was alleged to demonstrate that the principle she relied on was embedded in article 3 was [150]-[153] of the judgment of the ECtHR in Chahal. Those passages stress the need for a national remedy that is effective, and not just “as effective as can be”, and continue:

The principal claim, based directly on article 3, therefore fails.

The other aspect of the case was the proposition that the presence of the applicant throughout the process, and access by him and his advocates to all of the evidence, is a fundamental principle of legality in English domestic law, though rendered all the more important in this case by the engagement of the domestic courts’ obligations under article 3.

That principle and its importance is not in doubt. Whether it can prevail in the face of the statutory provisions is another matter.

For the contention that the statute and rules must be offset in some way or other the appellants relied on the classic statement of Lord Hoffman in Simms v SSHD [2000] 2 AC 115 (‘Simms’) at 131E:

There is no such difficulty because, when Parliament passed the statutory scheme, it was concerned, and well knew that it was concerned, with very specific legislation to address a clear and particular case. In creating SIAC and providing for its particular procedure, including the use of special advocates, Parliament did squarely confront what it was doing and accepted the political cost. There are no sensible grounds for fearing that the full implications of its words may have passed unnoticed in the democratic process. It must be obvious that Parliament was well aware that the SIAC procedure would be used in claims under the Convention, including in particular under article 3. The case is quite different from Simms, where a general and blanket ban on contacts between prisoners and journalists was held not to apply to contacts aimed at assisting in a further reference of the prisoner’s case to the Court of Appeal; or R (Morgan Grenfell) v Special Commissioner [2003] 1 AC 563, where a very general power to seek documents created by section 20(1) of the Taxes Management Act 1970 was held not to extend to documents covered by legal professional privilege.

In these circumstances we are unpersuaded that it is open to us to interfere with the statutory scheme: whether by selective interpretation; reading down; or (in respect, but only in respect, of that part of it that affects an article 3 issue) a declaration of incompatibility. The qualification in brackets in respect of a declaration of incompatibility reflects the acceptance by the appellants that even if all of their arguments were correct they could only seek a declaration of incompatibility in respect of the effect of the legislation on the issue as to safety on return, since the issue as to danger to national security was not in any event affected by article 3. That same differentiation was not accepted as applying in the case any exercise of selective interpretation on purely domestic law grounds. The Secretary of State widened that issue, to argue that it was irrational, and also that it produced an impossible exercise in interpreting what is a single set of rules, to seek to amend or interpret the rules in their application to safety on return when they plainly applied fully to the issue of danger to national security. We do not pursue that further argument because the Secretary of State would only need to rely on it if all or part of the appellants' case were otherwise correct: which premise we have held not to be established.

Very much as a second-best solution, but nonetheless strongly pressed, the appellants and Liberty said that, even if the whole SIAC procedure did not fail, within it SIAC must take steps to ensure that as much as possible of the “closed” material was made available to the applicants. That was particularly so in respect of material that was confidential because of diplomatic sensitivities, but which did not involve the operations of the security services, or state secrets in the usual sense of that expression.

Two procedures were suggested that might be employed. The first was that SIAC should conduct part of the proceedings in camera, with the applicant but not the press or public present. That, it was claimed, would adequately protect material that was restricted because of its diplomatic sensitivity if broadcast to the world. Rules as to contempt would adequately control the applicant and his advisers. This may be a possible option in some circumstances, and we were told that it had been used in some official secrets trials. It is however necessary to look at the reality of the actual case in issue. Whatever the correctness of the charges brought against these appellants, it is accepted that they are significant opponents of the Algerian state. As Mr Tam said, they are the last people who ought to be admitted to that state’s confidential diplomatic dealings. We are not clear whether SIAC was invited to adopt this option. It would have been entirely within its powers in rejecting it.

The second procedure was for information to be made available to each appellant’s (open) counsel, on terms that it could not be shown to or discussed with the client; and, presumably, that the client would have to leave court when counsel sought to deploy that material in argument. We were told, in general terms, that the Criminal Division of this court has approved arrangements in which a defendant in a criminal trial is precluded from knowing information disclosed to his legal representatives. The authority relied on for this proposition, R v Davis [2006] EWCA Crim 1155, is however of limited effect, addressing only a case where counsel could cross-examine a witness screened from the defendant on terms that information as to the witness’s identity was not passed to the defendant. That is far different from what is contemplated here, of the open advocate having a range of substantive material central to the issues in the case that he cannot discuss with or disclose to his client. In the present context, such an arrangement would seriously undermine the careful division between counsel appearing in the open proceedings and the special advocates. Counsel would be put in the impossible position of being asked to play, part-time, the role of the special advocate, but without the protection, and formal position, that the special advocates enjoy. That is objectionable not so much in the interests of counsel, but in the interests of not distorting the trial process. SIAC was right to hold that the disadvantages of this course far outweigh any benefit that might accrue to the applicants.

As to the identification of what material is in fact covered by rule 4, and more generally SIAC’s management of the open and the closed processes, Mr Tam told us, without contradiction, that SIAC takes rigorous care to ensure that the Secretary of State does not extend the closed process in an unreasonable way. That is a proper and sensible course, which properly mitigates the practical effect of a statutory scheme that is in law unchallengeable.

We accordingly reject all of the appellants’ arguments of principle. That does not, of course, mean that SIAC had to give anything other than anxious scrutiny to the material, closed as well as open, on which it did rely.

This is an appeal by Y against a decision made by SIAC on 24 August 2006 dismissing his appeal against a decision by the Secretary of State to deport him to Algeria. The appeal is brought in part with the permission of Ouseley J as chairman of SIAC and in part with the permission of this court, comprising Pill, Keene and Smith LJJ. There is no closed judgment in Y’s case.

The appeal raises three distinct questions:

whether SIAC erred in holding that Y would be able to take advantage of article 9 of the Ordonnance which implemented the Algerian Charter for Peace and National Reconciliation (‘the Charter’);

whether SIAC erred in placing any reliance upon closed material in its consideration of Y’s case on safety on return to Algeria; and

whether SIAC erred in concluding that article 1F(c) of the Refugee Convention 1951 (‘the Refugee Convention’) extended to acts committed by Y after his recognition as a refugee.

The facts are set out in very considerable detail in the judgment of SIAC, which runs to 416 paragraphs. They are not now in dispute. It is only necessary to refer briefly and in outline to some aspects of the facts. The reader should consult the judgment of SIAC for the detailed picture.

Y is a citizen of Algeria born in October 1969. He left Algeria in 1999 and arrived in the United Kingdom on 5 March 2000. He claimed asylum on 8 March 2000. His application was refused on 18 January 2001, but his appeal to an adjudicator was allowed on 1 June 2001. The Secretary of State did not appeal and Y was accordingly granted indefinite leave to remain in November 2001. He was arrested on 7 January 2003 on suspicion of being concerned in the instigation, preparation or commission of acts of terrorism contrary to section 41 of the Terrorism Act 2000 and was later charged with three offences. He was tried with four others at the Central Criminal Court between September 2004 and April 2005 at a trial which was known as the ‘ricin’ or ‘poisons plot’ trial. The indictment contained two counts: conspiracy to murder and conspiracy to cause a public nuisance. Y and three others were acquitted on all counts but one of the accused, Bourgass, was convicted of conspiracy to cause a public nuisance.

The Secretary of State gave notice to Y of his decision to make a deportation order against him on 15 September 2005, certifying under section 97(1)(a) of the 2002 Act that the decision was taken in the interests of national security. It followed that his appeal lay to SIAC and not to the Asylum and Immigration Tribunal. His Grounds of Appeal raised potential breaches of the Refugee Convention and of articles 2, 3, 5, 6 and 8 of the Convention. Even if not a breach of article 3, his expected treatment was cumulatively said to be sufficiently grave to cause SIAC to exercise its discretion under the Immigration Rules differently.

The case for the Secretary of State was that Y was a risk to national security, which Y denied. At [7] to [131] SIAC considered the question whether Y was such a risk in great detail and held that he was. Its conclusions are at [100] to [109]. We note that at [107] it said that its conclusions based on the open material were reinforced in each instance by the closed material. It concluded in that paragraph that Y is an Islamist extremist of long-standing, who has significant terrorist group connections, notably to Dhamat Houmet Daawa Salafia (‘DHDS’), and that his activities, by way of logistic support for those groups, and his presence as an active extremist supporter, showed that he is a risk to the UK’s national security and should be deported. Those conclusions are not challenged in this appeal.

Y’s case before SIAC was (among other things) that if he was deported to Algeria he would face a real risk of being subjected to treatment contrary to article 3 of the Convention and that it followed from the principles in Chahal that he could not be deported. SIAC rejected the submission that Y would face such a risk and dismissed his appeal against the decision to deport him.

SIAC held that it was safe for Y to be returned to Algeria because he would be entitled to the benefit of article 9 of the Ordonnance. Mr Starmer QC and Mr Husain submit on behalf of Y that in reaching that conclusion SIAC erred in law. We return below to the alleged error of law but consider first the facts relevant to this issue, including the nature of the Ordonnance. We do so with the assistance of a chronology of facts helpfully prepared by Mr Starmer and Mr Husain which are derived from the judgment and which are not in issue between the parties.

In June 1997 Y was convicted in his absence in Algeria of an offence of organising an armed group prejudicial to the security and integrity of the state. The offence was said to have been committed in March 1996. He was sentenced to life imprisonment. In February 1998 he was convicted of the same offence, again in his absence. That offence was said to have been committed in May 1996. He was sentenced to death.

In October 2001 the Foreign and Commonwealth Office (‘the FCO’) advised the Home Office that it should not attempt to deport terrorist suspects to Algeria. The Home Office had hoped that assurances could be obtained from Algeria which would ensure that there was no real risk of such suspects being subjected to treatment contrary to article 3 of the Convention. In December 2002 the FCO was asked to review its advice that such assurances should not be sought from Algeria. The FCO initially maintained its view but in May 2003 it agreed that specific and credible assurances might be acceptable. With the aim of obtaining such assurances, the Algerian government was approached in 2004 but little progress was made.

In January 2005 a further approach was made to the Algerian government and in July of that year the matter was raised during a visit to Algeria by a Minister of State. Political agreement was reached on the basis that deportation would be part of a larger package of judicial co-operation including an extradition treaty. There was contact between the Prime Minister and the President of Algeria. Mr Oakden, who gave evidence on behalf of the Secretary of State, co-chaired a British delegation with the Algerians on Deportation with Assurances at the end of August 2005.

