United Kingdom Supreme Court

Summary




Michaelmas Term
[2010] UKSC 49
On appeal from: [2010] EWCA Civ 426

JUDGMENT

MA (Somalia) (Respondent) v Secretary of State for the Home Department (Appellant)

before

Lord Phillips, President
Lord Walker
Lady Hale
Lord Mance
Sir John Dyson, SCJ

JUDGMENT GIVEN ON

24 November 2010

Heard on 11 October 2010

Appellant Elisabeth Laing QC Deok Joo Rhee (Instructed by Treasury Solicitor)

Respondent Richard Drabble QC Graham Denholm (Instructed by CLC Solicitors)

SIR JOHN DYSON SCJ (delivering the judgment of the court)

1. The issues raised by this appeal are whether the Court of Appeal (i) adopted the wrong approach to the assessment of the impact of MA's lies to the Asylum and Immigration Tribunal ("AIT") on his claim for international protection on the basis of Article 3 of the European Convention on Human Rights ("ECHR"); and (ii) impermissibly interfered with the assessment of the facts made by the AIT, including the impact of MA's lies on a relevant aspect of his claim. As will become apparent, there was little debate or disagreement between the parties to this appeal about the questions raised by the first issue, although they are unquestionably of general importance. The second issue raises the question of how far it is legitimate for an appeal court to interfere with the assessment of facts made by a specialist tribunal on the grounds of error of law.

The facts

2. MA is a citizen of Somalia. He is a member of the Isaaq clan, sub-clan Habr Yunis. He entered the United Kingdom illegally on 7 May 1995 and applied for asylum on 24 May 1995. That application was refused on 14 February 1996, but he was granted exceptional leave to remain until February 1997. He was then granted further leave until 14 February 2000.
3. On 23 July 1998, he was convicted of rape and indecency with a child. He was sentenced to eight years' imprisonment. On 21 May 2002 the Secretary of State for the Home Department ("the Secretary of State") served him with a notice of intention to make a deportation order.
4. MA appealed against the notice on human rights grounds. The Secretary of State decided (under the Immigration Rules) that the grounds of appeal amounted to a "fresh claim? for asylum; but refused the claim in a letter dated 26 June 2003. MA appealed. His appeal was dismissed by an adjudicator on 25 November 2003. On 5 April 2004, the Secretary of State made a deportation order, which was served on MA on 19 April 2004. On 4 March 2005, MA's solicitors made further representations to the Secretary of State, who decided that these did not amount to a "fresh claim". Removal directions were set for 29 November 2006, but MA applied for judicial review, raising issues under Article 3 of the ECHR.
5. Following further submissions, on 1 February 2007 the Secretary of State accepted that MA had made a "fresh claim? for asylum, but refused the claim. MA appealed again.
6. In a determination promulgated on 19 April 2007, the AIT allowed his appeal. They did so on human rights grounds only, as they held that the appellant was precluded by section 72 of the Nationality, Immigration and Asylum Act 2002 from claiming protection under the Refugee Convention, and, by paragraph 339D of the Immigration Rules, from claiming humanitarian protection.
7. The AIT accepted the concession made by the Secretary of State that MA was a member of the Isaaq clan. They also found that he was from Mogadishu, and that his parents were from Hargeisa in Somaliland. After considering evidence about the situation of the Isaaq clan in Mogadishu, they held that the Isaaq in Mogadishu were in the position of a minority clan who did not have protection, and that he would be at a real risk of physical violence which crossed the Article 3 threshold.
8. The Secretary of State applied for an order requiring the AIT to reconsider their decision. An order for reconsideration was made on 10 May 2007. At the first-stage reconsideration hearing on 28 February 2008, Senior Immigration Judge Spencer ordered a second-stage hearing at which the appeal would be determined afresh. He further ordered that the limited positive credibility findings made by the AIT about MA and their decision to prefer the evidence of Mr Höhne (MA's expert) to that relied on by the Secretary of State should be preserved.
9. MA's appeal was re-heard on 18 December 2008, and in a determination promulgated on 1 July 2009 the AIT dismissed the appeal. His application for permission to appeal to the Court of Appeal was granted by Sedley LJ on 18 December 2009. In a judgment delivered on 23 April 2010 [2010] EWCA Civ 426, the Court of Appeal allowed the appeal.
10. It will be necessary to examine parts of the AIT's determination of 1 July 2009 and the decision of the Court of Appeal in some detail. In short, the Court of Appeal held that, although the AIT directed themselves "impeccably", they did not apply that direction "properly" and they failed to take account of a material factor in reaching their conclusion.

