LORD JUSTICE BROOKE: This is an appeal by an asylum seeker from Burundi from a decision of an immigration judge on the reconsideration of her case made on 18 April 2005 whereby he dismissed her appeal from a refusal of the Secretary of State on 2 July 2003 to grant her asylum. An earlier decision by the adjudicator had been set aside by the Immigration Appeal Tribunal. The appellant is a Burundi national in her early 40s. She came to this country in May 2003 with her two teenage children and claimed asylum on arrival. Her brother had been granted asylum the previous year. She said that her husband had been abducted in January 2003 and that she feared persecution if she was returned to Burundi as a woman of mixed ethnicity who would be living on her own without a husband, particularly as her husband had been abducted on account of his political views.

She had a Hutu mother and a Tutsi father. Both her parents and one of her brothers were killed in the civil war in Burundi. She married her husband in 1988. He was a Tutsi but he joined a political party called FRODEBU, most of whose members are Hutu. She said that her husband disappeared on 18 January 2003 with two of his friends, and a third friend was found dead at the same time. She said that she received telephone threats in April 2003, and at the end of April, when she was away from her home, her house was attacked. She said that she had never in her life been arrested, detained or otherwise harmed although her mixed ethnicity was easily identifiable from her physical features. She said that both Hutus and Tutsis hate people of mixed ethnicity.

She had never been targeted before because she lived in a well-to-do area and was protected by her husband. She maintains that she was now at risk from Hutu and Tutsi extremists as a lone woman with two vulnerable daughters and no protection. Because she would have no support in Burundi she would be placed in a camp with her two daughters without her husband to protect her. In a camp her ethnicity would not be accepted and she and her daughters would be at risk of disease and rape. The gravamen of her complaints on this appeal is that the immigration judge gave no adequate reasons why he was unwilling to accept the report of an independent expert who supported her evidence in certain significant respects. When I gave directions in this matter last September, I commented that the simplest way of dealing with the matter in the light of the appellant’s complaints would be for me to remit it straight away to the immigration judge for him to reconsider the matter and give amplified reasons, now that certain evidence given by the expert had been highlighted by the appellant.

I therefore said that at the hearing of the appeal the court would wish to hear full argument as to whether it possessed this power in relation to decisions taken by the Asylum and Immigration Tribunal. I referred in this context to the recent decisions of this court in English v Emery Reimbold and Strick Ltd [2002] EWCA Civ 605, [2002] 1 WLR 249 and Barke v SEETEC Business Technology Centre Ltd [2005] EWCA Civ 1578. In the first of these cases, the Court of Appeal was concerned only with appeals from the decision of the judge in a lower court. Lord Phillips MR discussed the relevant considerations in paragraphs 15 to 25 of the judgment of the court. They cannot be readily transposed to an appeal from a statutory tribunal. More relevant in the present context is the Barke case where this court held that the Employment Appeal Tribunal, (“EAT”), did have power to direct an employment tribunal to supplement its reasons instead of having to allow a complaint about inadequacy of reasoning to go forward to a full appeal hearing.

The main reason for this decision lay in the language of rule 30 of the Employment Tribunals (Constitution and Rules of Procedure) Regulations 2004. These gave (by rule 30(3)(b)) a tribunal chairman an express duty to supply written reasons in relation to any judgment or order if requested by the EAT at any time. Dyson LJ, giving the judgment of the court, held at paragraph 22 that there was nothing in the language of the rule which prohibited the EAT from requesting further reasons, even when the tribunal had already given some reasons. He explained at paragraphs 23 to 28 how it was not correct to say that the tribunal was functus officio after giving its original reasons, because rules 34 to 37 set out a number of functions it was entitled to perform, including a review of a decision it had just made after its judgment and any written reasons had been entered in the register under rule 32.

It was not suggested that the Asylum and Immigration Tribunal (Procedure) Rules 2005 contained anything resembling that regulatory scheme. Indeed the AIT does not now even have the power which was formerly vested in the IAT to direct reconsideration when the matter would otherwise be on its way to the Court of Appeal. It now only has power of review for an administrative error. Neither Ms Plimmer, who appeared for the appellant, nor Mr Tam who appears for the Secretary of State were at all enthusiastic about this court having any such power, and Mr Tam informed me on instructions that the Asylum and Immigration Tribunal itself shared this general lack of enthusiasm.