The Charter was approved in a referendum in October 2005 and provided for an amnesty for individuals involved in earlier terrorist acts, excluding those involved in massacres, rapes and placing bombs in public places. As SIAC observed at [232], the preamble to the Charter explains its political background, namely ending the destructive tragedy which had engulfed Algeria, marking the defeat of terrorism and its abuse of Islam and furnishing the means for ensuring permanent peace and security through National Reconciliation, supported by the whole Algerian people. SIAC quoted “eight measures … aimed at consolidating peace” in the Charter as follows:

The position at this time was described by SIAC at [234] to [236] in this way:

Mr Oakden’s statement said that the British Government accepted that it could be at risk of breaching its obligations under the European Convention if it were to return Y without first obtaining assurances as to his treatment on return. Algeria’s human rights record had been criticised by NGOs and the international community. The concerns covered torture, arbitrary detention, extrajudicial executions and the right to a fair trial. On the other hand the British Government’s view, from diplomatic reporting, and other Governments’ published assessments was that the situation was improving. The improvement in human rights was directly linked to the recent political and security history of Algeria, as described in the two FCO papers.

It is clear from this “second” statement of Mr Oakden’s that as at November 2005, the SSHD was not seeking to return Y without assurances about his treatment in a variety of ways, assurances which he was seeking to make the subject of independent and credible monitoring.

In a note verbale dated 5 December 2005 the Algerian government explained what would happen upon Y’s return in the light of his convictions in absentia. He would be entitled to have his conviction and sentence set aside because he had been tried in his absence. He would be retried before a competent court. On arrival he would be arrested by the judicial police (as opposed to the DRS) and detained in a Ministry of Justice prison. He would have the following rights: to appear before a court for the purpose of obtaining a decision as to the legality of his arrest or detention, to be informed of the charges laid against him and to be informed of his right to be assisted by counsel of his choice and to make immediate contact with such counsel; to receive legal aid; not to be detained otherwise than by a competent judicial authority; to have the benefit of the presumption of innocence until his guilt was established lawfully; to inform one of his family or friends of his arrest or detention; to be visited by a doctor; to respect in any circumstances for his human dignity; if he was retried and found guilty and the death penalty was imposed, the 1993 moratorium on executions would apply; and, if he has not been previously involved in a collective massacre, rape or explosive attacks in public places he would be eligible to benefit from the provisions of the Charter and the subsequent legislation implementing it.

That subsequent legislation took the form of the Ordonnance, which was dated 27 February 2006 and approved by the Algerian Cabinet. The terms of the Ordonnance were set out in some detail in Appendix 1 to the judgment of SIAC. It is only necessary to set out the following provisions in order to determine this appeal:

to implement the provisions of the Charter for Peace and National Reconciliation, which is the expression of the sovereign will of the Algerian people

to give concrete expression to the determination of the Algerian people to put the final touches to the policy of peace and national reconciliation, which is essential for the Nation’s stability and development.

SIAC examined a number of exchanges between the FCO and the Algerian government on the question whether the amnesty provided for in the Ordonnance would apply to Y. For example, on 27 March and 6 April 2006 the Algerian Ministry of Justice explained how the Charter worked: see SIAC’s judgment at [273] and [274]. On 11 July 2006 there was an exchange of letters between the Prime Minister and the President of Algeria designed to strengthen co-operation between the two governments.

The case for the Secretary of State, supported by the evidence of Mr Oakden, was that Y would be entitled to the benefit of article 8 of the Ordonnance on his return to Algeria. It was put on this basis. The offences of which Y was convicted in his absence in 1997 and 1998 were offences contrary to article 87a of the Penal Code. Y was not, however, someone who had committed or acted as an accomplice in the commission of or instigated offences of collective massacre, rape or the use of explosives in public places. It followed that articles 5, 6, 8 and 9 of chapter II of the Ordonnance applied to him by reason of articles 2 and 10 (quoted above). So much was common ground. As to article 8 itself, the case for the Secretary of State was that, since Y was a person who was convicted and sentenced in his absence for an offence within article 2, the right to prosecute him again was extinguished.

The case for Y was that that would only be so if Y presented himself to the competent authorities in Algeria and made the declaration required by article 8. It was submitted to SIAC that Y could not make such a declaration in time because the time expired six months from the date of publication of the Ordonnance. It followed that time expired on 27 August 2006. Further and more importantly, Y’s case was that he was not guilty of the offences alleged against him and that in these circumstances he could not be expected to make the declaration. SIAC accepted at [279] that Y could not be expected to admit the serious offences of which he was convicted but which he denied.

SIAC then considered (at [279] et seq) the suggestion made by a seconded High Court judge that Y did not have to admit the offences in order to make the declaration. He could simply put “not applicable” in the section which required admissions of acts done. This suggestion was made at a comparatively late stage before Mr Oakden gave evidence before SIAC. Exchanges between the FCO and the Algerian government continued while the hearing before SIAC was proceeding. One such exchange included a document, which was disclosed on 21 June (shortly before final submissions were made on behalf of the Secretary of State on 28 June) and which related to a meeting in Algeria on 6 and 7 June: see [301] et seq.

SIAC analysed the evidence of Mr Oakden in considerable detail at [275] to [300] and the document just referred to at [301] et seq. It described the evolution of the Secretary of State’s case at [327] to [340], it discussed the political situation at [341] to [350] and it expressed its conclusions as to the application of the Ordonnance at [351] to [377]. The critical conclusion for the purpose of this appeal is at the end of [356], where SIAC said that, even if the article 8 declaration was made in time, “there would still be a risk that Article 8 would not be the route to the beneficial application of the Ordonnance to Y”.

The reasons for that conclusion can perhaps best be seen at [354] to [356], although the remaining paragraphs of SIAC’s judgment repay study. It said at [354] to [356]:

Notwithstanding its conclusion on article 8, SIAC held that there was no real risk to Y if he were returned to Algeria because “the right to bring a public prosecution” against him was or would be extinguished by article 9 of the Ordonnance. It did so on the basis of its own construction of article 9. That construction was simply this. Y is now held in custody, albeit in England, but in any event will be held in custody, albeit (on SIAC’s findings) for a short time, on his return to Algeria. In these circumstances, article 9 applies or will apply to Y because he “has not been finally sentenced for having committed or having acted as an accomplice in the commission” of the offences for which he was convicted in 1997 and 1998 and which are offences within article 2 of the Ordonnance.

That construction is entirely understandable but the problem is that, except perhaps at the very end of the case before SIAC, it was not part of the Secretary of State’s case. In particular, it was not part of the evidence of Mr Oakden upon which the Secretary of State relied. Although Mr Oakden did of course refer to article 9 because it was part of chapter II of the Ordonnance, he at no time suggested in evidence that, if article 8 was not applicable to Y because he would not make the declaration in time or because he could not reasonably be expected to make the relevant declaration, that did not matter because the same result would be achieved under article 9.

If Mr Oakden had thought that that might be the position he would surely have mentioned it in the course of his evidence, which was very extensive. Equally, if those advising the Secretary of State had thought that Y could safely be returned to Algeria because of article 9, they would have made article 9 part of the Secretary of State’s case from the outset. It did not occur to the Secretary of State that the issue of Y’s safety could be resolved simply by asking SIAC to construe articles 8 and 9 of the Ordonnance without the assistance of evidence. If the Secretary of State had thought of relying upon article 9 at the outset, he would surely have investigated the position under article 9 with the assistance of Mr Oakden, just as he did in the case of article 8. Just as SIAC said at [356] (quoted above) that the operation of the article 8 declaration was a matter upon which the Secretary of State was better placed than Y to obtain the necessary and conclusive information, so the same could in our opinion be said of article 9.

The hearing before SIAC took place on 25 to 28 April, 2 to 4 and 24 May and 26 June 2006. When written final submissions on behalf of Y were prepared on 15 May they contained no reference to article 9 because no reliance had been placed upon it on behalf of the Secretary of State. The Secretary of State served his written submissions on 26 May. There were four references to article 9, at paragraphs 4.2.17, 4.2.19, 4.2.26 and 4.2.27. We accept Mr Starmer’s submission that in those submissions the Secretary of State does not rely upon article 9 as a freestanding argument. In paragraph 4.2.17, article 9 is only mentioned in passing. Paragraph 4.19 simply describes the cases in which the right to prosecute is extinguished. It thus refers to article 4, to articles 5 to 8 and to article 9, without addressing the question which of those articles applies to Y. Paragraphs 4.26 and 4.27 appear in a section of the submissions which discusses the place of the declaration under article 8 and contain no clear statement that the Secretary of State was relying on article 9 if his case on article 8 was rejected.

We note that the memorandum of the meeting of 6 and 7 June mentioned above, which of course took place after the date of the submissions to which we have just referred, makes no reference to article 9. As Mr Starmer correctly submits, it is clear that even at that late stage the Algerian authorities were still being asked about the article 8 route and no one on either side thought to raise the possibility that article 9 might apply at the meeting.

We have seen a transcript of the submissions made by Mr Burnett QC on behalf of the Secretary of State on 28 June. He first referred to article 9 in the context of the correct approach to the nature of the declaration required under article 8. A little later Mr Burnett referred to article 9. His submissions contained the following:

In the course of his submissions in reply Mr Emmerson QC for Y referred to Mr Burnett floating the possibility that Y might be held not pursuant to article 8 but pursuant to article 9. He added:

There followed this exchange:

In summary, we consider that the Secretary of State only touched upon article 9 at the conclusion of his submissions and did so without any evidential basis for doing so. While Mr Emmerson did not expressly submit to SIAC that the point was not open to the Secretary of State, he did submit that there was no evidence to support the submission. SIAC did not address that submission.

SIAC concluded that article 9 did indeed apply to Y. It did so without any detailed analysis. At [353] it said that any expiry of the six month period under article 8 would put the case within article 9. At [357], immediately after [356] quoted above, SIAC said that on the basis that article 8 will not apply to Y for time reasons and might not apply because of a failure to make a declaration in accordance with article 8, Y would have to rely upon article 9, “which does not require a declaration but does suppose that the individual is in custody”. SIAC then considered at [357] to [363] and [364] and [376] whether Y might be tried for offences which were excluded from the Ordonnance under article 10 but concluded that he would not. That point would apply under both article 8 and article 9 and was not the subject of argument in this appeal.

