The decision

Field House (video-link to Manchester)

AS (Kirundi/Buyenzi – “country expert” evidence) Burundi [2005]
UKAIT 00172

Heard: 24.11.2005
Signed: 28.11.2005
Sent out: 08.11.2005


John Freeman (a senior immigration judge)
Mrs S Hussain JP and
Mr HG Jones MBE JP



Secretary of State for the Home Department,

Miss S Khan (counsel instructed by Howells, Sheffield) for the appellant
Mrs L Singh for the respondent


This case is reported on the following points:
a) the approach to be taken to “country expert” evidence (see 4 and 9);
b) the methodology of the ‘Ethnologue’ web-site, and the weight to be given to untested assertions of language skills (see 11);
c) the meaning of “race” in the Refugee Convention (see 14-15);
d) techniques of decision-writing (see 17); and
e) the approach to be taken to Foreign and Commonwealth Office travel advice (see 19).
It is not reported on the following points:
f) what is to be expected of persons claiming to be from Burundi by way of language skills (for which see SJ [2005] UKAIT 00134, explaining Rusiga [2005] EWCA Civ 407; or
g) what is the current situation for Hutus on return to Burundi, and in particular to the Buyenzi area of greater Bujumbura (for which see AM [2005] UKAIT 00123, confirming SS (Burundi) CG [2004] UKIAT 00290).