I am satisfied that the statutory scheme in the Employment Tribunals sector is totally different from the statutory scheme with which we are concerned in appeals in asylum matters. In contrast to the EAT having power to regulate its own procedure, which was another of the reasons which found favour with this court in Barke, the AIT is very tightly proscribed by the rules. In my judgment there is no power in this court to invite reasons to be supplemented.

I would add that I directed that this matter be considered because in the autumn of last year. I was concerned lest this court might be flooded by appeals from decisions of single judges and which never previously lay to this court in asylum matters. At the moment my concerns in that respect do not seem to have been justified in the event, and if in due course there are concerns on this score, it is clear to me that they would have to be resolved by an express rule rather than by a judgment such as this court gave in the English and Barke cases.

I turn then to the substance of the appeal. The adjudicator found that the appellant’s mother was Hutu and her father was Tutsi. She married a Tutsi. He said that there was no reasonable explanation why her Tutsi husband should join a political party which was overwhelmingly Hutu, particularly as Tutsis would view any Tutsi who joined that party as the worst kind of traitor. He did not accept that the appellant’s husband would allow the information that he had joined that party to spread abroad, and in general he was satisfied that the appellant was able to live on her own in a Tutsi neighbourhood given the fact that she had never been in any trouble in the past. He did not accept that the appellant’s husband had been abducted. He said that it was not plausible that he was a member of FRODEBU. If he had been abducted and the appellant had feared that Tutsis in authority might have been responsible, he did not accept that the appellant would ever have approached the authorities to help her in investigating his disappearance.

If she was wanted by extremist Tutsis or Hutus, he thought it implausible that she would go on living in the family house for four months without her husband. He did not accept that she would have been warned prior to an attack, and he did not accept that the rebels attacked her house because of her husband’s political views or her ethnicity, or that neighbours heard men attacking her house and asking where she was. He said that the men could have taken her whenever they wished, as opposed to attacking on an occasion when she was away from home, and they would not warn her by threatening calls. If she had received such calls, she would not have remained at home. He added that if her home had been attacked she could have sought the assistance of the authorities as she did when her husband disappeared, and he did not accept that she would be a target because of her mixed ethnicity. On the basis of the contemporary country background information at the time of his decision, he was satisfied that the situation in Burundi had dramatically improved. The appellant could resume a normal life in a Tutsi area comparable to the one in which she lived with her husband. She would not be forced to live in a displacement camp.

His conclusions in that regard are critically dependent on the soundness of his findings of fact. If those findings of fact are susceptible to legal challenge then they would have to be reconsidered afresh by the AIT.

In order to understand this case, it is necessary to say something about the politics of Burundi. There is a population of 6.7 million. It is an extremely poor country – I have seen in these papers that it was described on one occasion as the third poorest in the world – 90% of its population are dependent on subsistence agriculture.

The Hutu people constitute 85% or the population of Burundi and 15% are Tutsis. For very many years Tutsis dominated the political, social and economic system. Between the 1960s and 1990s the Tutsis were in the habit of suppressing and slaughtering the Hutus, but there was a popular election for the first time in 1993 which involved a democratic transfer of power to the Hutus. However, the elected Hutu president was murdered and as a consequence Hutu rebel groups took up arms against the Tutsi and 50,000 Tutsi people were killed in the ensuing massacre. Since then there has been a low grade civil war, at any rate until the period with which we are concerned in this case.

In 2000 a moderate Tutsi became president again and he negotiated a peace deal with most of the Hutu rebel groups, although two armed Hutu groups rejected the reconciliation agreement. The agreement led to a power-sharing government in November 2001. He led the government for 18 months and then at the end of April 2003, by arrangement he stood down and a Hutu president then took over. That was the situation at governmental level. During the decade in question between 1993 and 2003, in addition to the conflict which was going on between the largely Tutsi army and the Hutu rebel groups, there was considerable violence being carried on by both Hutu and Tutsi civilian militia. Extremist hard line Tutsi in the Piranha party were particularly involved in anti-Hutu violence. They often used violence to eliminate moderate Tutsis and to undermine the political attempts at reaching a compromise with the Hutus.