The relevant country guidance decision of the AIT

11. The relevant country guidance for Somalia is to be found in the AIT decision of AM and AM (armed conflict: risk categories)(Somalia) [2008] UKAIT 00091. At para 178, the AIT said:

"On the present evidence we consider that Mogadishu is no longer safe as a place to live for the great majority of its citizens. We do not rule out that notwithstanding the above there may be certain individuals who on the facts may be considered to be able to live safely in the city, for example if they are likely to have close connections with powerful actors in Mogadishu, such as prominent businessmen or senior figures in the insurgency or in powerful criminal gangs. However, barring cases of this kind, we consider that in the case of persons found to come from Mogadishu who are returnees from the UK, they would face on return to live there a real risk of persecution or serious harm and it is reasonably likely, if they tried staying there, that they would soon be forced to leave or that they would decide not to try and live there in the first place."

The Standard of Proof

12. It was not contended in the Court of Appeal or in this court that the AIT had applied the wrong standard of proof. It is well established that a breach of Article 3 of the ECHR is proved "where substantial grounds have been shown for believing that the person concerned faced a real risk of being subjected to torture or inhuman or degrading treatment" (Vilvarajah v UK (1991) 14 EHRR 248 para 103) (emphasis added). There was, however, some brief discussion before us on the question whether it is appropriate to apply the civil test of the balance of probabilities to some of the elements of what has to be proved in an Article 3 claim. This is a difficult topic which has occupied the attention of our courts in recent years in the analogous context of extradition and Refugee Convention cases.
13. It was authoritatively decided by the House of Lords in R v Secretary of State for the Home Department, ex p Sivakumaran [1988] AC 958 that in order for a fear to be well-founded for the purposes of the Refugee Convention, there must be a reasonable degree of likelihood that the applicant will be persecuted on return. It will be seen that this test is expressed in slightly different terms from the Article 3 test. But no argument was addressed to this court to suggest that there is a material difference between the two. Although it is not necessary for the determination of this appeal to decide whether there is any difference, we are inclined to the view that there is no practical difference between them. It would add considerably to the burdens of hard-pressed immigration judges, who are often called upon to decide claims based both on the Refugee Convention and the ECHR at the same time, if they were required to apply slightly different standards of proof to the same facts when considering the two claims.
14. The question that was touched on in argument is whether the same standard of proof should be applied in relation to the proof of past or existing facts as in relation to the assessment of future risk. In the extradition context, in Fernandez v Government of Singapore [1971] 1 WLR 987, the House of Lords had to interpret section 4(1)(c) of the Fugitive Offenders Act 1967, which entitled the applicant to avail himself of a prohibition on return if he "might" be restricted or detained if extradited. Lord Diplock said that the "balance of probabilities" was a convenient phrase to use in relation to the existence of facts; but was inappropriate when applied not to ascertaining what had already happened, but to prophesying what, if it happened at all, could only happen in the future (994A). In the latter situation, Lord Diplock found that a lesser degree of likelihood was sufficient (994G).
15. Prior to Sivakumaran, it seems that the general view in extradition and asylum cases was that past and existing facts should be determined according to the civil standard of proof (ie on the balance of probabilities); and the lower test propounded in Fernandez applied to assessing the risk of adverse treatment on the basis of those facts. An example of this approach is to be found in R v Immigration Appeal Tribunal, ex p Jonah [1985] Imm AR 7 (Nolan J).
16. Following Sivakumaran, it was unclear whether the "real risk/real possibility" test should be applied to the proof of past and existing facts. In Kaja v Secretary of State for the Home Department [1995] Imm AR 1 (IAT), the majority rejected a two-stage test of a determination of past and present facts on the balance of probabilities and an assessment of real risk in relation to future possibilities. They held that the test of reasonable degree of likelihood should be applied to all aspects of the determination. Following Kaja, the practice of the IAT was to apply the "real possibility" test to past and present facts. In Horvath v Secretary of State for the Home Department [2000] INLR 15, however, the Court of Appeal, obiter, favoured the Jonah approach. Although Horvath was appealed to the House of Lords, nothing was said by their Lordships as to the correctness of these observations.
17. The Court of Appeal considered the issue fully in the asylum context in Karanakaran v Secretary of State for the Home Department [2000] 3 All ER 449. We do not propose to examine the judgments of Brooke and Sedley LJJ in detail (with both of which Robert Walker LJ, as he then was, agreed). They endorsed the approach of the majority in Kaja. The degree of probability of the occurrence or non-occurrence of past events was no more than a relevant factor to be taken into account in deciding whether there was a well-founded fear of persecution. The decision-maker was bound to take account of all material considerations when making its assessment about the future.
18. GM (Eritrea), YT (Eritrea) and MY (Eritrea) v Secretary of State for the Home Department [2008] EWCA Civ 833 (which we shall refer to as "GM (Eritrea)") was a group of three asylum cases which we shall consider in some detail later in this judgment in the context of the question of the relevance of lies. But in relation to the standard of proof, it may be worth recording that the Court of Appeal stated that the applicants had to do no more than prove that there was a reasonable degree of likelihood that the past facts that they asserted (viz that they had left Eritrea illegally) were true.
19. This is consistent with the approach adopted by the Grand Chamber of the ECtHR in relation to Article 3 claims in Saadi v Italy (App no 37201/06, 28 February 2008, BAILII: [2008] ECHR 179):

"132. In cases where an applicant alleges that he or she is a member of a group systematically exposed to a practice of ill-treatment, the Court considers that the protection of Article 3 of the Convention enters into play when the applicant establishes …. that there are serious reasons to believe in the existence of the practice in question and his or her membership of the group concerned …" (emphasis added).