SIAC’s key conclusions with regard to article 9 are in [364] and [377] as follows:

SIAC then considered, under the heading of Detention, whether Y would be treated any differently under article 9 than it had held he would be treated under article 8. At [378] it noted that it had been anticipated that Y would only be in custody for a short period but that that was on the basis the he was returning to claim the benefit of article 8. At [379] it said that it did not know what the period in custody would be for someone returning to claim the benefit of article 9 but expressed the view that there was no real reason why that period should be any longer than the short period envisaged under article 8. SIAC nevertheless recognised that it might be longer, albeit no more than a few weeks. Moreover it considered whether he would be likely to be detained by the DRS and concluded at [381] that there was no more than a real possibility of that. It further concluded at [383] that if (as it put it) perchance he were to be in DRS custody while his position under the Ordonnance was being regularised he would not face a real risk of being tortured. The essential conclusion of SIAC in this regard was that, although the evidence did not address this question, there was no reason to think that a person entitled to the benefit of article 9 would be treated any differently from a person entitled to the benefit of article 8.

We are troubled that SIAC considered article 9 without the possibility that it might apply being covered in evidence. Mr Emmerson was justified in submitting in the passage from his oral argument quoted above that there was no evidence to support the conclusion that article 9 would apply to Y. Mr Tam submits that the question whether the article applies is a matter of fact. That is so but it is not, in our judgment, correct to treat it as simply a matter of Algerian law and thus, by our rules, a matter of fact and then to hold that this court cannot interfere with the conclusion of the SIAC because its jurisdiction is limited to matters of law.

The question is not simply one of Algerian law but of what is likely to happen on Y’s return to Algeria. That involves a consideration of what the Algerian authorities are likely to do. This was recognised by the Secretary of State in the context of article 8 and led to a great deal of evidence being given by Mr Oakden and considered in detail by SIAC. In our opinion, before the Secretary of State relied upon article 9 and in any event, before SIAC reached a conclusion on the applicability of article 9, the issue should have been explored in the evidence. In our judgment, viewed on the basis of the material available to SIAC when it gave its judgment on 24 August 2006, reliance upon article 9 without any evidential basis for it beyond the reading of the words of the article led to potential injustice for Y.

The above conclusion is reinforced by what happened subsequently. On 14 November 2006 a meeting took place at the Ministry of Justice in Algiers which was attended by among others, Mr Anthony Layden, who has replaced Mr Oakden as the person advising the Secretary of State and who has given evidence in the later Algerian cases, including those of BB and U. The Algerian delegation was led by Mr Amara, who was the seconded High Court Judge referred to above. The purpose of the meeting was to discuss the position subsequent to the judgment in Y. We have a detailed note of the meeting.

In paragraph 3 of the note, Mr Layden is recorded as explaining the next steps in Y and a possible time frame. The note, in which Mr Layden is referred to as ‘Anthony’ and MT as ‘Y’, continues:

We note that what Mr Layden said is consistent with the two central conclusions we have reached above. The first is that the Secretary of State had submitted to SIAC that Y would benefit from articles 8 and 13 of the Ordonnance. There is no suggestion that the Secretary of State had relied upon article 9. The second is that it was SIAC who decided the case on the basis of article 9. Mr Layden wanted advice on whether the Secretary of State was right to take the point he did and whether SIAC was correct to hold that Y would benefit from article 9.

Paragraph 4 of the note sets out the response of Mr Amara and an Algerian official on the article 8 point. On the first point, Mr Amara expressed his opinion, although he stressed that it would be a matter for a judge in Algeria and that he was expressing his view as a jurist and not as a judge or a government official. His view was that, if Y made a declaration without specifying his previous activities, the right to prosecute would be extinguished by articles 8 and 13. He and the Algerian official agreed that it now seemed clearer that the Secretary of State’s case on article 8 was correct.

Paragraph 5 of the note relates to the article 9 point and is in these terms:

As can be seen from paragraph 5 of the note, Mr Amara’s view was that SIAC’s construction of article 9 could be supported but that it was not what those drafting the Ordonnance had had in mind. An earlier part of the note shows that he was himself one of the draftsmen. Importantly, his view was that an Algerian judge would be unlikely to take the view of article 9 which had been taken by SIAC.

The Secretary of State did not object to the admissibility of this evidence in this appeal. He was, in our judgment, correct not do to so. The evidence reinforces the conclusion we had previously reached, namely that it was important for SIAC to have evidence from Algeria before reaching a conclusion of a part of the Ordonnance upon which the Secretary of State was not essentially relying. It is a reasonable inference from the note that, if such evidence had been sought and obtained, it would not have supported SIAC’s conclusion. Indeed, it may well be that the view expressed by Mr Amara, as evidenced in paragraph 5 of the note, was the view which Mr Oakden had already formed and is the reason that the Secretary of State did not rely upon article 9 in any part of his evidence.

There was discussion in the course of the argument as to whether the evidence now available demonstrates an error of law on the part of SIAC. Mr Starmer submits that it does. He relies upon a number of cases in which the courts have considered in what circumstances an error of fact can amount to an error of law. He relies in particular upon E v Secretary of State for the Home Department (‘E’) [2004] EWCA Civ 49, [2004] QB 49, Antonio Cabo Verde v SSHD [2004] EWCA Civ 1726, R (Iran) v SSHD [2005] EWCA Civ 982 and Shabana Shaheen v SSHD [2005] EWCA 1294.

The essential principle is stated thus in the E case by Carnwath LJ, giving the judgment of the court at [66]:

Mr Starmer submits that there was here a mistake as to an existing fact, namely that Y would benefit from article 9, that the fact that he would not has been established because of the opinion of Mr Amara which is now available, that neither Y nor his advisers were responsible for the mistake and that the mistake played a material part in the decision. There is undoubted force in those submissions. However, Mr Tam has drawn our attention to [7] of the speech of Lord Bingham in Runa Begum v Tower Hamlets LBC [2003] UKHL 5, [2003] 2 AC 430, where, as an example of a case in which an error of law might be established, he identified a case where the “decision maker is shown to have misunderstood or been ignorant of an established and relevant fact”. Mr Tam submits that the principle identified by Carnwath LJ only applies where there has been ignorance of “an established and relevant fact” which is (as Carnwath LJ put it) uncontentious and objectively verifiable. He submits that that is not the case here.

We certainly caution against the use of this principle to turn what is a simple error of fact into an error of law by asserting some new fact which is itself contentious. However, in our judgment it is not necessary to deploy this principle in order to conclude that the decision of SIAC on this point cannot stand. It is important to note that, although the appellant before SIAC was Y, all (or almost all) the information relevant to his safety on return to Algeria is in the hands of the FCO and the Secretary of State. The FCO has conducted many discussions with the Algerian authorities and was able to produce the detailed evidence of Mr Oakden. As we stated earlier, the Secretary of State could have asked Mr Oakden for his opinion on article 9, just as it did on article 8. We do not know whether Mr Oakden was asked about article 9 or not or, if he was, what his opinion was. It is a reasonable inference, in the light of the views of Mr Amara quoted above, that, if he had been asked he would have consulted Mr Amara or someone else with equivalent knowledge of the drafting of the Ordonnance. It may indeed be that Mr Oakden did consider article 9, either with or without taking advice from Algeria about it and that he formed the view that it would not avail Y. We do not know.

In these circumstances the process which led SIAC to conclude that Y would be entitled to the benefit of article 9 was not fair to him. It infringed the principles of natural justice in a way not dissimilar to the CICB case. It is common ground that cases like this, where it is said that there is a risk of a deportee being tortured in breach of article 3 of the Convention, require anxious scrutiny. In the absence of evidence about article 9, SIAC cannot properly be said to have given anxious scrutiny to the question whether it could safely hold that there was no real risk of Y being tortured on his return to Algeria on the basis that he was entitled to rely upon article 9. That failure to give anxious scrutiny to the point by insisting upon evidence about it is not an error of fact but an error of principle or law which entitles this court to interfere with SIAC’s conclusion that article 9 would apply to Y.

Mr Starmer invites us to remit the matter to SIAC, whereas Mr Tam submits that it is not necessary or appropriate to do so because of SIAC’s findings as to the value of the assurances given by the Algerian authorities. Mr Tam submits that, if SIAC had not decided Y’s appeal on the basis of article 9 of the Ordonnance it would have reached the same conclusion and dismissed the appeal.

We have reached the conclusion that we should not dismiss the appeal on the basis that, quite apart from the Ordonnance, it is safe to send Y back to Algeria. Whether it is safe to do so was (and is) essentially a matter for SIAC and not for us. It is common ground that each case must be decided on an individual basis. Thus SIAC must answer the question whether there is a real risk of a breach of article 3 separately in each case. That is because individual circumstances differ, as indeed can be seen from the differences between Y, BB and U in these appeals. That is so, notwithstanding the fact that some of the evidence and some of the conclusions may be common to all the Algerian cases.

Although SIAC did express some views in the course of its judgment in Y’s appeal, it is in our opinion plain that it did not address its mind to the question whether it would have dismissed the appeal whatever view it formed of the Ordonnance. That is clear from [398] and [399] under the heading “Other risks”:

The conclusion which we draw from those paragraphs that SIAC did not hold that it would have dismissed the appeal in any event is confirmed by the reaction of Ouseley J to the evidence of the meeting on 14 November 2006. In the course of giving his written reasons on the issue of permission to appeal on 28 November, Ouseley J, after referring to Mr Amara’s opinion that SIAC’s interpretation of article 9 was not what those who drafted the Ordonnance had in mind and that an Algerian judge would be unlikely to agree with SIAC’s interpretation, said this:

In these circumstances, we conclude that it would not be appropriate for us to analyse the facts and reach a conclusion on the issue of risk. That is a matter for SIAC. We therefore remit the matter for further consideration by SIAC.

For the reasons given above, we conclude that SIAC did not err in placing reliance on closed material. In any event, we are not persuaded that the closed material played any part in the decision in Y’s case: see eg [340] of the judgment.

At [132] to [155] and [402] to [404] SIAC considered a number of issues under the Refugee Convention. In the course of doing so, it considered the question whether Y has lost his status as a refugee by reason of article 1F(c) of that convention. SIAC held that he had, whereas Mr Husain submits that he had not. The issue is whether Article 1F(c) extends to acts committed after recognition of a person’s status as a refugee. SIAC held that it does. The question is whether it was correct so to hold.

The Refugee Convention provides, so far as relevant:

He has voluntarily re-availed himself of the protection of the country of his nationality; or

Having lost his nationality, he has voluntarily re-acquired it; or

He has acquired a new nationality, and enjoys the protection of the country of his new nationality; or

He has voluntarily re-established himself in the country which he left or outside which he remained owing to fear of persecution; or

He can no longer, because the circumstances in connexion with which he has been recognized as a refugee have ceased to exist, continue to refuse to avail himself of the protection of the country of his nationality:

he has committed a crime against peace, a war crime, or crime against humanity, as defined in the international instruments drawn up to make provision in respect of such crimes:

he has committed a serious non-political crime outside the country of refuge prior to his admission to that country as a refugee:

he has been guilty of acts contrary to the purposes and principles of the United Nations.