As a consequence of these developments there was an increasing segregation of Burundi’s formerly integrated population. Most of the cities are now primarily Tutsi. The countryside is mostly Hutu. The capital Bujambura now has Hutu neighbourhoods. There are conflicts within the Hutu neighbourhoods and Tutsi neighbourhoods and conflicts exist between the neighbourhoods. Although the level of violence was reduced following the peace settlement in 2000, the Hutu rebels continued to carry out raids and extremist Tutsi militias continue to do all they could to undermine the transition, particularly through targeted assassinations of moderates, especially moderate Tutsis.

That account of the matter was largely given by Professor Longman, an American professor who has vast experience as a researcher and consultant in Burundi, Rwanda and Congo. He has conducted extensive field research in all three countries since 1992. So far as the appellant is concerned, her father was an active member of the Piranha party, which is the opposition party, and she believes that he was killed by Tutsis. Her mother and one of her brothers were also killed. She believes they were killed by Hutus in May 2002.

She has always lived in the national capital. She was trained as a French teacher in 1982. She married her husband, who was a pharmacist, in 1988 and her full-time career as a teacher ended in 1989 when the first of her two children were born. After that she gave private lessons for children either at her home or at their home, apart from a period of six months in 2002 when she taught for two terms at a local state school. They were therefore a professional family. They lived in a zone of the capital called Kayanza, and they lived in the same house from 1991 to April 2003. It was a mixed Hutu and Tutsi neighbourhood and although there were tensions before 2002, her mixed ethnicity gave her no problems in that neighbourhood, which she described as a high class neighbourhood where people were sophisticated and civilised.

She used to be a member of the UPRONA political party which was mostly Tutsi. It was the majority party for much of the time, but she left it at the time of the popular elections in 1993 to 1994 and thereafter took no part in politics. Her account of her husband’s politics was that although a Tutsi, he had joined the FRODEBU party and continued to be active in that party after the 1993 to 1994 elections, although he did not want to show himself too much and he eventually belonged to a wing of the party which she identified.

She said he chose FRODEBU because that party was seeking a change of power away from the single Tutsi group which had always reigned in the country. He came from Muramuya, a province of Burundi, which was very antagonistic against the power of the Tutsi in the south. During the course of her evidence the appellant spoke of regional inequities between the South Burundi province and the rest of the country. The appellant said that her husband was never very open with her about his political involvement. Other Tutsis in his circle knew he was a member of that party because of his discussions in certain groups and knew of his friendships with Tutsis who were also very much implicated in that party. She heard everything about her husband’s involvement from outside, but she described an incident in 1993 when she was told by an intimate friend, whose husband was also a member of that party, that the appellant’s husband should refrain from involving himself too much. Her friend was going to make sure that her own husband stopped his activities at the earliest possible time.

The appellant says she was quite shocked, and she confronted her husband asking him to tell her the truth. He said that he was a member of the party but he was not involved actively and he could ask her friend to confirm this. She told him that she did not want him to continue his party membership, but he would not listen to her, and she thereafter did not question him about any of these issues. During the course of her evidence she said that this was a question of great conflict between her and her husband. On two occasions she experienced somebody breaking into her house when her husband was not there. She thought that it was somebody from one of the rebel groups. This happened on one occasion in 1996 and one in 2001. She gave much of that evidence in her answers at the interview. When she was cross-examined, she was asked why her husband as a Tutsi would join this party. She could not give a reason, but he believed that the party supported democracy and he had an open mind towards ethnicity.

The immigration judge treated this evidence in this way. In paragraph 52 of his determination he said the appellant had been unable to provide any reasonable explanation why her Tutsi husband should join what Professor Longman describes at paragraph 10 of his report as:

“… an overwhelmingly Hutu party - FRODEBU. Very few Tutsis have gone so far as to join FRODEBU. Many members of FRODEBU joined the anti-Government rebel movement. Extremist Tutsis believe that FRODEBU is a racist anti-Tutsi party. Tutsis would view any Tutsi who would join FRODEBU as the worst kind of traitor.”