20. Nevertheless, the approach in Jonah and Horvath to the ascertainment of past facts may also be seen as consistent with the requirement for "substantial grounds" or "serious reasons". The argument before us, however, proceeded on the basis that "real possibility" was the correct test to apply to past and present facts both in Refugee Convention and Article 3 cases. Without deciding the point, we are content to do the same in this appeal. We express no view on the issue which is both difficult and important. We think it would be desirable for the point to be decided authoritatively by this court on another occasion.

Relationship between lack of credibility and the assessment of risk

21. For appellants who appeal to the AIT in Refugee Convention or Article 3 cases, the stakes are often extremely high. The consequences of failure for those whose cases are genuine are usually grave. It is not, therefore, surprising that appellants frequently give fabricated evidence in order to bolster their cases. The task of sorting out truth from lies is indeed a daunting one. It is all too common for the AIT to find that an appellant's account is incredible. And yet there may be objective general undisputed evidence about the conditions in the country to which the Secretary of State wishes to send the appellant which shows that most of the persons who have the characteristics of, or fall into the category claimed by, the appellant would be at real risk of treatment contrary to Article 3 of the ECHR or persecution for a Refugee Convention reason (as the case may be), but that a minority of these, because of special circumstances, are not subject to such risk. How should the AIT approach such general evidence where they do not believe the evidence given by the appellant that bears on the question of whether such special circumstances apply? That was the problem which confronted the AIT in the present case. The Secretary of State wished to return MA to Somalia. This involved sending him to Mogadishu. The objective evidence about conditions in Somalia was that only a person who had close connections with powerful actors (such as prominent businessmen or senior figures in the insurgency or in powerful criminal gangs) was likely to be safe if returned to Mogadishu. MA gave a great deal of conflicting evidence to the effect that he had no connections in Mogadishu at all. The AIT found that he had not told them the truth about his links and circumstances in Mogadishu (para 109). But they were unable to find positively that he did have connections there, still less that he had close connections with "powerful actors".
22. A similar problem arose in GM (Eritrea). The undisputed objective evidence in these cases was that there was a reasonable likelihood that a person who left Eritrea illegally would be persecuted on return. The question for the AIT was whether there was a reasonable likelihood that the appellants had left Eritrea illegally. Each of the appellants gave an account about his or her exit from Eritrea which was almost entirely disbelieved. The issue was whether, in those circumstances and in the light of the objective background evidence, the AIT had been entitled to find that it was not reasonably likely that the appellants had left illegally. The objective evidence was that more people left Eritrea illegally than legally, but that there were classes of people who could leave legally (party activists, Ministers and ex-Ministers, persons over 40 who wished to visit relatives or go on Haj and government officials), and that those classes were not closed (eg it included those who had obtained student visas).
23. All three members of the Court agreed that two of the appeals (those of GM and YT) should be dismissed. In these cases, Buxton LJ (with whom Laws and Dyson LJJ agreed) stated:

"39. …. While they are unlikely to have fallen into any of the categories reported in para 9 above, they were of an age to have moved into the student category envisaged by the AIT. Since they put forward no truthful material about what they were doing in the relevant period, it is in my view impossible to say that there is a reasonable degree of likelihood that during that period the appellants did not move into the student category.

40. At the same time, it is equally impossible to say that it is likely that they did enter that category. That however is not the test. Mr Nicol was wrong in suggesting that it was for the Secretary of State to produce evidence to that effect. That would indeed be to reverse the burden of proof. As this court put it in Ariaya and Sammy v Secretary of State for the Home Department [2006] EWCA Civ 48, cited in para 12 above, it may not be necessary for the appellant in such circumstances to say much, but he must say something, adduce some evidence that puts him in a vulnerable position, before the effective burden of contradicting his case passes to the Secretary of State."