No Contracting State shall expel or return (“refouler”) a refugee in any manner whatsoever to the frontiers of territories where his life or freedom would be threatened on account of his race, religion, nationality, membership of a particular group or political opinion.

The benefit of the present provision may not, however, be claimed by a refugee whom there are reasonable grounds for regarding as a danger to the security of the country to which he is, or who, having been convicted by a final judgment of a particularly serious crime, constitutes a danger to the community of that country.”

Like any provision of a convention, the language of article 1F(c) of the Refugee Convention must be construed in accordance with the ordinary meaning to be given to the terms of the convention in their context and in the light of its object and purpose: see article 31(1) of the Vienna Convention on the Law of Treaties 1980, which codified pre-existing customary international law. Mr Husain submits that, so construed, article 1F(c) is limited to acts committed before an individual is recognised as a refugee. We are unable to accept that submission.

There is no support for that conclusion in the language of the article 1F. A comparison between the language of Article 1F(c) with that of Article 1F(b) suggests the contrary. Article 1F(b) is limited to a person who has committed a serious non-political crime outside the country of refuge “prior to his admission to that country as a refugee”. Article 1F(c) contains no limitation of time. It seems reasonable to suppose that, if the draftsman had intended a temporal limitation, he would have provided for one.

On the other hand we accept that differences of language do not resolve all questions. For example, article 1C expressly provides for circumstances in which the Convention “shall cease to apply” whereas articles 1D, 1E and 1F simply provide for circumstances in which the convention “shall not apply”. It might be argued that only cases within article 1C apply post-recognition. However, it is correctly accepted that there is no significance in that difference in language because it is plain that article 1E applies post-recognition. In our view, if article 1E applies post-recognition, there is no reason why article 1F(c) should not equally apply post-recognition.

As to policy, Mr Husain submits that the object and purpose of the article 1F was to deny access to refugee status at the point of application. He submits that that is true of article 1F(b) and that the explanation for the restriction in article 1F(b) is that it introduces a geographical limitation as well as a temporal limitation. This does not seem to us to be convincing. However, Mr Husain further submits that the general purpose of article 1F is not to protect the host state from dangerous refugees because that is the purpose served by article 33(2). He submits that article 33(2) governs post-recognition acts so as to deprive a recognised refugee of the protection against refoulement.

In support of his submission Mr Husain relies upon this passage in the judgment of Bastarache J for the majority of the Supreme Court of Canada in Pushpanathan v Canada (Minister of Citizenship and Immigration) [1999] INLR 36 at [58] (‘Pushpanathan’):

That statement undoubtedly affords Mr Husain’s argument some support. However, as Mr Tam observes, the Canadian Supreme Court was not addressing the issue which arises here. The issue was whether, in circumstances where, when his claim was being considered for the first time, the claimant had already been in the country of refuge for some time, acts committed within the country of refuge might fall within article 1F(c). It was argued on behalf of the claimant that only acts committed outside the country of refuge could be taken into account. That argument was rejected.

If the court had been considering the present issue it would no doubt have considered the difference in language between paragraphs (b) and (c) in the context of temporal limitations, just as it did in considering geographical limitations: see [58] quoted above. We do not know what conclusion it would have reached. We also note, as did SIAC at [139], that although Pushpanathan was considered in general terms by this court in A (Iraq) v SSHD [2005] EWCA Civ 1438 at [24], it did not consider the time issue raised in this case.

However that may be, SIAC preferred the reasoning of the Immigration Appeal Tribunal (‘IAT’) in KK v SSHD [2004] UKIAT 00101 as follows:

In any event, we accept Mr Tam’s submission that it is or may be significant that at [58] in Pushpanathan Bastarache J refers to the “general purpose” of article 1F as not being the protection of the country of refuge from dangerous individuals. We agree with Mr Tam that this is not inconsistent with the possibility of a further purpose, which is to ensure that if an individual successfully claims the benefit of refugee status and then, while enjoying the surrogate international protection of his human rights afforded by the country of refuge, commits acts which show that he no longer deserves the benefit of that status, that status can be removed from him, just as he would never have qualified for that status if he had committed those acts before making a claim for refugee status.

We agree with the conclusion of SIAC at [144] that being or becoming a ‘refugee’ as defined in the Refugee Convention does not require or start with a formal state act of recognition of status. A person simply is or is not a refugee within article 1A. He may lose the status of refugee in a number of circumstances as provided in articles 1C, D, E or F. We can see no reason to distinguish in terms of status between a person who, to take Mr Tam’s example, becomes a mercenary in a war and commits war crimes before or after he is (or would but for the war crimes be) recognised as a refugee.

As SIAC held, at [143] to [145], article 1 is concerned with the definition, not recognition of the status, of refugees. There is nothing in article 1F to indicate that it was intended that its provisions should only apply to those who had not hitherto been accorded the status of refuges. As to the role of article 33(2), we agree with the conclusions of SIAC at [146]:

In all these circumstances we conclude that SIAC was correct to hold that article 1F(c) applies both before and after recognition as a refugee. Since this is the only challenge to SIAC’s reasoning with regard to the Refugee Convention, we dismiss Y’s appeal in this regard. In these circumstances, it is not necessary for us to consider whether, if we had reached a different view, the position of Y would be improved or whether, as Mr Tam submits, it is academic.

We allow Y’s appeal because of SIAC’s inappropriate reliance upon article 9 of the Ordonnance and remit the matter to SIAC for further consideration.

We include this topic in this part of the judgment because its relevance is principally to the appeals of BB and U.

By section 7 of the 1997 Act and section 30(5) of the Antiterrorism Crime and Security Act 2001 (‘the 2001 Act’), an appeal from SIAC lies only on a point of law. No doubt with that limitation in mind, the complaints made in the grounds of appeal about SIAC’s conclusions were said to involve errors of law on SIAC’s part, though without further elaboration of what actual errors of law had been committed. It rapidly became clear in the course of the oral argument that what was in truth in issue were SIAC’s findings of fact. That question arose in the following way.

The first, and in the event most relied on, ground of appeal in BB was expressed in terms that SIAC erred in its approach to the assurances provided by the Algerian authorities as to whether, on return to Algeria, BB would be at risk of torture. That issue was central to SIAC’s judgment, because it was accepted that there was a previous history, though in SIAC’s view now significantly in decline, of the use of torture on detainees, particularly at the hands of the DRS, the internal security force. At [5] of its judgment SIAC set out the conditions that it considered must be met before assurances could be relied on as a basis for permitting a return to Algeria:

the assurances must be given in good faith;

there must be a sound objective basis for believing that the assurances will be fulfilled; and

fulfilment of the assurances must be capable of being verified.”

Rather, the complaint was not about the appropriateness of the questions, but about the answers that SIAC had given to those questions. That was apparent from paragraph 2.1.1 of the appellant’s skeleton argument:

When this difficulty was raised with them, the advocates representing both BB and U did not shrink from asserting that, whatever might be the ambit of the concept of error of law in the domestic jurisdiction, everything changed when the court was considering a complaint relating to article 3; or, it must necessarily follow, any other article of the Convention. That approach was indicated in oral argument on behalf of BB, and reinforced by a written submission after argument had closed in BB, which was signed by Mr Rabinder Singh QC, for BB; Mr Drabble QC, for U; and by Mr Hugh Southey who had settled the grounds of appeal in both of those cases. And thereafter, during argument in U, we received further submissions from Mr Drabble. The appellants’ case was summarised in paragraph 2 of the written submission:

In the domestic jurisdiction as previously understood the question of whether an applicant faces a real risk of being subjected to treatment contrary to article 3 (the issue as formulated in [3] of the judgment of SIAC, taken from [80] of the judgment of the ECtHR in Chahal) is a mixed question of fact and law. That expression is not here used, as it sometimes is, as a way of dressing up an issue of fact as an issue of law. Rather, it indicates that there are two discrete issues involved, one of fact and one of law. As Donaldson MR put it in O’Kelly v Trusthouse Forte [1984] QB 90 at [122H]-[123A]:

That is what Kennedy LJ had in mind in R(S) v Secretary of State [2004] HLR 17, a passage relied on by the appellants in our case, when he said at 20 that the issue of whether known facts amounted to a violation of article 3 was a question that the Court of Appeal was as well placed to answer as was a first instance court.

In the case of BB only the first of these issues is in dispute. That is because what BB claims to be at risk of is torture. If that claim is made good, then there is no dispute that that conduct falls within article 3. The position will be different when considering the claims of U (and also BB) under article 3 in respect of prison conditions in Algeria, and also the claims of U and BB under article 6. In both of those cases there are live and serious issues as to whether the facts complained of, even if made good, fall as a matter of law within the relevant reach of article 3 or article 6 respectively.

Because, therefore, of the limited nature of the contested article 3 enquiry in respect of BB it is not easy to see exactly what the appellant’s case is as to the court’s approach to that first issue, of whether there is a real risk of BB being tortured when he returns to Algeria. It would no doubt have been easier for the court to understand this dispute had it emerged earlier than half way through the first day of the hearing. Taken literally, the contention in the appellants’ formal submission that is set out in [96] above, that different considerations from those involved in finding primary facts apply when the facts found are then applied to determine whether there will be a violation of the Convention, says no more than that the second step in the enquiry is not an enquiry into fact at all: the analysis that is indeed adopted in [97] above. However, the further expansion of the argument, and the use sought to be made of it in this particular case, indicates that a good deal more was intended. The claim can only be understood as a claim that this court should treat the first step in the enquiry, the fact-finding process as to what will happen when BB returns to Algeria, if not exactly as a question of law, then certainly as a question on which this court can and should disagree with SIAC. The court’s power is therefore not limited to correcting exercises in irrationality in that fact-finding process: if the Court of Appeal thinks that SIAC reached the wrong conclusion, it should substitute its own view.

In oral argument the appellants’ case was sought to be advanced by analogy with the jurisprudence as to issues of proportionality. It is now clear, not least from [44] of the speech of Lord Bingham of Cornhill in A v SSHD (No1) [2004] UKHL 56, [2005] 2 AC 68 (‘A’), that questions of proportionality are not issues of pure fact, and therefore the conclusions of SIAC on those questions are not immune from reconsideration on appeal. The comparison is, however, misleading, and it received only a passing mention in the appellants’ more considered written submissions. As Lord Bingham made clear, an enquiry into proportionality requires the tribunal to exercise judgement and critical balance, an exercise appropriate for critical review by the appellate court of the kind to which Lord Bingham himself submitted the conclusions of SIAC in A. That assessment of proportionality is far different from the fact-finding exercise with which we are concerned in the present case.