The immigration judge said:

“It is implausible that the Appellant’s husband would join such a party but it is even more implausible that he would join such a party and his friends and neighbours would be aware of that fact. The Appellant has stated that her husband kept his membership of FRODEBU low profile. If that was the case, it is implausible that the Appellant would discover from other people that he had joined the Nyangoma Wing of FRODEBU. In interview, in answer to Q44, the Appellant said that she obtained this information from other people rather than from her husband. I do not accept that the Appellant’s husband would allow such information to spread, that he had joined the Hutu party and risk being viewed as a traitor by Tutsis.”

Ms Plimmer attacks this finding on the grounds that the appellant had given a very clear reason why her husband joined this party. She had explained that in the period up to 1993 when Burundi politics were dominated by the Tutsis, and particularly by the Tutsis in the southern province, there was very considerable antagonism towards this and as a result of that antagonism, particularly in the province from which her husband came, he chose to join a party which was opposed to the main Tutsi party in the south.

Mr Tam, who appears for the Secretary of State, submitted that the immigration judge was entitled to rely on the answer to a question which was given in cross-examination at the time of the hearing, which was to the effect, as I have said, that she could not give a reason why her husband as a Tutsi would join this party.

In my judgment that answer is quite inadequate to rule completely out of court the very full and clear answers the appellant had given in interview. It may be that the appellant was simply saying that she did not know why he joined FRODEBU as opposed to another party which was opposed to the Tutsis in the south, but in my judgment there are serious reasons for concern about the immigration judge’s statement that the appellant had been unable to provide any reasonable explanation why he should join FRODEBU. It may be that he had overlooked the clear evidence which the appellant gave in the course of the interview.

The second matter on which attention was focussed at the hearing of the appeal was the evidence relating to her husband’s disappearance. The appellant said in her various statements that her husband disappeared, as I have said, with two of his friends and another friend was found dead on 18 January. This was a Saturday. Her husband would spend the afternoon with people outside their home. It was customary to spend the afternoons and evenings drinking, particularly at weekends. On 19 January a male neighbour told her that they needed courage to go to the police station, and a male cousin went with her to the police station. On the way they received a call saying that a car had been seen in the bush. It turned out that her husband’s car was discovered on a street corner. It looked as if it had been hidden in grasses in the bush close to a bridge.

The appellant not only telephoned but also went to the office to say that her husband was missing and was told that they would make enquiries. The answer she gave in interview is not wholly clear. It either records “we” or “he”:

“… always thought that he was being looked for by soldiers and he had either been kidnapped or killed. It was a very delicate period of change. We were waiting for political change.”

No doubt she was referring to the change of power to a Hutu president, which was to take place in April. She did not really know why she was thinking of soldiers. It is not wholly clear from that evidence whether she was talking about armed forces or the militia, of which evidence was also given. At any rate, she said that she called the security agents that day because she did not suspect the government at that time to be responsible for his disappearance. She thought that by contacting the authorities at that point, they could actually help her to find out whether her husband had been the victim of a car accident or whether there had been some fight which required police attention. She said that she always maintained that her husband could have been killed, either as a result of his Tutsi ethnicity or his political opinion, by government agents or by FRODEBU party members.

Ms Plimmer not only attacks the premise of the immigration judge’s conclusion that the husband was not abducted, but she also submits that the immigration judge has not paid sufficient attention to the evidence. In my judgment that criticism is well founded. It is clear that the effect of what she was saying was that they initially went down to the police station or contacted the authorities because her husband had gone missing, and she wanted to know whether he had been the victim of a car accident or had been involved in a fight which the police might know about. As Professor Longman has said, there are moderates in the police force. It was only when it appeared that her husband had disappeared and she was concerned about the risk of getting involved with the authorities (where extremists might be involved) that she desisted from receiving help from the authorities.

Again I am of the view that the finding that she would not have approached the authorities if her story was true is susceptible of being unsound. I use this language because one really has to look at the totality of a decision of this kind and not focus on individual parts of it.