24. The third appeal concerned MY. The AIT found that MY (a 17 year old girl) had failed to show that she had left Eritrea illegally, because she had not given credible evidence as to how she had left the country. The objective evidence showed that there were categories of 17 year old girls who were allowed to leave the country legally. Buxton LJ would have allowed the third appeal, but Laws and Dyson LJJ agreed that it, too, should be dismissed.
25. In assessing the argument in the case of MY, Buxton LJ said (para 43) that her age alone made it very difficult indeed, even arguably, to fit her into any of the categories of person who might obtain exit visas, including the student category. That being so, he held (para 44) that the immigration judge should have considered, on the basis of all the evidence, "whether there was a reasonable degree of likelihood that during her residence in Eritrea MY did not fall into one of the categories that could or might leave the country legally". He said (para 45) that "the failure of the [evidential] case advanced by the appellant does not lead as a matter of necessity to the failure of her case if there is other evidence of general circumstances or probabilities against which what little is known about the applicant can be assessed". Finally, at para 46, he expressed himself in these terms:

"The evidence referred to above, and despite MY's failure to give truthful evidence either about her activities in Eritrea or about her actual exit from that country, drives me to the conclusion that even though I cannot say how MY actually left Eritrea, there must, if only by elimination of other possibilities, be a reasonable likelihood that she left illegally."

26. Laws LJ, with whom Dyson LJ agreed, said that the "concrete question for the immigration judge was whether there was a reasonable degree of likelihood that MY had left Eritrea illegally" (para 51); there may be cases where the appellant's testimony is disbelieved but other evidence proves his/her asylum claim (para 52). He continued:

"53. In my judgment that circumstance poses great difficulties for MY's case. The fact (if it be so) that it is reasonably likely that any 17 year old girl from Eritrea, about whom nothing else relevant is known, left the country illegally does not entail the conclusion that this particular 17 year old girl did so. The reason is that the probability that a particular person has or has not left illegally must depend on the particular facts of her case. Those facts may produce a conclusion quite different from that relating to illegal exit by members of such a class of persons about whose particular circumstances, however, the court knows nothing more than their membership of the class. There may indeed be a general probability of illegal exit by members of the class; but the particular facts may make all the difference. I think with respect that this consideration lies behind the observations approved by Richards LJ in Ariaya and Sammy v Secretary of State for the Home Department [2006] EWCA Civ 48, and para 449 in MA, which Buxton LJ cites at paras 12 and 13.

54. The position would only be otherwise if the general evidence was so solid as to admit of only fanciful exceptions; if the court or tribunal concluded that the 17 year old must have left illegally whatever the particular facts."

27. He then applied this approach to the facts of MY's case. At para 57 he said that, since her account of her departure had been rejected by the immigration judge, her claim could not succeed on the basis of "general" evidence unless "the possibility that the particular facts may make a difference is effectively excluded".
28. Dyson LJ agreed that MY's appeal should be dismissed substantially for the reasons given by Laws LJ. At para 61, he said:

"Unless it can safely be said that exit by any 17 year old girl is illegal, whether it is reasonably likely that the exit by an individual 17 year old girl was illegal will depend on the facts of her particular case. Her failure to give a credible account of those facts may lead to the conclusion that she has not shown that there is a reasonable likelihood that her exit was illegal. "

29. Like Laws LJ, he concluded on the basis of the general evidence that "it was entirely possible that MY left Eritrea legally" (para 64).
30. The appeal to this court has been conducted on the basis that the approach adopted by Laws and Dyson LJJ is substantially correct. But Mr Drabble questioned para 54 of Laws LJ's judgment. We think that what Laws LJ had in mind was a case where (i) the claimant's account is rejected as wholly incredible (it is riddled with contradictions and the tribunal is left in a state of being unable to believe anything that the claimant has said); but (ii) there is undisputed objective evidence about conditions in the relevant country which goes a long way to making good the shortcomings in the claimant's own evidence. In GM (Eritrea), for example, the AIT did not believe the account given to them by MY as to how she had left the country. They could not, therefore, rely on her account as a basis for concluding that she had left the country illegally. But if there had been objective evidence that no 17 year old girls were allowed to leave the country, her appeal would surely have succeeded despite her dishonest evidence. In fact, the objective evidence did not go nearly that far and the appeal was dismissed.
31. What Laws LJ was saying at para 54 was that, where a claimant tells lies on a central issue, his or her case will not be saved by general evidence unless that evidence is extremely strong. It is only evidence of that kind which will be sufficient to counteract the negative pull of the lie. But much depends on the bearing that the lie has on the case. The Court of Appeal correctly stated at para 104 of its judgment in the present case:

"The lie may have a heavy bearing on the issue in question, or the tribunal may consider that it is of little moment. Everything depends on the facts. For example, if in the Eritrea cases the Secretary of State had prima facie evidence that the appellants had left legally, the tribunal might think it appropriate to put considerable weight on the fact that the claimant told lies when seeking to counter that evidence. The lie might understandably carry far less weight where, as in YL itself, the judge is satisfied that the appellant has lied where the lie is against her interests."