In their written submissions, therefore, the appellants opened up a wider front. The argument had two limbs. First, Lord Bingham in Huang [2007] UKHL 11 at [8] (‘Huang’)had pointed out that public authorities, including the courts, acted unlawfully if they did not act compatibly with a person’s Convention rights:

The second limb of the argument is easily disposed of. Proceedings in the ECtHR are brought against the national state, and the issue is whether the state has acted consistently with the Convention right asserted by the applicant. Those are, therefore, primary, not appellate, proceedings. The ECtHR, while no doubt giving appropriate weight to the findings of the national court, can do no other than treat the case as what it is, a new case. The normal inhibitions on an appellate court in a national system therefore simply do not arise, and no sensible parallel can be drawn from practice in the ECtHR in order to inform, and much less to mandate, practice in the domestic appellate court.

As to the first limb of the argument, the contention appears to be that the Court of Appeal will act inconsistently with the appellant’s Convention rights if it proceeds on a finding of fact by SIAC that indicates that the appellant’s Convention rights are not going to be infringed, when further enquiry into the facts at the appellate level might lead to a different conclusion. It must, therefore, be open to the applicant to appeal SIAC’s findings of fact, even under a domestic rubic that limits such appeals to points of law.

That is completely to distort the principle stated by Lord Bingham. That requires no more than that issues arising in relation to the Convention should be properly considered within the national legal order. It does not require, or permit, disruption of the national legal order to the extent that, although consideration of an issue related to a Convention right is allocated to a particular court, other courts must also volunteer, contrary to the limits placed on their jurisdiction by the national legal order, to perform the same task. That can be further illustrated from the Convention’s own rules as to co-operation by the national state. The obligation of the national state under article 13 of the Convention is to provide an effective remedy for violations of Convention rights. In the present case, the prime instrument for providing that remedy is SIAC. SIAC is a court, not an administrative body, and therefore issues of deference and the like do not arise. The appellants’ argument amounts to saying that it was not open to the United Kingdom to provide that, in respect of the fact-finding part of the article 3 assessment, SIAC’s conclusion should be determinative in the absence of irrationality or perversity.

We are not aware of any other case in which it has been suggested that either the obligations of the courts under section 6 of the Human Rights Act, or the obligations of the state under article 13, require the courts to claim jurisdiction that is not provided by the national legal order. Nor does Lord Bingham say anything to suggest that his observations were intended to have that radical effect. Indeed, it will be recalled that his reason for holding in A that an appeal lay from SIAC on issues of proportionality was that such issues were not ones of pure fact. But the first of the two issues in an article 3 enquiry, that which we are presently concerned with, is an issue of pure fact. Stronger demonstration than is available would be required to show that two years later, in Huang, Lord Bingham had moved beyond that position, to hold that any factual decision in a Convention case must necessarily be appealable, whatever may be the rules of the domestic legal order.

There is a further reason why the appellants’ contention must be treated with extreme caution. We would not, rely on this point if it stood alone, but it certainly underlines the other difficulties of the submission. The appellants rely on the duty of the court, under section 6(1) of the Human Rights Act, not to act incompatibly with a Convention right. It is said that this court will act incompatibly with BB’s article 3 right if it does not admit the possibility of an appeal against SIAC’s factual findings. That proposition is extremely difficult in itself, because it assumes what it has to establish, that BB’s article 3 rights are indeed put at risk by SIAC’s findings. There is, however, a further difficulty in proceeding by way of section 6. Section 6(2)(a) provides that the rule in section 6(1) does not apply to an act if

Mr Drabble raised a further argument, that did not depend on this being a Convention case, but rather made claims about the nature of the fact-finding exercise in English law generally. He said that in the present case SIAC had to find the “primary” facts, which it was agreed involved only a fact-finding exercise. Then, however, it had to “assess” those facts (for instance, the reported facts about Algeria’s attitude to assurances about torture) to determine whether there was a sufficient risk of BB being tortured on return to Algeria. That latter exercise was not an exercise of pure fact-finding. Mr Drabble was perhaps reluctant to say that it was an exercise that of its nature involved questions of law; but he did say that it was an exercise that this court could perform again, in place of SIAC without exceeding its jurisdiction.

This argument misstates the factual issue that the court has to determine, and also misunderstands the nature of the fact-finding process. The issue of fact in this case is whether there is a sufficient risk of BB being tortured on return to Algeria. That is a single and undifferentiated question of fact, which it is for the fact-finding tribunal to determine. In making that determination the fact-finding court will no doubt assess the impact of other findings that it has made, just as it will assess the reliability of witnesses in deciding what findings to make. But that process is all part of the fact-finding process. That process of assessment is quite different from, and plays a quite different role from, for instance, the assessment that the court has to make, based on the found facts, of whether the defendant acted negligently: which is a matter of legal judgement, and not just a question of what is going to happen in certain circumstances in the future.

The effect of this argument would be to turn almost all issues of fact into questions of law; or, at least, into questions that could be reopened with impunity by an appellate tribunal. Once the proper use of the concept of assessment is understood, it can be seen that no such unacknowledged revolution has occurred.

We were much pressed with some observations of Auld LJ in this court in R (Bloggs 61) v Home Secretary [2003] 1 WLR 2724 (‘Bloggs’), in particular at [60]-[66]. The issue in that case was the risk to B’s life, under article 2, if as an informer he was placed in unprotected conditions in prison. This court held that the issue of risk to life was a single issue, and that it had to review the Prison Service’s decision on that issue with an appropriate intensity. That is indeed the exercise that the court undertook, not of retaking the decision, but of deciding whether the decision fell within the bounds of what was rational: as the court concluded, holding at [69] of the judgment that it could not interfere with the Prison Service’s decision. Thus as we read the decision in Bloggs it was a decision on rationality. It follows that in any event it is of no assistance in the instant case because it is not said that SIAC’s decision was irrational.

Finally, we should perhaps emphasise, in the context of an issue that we have considered in the Y at [66] to [69] above, that the issue that we have reviewed is different from the question of whether a mistake of law can occur when a tribunal proceeds in ignorance of, or by a mistaken understanding of, an existing and uncontentious fact: see the judgment of Carnwath LJ in E at [66]. As our judgment indicates, we accept that that is now an established, if limited, category of error of law. It has nothing to do with the general contentions as to jurisdiction just discussed, which claim that this court can become involved in the exercise of assessing whether a particular fact exists, as opposed to correcting the lower court for failure to act on a fact the existence of which is uncontentious.

We have gone into this dispute at some length both because of the weight placed on it by the appellants and because it is of some general importance. The appellants’ arguments cannot be accepted. However, the point is academic in the present appeals. In Y we have concluded that the appeal must succeed, but on a ground that does not engage the present issue. In BB and U, whatever attack is made on the factual findings of SIAC necessarily fails.

BB is a citizen of the Republic of Algeria, who arrived in this country as long ago as 4 May 1995. He claimed asylum on 26 February 1999, a claim that the Secretary of State refused on 29 September 2005. Meanwhile, however, on 15 September 2005 BB had been served with the Secretary of State’s decision to deport him on grounds of national security, the case then being certified under section 33 of the 2001 Act. It is that decision that was the subject of the judgments of SIAC that are under appeal.

For the reasons set out in its closed judgment SIAC found that BB is a danger to national security and that it would be in the public good of the United Kingdom for him to be deported. BB contests that conclusion. Our reasons for rejecting that part of his appeal are given in our own closed judgment of today. Additionally, BB complained of the form of SIAC’s open judgment on this point, which at [1] merely recorded the closed decision without further elaboration. For reasons that we have explained in our closed judgment we have decided to remit this case to SIAC for reconsideration of that point of form. That does not affect the substance of SIAC’s conclusion on the national security issue.

The present, open, judgment is therefore concerned with the second issue in the case, whether SIAC erred in law in concluding that there were no substantial grounds for believing that BB, if returned to Algeria, would face a real risk of treatment contrary to either article 3 or article 6 of the Convention.

The grounds of appeal extended to six heads. On an interlocutory application to a single judge of this court permission was granted in respect of only two of them, but the lord justice recognised that several of the grounds overlapped or could be argued to do so. We decided that, rather than hear discrete further argument directed at the permission stage, it would be most convenient to permit the appellant to develop before us the whole of his grounds.

Four further preliminary points should be noted. First, argument before us as to risks in Algeria under article 6 was addressed by counsel in the case of U, and we deal with it there. It was not suggested that, in that respect, there was any difference between BB and U. This judgment accordingly concentrates on the risk in relation to article 3 treatment. Second, the principal complaint under article 3 related to the risk of persons returning to Algeria being tortured and this judgment concentrates on that, very important, aspect of the article 3 issue. Further concern was however expressed as to whether prison conditions in Algeria involved a breach of article 3. That issue also arises in U and we deal with it in that judgment.

Third, SIAC, at its [3], appreciated that determination of that risk had to be fact-specific:

Fourth, there emerged in the course of argument an issue of principle as to the jurisdiction of this court in hearing appeals from SIAC. We have addressed that issue above.

The appellant’s case took the form of a series of criticisms of the conclusions of SIAC, some of them directly involving disputes of fact and some of them concentrating on the form of SIAC’s approach. We address the various elements in turn.

It will be recalled from [2.1.1] of the appellant’s skeleton argument, set out in paragraph 95 above, that a major complaint of the appellant was that any assurances given by the Algerian authorities were worthless because there was “uncontested evidence” that the state authorities were unable to control the DRS. Complaint was in particular made of SIAC’s analysis at [11] of its judgment of the evidence of Mr Anthony Layden, the Foreign Office’s representative in discussions with Algerian officials:

These contentions are simply wrong. They overlook the findings of SIAC in its judgment in Y as to the commitment of the Algerian state, led by President Bouteflika, to civilian control and the loosening of military power; its finding at [215] of Y that the President is master of his own domain and not a mere puppet of the military; and its finding, at [348] of Y, that the direction is sufficiently set for any likely successor to President Bouteflika to continue it should anything untoward happen. By pointing out that the DRS report to the President SIAC certainly did not therefore imply that the DRS were out of control, or not committed to the peace and reconciliation process. Indeed, quite the reverse, in view of SIAC’s findings as to the influence and commitment of the President. It was therefore entirely open to SIAC to accept the evidence that it summarised in [11] of its judgment.