The next area about which submissions were made relates to the period between January and April 2003. The appellant said that in February and March she stressed out. She was hoping to see her husband again, but in April she received telephone calls threatening her and saying that it would be her turn to be kidnapped or killed. On 28 April she went to a place called Kianza, close to the Rwanda border, to carry out some administrative formalities concerned with her family property after her father’s death. She left her children with some neighbours. On 29 April a neighbour telephoned her and advised her not to return because the house had been attacked. They would bring her children to Rwanda, and she joined them there. She said that the threatening calls started in the first and second weeks in April. A man over the telephone said:

“Do you recognise me? Do you want to get ready for my visit to your place as we will be there soon.”

They did not say who they were.

The appellant said that in principle when this happens it is Hutu rebels who do this, and she says it has happened to other people. She received three of these calls, the last one at the weekend of the second week in April. The neighbour told her that it was an attack by rebels who came down from the hills to their neighbourhood to attack houses and then return to the hills. She had always had feelings of insecurity due to her ethnic origins, and sometimes the Hutu rebels were opposed to Hutus who did not participate actively in their party. In cross-examination, she said that she stayed at home from January onwards because she hoped her husband had just been kidnapped and might be released. It was not possible for her to leave home after those calls. She would have had to move either to a Tutsi area or to a Hutu area and people would be cautious of her as being of mixed ethnicity. Her neighbour said that the attackers were saying, “Where is she?”

Other homes in her area were also targeted the same day. She believes her house was targeted and she named either the FDD or the FNL as the Hutu rebel group who could have been involved. Professor Longman said that to warn the appellant and to attack her home was consistent with the pattern of attack currently practised in Burundi by extremist Tutsis. He did not say anything in relation to the practice of Hutu rebel groups. The immigration judge said at paragraph 55:

“If the appellant was wanted by extremist Tutsis or Hutus, she was living in the family home for four months without her husband. I do not accept that she would have been warned. If she was a target she could have been attacked at any time.

“56.

I do not accept that the Appellant received threatening telephone calls from rebels or that they attacked her home because of her husband’s political views or her ethnicity. I reject her evidence that neighbours heard men attacking her home, asking where the Appellant was. Her case is that her house was specifically attacked. I do not accept that neighbours would be able to hear shouts directed at the Appellant when the Appellant’s home was the subject of a fierce attack. If men had wished to abduct or kill the appellant, they could have taken her when they wished, and I do not accept that they would attempt to do so when she was away from home on business. I do not accept that they would warn her by threatening telephone calls. If the Appellant did receive telephone calls threatening her, I do not accept that she would have remained at home. I reject her evidence that her home was specifically targeted. I find that if her home had been attacked, she could have sought the assistance of the authorities as she did when her husband disappeared.”

Ms Plimmer attacks this again on the basis that the immigration judge did not take into account the evidence given by the expert. In this sense, her case on this point is less strong because the expert was talking about the practice of extremist Tutsis, whereas the appellant was of the strong belief that these were Hutu rebel groups, who come from the hills to attack her home as well as other homes in the neighbourhood. But her point is a sound one in the sense that the professor was clearly familiar with the practice in this unhappy country of threatening telephone calls of this kind being made before attacks were made.

Mr Tam submits that the immigration judge was entitled to disbelieve the appellant on these matters. In this respect, I am of the view that his case is a sounder one and if this point stood alone, it would be unlikely that this court would be disposed to express concern about the findings of an immigration judge who had heard the evidence. But if the first two matters to which I have referred give rise to justifiable concern, to some extent they affect the third because the immigration judge’s approach to this evidence would have been premised on his earlier findings.

The fourth matter to which Ms Plimmer drew our attention is the approach of the immigration judge to the appellant’s evidence that as a person of mixed ethnicity she was at risk. In this regard the evidence was that a Danish fact-finding report in January 2000 contained evidence suggesting that ethnically mixed marriages in Burundi could cause problems, and that many couples in mixed ethnicity marriages had been forced to flee the country in 1993. The appellant said that when this happened she was living in Bukundira and these problems were more widely reported in the northern parts of the country. In 1993 she was living in this rather elitist district in the capital and as such, they did not experience the full extent of the ethnic cleansing which went on in 1993 following the assassination of the Hutu president.