32. Where the appellant has given a totally incredible account of the relevant facts, the tribunal must decide what weight to give to the lie, as well as to all the other evidence in the case, including the general evidence. Suppose, for example, that at the interview stage the appellant made an admission which, if true, would destroy his claim; and at the hearing before the AIT he withdraws the admission, saying that his answer at interview was wrongly recorded or that he misunderstood what he was being asked. If the AIT concludes that his evidence at the hearing on this point is dishonest, it is likely that his lies will assume great importance. They will almost certainly lead the tribunal to find that his original answers were true and dismiss his appeal. In other cases, the significance of an appellant's dishonest testimony may be less clear-cut. The AIT in the present case was rightly alive to the danger of falling into the trap of dismissing an appeal merely because the appellant had told lies. The dangers of that trap are well understood by judges who preside over criminal trials before juries. People lie for many reasons. In R v Lucas [1981] QB 720, the Court of Appeal had to consider whether a statement containing a lie was capable of amounting to corroboration. At p 724F, Lord Lane CJ said:

"To be capable of amounting to corroboration the lie told out of court must first of all be deliberate. Secondly, it must relate to a material issue. Thirdly, the motive for the lie must be a realisation of guilt and fear of the truth. The jury should in appropriate cases be reminded that people sometimes lie, for example, in an attempt to bolster up a just cause, or out of shame or out of a wish to conceal disgraceful behaviour from their family…."

33. Although the analogy is not exact, it is close enough for these words to be of relevance in the present context. So the significance of lies will vary from case to case. In some cases, the AIT may conclude that a lie is of no great consequence. In other cases, where the appellant tells lies on a central issue in the case, the AIT may conclude that they are of great significance. MA's appeal was such a case. The central issue was whether MA had close connections with powerful actors in Mogadishu. The AIT found that he had not told the truth about his links with Mogadishu. It is in such a case that the general evidence about the country may become particularly important. It will be a matter for the AIT to decide whether the general evidence is sufficiently strong to counteract what we have called the negative pull of the appellant's lies.

The AIT's determination in more detail

34. This was a second stage reconsideration at the behest of SIJ Spencer who had decided that there was a material error of law in the earlier decision in "failing to consider whether the appellant would be able to make arrangements for protection by the Hawiye through his connections with the Isaaq". The immigration judge added: "I take the view that in this regard the appellant had at least an evidential burden, which he failed to discharge, since what contacts and relationships he had was exclusively within his own knowledge and not that of the Secretary of State". Thus the nature of those connections was of central importance at the second stage reconsideration.
35. At para 17 of their determination, the AIT recorded that it was for MA to prove his case. He had to show that there was a real risk of his suffering ill treatment of such severity that his Article 3 rights would be breached. MA gave evidence to the tribunal. He did not need an interpreter. He gave conflicting accounts about his connections with Mogadishu. For example, para 31 of the determination records that his SEF form showed that his family tried to leave Somalia for Kenya, but had found that too risky. His evidence before the tribunal was that his family left Somalia for Kenya together in 1992 or 1993. But para 45 of the determination states that in his application form he said that he left his family behind in Somalia when he fled to Kenya in 1994. Para 48 records that MA said that he left his family behind in Somalia, but they followed him to Kenya later. Para 54 refers to his interview on 12 September 2002 when he said that, when he left Kenya for the UK, he believed that his family was in Somalia, although he did not know whether they were safe there. This is to be contrasted with his witness statement dated 27 July 2008 in which he said that he last saw his parents in Kenya when he was leaving for the UK.
36. He said that his mother was in South Africa and his father divided his time between South Africa and Kenya. He accepted that he had spoken to his sister about their parents while he was in prison. He said that she had told him that she did not know where they were. And yet in his witness statement he had said that he was in contact with his sister while he was in prison and she had told him that their parents were fine, but did not say where they were. Unsurprisingly, the AIT found that it was so unlikely that his sister would know that her parents were fine but not know where they were, that they found MA's account as to what he knew about the whereabouts of his parents incredible. MA told the AIT that he had not been in contact with his parents because he did not know how to contact them, but he was in contact with his sister. He gave no satisfactory answer as to why he did not contact his parents.
37. The AIT then reviewed the expert evidence of Marcus Höhne on behalf of MA. At para 88, they said:

"[W]e emphasise that although we have indicated our findings when we have considered the submissions, and, in places, the evidence, we did not make any findings until we had considered the entirety of the evidence in the round, with the submission made."

38. At para 104, they said that the submission made on behalf of the Secretary of State that MA was "not believable and could not prove his case" was "wholly justified". They continued:

"105. The Tribunal is not unfamiliar with the difficulties created by appellants who have not been truthful but who still may be at risk. This was considered by the Court of Appeal in GM (Eritrea) v Secretary of State for the Home Department [2008] EWCA Civ 833. We must be very careful not to dismiss an appeal just because an appellant has told lies. Even if very large parts of his story have been disbelieved it is still possible that the appellant has shown that he would be at risk on return. An appellant's own evidence has to be considered in the round with other evidence and that can include unimpeachable evidence from expert reports or country guidance cases or other evidence about the general state of affairs in that country.