In its [18] SIAC made further findings as to the interests of Algeria in maintaining good relations with the United Kingdom, and more particularly in being accepted by the international community as a normally-functioning civil society:

Although it does not seem to have been contested below, or in the Grounds of Appeal, that in principle it was open to SIAC to base its conclusion on assurances given by the receiving state, that issue did arise before us. The contention that assurances were or might never be sufficient was based on a case not available to SIAC, the judgment of de Montigny J in the Federal Court of Canada in Sing v Canada [2007] FC 361 (‘Sing’). At his [136] the judge cited with approval the report of the Special Rapporteur to the United Nations, UN Document A/59/324:

As the judge said, the logic of that view is easy to grasp. If a country is disrespectful of international norms and obligations, it is likely to be no less disrespectful of its obligations under a lower-level instrument such as a diplomatic note. But that is not this case. SIAC went to a good deal of trouble to establish that in Algeria there is not now a consistent pattern of gross, flagrant or mass violations of human rights or of a systematic practice of torture, nor anything like it; and that the Algerian authorities would not be contemptuous of the obligations that they had undertaken.

That the legitimacy of acceptance of assurances depends on the facts of each case can be demonstrated from Chahal. There, the ECtHR did not consider that the giving of assurances by the Indian government was sufficient protection for Mr Chahal; but it reached that conclusion after an analysis of the facts of the case and of the particular vulnerability of Mr Chahal, rather than by the application of any rule of law or thumb.

The assurance given by the Algerian authorities was, in relevant part:

Mr Rabinder Singh said that these complaints were open to him because an issue as to the construction of a document, such as the document containing the assurances, was always an issue of law. Even if that is correct as a universal proposition, which we doubt, the complaint in the present case is not about the meaning of the document, but about what the document does not say. But we can ignore these technical issues, because the complaint on any view borders on the fanciful. SIAC at its [15] pointed out that the reference to human dignity was precisely a reference to, amongst other things, the exclusion of torture, couched in universally understood diplomatic language. And, we may add, also in the language of commonsense. The suggestion that the Algerian authorities might in some way wish to hide behind provisions of Algerian law that permit torture when international law does not is equally unreal. Even if Algerian law were in theory to permit that course (an issue on which there was no evidence), the last thing that would be done by a country like Algeria, found by SIAC to wish to be accepted as a serious member of the international community, would be to point up in so stark a fashion the extent to which its domestic provisions lagged behind international norms.

At [143] of Sing the judge said:

The appellant said that the refusal of the Algerian authorities to accept independent monitoring, when the United Kingdom had wished them to do so, was significant. That argument ignored the finding of SIAC at [21] of BB, referring back to its findings at [335] and [336] of Y, that that diffidence was caused by the sensitivity of Algeria, as a recently post-colonial state, to any suggestion that it needed outside surveillance of its behaviour; and that the refusal of outside monitoring was not sinister so far as the interests of BB were concerned.

Whether the assurances can be properly policed is, again, an issue of fact and judgment, to be made in the circumstances of the particular country involved. SIAC thoroughly addressed the position with regard to Algeria in terms that cannot be challenged.

We have gone into the criticisms of SIAC’s factual findings in some detail even though that enquiry does not fall within our jurisdiction. On any view, the criticisms fail.

We now turn to a series of different complaints made by the appellant.

The Secretary of State has adopted a practice of “exculpatory disclosure”: that is, he will make available to an applicant any material that might assist the applicant’s case or undermine the Secretary of State’s case. The appellant said that this approach was too narrow, It did not respect the “cards on the table” principle in public law cases, as set out, for instance, by Donaldson MR in R v Lancashire CC ex p Huddleston [1986] 2 All ER 941. It is not easy, as so stated, to see what the difference between the two formulations is supposed to be. However, on exploring the issue it emerged that what is complained of is that by limiting disclosure to matters relevant to the applicant’s “case” the Secretary of State could avoid providing material that, although not relevant to a line of argument currently pursued by the applicant, might however suggest a different line of objection that it could be fruitful for him to pursue.

The short answer was given by Mr Tam. The applicant’s “case” is always going to be that he is at risk of suffering article 3 persecution on return. If the Secretary of State is loyal to his undertaking he will have to disclose anything touching on that general issue, whether or not it affects an argument already put by the applicant.

A subsidiary part of this issue was a complaint that the Secretary of State had not disclosed to BB the report of the discussion as to the ambit of the Ordonnance that took place after SIAC had delivered judgment in Y. That would, at best, be an example of failure to respect the policy, rather than a demonstration that the policy itself was flawed. Mr Tam said that the document had not been disclosed to BB because it concerned detention for serious offences, which was not expected to occur in BB’s case. It is arguable that scrupulous caution might have led to the document’s disclosure, but we cannot think that the issue is at all significant, and certainly does not indicate that the policy itself was unreasonable, let alone unlawful. It was never suggested to us that BB’s case had been in any way damaged by his not having seen the document.

Both BB and U raised as a ground of appeal the complaint that the prison conditions to which they would be subject if returned to Algeria gave rise to a real risk that their article 3 rights would be infringed. To avoid repetition we have dealt with this complaint in [169] to [173].

Both BB and U contended that there were substantial grounds for believing that if returned and if charged with any offence, they would suffer a ‘flagrant denial of a fair trial’. Again, to avoid repetition we have dealt with this complaint in [175] to [188].

Although some suggestion was made in the grounds of appeal of lack of reasoning, viewed as a separate issue, that complaint was not seriously pursued. And on the issues of substance raised in the open proceedings, and addressed above, we would see no reason for criticising SIAC’s conclusions. However, for reasons set out in our closed judgment, and on the case as a whole, we are persuaded that the case should be remitted to SIAC for further consideration.

U is an Algerian national, born in 1963. In November 1994, he came to the United Kingdom, claiming that he had fled from ill-treatment in Algeria. His application for asylum was not determined until 27 June 2000, when it was refused. Meanwhile, in 1996, U travelled to Afghanistan where he remained until 1999; he then returned to the UK. He was arrested in February 2001. In May 2001, he was released and rearrested on immigration grounds. Subsequently, the USA sought his extradition but abandoned its request in August 2005. In the same month, the Secretary of State decided that U posed a risk to national security and that he should be deported because his presence in the UK was not conducive to the public good.

U appealed to SIAC against the decision to deport him but he did not challenge the Secretary of State’s decision that he posed a risk to national security. He argued only that return to Algeria would infringe his Convention rights under articles 3, 5 and 6. If returned, there was a real risk that he would be subject to torture or inhuman and degrading treatment or punishment and, if put on trial, as was likely, the processes would not comply with articles 5 and 6.

The Secretary of State’s reasons for deciding that U posed a threat to national security were set out in some detail in SIAC’s open judgment. In summary, the Secretary of State considered that, while in the UK from 1996 until 2001, U was a leading organiser and facilitator of terrorist activity aimed mainly at overseas targets. He had formed and led a terrorist group. The UK Government’s views about U’s activities in Afghanistan were conveyed to the Algerians through diplomatic channels in 2006. These were that U had held a senior position in a Mujahedin training camp. He had had direct links with Usama Bin Laden and other senior Al Qaeda figures. He had been involved in supporting terrorists in the planned attack on the Strasbourg Christmas Market in 2000 and in an earlier plan to attack Los Angeles Airport. It was for that that the US Government had sought his extradition.

In its open judgment, dated 14 May 2007, SIAC concluded, on the balance of probabilities, that U had been involved in facilitating terrorist activity overseas. There was no evidence that, while detained in prison, he had in any way disavowed his former beliefs and associates. SIAC concluded that U was a significant risk to the UK’s national security.

As we have said, the issue on the appeal to SIAC was safety on return. Many of the issues which arose were similar to those which had arisen in the appeal of BB which had taken place in November 2006. U’s case was heard in two stages; on 8 days between 13 and 21 February 2007 and also on 17 and 18 April 2007. New evidence relevant to the risk on return had become available which had not been available at BB’s hearing. Thus, although the issues in U’s appeal were essentially the same as in BB’s and much of the evidence was common to both, there was additional evidence in U’s case. There was, however, an important factual difference between the two cases. BB, although held to be a security risk, could not be regarded as an important figure in the terrorist world; by contrast, U was. SIAC accepted that, if returned to Algeria, U would be of interest to the authorities and was likely to be charged with offences related to terrorism.

In its open judgment, SIAC first considered the assurance that had been provided by the Algerian Ministry of Justice in respect of U in August 2006. It was ;

SIAC referred to previous decisions in which it had considered the reliability of assurances given by the Algerian state. These were in the appeals of Y, BB and also G. SIAC adopted the conclusions reached in those appeals without repeating them. In summary, at [13], it said that Algeria was

SIAC referred to the approach that should be taken in an individual case to the question of whether the assurance could be relied on; that had been set out in BB and we have referred to it at [128] above. The same approach would be adopted in the case of U. However, since the hearings in Y and BB, information had become available about the way in which the Algerians had treated four men who had withdrawn their appeals to SIAC and had been deported to Algeria in late January 2007. One of them, Q, had not had the advantage of any individual assurance. Two of them, K and P had had assurances similar to that given to U. One, H, had had a differently-worded assurance. SIAC considered that information as to the way in which these men had been treated could be used as a guide to what would happen to U if returned and the reliability of the assurance in U’s and other cases.

Although SIAC analysed the information about the deportees in some detail, there is no need for us to do so. Suffice it to say that K and P had each been arrested for a short period on return and had then been released. They had not been charged with any offences and there was no suggestion that they had been ill-treated. SIAC concluded that, because both had been suspected of terrorist offences, their detention had been lawful. In their cases, the assurances had been honoured.

Q and H had also been arrested, detained and questioned but, unlike K and P, they had later been charged with offences of membership of a terrorist group acting abroad. They were now in custody awaiting trial. In each case, there was conflicting evidence as to the way in which they had been treated, both while in detention before charge and also since they had been charged and detained in Serkadji Prison in Algiers. SIAC evaluated that evidence and concluded that there had been no breach of any relevant part of the individual assurance given in respect of H. Also, in respect of both men, it appeared that Algerian law had been complied with. The time limits for detention before charge had been respected. Both men had been permitted family visits and had instructed a lawyer. There had been no ill-treatment. Although there was evidence that the men had been held incommunicado for some days before charge, SIAC was not prepared to hold that this had breached Algerian law. Detention incommunicado was permitted under Algerian law to ‘secure the secrecy of the investigation’. There was no evidence to show that the actions of the authorities had not been justified in these cases. SIAC dismissed as implausible a claim that H had heard the screams of a woman being ‘stretched’ in an adjacent cell.

Finally, SIAC considered reports that, while in detention, both Q and H had reported hearing the sounds of ill-treatment of others. It was suggested that they were being exposed to these sounds deliberately, in order to frighten them and weaken their moral resistance to interrogation. SIAC accepted that the deliberate exposure of a prisoner to sounds of ill-treatment with the intention of frightening him and weakening his moral resolve was capable of amounting to treatment infringing article 3. However, it held, on the balance of probabilities, that Q and H had not been deliberately exposed to the sounds of ill-treatment of others with the intention of weakening their resolve but it could not rule out the possibility that the men had heard such sounds.