This is a case where the Home Office were given permission to appeal from a decision of an adjudicator (Mr L Saffer), sitting at Bradford on 3 March 2005, allowing an asylum and human rights appeal by someone who claimed to be a Hutu citizen of Burundi. Under the transitional provisions of the 2004 Act, the case proceeds as if it were a reconsideration following review by the Asylum and Immigration Tribunal; but only (see r. 62.7 of the 2005 Procedure Rules) on the grounds on which permission was given. These were that the adjudicator
a) might have given undue weight to the report of a “country expert” (Professor James Fairhead of Sussex University), who had neither met or spoken to the appellant himself, nor given enough details as to the sources of his knowledge of the languages spoken in Burundi; and
b) might not have given enough reasons for finding that the general situation for Hutus in Burundi had got notably worse since the decision of the Immigration Appeal Tribunal in SS (Burundi) CG [2004] UKIAT 00290.
2. The language point (a) arose because the appellant was unable to speak Kirundi, the national language of Burundi, and common to both Hutus and Tutsis. She did claim to be able to understand it to an extent; but this was never put to the test, as she chose to be interviewed in Swahili, and her “country expert” as we have seen did not interview her at all. Swahili is of course the lingua franca of a number of surrounding countries, and in particular Tanzania. It is common ground however that it is spoken in some areas of Burundi, which for some reason generate what might appear a disproportionate number of the refugee claims made by persons claiming to be citizens of that country. This leads to credibility disputes as to whether such non-Kirundi speakers are citizens of Burundi, as they claim, at all.
3. Such cases have come before the appellate authorities and the courts in reported decisions several times already. In order of time, those referred to before us were: AR [2004] UKIAT 00225; but this was reversed in Agnes Rusiga [2005] EWCA Civ 407 (we should like to express our appreciation for the Court of Appeal’s avoidance of unnecessary anonymity), and the effect of that further explained in SJ [2005] UKAIT 00134. Of these, only AR was available to the adjudicator when he sat, though it had re-appeared as Rusiga by the time review of his decision was granted (which was no doubt why that grant did not extend to the criticism of him for not following it as AR); while SJ appeared only while the present reconsideration was pending. It follows that, contrary to the present position, there was no effective ‘country guidance’ case on the Kirundi problem when the adjudicator reached the decision now under challenge; and that our decision is of no relevance to that general question (on which decision-makers should turn first to SJ), but only to the very limited one of whether he went wrong in law on the information before him.
4. Professor Fairhead is a social anthropologist, who did fieldwork from 1986-89 in Rwanda and the DRC [Democratic Republic of the Congo, as it now is], during which time he also lectured at the University of Rwanda. He says he first visited Burundi in 1981: though we should have very much liked to know the date of his last visit to that country, Mrs Singh did not suggest that he is not to be regarded as an academic expert on it, and we are prepared to accept him as such. That does not of course mean that anything falling from his pen is to be accepted as Gospel; but that decision-makers are entitled to accept his reasoned conclusions on general questions, without detailed sourcing, where those do not go against other information from generally accepted background sources before them, or reported decisions of the Tribunal. (On the other hand, it would in our view be a wrong approach in law not to engage in vigorous critical analysis of “country experts”’ views, where those were out of line with such material).
5. In this case, the adjudicator found (at § 60) “… that it is perfectly feasible for someone from Burundi to speak Swahili as a first language and not to speak Kirundi”. That was much too rosy a general view of the background evidence, even at the date of the hearing, and, to the extent that it was not supported by anything in that, or even by Professor Fairhead, amounted to an error of law. However, that brings us to the crucial question of whether on all the evidence available at the time, including his, that finding did amount to a material error of law by the adjudicator.
6. What Professor Fairhead said about the language of the appellant’s claimed place of origin (starting at § 5 of his report) was this:
…from 1993, as the conflict in Burundi has become more entrenched, Buyenzi became an ‘ethnic ghetto’; a Hutu swahiliphone district with its own militia, defending itself against Tutsi militia and army units.
We shall return to what else Professor Fairhead said about Buyenzi when we come to ground b). The explanation he gives, at § 4, for Swahili having become the language of Buyenzi is through immigration, in the 1930s, mainly from Tanzania and the [then Belgian] Congo; coupled with return, in the 1980s, of those who had sought refuge in those countries in the 1970s. We cannot see how this could lead to acceptance as a Swahili-speaking native of Buyenzi of someone who does not seem to have given any family history of immigration or emigration such as Professor Fairhead describes; and this appellant claimed to have spent no more than a very limited time in a camp in what is now the DRC, between March and April 1994.
7. This is what Professor Fairhead said in his conclusions (at § 13) directed specifically to the languages used in Buyenzi:
…as a resident of Buyenzi, the appellant would speak Kiswahili as their first language, and not Kirundi. Whilst some Buyenzi residents speak Kirundi too, it would not be unusual for some not to speak it, or only to understand it. This is especially the case for the moslem educated residents such as the appellant.
We do not, as we have said, see any basis in Professor Fairhead’s reasoning at §§ 4-5 for this appellant or her family to have been Swahili-speakers in the first place. However, this point was neither pursued before the adjudicator, nor in the grounds of appeal which led to this reconsideration.