106. For reasons very properly emphasised by Mr Drabble and spelled out in the case of AM, Mogadishu is a desperately difficult place and it is probably not going too far to say that the respondent should think twice before making anybody go there against his or her will.

107. We do not believe the appellant but we have to decide if the background conditions show that he will be at risk. In para 178 of AM the Tribunal identified people as examples of people who may be safe in Mogadishu. The Tribunal said: "We do not rule out that notwithstanding the above there may be certain individuals who on the facts may be considered to be able to live safely in the city, for example if they are likely to have close connections with powerful actors in Mogadishu, such as prominent businessmen or senior figures in the insurgency or in powerful criminal gangs. However, barring cases of this kind, we consider that in the case of persons found to come from Mogadishu or returnees from the UK, they would face on return to live there a real risk of persecution or serious harm..."

108. We are not able to find positively that the appellant is a person with close connections with powerful actors in Mogadishu and so on.

109. The difficulty is that the appellant has not told us the truth about his links and circumstances in Mogadishu and we cannot exclude the possibility that he is a person with connections of this kind. The point is that it is not fanciful to say that he would not necessarily be at risk on return. Some people are not. Even though the appellant has to prove only a real risk rather than a probability of him being at risk we cannot make the necessary findings when he will not tell the truth about his connections and contacts there.

……………

118. [Mr Drabble] submitted that it was fanciful to suggest that the appellant came into any of the categories of people identified at para 178 of AM as being not at risk. The appellant would be at risk in Mogadishu and his appeal should be allowed.

119. Drawing all these things together we find that this appellant would be returned to Mogadishu. We find that he has links with Somaliland and would probably be accepted by the community there if he could get there. We find that it would be dangerous in fact too dangerous for him to travel from Mogadishu to Somaliland if he is telling the truth when he claims not to have any links with the country. We accept that some people do make the journey. There is no clear evidence about how they travelled. We cannot find that this appellant could follow their example. It is clearly the case (and no one has suggested otherwise) the appellant would be allowed through the airport at Hargeisa. Whilst we accept that the risks diminish as a traveller gets further away from Mogadishu a traveller has to get away from Mogadishu before that becomes an advantage and there are risks travelling around there. This appellant is going to be at risk if we accept his evidence of having no contacts there.

120. We do not accept his evidence about that. He was manifestly untruthful. We have reflected carefully on this because we are aware of the time that has elapsed, of which a full explanation has been given, since the appeal was heard. It is not a matter of nuance or inference. The appellant is blatantly untruthful and no passage of time has impacted our findings on that point.

121. Para 178 of AM does not give an exclusive list of people who are not at risk. It makes the point there are people who are not at risk. The burden is on the appellant and he has not told the truth about his links with Mogadishu and we are not able to say that he is a person who has shown he would be at risk there. He has stopped proper enquiry of a kind that might reveal the links and protection he would have. It would be very sad if, by so doing, the appellant has deprived himself of protection that he would otherwise need but he has told lies and must accept the consequence of that. It does diminish his credibility and makes it harder for him to prove his case.

122. In all the circumstances we dismiss the appeal."

Decision of the Court of Appeal

39. The court summarised the two submissions made by Mr Drabble. The first was that the AIT could not properly have concluded in the light of the country guidance given in AM and AM, as they did at para 109, that it was not "fanciful" to say that MA may not be at risk on return. The court rejected this submission. At para 110 of their judgment, they said that the AIT were entitled on the basis of AM and AM to conclude that this case "did not fall into the category of case identified by Laws LJ in GM where the general evidence would suffice, because anyone in MA's situation would necessarily be subject to persecution on return."
40. The second submission was that the AIT had misdirected themselves when considering the question of risk on return. They had focused on the difficulties caused by MA's failure to tell the truth, but they should have asked whether there was other evidence relating to MA's own particular situation, even if his own rejected testimony was left out of account, which would support his case. Mr Drabble submitted that there was such evidence which the Tribunal failed properly to evaluate. This was summarised by the Court of Appeal at para 113 in these terms:

"The evidence he relies upon is in particular the fact that the appellant has been in the UK for some 15 years, and that for almost all of the last 12 or so he has been in detention of one sort or another. In addition, his parents were from Hargeisa, not Mogadishu, and the evidence of Mr Höhne was to the effect that he would not get protection from the Isaaq clan in Mogadishu given the dramatic evacuation from that city. In the circumstances, Mr Drabble submits that it is fanciful to think that the appellant would be likely to fall into the exceptional category of persons with contacts in Mogadishu who could provide the requisite protection."