SIAC also considered the effect of the news of bomb outrages which had occurred in Algiers on 11 April 2007. The offices of the Prime Minister and Minister of the Interior had been attacked, as had a police station. 33 people had been killed and many injured. Responsibility for the attacks had been claimed by Al Qaeda of the Islamic Mahgreb (formerly the GSPC). SIAC observed that, so far as was known, the reaction of the Algerian authorities had not been to make mass arrests of known opponents of the state; nor had they reacted in any other way typical of a lawless authoritarian state. SIAC concluded that the occurrence of these attacks would mean that anyone connected with the GSPC (such as U) would be of interest to the authorities. It was contended on U’s behalf that these occurrences would increase the risk that U might be tortured by the DRS. But SIAC held that the fact that these outrages had occurred would not mean that there was a real risk that the DRS would subject U to torture or ill-treatment.

SIAC referred briefly to the communication arrangements (described in greater detail in SIAC’s open decision in Y), which the British Government had put in place so as to permit returnees who were at liberty and the relatives of those who were in detention to speak to Embassy officials about their concerns. It observed that it appeared that the British Government had taken sufficient active steps to ensure that the assurances given by the Algerians were fulfilled.

At [37], SIAC concluded that, although it could not wholly exclude the possibility that the assurances given in U’s case would not be breached or that he would be subjected to ill-treatment which would infringe article 3, it considered that that was no more than a ‘mere possibility’. There were no substantial grounds for believing that there was a real risk that such things would happen in the case of U.

SIAC then considered two further submissions: first that prison conditions in Algeria amounted to inhuman and degrading treatment and second that the legal system in Algeria could not provide U with a fair trial in accordance with article 6. They rejected both and concluded that, overall, there were no substantial grounds for believing that, if U were returned, the UK Government would be in breach of the Convention obligations.

Permission to appeal was granted in 6 of the 11 grounds originally advanced. Before this court, five grounds for which permission had been granted were pursued and a renewed application for permission was made in two more. Some of U’s grounds were also argued in BB’s appeal and have already been dealt with earlier in this judgment.

The first ground of appeal relevant only to U’s appeal relates to SIAC’s approach to its assessment of the facts when deciding whether there were substantial reasons to believe that there was a real risk of torture on return. At [30] SIAC accepted that it was bound, with (it said) one qualification, to approach this assessment as directed by this court in the context of an asylum case in Karanakaran v SSHD [2000] 3 All ER 449; [2000] Imm AR 271 (‘Karanakaran’). SIAC had in mind the passage at [468d]:

SIAC considered that it was entitled to apply a qualification to that guidance. Article 3 cases were different from refugee cases and it was not bound to follow the rule, applicable in a refugee case, that the decision-maker must not exclude any matters from its consideration unless it feels that it can safely discard them because it has no real doubt that they did not in fact occur. It then cited the Strasbourg case of Shameyev v Georgia and Russia, application number 36378/02 (‘Shameyev’) where it had been held that the ‘mere possibility of ill- treatment …. is not in itself sufficient to give rise to a breach of article 3’. SIAC continued:

Mr Drabble submitted that this approach was wrong in law. The approach in an article 3 case should be the same as in a refugee case. The process was similar and both types of case entailed the need to comply with international law. To support that proposition he relied on a starred but unreported determination of the IAT in Kajac v SSHD 21 May 2001. This approach, he submitted, had never been challenged. Moreover, the House of Lords had acknowledged the close relationship between article 3 and the Refugee Convention in R (Bagdanavicious) v SSHD [2005] 2 AC 668 at [30].

Mr Tam did not agree with this submission; he argued that there were differences between an asylum case and a case under article 3. However, his main argument was that, whatever qualification SIAC said it had applied, it had in fact applied Karanakaran correctly and had taken into account all the relevant factors, including its conclusion that Q and H might possibly have heard the sounds of others being ill-treated.

Mr Drabble submitted that this was not so. Moreover, his complaint was not limited to SIAC’s supposed failure to take into account the risk that U might be exposed to the sounds of ill-treatment. He produced a list of risk factors which he submitted SIAC should have taken into account but had not.

We would accept that the correct approach to the application of the Chahal test is that described in Karanakaran. The decision-maker should take a holistic approach; it should take account of all the relevant evidence and risk factors, giving to each matter such weight as it warrants, bearing in mind its importance in the context of the case and the extent to which it has been satisfactorily proved. It will be proper to exclude from consideration those matters which it can safely discard because it has no real doubt that they did not occur. The decision-maker should also take account of the absence of satisfactory information relating to matters of importance. If no evidence or information can be discovered on a matter of importance, its absence will be relevant to the assessment of future risk.

On the face of it, it appears that, by saying that it would exclude from consideration those factors which amounted to no more than a mere possibility, SIAC might have misunderstood the effect of the decision in Shamayev. That case is authority for the proposition that a mere possibility that the returnee might be subject to torture or inhuman or degrading treatment is not sufficient to justify a refusal to return. There must be more; there must be substantial grounds for believing that there is a real risk of such treatment. However, we do not think that SIAC did misunderstand Shamayev. It is clear at [37] that its conclusion was that, although there was a possibility that U might be treated in such a way as to infringe his article 3 rights, that was not enough; there were no substantial grounds for believing that there was a real risk that U would be treated in that way. That was the correct approach. It also appears to us that, although SIAC said that it was going to exclude from consideration the possibility that H and Q had heard the sounds of ill-treatment, it did in fact have that factor in mind when reaching its conclusion. It plainly did not think that that factor amounted to anything of significance. That being so, we are satisfied that SIAC did in fact adopt the approach required by Karanakaran.

As for Mr Drabble’s list of risk factors, which he submits should have been taken into account, it appears to us that, although the judgment does not set out a list of factors, each of these matters had in fact been considered and taken into account. We are quite satisfied that, at least so far as the open evidence was concerned, SIAC considered all relevant matters and applied the Chahal test as required by Karanakaran. The first ground of appeal fails.

Mr Drabble’s second submission was that SIAC had misunderstood the effect of Shamayev in another respect. He submitted that Shamayev demonstrates that the extradition (or, by analogy, return following deportation) of a person to a state where there is a history of torture can amount to a violation of article 3 if (a) there has been at least one previous extradition to that country and (b) the fate of the person extradited is not sufficiently clear because the receiving state will not permit information to be gathered.

We can deal with this point very briefly as it is quite without merit. Shameyev is not authority for the proposition claimed. In declaring that Shamayev’s extradition to Russia would infringe his article 3 rights, the Strasbourg court took into account the fact that conditions for detainees in Russia were very bad and that little was known about the fate of others who had been returned there some time earlier. As we have already said, the absence of information about conditions in a country will often be relevant to the decision on risk on return. However, as SIAC observed at [31], the propositions advanced on behalf of U sought to set up findings of fact about an assessment of risk in an individual case as a proposition of law. In our view, that is right. We entirely accept that lack of information about conditions within a country to which a person might be returned will be relevant to the assessment of risk on return. However, in U’s case, SIAC had taken account of the fact that there were some matters on which information was incomplete. It did not misunderstand Shamayev and this ground of appeal fails.

The third ground related to SIAC’s finding at [37] that the arrangements for the verification of the Algerian assurances were adequate. The arguments advanced were the same as those advanced in the appeal of BB. We rejected them there and do so again for the same reasons.

The contention here was the same as had been raised in the appeal of BB, namely that the language of the assurance in U’s case was inadequate. It was couched in terms of compliance with Algerian law rather than international obligations and it failed to provide a specific guarantee that U would not be tortured; referred only to respect for his human dignity. We have already dealt with this argument in our judgment in BB’s appeal. We rejected it there and reject it here for the same reasons.

Both BB and U raised as a ground of appeal their complaint that the prison conditions to which they would be subject if returned to Algeria gave rise to a real risk that their article 3 rights would be infringed. In neither case was permission to appeal case granted. The arguments were advanced to us as renewed applications. The argument was advanced by Mr Drabble and adopted by Mr Rabinder Singh.

The argument had been advanced to SIAC on the basis of evidence received about the conditions in which Q and H had been detained. It was reported that H’s beard had been shaved off by the prison authorities and he was very distressed about it. Q was said to be in a very small cell which he had to share with 2 others; it was dirty and there were shared toilet facilities. In addition, there was some evidence that, at Serkadji Prison, Q was sharing a dormitory cell with 25 others and was forced to take sleeping medication. One of the men had described conditions as ‘hell’ and had complained that he was only allowed to exercise for 10 minutes a day.

At [38], SIAC approached the question of prison conditions by reference to the decision of this court in Batayav v SSHD [2003] EWCA Civ 1489. SIAC summarised the position thus:

SIAC then considered the evidence before it. This came in part from the Algerian Government, which acknowledged that conditions in Serkadji jail were not ideal. It had been built in the 19^th century and was being gradually upgraded to bring it up to modern standards. Exercise periods varied between one and five hours. There were qualified prison officers and medical staff. The prison was subject to inspection by non-governmental organisations. SIAC also considered evidence that the prison had last been visited by the International Committee of the Red Cross on 26 February 2007. The latest US Department of State Country Report dated March 2007 noted that prison conditions in Algeria generally were ‘difficult’ but ‘improving’. Overcrowding was a problem in some prisons and the quality of medical care was uneven. However, independent human rights observers reported that conditions in prisons generally had improved.

SIAC concluded that, on the basis of this evidence, it could not sensibly be claimed that there was a consistent pattern of gross and systematic violation of article 3 rights in respect of prison conditions. Counsel for BB and U did not suggest that this conclusion was not open to SIAC on the evidence. Instead they submitted that SIAC should have judged the position by reference to the available evidence relating to Q and H. The problem with that submission is that, at [40], SIAC did so and concluded that the conditions reported in respect of Q and H were nothing like as serious as those reported in cases such as Shamayev or Kalashnikov v Russia (2003) 36 EHRR 34 or Peers v Greece (2001) 33 EHRR 51. Conditions did not approach those found in cases in which an article 3 infringement had been found. In so far as there was any shortage of evidence about the conditions in which Q and H were being detained, that was not the fault of the British Government. In any event, it could not be assumed that, if returned, U would be detained in the same conditions as Q and H. SIAC concluded that there were no substantial grounds for believing that there was a real risk that the conditions U would be detained in might infringe his article 3 rights. On the evidence, it was, in our view, entitled so to conclude. There was no misdirection of law. This ground of appeal is without merit and we refuse permission to appeal. In so far as BB sought to advance the same or similar arguments, it is without merit in his case too.