8. It follows that the adjudicator had before him a statement by Professor Fairhead, unexceptionable because in accordance with other background evidence, that people spoke Swahili in Buyenzi; and a contention of Professor Fairhead’s own that there would be those there who spoke only Swahili, for which he gave some apparently plausible, and so far unchallenged historical reasons (though our views on their proper application to the facts of this case will be clear). We note that the appellant had claimed to have been educated in an Arabic school (so avoiding any suggestion that she must have learnt Kirundi in the public schooling system). While her knowledge of Arabic has never been put to the test, there was no challenge to this part of her evidence, and the adjudicator was entitled to accept it.
9. A “country expert”, like any other expert witness, may (and to some extent needs to) work on the basis of the history he is given by those instructing him: what he must do is to make it plain what comes from them, and what from his own expertise. Usually in asylum and immigration cases individual history comes from the instructions given, and general background from the “country expert”’s own knowledge: the real expertise comes in weighing the plausibility of one against the other.
10. In this case Professor Fairhead cannot have meant to suggest he took any expert view as to this appellant’s individual linguistic competences, since he made it plain that he had never spoken to her; and there is nothing to show that the adjudicator was under any illusion about that. It was legitimate for Professor Fairhead, as for the adjudicator, to take account of the particular education the appellant claimed to have had; and, as we have seen, the adjudicator was entitled (without any challenge on the basis of the appellant’s lack of a significant family history of immigration or emigration) to accept his views, which were based on apparent expertise, on the languages generally spoken (or not) in Buyenzi.
11. The adjudicator went much too far in what he purported to take from Professor Fairhead’s evidence about the existence of Swahili-only speakers in Burundi generally, as we have explained at 5. The point he took (later in § 60) on the contents of the ‘Ethnologue’ web-site was misconceived: it should have been fairly clear that, while the 6,457,000 population of Burundi was a relatively exact census figure, the 6,000,000 Kirundi-speakers world-wide represented an estimate to the nearest complete million. There was no caution expressed by either Professor Fairhead or the adjudicator as to the appellant’s untested claim to understand, though not speak Kirundi. Professor Fairhead was entitled, not having met her, or indeed holding himself out as an expert in the languages themselves, rather than their social distribution, to accept that as part of his instructions. If it had been crucial to his findings, the adjudicator would have needed at least to make it clear that he himself realized that understanding any Kirundi represented a wholly untested assertion by the appellant herself.
12. However it would have been enough for the adjudicator to find in the appellant’s favour to accept what he correctly noted (at § 31) that Professor Fairhead said about there being people in Buyenzi who did not speak, or (our emphasis) who only understood Kirundi. This, as we have seen at 6 – 10, he was entitled to do on the way the case was presented before him. It follows that we do not see any material error of law in the adjudicator’s decision on the language point (a).
13. Turning to the Hutu situation (point b), the adjudicator correctly summarized the effect of SS 04-290 at § 43. The complaints made on this point in the grounds of appeal (drafted by the presenting officer who appeared before the adjudicator) were that:
5. he failed to identify the “objective evidence” which had led him to conclude that things had got significantly worse since SS 04-290; and
6. he “… identified a convention reason which is at best one based on ethnitiy [sic]”
Reconsideration was only granted on the point at 5: the senior immigration judge who did so was understandably too busy, or too polite to say anything about point 6. Since it may represent a misunderstanding of either the Refugee Convention or the English language on the part of a presenting officer (as well as a regrettable lack of spelling or proof-reading), we had better put in a few words about it. (Perhaps the draftsman only meant to distinguish between a claim based on the appellant’s individual history, and one on her tribal origins, in which case she need not take what we are about to say personally).
14. The Refugee Convention offers protection against persecution amongst other reasons because of a person’s race. That good simple English word covers any kind of tribal or ethnic identity they may be seen as having, and is most certainly not limited to such features as the colour of their skin. It was good enough for the framers of the Convention in 1951, and should be good enough for us now. We strongly suspect that transatlantic history and practice since then has been responsible both for the unnecessary “ethnicity”, and the misconception apparently represented by the presenting officer’s use of it.
15. The great struggle to defeat the colour bar in the United States in the 1950s and 60s led first to the false impression that race was only about colour; and then to the further delusion that not racialism, but the word “race” itself was the evil to be avoided. Persecution for reasons of race was of course the main historical evil against which the 1951 Convention was directed; and if it were thought that a claim can be belittled by describing it as “at best one based on ethnitiy”, that would show a lack of understanding of history, as well as of the words of the Convention.
16. This however has little to do with point b), on which reconsideration was granted. Here the adjudicator was criticized for not giving details of the “objective evidence” which had led him to his conclusion on it. What he had done, quite legitimately in our view, at § 68 was to refer back to it in the words “Having summarised the objective evidence above …”. The draftsman of the grounds of appeal has picked out the words which follow “… it is my judgment that unfortunately the situation in Burundi has deteriorated since [SS 04-290] …” without taking any account of that reference back.