41. At para 116, the court acknowledged that the AIT had directed themselves impeccably at para 105 of the determination. But they said that the tribunal had not properly applied that direction. They continued:

"117. We think, with respect to the Tribunal, that it is there adopting the wrong approach. Their analysis suggests that the fact that the appellant has lied has of itself disabled them from reaching a conclusion on the article 3 risk. They seem to be throwing up their hands in despair; since the appellant has concealed the truth, they cannot make any necessary findings. This is further confirmed by para 121 when they say that because his lying has prevented a full and proper inquiry, there is no relevant finding the Tribunal can make.

118. That does not, however, follow from GM. They first have to ask whether there is other evidence, independently of his unreliable testimony, casting light on the appellant's particular situation. If so, they must have regard to that evidence. As Buxton LJ put it in GM see para 98 above), there does not need to be much evidence, only sufficient to suggest that there is a real risk of persecution and thereby shift the burden to the Secretary of State to show otherwise. Nowhere does the Tribunal say that the only potential evidence is the appellant's rejected testimony and that without it there is no relevant evidence, and we do not think that it can be fairly inferred from their decision that this was how they approached the matter. For example, there is no reference in the whole judgment to the fact that the appellant has spent the best part of the last 12 years in prison or administrative detention in the UK. In our view that must on any view have relevance to the likelihood of this particular appellant having current contacts in Mogadishu which will afford him the necessary protection.

119. In any event, in our judgment, if they did analyse the issue in that way, we agree with Mr Drabble that it was not a conclusion open to them on the evidence. That evidence was that the appellant was from a clan which was in the minority in Mogadishu; that he had not been there for some 15 years; and that for most of that time he had been in detention. Whatever links might exceptionally exist to provide protection for an Isaaq returning to Mogadishu, there was in our view sufficient evidence adduced before the Tribunal at least to establish a real risk that it was unlikely to apply to him. He was not simply putting himself into the general category of persons returning to Mogadishu, nor even of a minority clan member taking that step, and then relying on the relevant statistics as to how such persons would in general be treated. There was the particular feature of his history in the UK -the lengthy period and the fact of detention - which constituted evidence relevant to the particular and specific risks which he faced and which enabled the court to make an assessment of risk on the basis of evidence independent of his own testimony.

120. We agree that the Tribunal ought to have made an assessment on the basis of that evidence, and had they done so, they must have concluded that there was a real risk that he would not obtain the relevant protection. Without it, in the light of AM and AM he was plainly at risk of adverse article 3 treatment, and therefore his deportation would be unlawful."

Should the Court of Appeal have interfered in this case?

42. It is important to note what the Court of Appeal did not say. They did not say that the AIT had misdirected themselves as to the correct test to be applied whether in relation to Article 3 cases generally or as to the impact of lies. They accepted that the AIT were right to direct themselves in accordance with the majority in GM (Eritrea). They acknowledged that the AIT did not commit the cardinal error of dismissing the appeal simply because MA had told lies. The error they identified in the AIT's approach was in relation to the application of GM (Eritrea). In summary, they made two criticisms. First, they said that paras 109 and 121 of the AIT's determination showed that they did dismiss the appeal simply because MA had told lies. As they put it at para 117 of their judgment, "[t]hey seem to be throwing up their hands in despair; since the appellant has concealed the truth, they cannot make any necessary findings". Secondly, they said that the AIT overlooked significant material which enabled the court to make an assessment of risk independent of MA's testimony and which, if it had been taken into account, must have led to the conclusion that there was a real risk that he would not obtain the relevant protection.
43. Before we examine these two criticisms, we need to make some general points about the proper role of the Court of Appeal in relation to appeals from specialist tribunals to it on the grounds of error of law. Although this is not virgin territory, the present case illustrates the need to reinforce what has been said on other occasions. The court should always bear in mind the remarks of Baroness Hale of Richmond in AH (Sudan) v Secretary of State for the Home Department [2007] UKHL 49, [2008] 1 AC 678 at para 30:

"This is an expert Tribunal charged with administering a complex area of law in challenging circumstances….[T]he ordinary courts should approach appeals from them with an appropriate degree of caution; it is probable that in understanding and applying the law in their specialised field the Tribunal will have got it right….They and they alone are judges of the facts…Their decisions should be respected unless it is quite clear that they have misdirected themselves in law. Appellate courts should not rush to find such misdirections simply because they might have reached a different conclusion on the facts or expressed themselves differently."