Her concerns were that her facial features would identify her as being of mixed ethnicity. She could not return to the Tutsi area without the support of her Tutsi husband and she said that there was systematic hatred against mixed race people. Professor Longman gave unequivocal evidence on this matter. He said:

“As someone with a Tutsi father and Hutu mother, she is in a small category of those with mixed ethnicities. There is little place in Burundi today for people who are ‘in between’ ethnicities. You must be either Hutu or Tutsi. As Burundi has become more segregated, those with mixed ethnicities have become increasingly vulnerable. Their mixed blood makes them suspect. They have been forced to prove their loyalties and those who have refused to side clearly with their ethnic groups have been seen as ethnic traitors. The appellant’s family would be unwelcome in both Hutu and Tutsi neighbourhoods and would be at risk in both.”

The appellant called a witness, who was a journalist working in this country who was of mixed ethnicity and came from Burundi. She told the immigration judge that people of mixed ethnicity would be persecuted and suffer isolation because of mixed ethnic background. She had seen them isolated and persecuted; both Hutus and Tutsis viewed people of mixed ethnicity as enemies and dangerous. She accepted that there were many mixed marriages in Burundi. On this issue, the immigration judge said:

“I asked Miss Plimmer to point me to evidence in the objective bundle submitted by the Appellant, or in the CIPU, that people of mixed ethnicity are targeted in Burundi. I have given this matter careful consideration. If 85% of the population are Hutus and 15% are Tutsis in a population of over 6 million, there must be many mixed marriages. This was confirmed in oral evidence by Ms Mossi. CIPU 2003 said that children of mixed marriages may face problems with Hutu or Tutsi extremists. Ms  Mossi is of the opinion that people of mixed ethnicity are persecuted. She witnessed people suffering isolation when she worked in Burundi. Professor Longman points out that people of mixed ethnicity might be seen as traitors in either Hutu or Tutsi neighbourhoods. There is no objective evidence to support the contention of the Appellant that as a person of mixed ethnicity she would be persecuted. Her children are three-quarters Tutsi. She lived in a Tutsi area. She worked as a teacher in a state school. I do not accept that with her qualifications and the physical appearance of her children, she would be a target because of her mixed ethnicity. This proposition is simply not supported by the objective evidence.”

Ms Plimmer complained strongly that the immigration judge was not entitled to say that there was no objective evidence to support this particular contention, unless the immigration judge was excluding the professor as giving objective evidence. If he had been excluding the professor as giving objective evidence, he ought to have said why he was ruling him out. Although Ms Plimmer did not take the point, in the Secretary of State’s own refusal letter it was clear that the Secretary of State was aware of the massacres of people of mixed ethnicity in 1993, and the report of the Danish fact finding mission to that effect. In my judgment on the totality of the evidence, this is another area in which the findings of the immigration judge do not appear to be soundly supported by the evidence.

Of course there must be many mixed marriages in Burundi that give rise to no difficulty, but that is no reason in my judgment, on the evidence which was before the immigration judge, to reject the evidence given by the appellant as a person who ought to know (unless the immigration judge rejected her credibility overall), supported as it was by both Ms Mossi, Professor Longman and the evidence of the Danish fact-finding mission. In my judgment this is another area in which there are reasons for concern.

Ms Plimmer has shown us extracts from the judgment in R v Immigration Appeal Tribunal, ex parte Haile (2001) EWCA Civ 663 in which Simon Brown LJ said that even if there was only one of six points on which the special adjudicator had made a mistake, it would leave a sense of deep injustice. Ms Plimmer says that this is a case in which an intelligent woman is left feeling that the immigration judge has rejected her case for asylum without giving satisfactory reasons for doing so, overriding quite a lot of evidence which supported her case without explaining why he was overriding it.

It is a rare occasion on which this court will be disposed to set aside the decision of an immigration judge on his treatment of the facts, particularly when he has clearly taken so much trouble over the case, as is apparent from the 25-page determination. However, in my judgment this is one of those rare cases where the appellant is entitled to have the matter reconsidered so that she can understand clearly why the findings adverse to her were made. I for my part would allow this appeal and remit the matter to the AIT for yet further reconsideration.

LORD JUSTICE JONATHAN PARKER: I agree.

LORD JUSTICE MAURICE KAY: I also agree.