17. This no doubt represents a misunderstanding of a common judicial technique, rather than any deliberate attempt to mislead the Tribunal; but it is a misunderstanding which should not be allowed to persist on either side in cases of this kind. It is enough if there is a proper analysis of evidence where it is set out (in other words one which gives the decision-maker’s own reasoned views on the relevant points), and a clear reference back to that in the conclusions, particularly (where relevant) on any credibility findings. So long as all that is done, no-one should try to blame decision-makers for not repeating in their conclusions an analysis already set out. For our part we should go further and say that any unnecessary repetition is so tedious as to be positively wrong in a class of case where decisions are sometimes unavoidably lengthy in any event.
18. It follows that what we are concerned with is whether the adjudicator’s analysis of the “objective evidence”, set out at §§ 53-55, 63 and 68, and including his references to Professor Fairhead’s evidence at §§ 32, was a proper one. This is not an exercize on which we (or the senior immigration judge who granted reconsideration) should have been invited to set out without any details of what was complained of (because of the misunderstanding referred to at 17); but because reconsideration was granted on this point we shall deal with it, as briefly as possible.
19. We have to say we discount what the adjudicator said about Foreign and Commonwealth Office advice not to travel to Burundi at § 54. Decision-makers are entitled to refer to such advice; but not in our view without reminding themselves that it is directed to citizens of this country, who have a free choice as to whether they travel to other ones, or not. It should be remembered that asylum-seekers are only entitled to be here as a matter of necessity, and not of choice. While the Foreign and Commonwealth Office advice, seen in that light, is of general background relevance to the conditions in the country in question, it could not on its own have entitled the adjudicator to take the view that there had been a significant change in those since the “country guidance” decision before him.
20. On the other hand, the adjudicator gave proper details at § 55 of the Human Rights Watch report of 13 January 2005, and the UNHCR one of 1 February. It might perhaps have been argued that these came too soon after the decision in SS 04-290 (issued 29 October 2004) for what they contained to represent a recent worsening of the position of Hutus (as opposed to a different view of the situation already considered in SS 04-290). However, the existence of these reports does make it clear that what Professor Fairhead had to say on this point was by no means out of line with the background evidence; so the adjudicator was entitled to rely on his summary of Professor Fairhead’s remarks at § 32, and to agree with them without giving further reasons, as he did at § 68.
21. So that those who read this can see what Professor Fairhead said, we shall pick out the salient points of it.
5. … whilst Buyenzi was still generally acknowledged as an ethnically mixed neighbourhood in 1995, the repeated waves of ethnic cleansing resulted in it acquiring a steadily stronger Hutu identity from then on. By the end of the 1990s it was a Hutu ghetto, subject to night-time terror attacks by Tutsi militias, and also by daytime reprisal and looting raids by Hutu insurgents from the nearby hills.
6. … In late July 2002, Buyenzi became an ethnic battleground. Since then it has remained at the centre of the Burundian war/peace process.
7. …
8. …
9. On surface impressions, Burundi is current [sic] in a period of peace-building and rapprochement between warring parties [but such attempts have come and gone before]. Moreover the fighting continues in the appellant’s home areas.
10. The continuing level of conflict in Burundi means that there is a very real risk that the appellant would suffer persecution due to her ethnicity whether by state agents acting in an unofficial capacity, or by ethnic militia that the authorities cannot control. …
23. We are very far from saying that we should have taken the same view on the situation on Professor Fairhead’s evidence as the adjudicator did; but in our view it is enough to cite the summary with which the recent decision of the Tribunal in AM [2005] UKAIT 00123 begins, to see that it remains a possible view in individual cases:
This case is reported to provide confirmation of the general guidance given in SS (Burundi) CG [2004] UKIAT 00290. Individual appellants who come from particular areas may still establish a well-founded fear of return.
As it happened, the area from which the appellant in AM 05-123 happened to come was once again Buyenzi, and on this basis the Tribunal quashed the adjudicator’s decision and allowed the original appeal. They made clear that their decision turned on its own facts; but those did of course include a good deal of background material.
24. It follows that we have reached the view that the adjudicator’s decision, both on point a) as to what the language spoken by the appellant had to say about her origin, and on point b), as to the general risk faced by Hutus in her home area, is one which he was entitled to reach at the time he wrote, and for the reasons he gave. It does not represent what we think the current approach to the factual questions in such cases should be, for which see SJ 05-134 (explaining Rusiga) and AM 05-123 (confirming SS 04-290); but the adjudicator was faced with a situation in which that guidance did not exist, and in our view made a determined and energetic attempt to grapple with the problems that posed, without running into any material error of law. There was a time when there was reason to suspect some adjudicators of making inadequately reasoned decisions to allow appeals as an easy way out of such problems; but that was very much not the case here.
The original Tribunal did not make a material error of law and the original determination of the appeal stands.

John Freeman
approved for electronic distribution
28 November 2005