44. Those general observations were made in a case where the Court of Appeal had allowed an appeal against a decision of the AIT. The role of the court is to correct errors of law. Examples of such errors include misinterpreting the ECHR (or in a refugee case, the Refugee Convention or the Qualification Directive); misdirecting themselves by propounding the wrong test on some legal question such as the burden or standard of proof; procedural impropriety such as a breach of the rules of natural justice; and the familiar errors of omitting a relevant factor or taking into account an irrelevant factor or reaching a conclusion on the facts which is irrational.
45. But the court should not be astute to characterise as an error of law what, in truth, is no more than a disagreement with the AIT's assessment of the facts. Moreover, where a relevant point is not expressly mentioned by the tribunal, the court should be slow to infer that it has not been taken into account.
46. We turn to the first of the Court of Appeal's criticisms. In our view, the court was wrong to interpret paras 109 and 121 of the determination as if the AIT were saying that they were dismissing the appeal because MA's account was incredible. In the light of the clear and impeccable self-direction set out only a few paragraphs earlier (at para 105), and having regard to the need for restraint to which we have referred, the court should surely have been very slow to reach the conclusion that it did. It should only have interpreted these paragraphs in the way that it did if there was no doubt that this is what they meant. It is often easy enough to find some ambiguity or obscurity in a judgment or determination, particularly in a field as difficult and complex as immigration, where the facts may be difficult to unravel and the law difficult to apply. If, as occurred in this case, a tribunal articulates a self-direction and does so correctly, the reviewing court should be slow to find that it has failed to apply the direction in accordance with its terms. All the more so where the effect of the failure to apply the direction is that the tribunal will be found to have done precisely the opposite of what it said it was going to do. The striking feature of the present case is that the Court of Appeal was of the view that at para 109, the AIT failed to apply the direction that they had set for themselves only four paragraphs earlier.
47. In our view, there was no need to interpret paras 109 and 121 in the way that the Court of Appeal did. There is an interpretation of these paragraphs which is consistent with the self-direction at para 105 and is unimpeachable. In our view, all that the AIT were saying at para 109 was that, because MA had not told the truth about his links and circumstances in Mogadishu, the possibility that he was a person with connections in Mogadishu could not be excluded. In other words, he had not discharged the burden of proof which the AIT had correctly said rested on him. The fact that the AIT were considering the burden of proof is demonstrated by the last sentence of para 109: "Even though the appellant has to prove only a real risk…we cannot make the necessary findings when he will not tell the truth…" (emphasis added). So too at para 121 which we have set out at para 38 above. The third sentence says: "The burden is on the appellant and he has not told the truth about his links with Mogadishu…". Later in the paragraph, the AIT says that MA has told lies: "It does diminish his credibility and makes it harder for him to prove his case" (emphasis added).
48. In our view, on a fair reading of paras 109 and 121 in the light of para 105, it is clear that the AIT were not throwing up their hands and rejecting MA's appeal because he had lied without more. They were saying that, because he had told lies, they were unable to make any relevant findings and the appeal failed because MA had not discharged the burden of proof.
49. We turn to the second criticism. The first limb of this criticism is that the AIT overlooked the fact that MA had spent the last 12 years in prison and administrative detention in the UK. It is true that there is no explicit reference to this fact in the determination. But the AIT were well aware of it. As we have already said, they considered the extent of his contact with his family when he was in prison during this period and set out MA's conflicting evidence at length at paras 35 to 65 of the determination. There are various references to his having been in prison (paras 35, 40, 59, 60, 61 and 64). In these circumstances, there is no warrant for holding that the AIT failed to have regard to the fact that MA was in custody or detention. They said in terms at para 88 that they had considered "the entirety of the evidence in the round". This was a detailed and careful determination running to 122 paragraphs. For that reason, as well as because of the need for restraint to which we have referred, the court should have been very slow to reach the conclusion that the AIT had not taken into account the fact that MA was in custody and detention for 12 years. It was obvious from his own evidence that MA was able to communicate with some members of his family, presumably by using his mobile phone. In these circumstances, the tribunal is likely to have thought that the fact that MA was deprived of his liberty for 12 years would not prevent him from maintaining his connections with the outside world. It is of some significance that in his detailed skeleton argument for the appeal to the AIT, Mr Drabble made the point at paras 20 to 22 that MA's case was that he had no knowledge of or contact with his family or friends in Somalia. But he did not say that MA would be unlikely to have such knowledge or contact because he was in prison and then administrative detention.
50. This brings us to the other limb of the second criticism made by the Court of Appeal, which is that, if the AIT had taken into account the fact that MA had been in prison and detention for 12 years, "they must have concluded that there was a real risk that he would not obtain the relevant protection." If this had been such an inevitable conclusion to draw from the fact of custody and detention for 12 years, it is indeed surprising that Mr Drabble did not invite the AIT to make it in his skeleton argument. Although they did not use the language of perversity, what the Court of Appeal were saying in effect was that it would have been perverse of the AIT not to draw this conclusion if they had thought about the significance of the custody and detention. In our view, this is quite untenable. It was possible that the deprivation of liberty for 12 years would have prevented MA from maintaining or developing the necessary protective links. But it was certainly not inevitable. It was for the AIT to assess the matter in the light of all the evidence.

Conclusion

51. We would, therefore, allow this appeal. The AIT did not adopt the wrong approach in their assessment of the impact of MA's lies and there was no error of law in their determination which warranted interference by the Court of Appeal.