It was contended on behalf of U that SIAC had erred by applying the Chahal test separately first to the risk of torture on return and then to the risk of exposure to inhuman and degrading prison conditions. This was not the right approach to a possible breach of article 3. Both matters should have been considered together as it was the cumulative risk that mattered. We agree that it does appear that SIAC considered these matters separately and we also agree that the correct approach is to consider the cumulative effect of the evidence. The Chahal test is whether there are substantial reasons for believing that there is a real risk of torture or inhuman or degrading treatment or punishment – all taken together. Strictly speaking, we accept that, in this case, SIAC should have considered the evidence about the risk of active, deliberate, ill-treatment aimed at U as an individual (including the possibility of his exposure to the sounds of the suffering of others) at the same time as the evidence of the risk of ill-treatment arising from prison conditions to which all Algerian prisoners are exposed by the system and also the evidence relating to Q and H. However, we do not think that, given the findings SIAC made on the separate issues, the result could have been any different if the risks had been considered cumulatively. We reject this ground.

Both U and BB contended that there were substantial grounds for believing that if returned and if charged with any offence, they would suffer a ‘flagrant denial of a fair trial’. As we have said, the position of the two appellants was rather different in that SIAC held that it was unlikely that BB would be of interest to the Algerian authorities and he was therefore unlikely to be charged. However, in U’s case, it held (at [44]) that U would be of interest and would probably be charged with an offence of membership of a terrorist organisation. He would then face a prolonged period in custody awaiting trial.

Before examining the evidence, SIAC considered the appropriate test for a decision on whether a person should not be returned because of the risk that there would not be a fair trial. At [45] it cited a passage from Soering v United Kingdom (1989) 11 EHRR 439 at [113] (‘Soering’):

However, SIAC considered that these words did not go so far as to say that in the event that such a risk arises, return is prohibited. It then cited Mamatkulov v Turkey (2005) 41 EHRR 25 at [494], as being the nearest that the Strasburg court had come to laying down a test:

In our view, that passage implies, although it does not expressly state, that return will be prohibited where there is a risk of a flagrant denial of justice. SIAC then referred to R(Ullah) v Special Adjudicator [2004] 2 AC 323 at [24], where Lord Bingham of Cornhill said:

At [44], Lord Steyn said:

SIAC observed that, from those statements of the law, it was not clear how article 6 is to be engaged or what the consequences of engagement are. However, in seeking to define the test that should be applied, it drew assistance from [86] of Soering where the court said:

After this consideration of authority, SIAC concluded that there was no coherent definable test. All it could do was to examine the Algerian procedures so as to determine whether U would, if tried, be subjected to a process which in its judgment would amount to a flagrant denial of justice. It has not been submitted to us that this approach was wrong.

SIAC then carefully examined the evidence available to it about the Algerian system of criminal justice. It did so in particular in the light of four grounds of concern raised by U’s counsel. These were that U would be detained for a lengthy period before trial; he would not be permitted access to a lawyer during the first 12 days of detention; the court of trial would not be independent or impartial and the evidence used against him might be tainted, as for example by being obtained by torture.

SIAC held that the conditions of detention during the first 12 days, although not article 6 compliant, could not, of themselves, be described as a flagrant denial of justice. It also considered that detention before trial for the maximum available period (which SIAC thought could be as much as 60 months) would not of itself amount to a flagrant denial of justice. Detention before trial was subject to judicial control. However, it accepted that both these factors must be taken into account in an overall assessment.

SIAC considered evidence relating to the independence and impartiality of the judiciary in considerable detail and at some length. Its conclusion was that the Algerian judiciary is both formally and effectively independent of the executive but that not all its members have yet developed the robust independence of mind which is the norm in countries where the rule of law is long established. It also observed that the concerns of Amnesty International have been mainly aimed at the reluctance of the judges to enquire into allegations of torture. SIAC noted that there was a credible report that, despite criticism of the attitudes of the Algerian courts, they do regularly acquit defendants in terrorist and national security cases. SIAC’s conclusion was that there was no real risk of a flagrant denial of justice by reason of the lack of independence of the judges.

Finally, SIAC considered whether there was a danger that U might be convicted on evidence obtained by torture. It recognised the possibility that tainted evidence might be used. It identified in particular evidence from a man named Meguerba, (in respect of whom there was some credible evidence that he had been tortured) who had made a written statement about U. However, SIAC considered that this evidence was of very limited effect. Further, if Meguerba (or indeed any other witness) had anything important to say about U’s activities, he would have to give oral evidence in the presence of the parties and could then be questioned about whether he had been tortured. There was no risk of a flagrant denial of justice. SIAC also considered the possibility that other potential sources of tainted evidence might be used against U but regarded the possibility as fanciful.

SIAC drew together all the strands of evidence about the criminal justice system in Algeria and said that it was unpersuaded that the shortcomings it had identified would create a real risk that U would be subjected to a flagrantly unfair trial.

Before us, it was submitted that SIAC’s finding in respect of the independence of the judiciary was not open to it in the light of the evidence that the President had said March 2006 that the justice system was ‘still dysfunctional’. It was submitted that this showed that Algeria could not provide an impartial and independent judiciary and was the clearest evidence of a flagrant breach of article 6. However, SIAC had specifically considered the President’s remarks and had attempted to put them into the context of his speech. This was to the effect that judges must be above corruption and not swayed by the military or politicians. Change would not occur overnight but ‘the choice made by the state was irrevocable’. What SIAC made of this speech was essentially a question of fact. In our view, it was open to SIAC to hold, as it did that, with some exceptions, the judiciary was both formally and effectively independent of the executive.

Second, it was submitted that SIAC’s finding that there was only a remote possibility that tainted evidence would have any bearing on U’s trial was not open to it. It was said that SIAC did not know what evidence would be used at U’s trial. We see the force of that. However, SIAC considered the rules by which evidence is admitted and while recognising that it was possible that a statement obtained by torture might be received in evidence, concluded that any such statement would have little bearing on the outcome of the trial. Any important evidence had to be presented orally and was open to challenge. In our view, this conclusion was open to SIAC on the evidence before it. Accordingly, we would refuse permission in respect of this ground. As BB adopted U’s submissions and as there is, on SIAC’s findings, little prospect that BB will ever face trial in Algeria, we refuse permission in his case too.

So far as the open evidence is concerned, we consider SIAC’s overall conclusion to be justified (on the facts found), namely that, in deporting U, the United Kingdom will not be in breach of its Convention obligations. However, having also considered the closed evidence and the arguments addressed to us by the special advocates, we cannot express the same degree of confidence. We cannot, of course, explain in any detail why we have reached that view. All we can say is that we have been shown closed evidence which is capable of undermining SIAC’s overall conclusion. We do not say that this evidence does in fact undermine its conclusion, only that it is capable of doing so. We do not consider that SIAC has dealt adequately, in its closed judgment, with some of the salient points raised by the special advocates. SIAC has not adequately explained why it concluded that the closed evidence did not undermine the conclusion it had reached in its open judgment. Accordingly, the appeal must be allowed and U’s case must be remitted to SIAC for it to reconsider the closed evidence and the effect, if any, it has upon the conclusion in its open judgment.

We summarise our conclusions as follows:

A number of points of principle were taken to the effect that SIAC had made inappropriate used of closed material. We have concluded that none of those points of principle is sound: see [6] to [23] above.

Y did not challenge SIAC’s conclusion that he is a danger to national security. His challenge was to the conclusion that there is no real risk of his being ill-treated contrary to article 3 of the Convention if he is returned to Algeria. SIAC’s conclusion was that Y would be entitled to rely upon article 9 of the Ordonnance. In our view, the process which led SIAC to reach that conclusion was not fair to him: see our conclusion at [70] and reasoning at [24] to [69]. The correct course is to remit Y’s case to SIAC in order to consider the Secretary of State’s alternative case that it would in any event be safe to send him back to Algeria. SIAC did not give detailed consideration to that question in Y’s case: see [71] to [75]. We have rejected Y’s challenge to SIAC’s decision that he has lost his refugee status by reason of Article 1F(c) of the Refugee Convention: see [77] to [90].

At [92] to [113] we consider the jurisdiction of this court in relation to appeals from SIAC on questions of law and distinguish them from questions of fact.

As to BB, we have rejected his appeal against SIAC’s decision that he is a danger to national security in a closed judgment given today, although we have remitted the matter to SIAC on a point of form: see [115]. As to his case that SIAC erred in law in concluding that there were no substantial grounds for concluding that, if returned to Algeria, he would face a real risk of treatment contrary to article 3 or 6 of the Convention, we have concluded, on the basis of the open material, that SIAC made no such error of law: see [116] to [139]. However, for reasons set out in our closed judgment, and on the case as a whole, we are persuaded that the case should be remitted to SIAC for further consideration: see [140].

In the case of U, he did not challenge the Secretary of State’s decision that he was a threat to national security: see [142] to [144]. The issue before SIAC was whether there were substantial grounds for concluding that, if returned to Algeria, he would face a real risk of treatment contrary to articles 3, 5 or 6 of the Convention. SIAC held that there were not. U challenged that decision. We consider that, so far as the open evidence is concerned, SIAC’s overall conclusion is justified (on the facts found), namely that, in deporting U, the United Kingdom will not be in breach of its Convention obligations. However, having also considered the closed evidence and the arguments addressed to us by the special advocates, we cannot express the same degree of confidence. We have been shown closed evidence which is capable of undermining SIAC’s overall conclusion. We do not say that this evidence does in fact undermine its conclusion, only that it is capable of doing so. We do not consider that SIAC has dealt adequately, in its closed judgment, with some of the salient points raised by the special advocates. SIAC has not adequately explained why it concluded that the closed evidence did not undermine the conclusion it had reached in its open judgment. Accordingly, we allow U’s appeal and remit his case to SIAC for it to reconsider the closed evidence and the effect, if any, it has upon the conclusion in its open judgment: see [189].

In the result, each of these cases must be remitted to SIAC for further consideration. This raises the question how that consideration will be carried out. That is of course a matter for SIAC but we wish to make it clear that we are not remitting each case to the same constitution of SIAC that heard each before. For understandable reasons each constitution was different. However, each of the cases raises questions which relate to the assurances given by the Algerian government to the United Kingdom government with regard to people returned to Algeria. In these circumstances, it seems to us to be desirable, if at all possible, for the cases now to be considered together (or perhaps one after the other) by the same constitution.

Finally, we have received applications for bail on behalf of Y and BB. Given that these cases are to be remitted to SIAC, we have reached the preliminary conclusion that the appropriate course would be for such applications to be made to SIAC and not to this court.